Issue in the retraining of the Pinsage model

[NomanSaleem4]

Hey Man first Thanks for quickly responsding every time.

So now, here is the error while retraining the pinsage model: IndexError: index copy (): Number of indices (334) should be equal to source.size(dim) (106)

Here are the old and new data infos:

old data_info:  n_users: 385, n_items: 334, data density: 0.3111 % 
new data_info:  n_users: 447, n_items: 366, data density: 0.2445 %

Error traceback is also attached.

[massquantity]

The error shows the new model only has 106 items. You can use print("new n_items: ", model.data_info.n_items) to verify it.

I suspect you used the previous data_info in the new model construction, but without seeing all the code I'm not sure about this.