chore(ollama,groq): Filtering Parameters in bind_tools for Ollama and Groq

[keenborder786]

• To prevent unsupported parameters from being passed to bind_tools, we now filter them out, which also resolves the error when using ProviderStrategy.
• Fixes Groq openai/gpt-oss-120b model incompatible with structured output (create_agent, ChatGroq.with_structured_output) #34155

[codspeed-hq]

CodSpeed Performance Report

Merging #34167 will not alter performance

_{Comparing keenborder786:filtering_extra_parameter (de477e8) with master (05ba853)}

Summary

✅ 1 untouched
⏩ 33 skipped¹

Footnotes

1. 33 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[keenborder786]

@ccurme is this looking good?

[ccurme]

Regarding Ollama (and possibly Groq): we should add handling for response_format, similar to what's done here for Anthropic:

langchain/libs/partners/anthropic/langchain_anthropic/chat_models.py

Lines 2189 to 2204 in 05ba853

	if "response_format" in payload:
	# response_format present when using agents.create_agent's ProviderStrategy
	# ---
	# ProviderStrategy converts to OpenAI-style format, which passes kwargs to
	# ChatAnthropic, ending up in our payload
	response_format = payload.pop("response_format")
	if (
	isinstance(response_format, dict)
	and response_format.get("type") == "json_schema"
	and "schema" in response_format.get("json_schema", {})
	):
	response_format = cast(dict, response_format["json_schema"]["schema"])
	# Convert OpenAI-style response_format to Anthropic's output_format
	payload["output_format"] = _convert_to_anthropic_output_format(
	response_format
	)

[ccurme]

Thanks!