Hi,
I tried to use my dataset but got error due to NaN after several epochs during training in the following variable:
after tracking this gives NaN:
- unwrapped_q_log_q
- batch_each_c_neg_entropy
- each_c_entropy
do you have suggestion on how to solve this?
thank you
Hi,
I tried to use my dataset but got error due to NaN after several epochs during training in the following variable:
after tracking this gives NaN:
do you have suggestion on how to solve this?
thank you