Reimplement training time and the performance on each task?

[Zzmonica]

As mentioned above.

[graykode]

Pre-training is Yes. But fine-tuning task is not Yet

[yuanjungod]

the q/k/v wrong!!

[xingchensong]

the q/k/v wrong!!

can u point out where the wrong code is? I compared this IMPL with ZihangDai'IMPL and didn't find anything wrong.

[xingchensong]

@graykode Hi~ GREAT THX for ur Pytorch IMPL of XLNet, I wonder whether u have a plan to impl fine-tuning part in pytorch?

[graykode]

@yuanjungod Could you tell me where is wrong point??
@xingchensong Thanks for your interest in my project. !! No not yet huggingface already did it!
https://github.com/huggingface/transformers < in here
You can see more easiliy as thought this project is shown about simple implementation such as how XLNet pre-training and pre-processing Thanks!