Community fine-tuning platform. Contribute training data, improve models. Collective intelligence.

Go to file

elmadani 43d15032ac Initial commit		2026-02-25 00:53:32 +00:00
ix_forge.py	Initial commit	2026-02-25 00:53:32 +00:00
LICENSE	Initial commit	2026-02-25 00:53:32 +00:00
README.md	Initial commit	2026-02-25 00:53:31 +00:00
requirements.txt	Initial commit	2026-02-25 00:53:32 +00:00

README.md

IX Forge

Part of the Inference-X Ecosystem

Community fine-tuning platform. Collective intelligence.

The Forge is where the community contributes training data, improves models, and shares fine-tuned adapters (LoRA/QLoRA format). Open. Transparent. Community-owned.

Deploy

pip install fastapi uvicorn
python ix_forge.py
# → API at http://localhost:7937

API

POST /contribute       — Submit training data (Q&A pairs)
GET  /datasets         — Browse approved datasets
POST /submit-adapter   — Submit a trained LoRA adapter
GET  /adapters         — Browse community adapters
GET  /leaderboard      — Top contributors by training pairs
GET  /health           — Service status

Contribution Format

{
  "type": "qa",
  "language": "en",
  "domain": "science",
  "contributor": "your_handle",
  "pairs": [
    {"q": "What is...", "a": "It is..."}
  ]
}

Supported Domains

science, code, math, multilingual, reasoning, creative, general

Built in Morocco for the world. 🇲🇦

License

Business Source License 1.1 Author: Salka Elmadani