Whisper: Is it possible to retrieve segments like no_speech_prob/compression_ratio/avg_log and so on?

Beginners
1

What is up guys, I have just build an API using OpenAI’s Whisper v3-lg model.

My idea is to have a JSON response similar to Open AI’s one, where segment parameters are available to set and are returned as well.

So far my response is almost like I want it:

{
	"status": 200,
	"msg": "Áudio: ",
	"data": {
		"task": "transcription",
		"duration": 14.957833333333333,
		"language": "italian",
		"text": " Questo example..",
		"segments": [
			{
				"id": 0,
				"seek": 0,
				"start": 0.0,
				"end": 4.24,
				"text": "text1",
				"tokens": [
					50258,
					50364,
					10204,
					1106,
					281,
					1914
				]
			},
			{
				"id": 1,
				"seek": 0,
				"start": 4.5,
				"end": 10.12,
				"text": "Ancora..",
				"tokens": [
					50258,
					50364,
					1057,
					3252,
					1106,
					41031,
					947,
					50257
				]
			}
		]
	}
}

Is it possible to define and retrieve response values for parameters like: temperature,avg_logprob, compression_ratio and etc?

Appreciate any input