Update requirements and improve model loading logic

switched to qwen

Refactor Gradio node cleanup and enhance OAuth integration. Updated _cleanup_stale_gradio_nodes to handle multiple patterns for orphaned processes. Improved demo assignment in _launch_with_oauth for better hot reload compatibility with Gradio Spaces.

Enhance model loading and configuration. Added BORDERLESS_PRELOAD_MODEL to .env.example and updated README.md for clarity. Refactored device resolution in model.py for improved GPU handling and logging. Added warning filter for UserWarning in ui/__init__.py.

Refactor ZeroGPU handling and update documentation. Reduced default BORDERLESS_GPU_DURATION to 60 seconds in .env.example and README.md. Enhanced error messaging in app.js and gradio_api.js for better user feedback on quota issues. Updated GPU_DURATION logic in config.py to enforce minimum and maximum limits.

Updated the chat UI and backend so MiniCPM thinking renders

Add warning filter for UserWarning related to 'json' field in firecrawl.v2.types

Update requirements.txt to modify transformers version and remove unused Hugging Face dependencies

MiniCPM5 ZeroGPU migration

Fix string formatting in tool_display_title function for URL handling

fix empty research findings

Compact Chat Tool UI is implemented

Todo schema updated, Carousel UI scroll fix

Richer Globe Research Popups is implemented

Profile-based chat titles are implemented

Chat history rename and delete are implemented

parallel research carousel UI

Formatted Chat And Map Details

Structured Final Synthesis

Robust Planner Discovery Workflow

nodes modularized

chore: update README to reflect new LangGraph workflow and enhanced research process

fix: messages shown in the chat UI

REMOVED: os.environ["GRADIO_NUM_PORTS"] = "20"

Update README.md

cleaned-up requirements.txt

Update requirements.txt

langgraph integration

todo: langgraph integration

langchain example

Add gradioStream SSE reader in gradio_api.js and wire runChat in app.js to live updates

Surface think/reasoning as Thinking messages and add tool args/results to metadata.log in tools.py and respond.py

Make /chat a generator: stream_chat in server_api.py, generator api_chat in app.py, chunked final-answer yields in streaming.py

Add Gradio API integration and refactor app.js. Introduced gradioPredict function for API calls, enhancing chat functionality and session management.

Implement versioned asset loading and enhance homepage response. Added API endpoint for intake choices.

Added header controls and client-side chat histor. Assistant messages now render as markdown

Clicking an example now fills the form fields instead of sending to chat.

Enhance form layout in index.html by adding example prompts section and adjusting persona list display. Update CSS for form examples and persona list styling to improve user experience.

auth button fixed

right panel now toggles between two full-height views

Removed the citizenship & languages field

Revise header section in index.html to enhance user experience. Updated the main title to a more descriptive format and added a brief description of the service offered. Removed redundant hero section from the main workspace.

errors fixed

Fix starlette pin for Gradio 6.16.0 on Spaces.

Update authentication prompts to direct users to the new Server-mode login link and adjust chat messages accordingly following the removal of the Gradio sidebar.

Point auth prompts at the Server-mode login link.

Add custom HTML/JS frontend for gradio.Server mode.

Add gradio.Server entrypoint and queued API adapters.

Bump Gradio to 6.16.0 for gradio.Server support.

Refactor response handling and tool execution in the agent module for improved clarity and maintainability.