config_vocals_segm_models.yaml

audio:
  chunk_size: 261632
  dim_f: 4096
  dim_t: 512
  hop_length: 512
  n_fft: 8192
  num_channels: 2
  sample_rate: 44100
  min_mean_abs: 0.001

model:
  encoder_name: tu-maxvit_large_tf_512 # look here for possibilities: https://github.com/qubvel/segmentation_models.pytorch#encoders-
  decoder_type: unet # unet, fpn
  act: gelu
  num_channels: 128
  num_subbands: 8

training:
  batch_size: 8
  gradient_accumulation_steps: 1
  grad_clip: 0
  instruments:
  - vocals
  - other
  lr: 5.0e-05
  patience: 2
  reduce_factor: 0.95
  target_instrument: null
  num_epochs: 1000
  num_steps: 2000
  augmentation: false # enable augmentations by audiomentations and pedalboard
  augmentation_type: simple1
  use_mp3_compress: false # Deprecated
  augmentation_mix: true # Mix several stems of the same type with some probability
  augmentation_loudness: true # randomly change loudness of each stem
  augmentation_loudness_type: 1 # Type 1 or 2
  augmentation_loudness_min: 0.5
  augmentation_loudness_max: 1.5
  q: 0.95
  coarse_loss_clip: true
  ema_momentum: 0.999
  optimizer: adamw
  other_fix: true # it's needed for checking on multisong dataset if other is actually instrumental

inference:
  batch_size: 1
  dim_t: 512
  num_overlap: 4