diff --git a/.gitignore b/.gitignore index af50500..b5918ee 100644 --- a/.gitignore +++ b/.gitignore @@ -4,4 +4,4 @@ logs .idea/ __pycache__/ knowledge_base/ -configs/model_config.py \ No newline at end of file +configs/*.py diff --git a/README.md b/README.md index 66c0c66..75507b8 100644 --- a/README.md +++ b/README.md @@ -181,9 +181,11 @@ $ git clone https://huggingface.co/moka-ai/m3e-base ### 3. 设置配置项 -复制文件 [configs/model_config.py.example](configs/model_config.py.example) 存储至项目路径下 `./configs` 路径下，并重命名为 `model_config.py`。 +复制模型相关参数配置模板文件 [configs/model_config.py.example](configs/model_config.py.example) 存储至项目路径下 `./configs` 路径下，并重命名为 `model_config.py`。 -在开始执行 Web UI 或命令行交互前，请先检查 `configs/model_config.py` 中的各项模型参数设计是否符合需求： +复制服务相关参数配置模板文件 [configs/server_config.py.example](configs/server_config.py.example) 存储至项目路径下 `./configs` 路径下，并重命名为 `server_config.py`。 + +在开始执行 Web UI 或命令行交互前，请先检查 `configs/model_config.py` 和 `configs/server_config.py` 中的各项模型参数设计是否符合需求： - 请确认已下载至本地的 LLM 模型本地存储路径写在 `llm_model_dict` 对应模型的 `local_model_path` 属性中，如: @@ -214,7 +216,6 @@ embedding_model_dict = { ```shell $ python init_database.py ``` - - 如果您是第一次运行本项目，知识库尚未建立，或者配置文件中的知识库类型、嵌入模型发生变化，需要以下命令初始化或重建知识库： ```shell @@ -244,6 +245,7 @@ $ python server/llm_api.py ``` 项目支持多卡加载，需在 llm_api.py 中修改 create_model_worker_app 函数中，修改如下三个参数: + ```python gpus=None, num_gpus=1, @@ -258,7 +260,7 @@ max_gpu_memory="20GiB" ##### 5.1.2 基于命令行脚本 llm_api_launch.py 启动 LLM 服务 -⚠️ **注意:** +⚠️ **注意:** **1.llm_api_launch.py脚本原生仅适用于linux,mac设备需要安装对应的linux命令,win平台请使用wls;** @@ -275,11 +277,13 @@ $ python server/llm_api_launch.py ```shell $ python server/llm_api_launch.py --model-path-address model1@host1@port1 model2@host2@port2 ``` + 如果出现server端口占用情况，需手动指定server端口,并同步修改model_config.py下对应模型的base_api_url为指定端口: ```shell $ python server/llm_api_launch.py --server-port 8887 ``` + 如果要启动多卡加载，示例命令如下： ```shell @@ -354,7 +358,6 @@ $ streamlit run webui.py --server.port 666 - Web UI 对话界面： ![](img/webui_0813_0.png) - - Web UI 知识库管理页面： ![](img/webui_0813_1.png) @@ -363,86 +366,38 @@ $ streamlit run webui.py --server.port 666 ### 6. 一键启动 -⚠️ **注意:** - -**1. 一键启动脚本仅原生适用于Linux,Mac 设备需要安装对应的linux命令, Winodws 平台请使用 WLS;** - -**2. 加载非默认模型需要用命令行参数 `--model-path-address` 指定模型，不会读取 `model_config.py` 配置。** - -#### 6.1 API 服务一键启动脚本 - -新增 API 一键启动脚本，可一键开启 FastChat 后台服务及本项目提供的 API 服务,调用示例： - -调用默认模型： +更新一键启动脚本 startup.py,一键启动所有 Fastchat 服务、API 服务、WebUI 服务，示例代码： ```shell -$ python server/api_allinone.py +$ python startup.py --all-webui ``` -加载多个非默认模型： +并可使用 `Ctrl + C` 直接关闭所有运行服务。 + +可选参数包括 `--all-webui`, `--all-api`, `--llm-api`, `--controller`, `--openai-api`, +`--model-worker`, `--api`, `--webui`，其中： + +- `--all-webui` 为一键启动 WebUI 所有依赖服务； + +- `--all-api` 为一键启动 API 所有依赖服务； + +- `--llm-api` 为一键启动 Fastchat 所有依赖的 LLM 服务； + +- `--openai-api` 为仅启动 FastChat 的 controller 和 openai-api-server 服务； + +- 其他为单独服务启动选项。 + +若想指定非默认模型，需要用 `--model-name` 选项，示例： ```shell -$ python server/api_allinone.py --model-path-address model1@host1@port1 model2@host2@port2 +$ python startup.py --all-webui --model-name Qwen-7B-Chat ``` -如果出现server端口占用情况，需手动指定server端口,并同步修改model_config.py下对应模型的base_api_url为指定端口: +**注意：** -```shell -$ python server/api_allinone.py --server-port 8887 -``` +**1. startup 脚本用多进程方式启动各模块的服务，可能会导致打印顺序问题，请等待全部服务发起后再调用，并根据默认或指定端口调用服务（默认 LLM API 服务端口：`127.0.0.1:8888`,默认 API 服务端口：`127.0.0.1:7861`,默认 WebUI 服务端口：`本机IP：8501`)** -多卡启动： - -```shell -python server/api_allinone.py --model-path-address model@host@port --num-gpus 2 --gpus 0,1 --max-gpu-memory 10GiB -``` - -其他参数详见各脚本及 FastChat 服务说明。 - -#### 6.2 webui一键启动脚本 - -加载本地模型： - -```shell -$ python webui_allinone.py -``` - -调用远程 API 服务： - -```shell -$ python webui_allinone.py --use-remote-api -``` -如果出现server端口占用情况，需手动指定server端口,并同步修改model_config.py下对应模型的base_api_url为指定端口: - -```shell -$ python webui_allinone.py --server-port 8887 -``` - -后台运行webui服务： - -```shell -$ python webui_allinone.py --nohup -``` - -加载多个非默认模型： - -```shell -$ python webui_allinone.py --model-path-address model1@host1@port1 model2@host2@port2 -``` - -多卡启动： - -```shell -$ python webui_alline.py --model-path-address model@host@port --num-gpus 2 --gpus 0,1 --max-gpu-memory 10GiB -``` - -其他参数详见各脚本及 Fastchat 服务说明。 - -上述两个一键启动脚本会后台运行多个服务，如要停止所有服务，可使用 `shutdown_all.sh` 脚本： - -```shell -bash shutdown_all.sh -``` +**2.服务启动时间示设备不同而不同，约 3-10 分钟，如长时间没有启动请前往 `./logs`目录下监控日志，定位问题。** ## 常见问题 diff --git a/configs/__init__.py b/configs/__init__.py index 0bed9b6..adec2c0 100644 --- a/configs/__init__.py +++ b/configs/__init__.py @@ -1 +1,4 @@ -from .model_config import * \ No newline at end of file +from .model_config import * +from .server_config import * + +VERSION = "v0.2.1-preview" diff --git a/configs/model_config.py.example b/configs/model_config.py.example index 8771cfc..ccfd1b6 100644 --- a/configs/model_config.py.example +++ b/configs/model_config.py.example @@ -1,14 +1,11 @@ import os import logging import torch -import argparse -import json # 日志格式 LOG_FORMAT = "%(asctime)s - %(filename)s[line:%(lineno)d] - %(levelname)s: %(message)s" logger = logging.getLogger() logger.setLevel(logging.INFO) logging.basicConfig(format=LOG_FORMAT) -import json # 在以下字典中修改属性值，以指定本地embedding模型存储位置 @@ -52,13 +49,13 @@ llm_model_dict = { "chatglm2-6b": { "local_model_path": "THUDM/chatglm2-6b", - "api_base_url": "http://localhost:8888/v1", # "name"修改为fastchat服务中的"api_base_url" + "api_base_url": "http://localhost:8888/v1", # URL需要与运行fastchat服务端的server_config.FSCHAT_OPENAI_API一致 "api_key": "EMPTY" }, "chatglm2-6b-32k": { "local_model_path": "THUDM/chatglm2-6b-32k", # "THUDM/chatglm2-6b-32k", - "api_base_url": "http://localhost:8888/v1", # "name"修改为fastchat服务中的"api_base_url" + "api_base_url": "http://localhost:8888/v1", # "URL需要与运行fastchat服务端的server_config.FSCHAT_OPENAI_API一致 "api_key": "EMPTY" }, diff --git a/configs/server_config.py.example b/configs/server_config.py.example new file mode 100644 index 0000000..5f37779 --- /dev/null +++ b/configs/server_config.py.example @@ -0,0 +1,100 @@ +from .model_config import LLM_MODEL, LLM_DEVICE + +# API 是否开启跨域，默认为False，如果需要开启，请设置为True +# is open cross domain +OPEN_CROSS_DOMAIN = False + +# 各服务器默认绑定host +DEFAULT_BIND_HOST = "127.0.0.1" + +# webui.py server +WEBUI_SERVER = { + "host": DEFAULT_BIND_HOST, + "port": 8501, +} + +# api.py server +API_SERVER = { + "host": DEFAULT_BIND_HOST, + "port": 7861, +} + +# fastchat openai_api server +FSCHAT_OPENAI_API = { + "host": DEFAULT_BIND_HOST, + "port": 8888, # model_config.llm_model_dict中模型配置的api_base_url需要与这里一致。 +} + +# fastchat model_worker server +# 这些模型必须是在model_config.llm_model_dict中正确配置的。 +# 在启动startup.py时，可用通过`--model-worker --model-name xxxx`指定模型，不指定则为LLM_MODEL +FSCHAT_MODEL_WORKERS = { + LLM_MODEL: { + "host": DEFAULT_BIND_HOST, + "port": 20002, + "device": LLM_DEVICE, + # todo: 多卡加载需要配置的参数 + "gpus": None, + "numgpus": 1, + # 以下为非常用参数，可根据需要配置 + # "max_gpu_memory": "20GiB", + # "load_8bit": False, + # "cpu_offloading": None, + # "gptq_ckpt": None, + # "gptq_wbits": 16, + # "gptq_groupsize": -1, + # "gptq_act_order": False, + # "awq_ckpt": None, + # "awq_wbits": 16, + # "awq_groupsize": -1, + # "model_names": [LLM_MODEL], + # "conv_template": None, + # "limit_worker_concurrency": 5, + # "stream_interval": 2, + # "no_register": False, + }, +} + +# fastchat multi model worker server +FSCHAT_MULTI_MODEL_WORKERS = { + # todo +} + +# fastchat controller server +FSCHAT_CONTROLLER = { + "host": DEFAULT_BIND_HOST, + "port": 20001, + "dispatch_method": "shortest_queue", +} + + +# 以下不要更改 +def fschat_controller_address() -> str: + host = FSCHAT_CONTROLLER["host"] + port = FSCHAT_CONTROLLER["port"] + return f"http://{host}:{port}" + + +def fschat_model_worker_address(model_name: str = LLM_MODEL) -> str: + if model := FSCHAT_MODEL_WORKERS.get(model_name): + host = model["host"] + port = model["port"] + return f"http://{host}:{port}" + + +def fschat_openai_api_address() -> str: + host = FSCHAT_OPENAI_API["host"] + port = FSCHAT_OPENAI_API["port"] + return f"http://{host}:{port}" + + +def api_address() -> str: + host = API_SERVER["host"] + port = API_SERVER["port"] + return f"http://{host}:{port}" + + +def webui_address() -> str: + host = WEBUI_SERVER["host"] + port = WEBUI_SERVER["port"] + return f"http://{host}:{port}" diff --git a/server/api.py b/server/api.py index 800680c..ecadd7c 100644 --- a/server/api.py +++ b/server/api.py @@ -4,7 +4,9 @@ import os sys.path.append(os.path.dirname(os.path.dirname(__file__))) -from configs.model_config import NLTK_DATA_PATH, OPEN_CROSS_DOMAIN +from configs.model_config import NLTK_DATA_PATH +from configs.server_config import OPEN_CROSS_DOMAIN +from configs import VERSION import argparse import uvicorn from fastapi.middleware.cors import CORSMiddleware @@ -14,11 +16,10 @@ from server.chat import (chat, knowledge_base_chat, openai_chat, from server.knowledge_base.kb_api import list_kbs, create_kb, delete_kb from server.knowledge_base.kb_doc_api import (list_docs, upload_doc, delete_doc, update_doc, download_doc, recreate_vector_store, - search_docs, DocumentWithScore) + search_docs, DocumentWithScore) from server.utils import BaseResponse, ListResponse, FastAPI, MakeFastAPIOffline from typing import List - nltk.data.path = [NLTK_DATA_PATH] + nltk.data.path @@ -27,7 +28,10 @@ async def document(): def create_app(): - app = FastAPI(title="Langchain-Chatchat API Server") + app = FastAPI( + title="Langchain-Chatchat API Server", + version=VERSION + ) MakeFastAPIOffline(app) # Add CORS middleware to allow all origins # 在config.py中设置OPEN_DOMAIN=True，允许跨域 @@ -75,10 +79,10 @@ def create_app(): )(create_kb) app.post("/knowledge_base/delete_knowledge_base", - tags=["Knowledge Base Management"], - response_model=BaseResponse, - summary="删除知识库" - )(delete_kb) + tags=["Knowledge Base Management"], + response_model=BaseResponse, + summary="删除知识库" + )(delete_kb) app.get("/knowledge_base/list_docs", tags=["Knowledge Base Management"], @@ -87,10 +91,10 @@ def create_app(): )(list_docs) app.post("/knowledge_base/search_docs", - tags=["Knowledge Base Management"], - response_model=List[DocumentWithScore], - summary="搜索知识库" - )(search_docs) + tags=["Knowledge Base Management"], + response_model=List[DocumentWithScore], + summary="搜索知识库" + )(search_docs) app.post("/knowledge_base/upload_doc", tags=["Knowledge Base Management"], @@ -99,10 +103,10 @@ def create_app(): )(upload_doc) app.post("/knowledge_base/delete_doc", - tags=["Knowledge Base Management"], - response_model=BaseResponse, - summary="删除知识库内指定文件" - )(delete_doc) + tags=["Knowledge Base Management"], + response_model=BaseResponse, + summary="删除知识库内指定文件" + )(delete_doc) app.post("/knowledge_base/update_doc", tags=["Knowledge Base Management"], diff --git a/server/api_allinone.py b/server/api_allinone_stale.py similarity index 95% rename from server/api_allinone.py rename to server/api_allinone_stale.py index 3be8581..78a7a6d 100644 --- a/server/api_allinone.py +++ b/server/api_allinone_stale.py @@ -15,7 +15,7 @@ import os sys.path.append(os.path.dirname(__file__)) sys.path.append(os.path.dirname(os.path.dirname(__file__))) -from llm_api_launch import launch_all, parser, controller_args, worker_args, server_args +from llm_api_stale import launch_all, parser, controller_args, worker_args, server_args from api import create_app import uvicorn diff --git a/server/llm_api_launch.py b/server/llm_api_stale.py similarity index 100% rename from server/llm_api_launch.py rename to server/llm_api_stale.py diff --git a/webui_allinone.py b/server/webui_allinone_stale.py similarity index 59% rename from webui_allinone.py rename to server/webui_allinone_stale.py index cea6895..627f956 100644 --- a/webui_allinone.py +++ b/server/webui_allinone_stale.py @@ -20,51 +20,50 @@ from webui_pages.utils import * from streamlit_option_menu import option_menu from webui_pages import * import os -from server.llm_api_launch import string_args, launch_all, controller_args, worker_args, server_args, LOG_PATH +from server.llm_api_stale import string_args,launch_all,controller_args,worker_args,server_args,LOG_PATH -from server.api_allinone import parser, api_args +from server.api_allinone_stale import parser, api_args import subprocess -parser.add_argument("--use-remote-api", action="store_true") -parser.add_argument("--nohup", action="store_true") -parser.add_argument("--server.port", type=int, default=8501) -parser.add_argument("--theme.base", type=str, default='"light"') -parser.add_argument("--theme.primaryColor", type=str, default='"#165dff"') -parser.add_argument("--theme.secondaryBackgroundColor", type=str, default='"#f5f5f5"') -parser.add_argument("--theme.textColor", type=str, default='"#000000"') -web_args = ["server.port", "theme.base", "theme.primaryColor", "theme.secondaryBackgroundColor", "theme.textColor"] +parser.add_argument("--use-remote-api",action="store_true") +parser.add_argument("--nohup",action="store_true") +parser.add_argument("--server.port",type=int,default=8501) +parser.add_argument("--theme.base",type=str,default='"light"') +parser.add_argument("--theme.primaryColor",type=str,default='"#165dff"') +parser.add_argument("--theme.secondaryBackgroundColor",type=str,default='"#f5f5f5"') +parser.add_argument("--theme.textColor",type=str,default='"#000000"') +web_args = ["server.port","theme.base","theme.primaryColor","theme.secondaryBackgroundColor","theme.textColor"] -def launch_api(args, args_list=api_args, log_name=None): +def launch_api(args,args_list=api_args,log_name=None): print("Launching api ...") print("启动API服务...") if not log_name: log_name = f"{LOG_PATH}api_{args.api_host}_{args.api_port}" print(f"logs on api are written in {log_name}") print(f"API日志位于{log_name}下，如启动异常请查看日志") - args_str = string_args(args, args_list) + args_str = string_args(args,args_list) api_sh = "python server/{script} {args_str} >{log_name}.log 2>&1 &".format( - script="api.py", args_str=args_str, log_name=log_name) + script="api.py",args_str=args_str,log_name=log_name) subprocess.run(api_sh, shell=True, check=True) print("launch api done!") print("启动API服务完毕.") - -def launch_webui(args, args_list=web_args, log_name=None): +def launch_webui(args,args_list=web_args,log_name=None): print("Launching webui...") print("启动webui服务...") if not log_name: log_name = f"{LOG_PATH}webui" - args_str = string_args(args, args_list) + args_str = string_args(args,args_list) if args.nohup: print(f"logs on api are written in {log_name}") print(f"webui服务日志位于{log_name}下，如启动异常请查看日志") webui_sh = "streamlit run webui.py {args_str} >{log_name}.log 2>&1 &".format( - args_str=args_str, log_name=log_name) + args_str=args_str,log_name=log_name) else: webui_sh = "streamlit run webui.py {args_str}".format( - args_str=args_str) + args_str=args_str) subprocess.run(webui_sh, shell=True, check=True) print("launch webui done!") print("启动webui服务完毕.") @@ -75,10 +74,10 @@ if __name__ == "__main__": print(f"开始启动webui_allinone,启动LLM服务需要约3-10分钟，请耐心等待，如长时间未启动，请到{LOG_PATH}下查看日志...") args = parser.parse_args() - print("*" * 80) + print("*"*80) if not args.use_remote_api: - launch_all(args=args, controller_args=controller_args, worker_args=worker_args, server_args=server_args) - launch_api(args=args, args_list=api_args) - launch_webui(args=args, args_list=web_args) + launch_all(args=args,controller_args=controller_args,worker_args=worker_args,server_args=server_args) + launch_api(args=args,args_list=api_args) + launch_webui(args=args,args_list=web_args) print("Start webui_allinone.py done!") - print("感谢耐心等待，启动webui_allinone完毕。") + print("感谢耐心等待，启动webui_allinone完毕。") \ No newline at end of file diff --git a/startup.py b/startup.py new file mode 100644 index 0000000..830c322 --- /dev/null +++ b/startup.py @@ -0,0 +1,456 @@ +from multiprocessing import Process, Queue +import multiprocessing as mp +import subprocess +import sys +import os +from pprint import pprint + +sys.path.append(os.path.dirname(os.path.dirname(__file__))) +from configs.model_config import EMBEDDING_DEVICE, EMBEDDING_MODEL, llm_model_dict, LLM_MODEL, LLM_DEVICE, LOG_PATH, \ + logger +from configs.server_config import (WEBUI_SERVER, API_SERVER, OPEN_CROSS_DOMAIN, FSCHAT_CONTROLLER, FSCHAT_MODEL_WORKERS, + FSCHAT_OPENAI_API, fschat_controller_address, fschat_model_worker_address, + fschat_openai_api_address, ) +from server.utils import MakeFastAPIOffline, FastAPI +import argparse +from typing import Tuple, List +from configs import VERSION + + +def set_httpx_timeout(timeout=60.0): + import httpx + httpx._config.DEFAULT_TIMEOUT_CONFIG.connect = timeout + httpx._config.DEFAULT_TIMEOUT_CONFIG.read = timeout + httpx._config.DEFAULT_TIMEOUT_CONFIG.write = timeout + + +def create_controller_app( + dispatch_method: str, +) -> FastAPI: + import fastchat.constants + fastchat.constants.LOGDIR = LOG_PATH + from fastchat.serve.controller import app, Controller + + controller = Controller(dispatch_method) + sys.modules["fastchat.serve.controller"].controller = controller + + MakeFastAPIOffline(app) + app.title = "FastChat Controller" + return app + + +def create_model_worker_app(**kwargs) -> Tuple[argparse.ArgumentParser, FastAPI]: + import fastchat.constants + fastchat.constants.LOGDIR = LOG_PATH + from fastchat.serve.model_worker import app, GptqConfig, AWQConfig, ModelWorker, worker_id + import argparse + import threading + import fastchat.serve.model_worker + + # workaround to make program exit with Ctrl+c + # it should be deleted after pr is merged by fastchat + def _new_init_heart_beat(self): + self.register_to_controller() + self.heart_beat_thread = threading.Thread( + target=fastchat.serve.model_worker.heart_beat_worker, args=(self,), daemon=True, + ) + self.heart_beat_thread.start() + + ModelWorker.init_heart_beat = _new_init_heart_beat + + parser = argparse.ArgumentParser() + args = parser.parse_args([]) + # default args. should be deleted after pr is merged by fastchat + args.gpus = None + args.max_gpu_memory = "20GiB" + args.load_8bit = False + args.cpu_offloading = None + args.gptq_ckpt = None + args.gptq_wbits = 16 + args.gptq_groupsize = -1 + args.gptq_act_order = False + args.awq_ckpt = None + args.awq_wbits = 16 + args.awq_groupsize = -1 + args.num_gpus = 1 + args.model_names = [] + args.conv_template = None + args.limit_worker_concurrency = 5 + args.stream_interval = 2 + args.no_register = False + + for k, v in kwargs.items(): + setattr(args, k, v) + + if args.gpus: + if args.num_gpus is None: + args.num_gpus = len(args.gpus.split(',')) + if len(args.gpus.split(",")) < args.num_gpus: + raise ValueError( + f"Larger --num-gpus ({args.num_gpus}) than --gpus {args.gpus}!" + ) + os.environ["CUDA_VISIBLE_DEVICES"] = args.gpus + + gptq_config = GptqConfig( + ckpt=args.gptq_ckpt or args.model_path, + wbits=args.gptq_wbits, + groupsize=args.gptq_groupsize, + act_order=args.gptq_act_order, + ) + awq_config = AWQConfig( + ckpt=args.awq_ckpt or args.model_path, + wbits=args.awq_wbits, + groupsize=args.awq_groupsize, + ) + + worker = ModelWorker( + controller_addr=args.controller_address, + worker_addr=args.worker_address, + worker_id=worker_id, + model_path=args.model_path, + model_names=args.model_names, + limit_worker_concurrency=args.limit_worker_concurrency, + no_register=args.no_register, + device=args.device, + num_gpus=args.num_gpus, + max_gpu_memory=args.max_gpu_memory, + load_8bit=args.load_8bit, + cpu_offloading=args.cpu_offloading, + gptq_config=gptq_config, + awq_config=awq_config, + stream_interval=args.stream_interval, + conv_template=args.conv_template, + ) + + sys.modules["fastchat.serve.model_worker"].worker = worker + sys.modules["fastchat.serve.model_worker"].args = args + sys.modules["fastchat.serve.model_worker"].gptq_config = gptq_config + + MakeFastAPIOffline(app) + app.title = f"FastChat LLM Server ({LLM_MODEL})" + return app + + +def create_openai_api_app( + controller_address: str, + api_keys: List = [], +) -> FastAPI: + import fastchat.constants + fastchat.constants.LOGDIR = LOG_PATH + from fastchat.serve.openai_api_server import app, CORSMiddleware, app_settings + + app.add_middleware( + CORSMiddleware, + allow_credentials=True, + allow_origins=["*"], + allow_methods=["*"], + allow_headers=["*"], + ) + + app_settings.controller_address = controller_address + app_settings.api_keys = api_keys + + MakeFastAPIOffline(app) + app.title = "FastChat OpeanAI API Server" + return app + + +def _set_app_seq(app: FastAPI, q: Queue, run_seq: int): + if run_seq == 1: + @app.on_event("startup") + async def on_startup(): + set_httpx_timeout() + q.put(run_seq) + elif run_seq > 1: + @app.on_event("startup") + async def on_startup(): + set_httpx_timeout() + while True: + no = q.get() + if no != run_seq - 1: + q.put(no) + else: + break + q.put(run_seq) + + +def run_controller(q: Queue, run_seq: int = 1): + import uvicorn + + app = create_controller_app(FSCHAT_CONTROLLER.get("dispatch_method")) + _set_app_seq(app, q, run_seq) + + host = FSCHAT_CONTROLLER["host"] + port = FSCHAT_CONTROLLER["port"] + uvicorn.run(app, host=host, port=port) + + +def run_model_worker( + model_name: str = LLM_MODEL, + controller_address: str = "", + q: Queue = None, + run_seq: int = 2, +): + import uvicorn + + kwargs = FSCHAT_MODEL_WORKERS[LLM_MODEL].copy() + host = kwargs.pop("host") + port = kwargs.pop("port") + model_path = llm_model_dict[model_name].get("local_model_path", "") + kwargs["model_path"] = model_path + kwargs["model_names"] = [model_name] + kwargs["controller_address"] = controller_address or fschat_controller_address() + kwargs["worker_address"] = fschat_model_worker_address() + + app = create_model_worker_app(**kwargs) + _set_app_seq(app, q, run_seq) + + uvicorn.run(app, host=host, port=port) + + +def run_openai_api(q: Queue, run_seq: int = 3): + import uvicorn + + controller_addr = fschat_controller_address() + app = create_openai_api_app(controller_addr) # todo: not support keys yet. + _set_app_seq(app, q, run_seq) + + host = FSCHAT_OPENAI_API["host"] + port = FSCHAT_OPENAI_API["port"] + uvicorn.run(app, host=host, port=port) + + +def run_api_server(q: Queue, run_seq: int = 4): + from server.api import create_app + import uvicorn + + app = create_app() + _set_app_seq(app, q, run_seq) + + host = API_SERVER["host"] + port = API_SERVER["port"] + + uvicorn.run(app, host=host, port=port) + + +def run_webui(q: Queue, run_seq: int = 5): + host = WEBUI_SERVER["host"] + port = WEBUI_SERVER["port"] + while True: + no = q.get() + if no != run_seq - 1: + q.put(no) + else: + break + q.put(run_seq) + p = subprocess.Popen(["streamlit", "run", "webui.py", + "--server.address", host, + "--server.port", str(port)]) + p.wait() + + +def parse_args() -> argparse.ArgumentParser: + parser = argparse.ArgumentParser() + parser.add_argument( + "--all-webui", + action="store_true", + help="run fastchat's controller/model_worker/openai_api servers, run api.py and webui.py", + dest="all_webui", + ) + parser.add_argument( + "--all-api", + action="store_true", + help="run fastchat's controller/model_worker/openai_api servers, run api.py and webui.py", + dest="all_api", + ) + parser.add_argument( + "--llm-api", + action="store_true", + help="run fastchat's controller/model_worker/openai_api servers, run api.py and webui.py", + dest="llm_api", + ) + parser.add_argument( + "-o", + "--openai-api", + action="store_true", + help="run fastchat controller/openai_api servers", + dest="openai_api", + ) + parser.add_argument( + "-m", + "--model-worker", + action="store_true", + help="run fastchat model_worker server with specified model name. specify --model-name if not using default LLM_MODEL", + dest="model_worker", + ) + parser.add_argument( + "-n" + "--model-name", + type=str, + default=LLM_MODEL, + help="specify model name for model worker.", + dest="model_name", + ) + parser.add_argument( + "-c" + "--controller", + type=str, + help="specify controller address the worker is registered to. default is server_config.FSCHAT_CONTROLLER", + dest="controller_address", + ) + parser.add_argument( + "--api", + action="store_true", + help="run api.py server", + dest="api", + ) + parser.add_argument( + "-w", + "--webui", + action="store_true", + help="run webui.py server", + dest="webui", + ) + args = parser.parse_args() + return args + + +if __name__ == "__main__": + import platform + import time + import langchain + import fastchat + from configs.server_config import api_address, webui_address + + mp.set_start_method("spawn") + queue = Queue() + args = parse_args() + if args.all_webui: + args.openai_api = True + args.model_worker = True + args.api = True + args.webui = True + + elif args.all_api: + args.openai_api = True + args.model_worker = True + args.api = True + args.webui = False + + elif args.llm_api: + args.openai_api = True + args.model_worker = True + args.api = False + args.webui = False + + logger.info(f"正在启动服务：") + logger.info(f"如需查看 llm_api 日志，请前往 {LOG_PATH}") + + processes = {} + + if args.openai_api: + process = Process( + target=run_controller, + name=f"controller({os.getpid()})", + args=(queue, len(processes) + 1), + daemon=True, + ) + process.start() + processes["controller"] = process + + process = Process( + target=run_openai_api, + name=f"openai_api({os.getpid()})", + args=(queue, len(processes) + 1), + daemon=True, + ) + process.start() + processes["openai_api"] = process + + if args.model_worker: + process = Process( + target=run_model_worker, + name=f"model_worker({os.getpid()})", + args=(args.model_name, args.controller_address, queue, len(processes) + 1), + daemon=True, + ) + process.start() + processes["model_worker"] = process + + if args.api: + process = Process( + target=run_api_server, + name=f"API Server{os.getpid()})", + args=(queue, len(processes) + 1), + daemon=True, + ) + process.start() + processes["api"] = process + + if args.webui: + process = Process( + target=run_webui, + name=f"WEBUI Server{os.getpid()})", + args=(queue, len(processes) + 1), + daemon=True, + ) + process.start() + processes["webui"] = process + + try: + # log infors + while True: + no = queue.get() + if no == len(processes): + time.sleep(0.5) + print("\n\n") + print("=" * 30 + "Langchain-Chatchat Configuration" + "=" * 30) + print(f"操作系统：{platform.platform()}.") + print(f"python版本：{sys.version}") + print(f"项目版本：{VERSION}") + print(f"langchain版本：{langchain.__version__}. fastchat版本：{fastchat.__version__}") + print("\n") + print(f"当前LLM模型：{LLM_MODEL} @ {LLM_DEVICE}") + pprint(llm_model_dict[LLM_MODEL]) + print(f"当前Embbedings模型： {EMBEDDING_MODEL} @ {EMBEDDING_DEVICE}") + print("\n") + print(f"服务端运行信息：") + if args.openai_api: + print(f" OpenAI API Server: {fschat_openai_api_address()}/v1") + print("请确认llm_model_dict中配置的api_base_url与上面地址一致。") + if args.api: + print(f" Chatchat API Server: {api_address()}") + if args.webui: + print(f" Chatchat WEBUI Server: {webui_address()}") + print("=" * 30 + "Langchain-Chatchat Configuration" + "=" * 30) + print("\n\n") + break + else: + queue.put(no) + + if model_worker_process := processes.get("model_worker"): + model_worker_process.join() + for name, process in processes.items(): + if name != "model_worker": + process.join() + except: + if model_worker_process := processes.get("model_worker"): + model_worker_process.terminate() + for name, process in processes.items(): + if name != "model_worker": + process.terminate() + +# 服务启动后接口调用示例： +# import openai +# openai.api_key = "EMPTY" # Not support yet +# openai.api_base = "http://localhost:8888/v1" + +# model = "chatglm2-6b" + +# # create a chat completion +# completion = openai.ChatCompletion.create( +# model=model, +# messages=[{"role": "user", "content": "Hello! What is your name?"}] +# ) +# # print the completion +# print(completion.choices[0].message.content) diff --git a/tests/api/stream_api_test.py b/tests/api/stream_api_test.py index 06a9654..2902c8a 100644 --- a/tests/api/stream_api_test.py +++ b/tests/api/stream_api_test.py @@ -28,4 +28,14 @@ if __name__ == "__main__": for line in response.iter_content(decode_unicode=True): print(line, flush=True) else: - print("Error:", response.status_code) \ No newline at end of file + print("Error:", response.status_code) + + + r = requests.post( + openai_url + "/chat/completions", + json={"model": LLM_MODEL, "messages": "你好", "max_tokens": 1000}) + data = r.json() + print(f"/chat/completions\n") + print(data) + assert "choices" in data + diff --git a/webui.py b/webui.py index 99db3f6..58fc0e3 100644 --- a/webui.py +++ b/webui.py @@ -9,6 +9,7 @@ from webui_pages.utils import * from streamlit_option_menu import option_menu from webui_pages import * import os +from configs import VERSION api = ApiRequest(base_url="http://127.0.0.1:7861", no_remote_api=False) @@ -17,6 +18,11 @@ if __name__ == "__main__": "Langchain-Chatchat WebUI", os.path.join("img", "chatchat_icon_blue_square_v2.png"), initial_sidebar_state="expanded", + menu_items={ + 'Get Help': 'https://github.com/chatchat-space/Langchain-Chatchat', + 'Report a bug': "https://github.com/chatchat-space/Langchain-Chatchat/issues", + 'About': f"""欢迎使用 Langchain-Chatchat WebUI {VERSION}！""" + } ) if not chat_box.chat_inited: @@ -35,7 +41,7 @@ if __name__ == "__main__": "func": knowledge_base_page, }, } - + with st.sidebar: st.image( os.path.join( @@ -44,6 +50,10 @@ if __name__ == "__main__": ), use_column_width=True ) + st.caption( + f"""

当前版本：{VERSION}

""", + unsafe_allow_html=True, + ) options = list(pages) icons = [x["icon"] for x in pages.values()]