stapledon-podcast

A pipeline that generates imaginative text responses and converts them to audio narration with customizable voice options.

Features

• Generates creative text responses based on your prompt
• Converts the text to audio with a choice of voice models
• Supports both elderly British male and modern male voices
• Adjustable speech parameters like speed

Models

Under the hood it uses these models:

• pipeline-examples/stapledon: A text generation model that creates imaginative, detailed responses
• jaaari/kokoro-82m: A text-to-speech model with customizable speed and voice options
• playht/play-dialog: A text-to-speech model with an elderly British narrative voice option

How it works

You provide a text prompt, and the pipeline first generates a creative response using the Stapledon model. This text is then converted to audio using your chosen voice model - either Kokoro (with the “bm_lewis” voice) or PlayHT (with the “Gideon” elderly male British narrative voice). The pipeline outputs the audio file of your generated content, ready for listening.