Skip to content

[CI]Add e2e test for 310p #1879

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

Jump to bottom
Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[CI]Add e2e test for 310p #1879

wants to merge 1 commit into from
+230 −2

Conversation

zhangxinyuehfad
Copy link
Contributor

@zhangxinyuehfad zhangxinyuehfad commented Jul 18, 2025 •
edited by github-actions bot
Loading

What this PR does / why we need it?

Add e2e test for 310p:
trigger conditions：tag, labels(ready-for-test, e2e-310p-test), schedule
image: m.daocloud.io/quay.io/ascend/cann:8.1.rc1-310p-ubuntu22.04-py3.10
runner: linux-aarch64-310p-1, linux-aarch64-310p-4
model: IntervitensInc/pangu-pro-moe-model, Qwen/Qwen3-0.6B-Base, Qwen/Qwen2.5-7B-Instruct

Does this PR introduce any user-facing change?

How was this patch tested?

@vllm-ascend-ci vllm-ascend-ci added ready-for-test start test by label for PR e2e-310p-test labels Jul 18, 2025
@codecov Codecov
Copy link

codecov bot commented Jul 18, 2025 •
edited
Loading

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.49%. Comparing base (ff97740) to head (c9039d5).
Report is 4 commits behind head on main.

Additional details and impacted files 
@@           Coverage Diff           @@
##             main    #1879   +/-   ##
=======================================
  Coverage   71.49%   71.49%           
=======================================
  Files          86       86           
  Lines        9131     9131           
=======================================
  Hits         6528     6528           
  Misses       2603     2603
Flag Coverage Δ
unittests 71.49% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@zhangxinyuehfad zhangxinyuehfad force-pushed the zxy_310p_ci branch 2 times, most recently from 4bc42e5 to 9a22d24 Compare July 20, 2025 04:39
@leo-pony
Copy link
Contributor

LGTM if CI passed

wangxiyuan
wangxiyuan reviewed Jul 21, 2025
View reviewed changes
tests/e2e/310p/test_offline_inference_multicard.py Outdated
@@ -0,0 +1,51 @@
#
Copy link
Collaborator

Choose a reason for hiding this comment

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

do not create a new forder. change to something like:
tests/e2e/singlecard/test_offiline_inference_310p.py

Copy link
Contributor Author

Choose a reason for hiding this comment

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

done

tests/e2e/310p/test_offline_inference_multicard.py Outdated
# See the License for the specific language governing permissions and
# limitations under the License.
# This file is a part of the vllm-ascend project.
# Adapted from vllm/tests/basic_correctness/test_basic_correctness.py
Copy link
Collaborator

Choose a reason for hiding this comment

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

useless 2 lines

Copy link
Contributor Author

Choose a reason for hiding this comment

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

done

tests/e2e/310p/test_offline_inference_multicard.py Outdated
#
"""Compare the short outputs of the Pangu (Ascend) model when using greedy sampling.

Run `pytest tests/e2e/test_offline_inference.py`.
Copy link
Collaborator

Choose a reason for hiding this comment

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

?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

delete it

@zhangxinyuehfad zhangxinyuehfad force-pushed the zxy_310p_ci branch 3 times, most recently from 67309fd to 63e626f Compare July 21, 2025 07:18
wangxiyuan
wangxiyuan reviewed Jul 22, 2025
View reviewed changes
.github/workflows/vllm_ascend_test_310p.yaml
@@ -0,0 +1,117 @@
#
Copy link
Collaborator

Choose a reason for hiding this comment

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

why create a new workflow? you can just add 2 jobs e2e-310p and e2e-4-cards-310p in vllm_ascend_test like other does.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

The startup method of this workflow is different with vllm_ascend_test , including label, schedule and tag.

Yikun
Yikun reviewed Jul 24, 2025
View reviewed changes
.github/workflows/vllm_ascend_test_310p.yaml Outdated
Comment on lines 60 to 61
# TODO(yikun): Remove m.daocloud.io prefix when infra proxy ready
image: m.daocloud.io/quay.io/ascend/cann:8.1.rc1-310p-ubuntu22.04-py3.10
Copy link
Collaborator

Choose a reason for hiding this comment

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

Change this according to: #1912

tests/e2e/multicard/test_offline_inference_310p.py Outdated
@pytest.mark.parametrize("model", MODELS)
@pytest.mark.parametrize("dtype", ["float16"])
@pytest.mark.parametrize("max_tokens", [5])
def test_pangu_model(model: str, dtype: str, max_tokens: int) -> None:
Copy link
Collaborator

Choose a reason for hiding this comment

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

It's better to also add torchair testcase

@Angazenn Please give some parameter examples here

Copy link
Contributor

Choose a reason for hiding this comment

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

sure, here is an example for pangu with torchair:

    additional_config = {
        "ascend_scheduler_config": {
            "enabled": True,
        },
        "torchair_graph_config": {
            "enabled": True,
        },
    }
    with VllmRunner(
            "vllm-ascend/pangu-pro-moe-pruing",
            dtype="half",
            tensor_parallel_size=4,
            distributed_executor_backend="mp",
            enforce_eager=False,
            additional_config=additional_config,
            enable_expert_parallel=True,
            compilation_config={
                "custom_ops":
               ["+unquantized_fused_moe"]
            }
    ) as vllm_model:

@zhangxinyuehfad
[CI]Add e2e test for 310p
c9039d5 
Signed-off-by: hfadzxy <starmoon_zhang@163.com>
@Yikun Yikun mentioned this pull request Jul 25, 2025
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@Yikun Yikun Yikun left review comments

@Angazenn Angazenn Angazenn left review comments

@wangxiyuan wangxiyuan wangxiyuan approved these changes

Assignees
No one assigned
Labels
e2e-310p-test module:tests ready-for-test start test by label for PR
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
6 participants
@zhangxinyuehfad @leo-pony @Yikun @wangxiyuan @Angazenn @vllm-ascend-ci