local-llm-server/llm_server/llm/generator.py

from llm_server import opts


def generator(request_json_body):
    if opts.mode == 'oobabooga':
        from .oobabooga.generate import generate
        return generate(request_json_body)
    elif opts.mode == 'vllm':
        from .vllm.generate import generate
        r = generate(request_json_body)
        return r
    else:
        raise Exception