BioBB PYTORCH Command Line Help

Generic usage:

biobb_command [-h] --config CONFIG --input_file(s) <input_file(s)> --output_file <output_file>

Train_mdae

Train a Molecular Dynamics AutoEncoder (MDAE) PyTorch model.

Get help

Command:

train_mdae -h

/bin/sh: train_mdae: command not found

I / O Arguments

Syntax: input_argument (datatype) : Definition

Config input / output arguments for this building block:

  • input_train_npy_path (string): Path to the input train data file. File type: input. Sample file. Accepted formats: NPY

  • output_model_pth_path (string): Path to the output model file. File type: output. Sample file. Accepted formats: PTH

  • input_model_pth_path (string): Path to the input model file. File type: input. Sample file. Accepted formats: PTH

  • output_train_data_npz_path (string): Path to the output train data file. File type: output. Sample file. Accepted formats: NPZ

  • output_performance_npz_path (string): Path to the output performance file. File type: output. Sample file. Accepted formats: NPZ

Config

Syntax: input_parameter (datatype) - (default_value) Definition

Config parameters for this building block:

  • latent_dimensions (integer): (2) min dimensionality of the latent space..

  • num_layers (integer): (4) number of layers in the encoder/decoder (4 to encode and 4 to decode)..

  • num_epochs (integer): (100) number of epochs (iterations of whole dataset) for training..

  • lr (number): (0.0001) learning rate..

  • self.lr_step_size (integer): (100) Period of learning rate decay..

  • self.gamma (number): (0.1) Multiplicative factor of learning rate decay..

  • checkpoint_interval (integer): (25) number of epochs interval to save model checkpoints o 0 to disable..

  • output_checkpoint_prefix (string): (checkpoint_epoch) prefix for the checkpoint files..

  • partition (number): (0.8) 0.8 = 80% partition of the data for training and validation..

  • batch_size (integer): (1) number of samples/frames per batch..

  • log_interval (integer): (10) number of epochs interval to log the training progress..

  • input_dimensions (integer): (None) input dimensions by default it should be the number of features in the input data (number of atoms * 3 corresponding to x, y, z coordinates)..

  • output_dimensions (integer): (None) output dimensions by default it should be the number of features in the input data (number of atoms * 3 corresponding to x, y, z coordinates)..

  • loss_function (string): (MSELoss) Loss function to be used. .

  • optimizer (string): (Adam) Optimizer algorithm to be used. .

YAML

Common config file

properties:
  num_epochs: 50

Command line

train_mdae --config config_train_mdae.yml --input_train_npy_path train_mdae_traj.npy --output_model_pth_path ref_output_model.pth --input_model_pth_path ref_output_model.pth --output_train_data_npz_path ref_output_train_data.npz --output_performance_npz_path ref_output_performance.npz

JSON

Common config file

{
  "properties": {
    "num_epochs": 50
  }
}

Command line

train_mdae --config config_train_mdae.json --input_train_npy_path train_mdae_traj.npy --output_model_pth_path ref_output_model.pth --input_model_pth_path ref_output_model.pth --output_train_data_npz_path ref_output_train_data.npz --output_performance_npz_path ref_output_performance.npz

Apply_mdae

Apply a Molecular Dynamics AutoEncoder (MDAE) PyTorch model.

Get help

Command:

apply_mdae -h

/bin/sh: apply_mdae: command not found

I / O Arguments

Syntax: input_argument (datatype) : Definition

Config input / output arguments for this building block:

  • input_data_npy_path (string): Path to the input data file. File type: input. Sample file. Accepted formats: NPY

  • input_model_pth_path (string): Path to the input model file. File type: input. Sample file. Accepted formats: PTH

  • output_reconstructed_data_npy_path (string): Path to the output reconstructed data file. File type: output. Sample file. Accepted formats: NPY

  • output_latent_space_npy_path (string): Path to the reduced dimensionality file. File type: output. Sample file. Accepted formats: NPY

Config

Syntax: input_parameter (datatype) - (default_value) Definition

Config parameters for this building block:

  • batch_size (integer): (1) number of samples/frames per batch..

  • latent_dimensions (integer): (2) min dimensionality of the latent space..

  • num_layers (integer): (4) number of layers in the encoder/decoder (4 to encode and 4 to decode)..

  • input_dimensions (integer): (None) input dimensions by default it should be the number of features in the input data (number of atoms * 3 corresponding to x, y, z coordinates)..

  • output_dimensions (integer): (None) output dimensions by default it should be the number of features in the input data (number of atoms * 3 corresponding to x, y, z coordinates)..

YAML

Common config file

properties:
  batch_size: 1

Command line

apply_mdae --config config_apply_mdae.yml --input_data_npy_path train_mdae_traj.npy --input_model_pth_path ref_output_model.pth --output_reconstructed_data_npy_path ref_output_reconstructed_data.npy --output_latent_space_npy_path ref_output_latent_space.npy

JSON

Common config file

{
  "properties": {
    "batch_size": 1
  }
}

Command line

apply_mdae --config config_apply_mdae.json --input_data_npy_path train_mdae_traj.npy --input_model_pth_path ref_output_model.pth --output_reconstructed_data_npy_path ref_output_reconstructed_data.npy --output_latent_space_npy_path ref_output_latent_space.npy