#podcast #transcription

app podcast2text

Cli tool for downloading episodes from rss and running transcription

3 releases

0.0.3 Nov 21, 2022
0.0.2 Nov 21, 2022
0.0.1 Nov 21, 2022

#3 in #transcription


352 lines

Jupiter Search

Crates.io MIT licensed APACHE 2 licensed Build Status

A showcase for indexing jupiter network podcasts using meilisearch. This repository is build in order to provide possible solution to following problems:


Warning! This is a work in progress version to showcase how indexing/transcription might work.


Project contains two main modules:

  • podcast2text a cli tool for downloading RSS feed and transcribing podcast episodes
  • search-load a cli tool for loading obtained transcriptions to instance of meilisearch


To build you would need following packages on your system:

  • cargo
  • pkg-config
  • openssl
  • ffmpeg

There is a nix flake configured to ship build dependencies just run direnv allow and run:

git submodule update --init --recursive
cargo build --release

To appease the gods of good taste please add following pre commit hook:

git config --local core.hooksPath .githooks


Run downloading podcasts

Process audio from RSS feed

  1. Download the whisper model
mkdir models
# this might be one of:
# "tiny.en" "tiny" "base.en" "base" "small.en" "small" "medium.en" "medium" "large"
curl --output models/model.bin https://ggml.ggerganov.com/ggml-model-whisper-$model.bin
  1. Run the inference on the RSS feed
# get information about the cli
docker run flakm/podcast2text --help

docker run \
    -v $PWD/models:/data/models \
    flakm/podcast2text \
    rss https://feed.jupiter.zone/allshows

Install meilisearch

docker pull getmeili/meilisearch:v0.29
docker run -it --rm \
    -p 7700:7700 \
    -v $(pwd)/meili_data:/meili_data \
    getmeili/meilisearch:v0.29 \
    meilisearch --env="development"

Run index creation and data loading

Running inference of some audio

  1. Download whisper model
mkdir models
curl --output models/ggml-$model.bin https://ggml.ggerganov.com/ggml-model-whisper-$model.bin
  1. Download the example audio from rss feed
curl https://feed.jupiter.zone/link/19057/15745245/55bb5263-04be-43a3-8b92-678072a9cfc8.mp3 -L -o action.mp3
  1. Install ffmpeg and put it on PATH variable.

  2. Run the inference example

cargo run --release --example=get_transcript -- models/ggml-medium.en.bin action_short.wav | tee output.txt


~659K SLoC