#8 causes NaNs almost immediately during training

openai / iaf

Code for reproducing key results in the paper "Improving Variational Inference with Inverse Autoregressive Flow"

https://arxiv.org/abs/1606.04934

MIT License

517 stars 130 forks source link

#8 causes NaNs almost immediately during training #11

Open craffel opened 6 years ago

craffel commented 6 years ago

If I run the TensorFlow version of this code (tf_train.py) with https://github.com/openai/iaf/pull/8 applied, I get a NaN within the first few iterations and training stops. If I remove that change, training proceeds fine. @pukkapies were you ever able to get the model training appropriately with your changes applied? If so, what hyperparameter settings were you using?