...
Results of fits
Objective fn. | Full | Un-penalized | |
Fit 0: TIG* | 5.4766E+035.4417E+03 | ||
Fit 1: TIG* without filtering ring-torsions | 5.4807E+03 | 5.4562E+03 | |
Fit 2: non-interpolated with 2 phases | 3.0181E+05 | 3.0181E+05 | |
Fit 3: non-interpolated with 1 phase running | running | ||
Chaya’s dataset only using fit0-FF | 3.2455E +02 3.1353E+02 | ||
OpenFF_1.3.0 _fb-fit2 | 1.5799E+03 | 1.5654E+03 | |
OpenFF_1.2.0 | 6.8768E+03 | 6.3742E+03 | |
OpenFF_1.0.0_RC2 | 4.5223E+03 | 4.2709E (Iter 0 on TIG dataset) | 5.9620e+03 |
For a first pass the objective function value didn’t look far away from 1.3.0. Some of the better looking TD curves are:
...