replicate / dreambooth

Train your own custom Stable Diffusion model using a small set of images

  • Public
  • 295.5K runs
  • A100 (80GB)
  • GitHub
  • Paper
  • License
  • Prediction

    replicate/dreambooth:a8ba568d
    ID
    4yyxeqh2cfcd7pueymhl6jhcra
    Status
    Succeeded
    Source
    Web
    Hardware
    Total duration
    Created
    by @bfirsh

    Input

    seed
    1337
    adam_beta1
    0.9
    adam_beta2
    0.999
    resolution
    512
    adam_epsilon
    1e-8
    class_prompt
    a cat
    lr_scheduler
    constant
    instance_data
    data.zip
    learning_rate
    0.000001
    max_grad_norm
    1
    n_save_sample
    4
    use_8bit_adam
    instance_prompt
    a cjw cat
    max_train_steps
    1000
    num_class_images
    100
    num_train_epochs
    1
    save_infer_steps
    50
    train_batch_size
    1
    adam_weight_decay
    0.01
    prior_loss_weight
    1
    sample_batch_size
    4
    train_text_encoder
    save_guidance_scale
    7.5
    gradient_checkpointing
    with_prior_preservation
    gradient_accumulation_steps
    1

    Output

    Generated in
  • Prediction

    replicate/dreambooth:a8ba568d
    ID
    h3wipflhzrexbasf3ngae6em64
    Status
    Succeeded
    Source
    API
    Hardware
    Total duration
    Created
    by @bfirsh

    Input

    class_prompt
    a photo of a man
    instance_data
    data.zip
    instance_prompt
    a photo of bfirsh
    max_train_steps
    2000

    Output

    Generated in

Want to make some of these yourself?

Run this model