harmoni_stt.deepspeech_client

Module Contents

Attributes

model_dir

class harmoni_stt.deepspeech_client.DeepSpeechClient(model_file_path, scorer_path, alpha, beta, beam_width, t_wait=0.5)
process_audio(self, in_data)

Passes a chunk of audio for the DeepSpeech stream to process

start_stream(self)

Starts DeepSpeech streaming inference state if it is not already open.

finish_stream(self)

Ends DeepSpeech streaming inference state if still open. Returns the transcribed text.

restart_stream(self)
property is_final(self)
property is_streaming(self)
harmoni_stt.deepspeech_client.model_dir