openvoice plugin

.ipynb_checkpoints/README-checkpoint.md

@@ -23,7 +23,7 @@ For more details, please check our interspeech paper: [DreamVoice](https://arxiv
 To listen to demos and download dataset, please check dreamvoice's homepage: [Homepage](https://haidog-yaqub.github.io/dreamvoice_demo/)
 
 
-# Model Usage
+# How to Use
 
 To load the models, you need to install packages:
 
@@ -91,6 +91,24 @@ gen_audio2, sr = dreamvoice.simplevc('examples/test2.wav', use_spk_cache=True)
 dreamvoice.save_audio('gen2.wav', gen_audio2, sr)
 ```
 
+# Training Guide
+
+1. download VCTK and LibriTTS-R
+2. extract speaker embeddings and cache in local path:
+```
+# example code
+dreamvoice/train_utils/prepare/prepare_se.py
+```
+3. modify trainning config and train your dreamvoice plugin:
+```
+# example code
+cd dreamvoice/train_utils/src
+accelerate launch train.py
+```
+
+
+# Extra Features
+
 - End-to-end DreamVoice VC Model
 
 ```python

.ipynb_checkpoints/update-checkpoint.py

@@ -1,26 +0,0 @@
-from huggingface_hub import HfApi, HfFolder, Repository
-
-# Authenticate with Hugging Face Hub
-api = HfApi()
-token = HfFolder.get_token()
-if token is None:
-    raise ValueError("Hugging Face token is not set. Please authenticate first.")
-
-# Create or clone the private repository using the token
-repo = Repository(
-    local_dir="DreamVoice",
-    clone_from="myshell-ai/DreamVoice",
-    use_auth_token=token
-)
-
-repo.lfs_track(".pt")
-repo.lfs_track(".png")
-
-# Add all changes to git
-repo.git_add()
-
-# Commit the changes with a message
-repo.git_commit("Initial commit")
-
-# Push the changes to the remote repository
-repo.git_push()

README.md

@@ -23,7 +23,7 @@ For more details, please check our interspeech paper: [DreamVoice](https://arxiv
 To listen to demos and download dataset, please check dreamvoice's homepage: [Homepage](https://haidog-yaqub.github.io/dreamvoice_demo/)
 
 
-# Model Usage
+# How to Use
 
 To load the models, you need to install packages:
 
@@ -91,6 +91,24 @@ gen_audio2, sr = dreamvoice.simplevc('examples/test2.wav', use_spk_cache=True)
 dreamvoice.save_audio('gen2.wav', gen_audio2, sr)
 ```
 
+# Training Guide
+
+1. download VCTK and LibriTTS-R
+2. extract speaker embeddings and cache in local path:
+```
+# example code
+dreamvoice/train_utils/prepare/prepare_se.py
+```
+3. modify trainning config and train your dreamvoice plugin:
+```
+# example code
+cd dreamvoice/train_utils/src
+accelerate launch train.py
+```
+
+
+# Extra Features
+
 - End-to-end DreamVoice VC Model
 
 ```python

dreamvoice/train_utils/prepare/prepare_se.py

@@ -92,10 +92,10 @@ if __name__ == '__main__':
     model = ToneColorConverter(f'{ckpt_converter}/config.json', device=device)
     model.load_ckpt(f'{ckpt_converter}/checkpoint.pth')
 
-    input_folder = '/home/jerry/Projects/Dataset/VCTK/24k/VCTK-Corpus/'
+    input_folder = 'your_path/VCTK-Corpus/'
     output_folder = 'spk/VCTK-Corpus/'
     process_audio_folder(input_folder, output_folder, model, device)
 
-    input_folder = '/home/jerry/Projects/Dataset/Speech/vctk_libritts/LibriTTS-R/train-clean-360'
+    input_folder = 'your_path/LibriTTS-R/train-clean-360'
     output_folder = 'spk/LibriTTS-R/train-clean-360/'
     process_audio_folder(input_folder, output_folder, model, device)