-
Notifications
You must be signed in to change notification settings - Fork 62
Kanvi/bfloat16 tests #185
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
Kanvi/bfloat16 tests #185
Conversation
shresthamalik
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Requested some changes
| ng_outval = sess_ng.run(ng_out[0], feed_dict=feed_dict) | ||
|
|
||
| # transpose TF output from NHWC to NCHW for comparison with ngraph output | ||
| assert (np.allclose( |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do we care about only the 1st output of the op?
remove fusedbatchnorm_inferece tests since those are not needed for resnet
shresthamalik
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
This PR implements the following new tests to test the bfloat16 conversions test_fusedbatchnorm_training_nhwc.py test_fusedbatchnorm_training_nchw.py test_maxpoolbackprop_nhwc.py test_maxpoolbackprop_nhwc.py The following existing tests are also fixed as part of this PR test_fusedbatchnorm.py test_l2loss.py test_maxpoolbackprop.py test_relugrad.py test_sparse_softmax_cross_entropy_with_logits.py Also, made changes to the cmake so the new folder is included.
This PR implements the following new tests to test the bfloat16 conversions
The following existing tests are also fixed as part of this PR
Also, made changes to the cmake so the new folder is included.