Open Daisling opened 6 months ago
Same problem for me. I couldn't find a solution yet. If you did, please tell me.
@gianscarpe @FGiuliari any hint?
@gianscarpe @FGiuliari any hint?
Are there any news?
Hi, we discovered the problem, and it's related to the optimizer Adafactor. We are trying to carefully define the environment we were using for running the experiment. You will have an update soon.
Thanks for your great work! But when I run this code I meet this problem: