Releases: TecEash1/Taurus
v4.5
v4.4
✨ Adds an option to disable/enable Gemini safety settings #41
✨Adds a dedicated 🧠 Models
section to the settings page
⚒️ Fixes generationConfig
and safetySettings
not being passed in properly
⚒️ Renames Load Balancing to Fallback System
Update Guide (From Previous Versions)
- Make sure you pull the settings database (This will wipe your old settings)
- Use the
/settings
command to add back your settings.
Full Changelog: v4.3...v4.4
v4.3
⚒️ Drastically improves the performance of the /image
command when NSFW Image Blocking
is enabled by caching the model and NSFW words. It now downsizes the image to be detected as well to improve performance.
⚒️ Adds asynchronous loading for model fetching
⚒️ Fixes the broken config.json
module in the image autocomplete, it now utilises the Settings Database
Full Changelog: v4.2...v4.3
v4.2
⚒️ Fixes a critical crashing error
Full Changelog: v4.1...v4.2
v4.1
⚒️ Fixes Console Webhook not updating live when submitted through /settings
(Fixes the need to restart the bot)
⚠️ This version has a critical error
Full Changelog: v4.0...v4.1
v4.0
✨ An all-new settings page, update your bot configuration live via modals and buttons inside of Discord!
🧠 Switch between Gemini Pro 1.5 and Google's new Gemini Flash 1.5 model
Gemini Flash 1.5 Flash is optimised for chatbots, it has larger rate limits and quicker response times
✨ Load Balancing (Between the Flash model and Pro models)
⚒️ Improved API Key Checking
⚒️ Improved Gemini Error Handling
⚒️ Moves the utils file #37
⚒️ Improves handlers and implements a Button Handler
📦 Updates Packages
Update Guide (From Previous Versions)
- Modify your
config.json
file to be in the same format as the example - Once the bot has loaded run the
/settings
slash command to edit your bots config (You will have to set your config up again)
Full Changelog: v3.7...v4.0
v3.7
✨ Adds 2x upscaling to image command
✨ Adds option to specify the height/width of the image
✨ Upgrades max image steps to 50
⚒️ Fixes max CFG scale
Full Changelog: v3.6...v3.7
v3.6
v3.5
⚒️ Fixes response.text is not a function
error. -> Fixes issues with that error embed being thrown when the user asks the model something that requires a long response.
✨ Adds auto-retrying with countdown when token limit is reached #31
📦 Updates Packages
Full Changelog: v3.4...v3.5