---
license: apache-2.0
tags:
- awq
- AMD
- Ryzen AI
---

# Optmized LLM models for Ryzen AI.

The models are quantized via [Ryzen AI SW 1.2](https://github.com/amd/RyzenAI-SW/tree/main/example/transformers).

The folder `quantized_models` contains a set of LLMs quantized via different algorithms.    
The folder `onnx_model_nodes` contains a set of `mlp.up_proj` nodes for -npugpu device parallel config.