-
Notifications
You must be signed in to change notification settings - Fork 249
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add documentation and helper functions to store and retrieve checkpoints #127
Comments
Is this issue free to work on? |
Yes, it is. However, parts may change with the ongoing refactoring so I'd recommend to work on something else atm. |
Sure
…On Mon, 20 Sep 2021, 12:27 Philipp Wirth, ***@***.***> wrote:
Yes, it is. However, parts may change with the ongoing refactoring so I'd
recommend to work on something else atm.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#127 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AJUIFVBJM2GUNHSNGHTBLC3UC3LL3ANCNFSM4XSGU2VA>
.
|
Has this been added to the documentation? |
Hi @shree-lily, we sadly did not add this to the documentation yet. You can still use either pytorch or pytorch lightning to save and load checkpoints:
Taking our example code for the SimCLR model here: https://docs.lightly.ai/examples/simclr.html You can save and load the model as follows:
It is common to drop the projection heads for downstream or inference tasks and only use the backbone instead. You can get the backbone with the |
We now have a tutorial on how to finetune checkpoints from lightly: https://docs.lightly.ai/self-supervised-learning/tutorials/package/tutorial_checkpoint_finetuning.html We also plan to upload the benchmark backbones: #1621 |
We have a model zoo but no documentation on how to load the models from it. We also have the PyTorch Lightning wrapper for training the models but no information on how to use the models afterward.
It would be great to provide the following (documentation and if it makes sense helper functions):
How to load a model that has been saved by PyTorch Lightning (e.g. I just want to get the ResNet backbone from it)?
How to manually save and load a model?
The text was updated successfully, but these errors were encountered: