torch text

data.csv

torchtext_str2int_labelfield.py

# credits to: bentrevett

import torch
import torchtext
from torchtext import data

str2int = lambda x : int(x)

NAME = data.Field()
AGE = data.LabelField(preprocessing=str2int)

fields = [('n', NAME), ('a', AGE)]

train, test = data.TabularDataset.splits(
                path = '',
                train = 'data.csv',
                test = 'data.csv',
                format = 'csv',
                fields = fields,
                skip_header = True,
)

print(vars(train[0]))

# output
# (torch)  λ python torchtext_str2int_labelfield.py 
# {'n': ['bob'], 'a': 47}
# (torch)  λ