Microsoft has recently introduced a new AI language model called Phi-3 mini, following the release of their Phi-2 model in late 2023. This latest edition boasts 3.8 billion parameters trained on 3.3 trillion tokens, offering similar performance to competitor products. Microsoft highlighted that Phi-3 mini was evaluated against academic benchmarks and internal testing, proving its ability to compete with models such as “Mixtral 8x7B and GPT-3.5”.
The Phi-3 mini model achieved a rating of 69 per cent in massive multitask language understanding (MMLU) and 8.38 in multi-turn (MT) benchmarks. Despite its capabilities, Microsoft acknowledged that Phi-3 mini is limited in size for certain tasks and lacks the capacity to store extensive factual knowledge. To address this limitation, Microsoft used a scaled-up dataset training approach with heavily filtered web and synthetic data to develop Phi-3 mini, building on the foundation set by Phi-2.
Microsoft emphasized that they developed Phi-3 mini with responsible AI principles in mind, 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.
