OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Audio & Voice AI

Spear Tts Pytorch

Spear Tts Pytorch is a Python/PyTorch implementation of a multi-speaker text-to-speech attention network. It provides a TextToSemantic module intended to supply conditioning for SoundStorm, alongside code for training, decoding and generating audio-text dataset files.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Spear Tts Pytorch — OpenViST project profile in Audio & Voice AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want to inspect or adapt a PyTorch implementation for text-to-semantic speech modeling, or follow its example dataset-generation code.

Who it is for

  • Researchers experimenting with multi-speaker text-to-speech and text-to-semantic modeling.
  • PyTorch developers exploring speech-model training, decoding or pseudo-labelled dataset workflows.

Documented uses

  • Build a TextToSemantic model using a HuBERT-with-k-means component.
  • Generate audio-text dataset files from a supplied dataset through the documented generator.
  • Explore a text-to-semantic component intended to condition a SoundStorm pipeline.
Before you choose

Requirements & limits

The repository is public and MIT-licensed. The README documents installation through pip.

  • The example depends on local HuBERT checkpoint and k-means files, plus audiolm-pytorch's HubertWithKmeans.
  • The README's dataset-generation demonstration uses MockDataset rather than a real speech corpus.
  • The repository lists cached key/value support and additional audio-text workflow work as unfinished.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up the documented example dataset workflow

Install the library and follow the README's example for generating and loading audio-text dataset files.

Before you start

  • Python environment with the documented packages available: PyTorch, audiolm-pytorch, and spear-tts-pytorch.
  • HuBERT checkpoint and k-means files at the paths used by the example.
  • A dataset; the example uses MockDataset.

Steps

  1. Install the package with the command below.

    Documented command
    pip install spear-tts-pytorch

  2. Initialize HubertWithKmeans using the checkpoint and k-means paths.

  3. Create TextToSemantic and pass it with a dataset to SemanticToTextDatasetGenerator, specifying an output folder.

  4. Call the generator with a maximum length, then load the output folder with GeneratedAudioTextDataset.

Expected outcome

The example produces files in the specified folder and creates a GeneratedAudioTextDataset from that folder.

Guide limitations

  • The README example uses MockDataset(10); it does not document a real-corpus preparation workflow.
  • The shown assertion is an example expectation, not an independently verified execution result.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
lucidrains/spear-tts-pytorch · Official homepage title · retained in its source language
How we review sourcesSuggest a correction