-
Notifications
You must be signed in to change notification settings - Fork 86
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ops test core dump #47
Comments
I have never tested it on 1.8, so it is possible this version introduced breaking C++ API changes. Unfortunately, I will not have time to look into this any time soon. |
Hi, what is the version of tf and cuda that you have tested? I make it to adapt tf2.1 ,but the gradient is not right.Thanks. |
@mygmyg Do you mean that tf2.1 produces incorrect results? I upgraded to tf1.14, also most of the tests gave 'core dump' errors but it still converges to reasonable output although I haven't checked if all values are correct. |
@eduardinjo I encountered |
I'm using 1.13. Not sure if is the same problem but I also met FATAL ERROR when using py.test to run the test.
I added if name == 'main' func in ops_test and used python to run it, which seemed to work. |
Hi:
I compiled the hdrnet ops under python 3.5.2 (virtualenv) tensorflow 1.8.0 CUDA 10 cuDNN 7.1.4 ubuntu 16.04. Then I test ops by test_ops.py. The first two tests are passed. But the third test 'test_grid_gradient' encountered a 'core dump' error. The detail infomation is as follows:
`test/ops_test.py Fatal Python error: Aborted
Thread 0x00007fde827c0700 (most recent call first):
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1409 in _call_tf_sessionrun
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1307 in _run_fn
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1322 in _do_call
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1316 in _do_run
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1135 in _run
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 900 in run
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/ops/gradient_checker.py", line 115 in _compute_theoretical_jacobian
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/ops/gradient_checker.py", line 237 in _compute_gradient
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/ops/gradient_checker.py", line 326 in compute_gradient
File "/data/mabing/PROGRAMFILES/tf18/lib/python3.5/site-packages/tensorflow/python/ops/gradient_checker.py", line 371 in compute_gradient_error
File "/data/mabing/CODES/HDRnets/tstops/test/ops_test.py", line 127 in test_grid_gradient
File "/usr/lib/python3.5/unittest/case.py", line 600 in run
File "/usr/lib/python3.5/unittest/case.py", line 648 in call`
I want to know if I should modify source code of bilateral slice to adapt to tf1.8 or just I tested it in a wrong way. Thanks in advance.
The text was updated successfully, but these errors were encountered: