New Feature: Add SambaNova as new AI connector

[snova-rodrigom]

---

Add SambaNova as new AI connector

It would be interesting to support SambaNova models in Semantic Kernel. Models there are very fast and precise.

I'd like to collaborate on adding SambaNova models to Semantic Kernel as new AI connectors. Do you have any docs related to how can someone help on adding new connectors? Thanks.

[markwallace-microsoft]

Hi @snova-rodrigom

I took a look at the SambaNova developer documentation and here is their Python example:

import os
import openai

client = openai.OpenAI(
    api_key=os.environ.get("SAMBANOVA_API_KEY"),
    base_url="https://api.sambanova.ai/v1",
)

response = client.chat.completions.create(
    model='Meta-Llama-3.1-8B-Instruct',
    messages=[{"role":"system","content":"You are a helpful assistant"},{"role":"user","content":"Hello"}],
    temperature =  0.1,
    top_p = 0.1
)

print(response.choices[0].message.content)

This uses the OpenAI client library (which also exists for .Net) so there is no need to create dedicated connector for this service.

[snova-rodrigom]

hey @markwallace-microsoft thank for your quick response. I work for SambaNova and I was just asking if adding a new connector would be the best way since I also see other LLM vendors there. I've already built an example using Semantic kernel and OpenAI connector pointing to our endpoints, and it's working great. If the latter is the approach we should take that's fine.

Let me know if you have any additional thoughts. Thanks!

[markwallace-microsoft]

Hi @snova-rodrigom
adding a concept sample would be the best option. Could you create a sample using the patterns demonstrated here: https://github.com/microsoft/semantic-kernel/tree/main/dotnet/samples/Concepts/ChatCompletion and update our list of samples here: https://github.com/microsoft/semantic-kernel/tree/main/dotnet/samples/Concepts#chatcompletion---examples-using-chatcompletion-messaging-capable-service-with-models.

Thanks!