from chatbug.inference import Inference
from chatbug.modelconfig import Modelconfig


def main():
    # Model size: 3.21B params
    Inference(Modelconfig("NousResearch/Hermes-3-Llama-3.2-3B", load_in_8bit=True))

    # Model size: 1.24B params
    Inference(Modelconfig("unsloth/Llama-3.2-1B", load_in_8bit=True))

    # Model size: 3.21B params
    Inference(Modelconfig("unsloth/Llama-3.2-3B-Instruct", load_in_8bit=True))

    # Model size: 4.65B params
    Inference(Modelconfig("unsloth/llama-3-8b-bnb-4bit", load_in_4bit=True))

    # Model size: 3.21B params
    Inference(Modelconfig("unsloth/Llama-3.2-3B-Instruct-GGUF", load_in_4bit=True))

    # Model size: 5.21B params
    Inference(Modelconfig("unsloth/gemma-2-9b-it-bnb-4bit", load_in_4bit=True))

    # Model size: 4.46B params
    Inference(Modelconfig("unsloth/Qwen2.5-7B-Instruct-bnb-4bit", load_in_4bit=True))

    # Model size: 3.09B params
    Inference(Modelconfig("unsloth/Qwen2.5-3B-Instruct", load_in_4bit=True))

    # Model size: 3.87B params
    Inference(Modelconfig("unsloth/mistral-7b-instruct-v0.3-bnb-4bit", load_in_4bit=True))


if __name__ == "__main__":
    main()