Universal training folder

operators.ml3d.universal_training_folder(client, folder_in_folder='/folder_in_folder', out_model_path='parameters_model_test', voxel_size=0.02, zero_centering='True', point_names='X,Y,Z', feature_names='', label_names='classification', num_classes=1, label_scales='0.01', learning_rate=3e-06, learning_decay=0.9999, num_epochs=200000, regularization_decay=1e-09, batch_size=2, save_after_epochs=100, backbone_type='MinkUNet14A', head_type='HeadPointwise', criterion_type='L1Sum', probabilistic='True', hidden_layers=8, store_in_memory='True', worker_instance_type='x2large', manager_instance_type='small', extension_in_folder='.laz', skip_existing_files=False)

universal_training_folder(client,
in_folder=’/in_folder’,
out_model_path=’parameters_model_test’,
voxel_size=0.02,
zero_centering=’True’,
point_names=’X,Y,Z’,
feature_names=’’,
label_names=’classification’,
num_classes=1,
label_scales=’0.01’,
learning_rate=3e-6,
learning_decay=0.9999,
num_epochs=200000,
regularization_decay=1e-9,
batch_size=2,
save_after_epochs=100,
backbone_type=’MinkUNet14A’,
head_type=’HeadPointwise’,
criterion_type=’L1Sum’,
probabilistic=’True’,
hidden_layers=8,
store_in_memory=’True’,
worker_instance_type=’x2large’,
manager_instance_type=”small”,
extension_in_folder=”.data_train”,
skip_existing_files = False )
Parameters:

  • out_model_path – model path

  • voxel_size – voxel size

  • zero_centering – zero centering

  • point_names – point names

  • feature_names – feature names

  • label_names – label names

  • num_classes – number of classes

  • label_scales – label scales

  • learning_rate – learning rate

  • learning_decay – learning rate decay

  • num_epochs – number of epochs

  • regularization_decay – regularization decay

  • batch_size – batch size for training

  • save_after_epochs – save after epochs

  • backbone_type – model type of backbone network

  • head_type – model type of head network

  • criterion_type – model type of criterion

  • probabilistic – estimate probabilities: labels in [0,1]

  • hidden_layers – number of hidden layers

  • store_in_memory – store training data in memory

  • folder_in_folder – input directory with training data

  • worker_instance_type – cloud instance type of worker nodes

  • manager_instance_type – cloud instance type of manager node

  • extension_in_folder – File extension of files in folder for folder_in_folder

  • skip_existing_files – skip files that already exist in the output folder