search

NoviScl / XLNet_DREAM

29 stars 6 forks source link

bug #2

Closed wind91725 closed 5 years ago

wind91725 commented 5 years ago

modeling_xlnet.py ,line 892 in forward mems_mask = torch.zeros([data_mask.shape[0],men,bsz]).to(data_mask)

RuntimeError : sizes must be non=negative

NoviScl commented 5 years ago

Hi,

Are you using a GPU to run? May I know the full command you use and the complete log output?

wind91725 commented 5 years ago

Hi，thank you for your reply. Here is full command and complete log output . On 08/6/2019 07:06，Chengleinotifications@github.com wrote：

Hi,

Are you using a GPU to run? May I know the full command you use and the complete log output?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

NoviScl commented 5 years ago

Note: this issue has been solved by updating the versions to: Python 3.6, Pytorch 1.0.1, cuda10