Two new models: Arcee-Spark and Arcee-Agent

Julien Simon
2 min readAug 6, 2024

Two new Arcee.ai model videos for you today 😀 As usual, notebooks are available. Have fun testing these models, and don’t forget to follow Arcee.ai to stay on top of the latest Small Language Model action!

⚡ Arcee-Spark

Arcee-Spark is an 8-billion parameter model created by Arcee.ai from Llama-3.1 8B. Spark outperforms its base model on many language and math benchmarks. In the video, I first run a quantized version locally. Then, I deploy the model on Amazon Web Services (AWS) with Amazon SageMaker. I run both synchronous and streaming inference. I also show you how to use the OpenAI Messages API, allowing you to invoke the model with the OpenAI prompting format and minimizing the migration cost. Finally, I compare ChatGPT and Arcee-Spark, and explain the benefits of open-source SLMs like Arcee-Spark compared to closed models.

⚡ Arcee-Agent

Arcee-Agent is a 7-billion parameter model created by Arcee.ai from Qwen2–7B. Arcee Agent is currently one of the top models for function calls and tool usage, outperforming several versions of GPT-3.5 and GPT-4o In the video, I build a financial agent able to invoke the Yahoo Finance API to answer questions on listed companies, such as “what’s the stock price?”, “who’s the CEO”, “what is this company doing?” and more.

Sign up to discover human stories that deepen your understanding of the world.

Free

Distraction-free reading. No ads.

Organize your knowledge with lists and highlights.

Tell your story. Find your audience.

Membership

Read member-only stories

Support writers you read most

Earn money for your writing

Listen to audio narrations

Read offline with the Medium app

Julien Simon
Julien Simon

No responses yet

What are your thoughts?