lincc-frameworks / lincc-frameworks/hyrax
Make Trace work with splits
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 41
- Forks
- 7
- Avg merge
- 5d 1h
- Merged PRs (30d)
- 8
Description
This is coming out of discussion on #783 Roughly we have an issue with trace mode's early stopping combined with splits.
Today it works only sometimes, but we would like to make it work all the time. The interface needed (probably on DataProvider) is an authoritative list of indexes that will be accessed by split (e.g. train: [1,2,4...] validate:[0,3,...])
Then when Trace is on, and it must do an early stop at a particular batch size (e.g. trace=2) we can go look at the first two of each index list, find the largest, and then limit DataProvider at that length. (e.g. with values above the max index is 3, so the length is 4)
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
Start with the discussion in #783, then locate the Trace and DataProvider entry points involved in split handling and early stopping. Confirm how split indexes are currently determined. Done means DataProvider exposes authoritative per-split indexes and Trace consistently limits processing to the required length for every split.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- data, machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100