Loading a csv containing a series of Graphs, Pytorch Geometric

I have a .csv file which gives the same graph with different node features, as there are 104 nodes, this means that every 104 rows in the .csv, is the same graph but with different values for the node features.
e.g.

node,feature1,feature2,feature3
0, 1, 2, 3
1, 6, 5, 7
...
104, 0, 68, 1
0, 7, 9, 76
1, 7, 8, 9

I want to load this .csv in as the multiple (a few thousand) graph states it represents (i.e. the edge_index is the same for all of these graphs, the only difference is the node feature values.)

How would I do this in Pytorch Geometric? (i.e. load it in as a list of Data() objects)

Regards,
Jordan McConnon