CircleCI tests improvement

[Didayolo]

E2E tests (#2042)

• E2E tests are failing on Django 4.2.0 pull request (Django to 4.2.0 #1959). We're remove it them but need to add them again.

• E2E tests are randomly failing (~ 1 time out of 4).

Podman test

• The tests only run Docker, not Podman (Podman updates & More Competition container security #1871). We need to add this.

More E2E tests

Add to the E2E tests:

• A Codabench results submission (Update E2E tests #2302)
• Another Codabench code submission, typically Mini-AutoML (Update E2E tests #2302)
• A failing submission (Update E2E tests #2302)
• Several submissions made to the same benchmark
• Add a "ingestion during scoring" test
• Multi-task tests are disabled on CI
• Test submission on custom queue
• All tests on codalab/codalab-legacy:py312? It is much lighter

E2E test files:
https://github.com/codalab/codabench/tree/develop/tests/test_files/competitions
https://github.com/codalab/codabench/tree/develop/tests/test_files/submissions

The test:
https://github.com/codalab/codabench/blob/develop/tests/test_submission.py

The templates:
https://github.com/codalab/competition-examples/

Improve E2E

• Wait until Finished or Failed, and then test if Finished. (Update E2E tests #2302)

This wait we do not wait in case of failure, and we have more informative tests.

[ObadaS]

The first and second points should be fixed after we merge #2042 (Playwright is ready to replace our Selenium tests, and it adds even more tests than before)

[Didayolo]

Current submissions tests:

test_submission.py::test_basic[firefox] PASSED                           [ 53%]
test_submission.py::test_v15[firefox] PASSED                             [ 61%]
test_submission.py::test_irisV15_code[firefox] PASSED                    [ 69%]
test_submission.py::test_irisV15_result[firefox] PASSED                  [ 76%]
test_submission.py::test_v18[firefox] PASSED                             [ 84%]
test_submission.py::test_v2_multiTask[firefox] SKIPPED (Works locall...) [ 92%]
test_submission.py::test_v2_multiTaskFactSheet[firefox] SKIPPED (Wor...) [100%]