coveralls
commented
3 years ago Pull Request Test Coverage Report for Build a45220a9-7dbf-47b5-9700-1e0f33a5d7a4
- 21 of 23 (91.3%) changed or added relevant lines in 5 files are covered.
- 2 unchanged lines in 1 file lost coverage.
- Overall coverage increased (+0.03%) to 60.7%
| Changes Missing Coverage | Covered Lines | Changed/Added Lines | % | ||
|---|---|---|---|---|---|
| wandb/lib/apikey.py | 5 | 6 | 83.33% | ||
| wandb/sdk/wandb_login.py | 10 | 11 | 90.91% | ||
| <!-- | Total: | 21 | 23 | 91.3% | --> |
| Files with Coverage Reduction | New Missed Lines | % | ||
|---|---|---|---|---|
| wandb/compat/tempfile.py | 2 | 59.09% | ||
| <!-- | Total: | 2 | --> |
| Totals | |
|---|---|
| Change from base Build e1e60f57-21e8-4cfc-9cc5-07858bcf0912: | 0.03% |
| Covered Lines: | 12795 |
| Relevant Lines: | 21079 |
This is getting us closer to where we need to be. The login refactor wont be so bad at this point.
The key part of this is making sure the wandb.setup singleton gets these early login flags. The way things were written it is too easy to assume you have the unified settings state.
There are definitely still issues hiding... Like if in a notebook you do: wandb.login(key="badkeey") Then wandb.login(key="goodkey")
this is because the singleton settings object is frozen and cant be updated with the goodkey. I think this should wait to bee fixed up later