Documentation in markdown for user custom dataset usage

if anyone wants to train on his data :

1. he puts the excel file into his project
2. he goes to dataset_description.py
3. he fills the datasetDescription with useful infos, for exemple

class Dataset:
    PUBLICATION_DATASET = ExcelDatasetDescriptor(
        "IRRADIATION",
        "DC_POWER",
        "pub/publishable_data.xlsx",
        "Sheet1"
    )

4. If he want to use the dataset he just described, he can now specify it in the config.py file (in the latest commit)

This guide needs to be reviewed, formatted and put in the README.md, maybe add some instructions for other file types

Start with the dataset_description.py example and the config.py usage described in the issue, then review the existing README.md. Document the custom Excel dataset workflow clearly, format the example, and determine whether guidance for other file types can be added. Done means the README contains an accurate, understandable guide for using a custom dataset.