https://github.com/fly-apps/whisper-example
Fly GPU Machines transcribing an mp3 file with Whisper
https://github.com/fly-apps/whisper-example
Last synced: over 1 year ago
JSON representation
Fly GPU Machines transcribing an mp3 file with Whisper
- Host: GitHub
- URL: https://github.com/fly-apps/whisper-example
- Owner: fly-apps
- Created: 2023-11-10T18:48:12.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-11-10T19:41:59.000Z (over 2 years ago)
- Last Synced: 2025-01-17T18:55:08.829Z (over 1 year ago)
- Language: Dockerfile
- Size: 2.75 MB
- Stars: 12
- Watchers: 9
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
# Run Whisper on Fly.io GPU Machines
I used [this article](https://nlpcloud.com/how-to-install-and-deploy-whisper-the-best-open-source-alternative-to-google-speech-to-text.html) for reference.
The easiest way to test this out (if your account [has GPU access](https://community.fly.io/t/fly-gpus-are-here/16110)):
```bash
# 1. Clone the repo
git clone git@github.com:fly-apps/whisper-example.git
cd whisper-example
# 2. Create an app
fly apps create
# 3. Edit fly.toml and add your app name there
# 4. Deploy a GPU Machine
# --vm-gpu-kind a100-pcie-40gb -> Create a a100 GPU machine, 40b memory
# --volume-initial-size 20 -> create a new volume on create
# --wait-timeout 600 -> Wait longer for the Machine to start
# --ha=false -> Only create one Machine instance
# --restart no -> Just run once (alternatively, use --rm)
fly deploy --vm-gpu-kind a100-pcie-40gb \
--volume-initial-size 20 \
--wait-timeout 600 --ha=false \
--restart no
```
File `audio.mp3` provided by https://www.flatworldsolutions.com/transcription/samples.php and converted to mp3 via:
```bash
ffmpeg -i Monologue.ogg -b:a 192k audio.mp3
```