[RL] logprob compute use the same method

[JunnYu]

Before submitting

• Lint code. If there are lint issues, please format the code first.

# Install and register `pre-commit` in the project folder
pip install pre-commit && pre-commit install

# Process previous code files separately
pre-commit run --file XXXX.py

• Add test cases into tests folder. If there are codecov issues, please add tests cases first.

PR types

Bug fixes

PR changes

Others

Description

问题：旧的代码在第一个step的时候，old_log_prob（actor计算时得到）、ref_log_prob（ref计算时得到）、log_prob(actor训练时候计算) 这三个值不一样，存在问题，需要修复，本PR进行了修复。
修复方法：

1. old_log_prob和ref_log_prob缺少了 amp_auto_cast包裹，而log_prob有amp_auto_cast包裹，因此需要添加这个操作。
2. 当开启fused head loss的时候，log_prob采用了切chunk以及fused head的形式进行计算，为了严格对齐，需要修改compute_log_prob函数，使用相同的实现方式。
3. 开启kl loss的时候在训练1000+步数的时候会出nan，这里定位到的原因是exp(delta)溢出了，因为delta = ref_log_prob - log_prob 当中包含了padding的token，而在padding token位置可能delta值比较大，大到比如30，40这么大。因此，我们需要先对ref_log_prob以及log_prob进行mask操作，随后进行相减然后求exp，在这种情况下，不会出现溢出问题了。

新增entropy loss的打印监控，用户可以监控entropy loss值

[paddle-bot]

Thanks for your contribution!

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 105 lines in your changes missing coverage. Please review.

Project coverage is 46.95%. Comparing base (6de50e6) to head (25825f9).
Report is 2 commits behind head on develop.

Files with missing lines	Patch %	Lines
paddlenlp/rl/trainer/actor_trainer.py	0.00%	86 Missing ⚠️
paddlenlp/rl/models/ppo_model_utils.py	0.00%	11 Missing ⚠️
paddlenlp/rl/trainer/ppo_trainer.py	0.00%	8 Missing ⚠️

❌ Your patch status has failed because the patch coverage (0.00%) is below the target coverage (80.00%). You can increase the patch coverage or adjust the target coverage.
❌ Your project status has failed because the head coverage (46.95%) is below the target coverage (58.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop   #10596      +/-   ##
===========================================
- Coverage    46.98%   46.95%   -0.04%     
===========================================
  Files          799      799              
  Lines       132255   132348      +93     
===========================================
  Hits         62139    62139              
- Misses       70116    70209      +93     

☔ 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.