Llama-3.2-Taiwan-3B-Instruct-Arbor-4bit

This repository contains a 4-bit Apple MLX quantized derivative of:
  lianghsun/Llama-3.2-Taiwan-3B-Instruct
  https://huggingface.co/lianghsun/Llama-3.2-Taiwan-3B-Instruct

The upstream Taiwan Instruct model is based on:
  meta-llama/Llama-3.2-3B-Instruct
  https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct

The non-instruct GGUF repository originally referenced was:
  QuantFactory/Llama-3.2-Taiwan-3B-GGUF
  https://huggingface.co/QuantFactory/Llama-3.2-Taiwan-3B-GGUF

The conversion in this repository uses the upstream safetensors model from lianghsun/Llama-3.2-Taiwan-3B-Instruct, not a GGUF-to-MLX round-trip.

Upstream license and terms
--------------------------
The original Meta Llama 3.2 model and derivatives are provided under the Llama 3.2 Community License. Use, copying, modification, redistribution, deployment, publication, hosting, and any other use of this derivative model must comply with the Llama 3.2 Community License and any additional terms, notices, access requirements, usage instructions, export-control, sanctions, and other applicable laws or restrictions shown on the upstream model pages.

This NOTICE is intended to preserve attribution and highlight upstream obligations. It does not replace the Llama 3.2 Community License or any terms published by the upstream model authors. If the upstream model pages, licenses, notices, or usage terms are updated, users and redistributors are responsible for reviewing and complying with those terms.

Derivative conversion notice
----------------------------
This repository redistributes a derivative model and includes quantized weights and model files converted for Apple MLX / MLX-LM. Quantization may change model behavior, output quality, numerical properties, robustness, and safety characteristics compared with the original upstream model. This repository does not claim improved accuracy, safety, suitability, or performance over the upstream model.

No affiliation or endorsement
-----------------------------
This repository is independently prepared and published by the repository owner. It is not affiliated with, sponsored by, approved by, or endorsed by Meta, Llama, lianghsun, QuantFactory, or their affiliates unless they explicitly state otherwise.

Trademarks
----------
Meta, Llama, and other related names, logos, and marks may be trademarks or registered trademarks of their respective owners. No trademark rights are granted by this repository except for reasonable descriptive use required to identify the upstream source/base models and preserve attribution. All other trademarks are the property of their respective owners.

Warranty and liability disclaimer
---------------------------------
This derivative model and associated files are provided as-is and without warranties or conditions of any kind, express or implied, including without limitation warranties of merchantability, fitness for a particular purpose, title, non-infringement, accuracy, availability, or error-free operation.

To the maximum extent permitted by applicable law, the repository owner is not liable for any direct, indirect, incidental, special, consequential, exemplary, punitive, or other damages arising from or related to use, redistribution, deployment, or outputs of this derivative model.

Users are solely responsible for evaluating model behavior, safety, compliance, and suitability for their intended use case and jurisdiction. Nothing in this NOTICE is legal advice.