Save checkpoint after certain number of iterations
sparshgarg23 opened this issue · comments
Currently checkpoint is being saved after 1 epoch.Since I am training on colab at a time I can only use one gpu which is causing the training to proceed very slowly.
With that in mind,what changes should I make to the config so as to save checkpoint after every 500 iterations?
I was able to discuss this at mmdetection.
it seems the above idea of saving checkpoint after N iterations can't be done because the mmcv version of this repository doesn't support by_epoch=False.
As such would it be possible to update the repository so that it's consistent with the current mmcv pytorch and mmdetection standard?