mindspore-Ecosystem
/
mindspore
21
0
Fork 27
Code Issues Pull Requests 6 Projects Releases Wiki Activity

!19914 Update loss value in bert precision test case 

Merge pull request !19914 from chenhaozhe/code_docs_update_bert_loss_r1.3
This commit is contained in:
i-robot 2021-07-10 07:24:55 +00:00 committed by Gitee
parent 070307634b 8a6ad41721
commit c592b4bcb2
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
tests/st/networks/models/bert/bert_precision/test_bert_tdt_lossscale.py
View File

@ -229,8 +229,8 @@ def test_bert_precision(enable_graph_kernel=False):
12.185522, 12.386192]
else:
assert np.allclose(loss_value[0], 12.2066, 0, 0.0005)
expect_loss_value = [12.206587, 11.966410, 11.965916, 11.975922, 11.970262, 12.608881, 12.174048, 12.840656,
12.407923, 12.631133]
expect_loss_value = [12.206557, 11.940715, 11.930809, 11.937332, 11.931781, 12.555943, 12.130274, 12.783247,
12.359643, 12.578059]
print("loss value: {}".format(loss_value))
assert np.allclose(loss_value, expect_loss_value, 0, 0.0005)