Term Frequency

1. Add a new term_frequencies attribute to your InvertedIndex class. It should be a dictionary mapping document IDs to Counter objects. A Counter is a dictionary optimized for counting.
2. Update your load and save methods to also load and save this new attribute into a new cache/term_frequencies.pkl file.
3. Update your __add_document method to update the term frequencies for each token in the document. For each token, increment its count in the Counter for that document ID.
4. Add a new get_tf(self, doc_id, term) method. It should return how many times the token appears in the document with the given ID. If the term doesn't appear in that document, return 0. Assume term is a single token.
5. Add a new helper function that tokenizes a single term.
   1. Use your existing tokenizer to tokenize the term.
   2. If the tokenizer doesn't return exactly one token, raise an exception.
   3. Return the single token.
6. Add a new tf command to the keyword_search_cli.py entrypoint script.
   1. It should take a document ID and a term as arguments.
   2. It should tokenize the term (using the new helper function) before looking up the term frequency.
   3. It should print the term frequency for that term in the document with the given ID.
   4. If the term doesn't exist in that document, it should just print 0.
7. Rerun uv run cli/keyword_search_cli.py build to store the term frequencies.

Run and submit the CLI tests.