Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add few-shot prompting to rai demo projects #346

Open
boczekbartek opened this issue Dec 11, 2024 · 0 comments
Open

add few-shot prompting to rai demo projects #346

boczekbartek opened this issue Dec 11, 2024 · 0 comments
Labels
enhancement New feature or request good first issue Good for newcomers llm This issue is related to Large Language Models

Comments

@boczekbartek
Copy link
Member

boczekbartek commented Dec 11, 2024

Is your feature request related to a problem? Please describe.
Currently, demo projects in rai use prompts that contain:

  • robot identity from rai_whoami (link)
  • general mission for the robot (link)
  • tips and guidelines about ros2 (link)
  • tips and guidelines about the task (link, link)
  • documentation of attached tools (link)

Describe the solution you'd like
Few-shot prompting (https://www.promptingguide.ai/techniques/fewshot) is a technique commonly used to improve LLM agents' performance. The current demo shows that LLMs can achieve good zero-shot capabilities, but with few-show prompting, they have the potential to be significantly improved.

Few-shot prompting can be applied by adding successful interactions to the system prompt. This can be a sequence of tool calls or tool calls with specific parameters.

Describe alternatives you've considered
Fine-tuning is a next step, but few-shot should be checked first, as much cheaper and easier.

Additional context

@boczekbartek boczekbartek added enhancement New feature or request good first issue Good for newcomers llm This issue is related to Large Language Models labels Dec 11, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request good first issue Good for newcomers llm This issue is related to Large Language Models
Projects
None yet
Development

No branches or pull requests

1 participant