Move settings initialization after module init phase (#3438)
### What problem does this PR solve? 1. Module init won't connect database any more. 2. Config in settings need to be used with settings.CONFIG_NAME ### Type of change - [x] Refactoring Signed-off-by: jinhai <haijin.chn@gmail.com>
This commit is contained in:
Jin Hai
241
api/settings.py
241
api/settings.py
@@ -30,114 +30,157 @@ LIGHTEN = int(os.environ.get('LIGHTEN', "0"))
|
||||
|
||||
REQUEST_WAIT_SEC = 2
|
||||
REQUEST_MAX_WAIT_SEC = 300
|
||||
|
||||
LLM = get_base_config("user_default_llm", {})
|
||||
LLM_FACTORY = LLM.get("factory", "Tongyi-Qianwen")
|
||||
LLM_BASE_URL = LLM.get("base_url")
|
||||
|
||||
CHAT_MDL = EMBEDDING_MDL = RERANK_MDL = ASR_MDL = IMAGE2TEXT_MDL = ""
|
||||
if not LIGHTEN:
|
||||
default_llm = {
|
||||
"Tongyi-Qianwen": {
|
||||
"chat_model": "qwen-plus",
|
||||
"embedding_model": "text-embedding-v2",
|
||||
"image2text_model": "qwen-vl-max",
|
||||
"asr_model": "paraformer-realtime-8k-v1",
|
||||
},
|
||||
"OpenAI": {
|
||||
"chat_model": "gpt-3.5-turbo",
|
||||
"embedding_model": "text-embedding-ada-002",
|
||||
"image2text_model": "gpt-4-vision-preview",
|
||||
"asr_model": "whisper-1",
|
||||
},
|
||||
"Azure-OpenAI": {
|
||||
"chat_model": "gpt-35-turbo",
|
||||
"embedding_model": "text-embedding-ada-002",
|
||||
"image2text_model": "gpt-4-vision-preview",
|
||||
"asr_model": "whisper-1",
|
||||
},
|
||||
"ZHIPU-AI": {
|
||||
"chat_model": "glm-3-turbo",
|
||||
"embedding_model": "embedding-2",
|
||||
"image2text_model": "glm-4v",
|
||||
"asr_model": "",
|
||||
},
|
||||
"Ollama": {
|
||||
"chat_model": "qwen-14B-chat",
|
||||
"embedding_model": "flag-embedding",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"Moonshot": {
|
||||
"chat_model": "moonshot-v1-8k",
|
||||
"embedding_model": "",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"DeepSeek": {
|
||||
"chat_model": "deepseek-chat",
|
||||
"embedding_model": "",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"VolcEngine": {
|
||||
"chat_model": "",
|
||||
"embedding_model": "",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"BAAI": {
|
||||
"chat_model": "",
|
||||
"embedding_model": "BAAI/bge-large-zh-v1.5",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
"rerank_model": "BAAI/bge-reranker-v2-m3",
|
||||
}
|
||||
}
|
||||
|
||||
if LLM_FACTORY:
|
||||
CHAT_MDL = default_llm[LLM_FACTORY]["chat_model"] + f"@{LLM_FACTORY}"
|
||||
ASR_MDL = default_llm[LLM_FACTORY]["asr_model"] + f"@{LLM_FACTORY}"
|
||||
IMAGE2TEXT_MDL = default_llm[LLM_FACTORY]["image2text_model"] + f"@{LLM_FACTORY}"
|
||||
EMBEDDING_MDL = default_llm["BAAI"]["embedding_model"] + "@BAAI"
|
||||
RERANK_MDL = default_llm["BAAI"]["rerank_model"] + "@BAAI"
|
||||
|
||||
API_KEY = LLM.get("api_key", "")
|
||||
PARSERS = LLM.get(
|
||||
"parsers",
|
||||
"naive:General,qa:Q&A,resume:Resume,manual:Manual,table:Table,paper:Paper,book:Book,laws:Laws,presentation:Presentation,picture:Picture,one:One,audio:Audio,knowledge_graph:Knowledge Graph,email:Email")
|
||||
|
||||
HOST = get_base_config(RAG_FLOW_SERVICE_NAME, {}).get("host", "127.0.0.1")
|
||||
HTTP_PORT = get_base_config(RAG_FLOW_SERVICE_NAME, {}).get("http_port")
|
||||
|
||||
SECRET_KEY = get_base_config(
|
||||
RAG_FLOW_SERVICE_NAME,
|
||||
{}).get("secret_key", str(date.today()))
|
||||
LLM = None
|
||||
LLM_FACTORY = None
|
||||
LLM_BASE_URL = None
|
||||
CHAT_MDL = ""
|
||||
EMBEDDING_MDL = ""
|
||||
RERANK_MDL = ""
|
||||
ASR_MDL = ""
|
||||
IMAGE2TEXT_MDL = ""
|
||||
API_KEY = None
|
||||
PARSERS = None
|
||||
HOST_IP = None
|
||||
HOST_PORT = None
|
||||
SECRET_KEY = None
|
||||
|
||||
DATABASE_TYPE = os.getenv("DB_TYPE", 'mysql')
|
||||
DATABASE = decrypt_database_config(name=DATABASE_TYPE)
|
||||
|
||||
# authentication
|
||||
AUTHENTICATION_CONF = get_base_config("authentication", {})
|
||||
AUTHENTICATION_CONF = None
|
||||
|
||||
# client
|
||||
CLIENT_AUTHENTICATION = AUTHENTICATION_CONF.get(
|
||||
"client", {}).get(
|
||||
"switch", False)
|
||||
HTTP_APP_KEY = AUTHENTICATION_CONF.get("client", {}).get("http_app_key")
|
||||
GITHUB_OAUTH = get_base_config("oauth", {}).get("github")
|
||||
FEISHU_OAUTH = get_base_config("oauth", {}).get("feishu")
|
||||
CLIENT_AUTHENTICATION = None
|
||||
HTTP_APP_KEY = None
|
||||
GITHUB_OAUTH = None
|
||||
FEISHU_OAUTH = None
|
||||
|
||||
DOC_ENGINE = os.environ.get('DOC_ENGINE', "elasticsearch")
|
||||
if DOC_ENGINE == "elasticsearch":
|
||||
docStoreConn = rag.utils.es_conn.ESConnection()
|
||||
elif DOC_ENGINE == "infinity":
|
||||
docStoreConn = rag.utils.infinity_conn.InfinityConnection()
|
||||
else:
|
||||
raise Exception(f"Not supported doc engine: {DOC_ENGINE}")
|
||||
DOC_ENGINE = None
|
||||
docStoreConn = None
|
||||
|
||||
retrievaler = search.Dealer(docStoreConn)
|
||||
kg_retrievaler = kg_search.KGSearch(docStoreConn)
|
||||
retrievaler = None
|
||||
kg_retrievaler = None
|
||||
|
||||
|
||||
def init_settings():
|
||||
global LLM, LLM_FACTORY, LLM_BASE_URL
|
||||
LLM = get_base_config("user_default_llm", {})
|
||||
LLM_FACTORY = LLM.get("factory", "Tongyi-Qianwen")
|
||||
LLM_BASE_URL = LLM.get("base_url")
|
||||
|
||||
global CHAT_MDL, EMBEDDING_MDL, RERANK_MDL, ASR_MDL, IMAGE2TEXT_MDL
|
||||
if not LIGHTEN:
|
||||
default_llm = {
|
||||
"Tongyi-Qianwen": {
|
||||
"chat_model": "qwen-plus",
|
||||
"embedding_model": "text-embedding-v2",
|
||||
"image2text_model": "qwen-vl-max",
|
||||
"asr_model": "paraformer-realtime-8k-v1",
|
||||
},
|
||||
"OpenAI": {
|
||||
"chat_model": "gpt-3.5-turbo",
|
||||
"embedding_model": "text-embedding-ada-002",
|
||||
"image2text_model": "gpt-4-vision-preview",
|
||||
"asr_model": "whisper-1",
|
||||
},
|
||||
"Azure-OpenAI": {
|
||||
"chat_model": "gpt-35-turbo",
|
||||
"embedding_model": "text-embedding-ada-002",
|
||||
"image2text_model": "gpt-4-vision-preview",
|
||||
"asr_model": "whisper-1",
|
||||
},
|
||||
"ZHIPU-AI": {
|
||||
"chat_model": "glm-3-turbo",
|
||||
"embedding_model": "embedding-2",
|
||||
"image2text_model": "glm-4v",
|
||||
"asr_model": "",
|
||||
},
|
||||
"Ollama": {
|
||||
"chat_model": "qwen-14B-chat",
|
||||
"embedding_model": "flag-embedding",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"Moonshot": {
|
||||
"chat_model": "moonshot-v1-8k",
|
||||
"embedding_model": "",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"DeepSeek": {
|
||||
"chat_model": "deepseek-chat",
|
||||
"embedding_model": "",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"VolcEngine": {
|
||||
"chat_model": "",
|
||||
"embedding_model": "",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
},
|
||||
"BAAI": {
|
||||
"chat_model": "",
|
||||
"embedding_model": "BAAI/bge-large-zh-v1.5",
|
||||
"image2text_model": "",
|
||||
"asr_model": "",
|
||||
"rerank_model": "BAAI/bge-reranker-v2-m3",
|
||||
}
|
||||
}
|
||||
|
||||
if LLM_FACTORY:
|
||||
CHAT_MDL = default_llm[LLM_FACTORY]["chat_model"] + f"@{LLM_FACTORY}"
|
||||
ASR_MDL = default_llm[LLM_FACTORY]["asr_model"] + f"@{LLM_FACTORY}"
|
||||
IMAGE2TEXT_MDL = default_llm[LLM_FACTORY]["image2text_model"] + f"@{LLM_FACTORY}"
|
||||
EMBEDDING_MDL = default_llm["BAAI"]["embedding_model"] + "@BAAI"
|
||||
RERANK_MDL = default_llm["BAAI"]["rerank_model"] + "@BAAI"
|
||||
|
||||
global API_KEY, PARSERS, HOST_IP, HOST_PORT, SECRET_KEY
|
||||
API_KEY = LLM.get("api_key", "")
|
||||
PARSERS = LLM.get(
|
||||
"parsers",
|
||||
"naive:General,qa:Q&A,resume:Resume,manual:Manual,table:Table,paper:Paper,book:Book,laws:Laws,presentation:Presentation,picture:Picture,one:One,audio:Audio,knowledge_graph:Knowledge Graph,email:Email")
|
||||
|
||||
HOST_IP = get_base_config(RAG_FLOW_SERVICE_NAME, {}).get("host", "127.0.0.1")
|
||||
HOST_PORT = get_base_config(RAG_FLOW_SERVICE_NAME, {}).get("http_port")
|
||||
|
||||
SECRET_KEY = get_base_config(
|
||||
RAG_FLOW_SERVICE_NAME,
|
||||
{}).get("secret_key", str(date.today()))
|
||||
|
||||
global AUTHENTICATION_CONF, CLIENT_AUTHENTICATION, HTTP_APP_KEY, GITHUB_OAUTH, FEISHU_OAUTH
|
||||
# authentication
|
||||
AUTHENTICATION_CONF = get_base_config("authentication", {})
|
||||
|
||||
# client
|
||||
CLIENT_AUTHENTICATION = AUTHENTICATION_CONF.get(
|
||||
"client", {}).get(
|
||||
"switch", False)
|
||||
HTTP_APP_KEY = AUTHENTICATION_CONF.get("client", {}).get("http_app_key")
|
||||
GITHUB_OAUTH = get_base_config("oauth", {}).get("github")
|
||||
FEISHU_OAUTH = get_base_config("oauth", {}).get("feishu")
|
||||
|
||||
global DOC_ENGINE, docStoreConn, retrievaler, kg_retrievaler
|
||||
DOC_ENGINE = os.environ.get('DOC_ENGINE', "elasticsearch")
|
||||
if DOC_ENGINE == "elasticsearch":
|
||||
docStoreConn = rag.utils.es_conn.ESConnection()
|
||||
elif DOC_ENGINE == "infinity":
|
||||
docStoreConn = rag.utils.infinity_conn.InfinityConnection()
|
||||
else:
|
||||
raise Exception(f"Not supported doc engine: {DOC_ENGINE}")
|
||||
|
||||
retrievaler = search.Dealer(docStoreConn)
|
||||
kg_retrievaler = kg_search.KGSearch(docStoreConn)
|
||||
|
||||
def get_host_ip():
|
||||
global HOST_IP
|
||||
return HOST_IP
|
||||
|
||||
|
||||
def get_host_port():
|
||||
global HOST_PORT
|
||||
return HOST_PORT
|
||||
|
||||
|
||||
class CustomEnum(Enum):
|
||||
|
||||
Reference in New Issue
Block a user