Scribe

KO | EN

gitlite — search

#python #react #llm #go #android #rust #deep-learning #windows #awesome #awesome-list #pytorch #ios #openai #open-source

(★ 10)

A compact (offline) GUI media transcriber that enables you to search for local content based on its spoken words.

Scribe 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

![Total Downloads](https://img.shields.io/github/downloads/o7q/Scribe/total?logo=github&label=Total%20Downloads&color=%232fd653)
![Code Size](https://img.shields.io/github/languages/code-size/o7q/Scribe?logo=github&label=Code%20Size&color=%23b65cff)

# [>> Download Latest](https://github.com/o7q/Scribe/releases/latest/download/Scribe.exe)
### Welcome! Scribe is a simple, compact, offline, GUI based media transcriber that enables you to search for local media content based on its spoken words.

---

---

# Overview
Scribe uses OpenAI's whisper model to process audio files.\
FFmpeg is used beforehand to handle a wide variety of media formats.

### **Powered by**
stable-ts: https://github.com/jianfch/stable-ts \
FFmpeg: https://ffmpeg.org

---

# Installation

1. Ensure you have Python 3.9 or 3.10 installed (not tested on newer versions) \
You can check your version by typing this command:
```powershell
python --version
```
2. Download the bootstrapper from https://github.com/o7q/Scribe/releases
3. Select your preferred device type (CPU or CUDA) \
You can check your CUDA version by typing this command:
```
nvidia-smi
```
4. Click install and wait for it to complete

# Side Note
By default, Scribe configures FFmpeg in the `PATH` environment variable. If you are having issues with Scribe not working correctly, check if FFmpeg was added to the `PATH` environment variable correctly.

---

## **Interface**

<details>
<summary>Selection</summary>

- **Add Folder** Adds a watch folder to scan from
- **Remove Folder** Removes the selected watch folder
- **Filetypes** Filetypes formats to scan for (separate with a comma)
- **Scan Subdirectories** Enable the scanning of subdirectories

</details>

<details>
<summary>Processing</summary>

- **Start With Windows** Makes the program start with your computer
- **Auto Update** Enable the automatic processing of watch folders
- **Use GPU (CUDA)** Enable the usage of supported NVIDIA CUDA GPUs
- **Model** The model for the whisper speech-to-text algorithm (default is: **small**)
- **Language** The language for the whisper speech-to-text algorithm (default is: **English**)
- **Media Buffer** The amount of media files Scribe will buffer before finalizing, the higher the number the faster the processing, but **I wouldn't recommend going above 30** because of limitations relating the the Windows Command Prompt. (this because the max command length is 8191)
- **Debug** Enable debug/verbose mode, this will display all intermediate steps during processing
- **Update** Manually process media
- **Abort** Cancels the process
- **Clean** Cleans storage files that no longer have the corresponding media file tied to it, it also cleans temporary files

</details>

<details>
<summary>Searching</summary>

- **Phrase Search** The phrase to search for
- **File Search** The specific file to search for (leave blank to show all files that contain the phrase)
- **Search** Searches with the specified settings
- **Auto Update** Automatically updates the search without having to click the **Search Button**
- **Case Sensitive** Enable case sensitivity
- **Open File** Opens the selected file
- **Browse File** Browses to the selected file in Windows Explorer

</details>

<details>
<summary>Viewing</summary>

- **View Panel** This will display all of the written text as well as their timestamps of the selection media, it will highlight every occurrence of the searched phrase

</details>

---

**Scribe** \
Written in C# using .NET 8.0

Scribe

Scribe 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Scribe

Scribe 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README