feature(nyz&dcy): add LLM/VLM reward model

[PaParaZz1]

Description

• math rule reward model
• math reward model like Qwen/Qwen2.5-Math-PRM-7B
• VLM reward model

Related Issue

#548

TODO

Check List

• merge the latest version source branch/repo, and resolve all the conflicts
• pass style check
• pass all the tests

[codecov]

Codecov Report

❌ Patch coverage is 18.73990% with 503 lines in your changes missing coverage. Please review.
✅ Project coverage is 74.82%. Comparing base (8f48cb1) to head (1cd74e5).
⚠️ Report is 14 commits behind head on main.

Files with missing lines	Patch %	Lines
ding/reward_model/math_rule_reward_model.py	10.00%	333 Missing ⚠️
...eward_model/tests/test_multi_modal_reward_model.py	34.37%	42 Missing ⚠️
ding/reward_model/math_reward_model.py	32.69%	35 Missing ⚠️
ding/reward_model/multi_modal_reward_model.py	31.25%	33 Missing ⚠️
ding/reward_model/tests/test_math_reward_model.py	17.50%	33 Missing ⚠️
.../reward_model/tests/test_math_rule_reward_model.py	35.71%	27 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #859      +/-   ##
==========================================
- Coverage   75.37%   74.82%   -0.56%     
==========================================
  Files         701      707       +6     
  Lines       57026    57955     +929     
==========================================
+ Hits        42982    43362     +380     
- Misses      14044    14593     +549     

Flag	Coverage Δ
unittests	74.82% <18.73%> (-0.54%)	⬇️

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.

[PaParaZz1]

English comments