fix(hybrid optim): fp32_grad not scaled when use offload_cpu

[fengsibo]

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

Bug fix: flat_fp32_avg_grads will not be scaled when hybridzerooptimizer use cpu_offload, detail see
#398.

Modification

Append truly used grad into single_grad_partition_groups, thus grad can be scaled in self._unscale_and_clip_grads whether on CPU or GPU.

BC-breaking (Optional)

Does the modification introduce changes that break the backward compatibility of the downstream repositories?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here and update the documentation.

Checklist

Before PR:

• Pre-commit or other linting tools are used to fix the potential lint issues.
• Bug fixes are fully covered by unit tests, the case that causes the bug should be added in the unit tests.
• The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
• The documentation has been modified accordingly, like docstring or example tutorials.

After PR:

• If the modification has potential influence on downstream or other related projects, this PR should be tested with those projects.
• CLA has been signed and all committers have signed the CLA in this PR.

[li126com]

看了应该没问题，flat_fp32_avg_grads.to(device)返回的是一个新的tensor，两边要对齐。但是格式检查还是要过一下的pre-commit run --files {code_path}