Meta Llama 3 70B Now Available on Private LLM for Apple Silicon Macs
We're excited to announce that Private LLM now offers support for downloading a 4-bit OmniQuant quantized version of the Meta Llama 3 70B Instruct model on Apple Silicon Macs with 48GB or more RAM. This powerful model, developed by Meta, is part of the Llama 3 family of large language models and has been optimized for dialogue use cases.
Meta Llama 3 70B Model Architecture
The Llama 3 70B model uses an auto-regressive language model with an optimized transformer architecture. It has been trained on over 15 trillion tokens collected from publicly available sources, including a significant amount of high-quality non-English data covering more than 30 languages. Meta has also employed a combination of supervised fine-tuning (SFT), rejection sampling, proximal policy optimization (PPO), and direct preference optimization (DPO) to align the model with human preferences for helpfulness and safety.
Other Llama 3 Fine-Tuned Models
In addition to the Llama 3 70B model, Private LLM offers several fine-tuned versions of the smaller Llama 3 8B model, such as Dolphin 2.9 Llama 3 8B, Llama 3 Smaug 8B, Llama 3 8B based OpenBioLLM-8B, and Hermes 2 Pro - Llama-3 8B, on both iOS and macOS. With the larger Llama 3 70B model now available for local and offline use, users can experience a model that rivals GPT-4 in performance, all while maintaining privacy and security on their devices.
Private LLM's Commitment to User Privacy
As always, Private LLM remains committed to providing users with a secure, private AI chatbot experience that works locally on their iPhone, iPad, and Mac, without the need for an internet connection or subscription fees. We're thrilled to bring the power of Meta's Llama 3 70B model to our users and look forward to seeing how it enhances their AI-driven workflows and interactions.
