System information
-PaddlePaddle version 2.1.1
-CPU: including CPUMKL/OpenBlas/MKLDNN version
-GPU: not used
-OS Platform and Distribution: Mac OS 10.15
-Python version: 3.9.0
Note: You can get most of the information by running summary_env.py.
To Reproduce
See code below. Paddlepaddle doesn't handle zero-dimensional tensors (same issue as in mxnet v1, now fixed)
Describe your current behavior
Code segfaults.
[1] 71657 segmentation fault ipython
Code to reproduce the issue
import paddle
paddle.zeros([]).transpose([]) # should return 0-dimensional tensor with 1 element, basically transpose should do nothing
paddle.zeros([]).reshape([]) # should return 0-dimensional tensor with 1 element, basically reshape should do nothing
Other info / logs
Support for zero-dimensional tensor is a hard requirement for inclusion for einops support.
Currently paddlepaddle can't pass the tests due to this issue
3条答案
按热度按时间dba5bblo1#
您好,我们已经收到了您的问题,会安排技术人员尽快解答您的问题,请耐心等待。请您再次检查是否提供了清晰的问题描述、复现代码、环境&版本、报错信息等。同时,您也可以通过查看官网API文档、常见问题、历史Issue、AI社区来寻求解答。祝您生活愉快~
Hi! We've received your issue and please be patient to get responded. We will arrange technicians to answer your questions as soon as possible. Please make sure that you have posted enough message to demo your request. You may also check out the API,FAQ,Github Issue and AI community to get the answer.Have a nice day!
kxe2p93d2#
@arogozhnikov thanks for providing details about processing zero-dimensional tensor in transpose (reshape). We will upgrade the ability of processing zero-dimensional tensor in future. :)
pdtvr36n3#
Hi, @joey12300. I am also very interested in intergrade
paddle
toeinops
.paddle.zeros([]).transpose([])
seems fixed inpaddlepaddle==2.2.0
, butpaddle.zeros([]).reshape([])
still get error: