zenaivn/dreambooth-training

zenaivn / dreambooth-training

Public
293 runs

Run zenaivn/dreambooth-training with an API

Use one of our client libraries to get started quickly. Clicking on a library will take you to the Playground tab where you can tweak different inputs, see the results, and copy the corresponding code to use in your own project.

Input schema

The fields you can use to run this model with an API. If you don't give a value for a field its default value will be used.

Field	Type	Default value	Description
train_dreambooth	boolean	True	Whether to train dreambooth or not.
train_lora	boolean	False	Whether to train lora or not.
train_lora_sliders	boolean	False	Whether to train lora sliders or not.
model_replicate_link	string		(Optional) Enter your model replicate link for lora sliders training. In case you don't have your own, here is a test link: https://replicate.delivery/pbxt/mu6jfnIIdQWIMKpBXRX3AkRltbjmwEX7kERNInQptcSM3VIJA/output.zip
breast	boolean	True	(Optional) Choose if you want to train lora slider for breast
butt	boolean	False	(Optional) Choose if you want to train lora slider for butt
weight	boolean	False	(Optional) Choose if you want to train lora slider for weight
class_name	string		Specify the class name (for example: man/woman/style/...).
identifier	string	ohwx	Unique identifier for the model instance.
num_repeats	string	1	Number of times the input images should be repeated.
output_name	string		Name of the model's output file. (for example: output_model)
instance_data	string		A ZIP file containing your training images (JPG, PNG, etc. size not restricted). These images contain your 'subject' that you want the trained model to embed in the output domain for later generating customized scenes beyond the training images. For best results, use images without noise or unrelated objects in the background.
class_data	string		An optional ZIP file containing the training data of class images. This corresponds to `class_prompt` above, also with the purpose of keeping the model generalizable. By default, the pretrained stable-diffusion model will generate N images (determined by the `num_class_images` you set) based on the `class_prompt` provided. But to save time or to have your preferred specific set of `class_data`, you can also provide them in a ZIP file.
using_crop_images	boolean	True	Whether to use crop images or not.
train_batch_size	integer	1	Batch size for training data loader, applied per device.
learning_rate	number	0.000001	Initial learning rate (after the potential warmup period) to use. For training LoRA, use 0.0005.
learning_rate_te	number	0.000001	Initial learning rate te (after the potential warmup period) to use. For training LoRA, use 0.0005
lr_scheduler	string (enum)	cosine Options: linear, cosine, cosine_with_restarts, polynomial, constant, constant_with_warmup	The scheduler type to use. For training LoRA, use constant
seed	integer	1337	Your seed for training script.

{
  "type": "object",
  "title": "Input",
  "required": [
    "class_name",
    "output_name",
    "instance_data"
  ],
  "properties": {
    "butt": {
      "type": "boolean",
      "title": "Butt",
      "default": false,
      "x-order": 5,
      "description": "(Optional) Choose if you want to train lora slider for butt"
    },
    "seed": {
      "type": "integer",
      "title": "Seed",
      "default": 1337,
      "x-order": 18,
      "description": "Your seed for training script."
    },
    "breast": {
      "type": "boolean",
      "title": "Breast",
      "default": true,
      "x-order": 4,
      "description": "(Optional) Choose if you want to train lora slider for breast"
    },
    "weight": {
      "type": "boolean",
      "title": "Weight",
      "default": false,
      "x-order": 6,
      "description": "(Optional) Choose if you want to train lora slider for weight"
    },
    "class_data": {
      "type": "string",
      "title": "Class Data",
      "format": "uri",
      "x-order": 12,
      "description": "An optional ZIP file containing the training data of class images. This corresponds to `class_prompt` above, also with the purpose of keeping the model generalizable. By default, the pretrained stable-diffusion model will generate N images (determined by the `num_class_images` you set) based on the `class_prompt` provided. But to save time or to have your preferred specific set of `class_data`, you can also provide them in a ZIP file."
    },
    "class_name": {
      "type": "string",
      "title": "Class Name",
      "x-order": 7,
      "description": "Specify the class name (for example: man/woman/style/...)."
    },
    "identifier": {
      "type": "string",
      "title": "Identifier",
      "default": "ohwx",
      "x-order": 8,
      "description": "Unique identifier for the model instance."
    },
    "train_lora": {
      "type": "boolean",
      "title": "Train Lora",
      "default": false,
      "x-order": 1,
      "description": "Whether to train lora or not."
    },
    "num_repeats": {
      "type": "string",
      "title": "Num Repeats",
      "default": "1",
      "x-order": 9,
      "description": "Number of times the input images should be repeated."
    },
    "output_name": {
      "type": "string",
      "title": "Output Name",
      "x-order": 10,
      "description": "Name of the model's output file. (for example: output_model)"
    },
    "lr_scheduler": {
      "enum": [
        "linear",
        "cosine",
        "cosine_with_restarts",
        "polynomial",
        "constant",
        "constant_with_warmup"
      ],
      "type": "string",
      "title": "lr_scheduler",
      "description": "The scheduler type to use. For training LoRA, use constant",
      "default": "cosine",
      "x-order": 17
    },
    "instance_data": {
      "type": "string",
      "title": "Instance Data",
      "x-order": 11,
      "description": "A ZIP file containing your training images (JPG, PNG, etc. size not restricted). These images contain your 'subject' that you want the trained model to embed in the output domain for later generating customized scenes beyond the training images. For best results, use images without noise or unrelated objects in the background."
    },
    "learning_rate": {
      "type": "number",
      "title": "Learning Rate",
      "default": 1e-06,
      "x-order": 15,
      "description": "Initial learning rate (after the potential warmup period) to use. For training LoRA, use 0.0005."
    },
    "learning_rate_te": {
      "type": "number",
      "title": "Learning Rate Te",
      "default": 1e-06,
      "x-order": 16,
      "description": "Initial learning rate te (after the potential warmup period) to use. For training LoRA, use 0.0005"
    },
    "train_batch_size": {
      "type": "integer",
      "title": "Train Batch Size",
      "default": 1,
      "x-order": 14,
      "description": "Batch size for training data loader, applied per device."
    },
    "train_dreambooth": {
      "type": "boolean",
      "title": "Train Dreambooth",
      "default": true,
      "x-order": 0,
      "description": "Whether to train dreambooth or not."
    },
    "using_crop_images": {
      "type": "boolean",
      "title": "Using Crop Images",
      "default": true,
      "x-order": 13,
      "description": "Whether to use crop images or not."
    },
    "train_lora_sliders": {
      "type": "boolean",
      "title": "Train Lora Sliders",
      "default": false,
      "x-order": 2,
      "description": "Whether to train lora sliders or not."
    },
    "model_replicate_link": {
      "type": "string",
      "title": "Model Replicate Link",
      "x-order": 3,
      "description": "(Optional) Enter your model replicate link for lora sliders training. In case you don't have your own, here is a test link: https://replicate.delivery/pbxt/mu6jfnIIdQWIMKpBXRX3AkRltbjmwEX7kERNInQptcSM3VIJA/output.zip"
    }
  }
}

Output schema

The shape of the response you’ll get when you run this model with an API.

Schema

{
  "type": "string",
  "title": "Output",
  "format": "uri"
}