nan or gradient explosion problem even using FP32
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 9k
- Forks
- 1.5k
- Avg merge
- 2d 4h
- Merged PRs (30d)
- 3
Description
Hi, I'm using Laplace loss for the offset vector regression task. Even if I manually set the loss to zero at each batch, the optimizer.step() still arise inf/nan problem. Actually, the laplace loss is limited and the training break at optimizer.step().
Could anyone give me some help? Thanks!
Contributor guide
No contributing guide indexed for this repository
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.
Research direction
No source file, test, or reproducible example is provided. Start by reproducing the FP32 Laplace-loss offset-regression case and inspect the optimizer.step() inputs for the first inf or nan; done means identifying a reproducible cause and documenting or fixing it.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100