Build GUI Chatbox for GPULlama3.java Inference with embedded GPU resource monitoring

[mikepapadim]

Description:

We want to create a user-friendly chat-style GUI for running LLM inference via our Java engine (powered by TornadoVM), inspired by the layout shown in the attached screenshot.

Desired Features

• Prompt Input

  • A text box where users can enter prompts for the LLM

• Dropdown Selectors

  • Engine: Choose between TornadoVM, JVM, etc.
  • Model: Select the model file to use for inference between Llama, Mistral etc

• File Picker

  • A "Browse" button to select .java files or model directories

• Run Button

  • A button to trigger inference (e.g., using tornado-llama-opencv)

• Output Display

  • A read-only area to show model responses and logs
  • Optional: System Monitoring Panel
  • Live GPU usage via nvtop or nvidia-smi
  • Live CPU & memory stats via htop or Java system metrics

Preferably, integrated with the core code base with JavaFX, or external with QT or python QT.

Example POC:

[svntax]

I'd like to take on this issue, but does this feature also depend on #17 being merged first?

[mikepapadim]

Hello @svntax, no this issue is completely indenpendent from other PRs.