I am trying to upload a model with several referenced clusters containing C# (there are many instances) but this operation makes the gh file heavier than uploading directly the GH without clusterizing the C#. Any hint or recommendation? Also the file takes more time to process when uploading in ShapeDiver.
Are you sure you need that many clusters? Could you use data trees instead? Could you explode the clusters before uploading to ShapeDiver? How many of them do you have?