← back to the directory
Library / SDK Fine-tuning & Training

Open-Instruct

1Recipes 2for 3training 4instruction-following 5language 6models.

the six Ws · specification

W1 Who

Built by the Allen Institute for AI.

W2 What

Open recipes and code for instruction tuning and RLHF on open models.

W3 Where

A Python codebase on GPU clusters.

W4 When

When reproducing or extending open post-training.

W5 Why

Transparent, reproducible post-training pipelines.

W6 With

PyTorch, Transformers, and vLLM.

W7 Watch

Apache-2.0, from AllenAI; research code, no credentials, runs on your hardware.

AllenAIInstruction-tuningRLHF

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.