| import gradio as gr |
| import assemblyai as aai |
| import re |
| import os |
| import tempfile |
|
|
| |
| aai.settings.api_key = os.getenv('ASSEMBLYAI_API_KEY') |
|
|
| def create_assembly_transcript(audio_file): |
| transcriber = aai.Transcriber() |
| transcript = transcriber.transcribe( |
| audio_file, config=aai.TranscriptionConfig(speaker_labels=True) |
| ) |
| return transcript |
|
|
| def transcript_to_string(transcript): |
| output = "" |
| for utterance in transcript.utterances: |
| name = f"SPEAKER {utterance.speaker}" |
| start_time = format_time(utterance.start) |
| output += f"{name} {start_time}\n{utterance.text}\n\n" |
| return output |
|
|
| def format_time(milliseconds): |
| seconds = milliseconds // 1000 |
| hours, seconds = divmod(seconds, 3600) |
| minutes, seconds = divmod(seconds, 60) |
| return f"{int(hours):02}:{int(minutes):02}:{int(seconds):02}" |
|
|
| def format_transcript_markdown(transcript_string): |
| speaker_label_pattern = r"^(.+?)(?=\s\d{2}:\d{2}:\d{2})" |
| timestamp_pattern = r"(\d{2}:\d{2}:\d{2})" |
| formatted_transcript = re.sub( |
| speaker_label_pattern, r"**\1**", transcript_string, flags=re.MULTILINE |
| ) |
| formatted_transcript = re.sub( |
| timestamp_pattern, r"_\1_", formatted_transcript, flags=re.MULTILINE |
| ) |
| return formatted_transcript |
|
|
| def transcribe_audio(audio_file): |
| if audio_file is None: |
| return "Please upload an audio file.", None |
|
|
| try: |
| transcript = create_assembly_transcript(audio_file) |
| |
| if transcript.error: |
| return f"An error occurred: {transcript.error}", None |
| |
| transcript_string = transcript_to_string(transcript) |
| md_transcript = format_transcript_markdown(transcript_string) |
| |
| |
| with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.md') as temp_file: |
| temp_file.write(md_transcript) |
| temp_file_path = temp_file.name |
| |
| return transcript_string, temp_file_path |
| except Exception as e: |
| return f"An error occurred: {str(e)}", None |
|
|
| def launch_app(): |
| iface = gr.Interface( |
| fn=transcribe_audio, |
| inputs=gr.Audio(type="filepath", label="Upload Audio File"), |
| outputs=[ |
| gr.Textbox(label="Transcript Preview", lines=10), |
| gr.File(label="Download Formatted Transcript") |
| ], |
| title="Audio Transcription App", |
| description="Upload an audio file to get a transcription with speaker labels. The preview shows plain text, while the download includes markdown formatting." |
| ) |
| iface.launch() |
|
|
| if __name__ == "__main__": |
| launch_app() |