qe_ar2_test.yaml

# @package _global_

# Test with length 131072, batch size 4, logger sampling_steps [3]

sampling_rate: 16000
length: 131072
channels: 2
log_every_n_steps: 1000


model:
  _target_: main.module_qe_ar2.Model
  lr: 1e-4
  lr_beta1: 0.95
  lr_beta2: 0.999
  lr_eps: 1e-6
  lr_weight_decay: 1e-3

  autoencoder_path: null
  codebook_size: 2048
  num_residuals: 3
  shared_codebook: False

  transformer:
    _target_: a_transformers_pytorch.transformers.Autoregressive
    num_tokens: 6144
    transformer:
      _target_: a_transformers_pytorch.transformers.Transformer
      features: 768
      max_length: 192
      num_layers: 8
      head_features: 64
      num_heads: 12
      multiplier: 4
      causal: True

datamodule:
  _target_: main.module_qe_ar2.Datamodule
  dataset:
    _target_: audio_data_pytorch.YoutubeDataset
    urls:
      - https://www.youtube.com/watch?v=_dEw2zQ7xcE # 1h GHØSTS
    root: ${data_dir}
    crop_length: 12 # seconds crops
    transforms:
      _target_: audio_data_pytorch.AllTransform
      source_rate: ${sampling_rate}
      target_rate: ${sampling_rate}
      random_crop_size: ${length}
      stereo: True
      loudness: -20
  val_split: 0.01
  batch_size: 4
  num_workers: 8
  pin_memory: True

callbacks:
  rich_progress_bar:
    _target_: pytorch_lightning.callbacks.RichProgressBar

  model_checkpoint:
    _target_: pytorch_lightning.callbacks.ModelCheckpoint
    monitor: "valid_loss"   # name of the logged metric which determines when model is improving
    save_top_k: 1           # save k best models (determined by above metric)
    save_last: True         # additionaly always save model from last epoch
    mode: "min"             # can be "max" or "min"
    verbose: False
    dirpath: ${logs_dir}/ckpts/${now:%Y-%m-%d-%H-%M-%S}
    filename: '{epoch:02d}-{valid_loss:.3f}'

  model_summary:
    _target_: pytorch_lightning.callbacks.RichModelSummary
    max_depth: 2

  audio_samples_logger:
    _target_: main.module_qe_ar2.SampleLogger
    num_items: 3
    channels: ${channels}
    sampling_rate: ${sampling_rate}
    length: ${length}

loggers:
  wandb:
    _target_: pytorch_lightning.loggers.wandb.WandbLogger
    project: ${oc.env:WANDB_PROJECT}
    entity: ${oc.env:WANDB_ENTITY}
    # offline: False  # set True to store all logs only locally
    job_type: "train"
    group: ""
    save_dir: ${logs_dir}

trainer:
  _target_: pytorch_lightning.Trainer
  gpus: 0 # Set `1` to train on GPU, `0` to train on CPU only, and `-1` to train on all GPUs, default `0`
  precision: 32 # Precision used for tensors, default `32`
  accelerator: null # `ddp` GPUs train individually and sync gradients, default `None`
  min_epochs: 0
  max_epochs: -1
  enable_model_summary: False
  log_every_n_steps: 1 # Logs metrics every N batches
  check_val_every_n_epoch: null
  val_check_interval: ${log_every_n_steps}