PLUGINS/MACHINE LEARNING - Revision history

Schlipf at 08:43, 16 March 2026

2026-03-16T08:43:32Z

← Older revision		Revision as of 08:43, 16 March 2026
Line 2:		Line 2:

	Description: {{TAG\|PLUGINS/MACHINE_LEARNING}} calls the Python plugin for the machine learning interface for each ionic relaxation step		Description: {{TAG\|PLUGINS/MACHINE_LEARNING}} calls the Python plugin for the machine learning interface for each ionic relaxation step
			{{NB\|deprecated\|We do not recommend using this interface anymore. There is a mismatch in the units of ASE and VASP that makes this hard to use. Please use {{TAG\|PLUGINS/FORCE_AND_STRESS}} in combination with {{TAG\|PLUGINS/ML_MODE\|run}}.}}
	----		----

Line 27:		Line 28:
	{{TAG\|PLUGINS/FORCE_AND_STRESS}},		{{TAG\|PLUGINS/FORCE_AND_STRESS}},
	{{TAG\|PLUGINS/LOCAL_POTENTIAL}},		{{TAG\|PLUGINS/LOCAL_POTENTIAL}},
			{{TAG\|PLUGINS/ML_MODE}},
	{{TAG\|PLUGINS/STRUCTURE}}		{{TAG\|PLUGINS/STRUCTURE}}

Huebsch at 13:10, 22 April 2025

2025-04-22T13:10:29Z

← Older revision		Revision as of 13:10, 22 April 2025
Line 13:		Line 13:
	The <code>machine_learning</code> Python function is different from the other plugins because it does not provide any input from VASP.		The <code>machine_learning</code> Python function is different from the other plugins because it does not provide any input from VASP.
	Instead the plugin infrastructure handles the communication with VASP.		Instead the plugin infrastructure handles the communication with VASP.
	The only requirement of the plugin is that the function returns a class compatible with an [https://wiki.fysik.dtu.dk/ase/development/calculators.html\|ASE Calculator].		The only requirement of the plugin is that the function returns a class compatible with an [https://wiki.fysik.dtu.dk/ase/development/calculators.html ASE Calculator].
	<syntaxhighlight lang="python" line>		<syntaxhighlight lang="python" line>
	def machine_learning():		def machine_learning():

Schlipf: /* Related tags and articles */

2024-12-19T08:21:46Z

Schlipf: Created page with "{{TAGDEF|PLUGINS/MACHINE_LEARNING| .True. {{!}} .False.|.False.}} Description: {{TAG|PLUGINS/MACHINE_LEARNING}} calls the Python plugin for the machine learning interface for each ionic relaxation step ---- When {{TAG|PLUGINS/MACHINE_LEARNING}}=.TRUE., VASP calls the `machine_learning` Python function at the end of each ionic relaxation step. You can use this tag to replace VASP forces and the stress tensor to represent an external machine-learned interatom..."

2024-12-19T08:07:38Z

Created page with "{{TAGDEF|PLUGINS/MACHINE_LEARNING| .True. {{!}} .False.|.False.}} Description: {{TAG|PLUGINS/MACHINE_LEARNING}} calls the Python plugin for the machine learning interface for each ionic relaxation step ---- When {{TAG|PLUGINS/MACHINE_LEARNING}}=.TRUE., VASP calls the <code>machine_learning</code> Python function at the end of each ionic relaxation step. You can use this tag to replace VASP forces and the stress tensor to represent an external machine-learned interatom..."

New page

{{TAGDEF|PLUGINS/MACHINE_LEARNING| .True. {{!}} .False.|.False.}}

Description: {{TAG|PLUGINS/MACHINE_LEARNING}} calls the Python plugin for the machine learning interface for each ionic relaxation step
----

When {{TAG|PLUGINS/MACHINE_LEARNING}}=.TRUE., VASP calls the <code>machine_learning</code> Python function at the end of each ionic relaxation step.
You can use this tag to replace VASP forces and the stress tensor to represent an external machine-learned interatomic potential.
This can be used if you want to combine an machine-learned interatomic potential with the different features of {{TAG|IBRION}}.
For example, you could compute the phonon frequencies of the interatomic potential and compare it to ab-initio data and be sure that the workflow is the same.

==Expected inputs==

The <code>machine_learning</code> Python function is different from the other plugins because it does not provide any input from VASP.
Instead the plugin infrastructure handles the communication with VASP.
The only requirement of the plugin is that the function returns a class compatible with an [https://wiki.fysik.dtu.dk/ase/development/calculators.html|ASE Calculator].
<syntaxhighlight lang="python" line>
def machine_learning():
# code to setup calculator
return calculator
</syntaxhighlight>
VASP will take care to convert its internal data to be compatible with the Calculator.
We compute the forces and stress using the calculator on this data.
VASP packages the results on the Python side and sends them back to Fortran where they replace the usual force and stress values.

== Related tags and articles ==
[[Plugins]],
{{TAG|PLUGINS/FORCE_AND_STRESS}},
{{TAG|PLUGINS/LOCAL_POTENTIAL}},
{{TAG|PLUGINS/STRUCTURE}}

[[Category:INCAR tag]]

← Older revision		Revision as of 08:21, 19 December 2024
Line 28:		Line 28:
	{{TAG\|PLUGINS/LOCAL_POTENTIAL}},		{{TAG\|PLUGINS/LOCAL_POTENTIAL}},
	{{TAG\|PLUGINS/STRUCTURE}}		{{TAG\|PLUGINS/STRUCTURE}}

			{{sc\|PLUGINS/MACHINE_LEARNING\|Examples\|Examples that use this tag}}

	[[Category:INCAR tag]]		[[Category:INCAR tag]]