These are chat archives for deeplearning4j/deeplearning4j/earlyadopters

Dec 2017
Dec 09 2017 20:59

hello @all
I recently got the Error "New shape length doesn't match original length" in my YOLOOutput Layer while training.
I guess something with my data is invalid, but I don't understand the error. Can someone help me understand this Message?

It happens in the Yolo2OutputLayer:
// ----- Step 2: apply activation functions to network output activations -----
//Reshape from [minibatch, B*(5+C), H, W] to [minibatch, B, 5+C, H, W]
INDArray input5 = input.dup('c').reshape('c', mb, b, 5+c, h, w);