You're looking at a specific version of this model. Jump to the model overview.
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 |
|---|---|---|---|
| input_images |
string
|
A zip file containing the images that will be used for training.
|
|
| instance_prompt |
string
|
Frog, yarn art style
|
Instance prompt to trigger the image generation
|
| resolution |
integer
(enum)
|
768
Options: 512, 768, 1024 |
The resolution for input images, all the images in the train/validation dataset will be resized to this
|
| max_train_steps |
integer
|
700
Min: 100 Max: 6000 |
Total number of training steps to perform
|
| rank |
integer
|
16
Min: 4 Max: 64 |
The dimension of the LoRA
|
| train_batch_size |
integer
|
1
Min: 1 Max: 8 |
Batch size for the training dataloader
|
| gradient_accumulation_steps |
integer
|
1
Min: 1 Max: 8 |
Number of updates steps to accumulate before performing a backward/update pass
|
| optimizer |
string
(enum)
|
AdamW
Options: AdamW, prodigy |
The optimizer type to use
|
| learning_rate |
number
|
0.0001
Min: 0.0001 Max: 1 |
Initial learning rate to use (1.0 for Prodigy)
|
| lr_scheduler |
string
(enum)
|
constant
Options: linear, cosine, cosine_with_restarts, polynomial, constant, constant_with_warmup |
'The scheduler type to use
|
| checkpointing_steps |
integer
|
Min: 100 Max: 6000 |
Save a checkpoint of the training state every X updates
|
| seed |
integer
|
Seed for reproducibility
|
|
| backend |
string
(enum)
|
no
Options: no, eager, aot_eager, inductor, nvfuser, aot_nvfuser, aot_cudagraphs, ofi, fx2trt, onnxrt, ipex |
Dynamo Backend
|
| hf_token |
string
|
Huggingface token (optional) with write access to upload to Hugging Face
|
|
| hub_model_id |
string
|
Huggingface model location for upload. Requires a HF token with write permissions. Ex: lucataco/SD3.5-Large-queso
|
|
| wandb_api_key |
string
|
Weights and Biases API key, if you'd like to log training progress to W&B.
|
Output schema
The shape of the response you’ll get when you run this model with an API.
{'format': 'uri', 'title': 'Output', 'type': 'string'}