Lightning-AI / Lightning-AI/pytorch-lightning
Load a checkpoint after warmup
@Felonious-Spellfire is already working on this.
Since Sep 4, 2022.
- Dominant language
- Python
- Stars
- 31.4k
- Forks
- 3.8k
- Avg merge
- 6d 7h
- Merged PRs (30d)
- 6
Description
## 📚 Documentation
https://pytorch-lightning.slack.com/team/U01H0HH3USF
Set's say I load a checkpoint, but then want to overwrite the learning rate schedulers. I'm using [pl_bolts.optimizers.lr](http://pl_bolts.optimizers.lr/)_scheduler.LinearWarmupCosineAnnealingLR which is deprecated :disappointed:
But basically it take 10 hours to do warmup. What I want to do is load the checkpoint from after warmup, and then just play with my cosine annealing scheduler. How can I do this? (I am worried there might be secret state in Lightning that I am not aware of.)
cc @borda @rohitgr7 @Felonious-Spellfire
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.