Update the default batch size to 16 in our benchmarking as that's the one we use most commonly.
Importantly, add the skip_postprocessing parameter which creates a custom Mesmer application with a no-op post-processing step. Note that this makes the results unintelligible. It's reflected in the results with a different deepcell version: _nopost.
Update the default batch size to 16 in our benchmarking as that's the one we use most commonly.
Importantly, add the
skip_postprocessing
parameter which creates a custom Mesmer application with a no-op post-processing step. Note that this makes the results unintelligible. It's reflected in the results with a different deepcell version:_nopost
.Paired with @WeihaoGe1009
Fixes #172