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

feat: support images understaning in debugging assistant for local model #440

Open
wants to merge 4 commits into
base: development
Choose a base branch
from

Conversation

boczekbartek
Copy link
Member

@boczekbartek boczekbartek commented Feb 26, 2025

Purpose

  • Support image understanding features for local models

Note

Please don't squash commits on merge, they are loosely related

Proposed Changes

  • add GetRos2Image tool to debugging_assistant
  • describe details of ollama setup with OpenAI API
  • describe ollama setup with local VLMs for tool calling
  • fix path to debugging_assistant in docs

Issues

Testing

  • ollama model shoud be created as described in updated docs
  • config shoud be changed as described in updated docs
  • streamlit should be run as described in updated docs
  • I started manipulation demo binary in the background.

This is the conversation print that I got:
ROS 2 Debugging Assistant - llama3.2-vision-tools.pdf

Please note that in my tests:

  • Model hallucinates and calls too many tools, but the setup works (it called tool to get an image, then looked and interpreted the image)
  • For better models and prompt engineering performance should be better, this PR is an example of rai setup

Summary by CodeRabbit

  • New Features

    • Enhanced the debugging assistant with image retrieval capabilities for improved troubleshooting workflows.
  • Documentation

    • Updated launch instructions for the debugging assistant.
    • Expanded vendor guides with detailed setup instructions for configuring the Ollama server, OpenAI-compatible APIs, and vision model examples.

@boczekbartek boczekbartek force-pushed the feat/support_images_understaning_in_debugging_assistant_for_local_models branch from 9224fa5 to 44160ed Compare February 26, 2025 16:54
@boczekbartek boczekbartek force-pushed the feat/support_images_understaning_in_debugging_assistant_for_local_models branch from 44160ed to 3c4b10a Compare February 27, 2025 16:01
@boczekbartek boczekbartek marked this pull request as ready for review February 27, 2025 16:02
@boczekbartek
Copy link
Member Author

@coderabbitai full review

Copy link
Contributor

coderabbitai bot commented Feb 27, 2025

Walkthrough

This pull request updates the documentation and the debugging assistant tool. The documentation now reflects a new path for launching the debugging assistant and includes additional instructions for configuring the Ollama server with OpenAI-compatible APIs and vision models. In the code, the initialize_graph function is enhanced by initializing the ROS 2 client library, creating a ROS2 connector, and integrating a new image retrieval tool for ROS 2, without altering the overall functionality.

Changes

File(s) Change Summary
docs/debugging_assistant.md
docs/vendors.md
docs/debugging_assistant.md: Updated the command path from src/rai/rai/tools/debugging_assistant.py to src/rai_core/rai/tools/debugging_assistant.py.
docs/vendors.md: Expanded instructions for configuring the Ollama server, including detailed sections for setting up OpenAI APIs and vision model configurations.
src/rai_core/.../debugging_assistant.py Modified the initialize_graph function by calling rclpy.init(), creating a new ROS2ARIConnector, and adding GetROS2ImageTool(connector) to the list of agent tools for ROS 2 image retrieval.

Sequence Diagram(s)

Loading
sequenceDiagram
    participant Main
    participant initGraph as initialize_graph()
    participant rclpy
    participant ROS2Connector as ROS2ARIConnector
    participant ImageTool as GetROS2ImageTool
    participant Agent

    Main->>initGraph: Call initialize_graph()
    initGraph->>rclpy: rclpy.init()
    initGraph->>ROS2Connector: Create new ROS2ARIConnector instance
    initGraph->>ImageTool: Instantiate GetROS2ImageTool(connector)
    initGraph->>Agent: Call create_conversational_agent(tools list)
    Agent-->>Main: Return agent instance

Possibly related PRs

  • feat: debugging assistant #326: Introduces an interactive interface for the debugging assistant, involving modifications to debugging_assistant.py, which relates to the command path updates and enhanced functionalities in this PR.
✨ Finishing Touches
  • 📝 Generate Docstrings

Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

❤️ Share
🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

  • Review comments: Directly reply to a review comment made by CodeRabbit. Example:
    • I pushed a fix in commit <commit_id>, please review it.
    • Generate unit testing code for this file.
    • Open a follow-up GitHub issue for this discussion.
  • Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
    • @coderabbitai generate unit testing code for this file.
    • @coderabbitai modularize this function.
  • PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
    • @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
    • @coderabbitai read src/utils.ts and generate unit testing code.
    • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
    • @coderabbitai help me debug CodeRabbit configuration file.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

  • @coderabbitai pause to pause the reviews on a PR.
  • @coderabbitai resume to resume the paused reviews.
  • @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
  • @coderabbitai full review to do a full review from scratch and review all the files again.
  • @coderabbitai summary to regenerate the summary of the PR.
  • @coderabbitai generate docstrings to generate docstrings for this PR.
  • @coderabbitai resolve resolve all the CodeRabbit review comments.
  • @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
  • @coderabbitai help to get help.

Other keywords and placeholders

  • Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
  • Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
  • Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

  • Visit our Documentation for detailed information on how to use CodeRabbit.
  • Join our Discord Community to get help, request features, and share feedback.
  • Follow us on X/Twitter for updates and announcements.

Copy link
Contributor

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Actionable comments posted: 0

🧹 Nitpick comments (3)
docs/vendors.md (3)

5-7: Improved installation instructions

The expanded instructions for starting the Ollama server are helpful. However, consider using proper Markdown link formatting rather than a bare URL.

-For installation see: https://ollama.com/. Then start
+For installation see: [Ollama's website](https://ollama.com/). Then start
🧰 Tools
🪛 markdownlint-cli2 (0.17.2)

5-5: Bare URL used
null

(MD034, no-bare-urls)


15-33: Clear documentation for OpenAI compatibility

The new section on configuring Ollama with OpenAI compatible APIs provides valuable information and a clear example of the configuration required.

-> Such setup might be more convenient if you frequently switch between OpenAI API and
-> Such a setup might be more convenient if you frequently switch between OpenAI API and
🧰 Tools
🪛 LanguageTool

[uncategorized] ~20-~20: You might be missing the article “a” here.
Context: ...penai-compatibility)). > [!TIP] > Such setup might be more convenient if you frequen...

(AI_EN_LECTOR_MISSING_DETERMINER_A)


35-60: Good instructions for vision model setup

The detailed instructions for setting up vision models with tool calling support the PR's main objective of adding image understanding capabilities.

Consider these minor improvements:

  1. Fix the grammar issues:
-> Such setup is not officially supported by Ollama and it's not guaranteed to be
-> Such a setup is not officially supported by Ollama, and it's not guaranteed to be
  1. Fix the step numbering - currently shows steps 1 and 3 but no step 2:
-3. Configure the model through OpenAI compatible API in [config.toml](../config.toml)
+2. Configure the model through an OpenAI compatible API in [config.toml](../config.toml)
🧰 Tools
🪛 LanguageTool

[uncategorized] ~42-~42: You might be missing the article “a” here.
Context: ...e` and load the model > [!NOTE] > Such setup is not officially supported by Ollama a...

(AI_EN_LECTOR_MISSING_DETERMINER_A)


[uncategorized] ~42-~42: Use a comma before ‘and’ if it connects two independent clauses (unless they are closely connected and short).
Context: ...up is not officially supported by Ollama and it's not guaranteed to be > working in ...

(COMMA_COMPOUND_SENTENCE)


[uncategorized] ~52-~52: You might be missing the article “an” here.
Context: ...ols ``` 3. Configure the model through OpenAI compatible API in [config.toml](../conf...

(AI_EN_LECTOR_MISSING_DETERMINER_AN)

📜 Review details

Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 0119d2b and 3c4b10a.

📒 Files selected for processing (3)
  • docs/debugging_assistant.md (1 hunks)
  • docs/vendors.md (1 hunks)
  • src/rai_core/rai/tools/debugging_assistant.py (2 hunks)
🧰 Additional context used
🪛 LanguageTool
docs/vendors.md

[uncategorized] ~20-~20: You might be missing the article “a” here.
Context: ...penai-compatibility)). > [!TIP] > Such setup might be more convenient if you frequen...

(AI_EN_LECTOR_MISSING_DETERMINER_A)


[uncategorized] ~42-~42: You might be missing the article “a” here.
Context: ...e` and load the model > [!NOTE] > Such setup is not officially supported by Ollama a...

(AI_EN_LECTOR_MISSING_DETERMINER_A)


[uncategorized] ~42-~42: Use a comma before ‘and’ if it connects two independent clauses (unless they are closely connected and short).
Context: ...up is not officially supported by Ollama and it's not guaranteed to be > working in ...

(COMMA_COMPOUND_SENTENCE)


[uncategorized] ~52-~52: You might be missing the article “an” here.
Context: ...ols ``` 3. Configure the model through OpenAI compatible API in [config.toml](../conf...

(AI_EN_LECTOR_MISSING_DETERMINER_AN)

🪛 markdownlint-cli2 (0.17.2)
docs/vendors.md

5-5: Bare URL used
null

(MD034, no-bare-urls)

🔇 Additional comments (5)
docs/debugging_assistant.md (1)

25-25: Path correction looks good

The updated path to the debugging assistant script correctly reflects the actual location of the file in the project structure.

src/rai_core/rai/tools/debugging_assistant.py (4)

15-15: Good addition of ROS2 client library

Adding rclpy import is necessary for ROS2 initialization and communication.


21-30: Appropriate imports for new functionality

The new imports for ROS2ARIConnector and GetROS2ImageTool properly support the image understanding capabilities being added.


36-40: Effective ROS2 initialization

The implementation correctly initializes the ROS2 client library with rclpy.init() and creates a connector instance to facilitate communication with the ROS2 system.


43-51: Well-structured tool integration

The code properly integrates the new GetROS2ImageTool into the existing tool list while improving readability by listing each tool on a separate line. The connector is correctly passed to the image tool.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant