Skip to content
Join the Network

The open frontier

Open infrastructure for collective frontier intelligence.

Contribute compute, train in the open and share every checkpoint. This dashboard shows the network as it really is: Run 0 is forming, and nothing has been trained yet.

Run Run 0 yourself Read the manifesto
  • 1Contributors on GitHub
  • 0GPUs in network
    none joined yet
  • 0Training runs
    Run 0 forming
  • 0Open models
    none released yet

Active training run

FormingDetails

OFL Run 0

Dense, about 1B parameters · OLMo-core · one GPU node

0% Not started: launcher, config and pinned upstreams are ready

  • 0 GPUs
  • 0 peers
  • CPU smoke passes

Your contribution

Sign up

0 GPU·h

Nodes are counted from Run 0 on, by GitHub handle, in the public compute ledger.

Recent activity

View all
  1. Lighthouse: system body font on phones, logo ratio, agent skillMerged in website #20
  2. Workspace: five-page contributor consoleMerged in website #19
  3. Add a contributors page and starter issues after sign-upMerged in website #18
  4. Add RFC, docs and starter-task issue forms; link DiscussionsMerged in .github #3
  5. Release: wait until the deploy token works for every scopeMerged in website #16
  6. Add contributor scaffolding: DCO, RFCs, CODEOWNERS, security, citationMerged in openfrontier #5

Global training network

No node has joined yet. The first lights appear when Run 0 opens.

  • 0Regions
  • 0Countries
  • 0Nodes online

Network utilization

Run 0: forming
  • NVIDIA (CUDA)0%
  • Apple Silicon0%
  • AMD (ROCm)0%
  • CPU / other0%

NVIDIA is the first target; Apple Silicon, AMD and the rest are planned.

Quick actions

Contribute computeSign up and join the waitlist Run trainingLaunch Run 0 on one GPU node Try the smoke testCPU only, no GPU needed See the dataDataset manifests and licences Read the designPhases, Run 0, joining a node

Every number on this page is real: GitHub data is read when the site is built, and everything else is zero until Run 0 starts. openfrontier-labs.com