Skip to content
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

CV2-5011 refactors for making alegre dual purpose on text encoding #103

Merged
merged 8 commits into from
Aug 16, 2024
Merged

CV2-5011 refactors for making alegre dual purpose on text encoding #103

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
3389935
CV2-5011 refactors for making alegre dual purpose on text encoding
DGaffney Aug 5, 2024
fb9db8b
resolve changes in tests
DGaffney Aug 5, 2024
cc051a8
refactor to accommodate erroring in transformers
DGaffney Aug 9, 2024
cc0a441
try moving case to audio model instead
DGaffney Aug 9, 2024
06395f9
remove bad imports, fix broken test cases
DGaffney Aug 9, 2024
a6f0083
Merge branch 'refs/heads/master' into cv2-5011-internal-refactor
ashkankzme Aug 14, 2024
5a25bd5
Merge remote-tracking branch 'origin/cv2-5011-internal-refactor' into…
ashkankzme Aug 14, 2024
24f3e86
Update test_queue.py
DGaffney Aug 15, 2024
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
2 changes: 1 addition & 1 deletion lib/model/generic_transformer.py
View file
Open in desktop
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I realize this file has existed for a while, but is there a reason we override the respond() function instead of process()?
this would make this model unique and different from other Presto endpoints, and disables cache and error handling bc the standard get_response() function won't be called from model.py.

But I may be missing something?

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes - the reason we override is because we want to, from the get-go, natively be able to process jobs in batch on transformers instead of just walking through items one-by-one like we do with the others. I'll look into a refactor to keep our caching and error handling working for text though - good catch

Original file line number Diff line number Diff line change
Expand Up @@ -33,4 +33,4 @@ def vectorize(self, texts: List[str]) -> List[List[float]]:
"""
Vectorize the text! Run as batch.
"""
return {"hash_value": self.model.encode(texts).tolist()}
return self.model.encode(texts).tolist()
2 changes: 1 addition & 1 deletion test/lib/model/test_fptg.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ def test_vectorize(self):
texts = [schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "Hello, how are you?"}, "model_name": "fptg__Model"}), schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "I'm doing great, thanks!"}, "model_name": "fptg__Model"})]
self.model.model = self.mock_model
self.model.model.encode = MagicMock(return_value=np.array([[4, 5, 6], [7, 8, 9]]))
vectors = self.model.vectorize(texts)["hash_value"]
vectors = self.model.vectorize(texts)
self.assertEqual(len(vectors), 2)
self.assertEqual(vectors[0], [4, 5, 6])
self.assertEqual(vectors[1], [7, 8, 9])
Expand Down
2 changes: 1 addition & 1 deletion test/lib/model/test_generic.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ def test_vectorize(self):
texts = [schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "Hello, how are you?"}, "model_name": "fptg__Model"}), schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "I'm doing great, thanks!"}, "model_name": "fptg__Model"})]
self.model.model = self.mock_model
self.model.model.encode = MagicMock(return_value=np.array([[4, 5, 6], [7, 8, 9]]))
vectors = self.model.vectorize(texts)["hash_value"]
vectors = self.model.vectorize(texts)
self.assertEqual(len(vectors), 2)
self.assertEqual(vectors[0], [4, 5, 6])
self.assertEqual(vectors[1], [7, 8, 9])
Expand Down
2 changes: 1 addition & 1 deletion test/lib/model/test_indian_sbert.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ def test_vectorize(self):
texts = [schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "Hello, how are you?"}, "model_name": "indian_sbert__Model"}), schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "I'm doing great, thanks!"}, "model_name": "indian_sbert__Model"})]
self.model.model = self.mock_model
self.model.model.encode = MagicMock(return_value=np.array([[4, 5, 6], [7, 8, 9]]))
vectors = self.model.vectorize(texts)["hash_value"]
vectors = self.model.vectorize(texts)
self.assertEqual(len(vectors), 2)
self.assertEqual(vectors[0], [4, 5, 6])
self.assertEqual(vectors[1], [7, 8, 9])
Expand Down
2 changes: 1 addition & 1 deletion test/lib/model/test_meantokens.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ def test_vectorize(self):
texts = [schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "Hello, how are you?"}, "model_name": "mean_tokens__Model"}), schemas.parse_message({"body": {"id": "123", "callback_url": "http://example.com/callback", "text": "I'm doing great, thanks!"}, "model_name": "mean_tokens__Model"})]
self.model.model = self.mock_model
self.model.model.encode = MagicMock(return_value=np.array([[4, 5, 6], [7, 8, 9]]))
vectors = self.model.vectorize(texts)["hash_value"]
vectors = self.model.vectorize(texts)
self.assertEqual(len(vectors), 2)
self.assertEqual(vectors[0], [4, 5, 6])
self.assertEqual(vectors[1], [7, 8, 9])
Expand Down
Loading