stable-diffusion-finetune/configs/stable-diffusion/txt2img-v2-clip-encoder-improved_aesthetics-512-dev.yaml

model:
  base_learning_rate: 1.0e-04
  target: ldm.models.diffusion.ddpm.LatentDiffusion
  params:

    linear_start: 0.00085
    linear_end: 0.0120
    num_timesteps_cond: 1
    log_every_t: 200
    timesteps: 1000
    first_stage_key: "jpg"
    cond_stage_key: "txt"
    image_size: 64
    channels: 4
    cond_stage_trainable: false   # Note: different from the one we trained before
    conditioning_key: crossattn
    monitor: val/loss_simple_ema
    scale_factor: 0.18215

    scheduler_config: # 10000 warmup steps
      target: ldm.lr_scheduler.LambdaLinearScheduler
      params:
        warm_up_steps: [ 10000 ]
        cycle_lengths: [ 10000000000000 ] # incredibly large number to prevent corner cases
        f_start: [ 1.e-6 ]
        f_max: [ 1. ]
        f_min: [ 1. ]

    unet_config:
      target: ldm.modules.diffusionmodules.openaimodel.UNetModel
      params:
        image_size: 64 # unused
        in_channels: 4
        out_channels: 4
        model_channels: 384
        attention_resolutions: [ 8, 4, 2, 1 ]
        num_res_blocks: [ 2, 2, 2, 2 ]
        channel_mult: [ 1, 2, 4, 4 ]
        disable_self_attentions: [ False, False, False, False ]  # converts the self-attention to a cross-attention layer if true
        num_heads: 8
        use_spatial_transformer: True
        transformer_depth: 1
        context_dim: 768
        use_checkpoint: True
        legacy: False

    first_stage_config:
      target: ldm.models.autoencoder.AutoencoderKL
      params:
        embed_dim: 4
        monitor: val/rec_loss
        ddconfig:
          double_z: true
          z_channels: 4
          resolution: 256
          in_channels: 3
          out_ch: 3
          ch: 128
          ch_mult:
          - 1
          - 2
          - 4
          - 4
          num_res_blocks: 2
          attn_resolutions: []
          dropout: 0.0
        lossconfig:
          target: torch.nn.Identity

    cond_stage_config:
      target: ldm.modules.encoders.modules.FrozenCLIPEmbedder


data:  # TODO
  target: main.DataModuleFromConfig
  params:
    batch_size: 1
    num_workers: 4
    wrap: false
    train:
      target: ldm.data.dummy.DummyData
      params:
        length: 20000
        size: [512, 512, 3]
    validation:
      target: ldm.data.dummy.DummyData
      params:
        length: 10000
        size: [512, 512, 3]

#data:
#  target: ldm.data.laion.WebDataModuleFromConfig
#  params:
#    tar_base: "pipe:aws s3 cp s3://s-datasets/laion-high-resolution/"
#    batch_size: 4
#    num_workers: 4
#    multinode: True
#    train:
#      shards: '{00000..17279}.tar -'
#      shuffle: 10000
#      image_key: jpg
#      image_transforms:
#      - target: torchvision.transforms.Resize
#        params:
#          size: 512
#          interpolation: 3
#      - target: torchvision.transforms.RandomCrop
#        params:
#          size: 512
#
#    # NOTE use enough shards to avoid empty validation loops in workers
#    validation:
#      shards: '{17280..17535}.tar -'
#      shuffle: 0
#      image_key: jpg
#      image_transforms:
#      - target: torchvision.transforms.Resize
#        params:
#          size: 512
#          interpolation: 3
#      - target: torchvision.transforms.CenterCrop
#        params:
#          size: 512


lightning:
  callbacks:
    image_logger:
      target: main.ImageLogger
      params:
        batch_frequency: 5  # TODO
        max_images: 4
        increase_log_steps: False
        log_first_step: False
        log_images_kwargs:
          use_ema_scope: False
          inpaint: False
          plot_progressive_rows: False
          plot_diffusion_rows: False
          N: 4
          unconditional_guidance_scale: 3.0
          unconditional_guidance_label: [""]

  trainer:
    #replace_sampler_ddp: False
    benchmark: True
    val_check_interval: 1000 # TODO: 5000000 # really sorry
    num_sanity_val_steps: 0
    accumulate_grad_batches: 2