Stable Beluga LLM - An Overview

Stable Beluga is a cutting-edge AI language model that has captured the attention of developers and researchers alike. Developed by Stability AI, this model is derived from the popular Llama 2 architecture and is fine-tuned on a unique (Orca-style dataset), enabling it to perform efficiently across various tasks.

System Requirements & Model Variants

Different models come with varying memory requirements:

• 7B models generally require at least 8GB of RAM.
• 13B models generally require at least 16GB of RAM.
• 70B models generally require at least 64GB of RAM.

By default, Braina AI utilizes 4-bit quantization for efficient performance.

Available Aliases:

• latest, 7b, 7b-q4_0
• 13b, 13b-q4_0
• 70b, 70b-q4_0

Higher quantization levels improve accuracy but at the expense of speed and increased memory consumption.