Skip to content

Commit

Permalink
#58 updating pyGSM parameters
Browse files Browse the repository at this point in the history
  • Loading branch information
joshkamm committed May 3, 2024
1 parent 6f9e4cd commit 6c5c00e
Showing 1 changed file with 3 additions and 3 deletions.
6 changes: 3 additions & 3 deletions examples/QChem/CDFT_MECP/MECP_Opt.py
Original file line number Diff line number Diff line change
Expand Up @@ -245,9 +245,9 @@ def main():

optimizer = eigenvector_follow.from_options(
Linesearch='backtrack', # a step size algorithm
OPTTHRESH=0.0005, # The gradrms threshold, this is generally easy to reach for large systems
OPTTHRESH=0.001, # The gradrms threshold, this is generally easy to reach for large systems
DMAX=0.01, # The initial max step size, will be adjusted if optimizer is doing well. Max is 0.5
conv_Ediff=0.1, # convergence of difference energy
conv_Ediff=1.0, # convergence of difference energy
conv_dE=0.1, # convergence of energy difference between optimization steps
conv_gmax=0.005, # convergence of max gradient
opt_cross=True, # use difference energy criteria to determine if you are at crossing
Expand All @@ -257,7 +257,7 @@ def main():
geoms, energies = optimizer.optimize(
molecule=initial,
refE=initial.energy,
opt_steps=150, # The max number of optimization steps, use a small number until you have your final sigma
opt_steps=250, # The max number of optimization steps, use a small number until you have your final sigma
verbose=True,
# opt_type='MECI', # JOSH
opt_type='UNCONSTRAINED', # JOSH
Expand Down

0 comments on commit 6c5c00e

Please sign in to comment.