1 versions
A lightweight AI model with 3.8 billion parameters with performance overtaking similarly and larger sized models.
Install our magic
package manager:
curl -ssL https://magic.modular.com/ | bash
Then run the source
command that's printed in your terminal.
Install Max Pipelines in order to run this model.
magic global install max-pipelines
Start a local endpoint for phi3.5/3.8b:
max-serve serve --huggingface-repo-id microsoft/Phi-3.5-mini-instruct
The endpoint is ready when you see the URI printed in your terminal:
Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
Now open another terminal to send a request using curl
:
curl -N http://0.0.0.0:8000/v1/chat/completions -H "Content-Type: application/json" -d '{
"model": "phi3.5/3.8b",
"stream": true,
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Who won the World Series in 2020?"}
]
}' | grep -o '"content":"[^"]*"' | sed 's/"content":"//g' | sed 's/"//g' | tr -d '
' | sed 's/\n/
/g'
🎉 Hooray! You’re running Generative AI. Our goal is to make this as easy as possible.
Phi-3.5-mini is a compact, cutting-edge open AI model built on datasets from Phi-3, which includes synthetic data and high-quality, reasoning-rich public web content. It emphasizes precision and safety through technologies like supervised fine-tuning, proximal policy optimization, and direct preference optimization.
Part of the Phi-3 family, this model supports an extensive 128K token context length, enabling it to tackle tasks such as long document summarization, question answering, and information retrieval with impressive efficiency. Its advanced design ensures both strong instruction adherence and robust performance across various scenarios.
With its 128K context length, Phi-3.5-mini excels at handling complex and extended content, making it suitable for applications involving long-form content processing.
DETAILS
MODULAR GITHUB
ModularCREATED BY
microsoft
MODEL
microsoft/Phi-3.5-mini-instruct
TAGS
@ Copyright - Modular Inc - 2024