Output the same for every input (survival curves)

Hello, so I've been trying to use Coxtime from your package to do churn analysis, but I'm getting some problems on the output, namely I'm getting the same output, in this case the same survival curve, for every instance. I don't know what's causing this, but my guess would be that the network's weights are all the same (couldn't find a way to check this).

Just for context, my dataset has 5 features and around 200k instances. For my network, I'm using 8 hidden layers with 256 nodes each, with batch normalization, dropout of 0.8, learning rate of 0.001 and 200 epochs (with early stop activated on the validation set). Have you had this problem before, and if so is there something that I should check to ensure that the network is being well trained?

Thank you in advance!

No file or test is named in the report. Start by locating Coxtime's training and prediction entry points, then reproduce the identical-curve result with the stated data and settings and check whether inputs and learned weights vary; done means the cause is isolated and a verified fix or clear diagnostic is available.

Written by the indexing model from the issue text.