fix: properly passing in the parameters from kwargs/smss into model l…

[themaherkhalil]

…imits calculation

[github-actions]

@CodiumAI-Agent /describe

[QodoAI-Agent]

Title

fix: properly passing in the parameters from kwargs/smss into model l…

---

User description

…imits calculation

---

PR Type

Bug fix

---

Description

• Deprecate max_new_tokens & max_completion_tokens args

• Comment out unused signature parameters

• Correct max_completion_tokens_smss source attribute

---

Changes walkthrough 📝

	Relevant files
Bug fix	chat.py Correct token limit calculation and signature                        py/genai_client/text_generation/openai_clients/operations/chat.py Comment out deprecated signature parameters in ask Pop and normalize kwargs unchanged Use tokenizer.max_completion_tokens for smss limit +3/-3

---

Need help?

Type /help how to ... in the comments thread for any questions about PR-Agent usage.

Check out the documentation for more information.

[github-actions]

@CodiumAI-Agent /review

[QodoAI-Agent]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Recommended focus areas for review Dead Commented Parameters The commented-out max_new_tokens and max_completion_tokens parameters add maintenance overhead and confusion. Consider removing or converting them to proper deprecation warnings or documentation. # max_new_tokens=None, # Deprecated # We dont use either of these?? I think they are passed in the kwargs # max_completion_tokens=None, # We dont use either of these?? I think they are passed in the kwargs Attribute Availability The new self.client.tokenizer.max_completion_tokens property must exist on the tokenizer. Verify the attribute is defined and handled in all supported tokenizer implementations to avoid runtime errors. max_completion_tokens_smss=self.client.tokenizer.max_completion_tokens,

[github-actions]

@CodiumAI-Agent /improve

[QodoAI-Agent]

PR Code Suggestions ✨

Category	Suggestion	Impact
Possible issue	Fallback and bound completion tokens Ensure you don’t assume the tokenizer always has a max_completion_tokens attribute and that it never exceeds max_tokens. Use getattr with a fallback and wrap in min() to enforce the upper bound. py/genai_client/text_generation/openai_clients/operations/chat.py [55] -max_completion_tokens_smss=self.client.tokenizer.max_completion_tokens, +max_completion_tokens_smss=min( + getattr(self.client.tokenizer, "max_completion_tokens", self.client.tokenizer.max_tokens), + self.client.tokenizer.max_tokens +), Suggestion importance[1-10]: 7 __ Why: The change adds a safe fallback with getattr and caps max_completion_tokens_smss using min(), improving robustness and preventing attribute errors.	Medium
General	Handle deprecated token alias Provide backward-compatibility for callers using the deprecated max_new_tokens key by falling back to it when max_completion_tokens is not present. py/genai_client/text_generation/openai_clients/operations/chat.py [53] -max_completion_tokens_call_param=kwargs.pop("max_completion_tokens", None), +max_completion_tokens_call_param=( + kwargs.pop("max_completion_tokens", None) + or kwargs.pop("max_new_tokens", None) +), Suggestion importance[1-10]: 6 __ Why: Introducing a fallback to max_new_tokens ensures backward compatibility for callers still using the deprecated key.	Low

[github-actions]

@CodiumAI-Agent /update_changelog