feat: add api for knowledge_based QA

.gitignore

@@ -164,4 +164,7 @@ cython_debug/
 output/*
 log/*
 .chroma
-vector_store/*
\ No newline at end of file
+vector_store/*
+
+llm/*
+embedding/*
\ No newline at end of file

configs/model_config.py

@@ -6,10 +6,11 @@ embedding_model_dict = {
     "ernie-tiny": "nghuyong/ernie-3.0-nano-zh",
     "ernie-base": "nghuyong/ernie-3.0-base-zh",
     "text2vec": "GanymedeNil/text2vec-large-chinese",
+    "local-text2vec": "./embedding/text2vec-large-chinese"
 }
 
 # Embedding model name
-EMBEDDING_MODEL = "text2vec"
+EMBEDDING_MODEL = "local-text2vec"
 
 # Embedding running device
 EMBEDDING_DEVICE = "cuda" if torch.cuda.is_available() else "mps" if torch.backends.mps.is_available() else "cpu"
@@ -19,10 +20,11 @@ llm_model_dict = {
     "chatglm-6b-int4-qe": "THUDM/chatglm-6b-int4-qe",
     "chatglm-6b-int4": "THUDM/chatglm-6b-int4",
     "chatglm-6b": "THUDM/chatglm-6b",
+    "local-chatglm-6b": "./llm/chatglm-6b"
 }
 
 # LLM model name
-LLM_MODEL = "chatglm-6b"
+LLM_MODEL = "local-chatglm-6b"
 
 # Use p-tuning-v2 PrefixEncoder
 USE_PTUNING_V2 = False