Filename explanations:
model_xl: the model trained on the cross-language dataset
model_en: the model trained on the RAVDESS dataset
model_cn: the model trained on the Chinese Mandarin speech data in the ESD dataset
model_de: the model trained on the EMO-DB dataset
model_fr: the model trained on the CaFE_48k dataset
model_pt: the model trained on the VERBO dataset
Notes:
The models are trained using the Learning API xgb.train().
The hyperparameters of the models are not saved with the JSON files; get_params() method will not work on these model files.
The hyperparameters can be found in the params.txt.