ROI-VARI ? Inference service

Model served by KServe ● Ready

Vegetation vigor analysis from an RGB image using the VARI index.

Details

Model	`roi-vari`
Format	`—`
Protocol	`v1`
Internal URL (cluster)	`http://roi-vari.mlops.internal.leelab.es`

How to send a request

Inference endpoint (v1):

POST https://roi-vari-mlops.leelab.es/v1/models/roi-vari:predict

curl example:

curl -X POST "https://roi-vari-mlops.leelab.es/v1/models/roi-vari:predict" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"instances": [{"image_b64": "<BASE64_ENCODED_PNG_OR_JPEG>"}]}'

Sample payload:

{
  "instances": [
    {
      "image_b64": "<BASE64_ENCODED_PNG_OR_JPEG>"
    }
  ]
}

Authentication

The public endpoint is protected by TLS + JWT (Keycloak). Add the Authorization: Bearer <token> header to every request sent to /v1 or /v2. This page (/) is public; the inference API is not.

Test the model

Paste your JWT and send the sample payload to verify that the model responds:

Not run yet.

Generated by mlops-backend.