Closed deepakdinesh1123 closed 8 months ago
I ran the benchmarks against the branches starting from stable/2.2.x till main, there were no regressions.
@deepakdinesh1123 -- did you want to have a go at rebasing this one?
@smithdc1 I ran the benchmark now and the time took was 7.29±0.1μs
should I add a few more model.full_clean()
calls?
Yes. I think that will be a good idea.
Having had a look at full_clean()
it doesn't look like there is any caching of the result which takes place so should be fine to rerun. Maybe check that if you add 10x calls the duration goes up by a similar ratio.
After adding more the calls the time took was 73.2±0.8μs
. I'll commit the changes
The model validation benchmark has been added to the models_bench directory.