Create an embedding_vector_conversion function

Explore embedding vector conversions with torch

Extend with more tests for the create dataloader function to show case different strides/max_length

Implement the GPTDatasetV1 and the create_dataloader_v1 method.

torch will need numpy, therefore add that to the project

Add pytorch as a dependency to the project

Add some example code for the data loader concept

Update the README with some references from chapter one

Create a test for the exercise 2.1 byte pair encoding of unknown words

Add the tiktoken library as a dependency to the project

Implement the SimpleTokenizerV2 including tests

Update the creation of the vocabulary so it will add the two special token, <|endoftext|> and <|unknown|>, to the list

Create a test that shows what happends when a word is missing from the vocabulary

Add a uv sync rule to the Makefile

Refactor the Makefile and add format and isort rules.