Phi3: Your Pocket-Sized Language Model

In an era where information is at our fingertips, the demand for accessible and efficient language models has never been greater. Enter Phi3, the pocket-sized language model changing how we interact with technology. Trained on an impressive 3.3 trillion tokens with 3.8 billion parameters, Phi3 is a powerhouse despite its compact size. Its performance rivals that of larger models like Mixtral 8x7B and GPT-3.5, making it a game-changer in the world of natural language processing.

One of the things that makes Phi-2 better than Meta's Llama 2 7B and other models is that its 2.7 billion parameter size is very well suited for fitting on a phone. For instance, Phi-3-Mini achieves 69% on the MMLU benchmark and 8.38 on the MT-bench, making it suitable for deployment on mobile phones.?

Many of us may only need small models. Phi-3-Mini is highly capable and can run locally on a cell phone. Its small size allows it to be quantized to 4 bits, occupying approximately 1.8GB of memory. Microsoft tested the quantized model by deploying Phi-3-Mini on an iPhone 14 with an A16 Bionic chip, running natively on the device and fully offline, achieving more than 12 tokens per second.

Interestingly, Phi3 is uniquely positioned for computations where you don't need to go to the cloud to get things done. With its mobile compatibility and robust performance, Phi3 empowers users to perform complex language tasks directly on their devices without relying on external servers or internet connectivity. This capability enhances privacy and security and enables faster and more efficient processing, even on resource-constrained devices like smartphones.

Phi3 represents a significant advancement in the field of natural language processing, offering unparalleled performance and versatility in a compact package. Whether you're a professional writer, a language enthusiast, or simply looking to enhance your productivity, Phi3 is the ultimate pocket-sized companion for all your linguistic needs.