feat: Verbose evals (#1558)

* Add `_verbose` flag to `BaseEvalModel`

* Start adding basic verbose-mode logging

* Add verbose mode to retries

* Only print when verbose flag is set

* Refactor verbose mode
- less noisy
- do not print OpenAI credentials information

* Continue refining verbose mode output

* Prefer absolute imports

* Fix type hint

* Try to clean up abstractions

* Add `printif` utility

* Prefer absolute imports

* Add `verbose` test for `llm_eval_binary`

* Test retrying with verbose mode

* Test that the "verbose" state does not get persisted

* Implement verbose flag as a context manager

* Add docstrings

* Improve verbosity statefulness test

* Lint imports

* Add blankline

* Shorten docstrings

* Enforce formatter settings

* Appease mypy

* Add verbose flag test for `generate`

* Use better dummy variable name

* Update src/phoenix/experimental/evals/models/base.py

Co-authored-by: Roger Yang <80478925+RogerHYang@users.noreply.github.com>

* Add more details to docstrings

* Update bedrock model

* Restore missing import

---------

Co-authored-by: Roger Yang <80478925+RogerHYang@users.noreply.github.com>