Training model to guess gender from first name: low accuracy and names become numbers?

Part 1 2022
1

I’ve done Part 1 lesson 1 of the course. I’m trying to train a model to guess the gender for a given first name.

I have found a dataset that lists genders and names which I’m using.

The code runs, but the accuracy isn’t great (60%), and when I learn.show_results() , the Name column is all numbers instead of the actual first name.

I’m not 100% sure what I’m doing, for example if I don’t have procs=[Categorify] , I’m getting a TypeError when trying to train the model but I don’t understand why. This is the notebook: firstname-recognizer | Kaggle.

Any pointers?

2

Hi,

Maybe this will help U:

Stay Cool,

MCZ