https://github.com/ging/tinyml-mlops-fiware-barrier
https://github.com/ging/tinyml-mlops-fiware-barrier
Last synced: about 1 year ago
JSON representation
- Host: GitHub
- URL: https://github.com/ging/tinyml-mlops-fiware-barrier
- Owner: ging
- License: mit
- Created: 2024-02-02T22:06:25.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-06-02T20:08:31.000Z (about 2 years ago)
- Last Synced: 2025-05-07T21:05:03.851Z (about 1 year ago)
- Language: Python
- Size: 495 KB
- Stars: 2
- Watchers: 10
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# FIWARE Machine Learning TinyML and MLOps - Barrier use case
Start base infraestructure
```
docker compose up -d
```
```
cd airflow
docker compose up -d
```
## With Airflow as orchestrator
- Access http://localhost:5000 to access MLFlow client
- Access http://localhost:8080 to access the Airflow Web UI (user: airflow, password: airflow)
- Initialize the dags:
- 1. `create_connection_dag` to create the connection to train server
- 2. `train_model` to train the model
Every 20 seconds the `urn:ngsi-ld:DensityDevice:1:Measurement:1` entity is updates, orion sends a notification to the predict system, who updates the `urn:ngsi-ld:DensityDevice:1:Prediction:1`
To get the entities:
```
curl localhost:1026/ngsi-ld/v1/entities/urn:ngsi-ld:DensityDevice:1:Measurement:1
```
```
curl localhost:1026/ngsi-ld/v1/entities/urn:ngsi-ld:DensityDevice:1:Prediction:1
```
```
curl localhost:1026/ngsi-ld/v1/subscriptions
```