configs/generate_diffree.yaml

# File modified by authors of InstructPix2Pix from original (https://github.com/CompVis/stable-diffusion).
# See more details in LICENSE.

model:
  base_learning_rate: 5.0e-05
  target: ldm.models.diffusion.ddpm_diffree.LatentDiffusion
  params:
    linear_start: 0.00085
    linear_end: 0.0120
    num_timesteps_cond: 1
    log_every_t: 200
    timesteps: 1000
    first_stage_key: edited
    cond_stage_key: edit
    first_stage_downsample: True
    # image_size: 64
    # image_size: 32
    image_size: 16
    channels: 4
    cond_stage_trainable: false   # Note: different from the one we trained before
    conditioning_key: hybrid
    monitor: val/loss_simple_ema
    scale_factor: 0.18215
    use_ema: true
    load_ema: true

    scheduler_config: # 10000 warmup steps
      target: ldm.lr_scheduler.LambdaLinearScheduler
      params:
        warm_up_steps: [ 0 ]
        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_diffree.UNetModel
      params:
        image_size: 32 # unused
        in_channels: 8
        # in_mask_channels: 12
        out_channels: 4
        model_channels: 320
        attention_resolutions: [ 4, 2, 1 ]
        num_res_blocks: 2
        channel_mult: [ 1, 2, 4, 4 ]
        num_heads: 8
        use_spatial_transformer: True
        transformer_depth: 1
        context_dim: 768
        use_checkpoint: True
        legacy: False
    
    omp_config:
      target: ldm.modules.diffusionmodules.openaimodel_diffree.OMPModule
      params:
        image_size: 32 # unused
        in_channels: 8
        # in_mask_channels: 12
        out_channels: 4
        model_channels: 320
        attention_resolutions: [ 4, 2, 1 ]
        num_res_blocks: 2
        channel_mult: [ 1, 2, 4, 4 ]
        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:
  target: main.DataModuleFromConfig
  params:
    batch_size: 128
    num_workers: 1
    wrap: false
    validation:
      target: edit_dataset_pam.EditDatasetMask
      params:
        path: data/clip-filtered-dataset
        cache_dir:  data/
        cache_name: data_10k
        split: val
        min_text_sim: 0.2
        min_image_sim: 0.75
        min_direction_sim: 0.2
        max_samples_per_prompt: 1
        min_resize_res: 512
        max_resize_res: 512
        crop_res: 512
        output_as_edit: False
        real_input: True