Sign inSign up

rnrk/failurescope

By rnrk

Updated 4 months ago

ML-based C182 failure cascade sim: physics+XGBoost/Transformer ML, NTSB-validated, Ollama mistral:7b

Image
Machine learning & AI
Developer tools
0

2.0K

rnrk/failurescope repository overview

FailureScope

Discovery engine for aircraft failure cascade analysis.

Requirements

  • Docker
  • Ollama running on host with mistral:7b pulled

Quick Start

ollama pull mistral:7b
ollama serve

docker run -p 8080:80 \
  --add-host=host.docker.internal:host-gateway \
  -v /opt/failurescope_data:/opt/failurescope_data:ro \
  rnrk/failurescope:latest

Open http://localhost:8080

What it does

Runs 5 sequential layers: NLP parsing → constraint graph → physics simulation → XGBoost + Transformer ML inference → NTSB validation. 14 validated failure modes, 21,761 NTSB records cross-referenced.

Source: https://github.com/Rohnnam/failurescope

Tag summary

Content type

Image

Digest

sha256:b0e4752f3

Size

3.3 GB

Last updated

4 months ago

docker pull rnrk/failurescope