> ## Documentation Index
> Fetch the complete documentation index at: https://docs.honeyhive.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# NVIDIA NeMo

> Learn how to integrate NVIDIA NeMo Models with HoneyHive

export const integrationType_1 = "NVIDIA NeMo"

export const integrationType_0 = "NVIDIA NeMo"

NVIDIA NeMo offers a suite of leading-edge NVIDIA-built and open-source generative AI models, meticulously fine-tuned for exceptional performance and efficiency. With the ability to deploy these models using NVIDIA NIM™ microservices and customize them through NeMo, developers can swiftly prototype and scale their AI applications.

With HoneyHive, you can trace all your {integrationType_0} operations using a single line of code. Find a list of all supported integrations [here](/introduction/troubleshooting#latest-package-versions-tested).

## HoneyHive Setup

Follow the [HoneyHive Installation Guide](/integrations/integration-prereqs) to get your API key and initialize the tracer.

## NeMo Setup

Go to the [NeMo Playground](https://build.nvidia.com/meta/llama-3_1-405b-instruct?api_key=true) to get your NVIDIA API key.

## Example

Here is an example of how to trace your {integrationType_1} code in HoneyHive.

<CodeGroup>
  ```python Python theme={null}

  # NVIDIA uses OpenAI client to interact with their API
  from openai import OpenAI  
  from honeyhive import HoneyHiveTracer

  # place the code below at the beginning of your application execution
  HoneyHiveTracer.init(
      api_key="MY_HONEYHIVE_API_KEY", # paste your API key here
      project="MY_HONEYHIVE_PROJECT_NAME", # paste your project name here
  )

  client = OpenAI(
      base_url="https://integrate.api.nvidia.com/v1",
      api_key="MY_NVIDIA_API_KEY",
  )

  completion = client.chat.completions.create(
      model="nvidia/mistral-nemo-minitron-8b-8k-instruct",
      messages=[
          {
              "role": "user",
              "content": "Write a limerick about the wonders of GPU computing.",
          }
      ],
      stream=True,
  )

  for chunk in completion:
      if chunk.choices[0].delta.content is not None:
          print(chunk.choices[0].delta.content, end="")
  ```

  ```tsx TypeScript theme={null}
  import OpenAI from 'openai';
  import { HoneyHiveTracer } from 'honeyhive';

  const tracer = await HoneyHiveTracer.init({
      apiKey: 'MY_HONEYHIVE_API_KEY',
      project: 'MY_HONEYHIVE_PROJECT_NAME',
      sessionName: 'test',
  });

  const openai = new OpenAI({
      apiKey: "MY_NVIDIA_API_KEY",
      baseURL: 'https://integrate.api.nvidia.com/v1',
  })

  async function main(input: string): Promise<string> {
      const completion = await openai.chat.completions.create({
          model: "nvidia/mistral-nemo-minitron-8b-8k-instruct",
          messages: [{ role: "user" as const, content: input }],
          temperature: 0.2,
          top_p: 0.7,
          max_tokens: 1024,
          stream: true
      });

      let ret = '';
      let content: string;
      
      for await (const chunk of completion) {
          content = chunk.choices[0]?.delta?.content ?? '\n';
          process.stdout.write(content);
          ret += content;
      }
      return ret;
  }

  const tracedMain = tracer.traceFunction()(main);
  await tracedMain("Write a limerick about the wonders of GPU computing.");
  ```
</CodeGroup>

## View your Traces

Once you run your code, you can view your execution trace in the HoneyHive UI by clicking the `Log Store` tab on the left sidebar.

<img src="https://mintcdn.com/honeyhiveai/oxFnPfpA79Ja1RXX/images/traces/nemo-trace.png?fit=max&auto=format&n=oxFnPfpA79Ja1RXX&q=85&s=48a8ae26d276fbf12c6a04b06b32672b" alt="NeMo Traces" width="1104" height="801" data-path="images/traces/nemo-trace.png" />