Improve logging in data valuation methods

Currently, there is almost no logging in the `src/pydvl/value` module. Potentially, this affects other modules as well.
Adding more logging can help with debugging and development. Therefore, I suggest doing it.

The issue mentions the `src/pydvl/value` module. Start by exploring this directory to understand the existing code structure and identify key functions where logging would be beneficial. Look for existing logging patterns in the project to maintain consistency. Determine appropriate log levels (INFO, DEBUG) and messages that would aid debugging. 'Done' means adding meaningful log statements to the main valuation methods without breaking existing tests.

Written by the indexing model from the issue text.