README.md

metadata

license: apache-2.0
pipeline_tag: text-generation
tags:
  - executorch

Introduction

This repository hosts the Bielik-v3.0 models for the React Native ExecuTorch library. It includes unquantized and quantized versions of the Bielik (instruct) model in .pte format, ready for use in the ExecuTorch runtime. Bielik is a model developed and trained on Polish text, created by Speakleash. It excels at understanding and generating Polish language.

If you'd like to run these models in your own ExecuTorch runtime, refer to the official documentation for setup instructions.

Compatibility

If you intend to use this model outside of React Native ExecuTorch, make sure your runtime is compatible with the ExecuTorch version used to export the .pte files. For more details, see the compatibility note in the ExecuTorch GitHub repository. If you work with React Native ExecuTorch, the constants from the library will guarantee compatibility with runtime used behind the scenes.

These models were exported using v1.0.0 version and no forward compatibility is guaranteed. Older versions of the runtime may not work with these files.

Repository Structure

The repository contains a single main directory:

• bielik-v3.0-1.5B

Each directory contains different versions of the model, including quantized, and the original models.

• The .pte file should be passed to the modelSource parameter.
• The tokenizer for the models is available within the repo root, under tokenizer.json and tokenizer_config.json