Closed JavierSanzCruza closed 2 years ago
There was an error on LightGCN which caused the dropout function to execute in evaluation time. I added a call to self.eval(), so training is equal to false each time the predict function is called.
self.eval()
predict
There was an error on LightGCN which caused the dropout function to execute in evaluation time. I added a call to
self.eval()
, so training is equal to false each time thepredict
function is called.