You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
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
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?
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?