Black Ink Guild Pernicious Prophecy 70B GGUF

🚀 Llamacpp imatrix Quantizations of Pernicious_Prophecy_70B by Black-Ink-Guild

This project provides Llama.cpp imatrix quantizations of the Pernicious_Prophecy_70B model, enabling efficient deployment and utilization on various hardware platforms.

🚀 Quick Start

Using llama.cpp release b4585 for quantization.

• Original model: https://huggingface.co/Black-Ink-Guild/Pernicious_Prophecy_70B
• All quants made using imatrix option with dataset from here
• Run them in LM Studio
• Run them directly with llama.cpp, or any other llama.cpp based project

✨ Features

• Multiple Quantization Types: Offers a wide range of quantization types, such as Q8_0, Q6_K, Q5_K_M, etc., to meet different performance and quality requirements.
• Optimized for Different Hardware: Some quantizations are optimized for ARM and AVX machines, and support online repacking to improve performance.
• Easy to Use: Can be easily run in LM Studio or directly with llama.cpp.

📦 Installation

Downloading using huggingface-cli

pip install -U "huggingface_hub[cli]"

huggingface-cli download bartowski/Black-Ink-Guild_Pernicious_Prophecy_70B-GGUF --include "Black-Ink-Guild_Pernicious_Prophecy_70B-Q4_K_M.gguf" --local-dir ./

huggingface-cli download bartowski/Black-Ink-Guild_Pernicious_Prophecy_70B-GGUF --include "Black-Ink-Guild_Pernicious_Prophecy_70B-Q8_0/*" --local-dir ./

💻 Usage Examples

Prompt format

<|begin_of_text|><|start_header_id|>system<|end_header_id|>

{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>

{prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

📚 Documentation

Download a file (not the whole branch) from below:

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 should I choose?

🔧 Technical Details

The project uses the llama.cpp library for quantization, and specifically the release b4585. The imatrix option is used for quantization, and the dataset for calibration is from here.

Some of the quantization types use Q8_0 for embed and output weights, which can improve the quality of the model in some cases.

The online repacking feature for weights in Q4_0 can improve performance on ARM and AVX machines.

📄 License

The project is licensed under llama3.3.

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. Thank you to LM Studio for sponsoring my work.

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