fix: pipecleaning 0.1.1 envs for rl

[fsiino-nvidia]

This PR contains infra + environment-specific changes made during pipecleaning efforts to get the respective environments to run successfully in nemo-rl.

rollout_collection:
Any server crash, network timeout, model error or resource exhaustion can cause a rollout to fail. Any single failed row kills the entire batch. We return a zero-reward fallback after waiting for ROLLOUT_ROW_TIMEOUT_SECONDS.

equivalence_llm_judge:
Previously any overly long generated answers from a thinking model would overflow the judge model's context window, crash the judge call and lead to cascading rollout failures. The implementation allows opt-in via max_judge_input_tokens and chars_per_token_estimate to truncate the generated answer before passing it to the judge model.

math_with_code:
Adds fallback for \boxed{} answer extraction. Previously only assistant text messages were searched for the answer. If a model were to answer via code execution, it ended up getting a zero reward. This fallback also searches function_call_output so an answer can be extracted from the tool output if the model prints it inside the executed code.

math_with_judge:
- Add config fields for judge truncation, similar to equivalence_llm_judge.

• Similar to math_with_code:
  • Send extracted answer parsed from \boxed{} to judge model instead of the raw generated answer to avoid overwhelming the judge's context window.
  • Expected answers wrapped in \(...\) produce \boxed{\42\)} which is not parsable by the math_verify library. This strips the outer delimiters to fix parsing.
• Prevent crash if the extracted answer is empty.

mcqa:
Address null option values in the dataset. Previously these were skipped which led to invalid rows being collected and training crashes.

mini_swe_agent:
Add a standalone data processing script to convert raw SWE-Bench/SWE-Gym data into the nemo-gym training format.

[bxyu-nvidia]

can we revert this pls given we don't want to artificially limit the rollout collection time?