- original algorithm repositoryBert-VITS2(v2.0.2)
- all to trtfp16 to get 9x faster.
- after you get your own trained model. (train_ms.py)
0.1 prepare your own 'config.yml' as the 'default_config.yml' - sh bert_to_onnx.sh
- python infer_torch_export_onnx.py (first comment line: g_model_name = None)
- see inputs.py to get trt.engine
- python infer_backend_mix.py (test trt infer)
- change the training forward to merge the all 6 submodel to one model to get VITS2.engine.
- improve the performance of VITS2 on emotion.
- when some submodel severe performance loss, you can replece the trt to onnx.