map max_tokens to max_completion_tokens internally

[asamal4]

Description

Use max_completion_tokens instead of deprecated max_tokens

Type of change

• Refactor
• New feature
• Bug fix
• CVE fix
• Optimization
• Documentation Update
• Configuration Update
• Bump-up service version
• Bump-up dependent library
• Bump-up library or tool used for development (does not change the final image)
• CI configuration change
• Unit tests improvement

Tools used to create PR

Identify any AI code assistants used in this PR (for transparency and review context)

• Assisted-by: N/A

Related Tickets & Documents

• Related Issue #
• Closes #

Checklist before requesting a review

• I have performed a self-review of my code.
• PR has passed all pre-merge test jobs.
• If it is a core feature, I have added thorough tests.

Testing

• Please provide detailed steps to perform tests related to this code change.
• How were the fix/results from this change verified? Please provide relevant screenshots or results.

Summary by CodeRabbit

• Refactor
  • Updated LLM parameter naming from "max_tokens" to "max_completion_tokens" across the system for improved API alignment.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Walkthrough

The parameter key for LLM maximum completion tokens is renamed from "max_tokens" to "max_completion_tokens" consistently across the custom LLM handler, DeepEval manager, base LLM manager, and all corresponding test files. No logic or control flow changes are introduced.

Changes

Cohort / File(s)	Summary
Core LLM Implementation src/lightspeed_evaluation/core/llm/custom.py, src/lightspeed_evaluation/core/llm/deepeval.py, src/lightspeed_evaluation/core/llm/manager.py	Parameter key renamed from max_tokens to max_completion_tokens in litellm.completion payload construction, LiteLLMModel initialization, and public get_llm_params()/get_model_info() return dictionaries.
Test Files tests/unit/core/llm/test_deepeval_manager.py, tests/unit/core/llm/test_llm_manager.py, tests/unit/core/llm/test_manager.py	Test fixtures and assertions updated to validate max_completion_tokens key instead of max_tokens in llm_params dictionaries and model info outputs.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• add common custom llm #70 — Main PR completing the LLM parameter key rename from "max_tokens" to "max_completion_tokens" across all LLM implementations and tests, directly addressing the parameter naming introduced in prior work on BaseCustomLLM and LLM parameter standardization.

Suggested reviewers

• VladimirKadlec
• tisnik

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main change: replacing the deprecated max_tokens parameter with max_completion_tokens across the LLM configuration code, which is the primary objective of the PR.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 0

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

src/lightspeed_evaluation/core/llm/deepeval.py (1)

19-26: Fix DeepEval LiteLLMModel parameter: use max_tokens instead of max_completion_tokens.

DeepEval's LiteLLMModel constructor accepts max_tokens, not max_completion_tokens. The current parameter is silently ignored and won't configure the token limit as intended.

Update line 23 to:

max_tokens=llm_params.get("max_completion_tokens"),

Also, replace the print() statement on line 28 with structured logging (required by coding guidelines for src/lightspeed_evaluation/**/*.py files).

🧹 Nitpick comments (1)

src/lightspeed_evaluation/core/llm/custom.py (1)

119-128: Parameter name is correct; consider optional backward compatibility fallback.
LiteLLM's completion API supports max_completion_tokens, so the current implementation is valid. However, if callers may pass max_tokens (older convention), the suggested fallback pattern provides safer compatibility:

💡 Optional backward compatibility improvement

-        "max_completion_tokens": self.llm_params.get("max_completion_tokens"),
+        "max_completion_tokens": self.llm_params.get(
+            "max_completion_tokens", self.llm_params.get("max_tokens")
+        ),

[VladimirKadlec]

LGTM