Are you planning to support other STT systems, like Baidu's DeepSpeech 2? This would make it possible to use their pretrained models, which are said to be very good for Chinese.
There are no such plans yet. For now you could use another STT system in combination with some script to produce .tlog files. This could be a viable option for an actual integration.
Are you planning to support other STT systems, like Baidu's DeepSpeech 2? This would make it possible to use their pretrained models, which are said to be very good for Chinese.