Fastai ML course using Google Colab

Hi i am using Google colab for the ML course. I am using fasti library 0.7. I have the following issues .

When i try running the random forest regressor and calculate rmse with math.sqrt i get the error

Here is the code:
def rmse(x,y): return math.sqrt(((x-y)**2).mean())
Error -
/usr/local/lib/python3.6/dist-packages/pandas/core/ in wrapper(self) 110 return converter(self.iloc[0]) 111 raise TypeError("cannot convert the series to " --> 112 “{0}”.format(str(converter))) 113 114 return wrapper

TypeError: cannot convert the series to <class ‘float’>.

I tried using numpy.sqrt instead of math.sqrt and i am not getting any error.
Code using numpy.sqrt-
def rmse(x,y): return np.sqrt(((x-y)**2).mean())

However instead of getting one value for rmse i am getting the following output.

[SalesID 16167.232432
MachineID 16831.173108
ModelID 2101.414677
datasource 0.207197
auctioneerID 4.774846
YearMade 110.080494
MachineHoursCurrentMeter 3215.001167
UsageBand 0.253528
fiModelDesc 366.010106
fiBaseModel 135.708634
fiSecondaryDesc 13.634496
fiModelSeries 10.350931
fiModelDescriptor 10.571506
ProductSize 0.712905
fiProductClassDesc 7.074797
state 3.589563
ProductGroup 0.526127
ProductGroupDesc 0.526127
and so on …

Any help would be appreciated. The notebook is available at

Any help would be appreciated