After big change about level of theory, there are some missing variables.
I fix the orca but I'm not sure if these modifies are correct or may cause something wrong.
What I am worried about is the speed.
I compare the two version with the same setting(cores, reactant, level of theory and coordinate ...).
first one: latest version until '9c07d6db570f51c6965115e45c4dd74a8313f809' with my modify.
Total time cost: ~54 mins
opt_iter: 0 totalgrad: 0.215 gradrms: 0.007427 max E(1) 54.77
opt_iter: 1 totalgrad: 0.0899 gradrms: 0.003027 max E(1) 68.56
opt_iter: 2 totalgrad: 0.0512 gradrms: 0.001808 max E(1) 68.53
opt_iter: 3 totalgrad: 0.116 gradrms: 0.004548 max E(1) 72.66
opt_iter: 4 totalgrad: 0.0743 gradrms: 0.002792 max E(1) 71.04
opt_iter: 5 totalgrad: 0.045 gradrms: 0.001614 max E(1) 69.1
compared one: 'Fix typo in base_Gsm' '2745fb2417f21466ba3d33d20c6c0df11ce47fa3'
Total time cost: ~9 mins
opt_iter: 0 totalgrad: 0.217 gradrms: 0.007518 max E(1) 54.6
opt_iter: 1 totalgrad: 0.106 gradrms: 0.003661 max E(1) 68.84
opt_iter: 2 totalgrad: 0.0476 gradrms: 0.001667 max E(1) 70.74
opt_iter: 3 totalgrad: 0.0577 gradrms: 0.002293 max E(1) 72.13
opt_iter: 4 totalgrad: 0.0564 gradrms: 0.002127 max E(1) 70.85
opt_iter: 5 totalgrad: 0.0361 gradrms: 0.001351 max E(1) 69.27
opt_iter: 6 totalgrad: 0.0796 gradrms: 0.003783 max E(1) 68.42
Though the latest one have about the same energy and lower gradient, it also cost ~6 times more time.
I am not sure it cause by what I modify, some variable refactor or something else.
After big change about level of theory, there are some missing variables. I fix the orca but I'm not sure if these modifies are correct or may cause something wrong.
What I am worried about is the speed. I compare the two version with the same setting(cores, reactant, level of theory and coordinate ...). first one: latest version until '9c07d6db570f51c6965115e45c4dd74a8313f809' with my modify. Total time cost: ~54 mins
compared one: 'Fix typo in base_Gsm' '2745fb2417f21466ba3d33d20c6c0df11ce47fa3' Total time cost: ~9 mins
Though the latest one have about the same energy and lower gradient, it also cost ~6 times more time. I am not sure it cause by what I modify, some variable refactor or something else.