feat: add completion signal to stream response and update frontend to handle termination

refactor: migrate generation event handling to async iteration for improved control flow and cleanup

refactor: transition job processing from async iteration to event-based handling for improved lifecycle management

refactor: update reasoning content handling to support both string and list accumulation types

feat: implement streaming response support for real-time token generation and UI updates

feat: increase default media trim duration from 5 to 10 seconds

feat: replace text-only prompt cards with a visual grid of example media cards

feat: integrate multimodal example prompts into UI cards with automatic submission functionality

refactor: update audio message processing to use input_audio format with base64 data

fix: update default audio extension from mp3 to wav for processing

refactor: update default media trim duration from 10 to 5 seconds

refactor: replace deprecated subclip method with subclipped for video and audio processing

fix: update moviepy import path to support v2.0.0+ syntax

refactor: remove conditional import and optimize whitespace in media trimming utility

feat: add media trimming functionality for video and audio inputs using moviepy

feat: add interactive example media loading buttons for images, video, and audio

feat: implement multimodal file attachment support for images, audio, and video via file picker and drag-and-drop

feat: add model configuration to README metadata

refactor: redesign UI shell and messaging layout with updated color palette and modern component styling

feat: implement Nemotron-3-Nano-Omni chat interface with reasoning support using FastAPI and Gradio

initial commit