Closed youli-jlu closed 3 months ago
Hi youli-jlu,
If I understand correctly, the issue is rather with ModelCheckpoint()
than with the optimizer. Since I don't have any experience with the lightning.pytorch
package, I don't think I can help you. Sorry!
Thank you for your quick response. I am going to submit an issue to lightning.
Hi ltatzel ,
Your Hessian-Free LM optimizer performs very well, and I want to use it to replace the pytorch L-BFGS optimizer. However, the model can't be saved normally if I use the ModelCheckpoint() in pytorch-lighting. You can find my test python file in the attachment. Could you give me some suggestions to handle this problem?
Thanks.