Preferably a research-quality text. It doesn't need to be edited in English necessarily; and can be based on any dialect or standardization since the republics founding.
I'm not a native speaker, so I wouldn't trust my own interpretation of the phonemes for transcribing. And I don't consider the Turkish-Latin script to be sufficient for my purposes. (For instance, I would like the distinction between l and velarized ɫ for neighboring front/back vowels 1.) Thx!