Problem: NaN or Inf in w[0]. : Tensor had NaN values

geffy / tffm

TensorFlow implementation of an arbitrary order Factorization Machine

MIT License

780 stars 176 forks source link

Problem: NaN or Inf in w[0]. : Tensor had NaN values #50

Open yy1987oo opened 5 years ago

yy1987oo commented 5 years ago

why no value in the bias tensor?

bigheiniu commented 5 years ago

Hello @yy1987oo, I have the same problem, did you solve this problem? Thanks

Noohshiny commented 4 years ago

I have the same problem. Any solution?

sajjadGG commented 4 years ago

It is because of the learning rate It happened for me too when I set it to 0.05 but when I lower it to 0.001 it was solved. I think there is a bug in the implementation.

lavafool commented 4 years ago

I delete the numerical feature in my train_sparse_data, then it doesn't happen anymore. But i don't know why..