Create a `samples` directory, with sample data in text and/or JSON

[aih]

We have the whole dataset stored on Huggingface and in Google Drive. Perhaps we can describe a testing process for a new model, or include a local set of ~ 10 bill texts to test any summarization model against. These bills should include the three that are described in the Wiki.

[BorodaUA]

The example dataset added in the samples folder https://github.com/dreamproit/BillML/blob/main/samples/wiki_9_bills_example_dataset.jsonl