########################################################
README for the evaluation scripts that we need. Again, the goal is to make them
act as "functions" (inputs discussed below) to make it easy to run multiple things 
with each other. That being said, would be a good idea to make different folders
for each type of evaluation (RMSE, MAPE, etc) in case different models need
different scripts for the same evaluation type.
########################################################

toy_eval.py: contains evaluation functions (spearmans, pearsons) checked that weren't given by tensorflow.
