Skip to content

Major Update Proposal : Vision support, File attachments, Context fixes #17

@Lognews

Description

@Lognews

Hello!

I've been using your GUI and I really appreciate your work. I decided to improve the script a bit to add support for multimodal models and fix some context/layout issues.

Here is a significantly updated version (v1.3

ollama_gui_1.3.0.zip

.0).

Changelog:

Multimodal Support: Added a button to attach Images. Supports models like llava, moondream, etc. (Requires Pillow).
File Attachments: Added support for attaching Text Files (.txt, .py, etc.). The content is automatically read and sent to the model hidden from the chat UI to keep it clean.
Smart Context Window: Implemented a rolling window mechanism. It prunes old messages from the prompt if the conversation exceeds the num_ctx limit, preventing crashes on long chats.
UI Improvements:
Added a Context Usage indicator (shows %).
Added a Tokens/sec counter.
Fixed scrolling behavior (no more jumping when text is generated).
Bug Fixes:
Fixed Russian keyboard layout hotkeys (Ctrl+C/V, Delete) using keycodes.
Fixed settings saving (UTF-8 support for config).
Model Settings: Added a UI to configure temperature and system prompt per model.
I would be happy if you could integrate these changes into the main repository!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions