Skip to content

GitLab

Commit 1c378970 authored by Gao, Xiang's avatar Gao, Xiang Committed by Farhad Ramezanghorbani
Browse files

Force coefficient is better to be 0.1 (#249) 

Being 1 does not work well
parent e1c78efe
Show whitespace changes
Inline Side-by-side
Showing with 1 addition and 1 deletion
examples/nnp_training_force.py
View file @ 1c378970
...@@ -201,7 +201,7 @@ if 'scheduler' in checkpoint: ...@@ -201,7 +201,7 @@ if 'scheduler' in checkpoint:
print("training starting from epoch", scheduler.last_epoch + 1) print("training starting from epoch", scheduler.last_epoch + 1)
max_epochs = 20 max_epochs = 20
early_stopping_learning_rate = 1.0E-5 early_stopping_learning_rate = 1.0E-5
force_coefficient = 1 # controls the importance of energy loss vs force loss force_coefficient = 0.1 # controls the importance of energy loss vs force loss
best_model_checkpoint = 'force-training-best.pt' best_model_checkpoint = 'force-training-best.pt'
for _ in range(scheduler.last_epoch + 1, max_epochs): for _ in range(scheduler.last_epoch + 1, max_epochs):
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment