Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / LarsKue/lightning-trainable issues and pull requests
#26 - omegaconf
Issue -
State: closed - Opened by ThisIsForReview 11 months ago
- 1 comment
#25 - No last.ckpt
Issue -
State: closed - Opened by ThisIsForReview 11 months ago
- 1 comment
Labels: help wanted, question
#24 - Cannot run lightning-trainable example
Issue -
State: closed - Opened by ThisIsForReview 11 months ago
- 6 comments
Labels: documentation, help wanted
#23 - Allow single-layer (linear) `FullyConnectedNetwork`
Issue -
State: closed - Opened by LarsKue 12 months ago
- 1 comment
Labels: enhancement
#22 - No support for cpu only
Issue -
State: closed - Opened by thelostscout 12 months ago
- 2 comments
Labels: enhancement
#21 - HParams with dicts (i.e. almost all) store dicts as special AttributeDict type in `hparams.yaml`
Issue -
State: closed - Opened by fdraxler 12 months ago
- 3 comments
Labels: bug
#20 - HParams validation batch has `self.training = False`
Issue -
State: closed - Opened by fdraxler 12 months ago
- 1 comment
Labels: bug
#19 - Test case for workaround for hparams loading issues
Pull Request -
State: closed - Opened by fdraxler 12 months ago
- 2 comments
#18 - load_from_checkpoint loads hparams as dict instead of class HParams
Issue -
State: closed - Opened by thelostscout about 1 year ago
- 6 comments
Labels: question
#17 - Cannot assign new keys to hparams of type dict
Issue -
State: closed - Opened by thelostscout about 1 year ago
#16 - on_train_start does not transfer batch to self.device
Issue -
State: closed - Opened by LarsKue about 1 year ago
Labels: bug
#15 - Extract PyTorch Lightning LR scheduler kwargs from scheduler kwargs
Pull Request -
State: closed - Opened by fdraxler about 1 year ago
- 1 comment
#14 - More Modules, Metrics, Bugfixes, and more
Pull Request -
State: closed - Opened by LarsKue about 1 year ago
#13 - Copying HParams returns dict
Issue -
State: closed - Opened by thelostscout about 1 year ago
- 1 comment
Labels: bug
#12 - How to validate without logger?
Issue -
State: closed - Opened by fdraxler about 1 year ago
- 5 comments
Labels: enhancement
#11 - Use default engine for resolving class specs for loggers as well
Issue -
State: closed - Opened by fdraxler about 1 year ago
- 3 comments
#10 - Continued training `model.fit()` creates spurious validation entries in external log
Issue -
State: closed - Opened by fdraxler about 1 year ago
- 5 comments
Labels: bug
#9 - Consistent class loading by name, support other loggers
Pull Request -
State: closed - Opened by fdraxler about 1 year ago
#8 - Progressbar confusing when continuing training of a loaded model.
Issue -
State: closed - Opened by HeiseMax over 1 year ago
- 1 comment
Labels: bug, enhancement, good first issue
#7 - pip doesn't install the dependency tomli
Issue -
State: closed - Opened by Fjallripa over 1 year ago
- 1 comment
#6 - Don't modify originally passed hparams
Pull Request -
State: closed - Opened by fdraxler over 1 year ago
#5 - Start from checkpoint
Pull Request -
State: closed - Opened by fdraxler over 1 year ago
#4 - Collection of launcher scripts
Pull Request -
State: closed - Opened by fdraxler over 1 year ago
#3 - Better Optimizer and LR Scheduler kwargs specification + DenseTrainable
Pull Request -
State: closed - Opened by RussellALA over 1 year ago
#2 - Automatically convert dicts to nested HParams
Pull Request -
State: closed - Opened by fdraxler over 1 year ago
- 7 comments
#1 - Suggestion: Use epoch progress bar
Pull Request -
State: closed - Opened by fdraxler almost 2 years ago
- 1 comment