Changing CLIP model not processed

[imc67]

Here it says you have to change the json: https://docs.cloudron.io/packages/immich/

Here it has the options: https://docs.immich.app/install/config-file

I wanted to change the Smart Search CLIP model from default ViT-B-32__openai to ViT-SO400M-16-SigLIP2-384__webli with this additional code to the json, but it doesn’t work (restart app):

{
  "machineLearning": {
    "availabilityChecks": {
      "enabled": true,
      "interval": 30000,
      "timeout": 2000
    },
    "clip": {
      "enabled": true,
      "modelName": "ViT-SO400M-16-SigLIP2-384__webli"
    },
    "duplicateDetection": {
      "enabled": true,
      "maxDistance": 0.01
    },
    "enabled": true,
    "facialRecognition": {
      "enabled": true,
      "maxDistance": 0.5,
      "minFaces": 3,
      "minScore": 0.7,
      "modelName": "buffalo_l"
    },
    "ocr": {
      "enabled": true,
      "maxResolution": 736,
      "minDetectionScore": 0.5,
      "minRecognitionScore": 0.8,
      "modelName": "PP-OCRv5_mobile"
    },
    "urls": ["http://127.0.0.1:3003"]
  },
}

[imc67]

just tried again but without the very last comma .... and now it works.
poeh, what a comma can do ... or not