Crynux Network
HomeBlogGitHubTwitterDiscordNetstats
  • Crynux Network
  • Releases
    • Helium Network
    • Hydrogen Network
  • System Design
    • Network Architecture
    • Consensus Protocol
      • Inference Task Validation
      • Training/FT Task Validation
    • Verifiable Secret Sampling
    • Task Lifecycle
      • Task State Transitions
    • Task Dispatching
    • Task Pricing
    • Quality of Service (QoS)
    • Model Distribution
  • Node Hosting
    • Start a Node
      • Start a Node - Windows
      • Start a Node - Mac
      • Start a Node - Linux
      • Start a Node - Docker
      • Start a Node - Vast
    • Get the Test CNX Tokens
    • Private Key Security
    • Assign GPU to the Node
    • Proxy Settings
    • Docker Compose Options
    • Advanced Configuration
  • Application Development
    • How to Run LLM using Crynux Network
    • Application Workflow
    • Execute Tasks
      • Text-to-Image Task
      • Text-to-Text Task
      • Text-to-Music Task
      • Text-to-Video Task
      • Fine-Tuning Task
    • Crynux Bridge
    • API Specification of the Relay
    • Crynux SDK
  • Crynux Token
    • Wallet Configuration
      • Metamask
  • Troubleshooting
    • FAQ
    • Locate the Error Message
    • Exceptions in WebUI
  • Misc
    • Privacy Policy
Powered by GitBook
On this page
Edit on GitHub
  1. System Design
  2. Consensus Protocol

Training/FT Task Validation

PreviousInference Task ValidationNextVerifiable Secret Sampling

Last updated 7 months ago

Stable Diffusion Model Fine-tuning

The SD model fine-tuning task could be executed using a combination of all types of GPU models.

Rather than directly validating the result models, multiple images are produced using the models and a random prompt (seed) provided by the blockchain. for validating image generation tasks is applied to assess the similarity between images created by two models.

The average similarity score of these images serves as the measure of similarity between the two models. And models with the similarity score under a given threshold is considered the same model.

The method