error with check.py

[jih189]

PyTorch GitHub Issues Guidelines

• OS: Ubuntu 18.04
• PyTorch version: 1.4.0
• How you installed PyTorch (conda, pip, source): conda
• Python version: 3.6
• CUDA/cuDNN version:10.0
• GPU models and configuration: Geforce RTX 2080 ti
• GCC version (if compiling from source): 7.5.0

when I run python check.py forward --cuda,
I got following error

Forward: Baseline (Python) vs. C++ ... Ok
Forward: Baseline (Python) vs. CUDA ... Traceback (most recent call last):
File "check.py", line 104, in
check_forward(variables, options.cuda, options.verbose)
File "check.py", line 45, in check_forward
check_equal(baseline_values, cuda_values, verbose)
File "check.py", line 22, in check_equal
np.testing.assert_allclose(x, y, err_msg="Index: {}".format(i))
File "/home/cogrob-wrc/.local/lib/python3.6/site-packages/numpy/testing/_private/utils.py", line 1533, in assert_allclose
verbose=verbose, header=header, equal_nan=equal_nan)
File "/home/cogrob-wrc/.local/lib/python3.6/site-packages/numpy/testing/_private/utils.py", line 846, in assert_array_compare
raise AssertionError(msg)
AssertionError:
Not equal to tolerance rtol=1e-07, atol=0
Index: 0
Mismatched elements: 15 / 15 (100%)
Max absolute difference: 0.91315287
Max relative difference: inf
x: array([[-0.913153, 0.066085, -0.006637, -0.306991, 0.552139],
[-0.28065 , -0.752348, 0.705655, 0.089504, 0.06883 ],
[-0.200587, 0.048015, 0.003387, 0.100579, -0.40925 ]])
y: array([[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.]])
[1] 4035 segmentation fault (core dumped) python check.py forward --cuda

need help!!!!