Modal Configuration

ModalApp

This module defines the modal integration and the code to be run in Modal's container.

Attributes:

Name	Type	Description
LOGGER	Logger	Module's logging object.
MODAL_IMAGE	str	Docker Repository URL for pre-built Image acquired from environment variable. Defaults to docker.io/svdc1/mirumoji-modal-gpu:latest
GPU	str	Which Modal GPU to use from environment variable. Defaults to A10G

transcribe_srt_job(media_fp, fwhisper_kwargs={}, transcribe_kwargs={}, fix_with_chat_gpt=True)

Runs Whisper transcription on media_fp, fixes with GPT, and returns SRT string.

Parameters:

Name	Type	Description	Default
media_fp	Union[str, Path]	Path to the video for transcription.	required
fwhisper_kwargs	dict	Additional arguments for FWhisperWrapper	{}
transcribe_kwargs	dict	Additional arguments for FWhisperWrapper.transcribe	{}
fix_with_chat_gpt	bool	If True request ChatGPT to fix transcription. Defaults to True	True

Returns:

Name	Type	Description
str	Union[str, None]	Transcription in form of SRT string.

transcribe_to_string_job(audio_fp, fwhisper_kwargs={}, transcribe_kwargs={})

Transcribe audio to string using Faster Whisper.

Parameters:

Name	Type	Description	Default
audio_fp	Union[str, Path]	Path to the audio for transcription.	required
fwhisper_kwargs	dict	Additional arguments for FWhisperWrapper	{}
transcribe_kwargs	dict	Additional arguments for FWhisperWrapper.transcribe	{}

Returns:

Name	Type	Description
str	Union[str, None]	Transcription in form of string.

video_conversion_job(video_fp, to_mp4_kwargs={})

Converts video_fp to MP4 using NVENC and returns the video content as bytes.

Parameters:

Name	Type	Description	Default
video_fp	Union[str, Path]	Path to the video for conversion.	required
to_mp4_kwargs	dict	Additional arguments for AudioTools.to_mp4.	{}

Yields:

Name	Type	Description
bytes	bytes	The converted video chunks.