model comparison by param enumeration

[judithfan]

• [x] Set up bda-enumerate.wppl to sweep through the following grid, for each perception module ('human','mm_fc6','mm_conv42','mm_pool1'), for each split (1-5). Do this by writing wrapper that spawns different bda-enumerate.wppl processes for sub-regions of grid.
  • costWeight: [0,12] in steps of 0.5
  • simScaling: [0,200] in steps of 2
  • pragWeight: [0,1] in steps of 0.5
  • infWeight: [0,50] in steps of 1.