Microsoft introduces cost-effective Phi-3-Mini language model series
MarTech — April 26, 2024, 05:00 PM UTC
Summary: Microsoft introduces Phi-3-Mini, the first in a series of small language models. Phi-3 models are cost-effective and outperform larger models in various benchmarks. Phi-3-Mini has 3.8 billion parameters compared to ChatGPT's trillion. It was trained on a dataset of 3.3 trillion tokens. These models may enable local use on devices like smartphones. The affordability of Phi-3 models could democratize AI access for small businesses.
Article metrics
Significance5.6
Scale7.0
Magnitude7.5
Potential8.5
Novelty8.0
Actionability9.0
Immediacy9.5
Positivity8.0
Credibility6.5