ANCOR_Anaphora
The ANCOR project on anaphora and coreference in French Spoken Corpora can be downloaded here: https://www.ortolang.fr/market/corpora/ortolang-000903#!.
This project provides the first large corpus of conversational speech for French (488.000 words; 30,5 hours of recordings) which:
is annotated in anaphora and coreference;
is freely available (CC).