Dippy SN11: Enterprise Grade Generative Media Platform on Bittensor

Please check our Website for more details on our vision.

)

---

• Introduction
• Roadmap
• Overview of Miner and Validator Functionality
  • Miner
  • Validator
• Running Miners and Validators
  • Running a Miner
  • Running a Validator
• License

---

Introduction

Note: The following documentation assumes you are familiar with basic Bittensor concepts: Miners, Validators, and incentives. If you need a primer, please check out Bittensor Docs

Dippy is one of the world's leading AI companion apps with 7M+ users. The app has ranked #3 on the App Store in countries like Germany, been covered by publications like Wired magazine and the average Dippy user spends 1+ hour on the app.

When Dippy moved beyond text inference, we discovered that Bittensor still doesn’t have production-ready subnets for media inference. The Dippy Studio subnet aims to address this gap in the market and power the fastest, media generation platform on Bittensor. It's scalable, multi-modal, and ready to serve millions of users on Day 1. Our custom, in-house Dippy Inference Engine means that our users can reliably expect the fastest and cheapest inference available on the market.

Roadmap

Given the complexity of starting an enterprise grade inference engine from scratch, we plan to divide the process into 3 distinct phases.

Phase 1:

• Subnet launch with deterministic inference from Day 1
• Dippy Studio powers all Dippy in-chat images using Dippy Inference Engine serving millions of users
• Beta version of self-serve API access for external builders to start building

Phase 2:

• Release dashboard with live statistics on every model
• Set up our in-house Dippy Inference Engine pipeline to optimize any image model
• On-board 50+ AI startups with Dippy Studio's image inference

Phase 3:

• Expand modalities to audio, video, 3D etc while maintaining our speed advantage
• Power 100+ AI startups with fully decentralized media infrastructure

Overview of Miner and Validator Functionality

Note that this repository mainly serves for validator code. There are two critical sub components - orchestrator and miner - which are linked here but have their own respective repositories.

Miners Run inference on an optimized version of the image and other media models using Dippy's Inference Engine.

Validators evaluate and assess the output quality, consistency and latency of miner outputs. Since outputs are deterministic,

Running Miners and Validators

Running a Miner

For detailed miner setup instructions, please see Miner Documentation.

Running a Validator

For detailed validator setup instructions, please see Validator Documentation.

License

The Dippy Bittensor subnet is released under the MIT License.

Project Structure Overview

Core Components

1. Miner Implementation

• dippy-studio-bittensor-miner/ - Main miner application

2. Ochestrator Implemetnation

• dippy-studio-bittensor-orchestrator/ - Orchestrator system

3. Documentation

• docs/ - Project documentation
  • miner.md - Miner setup and usage guide
  • validator.md - Validator setup and usage guide