vihangd/open_llama_7b_700bt_ggml

Open Llama 7b GGML Models

This repository contains the GGML models for the latest 700 billion token (700bt) preview of the Open Llama 7b model.

The following files can be used with llama.cpp or kobold.cpp:

• ggml-model-q4_0.bin: Quantized model, 3.79 GB
• ggml-model-q4_1.bin: Quantized model, 4.21 GB
• ggml-model-q5_0.bin: Quantized model, 4.63 GB
• ggml-model-q5_1.bin: Quantized model, 5.06 GB
• ggml-model-q8_0.bin: Quantized model, 7.16 GB