Aura 4B GGUF

🚀 Llamacpp imatrix Quantizations of Aura-4B

This project offers Llama.cpp imatrix quantizations of the Aura-4B model. It provides various quantization types to balance between model quality and resource usage, enabling users to run the model efficiently on different hardware.

🚀 Quick Start

• Quantization: We use llama.cpp release b4381 for quantization.
• Original Model: You can access the original model at Aura-4B.
• Running the Model: You can run the quantized models in LM Studio.

✨ Features

• Multiple Quantization Types: Offers a wide range of quantization types, such as f16, Q8_0, Q6_K_L, etc., to meet different performance and quality requirements.
• Optimized Embedding and Output Weights: Some quantizations use Q8_0 for embedding and output weights to improve quality.
• Online Repacking: Supports online repacking of weights for better performance on ARM and AVX machines.

📦 Installation

Installing huggingface-cli

First, ensure you have huggingface-cli installed:

pip install -U "huggingface_hub[cli]"

Downloading a Specific File

To download a specific file, use the following command:

huggingface-cli download bartowski/Aura-4B-GGUF --include "Aura-4B-Q4_K_M.gguf" --local-dir ./

Downloading Split Files

If the model is split into multiple files (models larger than 50GB), you can download them all to a local folder:

huggingface-cli download bartowski/Aura-4B-GGUF --include "Aura-4B-Q8_0/*" --local-dir ./

💻 Usage Examples

Prompt Format

<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

📚 Documentation

Downloadable Files

Embed/Output Weights

Some of these quants (Q3_K_XL, Q4_K_L etc) are the standard quantization method with the embeddings and output weights quantized to Q8_0 instead of what they would normally default to.

ARM/AVX Information

Previously, you would download Q4_0_4_4/4_8/8_8, and these would have their weights interleaved in memory in order to improve performance on ARM and AVX machines by loading up more data in one pass.

Now, however, there is something called "online repacking" for weights. details in this PR. If you use Q4_0 and your hardware would benefit from repacking weights, it will do it automatically on the fly.

As of llama.cpp build b4282 you will not be able to run the Q4_0_X_X files and will instead need to use Q4_0.

Additionally, if you want to get slightly better quality for , you can use IQ4_NL thanks to this PR which will also repack the weights for ARM, though only the 4_4 for now. The loading time may be slower but it will result in an overall speed incrase.

Which File to Choose

🔧 Technical Details

Benchmarks on an AVX2 system (EPYC7702)

📄 License

This project is licensed under the Apache-2.0 License.

Credits

Thank you kalomaze and Dampf for assistance in creating the imatrix calibration dataset.

Thank you ZeroWw for the inspiration to experiment with embed/output.

Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski