Loading...
Loading
ج / τ Price
0.01483 τ
+0.11%24h−0.59%7d
Overview
Registration Activity
Neuron registration trends
This Interval
0
24h: 0 registrations (0 burn · 0 PoW)
Target / Interval
1
Slots Available
Competitive
256/256
Neurons
256
Loading
Neuron registration trends
The alpha price in τ, day by day, and what moved it each way.
Log change over the last 30 days −8.00%: protocol buys +6.57%, root sells −1.16%, users net −12.68%, cross-subnet −0.90%, unattributed +0.17%.
The chart follows the alpha price in τ. Each period starts at its opening price and steps through the mechanisms below, in a fixed order, to its closing price: a step that rises pushed the price up, one that falls pushed it down, and its height is how far it moved the price. A thin line carries each close into the next period’s open; it is dashed where the two differ. The faint candle behind each period is the market’s own open, high, low and close, for context.
With many periods on screen (90 days, or hours), the steps would be too thin to read, so the chart draws the price as a line instead and colours each period by the mechanism that moved it most. Hover a period for the full step-by-step breakdown.
Not here, because they do not move the spot price: the per-block TAO injection into the pool, and burns.
All figures are log changes, not simple percentage changes, so the parts add up exactly to the total; on big moves the two differ (a 10% log fall is a 9.5% price fall).
2026-10-14 UTC, as of ≈06:18 UTC (block 9,279,675)
The day isn’t over yet, so these figures keep changing. They are in no daily total, average or rank.
Holder & total-staked metrics as of (daily close)
Training teams and researchers today either pay centralized cloud providers for GPU compute, or manage their own infrastructure at significant operational overhead. There is no standardized way to benchmark training improvements in a peer-reviewed, reproducible environment, and no marketplace that fairly compensates contributors for superior training methods.
A decentralized network can solve this by creating recurring tournaments where contributors submit training code, a distributed set of validators runs that code on shared GPU infrastructure using identical evaluation protocols, and a transparent scoring system ranks submissions and distributes rewards to top performers. This removes the need for individual compute ownership, creates a permanent record of what training strategies work best, and aligns financial incentives with actual training quality.
G.O.D (Gradients on Demand) on Bittensor implements this as a tournament-based subnet. Miners submit training repositories for three independent tournament types: text (instruct, DPO, GRPO, chat), image (diffusion/LoRA), and environment (reinforcement learning).
Validators automatically clone the submitted code at a specified commit, validate its structure, run it on tournament-managed A100/H100 GPUs using standardized datasets and timeouts, upload the resulting models to Hugging Face, and evaluate them against all other submissions using identical held-out test sets. Winners advance through group, knockout, and boss-round stages; tournament performance directly feeds validator weight allocation, so consistent training improvements translate to measurable economic reward.
Auto-generated from the GitHub repository and may not reflect the latest changes.
G.O.D (Gradients on Demand) is a Bittensor subnet that runs competitive training tournaments where miners submit open-source training code, validators execute it on managed GPU infrastructure, evaluate the resulting models, and advance winners through tournament rounds. Miners expose training repositories via a miner endpoint, and validators coordinate task creation, evaluation, scoring, and weight distribution across text, image, and environment tournament types.
The subnet consists of three main services: Miner (port 7999) exposes a repository endpoint that returns GitHub URL, commit SHA, and optional token for tournament registration; Trainer (port 8001) runs miner-submitted Docker containers on validator-controlled GPUs, manages model/dataset caching, and uploads outputs to Hugging Face; Validator (port 9001) orchestrates tournaments, coordinates task creation/distribution, evaluates submissions via Docker and Basilica, scores results, and sets subnet weights. Validators depend on Postgres (state), Redis (caching), S3-compatible storage (models/datasets), Fiber (Bittensor communication), and external services (LLM for obfuscation detection, SWE Infinite server for environment tasks). Training containers run in isolated Docker with assigned GPUs, read-only cache volumes, and mounted checkpoint directories, receiving standardized CLI arguments and environment variables defining tasks, datasets, and timeouts.
Tournament structure varies by type: text uses group/knockout/boss rounds with per-sample comparisons on instruct/DPO/chat tasks and score-based GRPO; image uses Flux/Ideogram/Z-Image in round 1, pairwise knockout in round 2, and 6-task boss round; environment uses PvP-style groups and environment-specific evaluation with swe_infinite in final rounds. Champions cannot decay over time and earn the same weight whether long-reigning or freshly-crowned. Boss-round win conditions require beating the defending champion by strict thresholds (55% sample win rate + mean gap for text/image; 1% score margin for GRPO/environment).
Challenger must win 5 of 6 boss tasks and all continuous-SFT tasks in text, or all but one of 6 image tasks, to dethrone. Environment challenger must have no losses and tie the boss on swe_infinite.
The sections above were auto-generated from the GitHub repository and may not reflect the latest changes. View repository (github.com, opens in a new tab, external site)
Best AutoML plaftorm in the world
This information was auto-extracted from the GitHub repository and may be incomplete or out of date. Always refer to the subnet's GitHub repository (github.com, opens in a new tab, external site) for the latest instructions.
Not yet verified: Notes
Not found in the repository docs.
Machine-extracted from the subnet's repository docs; may be wrong or out of date. Check the repository before buying or renting hardware. Open the repository (github.com, opens in a new tab, external site) · Extracted 28 Sep 2026
Clone the G.O.D repository
git clone https://github.com/rayonlabs/G.O.D.git && cd G.O.D
Install system dependencies
task bootstrap
Install the Python package
task install
Generate miner configuration (prompts for wallet, hotkey, network, and stake threshold)
task miner-config
Start the miner server on port 7999
task miner
Register your hotkey on the subnet
Register hotkey with: btcli s register --netuid 56 (mainnet) or btcli s register --network test --netuid 241 (testnet)
Post your miner IP to the metagraph
fiber-post-ip --netuid 56 --subtensor.network finney --external_port 7999 --wallet.name default --wallet.hotkey default --external_ip YOUR_PUBLIC_IP
Transfer tournament participation fee to collection address (0.7 TAO for text, 0.4 TAO for image, 0.6 TAO for environment)
Transfer TAO to 5Ef5JgNv14LY4UEQFHbRQkf8TnegDV3AfAbcsJe5T2w6VQdo
Configure your training repository response (GitHub repo URL, commit SHA, optional token, and requested datasets)
Edit miner/endpoints/training_repo.py
Clone the G.O.D repository
git clone https://github.com/rayonlabs/G.O.D.git && cd G.O.D
Install system dependencies
task bootstrap
Install the Python package
task install
Generate validator configuration (creates .vali.env with wallet, hotkey, network, DB, S3, and other settings)
task config
Add model hash salt to configuration
echo "MODEL_HASH_SALT=$(openssl rand -hex 32)" >> .vali.env
Start the validator (runs compose stack, Postgres, Redis, and validator service on port 9001)
task validator
Configure optional content service API key(optional)
Optional: echo "NINETEEN_API_KEY=<key>" >> .vali.env (for content service features)
Configure optional SWE Infinite server URL for environment tournament evaluation(optional)
Optional: echo "SWE_INFINITE_SERVER_BASE_URL=<url>" >> .vali.env (for environment tournaments)
| Day (UTC) | Share |
|---|---|
| 2026-10-13 | 1.68% |
| 2026-10-12 | 1.68% |
| 2026-10-11 | 1.68% |
| 2026-10-10 | 1.69% |
| 2026-10-09 | 1.70% |
| 2026-10-08 | 1.70% |
| 2026-10-07 | 1.72% |
| 2026-10-06 | 1.74% |
| 2026-10-05 | 1.73% |
| 2026-10-04 | 1.71% |
| 2026-10-03 | 1.77% |
| 2026-10-02 | 1.81% |
| 2026-10-01 | 1.76% |
| 2026-09-30 | 1.89% |
| 2026-09-29 | 1.83% |
| 2026-09-28 | 1.82% |
| 2026-09-27 | 1.86% |
| 2026-09-26 | 2.00% |
| 2026-09-25 | 1.97% |
| 2026-09-24 | 1.95% |
| 2026-09-23 | 1.93% |
| 2026-09-22 | 1.95% |
| 2026-09-21 | 1.93% |
| 2026-09-20 | 1.94% |
| 2026-09-19 | 1.88% |
| 2026-09-18 | 1.83% |
| 2026-09-17 | 1.85% |
| 2026-09-16 | 1.89% |
| 2026-09-15 | 1.92% |
| 2026-09-14 | 1.98% |
| 2026-09-13 | 1.96% |
| 2026-09-12 | 1.97% |
| 2026-09-11 | 1.95% |
| 2026-09-10 | 1.91% |
| 2026-09-09 | 2.05% |
| 2026-09-08 | 2.31% |
| 2026-09-07 | 2.37% |
| 2026-09-06 | 2.43% |
| 2026-09-05 | 0.86% |
| 2026-09-04 | 0.08% |
| 2026-09-03 | 0.08% |
| 2026-09-02 | 0.08% |
| 2026-09-01 | 0.09% |
| 2026-08-31 | 0.09% |
| 2026-08-30 | 0.10% |
| 2026-08-29 | 0.11% |
| 2026-08-28 | 0.11% |
| 2026-08-27 | 0.12% |
| 2026-08-26 | 0.08% |
| 2026-08-25 | 0.09% |
| 2026-08-24 | 0.09% |
| 2026-08-23 | 0.09% |
| 2026-08-22 | 0.11% |
| 2026-08-21 | 0.13% |
| 2026-08-20 | 0.12% |
| 2026-08-19 | 0.13% |
| 2026-08-18 | 0.17% |
| 2026-08-17 | 0.21% |
| 2026-08-16 | 0.21% |
| 2026-08-15 | 0.23% |
| 2026-08-14 | 0.24% |
| 2026-08-13 | 0.26% |
| 2026-08-12 | 0.30% |
| 2026-08-11 | 0.15% |
| 2026-08-10 | 0.17% |
| 2026-08-09 | 0.20% |
| 2026-08-08 | 0.18% |
| 2026-08-07 | 0.17% |
| 2026-08-06 | 0.19% |
| 2026-08-05 | 0.18% |
| 2026-08-04 | 0.24% |
| 2026-08-03 | 0.30% |
| 2026-08-02 | 0.29% |
| 2026-08-01 | 0.25% |
| 2026-07-31 | 0.29% |
| 2026-07-30 | 0.30% |
| 2026-07-29 | 0.37% |
| 2026-07-28 | 0.20% |
| 2026-07-27 | 0.46% |
| 2026-07-26 | 0.68% |
| 2026-07-25 | 0.69% |
| 2026-07-24 | 0.70% |
| 2026-07-23 | 0.75% |
| 2026-07-22 | 0.72% |
| 2026-07-21 | 0.70% |
| 2026-07-20 | 0.71% |
| 2026-07-19 | 0.73% |
| 2026-07-18 | 0.76% |
| 2026-07-17 | 0.77% |
| 2026-07-16 | 0.53% |
