77 lines
2.2 KiB
Markdown
77 lines
2.2 KiB
Markdown
# Whisper API Server (Go)
|
|
|
|
## ⚠️ This project is a work in progress (WIP).
|
|
|
|
This API server enables audio transcription using the OpenAI Whisper models.
|
|
|
|
# Setup
|
|
|
|
- Download `.exe` from [Releases](https://github.com/xzeldon/whisper-api-server/releases/latest)
|
|
- Just run it!
|
|
|
|
# Build from source
|
|
|
|
Before build make sure that **CGO_ENABLED** env is set to **1**
|
|
|
|
```
|
|
$env:CGO_ENABLED = "1"
|
|
```
|
|
|
|
you can check this with this command
|
|
|
|
```
|
|
go env
|
|
```
|
|
|
|
Also you have to have installed gcc x64 i.e. by MYSYS
|
|
|
|
Download the sources and use `go build`.
|
|
For example, you can build using the following command:
|
|
|
|
```bash
|
|
go build -ldflags "-s -w" -o server.exe main.go
|
|
```
|
|
|
|
# Usage example
|
|
|
|
Make a request to the server using the following command:
|
|
|
|
```sh
|
|
curl http://localhost:3000/v1/audio/transcriptions \
|
|
-H "Content-Type: multipart/form-data" \
|
|
-F file="@/path/to/file/audio.mp3" \
|
|
```
|
|
|
|
Receive a response in JSON format:
|
|
|
|
```json
|
|
{
|
|
"text": "Imagine the wildest idea that you've ever had, and you're curious about how it might scale to something that's a 100, a 1,000 times bigger. This is a place where you can get to do that."
|
|
}
|
|
```
|
|
|
|
# Usage with [Obsidian](https://obsidian.md/)
|
|
|
|
1. Install [Obsidian voice recognotion plugin](https://github.com/nikdanilov/whisper-obsidian-plugin)
|
|
2. Open the plugin's settings.
|
|
3. Set the following values:
|
|
- API KEY: `sk-1`
|
|
- API URL: `http://localhost:3000/v1/audio/transcriptions`
|
|
- Model: `whisper-1`
|
|
|
|
# Roadmap
|
|
|
|
- [x] Implement automatic model downloading from [huggingface](https://huggingface.co/ggerganov/whisper.cpp/tree/main)
|
|
- [x] Implement automatic `Whisper.dll` downloading from [Guthub releases](https://github.com/Const-me/Whisper/releases)
|
|
- [x] Provide prebuilt binaries for Windows
|
|
- [ ] Include instructions for running on Linux with Wine (likely possible).
|
|
- [ ] Use flags to override the model path
|
|
- [ ] Use flags to override the model type (when downloading the model)
|
|
- [ ] Use flags to override the port
|
|
|
|
# Credits
|
|
|
|
- [Const-me/Whisper](https://github.com/Const-me/Whisper) project
|
|
- [goConstmeWhisper](https://github.com/jaybinks/goConstmeWhisper) for the remarkable Go bindings for [Const-me/Whisper](https://github.com/Const-me/Whisper)
|
|
- [Georgi Gerganov](https://github.com/ggerganov) for GGML models
|