How to export fp32 master weights with opt_level O2
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 9k
- Forks
- 1.5k
- Avg merge
- 2d 4h
- Merged PRs (30d)
- 3
Description
My model is training with opt_level O2, but it will be used in fp32 mode. How can I save the state_dict of model?
For now, I just use
model, optimizer = amp.initialize(model, optimizer, opt_level="O2"
# train the model
torch.save(model.state_dict(), "model.pth")
# use the model in float32 mode
Should I do something to save the fp32 master weights rather than the model.state_dict() which has fp16 weights?
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 files or tests are named. Start with the Apex AMP entry point amp.initialize at opt_level="O2" and the model.state_dict() call, then check the project's guidance for checkpointing and fp32 master weights. Done means a documented, verified procedure for exporting weights that can be loaded for fp32 use.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Documentation
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100