Hi,
I am trying to have a minimum footprint deployment of a tensorflow gnn model. Seems that the tf-GNN models take in graphTensor as input which would require both tensorflow and tfgnn libraries to be present. Is there a way to have tf-GNN lite model to run inference without needing to import the tfgnn library? Thanks!
Hi, I am trying to have a minimum footprint deployment of a tensorflow gnn model. Seems that the tf-GNN models take in graphTensor as input which would require both tensorflow and tfgnn libraries to be present. Is there a way to have tf-GNN lite model to run inference without needing to import the tfgnn library? Thanks!