Machine learning force field calculations: Basics - Revision history

Karsai: /* Step-by-step instructions */

2026-03-20T07:54:42Z

Step-by-step instructions

← Older revision		Revision as of 07:54, 20 March 2026
Line 4:		Line 4:

	On-the-fly training is based on molecular-dynamics (MD) simulations to sample training structures. Piece by piece a data set is automatically assembled and used to generate an MLFF whenever feasible. Conversely, at each time step the current force field predicts energy, forces and the corresponding Bayesian error estimations. Simply put, if the error is above a certain threshold another ab-initio calculation is performed and the reference energy and forces are added to the training data set. In the opposite case, the ab-initio step is omitted and the system is propagated via MLFF predictions. As the force field gets better along the trajectory many ab-initio steps can be avoided and the MD simulation is significantly accelerated. Ultimately, the on-the-fly training results in an MLFF which is ready for production, i.e., running an MD simulation in prediction-only mode.		On-the-fly training is based on molecular-dynamics (MD) simulations to sample training structures. Piece by piece a data set is automatically assembled and used to generate an MLFF whenever feasible. Conversely, at each time step the current force field predicts energy, forces and the corresponding Bayesian error estimations. Simply put, if the error is above a certain threshold another ab-initio calculation is performed and the reference energy and forces are added to the training data set. In the opposite case, the ab-initio step is omitted and the system is propagated via MLFF predictions. As the force field gets better along the trajectory many ab-initio steps can be avoided and the MD simulation is significantly accelerated. Ultimately, the on-the-fly training results in an MLFF which is ready for production, i.e., running an MD simulation in prediction-only mode.
	{{NB\|warning\|When compiling with shared memory MPI support (-Duse_shmem), it is utterly important to pin the MPI ranks to the physical cores of the node.For guidance on how to understand the hardware topology of your system and how to correctly set up rank pinning accordingly, refer to [[Optimizing the parallelization#Understanding the hardware\|here]].}}		{{NB\|warning\|When compiling with shared memory MPI support (-Duse_shmem), it is utterly important to pin the MPI ranks to the physical cores of the node. For guidance on how to understand the hardware topology of your system and how to correctly set up rank pinning accordingly, refer to [[Optimizing the parallelization#Understanding the hardware\|here]].}}

	The following steps outline the path from start to production run:		The following steps outline the path from start to production run:

Karsai: /* Step-by-step instructions */

2026-03-20T07:41:50Z

Step-by-step instructions

← Older revision		Revision as of 07:41, 20 March 2026
Line 4:		Line 4:

	On-the-fly training is based on molecular-dynamics (MD) simulations to sample training structures. Piece by piece a data set is automatically assembled and used to generate an MLFF whenever feasible. Conversely, at each time step the current force field predicts energy, forces and the corresponding Bayesian error estimations. Simply put, if the error is above a certain threshold another ab-initio calculation is performed and the reference energy and forces are added to the training data set. In the opposite case, the ab-initio step is omitted and the system is propagated via MLFF predictions. As the force field gets better along the trajectory many ab-initio steps can be avoided and the MD simulation is significantly accelerated. Ultimately, the on-the-fly training results in an MLFF which is ready for production, i.e., running an MD simulation in prediction-only mode.		On-the-fly training is based on molecular-dynamics (MD) simulations to sample training structures. Piece by piece a data set is automatically assembled and used to generate an MLFF whenever feasible. Conversely, at each time step the current force field predicts energy, forces and the corresponding Bayesian error estimations. Simply put, if the error is above a certain threshold another ab-initio calculation is performed and the reference energy and forces are added to the training data set. In the opposite case, the ab-initio step is omitted and the system is propagated via MLFF predictions. As the force field gets better along the trajectory many ab-initio steps can be avoided and the MD simulation is significantly accelerated. Ultimately, the on-the-fly training results in an MLFF which is ready for production, i.e., running an MD simulation in prediction-only mode.

	{{NB\|warning\|When compiling with shared memory MPI support (-Duse_shmem), it is utterly important to pin the MPI ranks to the physical cores of the node.For guidance on how to understand the hardware topology of your system and how to correctly set up rank pinning accordingly, refer to [[Optimizing the parallelization#Understanding the hardware\|here]].}}		{{NB\|warning\|When compiling with shared memory MPI support (-Duse_shmem), it is utterly important to pin the MPI ranks to the physical cores of the node.For guidance on how to understand the hardware topology of your system and how to correctly set up rank pinning accordingly, refer to [[Optimizing the parallelization#Understanding the hardware\|here]].}}

Karsai: /* Step-by-step instructions */

2026-03-20T07:40:46Z

Step-by-step instructions

← Older revision		Revision as of 07:40, 20 March 2026
Line 3:		Line 3:
	== Step-by-step instructions ==		== Step-by-step instructions ==

	On-the-fly training is based on molecular-dynamics (MD) simulations to sample training structures. Piece by piece a data set is automatically assembled and used to generate an MLFF whenever feasible. Conversely, at each time step the current force field predicts energy, forces and the corresponding Bayesian error estimations. Simply put, if the error is above a certain threshold another ab-initio calculation is performed and the reference energy and forces are added to the training data set. In the opposite case, the ab-initio step is omitted and the system is propagated via MLFF predictions. As the force field gets better along the trajectory many ab-initio steps can be avoided and the MD simulation is significantly accelerated. Ultimately, the on-the-fly training results in an MLFF which is ready for production, i.e., running an MD simulation in prediction-only mode. The following steps outline the path from start to production run:		On-the-fly training is based on molecular-dynamics (MD) simulations to sample training structures. Piece by piece a data set is automatically assembled and used to generate an MLFF whenever feasible. Conversely, at each time step the current force field predicts energy, forces and the corresponding Bayesian error estimations. Simply put, if the error is above a certain threshold another ab-initio calculation is performed and the reference energy and forces are added to the training data set. In the opposite case, the ab-initio step is omitted and the system is propagated via MLFF predictions. As the force field gets better along the trajectory many ab-initio steps can be avoided and the MD simulation is significantly accelerated. Ultimately, the on-the-fly training results in an MLFF which is ready for production, i.e., running an MD simulation in prediction-only mode.

			{{NB\|warning\|When compiling with shared memory MPI support (-Duse_shmem), it is utterly important to pin the MPI ranks to the physical cores of the node.For guidance on how to understand the hardware topology of your system and how to correctly set up rank pinning accordingly, refer to [[Optimizing the parallelization#Understanding the hardware\|here]].}}

			The following steps outline the path from start to production run:

	''' Step 1: Prepare a molecular dynamics run '''		''' Step 1: Prepare a molecular dynamics run '''

Wolloch at 08:56, 16 March 2026

2026-03-16T08:56:22Z

← Older revision		Revision as of 08:56, 16 March 2026
Line 60:		Line 60:
	== Parallelization ==		== Parallelization ==

	At present, {{VASP}} provides only MPI-based parallelization for the MLFF feature. Therefore, any operational mode relying exclusively on MLFF code - such as predictive MD simulations ({{TAG\|ML_MODE}} = <code>run</code>) and local reference configuration selection ({{TAG\|ML_MODE}} = <code>select</code>) - cannot leverage alternative forms of parallelization like [[~~OpenACC~~ GPU ~~port~~ of VASP\|~~OpenACC~~ offloading to GPUs]] or [[Combining MPI and OpenMP\|an MPI/OpenMP hybrid approach]]. Conversely, a usual on-the-fly training involves both MLFF generation and ab-initio computations. When the latter component predominates in terms of computational demand, utilizing non-MPI parallelization remains practical.		At present, {{VASP}} provides only MPI-based parallelization for the MLFF feature. Therefore, any operational mode relying exclusively on MLFF code - such as predictive MD simulations ({{TAG\|ML_MODE}} = <code>run</code>) and local reference configuration selection ({{TAG\|ML_MODE}} = <code>select</code>) - cannot leverage alternative forms of parallelization like [[GPU ports of VASP\|offloading to GPUs]] or [[Combining MPI and OpenMP\|an MPI/OpenMP hybrid approach]]. Conversely, a usual on-the-fly training involves both MLFF generation and ab-initio computations. When the latter component predominates in terms of computational demand, utilizing non-MPI parallelization remains practical.

	[[Category:Machine-learned force fields]]		[[Category:Machine-learned force fields]]
	[[Category:Howto]]		[[Category:Howto]]

Huebsch: /* Step-by-step instructions */

2025-10-16T12:38:42Z

Step-by-step instructions

← Older revision		Revision as of 12:38, 16 October 2025
Line 7:		Line 7:
	''' Step 1: Prepare a molecular dynamics run '''		''' Step 1: Prepare a molecular dynamics run '''

	Prepare an [[#Category:Calculation setup\|ab-initio]] [[#Category:Molecular dynamics\|MD]] run with your desired {{FILE\|POSCAR}} starting configuration and an appropriate setup in {{FILE\|INCAR}}, {{FILE\|KPOINTS}} and {{FILE\|POTCAR}} files.		Prepare an [[:Category:Calculation setup\|ab-initio]] [[:Category:Molecular dynamics\|MD]] run with your desired {{FILE\|POSCAR}} starting configuration and an appropriate setup in {{FILE\|INCAR}}, {{FILE\|KPOINTS}} and {{FILE\|POTCAR}} files.

	''' Step 2: Start on-the-fly training from scratch '''		''' Step 2: Start on-the-fly training from scratch '''

Singraber at 09:40, 16 October 2025

2025-10-16T09:40:54Z

← Older revision		Revision as of 09:40, 16 October 2025
Line 41:		Line 41:
	In the {{FILE\|INCAR}} file set		In the {{FILE\|INCAR}} file set
	{{TAGBL\|ML_MODE}} = run		{{TAGBL\|ML_MODE}} = run
	With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries. However, the reverse is not guaranteed: if you provide an {{FILE\|ML_FF}} file to an older {{VASP}} version you may receive an error message stating that the file is too new and not supported. In order to check ~~for compatibility the force field files themselves are versioned and the version can be extracted from the~~ [[ML_FFN#File header\|file header]].		With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries (for details check [[ML_FFN#File header\|the ML_FFN file header]]).

	== Advice ==		== Advice ==

Singraber: /* Step-by-step instructions */

2025-10-16T08:58:56Z

Step-by-step instructions

← Older revision		Revision as of 08:58, 16 October 2025
Line 27:		Line 27:
	Leave <code>{{TAG\|ML_MODE}} = train</code> unchanged and restart {{VASP}}. The log file will contain a section describing the existing data set and after initial generation of a force field the regular on-the-fly procedure continues. In the end, the resulting {{FILE\|ML_ABN}} will contain the training structures from both on-the-fly runs. Similarly, the {{FILE\|ML_FFN}} file is a combined force field. In the presence of an {{FILE\|ML_AB}} file the <code>train</code> mode will always perform a continuation run. If you would like to start from scratch just remove the {{FILE\|ML_AB}} file from the execution directory. {{NB\|tip\|Apply this strategy repeatedly in order to systematically improve your MLFF, e.g., first train on water only, then on sodium chloride and finally, train on the combination of both.}}		Leave <code>{{TAG\|ML_MODE}} = train</code> unchanged and restart {{VASP}}. The log file will contain a section describing the existing data set and after initial generation of a force field the regular on-the-fly procedure continues. In the end, the resulting {{FILE\|ML_ABN}} will contain the training structures from both on-the-fly runs. Similarly, the {{FILE\|ML_FFN}} file is a combined force field. In the presence of an {{FILE\|ML_AB}} file the <code>train</code> mode will always perform a continuation run. If you would like to start from scratch just remove the {{FILE\|ML_AB}} file from the execution directory. {{NB\|tip\|Apply this strategy repeatedly in order to systematically improve your MLFF, e.g., first train on water only, then on sodium chloride and finally, train on the combination of both.}}

	''' Step 4: Refit for fast prediction mode ''' (available as of {{VASP}}.6.4.0)		''' Step 4: Refit for fast prediction mode ''' (available as of {{VASP}} 6.4.0)

	When on-the-fly training succeeded and the result matches your expectations with respect to applicability and residual errors there is one final step required before the force field should be applied in prediction-only MD runs: refitting for fast prediction mode. Copy once again the final data set to {{FILE\|ML_AB}}:		When on-the-fly training succeeded and the result matches your expectations with respect to applicability and residual errors there is one final step required before the force field should be applied in prediction-only MD runs: refitting for fast prediction mode. Copy once again the final data set to {{FILE\|ML_AB}}:

Singraber: /* Step-by-step instructions */

2025-10-16T08:58:09Z

Step-by-step instructions

← Older revision		Revision as of 08:58, 16 October 2025
Line 33:		Line 33:
	Also, set in the {{FILE\|INCAR}} file:		Also, set in the {{FILE\|INCAR}} file:
	{{TAGBL\|ML_MODE}} = refit		{{TAGBL\|ML_MODE}} = refit
	Running {{VASP}} will create a new {{FILE\|ML_FFN}} which finally can be used for production. {{NB\|important\|Although it is technically possible to continue directly with step 5 given a {{FILE\|ML_FFN}} file from steps 2 or 3 it is strongly discouraged. Without the refitting step {{VASP}} cannot enable the fast prediction mode which comes with speedup factor of approximately 20 to 100. You can check the {{FILE\|ML_FFN}} ASCII header to be sure whether the contained force field supports fast prediction.}}		Running {{VASP}} will create a new {{FILE\|ML_FFN}} which finally can be used for production. {{NB\|important\|Although it is technically possible to continue directly with step 5 given a {{FILE\|ML_FFN}} file from steps 2 or 3 it is strongly discouraged. Without the refitting step {{VASP}} cannot enable the fast prediction mode which comes with speedup factor of approximately 20 to 100. You can check the {{FILE\|ML_FFN}} ASCII [[ML_FFN#Fast_prediction_mode\|header]] to be sure whether the contained force field supports fast prediction.}}

	''' Step 5: Applying the machine-learned force field in production runs '''		''' Step 5: Applying the machine-learned force field in production runs '''
Line 41:		Line 41:
	In the {{FILE\|INCAR}} file set		In the {{FILE\|INCAR}} file set
	{{TAGBL\|ML_MODE}} = run		{{TAGBL\|ML_MODE}} = run
	With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries. However, the reverse is not guaranteed: if you provide an {{FILE\|ML_FF}} file to an older {{VASP}} version you may receive an error message stating that the file is too new and not supported. In order to check for compatibility the force field files themselves are versioned and the version can be extracted from the [[ML_FFN\|file header]].		With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries. However, the reverse is not guaranteed: if you provide an {{FILE\|ML_FF}} file to an older {{VASP}} version you may receive an error message stating that the file is too new and not supported. In order to check for compatibility the force field files themselves are versioned and the version can be extracted from the [[ML_FFN#File header\|file header]].

	== Advice ==		== Advice ==

Singraber: /* Step-by-step instructions */

2025-10-15T09:00:01Z

Step-by-step instructions

← Older revision		Revision as of 09:00, 15 October 2025
Line 41:		Line 41:
	In the {{FILE\|INCAR}} file set		In the {{FILE\|INCAR}} file set
	{{TAGBL\|ML_MODE}} = run		{{TAGBL\|ML_MODE}} = run
	With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}}		With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries. However, the reverse is not guaranteed: if you provide an {{FILE\|ML_FF}} file to an older {{VASP}} version you may receive an error message stating that the file is too new and not supported. In order to check for compatibility the force field files themselves are versioned and the version can be extracted from the [[ML_FFN\|file header]].

	== Advice ==		== Advice ==

Singraber at 13:19, 24 January 2025

2025-01-24T13:19:22Z

← Older revision		Revision as of 13:19, 24 January 2025
Line 27:		Line 27:
	Leave <code>{{TAG\|ML_MODE}} = train</code> unchanged and restart {{VASP}}. The log file will contain a section describing the existing data set and after initial generation of a force field the regular on-the-fly procedure continues. In the end, the resulting {{FILE\|ML_ABN}} will contain the training structures from both on-the-fly runs. Similarly, the {{FILE\|ML_FFN}} file is a combined force field. In the presence of an {{FILE\|ML_AB}} file the <code>train</code> mode will always perform a continuation run. If you would like to start from scratch just remove the {{FILE\|ML_AB}} file from the execution directory. {{NB\|tip\|Apply this strategy repeatedly in order to systematically improve your MLFF, e.g., first train on water only, then on sodium chloride and finally, train on the combination of both.}}		Leave <code>{{TAG\|ML_MODE}} = train</code> unchanged and restart {{VASP}}. The log file will contain a section describing the existing data set and after initial generation of a force field the regular on-the-fly procedure continues. In the end, the resulting {{FILE\|ML_ABN}} will contain the training structures from both on-the-fly runs. Similarly, the {{FILE\|ML_FFN}} file is a combined force field. In the presence of an {{FILE\|ML_AB}} file the <code>train</code> mode will always perform a continuation run. If you would like to start from scratch just remove the {{FILE\|ML_AB}} file from the execution directory. {{NB\|tip\|Apply this strategy repeatedly in order to systematically improve your MLFF, e.g., first train on water only, then on sodium chloride and finally, train on the combination of both.}}

	''' Step 4: Refit for fast prediction mode '''		''' Step 4: Refit for fast prediction mode ''' (available as of {{VASP}}.6.4.0)

	When on-the-fly training succeeded and the result matches your expectations with respect to applicability and residual errors there is one final step required before the force field should be applied in prediction-only MD runs: refitting for fast prediction mode. Copy once again the final data set to {{FILE\|ML_AB}}:		When on-the-fly training succeeded and the result matches your expectations with respect to applicability and residual errors there is one final step required before the force field should be applied in prediction-only MD runs: refitting for fast prediction mode. Copy once again the final data set to {{FILE\|ML_AB}}:

← Older revision		Revision as of 08:58, 16 October 2025
Line 33:		Line 33:
	Also, set in the {{FILE\|INCAR}} file:		Also, set in the {{FILE\|INCAR}} file:
	{{TAGBL\|ML_MODE}} = refit		{{TAGBL\|ML_MODE}} = refit
	Running {{VASP}} will create a new {{FILE\|ML_FFN}} which finally can be used for production. {{NB\|important\|Although it is technically possible to continue directly with step 5 given a {{FILE\|ML_FFN}} file from steps 2 or 3 it is strongly discouraged. Without the refitting step {{VASP}} cannot enable the fast prediction mode which comes with speedup factor of approximately 20 to 100. You can check the {{FILE\|ML_FFN}} ASCII header to be sure whether the contained force field supports fast prediction.}}		Running {{VASP}} will create a new {{FILE\|ML_FFN}} which finally can be used for production. {{NB\|important\|Although it is technically possible to continue directly with step 5 given a {{FILE\|ML_FFN}} file from steps 2 or 3 it is strongly discouraged. Without the refitting step {{VASP}} cannot enable the fast prediction mode which comes with speedup factor of approximately 20 to 100. You can check the {{FILE\|ML_FFN}} ASCII [[ML_FFN#Fast_prediction_mode\|header]] to be sure whether the contained force field supports fast prediction.}}

	''' Step 5: Applying the machine-learned force field in production runs '''		''' Step 5: Applying the machine-learned force field in production runs '''
Line 41:		Line 41:
	In the {{FILE\|INCAR}} file set		In the {{FILE\|INCAR}} file set
	{{TAGBL\|ML_MODE}} = run		{{TAGBL\|ML_MODE}} = run
	With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries. However, the reverse is not guaranteed: if you provide an {{FILE\|ML_FF}} file to an older {{VASP}} version you may receive an error message stating that the file is too new and not supported. In order to check for compatibility the force field files themselves are versioned and the version can be extracted from the [[ML_FFN\|file header]].		With this choice {{VASP}} will only use the predictions from the MLFF, no ab-initio calculations are performed. The execution time per time step will be orders of magnitude lower if compared with corresponding ab-initio runs. {{NB\|tip\|The MLFF can be transferred to larger system sizes, i.e., you may duplicate your simulation box to benefit from improved statistics. Because the method scales linearly with the number of atoms you can easily estimate the impact on computational demand.}} {{NB\|warning\|Please have a look at a [[Known issues\|known issue]] (from date 2023-05-17) if you intend to use triclinic geometries with small or large lattice angles.}} Usually, VASP machine-learned force field files {{FILE\|ML_FFN}} are backward compatible, i.e., files generated by an older version of {{VASP}} can still be used with newer {{VASP}} binaries. However, the reverse is not guaranteed: if you provide an {{FILE\|ML_FF}} file to an older {{VASP}} version you may receive an error message stating that the file is too new and not supported. In order to check for compatibility the force field files themselves are versioned and the version can be extracted from the [[ML_FFN#File header\|file header]].

	== Advice ==		== Advice ==