Command-line interface¶
sGDML provides a fully featured command-line interface (CLI) for all tasks related to force field reconstruction. You can get help for any command with the -h
flag:
$ sgdml <command> -h
List of commands¶
Task |
Command |
Reconstruct a force field from beginning to end |
|
Create training tasks |
|
Train models from training tasks |
|
Validate models |
|
Select best performing model |
|
Test a model |
|
Show details for dataset, task or model file |
|
Purge all caches |
|
Tip
Dataset files can also be referenced by their fingerprint instead of their file name, e.g. ./d_ethanol.npz
is equivalent to ./f03a68c944d70bd7083c951e7f77aaac
within the CLI. Since fingerprints are guaranteed to be unique, this practice is less prone to user input error when dealing with multiple similar datasets.
List of optional arguments¶
Some commands have optional arguments to configure model/training parameters, file handling or to manage compute resources. Please use the -h
flag to see which commands support what arguments.
Generic arguments¶
Option |
Argument |
Print a description of all command line options |
|
Print the sGDML version number and exit |
|
Model/training configuration¶
Option |
Argument |
Path to a separate validation dataset file |
|
Path to separate test dataset file |
|
Integer list and/or range |
|
Ignore symmetries in the model (as in the orginal GDML variant) |
|
Take permutations from existing file (e.g. other |
|
Reconstruct force field w/o its corresponding potential energy surface |
|
Include the energy constraints in the kernel (CPU-implementation only) |
|
Give up on unfinished tasks (e.g. due to timeouts, crashes, etc.), if multiple are trained |
|
File management¶
Option |
Argument |
Specify a custom training directory |
|
Specify a custom model output file |
|
Overwrite outputs/allow changing existing files |
|
Compute resource management¶
Option |
Argument |
Limit memory usage (whenever possible) [GB] |
|
Limit number of processes |
|
Use CPU implementation (no PyTorch dependency) |
|