Lexless | Automatically Remove Interviewer Segments From Podcasts with Python AI
Turn long podcasts into guest-only audio. Use Python and speaker diarization to remove interviewer clips and keep just the guest content.
---
What's in this video
00:00 Intro: what Lexless does
01:15 Prerequisites and setup
05:00 Configuring the Hugging Face token
07:23 Processing local audio files
10:20 How it works
11:00 Wrap-up
---
Example result: https://drive.google.com/file/d/1gqbX...
About Lexless
Lexless flags and removes interviewer segments, keeping only guest audio.
AI speaker detection with pyannote.audio
Supports local file processing
Smooth crossfades between cuts
Normalization options
Works on Linux, macOS, and Windows
---
Quick start
*1. Install:*
```bash
git clone https://github.com/mikael-codes/lexle...
cd lexless
pip install -e .
```
*2. Hugging Face setup:*
Create a token: https://huggingface.co/settings/tokens
Accept model terms: https://huggingface.co/pyannote/speak...
Accept model terms: https://huggingface.co/pyannote/segme...
*3. Run:*
```bash
lexless PATH_TO_FILE --file --hf-token YOUR_TOKEN
```
---
Topics
speaker diarization with pyannote.audio
automatic segment removal
audio editing in Python (librosa)
API setup (Hugging Face)
Python CLI tools (Click)
audio processing and fades
---
Links
GitHub: https://github.com/mikael-codes/lexless
Docs: HowItWorks.md and README.md in the repo
Models: pyannote.audio
---
Subscribe for more tutorials and fun
---
Tags
#Python #AI #HuggingFace #MachineLearning #AudioAI #Podcast #Automation
---
Leave a comment if you want to see more automation topics or have questions about Lexless.