Size of tensors much match runtime error

[demongolem-biz]

When I load the embedding matrix at emb_matrix = np.load(emb_file) I see the shape is (39,300). When I go to train I get the error RuntimeError: Sizes of tensors must match except in dimension 1. Got 9 and 10 (The offending index is 0). Before erroring, I see the shape of emb_matrix is (torch.Size([50, 10, 300]). I think this is somehow that 10 does not divide 39, but being new to this code I don't quite get it. What would I need to do to get it to train?

[demongolem-biz]

I think where I got confused is as follows. In data/loader.py, the subject start/end and object start/end are used to create the size of the tokens structure. I thought that within the Stanford ecosystem, these values were 1-based. However, in this file, it appears a 0-based scheme. So if I subtracted one off of these 4 values that I feed in as input data, that would appear to correct it, but I am still confused because it seems to disagree with what was done in Stanford CoreNLP (and then possibly Stanza?)