Skip to content

After the multi-temperature zone convergence of dpgen2 based on a pre-trained model, is it necessary to collect all the configurations labeled in previous rounds for long training? #342

@sau-wen

Description

@sau-wen

After multi-temperature zone convergence, is it necessary to collect all the configurations labeled in all previous iteration rounds for the final long training? Or can the converged model be directly used for production?

Metadata

Metadata

Assignees

No one assigned

    Labels

    questionFurther information is requested

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions