1

I am coding one neural network Quantum Monte Carlo software. I already finished all modules. However, when I try to use kfac_jax to optimize the parameters, I meet one bug,

jaxlib.xla_extension.XlaRuntimeError: INVALID_ARGUMENT: Attempt to donate the same buffer twice in Execute() (second use: flattened argument 1, replica 0). Toy example for this bug: f(donate(a), donate(a)).

I can run optimizer.init successfully.

anyone has ideas about this?

I dont konw how to find the sources of this bug.

Improve this question
1
  • Please edit your question to add a minimal reproducible example demonstrating your issue. Without this, it's unlikely that anybody will be able to answer your question.
    – jakevdp
    Commented Nov 11 at 13:14
Related questions
162
Many to one and many to many LSTM examples in Keras
1
How to take a derivative of one of the outputs of a neural network (involving batched inputs) with respect to inputs?
2
"The truth value of an array with more than one element is ambiguous" when trying to train a new JAX+Equinox model a second time
Load 7 more related questions Show fewer related questions

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.