My code won’t pass the unit test, I get this assertion error:
I’ve been stuck for quite a while, can anyone figure out what’s wrong?
[edit] Solved: The argument of layernorm2 should be ffn_output + out1
My code won’t pass the unit test, I get this assertion error:
[edit] Solved: The argument of layernorm2 should be ffn_output + out1
Do not use the training argument on self.mha()
in self.layernorm2(), use out1 instead of attn_output.
Also, please edit your message to remove the code. That will clear the Honor Code.