Alpaca vs. Phi

Overview

Alpaca is an instruction-finetuned LLM based off of LLaMA.

The first of many instruct-finetuned versions of LLaMA, Alpaca is an instruction-following model introduced by Stanford researchers. Impressively, with only $600 of compute spend, the researchers demonstrated that on qualitative benchmarks Alpaca performed similarly to OpenAI's text-davinci-003, a significantly larger model.

Initial release: 2023-03-13

Reference

https://crfm.stanford.edu/2023/03/13/alpaca.html

Further Reading

Overview

Phi is a series of compact language models developed by Microsoft using textbooks and synthetic data.

Phi-1 and Phi-2 are 1.3B and 2.7B parameter language models, respectively, developed by Microsoft to demonstrate the ability of smaller language models trained on high-quality data. Despite its size and not having an instruct finetuned counterpart, the Phi-2 model is well-suited to research and experimentation given its size and MIT license.

Initial release: 2023-06-20

Reference

https://ai.azure.com/explore/models/microsoft-phi-2/version/4/registry/azureml-msr

Further Reading

• https://huggingface.co/microsoft/phi-2
• https://arxiv.org/abs/2306.11644