Microsoft has recently unveiled a new lightweight AI language model called Phi-3 mini. This latest addition to their product line was developed following the release of their Phi-2 model in late 2023. The new model boasts 3.8 billion parameters trained on 3.3 trillion tokens, offering similar performance to competitor products.
The Phi-3 mini model was evaluated against academic benchmarks and internal testing, and it achieved a rating of 69 per cent in massive multitask language understanding (MMLU) and 8.38 in multi-turn (MT) benchmarks.
Microsoft used a scaled-up dataset training approach with heavily filtered web and synthetic data to develop the Phi-3 mini model, building on the foundation set by Phi-2. Despite its capabilities, the model is limited in size for certain tasks and lacks the capacity to store extensive factual knowledge.
The company stressed that they developed the Phi-3 mini model in alignment with their responsible AI principles, with input from internal teams helping to curate additional datasets tailored to address specific insights. This approach resulted in significantly reduced harmful response rates, demonstrating Microsoft’s commitment to ethical AI development.
The Phi-3 mini model is currently available through Microsoft’s Azure cloud platform, providing users with access to its advanced features and capabilities.