# general settings
name: InterLCM
model_type: CodeFormerJointModel
num_gpu: 1
manual_seed: 0

# dataset and data loader settings
datasets:
  train:
    name: FFHQ
    type: FFHQBlindJointDataset
    dataroot_gt: /data/dataset/ffhq/ffhq_512
    filename_tmpl: '{}'
    io_backend:
      type: disk

    in_size: 512
    gt_size: 512
    mean: [0.5, 0.5, 0.5]
    std: [0.5, 0.5, 0.5]
    use_hflip: true
    use_corrupt: true

    blur_kernel_size: 41
    use_motion_kernel: false
    motion_kernel_prob: 0.001
    kernel_list: ['iso', 'aniso']
    kernel_prob: [0.5, 0.5]
    # small degradation
    blur_sigma: [0.1, 10]
    downsample_range: [1, 12]
    noise_range: [0, 15]
    jpeg_range: [60, 100]
    # large degradation
    blur_sigma_large: [1, 15]
    downsample_range_large: [4, 30]
    noise_range_large: [0, 20]
    jpeg_range_large: [30, 80]

    # data loader
    num_worker_per_gpu: 1
    batch_size_per_gpu: 1
    dataset_enlarge_ratio: 100
    prefetch_mode: ~

#   val:
#     name: CelebA-HQ-512
#     type: PairedImageDataset
#     dataroot_lq: datasets/faces/validation/lq
#     dataroot_gt: datasets/faces/validation/gt
#     io_backend:
#       type: disk
#     mean: [0.5, 0.5, 0.5]
#     std: [0.5, 0.5, 0.5]
#     scale: 1
    
# network structures
visual_encoder:
  type: VisualEncoder
  nf: 64
  emb_dim: 197
  ch_mult: [2, 4, 8]
  res_blocks: 2
  img_size: 512

lcm:
  type: LCM
  pretrained_model: 'SimianLuo/LCM_Dreamshaper_v7'
  num_inference_steps: 4

spatial_encoder:
  pretrained_model: 'runwayml/stable-diffusion-v1-5'

network_vqgan: # this config is needed if no pre-calculated latent
  type: VQAutoEncoder
  img_size: 512
  nf: 64
  ch_mult: [1, 2, 2, 4, 4, 8]
  quantizer: 'nearest'
  codebook_size: 1024

network_d:
  type: VQGANDiscriminator
  nc: 3
  ndf: 64
  n_layers: 4

# path
path:
  pretrain_network_ve: ~
  pretrain_network_se: ~
  param_key_ve: params_ema
  param_key_se: params_ema
  strict_load_ve: true
  strict_load_se: true
  pretrain_network_d: ~
  resume_state: ~

# base_lr(4.5e-6)*bach_size(4)
train:
  use_hq_feat_loss: true
  feat_loss_weight: 1.0
  cross_entropy_loss: true
  entropy_loss_weight: 0.5
  scale_adaptive_gan_weight: 0.1

  # Visual Encoder and Spatial Encoder
  optim_vse:
    type: Adam
    lr: !!float 5e-5
    weight_decay: 0
    betas: [ 0.9, 0.99 ]
  optim_d:
    type: Adam
    lr: !!float 5e-5
    weight_decay: 0
    betas: [0.9, 0.99]

  scheduler:
    type: CosineAnnealingRestartLR
    periods: [150000]
    restart_weights: [1]
    eta_min: !!float 2e-5


  total_iter: 15001

  warmup_iter: -1  # no warm up
  ema_decay: 0.997

  pixel_opt:
    type: L1Loss
    loss_weight: 1.0
    reduction: mean

  perceptual_opt:
    type: LPIPSLoss
    loss_weight: 1.0
    use_input_norm: true
    range_norm: true

  gan_opt:
    type: GANLoss
    gan_type: hinge
    loss_weight: !!float 1.0 # adaptive_weighting

  use_adaptive_weight: true

  net_d_iters: 1
  net_d_start_iter: 10001
  manual_seed: 0

# validation settings
val:
  val_freq: !!float 5e10 # no validation
  save_img: true

  metrics:
    psnr: # metric name, can be arbitrary
      type: calculate_psnr
      crop_border: 4
      test_y_channel: false

# logging settings
logger:
  print_freq: 100
  save_checkpoint_freq: !!float 5e3
  use_tb_logger: true
  wandb:
    project: ~
    resume_id: ~

# dist training settings
dist_params:
  backend: nccl
  port: 29413

find_unused_parameters: true