Roadmap

April 2024:

• Implemented foundational image generation network, integrating cosine similarity for image-text validation and reward distribution.

• Launched Web App, enabling direct utilization of subnet model outputs through a web interface.

May 2024:

• Introduced Magic Prompt feature, refining LLM to enhance user input prompts, resulting in higher quality images.

• Launched TG Bot + Discord Bot for enhanced accessibility.

July 2024:

• Introduced Magic Model feature, automatically matching appropriate models based on user prompts.

• Enabled model selection for users to choose preferred styles for image generation.

• Enhanced scheduling algorithms and validation algorithm to optimize user experience, service reliability, and image quality. This includes considerations for dimensions such as quality, response time, and relevance.

Q3 2024

• Enabled fine-tuning capabilities for enhanced customization.

• Decoupled miners from validators by integrating with an external model registry for improved efficiency.