diarize_speech
Transcribe audio files to text with speaker identification, saving results to a specified directory. Supports custom models, maximum speaker limits, and word boosting for accurate recognition. Uses Whissle API for processing.
Instructions
Convert speech to text with speaker diarization and save the output text file to a given directory. Directory is optional, if not provided, the output file will be saved to $HOME/Desktop.
Input Schema
Name | Required | Description | Default |
---|---|---|---|
audio_file_path | Yes | ||
boosted_lm_score | No | ||
boosted_lm_words | No | ||
max_speakers | No | ||
model_name | No | en-NER |