Hi,
When I tried to train the model, I got the below error("index out of range in self).
I guess it is due to embedding size. However I don’t know why this issue occur.
This is the projection layer code.
I’m using two node types (user and business). so I checked the data size used input tensor.
Could you give me any advice?
Could you also print the configuration (namely the number of embeddings) of the embedding layer that threw the error?
Sorry, but how can I print the configuration of the embedding layer?..
(fyr. I haven’t changed the layers and architecture of the pinsage model you have published on your github. and the input data shape is same with below screenshot)
Printing the layer itself should be fine.
print(emb_layer)
This usually gives something like
Embedding(20, 30)
where the first number is the number of embeddings created and the second number is the embedding size.
I really thank you for your response…
I don’t know embedding layer names… so just put the ‘emb_layer’ but it didn’t printed… 
but when I try to print model, I found that the embedding size is printed.
printed embedding size is all same (100)
I don’t know still why this issue occur.('index out of range in self)
