sovits-ai-voice-for-lin-manuel-miranda's Introduction

Sovits-AI-voice-for-Lin-Manuel-Miranda

The singing voice conversion model trained with the famous musical playwright Lin Manuel Miranda's voice dataset.LMM (Lin Manuel Miranda) AI voice conversion, the project discloses the training data set of Lin's voice made by myself and the original audio of its source, a set of models trained for 3000+ epochs, and some songs from my own inference.

林漫威（Lin Manuel Miranda）AI语音转换，项目里公开了我自己制作的林聚聚语音的训练数据集以及其出处的原音频、一组训练了3000+ epoch的模型、一些自己inference出来的歌曲。

使用数据集或模型请注明出处。

folder "vocal_resource"(文件夹“vocal_resource”)

clip_output（raw_dataset）文件夹是完全处理完成的数据集，可以直接丢到模型训练的dataset_raw下面训练用。（可以直接拿去训练，使用请注明出处） dry_output、edited_output是处理中间的文件 original是音频数据源文件 Inference_original 是用于推理的音频歌曲。最终的推理结果在Inference_output

The clip_output (raw_dataset) folder is a fully processed dataset, which can be directly dropped into the dataset_raw for model training for training. (It can be used directly for training, please indicate the source when using it) dry_output, edited_output are files in the middle of processing original is the audio data source file Inference_original is the audio song used for inference.

folder "trained_models" (文件夹“trained_models”)

训练结果模型（可以直接拿去推理，使用请注明出处）

Training result model (can be directly used for reasoning, please indicate the source)

Recommend Projects

qianyi-duan2023 / sovits-ai-voice-for-lin-manuel-miranda Goto Github PK