[trainer] fix: Address feedback on critic update logic from commit ff…

[xingyunjohn1]

[trainer] fix: Address feedback on critic update logic from commit ff750e2

What does this PR do?

Thanks @vermouth1992 !
This PR addresses the feedback provided in commit ff750e2 regarding the actor-critic update logic in examples/split_placement/split_monkey_patch.py.

It ensures that the actor model is updated, and that this update occurs after the critic model. Additionally, the critic update is now wrapped in a conditional block to make its usage optional via the self.use_critic flag.

Checklist Before Starting

• Search for similar PRs. Paste at least one query link here: [trainer] fix: fix split placement #2227
• Format the PR title as [{modules}] {type}: {description} (This will be checked by the CI)
  • {modules} include fsdp, megatron, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data
  • If this PR involves multiple modules, separate them with , like [megatron, fsdp, doc]
  • {type} is in feat, fix, refactor, chore, test
  • If this PR breaks any API (CLI arguments, config, function signature, etc.), add [BREAKING] to the beginning of the title.
  • Example: [BREAKING][fsdp, megatron] feat: dynamic batching

Test

For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc.

API and Usage Example

Demonstrate how the API changes if any, and provide usage example(s) if possible.

# Add code snippet or script demonstrating how to use this

High-Level Design

Demonstrate the high-level design if this PR is complex.

Specific Changes

List the specific changes.

Checklist Before Submitting

Important

Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review.

• Read the Contribute Guide.
• Apply pre-commit checks: pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always
• Add / Update the documentation.
• Add unit or end-to-end test(s) to the CI workflow to cover all the code. If not feasible, explain why: ...
• Once your PR is ready for CI, send a message in the ci-request channel in the verl Slack workspace.

[vermouth1992]

This is still not correct. See #2227

[xingyunjohn1]

Thank you for pointing out the issues in my code.

After checking the original codebase, I can confirm that the condition if self.config.trainer.critic_warmup <= self.global_steps: should indeed be added at L188 to ensure the critic warmup is implemented.

I also noted that PR #2227 moved the actor update to occur before the critic update, which is inconsistent with the logic in verl/trainer/ppo/ray_trainer.py.