asadullahdogarr/adaption-depin-orchestrator-instruct

This dataset is a remastered version prepared using Adaption's Adaptive Data platform.

adaption-DePIN-Orchestrator-Instruct

A specialized instruction-tuning dataset engineered to train AI agents in orchestrating Decentralized AI (DeFAI) microservices over DePIN infrastructure. It rigorously tests agentic reasoning through complex backend scenarios, including asynchronous FastAPI failure recovery, Redis token revocation, and dynamic inference routing between local PyTorch and Ollama. Designed to push model logic limits, the dataset injects contradictory user constraints requiring advanced conflict resolution. All target outputs are strictly enforced as JSON objects (userinstruction, agentreasoninglog, finalexecution_state).

Dataset size

There are 1,000 data points in this dataset. This is an instruction tuning dataset.

Quality of Remastered Dataset

The final quality is A, with a relative quality improvement of -5.0%.

Domain

• —Code (88%)
• —Technology (10%)
• —Governance (2%)

Language

• —English (100%)

Tone

• —Technical (96%)
• —Analytical (4%)

Evaluation Results

• —Quality Gains: <img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/1deda32a-726e-46de-aac9-4f9a3e5268d4.png" alt="QualityGains" style="max-width: 50%; display: block; margin-left: auto; margin-right: auto;" />

• —Grade Improvement: <img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/c0e4813a-8e51-41b9-aef2-6fef48049731.png" alt="Grade" style="max-width: 50%; display: block; margin-left: auto; margin-right: auto;" />

• —Percentile Chart: <img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/38dbaff2-6fa4-4dfa-8052-d964509c0aac.png" alt="Percentile Chart" style="max-width: 50%; display: block; margin-left: auto; margin-right: auto;" />