Overview

Phi 3.5, Microsoft's small language model (SLM) created in August 2024, is an open source state of the art model. Phi 3.5 comes in two models: Phi-3.5 Mini, and Phi-3.5-MoE.

Compared to various other SLM's released in 2024, Phi-3.5 Mini is by far the smallest. Despite its size, it still manages to perform above par in terms of accuracy on the MMLU benchmark. Meanwhile Phi-3.5-MoE is highly competent on the benchmark, beating out models exponentially larger than itself, such as Mistral.

‍

Phi-3.5 License

Phi-3.5

is licensed under a

license.

Phi-3.5 Alternatives

Explore alternatives to

Phi-3.5

Vision Transformer

MetaCLIP

YOLOS

MobileNet SSD v2

YOLOX

YOLOv9

CoDet

YOLOv10

Faster R-CNN

YOLOv3 PyTorch

Performance

Deploy a Phi-3.5 API

You can use Roboflow Inference to deploy a

Phi-3.5

API on your hardware. You can deploy the model on CPU (i.e. Raspberry Pi, AI PCs) and GPU devices (i.e. NVIDIA Jetson, NVIDIA T4).

Below are instructions on how to deploy your own model API.

Label Data Automatically with Phi-3.5

You can automatically label a dataset using Phi-3.5 with help from Autodistill, an open source package for training computer vision models. You can label a folder of images automatically with only a few lines of code. Below, see our tutorials that demonstrate how to use Phi-3.5 to train a computer vision model.

No items found.