deep-voice-cloning / scripts /training_config.json
konverner's picture
Initial commit
899cf32
raw
history blame
249 Bytes
{
"audio_path": "/content/deep-voice-cloning/scripts/input/hank_hill.mp3",
"output_dir": "/content/deep-voice-cloning/models",
"lang": "en",
"batch_size": 2,
"learning_rate": 1e-4,
"max_steps": 1500,
"warmup_steps": 250
}