Describe the change
Openai has by default a parallel_tool_calls set to true, which allows the llm to make multiple tool calls in one response.

This can be opted out of with setting it to false, and ensures the llm only responds with one tool call at a time.

Provide OpenAI documentation link
https://platform.openai.com/docs/guides/function-calling/parallel-function-calling

Describe your solution
We only need to set the field to false, and true for being certain it is true, we can not just use a bool type.
As that would require a custom marshaller for not including it when not set.

I think a *bool works oke here, it is a bit annoying to set a *bool value. Which is why the helper functions have been given: ParallelOptionFalse() and ParallelOptionTrue().

Intended use:

request := openai.ChatCompletionRequest{}
request.Model = completeChatModel
	
request.ParallelToolCalls = openai.ParallelOptionFalse()	

This seemed the best option without resorting to custom types with custom marshallers, or creating a custom marshal for ChatCompletionRequest

Tests
Currently using a version with bool type and no omitempty which has stopped gpt responding with multiple function calls.