How to preprocess address, latitude and longitude features in tabular data for Pytorch

Question

I have cleared my data into the next 6 columns that you see ahead that it is my input data. I split the dataset to have the label in another variable Y.

My main problem: I don't know how to preprocess the data to have a good input to any model.

My dataset X looks like this:

desc	tipo	address	region	latitude	longitude
Galpón	Industria	Subdivisión de la Finca Denominada Violeta S/N	Región de Arica y Parinacota-Arica	-19.423411	-11.371551

• desc - string
• tipo - string
• address - string
• region - string
• latitude - string
• longitude - string

My dataset Y looks like

CIO
169379

What I tried

I have followed this tutorial that allows me to comprehend a little bit about tabular data, but the data is completely different and I don't know if it fits me as well. So, my code transformed all the data to a LabelEncoder, but it is obvious that doesn't apply to latitude and longitude.

for col in df.columns:
    if df.dtypes[col] == "object":
        df[col] = df[col].fillna("NA")
    else:
        df[col] = df[col].fillna(0)
    df[col] = LabelEncoder().fit_transform(df[col])

for col in df.columns:
    df[col] = df[col].astype('category')

Also, the author used some Categorical Embedding that I don't know if works properly with my kind of data too.