train_pointnet_clm.yaml
=======================

.. code-block:: yaml
  :linenos:

  # train ae or vae
  #### a template for pcd encoder
  mode: train
  model:
    # can be AE, VAE, SeM and DDPM
    name: ClM
    # encoder config
    encoder:
      name: PointNet
      in_channel: 3
      out_channel: 46
      point_num: 2048
      building_block: dense
      # num_neighbors_k: 20
      local_feature_channels: [64, 64, 128, 256, 512]
      dense_channels: [1024, 512, 256]
      activation: lrelu
      normalization: group
    classification_losses: 
      - name: CE
  loader:
    dataset: 
      name: PcdClsDataset
      data_suffix: .ply
      cls_label: MedPointS
    data_path_list: 
      - path/to/datasets/pcd/MedPointS/classification/fold1
      - path/to/datasets/pcd/MedPointS/classification/fold2
      - path/to/datasets/pcd/MedPointS/classification/fold3
    batch_size: 16
    num_workers: 8
    shuffle: true
    data_transforms:
      - name: Normalize
      - name: FixedPoints
        num: 2048
      - name: ToTensor
        dtype: float
    label_transforms:
      - name: ToOneHot
        num_classes: 47
        ignore_background: true
      - name: ToTensor
        dtype: float
  val_loader:
    dataset: 
      name: PcdClsDataset
      data_suffix: .ply
      cls_label: MedPointS
    data_path_list: 
      - path/to/datasets/pcd/MedPointS/classification/fold4
    batch_size: 32
    num_workers: 8
    shuffle: true
    data_transforms:
      - name: Normalize
      - name: FixedPoints
        num: 2048
      - name: ToTensor
        dtype: float
    label_transforms:
      - name: ToOneHot
        num_classes: 47
        ignore_background: true
      - name: ToTensor
        dtype: float
  check_point_dir: path/to/flemme-ckp/MedPointS/PointNet/PointNet_CLS
  ### parameter for optimizer
  optimizer:
    name: Adam
    lr: 0.0001
    weight_decay: 0.00000001
  evaluation_metrics:
    cls:
      - name: ACC
  score_metric:
    name: ACC
  ### scheduler for learning rate
  lr_scheduler: 
    name: LinearLR
    start_factor: 1.0
    end_factor: 0.01
  max_epoch: 100
  write_after_iters: 20
  save_after_epochs: 1