mlcommons / mlcommons/algorithmic-efficiency
Test resetting model parameters via `workload.init_model_fn()` during training
Open
@pomonam is already working on this.
Since Feb 17, 2023.
- Dominant language
- Python
- Stars
- 425
- Forks
- 78
- PR merge metrics
- No merged PRs in 30d
Description
According to the rules, it is expected that the model parameters can be reset by calling workload.init_model_fn(). However, creating a new model instance, like it is currently done in the MNIST PyTorch workload, might lead to out-of-memory errors, if the loss/optimizer and potentially other objects are not deleted properly (also described here). Hence, resetting model paramerters during training should be tested and workload.init_model_fn() should potentially be adjusted.
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.