Engineer - FR (TF2)

Train a voice

No model is associated with this dataset yet. The format is LJSpeech (metadata.csv and a wavs/ folder). Training takes 1 to 6 hours in the Kaggle notebook and produces model.onnx and model.onnx.json.