metadata

title: TranscriptTool - smolagent Transcription Tool
emoji: 💬
colorFrom: green
colorTo: green
sdk: gradio
sdk_version: 5.13.1
app_file: app.py
pinned: false
license: apache-2.0
short_description: smolagent tool to transcribe audio & video files

TranscriptTool: A smolagent Tool for Audio/Video Transcription

Overview

TranscriptTool is a smolagent tool designed to transcribe audio and video files into text. Implementing OpenAI's Whisper and ffmpeg, this tool enables agents to process multimedia inputs efficiently. It supports robust file handling, including format conversion to WAV, dynamic device selection (CPU or GPU), and easy use within smolagents via the Hugging Face API.

The repository contains three main components:

transcription_tool.py: The core smolagent tool for transcription.
app.py: A Gradio-powered web app to test and use the tool interactively.
example_smolagent.py: Toy demonstration of how the tool operates within a smolagent framework.

Installation

Clone this repository:

git clone https://huggingface.co/spaces/maguid28/TranscriptTool
cd TranscriptTool

Install dependencies:
```
pip install -r requirements.txt
```

Usage

Testing with Gradio (app.py)

To quickly test and use the transcription tool, run the provided Gradio app:

python app.py

This launches a local Gradio interface. Upload an audio or video file to transcribe it directly.

Running example SmolAgent (example_smolagent.py)

To see how TranscriptTool operates within a SmolAgent framework:

python example_smolagent.py

Access via Hugging Face API

The TranscriptTool is also available as a tool through the Hugging Face API.

How to Use the Tool via Hugging Face API (Currently not working)

Install SmolAgents

Ensure you have the SmolAgents library installed:
```
pip install smolagents
```
Load the Tool from the Hugging Face Hub

You can load the tool directly using the Hugging Face API.

from smolagents import Tool
transcript_tool = Tool.from_space(
 "maguid28/TranscriptTool",
 name="TranscriptTool",
 description="""
 A smolagent tool for transcribing audio and video files into text. This tool utilises Whisper for transcription 
 and ffmpeg for media conversion, enabling agents to process multimedia inputs into text. The tool supports robust 
 file handling, including format conversion to WAV and dynamic device selection for optimal performance.
 """
 )

License

This project is licensed under the Apache-2.0 License. See the LICENSE file for more details.

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any improvements or bug fixes.