local-llm-server/llm_server/opts.py

# Read-only global variables

# TODO: rewrite the config system so I don't have to add every single config default here

running_model = 'none'
concurrent_gens = 3
mode = 'oobabooga'
backend_url = None
context_size = 5555
max_new_tokens = 500
database_path = './proxy-server.db'
auth_required = False
log_prompts = False
frontend_api_client = ''
base_client_api = None
http_host = None
verify_ssl = True
show_num_prompts = True
show_uptime = True
average_generation_time_mode = 'database'
show_total_output_tokens = True
netdata_root = None
simultaneous_requests_per_ip = 3
show_backend_info = True
manual_model_name = None
llm_middleware_name = ''
enable_openi_compatible_backend = True
openai_system_prompt = """You are an assistant chatbot. Your main function is to provide accurate and helpful responses to the user's queries. You should always be polite, respectful, and patient. You should not provide any personal opinions or advice unless specifically asked by the user. You should not make any assumptions about the user's knowledge or abilities. You should always strive to provide clear and concise answers. If you do not understand a user's query, ask for clarification. If you cannot provide an answer, apologize and suggest the user seek help elsewhere.\nLines that start with "### ASSISTANT" were messages you sent previously.\nLines that start with "### USER" were messages sent by the user you are chatting with.\nYou will respond to the "### RESPONSE:" prompt as the assistant and follow the instructions given by the user.\n\n"""
expose_openai_system_prompt = True
enable_streaming = True
openai_api_key = None
backend_request_timeout = 30
backend_generate_request_timeout = 95