SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

[tests] tag serve tests as slow (#39343)

Browse Source 
* maybe they need more cpu resources?

* add todo
This commit is contained in:
Joao Gante
2025-07-10 16:40:08 +01:00
committed by GitHub
parent 36e80a18da
commit df49b399dc
1 changed files with 2 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/commands/test_serving.py
View File

@@ -197,6 +197,7 @@ class ServeCompletionsMixin:
# TODO: speed-based test to confirm that KV cache is working across requests # TODO: speed-based test to confirm that KV cache is working across requests
@slow # TODO (joao): this shouldn't be needed
class ServeCompletionsGenerateTest(ServeCompletionsMixin, unittest.TestCase): class ServeCompletionsGenerateTest(ServeCompletionsMixin, unittest.TestCase):
"""Tests the `generate` version of the Completions API.""" """Tests the `generate` version of the Completions API."""
@@ -286,6 +287,7 @@ class ServeCompletionsGenerateTest(ServeCompletionsMixin, unittest.TestCase):
self.assertTrue(all(reason is None for reason in finish_reasons[:-1])) self.assertTrue(all(reason is None for reason in finish_reasons[:-1]))
@slow # TODO (joao): this shouldn't be needed
class ServeCompletionsContinuousBatchingTest(ServeCompletionsMixin, unittest.TestCase): class ServeCompletionsContinuousBatchingTest(ServeCompletionsMixin, unittest.TestCase):
"""Tests the `continuous_batching` version of the Completions API.""" """Tests the `continuous_batching` version of the Completions API."""