Error in transformed variables with manual model vectorization

[jdehning]

Error in transformed variables with manual model vectorization

[jdehning]

I think I have some ideas on how to fix it. Probably next week I will have some time to look deeper into it and open a PR, if you are okay with it...

[lucianopaz]

Thanks for reporting this @jdehning. I can reproduce the exception you are running into and can confirm that its coming from the gradient computation. I still have to investigate more before confirming that the culprits are inverse_log_det_jacobian and forward_log_det_jacobian, but what you posted was very helpful to pinpoint the cause thus far. I'm not sure when I'll get a chance to fix this. Maybe @ferrine has some time to look into the problem a bit more.

[ferrine]

Hmm, I can have a look after weekends

[ferrine]

Looks like I failed to fix the error quickly, I've tried to figure out if it is the wrong usage of reinterpreted_batch_ndims=2 and similar arguments but it was not successful.