Gradients on Demand (G.O.D) runs competitive tournaments where miners' open-source AutoML scripts train LLMs and diffusion models on dedicated validator infrastructure. Tournaments last 4-7 days with 72-hour gaps between events, offering exponentially higher weight potential for top performers. First-place winning scripts are released to github.com/gradients-opensource, advancing open-source AI development.
0.0255This subnet is currently safe from deregistration.
No contribution data available
Yuma Pulse™
Tournament Competition
4-7 day competitive events where validators execute miners' AutoML scripts on dedicated training infrastructure
Open Source Winners
First-place tournament scripts published to gradients-opensource repository, benefiting the broader AI community
Dual Model Support
Competitions for both LLM fine-tuning and diffusion model training with separate Docker environments
Exponential Rewards
Top performers receive exponentially higher weight allocation, creating strong incentives for breakthrough approaches