whisper-th-medium-combined / preprocessor_config.json
tensorops's picture
update model to latest
eebf842
{
"chunk_length": 30,
"feature_extractor_type": "WhisperFeatureExtractor",
"feature_size": 80,
"hop_length": 160,
"n_fft": 400,
"n_samples": 480000,
"nb_max_frames": 3000,
"padding_side": "right",
"padding_value": 0.0,
"processor_class": "WhisperProcessor",
"return_attention_mask": false,
"sampling_rate": 16000
}