# Training

Step 1: Dataset Preparation
  1. Upload your audio in .wav format using the Dataset Maker (if it is a single audio) or setup it manually (various audios) to applio/assets/datasets creating inside a folder for the program to read it.

Step 2: Dataset Processing
  1. Once the model is named and the dataset selected press "Prepocess Dataset" and wait for the message in the CMD.

Step 3: Feature Extraction
  1. Select an F0 method that suits your needs.
  2. (optional) modify Hop lenght, lower value, higher smoothness of pitch change but slower training and vice versa.
  3. (optional) select the Embedder model (hubert or contentvec)

Step 4: Model Training

Configure the training parameters according to your needs.

  • Save Every Epoch: Set this value between 10 and 50 to determine how often the model's state is saved during training.

  • Total Epochs: The number of epochs needed varies based on your dataset. Monitor progress using TensorBoard; typically, models perform well around 200-400 epochs.

  • Batch Size: Adjust based on your GPU's VRAM. For 8 GB VRAM, use a batch size between 6 and 8. Consider CUDA cores when experimenting with higher batch sizes.

# Other Options

  • Pitch Guidance: Gives variation of pitch.
  • Pretrained: Uses the RVC pretrained.
  • Save Only Latest: Save a single D/G file with information.
  • Save Every Weights: Save the weights of the model when a cycle of 'Save Every Epoch' is completed.
  • Custom Pretrained: Uses the Custom Pretrained that are loaded.
  • GPU Settings: Allows to choose GPUs (only for users who have more than one GPU).
  • Overtraining Detector: Mark it only if you will train for more than 200 epochs.
  • Overtraining Threshold: Set the maximum number of epochs you want your model to stop training if no improvement is detected.

Once configured, press 'Start training' to start the process, everything is registered in the CMD.

Final Step: Model Saving and Index File Generation
  1. Once training is completed, generate the index file by clicking the "Train Feature Index" button.
  2. Your trained model is located in the logs/model folder, and the .pth files are in the logs/zips folder.

Optional Step: Resume training
  1. Open Applio if you have closed it.
  2. Then, in the Applio interface, input your model name, use the same sample rate, and proceed to the last part of the train tab. Set the same batch size, pretrained (if you used) and increase the number of epochs you want to train.
  3. Once configured, press 'Start training' to start the process, everything is registered in the CMD.