ML ABN: Difference between revisions
No edit summary |
No edit summary |
||
(2 intermediate revisions by 2 users not shown) | |||
Line 1: | Line 1: | ||
{{DISPLAYTITLE:ML_ABN}} | {{DISPLAYTITLE:ML_ABN}} | ||
In context of the machine-learned force fields method this file serves as an output for the collected training data set. When running {{VASP}} with {{TAGO|ML_MODE|train, select}} it is automatically written out whenever a learning step is performed. The file format is shared with the {{FILE|ML_AB}} input file, for further details please have a look at the {{FILE|ML_AB}} Wiki entry. An {{FILE|ML_ABN}} file can be reused as a starting point for continuation runs ({{TAGO|ML_MODE|train}}) and local reference configuration re-selection runs ({{TAGO|ML_MODE|select}}) simply by renaming (or copying) it to {{FILE|ML_AB}}. | |||
---- | ---- | ||
[[Category:Files]][[Category:Machine | [[Category:Files]][[Category:Machine-learned force fields]][[Category:Output Files]] |
Latest revision as of 14:35, 19 October 2023
In context of the machine-learned force fields method this file serves as an output for the collected training data set. When running VASP with ML_MODE = train, select
it is automatically written out whenever a learning step is performed. The file format is shared with the ML_AB input file, for further details please have a look at the ML_AB Wiki entry. An ML_ABN file can be reused as a starting point for continuation runs (ML_MODE = train
) and local reference configuration re-selection runs (ML_MODE = select
) simply by renaming (or copying) it to ML_AB.