Skip to content

Added integration tests for Pytorch implementation of the ALBERT model #9980

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.

Sign up for GitHub

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

Merged
LysandreJik merged 4 commits into from
Feb 3, 2021
Merged

Added integration tests for Pytorch implementation of the ALBERT model #9980

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
186556e
Merge pull request #16 from huggingface/master
spatil6 Jan 5, 2021
2dcf180
Merge pull request #21 from huggingface/master
spatil6 Feb 3, 2021
95bac7e
Alber model integration testing added
spatil6 Feb 3, 2021
61ff8e3
Merge branch 'MPNet_pt_test' of https://github.com/spatil6/transforme…
spatil6 Feb 3, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 16 additions & 0 deletions tests/test_modeling_albert.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -285,3 +285,19 @@ def test_model_from_pretrained(self):
for model_name in ALBERT_PRETRAINED_MODEL_ARCHIVE_LIST[:1]:
model = AlbertModel.from_pretrained(model_name)
self.assertIsNotNone(model)


@require_torch
class AlbertModelIntegrationTest(unittest.TestCase):
@slow
def test_inference_no_head_absolute_embedding(self):
model = AlbertForPreTraining.from_pretrained("albert-base-v2")
input_ids = torch.tensor([[0, 345, 232, 328, 740, 140, 1695, 69, 6078, 1588, 2]])
output = model(input_ids)[0]
expected_shape = torch.Size((1, 11, 30000))
self.assertEqual(output.shape, expected_shape)
expected_slice = torch.tensor(
[[[4.6061, 0.7321, -1.7725], [4.6061, 0.7323, -1.7727], [4.6061, 0.7323, -1.7727]]]
)

self.assertTrue(torch.allclose(output[:, :3, :3], expected_slice, atol=1e-4))