No, I have not tested it before. However, you can check the following repositories on how to fine-tune the whole model on other languages (include custom text encoder):

Custom chinese text encoder:

https://github.com/IDEA-CCNL/Fengshenbang-LM/tree/main/fengshen/examples/stable_diffusion_chinese

Custom multilingual text encoder:

https://github.com/FlagAI-Open/FlagAI/tree/master/examples/AltDiffusion