- PuLID Workflow
- Now supports PuLID workflow
- Upload a reference image to generate an image using a PuLID just fill in style information
- example: image of a woman wearing a yellow dress, "/pulid select your resolution, enter in the prompt " woman with a red dress"
- Security Implementation
- Implemented a security middleware to prevent abuse and protect against malicious requests.
- Added a permanent block list to block IP addresses that have been repeatedly tried to access restricted endpoints.
- you can check the security folder for the permanent block list and see what they were requesting.
- AI Prompt Enhancements
- Now supports Google Gemini 2.0 Flash
Banned words list
- Updated the banned words list to include a list of common profanities and sensitive terms to avoid discord flagging and banning.
- User can add new words to the list or use the
/add_banned_word
command in discord to add words manually. - Remember USERS ARE 100% RESPONSIBLE FOR WHAT THEY POST IN DISCORD, Generated images that contain sensitive content may result in a permanent ban from discord.
- Please make sure your users are well aware of the banned words list and discords policies.
-
Flux Redux
-
/redux command to generate an image using two reference images
-
/reduxprompt command to generate an image using a reference image and prompt
-
redux first image will be the main image, 2nd image is style reference
-
reduxprompt image allows user to use a prompt to generate an image from a reference image with a style prompt
-
Various fixes and improvements
-
resolved setup tool not saving server_address
-
expanded fields in setup tool to allow better view of servers and channels
-
fixed issues banned and why banned giving errors
-
-
LMStudio / Ai Prompt Enhancements
- Now you can select LMStudio, xAI or OpenAI with in the setup tool
- When generating prompts a modal will appear to select Creativity level
- Creativity level can be adjusted from 1-10 with 1 not changing your prompt and 10 extremely creative.
- can enable or disable AI Prompt Enhancements using the setup tool
-
Various fixes and improvements
- Added features to the setup tool for AI Prompt Enhancements
- updated workflows in Datasets to support new clip models
- updated Lora Editor and made it more user friendly and robust
-
Enhanced LoRA Weight Management
- Automatic weight scaling for multiple LoRAs
- Full weight utilization for single LoRA selections
-
LoRA Info Command
- Detailed information display
- URL links to example images
- Visual reference guides
-
Improved LoRA Editor
- Modernized user interface
- Double-click editing functionality
- Automatic save feature
- Enhanced user experience
-
Numpy 2 Compatibility
- Fixed workflow loading errors
- Modified file system for compatibility
- Maintained backward compatibility
-
Setup Tool Improvements
- Added folder validation system
- Enhanced base directory specification
- Example path:
C:/Comfyui_windows_portable/ComfyUI
-
Lora Monitor Enhancement
- Improved timing for LoRA list updates
- Better update status verification
- Enhanced reliability
[To be added]
[To be added]
- All updates are tested thoroughly before release
- Backup your configurations before major updates
- Check compatibility with your current setup