I set 100 epochs but training stops when I reach 10 epochs?

[PhanTung-06]

In configs file train_abinet.yaml, I set training epochs = 100 or 1000, but training stops when it reach 10 epochs?

[PhanTung-06]

[PhanTung-06]

[sneedgers]

Same issue here

[anandpawara]

What was the solution? facing same issue

[PhanTung-06]

What was the solution? facing same issue

I used python 3.7 and error. Upgrade to python 3.8 to work.

[nguyennampfiev]

I have same issue, @PhanTung-06 You can training over 10 epochs by just update python 3.8.

[anandpawara]

I guess it's nothing to do with python version, it didn't work for me as well Try updating these values: scheduler: { periods: [6, 2], gamma: 0.1, } Eg: suppose if you are training the model for 100 epochs, then update values in periods as [80,20] split which is 80+20 = 100, this should work, which means 80 epochs for training and 20 epochs for evaluation, correct me if I am wrong