Using Cloudera Copilot

You can use Cloudera Copilot to assist you in code generation and code completion thus providing flexibility and efficiency in the Cloudera AI workflows.

Start a session within your project and select JupyterLab as your editor.
Click the Jupyter AI Chat icon in the left navigation pane.
Start typing in the prompt and the Cloudera Copilot will assist you.
For example, enter Help me write a fibonacci function.
important
Embedding models are not selected by default in the Copilot UI. You can click the icon from the Cloudera Copilot chat interface and explicitly select the embedding model.

Example use cases for Cloudera Copilot:
- Debugging/fixing code
- Code completion
- Explaining code or explaining errors
- Code generation
- Refer to Jupyter AI documentation for instructions on /ask /learn /fix and /generate commands.
Magic commands:
- In addition to using the chat interface, you can also use Cloudera Copilot by calling Magic Commands within your notebook. Magic commands are special lines of code that will make calls to a language model that you have configured.
- Example syntax:
  - %load_ext jupyter_ai_magics
  - %%ai anthropic.claude-3-sonnet-20240229-v1:0
  - def fibonacci(
- Finding model ID strings for magic commands.
  - The Amazon Bedrock models recommended for Cloudera Copilot are:
    - anthropic.claude-3-sonnet-20240229-v1:0
    - anthropic.claude-3-opus-20240229-v1:0
    - mistral.mixtral-8x7b-instruct-v0:1
    - meta.llama3-70b-instruct-v1:0
    - amazon.titan-embed-text-v2:0
  - For Cloudera AI Inference service, the model ID string will be of the form:
    - cloudera:<model_name>. Here, the model_name is the same model ID from the Model Endpoints details page that you added when you configured Cloudera Copilot.