Guanaco vs. Phi

Overview

Guanaco is an LLM based off the QLoRA 4-bit finetuning method developed by Tim Dettmers et. al. in the UW NLP group. Guanaco achieves 99% ChatGPT performance on the Vicuna benchmark.

Guanaco is an LLM that uses a finetuning method called LoRA that was developed by Tim Dettmers et. al. in the UW NLP group. With QLoRA, it becomes possible to finetune up to a 65B parameter model on a 48GB GPU without loss of performance relative to a 16-bit model. The Guanaco model family outperforms all previously released models on the Vicuna benchmark. However, given the models are based off of the LLaMA model family, commercial use is not permitted.

Initial release: 2023-05-23

Reference

https://github.com/artidoro/qlora

Further Reading

• https://arxiv.org/abs/2305.14314
• https://twitter.com/Tim_Dettmers/status/1661379354507476994

Overview

Phi is Microsoft's family of compact open models designed for efficient local and edge deployment. Phi-4 is the current generation.

Microsoft introduced Phi-1 and Phi-2 to demonstrate how carefully curated and synthetic training data can make small models unusually capable. Phi-3 expanded the family for mobile and multimodal use. The Phi-4 generation includes a 14B general model and reasoning variants, plus smaller and multimodal checkpoints. Phi models are released under the MIT license and are designed for research, local inference, and resource-constrained production workloads.

Initial release: 2023-06-20

Current generation: Phi-4

Reference

https://azure.microsoft.com/en-us/products/phi

Further Reading

• https://huggingface.co/microsoft/phi-2
• https://arxiv.org/abs/2306.11644
• https://huggingface.co/microsoft/phi-4