Update README.md

README.md

@@ -6,18 +6,14 @@ tags:
 - fairseq
 - audio
 - text-to-speech
-language: en
-datasets:
-- mtedx
-- covost2
-- europarl_st
-- voxpopuli
+language: hk
 ---
 ## unit_hifigan_HK_layer12.km2500_frame_TAT-TTS
 
-Speech-to-speech translation model from fairseq S2UT ([paper](https://arxiv.org/abs/2204.02967)/[code](https://github.com/facebookresearch/fairseq/blob/main/examples/speech_to_speech/docs/enhanced_direct_s2st_discrete_units.md)):
-- Spanish-English
-- Trained on mTEDx, CoVoST 2, Europarl-ST and VoxPopuli
+Speech-to-speech translation model from fairseq:
+- English-Hokkien
+- Trained with supervised data in TED domain, and weakly supervised data in TED and Audiobook domain. See [here]( https://research.facebook.com/publications/hokkien-direct-speech-to-speech-translation) 
+for training details
 
 ## Usage