mirror of
https://github.com/RYDE-WORK/Langchain-Chatchat.git
synced 2026-01-28 01:33:17 +08:00
450 lines
19 KiB
Python
450 lines
19 KiB
Python
import os
|
||
import logging
|
||
import sys
|
||
from pathlib import Path
|
||
from typing import Any, Optional, List, Dict
|
||
|
||
from dataclasses import dataclass
|
||
|
||
sys.path.append(str(Path(__file__).parent))
|
||
import _core_config as core_config
|
||
|
||
logger = logging.getLogger()
|
||
|
||
|
||
class ConfigModel(core_config.Config):
|
||
DEFAULT_LLM_MODEL: Optional[str] = None
|
||
"""默认选用的 LLM 名称"""
|
||
DEFAULT_EMBEDDING_MODEL: Optional[str] = None
|
||
"""默认选用的 Embedding 名称"""
|
||
Agent_MODEL: Optional[str] = None
|
||
"""AgentLM模型的名称 (可以不指定,指定之后就锁定进入Agent之后的Chain的模型,不指定就是LLM_MODELS[0])"""
|
||
HISTORY_LEN: Optional[int] = None
|
||
"""历史对话轮数"""
|
||
MAX_TOKENS: Optional[int] = None
|
||
"""大模型最长支持的长度,如果不填写,则使用模型默认的最大长度,如果填写,则为用户设定的最大长度"""
|
||
TEMPERATURE: Optional[float] = None
|
||
"""LLM通用对话参数"""
|
||
SUPPORT_AGENT_MODELS: Optional[List[str]] = None
|
||
"""支持的Agent模型"""
|
||
LLM_MODEL_CONFIG: Optional[Dict[str, Dict[str, Any]]] = None
|
||
"""LLM模型配置,包括了不同模态初始化参数"""
|
||
MODEL_PLATFORMS: Optional[List[Dict[str, Any]]] = None
|
||
"""模型平台配置"""
|
||
MODEL_PROVIDERS_CFG_PATH_CONFIG: Optional[str] = None
|
||
"""模型平台配置文件路径"""
|
||
MODEL_PROVIDERS_CFG_HOST: Optional[str] = None
|
||
"""模型平台配置服务host"""
|
||
MODEL_PROVIDERS_CFG_PORT: Optional[int] = None
|
||
"""模型平台配置服务port"""
|
||
TOOL_CONFIG: Optional[Dict[str, Any]] = None
|
||
"""工具配置项"""
|
||
|
||
@classmethod
|
||
def class_name(cls) -> str:
|
||
return cls.__name__
|
||
|
||
def __str__(self):
|
||
return self.to_json()
|
||
|
||
|
||
@dataclass
|
||
class ConfigModelFactory(core_config.ConfigFactory[ConfigModel]):
|
||
"""ConfigModel工厂类"""
|
||
|
||
def __init__(self):
|
||
# 默认选用的 LLM 名称
|
||
self.DEFAULT_LLM_MODEL = "chatglm3-6b"
|
||
|
||
# 默认选用的 Embedding 名称
|
||
self.DEFAULT_EMBEDDING_MODEL = "bge-large-zh-v1.5"
|
||
|
||
# AgentLM模型的名称 (可以不指定,指定之后就锁定进入Agent之后的Chain的模型,不指定就是LLM_MODELS[0])
|
||
self.Agent_MODEL = None
|
||
|
||
# 历史对话轮数
|
||
self.HISTORY_LEN = 3
|
||
|
||
# 大模型最长支持的长度,如果不填写,则使用模型默认的最大长度,如果填写,则为用户设定的最大长度
|
||
self.MAX_TOKENS = None
|
||
|
||
# LLM通用对话参数
|
||
self.TEMPERATURE = 0.7
|
||
# TOP_P = 0.95 # ChatOpenAI暂不支持该参数
|
||
|
||
self.SUPPORT_AGENT_MODELS = [
|
||
"chatglm3-6b",
|
||
"openai-api",
|
||
"Qwen-14B-Chat",
|
||
"Qwen-7B-Chat",
|
||
"qwen-turbo",
|
||
]
|
||
|
||
self.MODEL_PROVIDERS_CFG_PATH_CONFIG = os.path.join(os.path.dirname(os.path.abspath(__file__)),
|
||
"model_providers.yaml")
|
||
self.MODEL_PROVIDERS_CFG_HOST = "127.0.0.1"
|
||
|
||
self.MODEL_PROVIDERS_CFG_PORT = 20000
|
||
|
||
self._init_llm_work_config()
|
||
|
||
def _init_llm_work_config(self):
|
||
"""初始化知识库runtime的一些配置"""
|
||
|
||
self.LLM_MODEL_CONFIG = {
|
||
# 意图识别不需要输出,模型后台知道就行
|
||
"preprocess_model": {
|
||
self.DEFAULT_LLM_MODEL: {
|
||
"temperature": 0.05,
|
||
"max_tokens": 4096,
|
||
"history_len": 100,
|
||
"prompt_name": "default",
|
||
"callbacks": False
|
||
},
|
||
},
|
||
"llm_model": {
|
||
self.DEFAULT_LLM_MODEL: {
|
||
"temperature": 0.9,
|
||
"max_tokens": 4096,
|
||
"history_len": 10,
|
||
"prompt_name": "default",
|
||
"callbacks": True
|
||
},
|
||
},
|
||
"action_model": {
|
||
self.DEFAULT_LLM_MODEL: {
|
||
"temperature": 0.01,
|
||
"max_tokens": 4096,
|
||
"prompt_name": "ChatGLM3",
|
||
"callbacks": True
|
||
},
|
||
},
|
||
"postprocess_model": {
|
||
self.DEFAULT_LLM_MODEL: {
|
||
"temperature": 0.01,
|
||
"max_tokens": 4096,
|
||
"prompt_name": "default",
|
||
"callbacks": True
|
||
}
|
||
},
|
||
"image_model": {
|
||
"sd-turbo": {
|
||
"size": "256*256",
|
||
}
|
||
}
|
||
}
|
||
|
||
# 可以通过 model_providers 提供转换不同平台的接口为openai endpoint的能力,启动后下面变量会自动增加相应的平台
|
||
# ### 如果您已经有了一个openai endpoint的能力的地址,可以在这里直接配置
|
||
# - platform_name 可以任意填写,不要重复即可
|
||
# - platform_type 以后可能根据平台类型做一些功能区分,与platform_name一致即可
|
||
# - 将框架部署的模型填写到对应列表即可。不同框架可以加载同名模型,项目会自动做负载均衡。
|
||
|
||
# 创建一个全局的共享字典
|
||
self.MODEL_PLATFORMS = [
|
||
|
||
{
|
||
"platform_name": "oneapi",
|
||
"platform_type": "oneapi",
|
||
"api_base_url": "http://127.0.0.1:3000/v1",
|
||
"api_key": "sk-",
|
||
"api_concurrencies": 5,
|
||
"llm_models": [
|
||
# 智谱 API
|
||
"chatglm_pro",
|
||
"chatglm_turbo",
|
||
"chatglm_std",
|
||
"chatglm_lite",
|
||
# 千问 API
|
||
"qwen-turbo",
|
||
"qwen-plus",
|
||
"qwen-max",
|
||
"qwen-max-longcontext",
|
||
# 千帆 API
|
||
"ERNIE-Bot",
|
||
"ERNIE-Bot-turbo",
|
||
"ERNIE-Bot-4",
|
||
# 星火 API
|
||
"SparkDesk",
|
||
],
|
||
"embed_models": [
|
||
# 千问 API
|
||
"text-embedding-v1",
|
||
# 千帆 API
|
||
"Embedding-V1",
|
||
],
|
||
"image_models": [],
|
||
"reranking_models": [],
|
||
"speech2text_models": [],
|
||
"tts_models": [],
|
||
},
|
||
|
||
{
|
||
"platform_name": "xinference",
|
||
"platform_type": "xinference",
|
||
"api_base_url": "http://127.0.0.1:9997/v1",
|
||
"api_key": "EMPTY",
|
||
"api_concurrencies": 5,
|
||
"llm_models": [
|
||
"glm-4",
|
||
"qwen2-instruct",
|
||
"qwen1.5-chat",
|
||
],
|
||
"embed_models": [
|
||
"bge-large-zh-v1.5",
|
||
],
|
||
"image_models": [],
|
||
"reranking_models": [],
|
||
"speech2text_models": [],
|
||
"tts_models": [],
|
||
},
|
||
|
||
]
|
||
# 工具配置项
|
||
self.TOOL_CONFIG = {
|
||
"search_local_knowledgebase": {
|
||
"use": False,
|
||
"top_k": 3,
|
||
"score_threshold": 1.0,
|
||
"conclude_prompt": {
|
||
"with_result":
|
||
'<指令>根据已知信息,简洁和专业的来回答问题。如果无法从中得到答案,请说 "根据已知信息无法回答该问题",'
|
||
'不允许在答案中添加编造成分,答案请使用中文。 </指令>\n'
|
||
'<已知信息>{{ context }}</已知信息>\n'
|
||
'<问题>{{ question }}</问题>\n',
|
||
"without_result":
|
||
'请你根据我的提问回答我的问题:\n'
|
||
'{{ question }}\n'
|
||
'请注意,你必须在回答结束后强调,你的回答是根据你的经验回答而不是参考资料回答的。\n',
|
||
}
|
||
},
|
||
"search_internet": {
|
||
"use": False,
|
||
"search_engine_name": "bing",
|
||
"search_engine_config":
|
||
{
|
||
"bing": {
|
||
"result_len": 3,
|
||
"bing_search_url": "https://api.bing.microsoft.com/v7.0/search",
|
||
"bing_key": "",
|
||
},
|
||
"metaphor": {
|
||
"result_len": 3,
|
||
"metaphor_api_key": "",
|
||
"split_result": False,
|
||
"chunk_size": 500,
|
||
"chunk_overlap": 0,
|
||
},
|
||
"duckduckgo": {
|
||
"result_len": 3
|
||
}
|
||
},
|
||
"top_k": 10,
|
||
"verbose": "Origin",
|
||
"conclude_prompt":
|
||
"<指令>这是搜索到的互联网信息,请你根据这些信息进行提取并有调理,简洁的回答问题。如果无法从中得到答案,请说 “无法搜索到能回答问题的内容”。 "
|
||
"</指令>\n<已知信息>{{ context }}</已知信息>\n"
|
||
"<问题>\n"
|
||
"{{ question }}\n"
|
||
"</问题>\n"
|
||
},
|
||
"arxiv": {
|
||
"use": False,
|
||
},
|
||
"shell": {
|
||
"use": False,
|
||
},
|
||
"weather_check": {
|
||
"use": False,
|
||
"api_key": "S8vrB4U_-c5mvAMiK",
|
||
},
|
||
"search_youtube": {
|
||
"use": False,
|
||
},
|
||
"wolfram": {
|
||
"use": False,
|
||
"appid": "",
|
||
},
|
||
"calculate": {
|
||
"use": False,
|
||
},
|
||
"vqa_processor": {
|
||
"use": False,
|
||
"model_path": "your model path",
|
||
"tokenizer_path": "your tokenizer path",
|
||
"device": "cuda:1"
|
||
},
|
||
"aqa_processor": {
|
||
"use": False,
|
||
"model_path": "your model path",
|
||
"tokenizer_path": "yout tokenizer path",
|
||
"device": "cuda:2"
|
||
},
|
||
"text2images": {
|
||
"use": False,
|
||
},
|
||
# text2sql使用建议
|
||
# 1、因大模型生成的sql可能与预期有偏差,请务必在测试环境中进行充分测试、评估;
|
||
# 2、生产环境中,对于查询操作,由于不确定查询效率,推荐数据库采用主从数据库架构,让text2sql连接从数据库,防止可能的慢查询影响主业务;
|
||
# 3、对于写操作应保持谨慎,如不需要写操作,设置read_only为True,最好再从数据库层面收回数据库用户的写权限,防止用户通过自然语言对数据库进行修改操作;
|
||
# 4、text2sql与大模型在意图理解、sql转换等方面的能力有关,可切换不同大模型进行测试;
|
||
# 5、数据库表名、字段名应与其实际作用保持一致、容易理解,且应对数据库表名、字段进行详细的备注说明,帮助大模型更好理解数据库结构;
|
||
# 6、若现有数据库表名难于让大模型理解,可配置下面table_comments字段,补充说明某些表的作用。
|
||
"text2sql": {
|
||
"use": False,
|
||
# SQLAlchemy连接字符串,支持的数据库有:
|
||
# crate、duckdb、googlesql、mssql、mysql、mariadb、oracle、postgresql、sqlite、clickhouse、prestodb
|
||
# 不同的数据库请查询SQLAlchemy,修改sqlalchemy_connect_str,配置对应的数据库连接,如sqlite为sqlite:///数据库文件路径,下面示例为mysql
|
||
# 如提示缺少对应数据库的驱动,请自行通过poetry安装
|
||
"sqlalchemy_connect_str": "mysql+pymysql://用户名:密码@主机地址/数据库名称e",
|
||
# 务必评估是否需要开启read_only,开启后会对sql语句进行检查,请确认text2sql.py中的intercept_sql拦截器是否满足你使用的数据库只读要求
|
||
# 优先推荐从数据库层面对用户权限进行限制
|
||
"read_only": False,
|
||
# 限定返回的行数
|
||
"top_k": 50,
|
||
# 是否返回中间步骤
|
||
"return_intermediate_steps": True,
|
||
# 如果想指定特定表,请填写表名称,如["sys_user","sys_dept"],不填写走智能判断应该使用哪些表
|
||
"table_names": [],
|
||
# 对表名进行额外说明,辅助大模型更好的判断应该使用哪些表,尤其是SQLDatabaseSequentialChain模式下,是根据表名做的预测,很容易误判。
|
||
"table_comments": {
|
||
# 如果出现大模型选错表的情况,可尝试根据实际情况填写表名和说明
|
||
# "tableA":"这是一个用户表,存储了用户的基本信息",
|
||
# "tanleB":"角色表",
|
||
}
|
||
},
|
||
}
|
||
|
||
def default_llm_model(self, llm_model: str):
|
||
self.DEFAULT_LLM_MODEL = llm_model
|
||
|
||
def default_embedding_model(self, embedding_model: str):
|
||
self.DEFAULT_EMBEDDING_MODEL = embedding_model
|
||
|
||
def agent_model(self, agent_model: str):
|
||
self.Agent_MODEL = agent_model
|
||
|
||
def history_len(self, history_len: int):
|
||
self.HISTORY_LEN = history_len
|
||
|
||
def max_tokens(self, max_tokens: int):
|
||
self.MAX_TOKENS = max_tokens
|
||
|
||
def temperature(self, temperature: float):
|
||
self.TEMPERATURE = temperature
|
||
|
||
def support_agent_models(self, support_agent_models: List[str]):
|
||
self.SUPPORT_AGENT_MODELS = support_agent_models
|
||
|
||
def model_providers_cfg_path_config(self, model_providers_cfg_path_config: str):
|
||
self.MODEL_PROVIDERS_CFG_PATH_CONFIG = model_providers_cfg_path_config
|
||
|
||
def model_providers_cfg_host(self, model_providers_cfg_host: str):
|
||
self.MODEL_PROVIDERS_CFG_HOST = model_providers_cfg_host
|
||
|
||
def model_providers_cfg_port(self, model_providers_cfg_port: int):
|
||
self.MODEL_PROVIDERS_CFG_PORT = model_providers_cfg_port
|
||
|
||
def get_config(self) -> ConfigModel:
|
||
config = ConfigModel()
|
||
config.DEFAULT_LLM_MODEL = self.DEFAULT_LLM_MODEL
|
||
config.DEFAULT_EMBEDDING_MODEL = self.DEFAULT_EMBEDDING_MODEL
|
||
config.Agent_MODEL = self.Agent_MODEL
|
||
config.HISTORY_LEN = self.HISTORY_LEN
|
||
config.MAX_TOKENS = self.MAX_TOKENS
|
||
config.TEMPERATURE = self.TEMPERATURE
|
||
config.SUPPORT_AGENT_MODELS = self.SUPPORT_AGENT_MODELS
|
||
config.LLM_MODEL_CONFIG = self.LLM_MODEL_CONFIG
|
||
config.MODEL_PLATFORMS = self.MODEL_PLATFORMS
|
||
config.MODEL_PROVIDERS_CFG_PATH_CONFIG = self.MODEL_PROVIDERS_CFG_PATH_CONFIG
|
||
config.MODEL_PROVIDERS_CFG_HOST = self.MODEL_PROVIDERS_CFG_HOST
|
||
config.MODEL_PROVIDERS_CFG_PORT = self.MODEL_PROVIDERS_CFG_PORT
|
||
config.TOOL_CONFIG = self.TOOL_CONFIG
|
||
|
||
return config
|
||
|
||
|
||
class ConfigModelWorkSpace(core_config.ConfigWorkSpace[ConfigModelFactory, ConfigModel]):
|
||
"""
|
||
工作空间的配置预设, 提供ConfigModel建造方法产生实例。
|
||
"""
|
||
config_factory_cls = ConfigModelFactory
|
||
|
||
def __init__(self):
|
||
super().__init__()
|
||
|
||
def _build_config_factory(self, config_json: Any) -> ConfigModelFactory:
|
||
|
||
_config_factory = self.config_factory_cls()
|
||
if config_json.get("DEFAULT_LLM_MODEL"):
|
||
_config_factory.default_llm_model(config_json.get("DEFAULT_LLM_MODEL"))
|
||
if config_json.get("DEFAULT_EMBEDDING_MODEL"):
|
||
_config_factory.default_embedding_model(config_json.get("DEFAULT_EMBEDDING_MODEL"))
|
||
if config_json.get("Agent_MODEL"):
|
||
_config_factory.agent_model(config_json.get("Agent_MODEL"))
|
||
if config_json.get("HISTORY_LEN"):
|
||
_config_factory.history_len(config_json.get("HISTORY_LEN"))
|
||
if config_json.get("MAX_TOKENS"):
|
||
_config_factory.max_tokens(config_json.get("MAX_TOKENS"))
|
||
if config_json.get("TEMPERATURE"):
|
||
_config_factory.temperature(config_json.get("TEMPERATURE"))
|
||
if config_json.get("SUPPORT_AGENT_MODELS"):
|
||
_config_factory.support_agent_models(config_json.get("SUPPORT_AGENT_MODELS"))
|
||
if config_json.get("MODEL_PROVIDERS_CFG_PATH_CONFIG"):
|
||
_config_factory.model_providers_cfg_path_config(config_json.get("MODEL_PROVIDERS_CFG_PATH_CONFIG"))
|
||
if config_json.get("MODEL_PROVIDERS_CFG_HOST"):
|
||
_config_factory.model_providers_cfg_host(config_json.get("MODEL_PROVIDERS_CFG_HOST"))
|
||
if config_json.get("MODEL_PROVIDERS_CFG_PORT"):
|
||
_config_factory.model_providers_cfg_port(config_json.get("MODEL_PROVIDERS_CFG_PORT"))
|
||
|
||
return _config_factory
|
||
|
||
@classmethod
|
||
def get_type(cls) -> str:
|
||
return ConfigModel.class_name()
|
||
|
||
def get_config(self) -> ConfigModel:
|
||
return self._config_factory.get_config()
|
||
|
||
def set_default_llm_model(self, llm_model: str):
|
||
self._config_factory.default_llm_model(llm_model)
|
||
self.store_config()
|
||
|
||
def set_default_embedding_model(self, embedding_model: str):
|
||
self._config_factory.default_embedding_model(embedding_model)
|
||
self.store_config()
|
||
|
||
def set_agent_model(self, agent_model: str):
|
||
self._config_factory.agent_model(agent_model)
|
||
self.store_config()
|
||
|
||
def set_history_len(self, history_len: int):
|
||
self._config_factory.history_len(history_len)
|
||
self.store_config()
|
||
|
||
def set_max_tokens(self, max_tokens: int):
|
||
self._config_factory.max_tokens(max_tokens)
|
||
self.store_config()
|
||
|
||
def set_temperature(self, temperature: float):
|
||
self._config_factory.temperature(temperature)
|
||
self.store_config()
|
||
|
||
def set_support_agent_models(self, support_agent_models: List[str]):
|
||
self._config_factory.support_agent_models(support_agent_models)
|
||
self.store_config()
|
||
|
||
def set_model_providers_cfg_path_config(self, model_providers_cfg_path_config: str):
|
||
self._config_factory.model_providers_cfg_path_config(model_providers_cfg_path_config)
|
||
self.store_config()
|
||
|
||
def set_model_providers_cfg_host(self, model_providers_cfg_host: str):
|
||
self._config_factory.model_providers_cfg_host(model_providers_cfg_host)
|
||
self.store_config()
|
||
|
||
def set_model_providers_cfg_port(self, model_providers_cfg_port: int):
|
||
self._config_factory.model_providers_cfg_port(model_providers_cfg_port)
|
||
self.store_config()
|
||
|
||
|
||
config_model_workspace: ConfigModelWorkSpace = ConfigModelWorkSpace() |