src/generation.py

FIXED_GENERATION_CONFIG = dict(
    max_completion_tokens=1024,
    top_k=50,
    length_penalty=1.0,
    seed=42
)

MAX_AUDIO_LENGTH = 120



def prepare_multimodal_content(text_input, base64_audio_input):
    return [
        {
            "type": "text",
            "text": f"Text instruction: {text_input}"
        },
        {
            "type": "audio_url",
            "audio_url": {
                "url": f"data:audio/ogg;base64,{base64_audio_input}"
            },
        },
    ]


def change_multimodal_content(
        original_content, 
        text_input="", 
        base64_audio_input=""):
    
    # Since python 3.7 dictionary is ordered. 
    if text_input:
        original_content[0] = {
            "type": "text",
            "text": f"Text instruction: {text_input}"
        }

    if base64_audio_input:
        original_content[1] = {
            "type": "audio_url",
            "audio_url": {
                "url": f"data:audio/ogg;base64,{base64_audio_input}"
            }
        }

    return original_content