升级langchain等依赖同时调整llm_config相关配置

Author: lightislost

.gitignore

@@ -14,4 +14,5 @@ setup_test.py
 build
 *egg-info
 dist
-.ipynb_checkpoints
+.ipynb_checkpoints
+zdatafront*

muagent/chat/search_chat.py

@@ -5,7 +5,7 @@
 from langchain.callbacks import AsyncIteratorCallbackHandler
 from langchain.utilities import BingSearchAPIWrapper, DuckDuckGoSearchAPIWrapper
 from langchain.prompts.chat import ChatPromptTemplate
-from langchain.docstore.document import Document
+from langchain_community.docstore.document import Document
 
 # from configs.model_config import (
 #     PROMPT_TEMPLATE, SEARCH_ENGINE_TOP_K, BING_SUBSCRIPTION_KEY, BING_SEARCH_URL,

muagent/codechat/code_search/code_search.py

@@ -15,7 +15,7 @@
 
 from muagent.codechat.code_search.cypher_generator import CypherGenerator
 from muagent.codechat.code_search.tagger import Tagger
-from muagent.embeddings.get_embedding import get_embedding
+from muagent.llm_models.get_embedding import get_embedding
 from muagent.llm_models.llm_config import LLMConfig, EmbedConfig
 
 

muagent/codechat/code_search/cypher_generator.py

@@ -5,7 +5,7 @@
 @time: 2023/11/24 上午10:17
 @desc:
 '''
-from langchain import PromptTemplate
+from langchain.prompts import PromptTemplate
 from loguru import logger
 
 from muagent.llm_models.openai_model import getChatModelFromConfig
@@ -14,7 +14,8 @@
 from langchain.schema import (
     HumanMessage,
 )
-from langchain.chains.graph_qa.prompts import NGQL_GENERATION_PROMPT, CYPHER_GENERATION_TEMPLATE
+# from langchain.chains.graph_qa.prompts import NGQL_GENERATION_PROMPT, CYPHER_GENERATION_TEMPLATE
+from langchain_community.chains.graph_qa.prompts import CYPHER_GENERATION_TEMPLATE
 
 schema = '''
 Node properties: [{'tag': 'package', 'properties': []}, {'tag': 'class', 'properties': []}, {'tag': 'method', 'properties': []}]

muagent/codechat/codebase_handler/code_importer.py

@@ -12,7 +12,7 @@
 
 from muagent.db_handler.graph_db_handler.nebula_handler import NebulaHandler
 from muagent.db_handler.vector_db_handler.chroma_handler import ChromaHandler
-from muagent.embeddings.get_embedding import get_embedding
+from muagent.llm_models.get_embedding import get_embedding
 from muagent.llm_models.llm_config import EmbedConfig
 
 

muagent/connector/memory_manager.py

@@ -6,14 +6,14 @@
 from loguru import logger
 import numpy as np
 
-from langchain.docstore.document import Document
+from langchain_community.docstore.document import Document
 
 
 from .schema import Memory, Message
 from muagent.service.service_factory import KBServiceFactory
 from muagent.llm_models import getChatModelFromConfig
 from muagent.llm_models.llm_config import EmbedConfig, LLMConfig
-from muagent.embeddings.utils import load_embeddings_from_path
+from muagent.retrieval.utils import load_embeddings_from_path
 from muagent.utils.common_utils import *
 from muagent.connector.configs.prompts import CONV_SUMMARY_PROMPT_SPEC
 from muagent.orm import table_init
@@ -489,7 +489,7 @@ def check_chat_index(self, chat_index: str):
 
 
 from muagent.utils.tbase_util import TbaseHandler
-from muagent.embeddings.get_embedding import get_embedding
+from muagent.llm_models.get_embedding import get_embedding
 from redis.commands.search.field import (
     TextField,
     NumericField,

muagent/embeddings/__init__.py renamed to muagent/llm_models/embeddings/__init__.py

@@ -8,8 +8,8 @@
 from loguru import logger
 
 # from configs.model_config import EMBEDDING_MODEL
-from muagent.embeddings.openai_embedding import OpenAIEmbedding
-from muagent.embeddings.huggingface_embedding import HFEmbedding
+from muagent.llm_models.openai_embedding import OpenAIEmbedding
+from muagent.llm_models.huggingface_embedding import HFEmbedding
 from muagent.llm_models.llm_config import EmbedConfig
 
 def get_embedding(
@@ -35,7 +35,7 @@ def get_embedding(
         oae = OpenAIEmbedding()
         emb_res = oae.get_emb(text_list)
     elif engine == 'model':
-        hfe = HFEmbedding(model_path, embedding_device)
+        hfe = HFEmbedding(model_path, embed_config.model_device)
         emb_res = hfe.get_emb(text_list)
 
     return emb_res

muagent/embeddings/get_embedding.py renamed to muagent/llm_models/get_embedding.py

@@ -8,7 +8,7 @@
 from loguru import logger
 # from configs.model_config import EMBEDDING_DEVICE
 # from configs.model_config import embedding_model_dict
-from muagent.embeddings.utils import load_embeddings, load_embeddings_from_path
+from muagent.retrieval.utils import load_embeddings, load_embeddings_from_path
 
 
 class HFEmbedding:

muagent/embeddings/huggingface_embedding.py renamed to muagent/llm_models/huggingface_embedding.py

@@ -11,6 +11,7 @@ class LLMConfig:
     def __init__(
             self,
             model_name: str = "gpt-3.5-turbo",
+            model_engine: str = "openai",
             temperature: float = 0.25, 
             stop: Union[List[str], str] = None,
             api_key: str = "",
@@ -19,12 +20,15 @@ def __init__(
             llm: LLM = None,
             **kwargs
         ):
-
+        # only support http connection with others
+        # llm_model init config
         self.model_name: str = model_name
+        self.model_engine: str = model_engine
         self.temperature: float = temperature
         self.stop: Union[List[str], str] = stop
         self.api_key: str = api_key
         self.api_base_url: str = api_base_url
+        # custom llm
         self.llm: LLM = llm
         # 
         self.check_config()
@@ -55,7 +59,7 @@ def __init__(
         self.model_device: str = model_device
         self.api_key: str = api_key
         self.api_base_url: str = api_base_url
-        # 
+        # custom embeddings
         self.langchain_embeddings = langchain_embeddings
         # 
         self.check_config()