* a draft of trainer.test()

* polish trainer.test()

* polish trainer.test()

* update code format

* update

* polish code

* polish code

* polish code

* Make trainer.test follow the rule of returning [loss, metric, metric, ..]

* Add FP16 option in save_combine_op

* Fix casting issue

Clean Parallel exe

Refine serde code

* add rdma train doc

* update by comment

* fix table

Rewrite Matmul, make code cleaner

add prependAllocatedop

Add float16 support to load op

fix tensorrt_engine compiler error

* "add float64 tests"

* "fix based comment"

* "fixed based comment"

Our tests interfere with each other and cause random failure.

save metrics data to csv file

Dockerfile: fix production dockefile when WITH_GOLANG=OFF.

decrease the pass number of test_label_semantic_roles

Revert "CI: rerun failed tests. (#10536)"

We should improve our test isolation before raising the
level of parallelism.

This reverts commit 0446220e.

Reason:

Rerun failed test hides flaky test.
Flaky test can be bugs, for example, race condition.
Test shouldn't be flaky, if a test is flaky, it should be fixed.

* CI: rerun failed tests.

* fix check style error.

* Put all new tests under high-level-api

* be consistent with the naming