How to use predict func after training model ?

[monkey1302]

After training step, I got a 'model' object . And then how to use it to predict new data?

I tried model.predict([X, A],verbose=0) and got some error
InvalidArgumentError: Incompatible shapes: [32,32] vs. [32,9928] [[node model/graph_attention/add_15 (defined at <ipython-input-10-0d3199260314>:122) ]] [Op:__inference_predict_function_131650]

[danielegrattarola]

Hi,

like when using model.fit, you have to specify the batch_size to be equal to the number of nodes.

Btw, this repo is quite old and I suggest you look at the more up-to-date examples on the Spektral library:

https://github.com/danielegrattarola/spektral/blob/master/examples/node_prediction/citation_gat.py

Cheers

[XuRuiAngel]

Can model. predict() be used to process multiple graphs at once, receiving X with the shape of (batch_size, N, F) and A with the shape of (batch_size, N, N)?