github/DB-GPT
1
0
Fork 0
mirror of https://github.com/csunny/DB-GPT.git synced 2025-07-29 14:57:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

doc:add llm management and deployment documents

Browse Source
This commit is contained in:
aries_ckt 2023-10-17 19:46:34 +08:00
parent 806612ad7d
commit 69f68fffb2
11 changed files with 1106 additions and 242 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
docs/_static/img/muti-model-cluster-overview.png vendored
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 361 KiB

After

Width:  |  Height:  |  Size: 668 KiB

3
docs/getting_started/application.rst
View File

@ -19,4 +19,5 @@ DB-GPT product is a Web application that you can chat database, chat knowledge,
./application/chatdb/chatdb.md
./application/kbqa/kbqa.md
./application/dashboard/dashboard.md
./application/chatexcel/chatexcel.md
./application/chatexcel/chatexcel.md
./application/model/model.md

61
docs/getting_started/application/model/model.md Normal file
View File

@ -0,0 +1,61 @@
Model Management
==================================
![model](https://github.com/eosphoros-ai/DB-GPT/assets/13723926/4b160ee7-2e2a-4502-bd54-d7daa14b23e5)
DB-GPT Product Provides LLM Model Management in web interface.Including LLM Create, Start and Stop.
Now DB-GPT support LLMs:
```{admonition} Support LLMs
* Multi LLMs Support, Supports multiple large language models, currently supporting
* [meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf)
* [baichuan2-7b/baichuan2-13b](https://huggingface.co/baichuan-inc)
* [internlm/internlm-chat-7b](https://huggingface.co/internlm/internlm-chat-7b)
* [Qwen/Qwen-7B-Chat/Qwen-14B-Chat](https://huggingface.co/Qwen/)
* [Vicuna](https://huggingface.co/Tribbiani/vicuna-13b)
* [BlinkDL/RWKV-4-Raven](https://huggingface.co/BlinkDL/rwkv-4-raven)
* [camel-ai/CAMEL-13B-Combined-Data](https://huggingface.co/camel-ai/CAMEL-13B-Combined-Data)
* [databricks/dolly-v2-12b](https://huggingface.co/databricks/dolly-v2-12b)
* [FreedomIntelligence/phoenix-inst-chat-7b](https://huggingface.co/FreedomIntelligence/phoenix-inst-chat-7b)
* [h2oai/h2ogpt-gm-oasst1-en-2048-open-llama-7b](https://huggingface.co/h2oai/h2ogpt-gm-oasst1-en-2048-open-llama-7b)
* [lcw99/polyglot-ko-12.8b-chang-instruct-chat](https://huggingface.co/lcw99/polyglot-ko-12.8b-chang-instruct-chat)
* [lmsys/fastchat-t5-3b-v1.0](https://huggingface.co/lmsys/fastchat-t5)
* [mosaicml/mpt-7b-chat](https://huggingface.co/mosaicml/mpt-7b-chat)
* [Neutralzz/BiLLa-7B-SFT](https://huggingface.co/Neutralzz/BiLLa-7B-SFT)
* [nomic-ai/gpt4all-13b-snoozy](https://huggingface.co/nomic-ai/gpt4all-13b-snoozy)
* [NousResearch/Nous-Hermes-13b](https://huggingface.co/NousResearch/Nous-Hermes-13b)
* [openaccess-ai-collective/manticore-13b-chat-pyg](https://huggingface.co/openaccess-ai-collective/manticore-13b-chat-pyg)
* [OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5](https://huggingface.co/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5)
* [project-baize/baize-v2-7b](https://huggingface.co/project-baize/baize-v2-7b)
* [Salesforce/codet5p-6b](https://huggingface.co/Salesforce/codet5p-6b)
* [StabilityAI/stablelm-tuned-alpha-7b](https://huggingface.co/stabilityai/stablelm-tuned-alpha-7b)
* [THUDM/chatglm-6b](https://huggingface.co/THUDM/chatglm-6b)
* [THUDM/chatglm2-6b](https://huggingface.co/THUDM/chatglm2-6b)
* [tiiuae/falcon-40b](https://huggingface.co/tiiuae/falcon-40b)
* [timdettmers/guanaco-33b-merged](https://huggingface.co/timdettmers/guanaco-33b-merged)
* [togethercomputer/RedPajama-INCITE-7B-Chat](https://huggingface.co/togethercomputer/RedPajama-INCITE-7B-Chat)
* [WizardLM/WizardLM-13B-V1.0](https://huggingface.co/WizardLM/WizardLM-13B-V1.0)
* [WizardLM/WizardCoder-15B-V1.0](https://huggingface.co/WizardLM/WizardCoder-15B-V1.0)
* [baichuan-inc/baichuan-7B](https://huggingface.co/baichuan-inc/baichuan-7B)
* [HuggingFaceH4/starchat-beta](https://huggingface.co/HuggingFaceH4/starchat-beta)
* [FlagAlpha/Llama2-Chinese-13b-Chat](https://huggingface.co/FlagAlpha/Llama2-Chinese-13b-Chat)
* [BAAI/AquilaChat-7B](https://huggingface.co/BAAI/AquilaChat-7B)
* [all models of OpenOrca](https://huggingface.co/Open-Orca)
* [Spicyboros](https://huggingface.co/jondurbin/spicyboros-7b-2.2?not-for-all-audiences=true) + [airoboros 2.2](https://huggingface.co/jondurbin/airoboros-l2-13b-2.2)
* [VMware's OpenLLaMa OpenInstruct](https://huggingface.co/VMware/open-llama-7b-open-instruct)
* Support API Proxy LLMs
* [ChatGPT](https://api.openai.com/)
* [Tongyi](https://www.aliyun.com/product/dashscope)
* [Wenxin](https://cloud.baidu.com/product/wenxinworkshop?track=dingbutonglan)
* [ChatGLM](http://open.bigmodel.cn/)
```
### Create && Start LLM Model
```{note}
Make sure your LLM Model file is downloaded or LLM Model Proxy api service is ready.
```
![model-start](https://github.com/eosphoros-ai/DB-GPT/assets/13723926/dacabcb9-92c6-43eb-95ed-8cabaa2d18e6)
When create success, you can see:
![image](https://github.com/eosphoros-ai/DB-GPT/assets/13723926/1b69bff6-8b37-493d-b6be-38f7b6e8ae2d)
Then you can choose and switch llm model service to chat.
![image](https://github.com/eosphoros-ai/DB-GPT/assets/13723926/2d20eb6b-8976-4731-b433-373ac3383602)
### Stop LLM Model
![image](https://github.com/eosphoros-ai/DB-GPT/assets/13723926/a21278d9-7bef-487b-bef1-460ce516b2f5)

64
docs/getting_started/install/cluster/cluster.rst
View File

@ -1,5 +1,8 @@
Cluster deployment
LLM Deployment
==================================
In the exploration and implementation of AI model applications, it can be challenging to directly integrate with model services. Currently, there is no established standard for deploying large models, and new models and inference methods are constantly being released. As a result, a significant amount of time is spent adapting to the ever-changing underlying model environment. This, to some extent, hinders the exploration and implementation of AI model applications.
We divide the deployment of large models into two layers: the model inference layer and the model deployment layer. The model inference layer corresponds to model inference frameworks such as vLLM, TGI, and TensorRT. The model deployment layer interfaces with the inference layer below and provides model serving capabilities above. We refer to this layer's framework as the model deployment framework. Positioned above the inference frameworks, the model deployment framework offers capabilities such as multiple model instances, multiple inference frameworks, multiple service protocols, multi-cloud support, automatic scaling, and observability.
In order to deploy DB-GPT to multiple nodes, you can deploy a cluster. The cluster architecture diagram is as follows:
@ -7,8 +10,64 @@ In order to deploy DB-GPT to multiple nodes, you can deploy a cluster. The clust
<img src="../../../_static/img/muti-model-cluster-overview.png" />
Design of DB-GPT:
-----------------
DB-GPT is designed as a llm deployment framework, taking into account the above design objectives.
- Support for llm and inference frameworks: DB-GPT supports the simultaneous deployment of llm and is compatible with multiple inference frameworks such as vLLM, TGI, and TensorRT.
- Scalability and stability: DB-GPT has good scalability, allowing easy addition of new models and inference frameworks. It utilizes a distributed architecture and automatic scaling capabilities to handle high concurrency and large-scale requests, ensuring system stability.
- Performance optimization: DB-GPT undergoes performance optimization to provide fast and efficient model inference capabilities, preventing it from becoming a performance bottleneck during inference.
- Management and observability capabilities: DB-GPT offers management and monitoring functionalities, including model deployment and configuration management, performance monitoring, and logging. It can generate reports on model performance and service status to promptly identify and resolve issues.
- Lightweight: DB-GPT is designed as a lightweight framework to improve deployment efficiency and save resources. It employs efficient algorithms and optimization strategies to minimize resource consumption while maintaining sufficient functionality and performance.
1.Support for multiple models and inference frameworks
-----------------
The field of large models is evolving rapidly, with new models being released and new methods being proposed for model training and inference. We believe that this situation will continue for some time.
For most users exploring and implementing AI applications, this situation has its pros and cons. The benefits are apparent, as it brings new opportunities and advancements. However, one drawback is that users may feel compelled to constantly try and explore new models and inference frameworks.
In DB-GPT, seamless support is provided for FastChat, vLLM, and llama.cpp. In theory, any model supported by these frameworks is also supported by DB-GPT. If you have requirements for faster inference speed and concurrency, you can directly use vLLM. If you want good inference performance on CPU or Apple's M1/M2 chips, you can use llama.cpp. Additionally, DB-GPT also supports various proxy models from OpenAI, Azure OpenAI, Google BARD, Wenxin Yiyan, Tongyi Qianwen, and Zhipu AI, among others.
2.Have good scalability and stability
-----------------
A comprehensive model deployment framework consists of several components: the Model Worker, which directly interfaces with the underlying inference frameworks; the Model Controller, which manages and maintains multiple model components; and the Model API, which provides external model serving capabilities.
The Model Worker plays a crucial role and needs to be highly extensible. It can be specialized for deploying large language models, embedding models, or other types of models. The choice of Model Worker depends on the deployment environment, such as a regular physical server environment, a Kubernetes environment, or specific cloud environments provided by various cloud service providers.
Having different Model Worker options allows users to select the most suitable one based on their specific requirements and infrastructure. This flexibility enables efficient deployment and utilization of models across different environments.
The Model Controller, responsible for managing model metadata, also needs to be scalable. Different deployment environments and model management requirements may call for different choices of Model Controllers.
Furthermore, I believe that model serving shares many similarities with traditional microservices. In microservices, a service can have multiple instances, and all instances are registered in a central registry. Service consumers retrieve the list of instances based on the service name from the registry and select a specific instance for invocation using a load balancing strategy.
Similarly, in model deployment, a model can have multiple instances, and all instances can be registered in a model registry. Model service consumers retrieve the list of instances based on the model name from the registry and select a specific instance for invocation using a model-specific load balancing strategy.
Introducing a model registry, responsible for storing model instance metadata, enables such an architecture. The model registry can leverage existing service registries used in microservices (such as Nacos, Eureka, etcd, Consul, etc.) as implementations. This ensures high availability of the entire deployment system.
3.High performance for framework.
------------------
and optimization are complex tasks, and inappropriate framework designs can increase this complexity. In our view, to ensure that the deployment framework does not lag behind in terms of performance, there are two main areas of focus:
Avoid excessive encapsulation: The more encapsulation and longer the chain, the more challenging it becomes to identify performance issues.
High-performance communication design: High-performance communication involves various aspects that cannot be elaborated in detail here. However, considering that Python occupies a prominent position in current AIGC applications, asynchronous interfaces are crucial for service performance in Python. Therefore, the model serving layer should only provide asynchronous interfaces and be compatible with the layers that interface with the model inference framework. If the model inference framework offers asynchronous interfaces, direct integration should be implemented. Otherwise, synchronous-to-asynchronous task conversion should be used to provide support.
4.Management and monitoring capabilities.
------------------
In the exploration or production implementation of AIGC (Artificial Intelligence and General Computing) applications, it is important for the model deployment system to have certain management capabilities. This involves controlling the deployed model instances through APIs or command-line interfaces, such as for online/offline management, restarting, and debugging.
Observability is a crucial capability in production systems, and I believe it is equally, if not more, important in AIGC applications. This is because user experiences and interactions with the system are more complex. In addition to traditional observability metrics, we are also interested in user input information and corresponding contextual information, which specific model instance and parameters were invoked, the content and response time of model outputs, user feedback, and more.
By analyzing this information, we can identify performance bottlenecks in model services and gather user experience data (e.g., response latency, problem resolution, and user satisfaction extracted from user content). These insights serve as important foundations for further optimizing the entire application.
* On :ref:`Deploying on standalone mode <standalone-index>`. Standalone Deployment.
* On :ref:`Deploying on cluster mode <local-cluster-index>`. Cluster Deployment.
* On :ref:`Deploying on local machine <local-cluster-index>`. Local cluster deployment.
.. toctree::
:maxdepth: 2
@ -16,4 +75,5 @@ In order to deploy DB-GPT to multiple nodes, you can deploy a cluster. The clust
:name: cluster_deploy
:hidden:
./vms/standalone.md
./vms/index.md

2
docs/getting_started/install/cluster/vms/index.md
View File

@ -1,4 +1,4 @@
Local cluster deployment
Cluster Deployment
==================================
(local-cluster-index)=
## Model cluster deployment

19
docs/getting_started/install/deploy/deploy.md
View File

@ -6,8 +6,19 @@ This tutorial gives you a quick walkthrough about use DB-GPT with you environmen
To get started, install DB-GPT with the following steps.
### 1. Hardware Requirements
As our project has the ability to achieve ChatGPT performance of over 85%, there are certain hardware requirements. However, overall, the project can be deployed and used on consumer-grade graphics cards. The specific hardware requirements for deployment are as follows:
### 1. Hardware Requirements
DB-GPT can be deployed on servers with low hardware requirements or on servers with high hardware requirements.
##### Low hardware requirements
The low hardware requirements mode is suitable for integrating with third-party LLM services' APIs, such as OpenAI, Tongyi, Wenxin, or Llama.cpp.
DB-GPT provides set proxy api to support LLM api.
As our project has the ability to achieve ChatGPT performance of over 85%,
##### High hardware requirements
The high hardware requirements mode is suitable for independently deploying LLM services, such as Llama series models, Baichuan, ChatGLM, Vicuna, and other private LLM service.
there are certain hardware requirements. However, overall, the project can be deployed and used on consumer-grade graphics cards. The specific hardware requirements for deployment are as follows:
| GPU | VRAM Size | Performance |
|----------|-----------| ------------------------------------------- |
@ -16,7 +27,7 @@ As our project has the ability to achieve ChatGPT performance of over 85%, there
| V100 | 16 GB | Conversation inference possible, noticeable stutter |
| T4 | 16 GB | Conversation inference possible, noticeable stutter |
if your VRAM Size is not enough, DB-GPT supported 8-bit quantization and 4-bit quantization.
If your VRAM Size is not enough, DB-GPT supported 8-bit quantization and 4-bit quantization.
Here are some of the VRAM size usage of the models we tested in some common scenarios.
@ -64,7 +75,7 @@ Notice make sure you have install git-lfs
centos:yum install git-lfs
ubuntu:app-get install git-lfs
ubuntu:apt-get install git-lfs
macos:brew install git-lfs
```

346
docs/locales/zh_CN/LC_MESSAGES/getting_started/application/model/model.po Normal file
View File

@ -0,0 +1,346 @@
# SOME DESCRIPTIVE TITLE.
# Copyright (C) 2023, csunny
# This file is distributed under the same license as the DB-GPT package.
# FIRST AUTHOR <EMAIL@ADDRESS>, 2023.
#
#, fuzzy
msgid ""
msgstr ""
"Project-Id-Version: DB-GPT 👏👏 0.3.9\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2023-10-17 19:39+0800\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language: zh_CN\n"
"Language-Team: zh_CN <LL@li.org>\n"
"Plural-Forms: nplurals=1; plural=0;\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=utf-8\n"
"Content-Transfer-Encoding: 8bit\n"
"Generated-By: Babel 2.12.1\n"
#: ../../getting_started/application/model/model.md:1
#: 9d942556958a4a83ba09229f08774e18
msgid "Model Management"
msgstr "模型服务管理"
#: ../../getting_started/application/model/model.md:3
#: 5b8688af589b4ad1ab6fb0ec8a3a664f
msgid ""
"![model](https://github.com/eosphoros-ai/DB-"
"GPT/assets/13723926/4b160ee7-2e2a-4502-bd54-d7daa14b23e5) DB-GPT Product "
"Provides LLM Model Management in web interface.Including LLM Create, "
"Start and Stop. Now DB-GPT support LLMs:"
msgstr "![model](https://github.com/eosphoros-ai/DB-"
"GPT/assets/13723926/4b160ee7-2e2a-4502-bd54-d7daa14b23e5) DB-GPT "
"在web界面上提供模型管理能力.包括模型创建、启动、停止。目前支持的模型："
#: ../../getting_started/application/model/model.md:3
#: 023da40dc5334f93948f429b1360ff50
msgid "model"
msgstr "model"
#: ../../getting_started/application/model/model.md:6
#: 77fbd490f31946f3af627a6575b04f95
msgid "Support LLMs"
msgstr "支持的模型"
#: ../../getting_started/application/model/model.md:7
#: 1fa23499c6bd4a498880f007341601e3
msgid ""
"Multi LLMs Support, Supports multiple large language models, currently "
"supporting"
msgstr "支持的模型类型："
#: ../../getting_started/application/model/model.md:8
#: 20ad38e471c241ab9f162db9acfdbefa
msgid ""
"[meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2"
"-7b-chat-hf)"
msgstr ""
#: ../../getting_started/application/model/model.md:9
#: 1984188d04d74e7e93cdae6c8f1e00a8
msgid "[baichuan2-7b/baichuan2-13b](https://huggingface.co/baichuan-inc)"
msgstr ""
#: ../../getting_started/application/model/model.md:10
#: 3b4aa596176241ca94a9c023edf911b8
msgid ""
"[internlm/internlm-chat-7b](https://huggingface.co/internlm/internlm-"
"chat-7b)"
msgstr ""
#: ../../getting_started/application/model/model.md:11
#: 2608358ee0284d26b64b18ca86c584c5
msgid "[Qwen/Qwen-7B-Chat/Qwen-14B-Chat](https://huggingface.co/Qwen/)"
msgstr ""
#: ../../getting_started/application/model/model.md:12
#: 0153627d59fb4268bfd2b8f2be0f8257
msgid "[Vicuna](https://huggingface.co/Tribbiani/vicuna-13b)"
msgstr ""
#: ../../getting_started/application/model/model.md:13
#: 930768d5e2ca4c9891d81ed9d6e4ad8d
msgid "[BlinkDL/RWKV-4-Raven](https://huggingface.co/BlinkDL/rwkv-4-raven)"
msgstr ""
#: ../../getting_started/application/model/model.md:14
#: 2b4ab4bf8b604ba082c382bd8d4df40c
msgid ""
"[camel-ai/CAMEL-13B-Combined-Data](https://huggingface.co/camel-ai/CAMEL-"
"13B-Combined-Data)"
msgstr ""
#: ../../getting_started/application/model/model.md:15
#: a9b65b74c53d464988399baba8d35684
msgid "[databricks/dolly-v2-12b](https://huggingface.co/databricks/dolly-v2-12b)"
msgstr ""
#: ../../getting_started/application/model/model.md:16
#: 96703db7e5764b998b87f3dc8932745b
msgid ""
"[FreedomIntelligence/phoenix-inst-chat-"
"7b](https://huggingface.co/FreedomIntelligence/phoenix-inst-chat-7b)"
msgstr ""
#: ../../getting_started/application/model/model.md:17
#: de76d7426b4846d495aa8631ce0d9d20
msgid ""
"[h2oai/h2ogpt-gm-oasst1-en-2048-open-llama-"
"7b](https://huggingface.co/h2oai/h2ogpt-gm-oasst1-en-2048-open-llama-7b)"
msgstr ""
#: ../../getting_started/application/model/model.md:18
#: 7b49bcb8879b4ab4bf5f458709c3d695
msgid ""
"[lcw99/polyglot-ko-12.8b-chang-instruct-"
"chat](https://huggingface.co/lcw99/polyglot-ko-12.8b-chang-instruct-chat)"
msgstr ""
#: ../../getting_started/application/model/model.md:19
#: 1ec742759fee4fbf98cc2839c72f70d9
msgid "[lmsys/fastchat-t5-3b-v1.0](https://huggingface.co/lmsys/fastchat-t5)"
msgstr ""
#: ../../getting_started/application/model/model.md:20
#: bc1a8be505684b728e6fb7f758d75ae2
msgid "[mosaicml/mpt-7b-chat](https://huggingface.co/mosaicml/mpt-7b-chat)"
msgstr ""
#: ../../getting_started/application/model/model.md:21
#: 41496816f052446da7364d39839f5135
msgid "[Neutralzz/BiLLa-7B-SFT](https://huggingface.co/Neutralzz/BiLLa-7B-SFT)"
msgstr ""
#: ../../getting_started/application/model/model.md:22
#: 35ecfef506e040a8bac2102c28b977dd
msgid ""
"[nomic-ai/gpt4all-13b-snoozy](https://huggingface.co/nomic-ai/gpt4all-"
"13b-snoozy)"
msgstr ""
#: ../../getting_started/application/model/model.md:23
#: 1d8704b20f2c4dc5a8974ab813139a62
msgid ""
"[NousResearch/Nous-Hermes-13b](https://huggingface.co/NousResearch/Nous-"
"Hermes-13b)"
msgstr ""
#: ../../getting_started/application/model/model.md:24
#: a9a04ac4211f4001912073668b5baf60
msgid ""
"[openaccess-ai-collective/manticore-13b-chat-pyg](https://huggingface.co"
"/openaccess-ai-collective/manticore-13b-chat-pyg)"
msgstr ""
#: ../../getting_started/application/model/model.md:25
#: 0c1106b0e03a41c98b03dc17baa1298b
msgid ""
"[OpenAssistant/oasst-sft-4-pythia-12b-"
"epoch-3.5](https://huggingface.co/OpenAssistant/oasst-sft-4-pythia-12b-"
"epoch-3.5)"
msgstr ""
#: ../../getting_started/application/model/model.md:26
#: b7bccd9805914aeba74685d837fa367c
msgid ""
"[project-baize/baize-v2-7b](https://huggingface.co/project-"
"baize/baize-v2-7b)"
msgstr ""
#: ../../getting_started/application/model/model.md:27
#: 522805e911364c6ebdf7e56d674a7378
msgid "[Salesforce/codet5p-6b](https://huggingface.co/Salesforce/codet5p-6b)"
msgstr ""
#: ../../getting_started/application/model/model.md:28
#: 19571aa2347347dabdc03fe1d4656be7
msgid ""
"[StabilityAI/stablelm-tuned-alpha-7b](https://huggingface.co/stabilityai"
"/stablelm-tuned-alpha-7b)"
msgstr ""
#: ../../getting_started/application/model/model.md:29
#: fba42e0373e743b08c0c49899ab970b9
msgid "[THUDM/chatglm-6b](https://huggingface.co/THUDM/chatglm-6b)"
msgstr ""
#: ../../getting_started/application/model/model.md:30
#: d518e63d906e41909c95d62c441aa746
msgid "[THUDM/chatglm2-6b](https://huggingface.co/THUDM/chatglm2-6b)"
msgstr ""
#: ../../getting_started/application/model/model.md:31
#: 48d6eacaabd14992be49428c3d5d205a
msgid "[tiiuae/falcon-40b](https://huggingface.co/tiiuae/falcon-40b)"
msgstr ""
#: ../../getting_started/application/model/model.md:32
#: 6594a1859162405bb3df610d0982f71e
msgid ""
"[timdettmers/guanaco-33b-merged](https://huggingface.co/timdettmers"
"/guanaco-33b-merged)"
msgstr ""
#: ../../getting_started/application/model/model.md:33
#: b146beb3e6d043c38dd6211bddc9f0b4
msgid ""
"[togethercomputer/RedPajama-INCITE-7B-"
"Chat](https://huggingface.co/togethercomputer/RedPajama-INCITE-7B-Chat)"
msgstr ""
#: ../../getting_started/application/model/model.md:34
#: 65f4089de9eb4056b90212efc1995db7
msgid ""
"[WizardLM/WizardLM-13B-V1.0](https://huggingface.co/WizardLM/WizardLM-"
"13B-V1.0)"
msgstr ""
#: ../../getting_started/application/model/model.md:35
#: 6184f69597b9446daa869d03b8de2f55
msgid ""
"[WizardLM/WizardCoder-15B-V1.0](https://huggingface.co/WizardLM"
"/WizardCoder-15B-V1.0)"
msgstr ""
#: ../../getting_started/application/model/model.md:36
#: 74c552cf701a4b179065e0cb62c15f0c
msgid ""
"[baichuan-inc/baichuan-7B](https://huggingface.co/baichuan-inc/baichuan-"
"7B)"
msgstr ""
#: ../../getting_started/application/model/model.md:37
#: aa5bf7e5f91c417598a298b06d4fa8b1
msgid ""
"[HuggingFaceH4/starchat-beta](https://huggingface.co/HuggingFaceH4"
"/starchat-beta)"
msgstr ""
#: ../../getting_started/application/model/model.md:38
#: 507210a5746746f8bfc9ef5421e0f712
msgid ""
"[FlagAlpha/Llama2-Chinese-13b-"
"Chat](https://huggingface.co/FlagAlpha/Llama2-Chinese-13b-Chat)"
msgstr ""
#: ../../getting_started/application/model/model.md:39
#: 8a45141e06584b23bfae1eb1d7b3152a
msgid "[BAAI/AquilaChat-7B](https://huggingface.co/BAAI/AquilaChat-7B)"
msgstr ""
#: ../../getting_started/application/model/model.md:40
#: 78f34f97c6a441a29d20d0ec3164a466
msgid "[all models of OpenOrca](https://huggingface.co/Open-Orca)"
msgstr ""
#: ../../getting_started/application/model/model.md:41
#: b7b0f7867d61464faa7b4913daaaca5c
msgid ""
"[Spicyboros](https://huggingface.co/jondurbin/spicyboros-7b-2.2?not-for-"
"all-audiences=true) + [airoboros "
"2.2](https://huggingface.co/jondurbin/airoboros-l2-13b-2.2)"
msgstr ""
#: ../../getting_started/application/model/model.md:42
#: c3750533dfc64a9fb37044eb83c857dc
msgid ""
"[VMware&#39;s OpenLLaMa OpenInstruct](https://huggingface.co/VMware/open-"
"llama-7b-open-instruct)"
msgstr ""
#: ../../getting_started/application/model/model.md:44
#: 325d5b147abe4360997d2d6cbd6da986
msgid "Support API Proxy LLMs"
msgstr "支持第三方模型服务"
#: ../../getting_started/application/model/model.md:45
#: 426d835ec30c4171b05817fc0bbafeb6
msgid "[ChatGPT](https://api.openai.com/)"
msgstr ""
#: ../../getting_started/application/model/model.md:46
#: 659eefe6896c49e09d81edd6e3c36afe
msgid "[Tongyi](https://www.aliyun.com/product/dashscope)"
msgstr ""
#: ../../getting_started/application/model/model.md:47
#: ff24d85761c6492bbd546ecda67f38ea
msgid "[Wenxin](https://cloud.baidu.com/product/wenxinworkshop?track=dingbutonglan)"
msgstr ""
#: ../../getting_started/application/model/model.md:48
#: a798259b3694447091d6f1cee969e6af
msgid "[ChatGLM](http://open.bigmodel.cn/)"
msgstr ""
#: ../../getting_started/application/model/model.md:50
#: e053ece66b444b37a4bced21a4353c95
msgid "Create && Start LLM Model"
msgstr "创建并启动模型服务"
#: ../../getting_started/application/model/model.md:52
#: 58c44380ae2946d4a3ecd22dd0ae47ac
msgid ""
"Make sure your LLM Model file is downloaded or LLM Model Proxy api "
"service is ready."
msgstr "需要事先下载模型文件或者准备好第三方模型服务api"
#: ../../getting_started/application/model/model.md:54
#: 659e756c822d4ef68b086a6e69d3ed9f
msgid ""
"![model-start](https://github.com/eosphoros-ai/DB-"
"GPT/assets/13723926/dacabcb9-92c6-43eb-95ed-8cabaa2d18e6) When create "
"success, you can see: ![image](https://github.com/eosphoros-ai/DB-"
"GPT/assets/13723926/1b69bff6-8b37-493d-b6be-38f7b6e8ae2d) Then you can "
"choose and switch llm model service to chat. ![image](https://github.com"
"/eosphoros-ai/DB-"
"GPT/assets/13723926/2d20eb6b-8976-4731-b433-373ac3383602)"
msgstr ""
#: ../../getting_started/application/model/model.md:54
#: e662c48673a94b0e9f4343f900a51af2
msgid "model-start"
msgstr ""
#: ../../getting_started/application/model/model.md:54
#: ../../getting_started/application/model/model.md:60
#: 38dd8f0de66740b3961fbd9444437d76 4190bd7bb2d34a6688382df6ee6ac610
#: f60becdb866a411babe22ad27922fdcc
msgid "image"
msgstr ""
#: ../../getting_started/application/model/model.md:59
#: 61bfa8fa14ec40e2ba56a6bc65fda9df
msgid "Stop LLM Model"
msgstr ""
#: ../../getting_started/application/model/model.md:60
#: 98188d79b0034913a500f1c0da603741
msgid ""
"![image](https://github.com/eosphoros-ai/DB-GPT/assets/13723926/a21278d9"
"-7bef-487b-bef1-460ce516b2f5)"
msgstr ""

410
docs/locales/zh_CN/LC_MESSAGES/getting_started/install/cluster/cluster.po
View File

@ -8,7 +8,7 @@ msgid ""
msgstr ""
"Project-Id-Version: DB-GPT 👏👏 0.3.6\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2023-09-13 10:11+0800\n"
"POT-Creation-Date: 2023-10-17 19:39+0800\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language: zh_CN\n"
@ -19,24 +19,412 @@ msgstr ""
"Content-Transfer-Encoding: 8bit\n"
"Generated-By: Babel 2.12.1\n"
#: ../../getting_started/install/cluster/cluster.rst:2
#: ../../getting_started/install/cluster/cluster.rst:13
#: 69804208b580447798d6946150da7bdf
#: ../../getting_started/install/cluster/cluster.rst:72
msgid "Cluster deployment"
msgstr "集群部署"
#: ../../getting_started/install/cluster/cluster.rst:4
#: fa3e4e0ae60a45eb836bcd256baa9d91
#: ../../getting_started/install/cluster/cluster.rst:2
#: bc5bb85c846b4ad19aeeccdd016f3ce8
#, fuzzy
msgid "LLM Deployment"
msgstr "集群部署"
#: ../../getting_started/install/cluster/cluster.rst:3
#: e1cebf0518db423fbc78e39945a423fa
msgid ""
"In the exploration and implementation of AI model applications, it can be"
" challenging to directly integrate with model services. Currently, there "
"is no established standard for deploying large models, and new models and"
" inference methods are constantly being released. As a result, a "
"significant amount of time is spent adapting to the ever-changing "
"underlying model environment. This, to some extent, hinders the "
"exploration and implementation of AI model applications."
msgstr ""
"在AIGC应用探索和生产落地中，难以避免直接与模型服务对接，但是目前大模型的推理部署目前还没有一个事实标准，不断有新的模型发布，也不断有新的训练和推理方法被提出，而我们就不得不花费相当一部分时间来适配多变的底层模型环境，而这在一定程度上制约了"
" AIGC 应用的探索和落地。"
#: ../../getting_started/install/cluster/cluster.rst:5
#: c6179ac327734b7ca7b87612988dad29
msgid ""
"We divide the deployment of large models into two layers: the model "
"inference layer and the model deployment layer. The model inference layer"
" corresponds to model inference frameworks such as vLLM, TGI, and "
"TensorRT. The model deployment layer interfaces with the inference layer "
"below and provides model serving capabilities above. We refer to this "
"layer's framework as the model deployment framework. Positioned above the"
" inference frameworks, the model deployment framework offers capabilities"
" such as multiple model instances, multiple inference frameworks, "
"multiple service protocols, multi-cloud support, automatic scaling, and "
"observability."
msgstr ""
"我们将大模型推理部署分为两层：模型推理层、模型部署层。模型推理层，对应模型推理框架 vLLM、TGI 和 TensorRT "
"等。模型部署层向下对接推理层，向上提供模型服务能力，这一层的框架我们称为模型部署框架，模型部署框架在推理框架之上，提供了多模型实例、多推理框架、多服务协议、多云、自动扩缩容和可观测性等能力。"
#: ../../getting_started/install/cluster/cluster.rst:7
#: 61bae2fc8e3347248ecf084a3977e448
msgid ""
"In order to deploy DB-GPT to multiple nodes, you can deploy a cluster. "
"The cluster architecture diagram is as follows:"
msgstr "为了能将 DB-GPT 部署到多个节点上，你可以部署一个集群，集群的架构图如下:"
msgstr "为了能将DB-GPT部署到多个节点上，你可以部署一个集群，集群的架构图如下:"
#: ../../getting_started/install/cluster/cluster.rst:11
#: e739449099ca43cabe9883233ca7e572
#: ../../getting_started/install/cluster/cluster.rst:14
#: af8d74ac3c5747b3934d02200afbb4ba
msgid "Design of DB-GPT:"
msgstr "设计目标"
#: ../../getting_started/install/cluster/cluster.rst:16
#: ab9f332105ac490097501798d7b6cf15
msgid ""
"DB-GPT is designed as a llm deployment framework, taking into account the"
" above design objectives."
msgstr "支持多模型和多推理框架"
#: ../../getting_started/install/cluster/cluster.rst:18
#: 281c38e2e84940098eeeb435db6d1f05
msgid ""
"Support for llm and inference frameworks: DB-GPT supports the "
"simultaneous deployment of llm and is compatible with multiple inference "
"frameworks such as vLLM, TGI, and TensorRT."
msgstr ""
"在 DB-GPT 中，直接提供了对 FastChat、vLLM和 llama.cpp 的无缝支持，理论上它们支持模型 DB-GPT "
"都支持，如果您对推理速度和并发能力有需求，可以直接使用 vLLM，如果您希望 CPU 或者 mac 的 "
"m1/m2性能也获得不错的推理性能，可以使用 llama.cpp，此外，DB-GPT 还支持了很多代理模型（openai、azure "
"openai、google bard、文心一言、通义千问和智谱AI等）。"
#: ../../getting_started/install/cluster/cluster.rst:20
#: ec7e111f2db64c7fa926b1491020ae73
msgid ""
"Scalability and stability: DB-GPT has good scalability, allowing easy "
"addition of new models and inference frameworks. It utilizes a "
"distributed architecture and automatic scaling capabilities to handle "
"high concurrency and large-scale requests, ensuring system stability."
msgstr "良好的扩展性和稳定性”"
#: ../../getting_started/install/cluster/cluster.rst:22
#: 49566c3e708c4ef3a6135ea6245a5417
msgid ""
"Performance optimization: DB-GPT undergoes performance optimization to "
"provide fast and efficient model inference capabilities, preventing it "
"from becoming a performance bottleneck during inference."
msgstr "框架性能 “不拖后腿”"
#: ../../getting_started/install/cluster/cluster.rst:24
#: 0ae41617a7904dcfadd64ec921d3987e
msgid ""
"Management and observability capabilities: DB-GPT offers management and "
"monitoring functionalities, including model deployment and configuration "
"management, performance monitoring, and logging. It can generate reports "
"on model performance and service status to promptly identify and resolve "
"issues."
msgstr "管理与可观测性能力"
#: ../../getting_started/install/cluster/cluster.rst:26
#: 7c7c762642754c8d8e8b7d4eaad55384
msgid ""
"Lightweight: DB-GPT is designed as a lightweight framework to improve "
"deployment efficiency and save resources. It employs efficient algorithms"
" and optimization strategies to minimize resource consumption while "
"maintaining sufficient functionality and performance."
msgstr "轻量化"
#: ../../getting_started/install/cluster/cluster.rst:29
#: 32c1d24c20ed4155ad05c505a355ebaf
msgid "1.Support for multiple models and inference frameworks"
msgstr "1.支持多模型和多推理框架"
#: ../../getting_started/install/cluster/cluster.rst:30
#: b0d80a26a0d14ab4a6a82bbdc693a9cc
msgid ""
"The field of large models is evolving rapidly, with new models being "
"released and new methods being proposed for model training and inference."
" We believe that this situation will continue for some time."
msgstr "当前大模型领域发展可谓日新月异，不断有新的模型发布，在模型训练和推理方面，也不断有新的方法被提出。我们判断，这样情况还会持续一段时间。"
#: ../../getting_started/install/cluster/cluster.rst:32
#: 8b7c3830e5d64567bef9244bd0c4442d
msgid ""
"For most users exploring and implementing AI applications, this situation"
" has its pros and cons. The benefits are apparent, as it brings new "
"opportunities and advancements. However, one drawback is that users may "
"feel compelled to constantly try and explore new models and inference "
"frameworks."
msgstr ""
"大于大部分 AIGC "
"应用场景探索和落地的用户来说，这种情况有利也有弊，利无需多言，而弊端之一就在于被“牵着鼻子走”，需要不断去尝试和探索新的模型、新的推理框架。"
#: ../../getting_started/install/cluster/cluster.rst:34
#: b2f7bf8f9ef4406989a366d66e59794b
msgid ""
"In DB-GPT, seamless support is provided for FastChat, vLLM, and "
"llama.cpp. In theory, any model supported by these frameworks is also "
"supported by DB-GPT. If you have requirements for faster inference speed "
"and concurrency, you can directly use vLLM. If you want good inference "
"performance on CPU or Apple's M1/M2 chips, you can use llama.cpp. "
"Additionally, DB-GPT also supports various proxy models from OpenAI, "
"Azure OpenAI, Google BARD, Wenxin Yiyan, Tongyi Qianwen, and Zhipu AI, "
"among others."
msgstr ""
"在 DB-GPT 中，直接提供了对 FastChat、vLLM和 llama.cpp 的无缝支持，理论上它们支持模型 DB-GPT "
"都支持，如果您对推理速度和并发能力有需求，可以直接使用 vLLM，如果您希望 CPU 或者 mac 的 "
"m1/m2性能也获得不错的推理性能，可以使用 llama.cpp，此外，DB-GPT 还支持了很多代理模型（openai、azure "
"openai、google bard、文心一言、通义千问和智谱AI等）。"
#: ../../getting_started/install/cluster/cluster.rst:37
#: 9f894d801c364d58814f295222567992
msgid "2.Have good scalability and stability"
msgstr "2.扩展性和稳定性要足够好"
#: ../../getting_started/install/cluster/cluster.rst:38
#: 5423f1f5f0e94804becd3caa500b4046
msgid ""
"A comprehensive model deployment framework consists of several "
"components: the Model Worker, which directly interfaces with the "
"underlying inference frameworks; the Model Controller, which manages and "
"maintains multiple model components; and the Model API, which provides "
"external model serving capabilities."
msgstr ""
"一个比较完善模型部署框架需要多个部分组成，与底层推理框架直接对接的 Model Worker，管理和维护多个模型组件的 Model "
"Controller 以及对外提供模型服务能力的 Model API。"
#: ../../getting_started/install/cluster/cluster.rst:40
#: 6e0948e9a239405ca7d90543569f35fa
msgid ""
"The Model Worker plays a crucial role and needs to be highly extensible. "
"It can be specialized for deploying large language models, embedding "
"models, or other types of models. The choice of Model Worker depends on "
"the deployment environment, such as a regular physical server "
"environment, a Kubernetes environment, or specific cloud environments "
"provided by various cloud service providers."
msgstr ""
"其中 Model Worker 必须要可以扩展，可以是专门部署大语言模型的 Model Worker，也可以是用来部署 Embedding 模型的"
" Model Worker。"
#: ../../getting_started/install/cluster/cluster.rst:42
#: 54eba96c95c847e6af77ba94114419ab
msgid ""
"Having different Model Worker options allows users to select the most "
"suitable one based on their specific requirements and infrastructure. "
"This flexibility enables efficient deployment and utilization of models "
"across different environments."
msgstr "当然也可以根据部署的环境，如普通物理机环境、kubernetes 环境以及一些特定云服务商提供的云环境等来选择不同 Model Worker"
#: ../../getting_started/install/cluster/cluster.rst:44
#: 693ec0c1b9274f64a1d8fcbd5a8a273d
msgid ""
"The Model Controller, responsible for managing model metadata, also needs"
" to be scalable. Different deployment environments and model management "
"requirements may call for different choices of Model Controllers."
msgstr ""
"用来管理模型元数据的 Model Controller 也需要可扩展，不同的部署环境已经不同的模型管控要求来选择不同的 Model "
"Controller。"
#: ../../getting_started/install/cluster/cluster.rst:46
#: 616c0dc43dd84069bde396f1cc99e316
msgid ""
"Furthermore, I believe that model serving shares many similarities with "
"traditional microservices. In microservices, a service can have multiple "
"instances, and all instances are registered in a central registry. "
"Service consumers retrieve the list of instances based on the service "
"name from the registry and select a specific instance for invocation "
"using a load balancing strategy."
msgstr "另外，在我看来，模型服务与传统的微服务有很多共通之处，在微服务中，微服务中某个服务可以有多个服务实例，所有的服务实例都统一注册到注册中心，服务调用方根据服务名称从注册中心拉取该服务名对应的服务列表，然后根据一定的负载均衡策略选择某个具体的服务实例去调用。"
#: ../../getting_started/install/cluster/cluster.rst:48
#: 83389d65894f44598a0eda3984a41cb3
msgid ""
"Similarly, in model deployment, a model can have multiple instances, and "
"all instances can be registered in a model registry. Model service "
"consumers retrieve the list of instances based on the model name from the"
" registry and select a specific instance for invocation using a model-"
"specific load balancing strategy."
msgstr "而在模型部署中，也可以考虑这样的架构，某一个模型可以有多个模型实例，所有的模型实例都统一注册到模型注册中心，然后模型服务调用方根据模型名称到注册中心去拉取模型实例列表，然后根据模型的负载均衡策略去调用某个具体的的模型实例。"
#: ../../getting_started/install/cluster/cluster.rst:50
#: 8524b6f0536446a6900715aaefcdee98
msgid ""
"Introducing a model registry, responsible for storing model instance "
"metadata, enables such an architecture. The model registry can leverage "
"existing service registries used in microservices (such as Nacos, Eureka,"
" etcd, Consul, etc.) as implementations. This ensures high availability "
"of the entire deployment system."
msgstr ""
"这里我们引入模型注册中心，它负责存储 Model Controller 中的模型实例元数据，它可以直接使用微服务中的注册中心作为实现（如 "
"nacos、eureka、etcd 和 consul 等），这样整个部署系统便可以做到高可用。"
#: ../../getting_started/install/cluster/cluster.rst:53
#: ff904ff9192248bda12b5ccae28df26f
msgid "3.High performance for framework."
msgstr "3.框架性能“不拖后腿”"
#: ../../getting_started/install/cluster/cluster.rst:54
#: c6baf9f2a059487bbf7c3996e401effb
msgid ""
"and optimization are complex tasks, and inappropriate framework designs "
"can increase this complexity. In our view, to ensure that the deployment "
"framework does not lag behind in terms of performance, there are two main"
" areas of focus:"
msgstr "框架层不应该成为模型推理性能的瓶颈，大部分情况下，硬件及推理框架决定了模型服务的服务能力，模型的推理部署和优化是一项复杂的工程，而不恰当的框架设计却可能增加这种复杂度，在我们看来，部署框架为了在性能上“不拖后腿”，有两个主要关注点："
#: ../../getting_started/install/cluster/cluster.rst:56
#: f74418d5394b4afd96578c28ff306116
msgid ""
"Avoid excessive encapsulation: The more encapsulation and longer the "
"chain, the more challenging it becomes to identify performance issues."
msgstr "避免过多的封装：封装越多、链路越长，性能问题越难以排查。"
#: ../../getting_started/install/cluster/cluster.rst:58
#: 692bc702a7f54d48b67c36ac1dc38867
msgid ""
"High-performance communication design: High-performance communication "
"involves various aspects that cannot be elaborated in detail here. "
"However, considering that Python occupies a prominent position in current"
" AIGC applications, asynchronous interfaces are crucial for service "
"performance in Python. Therefore, the model serving layer should only "
"provide asynchronous interfaces and be compatible with the layers that "
"interface with the model inference framework. If the model inference "
"framework offers asynchronous interfaces, direct integration should be "
"implemented. Otherwise, synchronous-to-asynchronous task conversion "
"should be used to provide support."
msgstr ""
"高性能的通信设计：高性能通信涉及的点很多，这里不做过多阐述。由于目前 AIGC 应用中，Python 占据领导地位，在 Python "
"中，异步接口对于服务的性能至关重要，因此，模型服务层只提供异步接口，与模型推理框架对接的层做兼容，如果模型推理框架提供了异步接口则直接对接，否则使用同步转异步的任务的方式支持。"
#: ../../getting_started/install/cluster/cluster.rst:61
#: 3b2bed671a264a13a61b7337e4577185
msgid "4.Management and monitoring capabilities."
msgstr "4.具备一定的管理和监控能力"
#: ../../getting_started/install/cluster/cluster.rst:62
#: 010c26d97cb748d28f86d9d58bdb3c6d
msgid ""
"In the exploration or production implementation of AIGC (Artificial "
"Intelligence and General Computing) applications, it is important for the"
" model deployment system to have certain management capabilities. This "
"involves controlling the deployed model instances through APIs or "
"command-line interfaces, such as for online/offline management, "
"restarting, and debugging."
msgstr ""
"在 AIGC 应用探索中或者 AIGC 应用生产落地中，我们需要模型部署系统能具备一定管理能力：通过 API "
"或者命令行等对部署的模型实例进行一定管控（如上线、下线、重启和 debug 等）。"
#: ../../getting_started/install/cluster/cluster.rst:64
#: 49f450fdb5f24d578b4cb8427e57ec15
msgid ""
"Observability is a crucial capability in production systems, and I "
"believe it is equally, if not more, important in AIGC applications. This "
"is because user experiences and interactions with the system are more "
"complex. In addition to traditional observability metrics, we are also "
"interested in user input information and corresponding contextual "
"information, which specific model instance and parameters were invoked, "
"the content and response time of model outputs, user feedback, and more."
msgstr ""
"可观测性是生产系统的一项重要能力，个人认为在 AIGC "
"应用中，可观测性同样重要，甚至更加重要，因为用户的体验、用户与系统的交互行为更复杂，除了传统的观测指标外，我们还更加关心用户的输入信息及其对应的场景上下文信息、调用了哪个模型实例和模型参数、模型输出的内容和响应时间、用户反馈等等。"
#: ../../getting_started/install/cluster/cluster.rst:66
#: 18c940b2e65d4f57ba54b9671ac02254
msgid ""
"By analyzing this information, we can identify performance bottlenecks in"
" model services and gather user experience data (e.g., response latency, "
"problem resolution, and user satisfaction extracted from user content). "
"These insights serve as important foundations for further optimizing the "
"entire application."
msgstr "我们可以从这些信息中发现一部分模型服务的性能瓶颈，以及一部分用户体验数据（响应延迟如何？是否解决了用户的问题也及用户内容中提取出用户满意度等等），这些都是整个应用进一步优化的重要依据。"
#: ../../getting_started/install/cluster/cluster.rst:68
#: a1aa65d7b0694b75a8a298090b3cbfac
#, fuzzy
msgid ""
"On :ref:`Deploying on local machine <local-cluster-index>`. Local cluster"
" deployment."
"On :ref:`Deploying on standalone mode <standalone-index>`. Standalone "
"Deployment."
msgstr "关于 :ref:`在本地机器上部署 <local-cluster-index>`。本地集群部署。"
#: ../../getting_started/install/cluster/cluster.rst:69
#: 2d74de97891c4a31806ce286c3818631
#, fuzzy
msgid ""
"On :ref:`Deploying on cluster mode <local-cluster-index>`. Cluster "
"Deployment."
msgstr "关于 :ref:`在本地机器上部署 <local-cluster-index>`。本地集群部署。"
#~ msgid ""
#~ "When it comes to model deployment, "
#~ "performance is of utmost importance. The"
#~ " framework should be optimized to "
#~ "ensure efficient and fast model "
#~ "inference capabilities. It should not "
#~ "become a performance bottleneck and "
#~ "should be capable of handling high "
#~ "volumes of requests without compromising "
#~ "response times."
#~ msgstr "框架层不应该成为模型推理性能的瓶颈，大部分情况下，硬件及推理框架决定了模型服务的服务能力，模型的推理部署和优化是一项复杂的工程，而不恰当的框架设计却可能增加这种复杂度，在我们看来，部署框架为了在性能上“不拖后腿”，有两个主要关注点："
#~ msgid ""
#~ "To achieve this, the framework can "
#~ "employ various performance optimization "
#~ "techniques. This may include utilizing "
#~ "efficient algorithms, leveraging hardware "
#~ "acceleration (such as GPUs or "
#~ "specialized AI chips), optimizing memory "
#~ "usage, and implementing parallel processing"
#~ " techniques to maximize throughput."
#~ msgstr ""
#~ msgid ""
#~ "By prioritizing performance optimization, the"
#~ " framework can provide seamless and "
#~ "efficient model inference, enabling real-"
#~ "time and high-performance applications "
#~ "without impeding the overall system "
#~ "performance."
#~ msgstr ""
#~ msgid ""
#~ "To ensure the stability and reliability"
#~ " of model deployment, the framework "
#~ "needs to provide management and "
#~ "monitoring functionalities. This includes "
#~ "managing the lifecycle of models, such"
#~ " as model registration, updates, and "
#~ "deletion. Additionally, the framework should"
#~ " offer monitoring and logging of "
#~ "performance metrics, resource utilization, and"
#~ " system health to promptly identify "
#~ "and resolve potential issues."
#~ msgstr ""
#~ "在 AIGC 应用探索中或者 AIGC "
#~ "应用生产落地中，我们需要模型部署系统能具备一定管理能力：通过 API "
#~ "或者命令行等对部署的模型实例进行一定管控（如上线、下线、重启和 debug 等）。"
#~ msgid ""
#~ "Management capabilities may involve user "
#~ "permission management, model versioning, and"
#~ " configuration management to facilitate "
#~ "team collaboration and manage multiple "
#~ "versions and configurations of models."
#~ msgstr ""
#~ msgid ""
#~ "Monitoring capabilities can include real-"
#~ "time monitoring of model performance "
#~ "metrics such as inference latency and"
#~ " throughput. Furthermore, monitoring system "
#~ "resource usage, such as CPU, memory, "
#~ "network, and system health, along with"
#~ " error logging, can be valuable for"
#~ " diagnostics and troubleshooting."
#~ msgstr ""
#~ "可观测性是生产系统的一项重要能力，个人认为在 AIGC "
#~ "应用中，可观测性同样重要，甚至更加重要，因为用户的体验、用户与系统的交互行为更复杂，除了传统的观测指标外，我们还更加关心用户的输入信息及其对应的场景上下文信息、调用了哪个模型实例和模型参数、模型输出的内容和响应时间、用户反馈等等。"
#~ msgid ""
#~ "By providing management and monitoring "
#~ "capabilities, the framework can assist "
#~ "users in effectively managing and "
#~ "maintaining deployed models, ensuring system"
#~ " stability and reliability, and enabling"
#~ " timely responses to and resolution "
#~ "of issues, thus enhancing overall system"
#~ " efficiency and availability."
#~ msgstr ""

66
docs/locales/zh_CN/LC_MESSAGES/getting_started/install/cluster/vms/index.po
View File

@ -8,7 +8,7 @@ msgid ""
msgstr ""
"Project-Id-Version: DB-GPT 👏👏 0.3.6\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2023-09-20 17:34+0800\n"
"POT-Creation-Date: 2023-10-17 17:24+0800\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language: zh_CN\n"
@ -20,22 +20,23 @@ msgstr ""
"Generated-By: Babel 2.12.1\n"
#: ../../getting_started/install/cluster/vms/index.md:1
#: 48c062c146cd42b48c248ae590d386df
msgid "Local cluster deployment"
msgstr "本地集群部署"
#: b23e82d177c443ca8a36b94343ce2173
#, fuzzy
msgid "Cluster Deployment"
msgstr "模型集群部署"
#: ../../getting_started/install/cluster/vms/index.md:4
#: ce59bbbc9c294cafa6df8165de61967f
#: 47ba242f687a41438f1fa41febbe81a3
msgid "Model cluster deployment"
msgstr "模型集群部署"
#: ../../getting_started/install/cluster/vms/index.md:7
#: 51650b41f4974f819a623db1e97764c7
#: 077917ec4fa940689ec2e08e3a000578
msgid "**Installing Command-Line Tool**"
msgstr "**安装命令行工具**"
#: ../../getting_started/install/cluster/vms/index.md:9
#: 64fcb0e3ec8d491aa9d15f783823e579
#: ad498ea7e59f4126838d0a6760da41a3
#, fuzzy
msgid ""
"All operations below are performed using the `dbgpt` command. To use the "
@ -47,129 +48,132 @@ msgstr ""
".`。或者，您可以使用 `python pilot/scripts/cli_scripts.py` 作为 `dbgpt` 命令的替代。"
#: ../../getting_started/install/cluster/vms/index.md:11
#: 572f2d79178a4e6780799dd8bc0867f9
#: 33e6fa8572054ed1b7e92e14487ef044
msgid "Launch Model Controller"
msgstr "启动 Model Controller"
#: ../../getting_started/install/cluster/vms/index.md:17
#: 66cfeb3d834c4f7b87bb3180ae447203
#: 2016f7400d9c4013a2da40e3ecfbe02c
msgid "By default, the Model Controller starts on port 8000."
msgstr "默认情况下，Model Controller 启动在 8000 端口。"
#: ../../getting_started/install/cluster/vms/index.md:20
#: cddb3dbc31734462b6aa3c63e3c76fe2
#: 82338f543db649c1adc2dc57867e2094
msgid "Launch LLM Model Worker"
msgstr "启动 LLM Model Worker"
#: ../../getting_started/install/cluster/vms/index.md:22
#: 953eeafd791942e895833bce2a4d755f
#: 49c2a89381be4fdda17d3cb002899d1f
msgid "If you are starting `chatglm2-6b`:"
msgstr "如果您启动的是 `chatglm2-6b`："
#: ../../getting_started/install/cluster/vms/index.md:31
#: 779d8daa394b4731bc74a93c077961e1
#: 5c8b223521d640d9a18b169924225510
msgid "If you are starting `vicuna-13b-v1.5`:"
msgstr "如果您启动的是 `vicuna-13b-v1.5`："
#: ../../getting_started/install/cluster/vms/index.md:40
#: ../../getting_started/install/cluster/vms/index.md:53
#: 736b34df46e640fbbf3eb41ff5f44cc2 b620ee13d10748e6a89c67a9bfb5a53b
#: 1ad98a11e3f6488cad3d6f7349d4ff70 64b71a7581c34a0d9ba0c9455167b81d
msgid "Note: Be sure to use your own model name and model path."
msgstr "注意：确保使用您自己的模型名称和模型路径。"
#: ../../getting_started/install/cluster/vms/index.md:42
#: d1f48ab4090d4344aa2a010cdc88a28e
#: 5929f47166b241fa9988f1ecb1e45186
msgid "Launch Embedding Model Worker"
msgstr "启动 Embedding Model Worker"
#: ../../getting_started/install/cluster/vms/index.md:55
#: 0b4c6d2ff51c4167b553a6255ce268ba
#: db56788d6758451a823f5b1c91719b56
msgid "Check your model:"
msgstr "检查您的模型："
#: ../../getting_started/install/cluster/vms/index.md:61
#: defd23cef23a4e74b150b7b49b99d333
#: e0dae6b3b0c84b5ba24194dffee8c919
msgid "You will see the following output:"
msgstr "您将看到以下输出："
#: ../../getting_started/install/cluster/vms/index.md:75
#: aaa86e08b60e46ddae52a03f25812f24
#: 9806216c698b44909b3664c72cc09710
msgid "Connect to the model service in the webserver (dbgpt_server)"
msgstr "在 webserver (dbgpt_server) 中连接到模型服务 (dbgpt_server)"
#: ../../getting_started/install/cluster/vms/index.md:77
#: 7fd7b622b2f649d0b7d9b51a998a038c
#: 25fb95f7850a4b0e90f6d949bf440f86
msgid ""
"**First, modify the `.env` file to change the model name and the Model "
"Controller connection address.**"
msgstr "**首先，修改 `.env` 文件以更改模型名称和模型控制器连接地址。**"
#: ../../getting_started/install/cluster/vms/index.md:85
#: 7ce03ec66f624d0eabd5a2fbe2efcbcc
#: 4f66546f32934c5080ca5b7044eeffb8
msgid "Start the webserver"
msgstr "启动 webserver"
#: ../../getting_started/install/cluster/vms/index.md:91
#: 9e1e2b7925834d6b9140633db1082032
#: 4cc99c718b6c470e93d3e5016cdb5be9
msgid "`--light` indicates not to start the embedded model service."
msgstr "`--light` 表示不启动嵌入式模型服务。"
#: ../../getting_started/install/cluster/vms/index.md:93
#: 4d47f76763914a78a89d62f0befa3fd9
#: 4242d989fec249c98a53bdf8a776a103
msgid ""
"Alternatively, you can prepend the command with `LLM_MODEL=chatglm2-6b` "
"to start:"
msgstr "或者，您可以在命令前加上 `LLM_MODEL=chatglm2-6b` 来启动："
#: ../../getting_started/install/cluster/vms/index.md:100
#: 28408fe554dd411c9ca672466d5563b6
#: b50e829504b24d64ac9bb3c96bba0271
msgid "More Command-Line Usages"
msgstr "更多命令行用法"
#: ../../getting_started/install/cluster/vms/index.md:102
#: 8e0aa88d092d49fdb0fa849c83565a41
#: 332f11f9f2f24039a0e512cac2672ded
msgid "You can view more command-line usages through the help command."
msgstr "您可以通过帮助命令查看更多命令行用法。"
#: ../../getting_started/install/cluster/vms/index.md:104
#: f307f82ced7947f980bb65b3543580d1
#: 89676ed8d92d4d008183aff4c156bcfe
msgid "**View the `dbgpt` help**"
msgstr "**查看 `dbgpt` 帮助**"
#: ../../getting_started/install/cluster/vms/index.md:109
#: 7564aed77a7d43e6878b506c6a9788a2
#: 384b15e4ee434026814f72044f2eae20
msgid "You will see the basic command parameters and usage:"
msgstr "您将看到基本的命令参数和用法："
#: ../../getting_started/install/cluster/vms/index.md:127
#: 569f2b9e62e44179ae8dcf5b05a1f3e8
#: 342307aee3a74b5b80c948a53ec4c99f
msgid "**View the `dbgpt start` help**"
msgstr "**查看 `dbgpt start` 帮助**"
#: ../../getting_started/install/cluster/vms/index.md:133
#: 43e8747d136d4f6cab83c1b1beaa32b0
#: 7a3e29aa9caf49ac885e5842606a3d00
msgid "Here you can see the related commands and usage for start:"
msgstr "在这里，您可以看到启动的相关命令和用法："
#: ../../getting_started/install/cluster/vms/index.md:150
#: 25e37b3050b348ec9a5c96d9db515e9b
#: d64bac2b25ec4c619f74a7209e634ff3
msgid "**View the `dbgpt start worker`help**"
msgstr "**查看 `dbgpt start worker` 帮助**"
#: ../../getting_started/install/cluster/vms/index.md:156
#: 8e959fd455ca45f9a5e69e0af9b764a4
#: 4bb1293ffd6e40f7923943f62c452925
msgid "Here you can see the parameters to start Model Worker:"
msgstr "在这里，您可以看到启动 Model Worker 的参数："
#: ../../getting_started/install/cluster/vms/index.md:215
#: 374d274c7a254533900145ef17bb24fb
#: 110dd6d71c2845afbe8550d1de9393de
msgid "**View the `dbgpt model`help**"
msgstr "**查看 `dbgpt model` 帮助**"
#: ../../getting_started/install/cluster/vms/index.md:221
#: 19bcc9abe62f490d9c3c092c5deea24a
#: b7ac90dffb84457f8dd87a531ddb72a2
msgid ""
"The `dbgpt model ` command can connect to the Model Controller via the "
"Model Controller address and then manage a remote model:"
msgstr "`dbgpt model` 命令可以通过 Model Controller 地址连接到 Model Controller，然后管理远程模型："
#~ msgid "Local cluster deployment"
#~ msgstr "本地集群部署"

276
docs/locales/zh_CN/LC_MESSAGES/getting_started/install/deploy/deploy.po
View File

@ -8,7 +8,7 @@ msgid ""
msgstr ""
"Project-Id-Version: DB-GPT 👏👏 0.3.5\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2023-08-29 20:50+0800\n"
"POT-Creation-Date: 2023-10-17 14:35+0800\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language: zh_CN\n"
@ -20,212 +20,249 @@ msgstr ""
"Generated-By: Babel 2.12.1\n"
#: ../../getting_started/install/deploy/deploy.md:1
#: b4f766ca21d241e2849ee0a277a0e8f0
#: 73f932b662564edba45fbd711fd19005
msgid "Installation From Source"
msgstr "源码安装"
#: ../../getting_started/install/deploy/deploy.md:3
#: 9cf72ef201ba4c7a99da8d7de9249cf4
#: 70b623827a26447cb9382f1cb568b93c
msgid ""
"This tutorial gives you a quick walkthrough about use DB-GPT with you "
"environment and data."
msgstr "本教程为您提供了关于如何使用DB-GPT的使用指南。"
#: ../../getting_started/install/deploy/deploy.md:5
#: b488acb9552043df96e9f01277375b56
#: 6102ada4b19a4062947ad0ee5305dad5
msgid "Installation"
msgstr "安装"
#: ../../getting_started/install/deploy/deploy.md:7
#: e1eb3aafea0c4b82b8d8163b947677dd
#: 7c006c0c72944049bba43fd95daf1bd1
msgid "To get started, install DB-GPT with the following steps."
msgstr "请按照以下步骤安装DB-GPT"
#: ../../getting_started/install/deploy/deploy.md:9
#: 4139c4e62e874dc58136b1f8fe0715fe
#: eac8c7f921a042b79b4d0032c01b095a
msgid "1. Hardware Requirements"
msgstr "1. 硬件要求"
#: ../../getting_started/install/deploy/deploy.md:10
#: c34a204cfa6e4973bfd94e683195c17b
#: 8c430e2db5ce41e8b9d22e6e13c62cb3
msgid ""
"As our project has the ability to achieve ChatGPT performance of over "
"85%, there are certain hardware requirements. However, overall, the "
"project can be deployed and used on consumer-grade graphics cards. The "
"specific hardware requirements for deployment are as follows:"
msgstr "由于我们的项目有能力达到85%以上的ChatGPT性能，所以对硬件有一定的要求。但总体来说，我们在消费级的显卡上即可完成项目的部署使用，具体部署的硬件说明如下:"
"DB-GPT can be deployed on servers with low hardware requirements or on "
"servers with high hardware requirements."
msgstr "DB-GPT可以部署在对硬件要求不高的服务器，也可以部署在对硬件要求高的服务器"
#: ../../getting_started/install/deploy/deploy.md:12
#: a6b042509e1149fa8213a014e42eaaae
#, fuzzy
msgid "Low hardware requirements"
msgstr "1. 硬件要求"
#: ../../getting_started/install/deploy/deploy.md:13
#: 577c8c4edc2e4f45963b2a668385852f
msgid ""
"The low hardware requirements mode is suitable for integrating with "
"third-party LLM services' APIs, such as OpenAI, Tongyi, Wenxin, or "
"Llama.cpp."
msgstr "Low hardware requirements模式适用于对接第三方模型服务的api,比如OpenAI, 通义千问, 文心.cpp。"
#: ../../getting_started/install/deploy/deploy.md:15
#: 384475d3a87043eb9eebc384052ac9cc
msgid "DB-GPT provides set proxy api to support LLM api."
msgstr "DB-GPT可以通过设置proxy api来支持第三方大模型服务"
#: ../../getting_started/install/deploy/deploy.md:17
#: e5bd8a999adb4e07b8b5221f1893251d
msgid "As our project has the ability to achieve ChatGPT performance of over 85%,"
msgstr "由于我们的项目有能力达到85%以上的ChatGPT性能"
#: ../../getting_started/install/deploy/deploy.md:19
#: 6a97ed5893414e17bb9c1f8bb21bc965
#, fuzzy
msgid "High hardware requirements"
msgstr "1. 硬件要求"
#: ../../getting_started/install/deploy/deploy.md:20
#: d0c248939b4143a2b01afd051b02ec12
#, fuzzy
msgid ""
"The high hardware requirements mode is suitable for independently "
"deploying LLM services, such as Llama series models, Baichuan, ChatGLM, "
"Vicuna, and other private LLM service. there are certain hardware "
"requirements. However, overall, the project can be deployed and used on "
"consumer-grade graphics cards. The specific hardware requirements for "
"deployment are as follows:"
msgstr "High hardware requirements模式适用于需要独立部署私有大模型服务，比如Llama系列模型，Baichuan, chatglm，vicuna等私有大模型所以对硬件有一定的要求。但总体来说，我们在消费级的显卡上即可完成项目的部署使用，具体部署的硬件说明如下:"
#: ../../getting_started/install/deploy/deploy.md
#: 3a92203e861b42c9af3d4b687d83de5e
#: 2ee432394f6b4d9cb0a424f4b99bf3be
msgid "GPU"
msgstr "GPU"
#: ../../getting_started/install/deploy/deploy.md
#: 6050741571574eb8b9e498a5b3a7e347 c0a7e2aecb4b48949c3e5a4d479ee7b5
#: 4cd716486f994080880f84853b047a5d
msgid "VRAM Size"
msgstr "显存"
#: ../../getting_started/install/deploy/deploy.md
#: 247159f568e4476ca6c5e78015c7a8f0
#: d1b33d0348894bfc8a843a3d38c6daaa
msgid "Performance"
msgstr "Performance"
#: ../../getting_started/install/deploy/deploy.md
#: 871113cbc58743ef989a366b76e8c645
#: d5850bbe7d0a430d993b7e6bd1f24bff
msgid "RTX 4090"
msgstr "RTX 4090"
#: ../../getting_started/install/deploy/deploy.md
#: 81327b7e9a984ec99cae779743d174df c237f392162c42d28ec694d17c3f281c
#: c7d15be08ac74624bbfb5eb4554fc7ff
msgid "24 GB"
msgstr "24 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 6e19f23bae05467ba03f1ebb194e0c03
#: 219dff2fee83460da55d9d628569365e
msgid "Smooth conversation inference"
msgstr "Smooth conversation inference"
msgstr "丝滑的对话体验"
#: ../../getting_started/install/deploy/deploy.md
#: 714a48b2c4a943819819a6af034f1998
#: 56025c5f37984963943de7accea85850
msgid "RTX 3090"
msgstr "RTX 3090"
#: ../../getting_started/install/deploy/deploy.md
#: 06dae55d443c48b1b3fbab85222c3adb
#: 8a0b8a0afa0c4cc39eb7c2271775cf60
msgid "Smooth conversation inference, better than V100"
msgstr "Smooth conversation inference, better than V100"
msgstr "丝滑的对话体验，性能好于V100"
#: ../../getting_started/install/deploy/deploy.md
#: 5d50db167b244d65a8be1dab4acda37d
#: 2fc5e6ac8a6b4c508944c659adffa0c1
msgid "V100"
msgstr "V100"
#: ../../getting_started/install/deploy/deploy.md
#: 0d72262c85d148d8b1680d1d9f8fa2c9 e10db632889444a78e123773a30f23cf
#: f92a1393539a49db983b06f7276f446b
msgid "16 GB"
msgstr "16 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 1c0379e653cf46f19d83535c568c54c8 aee8eb48e7804572af351dcfaea5b0fb
#: 4e7de52a58d24a0bb10e45e1435128a6
msgid "Conversation inference possible, noticeable stutter"
msgstr "Conversation inference possible, noticeable stutter"
#: ../../getting_started/install/deploy/deploy.md
#: 5bc90343dcef48c197438f01efe52bfc
#: 217fe55f590a497ba6622698945e7be8
msgid "T4"
msgstr "T4"
#: ../../getting_started/install/deploy/deploy.md:19
#: c9b5f973d19645d39b1892c00526afa7
#: ../../getting_started/install/deploy/deploy.md:30
#: 30ca67fe27f64df093a2d281e1288c5c
#, fuzzy
msgid ""
"if your VRAM Size is not enough, DB-GPT supported 8-bit quantization and "
"If your VRAM Size is not enough, DB-GPT supported 8-bit quantization and "
"4-bit quantization."
msgstr "如果你的显存不够，DB-GPT支持8-bit和4-bit量化版本"
#: ../../getting_started/install/deploy/deploy.md:21
#: 5e488271eede411d882f62ec8524dd4a
#: ../../getting_started/install/deploy/deploy.md:32
#: fc0c3a0730d64e9e98d1b25f4dd5db34
msgid ""
"Here are some of the VRAM size usage of the models we tested in some "
"common scenarios."
msgstr "这里是量化版本的相关说明"
#: ../../getting_started/install/deploy/deploy.md
#: 2cc65f16fa364088bedd0e58b6871ec8
#: 1f1f6c10209b446f99d520fdb68e0f5d
msgid "Model"
msgstr "Model"
#: ../../getting_started/install/deploy/deploy.md
#: d0e1a0d418f74e4b9f5922b17f0c8fcf
#: 18e3240d407e41f88028b24aeced1bf4
msgid "Quantize"
msgstr "Quantize"
#: ../../getting_started/install/deploy/deploy.md
#: 460b418ab7eb402eae7a0f86d1fda4bf 5e456423a9fa4c0392b08d32f3082f6f
#: 03aa79d3c3f54e3c834180b0d1ed9a5c
msgid "vicuna-7b-v1.5"
msgstr "vicuna-7b-v1.5"
#: ../../getting_started/install/deploy/deploy.md
#: 0f290c12b9324a07affcfd66804b82d7 29c81ce163e749b99035942a3b18582a
#: 3a4f4325774d452f8c174cac5fe8de47 584f986a1afb4086a0382a9f7e79c55f
#: 994c744ac67249f4a43b3bba360c0bbf aa9c82f660454143b9212842ffe0e0d6
#: ac7b00313284410b9253c4a768a30f0c
#: 09419ad0a88c4179979505ef71204fd6 1b4ab0186184493d895eeec12d078c52
#: 6acec7b76e604343885aa71d92b04d1e 9b73ca1c18d14972b894db69438e3fb2
#: b869995505ae4895b9f13e271470e5cb c9eaf983eeb2486da08e628728ae301f
#: ff0a86dc63ce4cd580f354d15d333501
msgid "4-bit"
msgstr "4-bit"
#: ../../getting_started/install/deploy/deploy.md
#: 27401cbb0f2542e2aaa449a586aad2d1 2a1d2d10001f4d9f9b9961c28c592280
#: b69a59c6e4a7458c91be814a98502632
#: d0d959f022f44bbeb34d67ccf49ba3bd
msgid "8 GB"
msgstr "8 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 0a15518df1b94492b610e47f3c7bb4f6 1f1852ceae0b4c21a020dc9ef4f8b20b
#: 89ad803f6bd24b5d9708a6d4bd48a54f ac7c222678d34637a03546dcb5949668
#: b12e1599bdcb4d27ad4e4a83f12de916 c80ba4ddc1634093842a6f284b7b22bb
#: f63b900e4b844b3196c4c221b36d31f7
#: 01cb7be0064940e8a637df7ed8e15310 13568d8a793b4c2db655f89dc690929a
#: 28ce31711f91455b9b910276fa059c65 2dddf2e87a70452fb27a627d62464346
#: 3f3f4dc00acb43258dce311f144e0fd7 5aa76fd2fb35474e8d06795e7369ceb4
#: d660be499efc4b6ca61da0d5af758620
msgid "8-bit"
msgstr "8-bit"
#: ../../getting_started/install/deploy/deploy.md
#: 02f72ed48b784b05b2fcaf4ea33fcba8 17285314376044bf9d9a82f9001f39dc
#: 403178173a784bdf8d02fe856849a434 4875c6b595484091b622602d9ef0d3e8
#: 4b11125d4b0c40c488bffb130f4f2b9f e2418c76e7e04101821f29650d111a4a
#: 3b963a1ce6934229ba7658cb407b6a52
msgid "12 GB"
msgstr "12 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 01dfd16f70cf4128a49ca7bc79f77042 a615efffecb24addba759d05ef61a1c0
#: 30d28dcaa64545198aaa20fe4562bb6d
msgid "vicuna-13b-v1.5"
msgstr "vicuna-13b-v1.5"
#: ../../getting_started/install/deploy/deploy.md
#: 412ddfa6e6fb4567984f757cf74b3bfc 529650341d96466a93153d58ddef0ec9
#: 6176929d59bb4e31a37cbba8a81a489f
#: 28de25a1952049d2b7aff41020e428ff
msgid "20 GB"
msgstr "20 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 566b7aa7bc88421a9364cef6bfbeae48 ae32a218d07e44c796ca511972ea2cb0
#: 535974c886b14c618ca84de1fe63d5e4
msgid "llama-2-7b"
msgstr "llama-2-7b"
#: ../../getting_started/install/deploy/deploy.md
#: 1ac748eb518b4017accb98873fe1a8e5 528109c765e54b3caf284e7794abd468
#: cc04760a8b9e4a79a7dada9a11abda2c
msgid "llama-2-13b"
msgstr "llama-2-13b"
#: ../../getting_started/install/deploy/deploy.md
#: dfb5c0fa9e82423ab1de9256b3b3f215 f861be75871d40849f896859d0b8be4c
#: 9e83d8d5ae44411dba4cc6c2d796b20f
msgid "llama-2-70b"
msgstr "llama-2-70b"
#: ../../getting_started/install/deploy/deploy.md
#: 5568529a82cd4c49812ab2fd46ff9bf0
#: cb6ce389adfc463a9c851eb1e4abfcff
msgid "48 GB"
msgstr "48 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 4ba730f4faa64df9a0a9f72cb3eb0c88
#: 906d664156084223a4efa0ae9804bd33
msgid "80 GB"
msgstr "80 GB"
#: ../../getting_started/install/deploy/deploy.md
#: 47221748d6d5417abc25e28b6905bc6f 6023d535095a4cb9a99343c2dfddc927
#: 957fb0c6f3114a63ba33a1cfb31060e3
msgid "baichuan-7b"
msgstr "baichuan-7b"
#: ../../getting_started/install/deploy/deploy.md
#: 55011d4e0bed451dbdda75cb8b258fa5 bc296e4bd582455ca64afc74efb4ebc8
#: 5f3bc4cf57d946cfb38a941250685151
msgid "baichuan-13b"
msgstr "baichuan-13b"
#: ../../getting_started/install/deploy/deploy.md:40
#: 4bfd52634a974776933c93227f419cdb
#: ../../getting_started/install/deploy/deploy.md:51
#: 87ae8c58df314b69ae119aa831cb7dd5
msgid "2. Install"
msgstr "2. Install"
#: ../../getting_started/install/deploy/deploy.md:45
#: 647f09001d4c4124bed11da272306946
#: ../../getting_started/install/deploy/deploy.md:56
#: 79cdebf089614761bf4299a9ce601b81
msgid ""
"We use Sqlite as default database, so there is no need for database "
"installation. If you choose to connect to other databases, you can "
@ -239,49 +276,49 @@ msgstr ""
"GPT快速部署不需要部署相关数据库服务。如果你想使用其他数据库，需要先部署相关数据库服务。我们目前使用Miniconda进行python环境和包依赖管理[安装"
" Miniconda](https://docs.conda.io/en/latest/miniconda.html)"
#: ../../getting_started/install/deploy/deploy.md:54
#: bf9fcf320ca94dbd855016088800b1a9
#: ../../getting_started/install/deploy/deploy.md:65
#: 03ff2f444721454588095bb348220276
msgid "Before use DB-GPT Knowledge"
msgstr "在使用知识库之前"
#: ../../getting_started/install/deploy/deploy.md:60
#: e0cb6cb46a474c4ca16edf73c82b58ca
#: ../../getting_started/install/deploy/deploy.md:71
#: b6faa4d078a046d6a7c0313e8deef0f3
msgid ""
"Once the environment is installed, we have to create a new folder "
"\"models\" in the DB-GPT project, and then we can put all the models "
"downloaded from huggingface in this directory"
msgstr "如果你已经安装好了环境需要创建models, 然后到huggingface官网下载模型"
#: ../../getting_started/install/deploy/deploy.md:63
#: 03b1bf35528d4cdeb735047aa840d6fe
#: ../../getting_started/install/deploy/deploy.md:74
#: f43fd2b74d994bf6bb4016e88c43d51a
msgid "Notice make sure you have install git-lfs"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:65
#: f8183907e7c044f695f86943b412d84a
#: ../../getting_started/install/deploy/deploy.md:76
#: f558a7ee728a4344af576aa375b43092
msgid "centos:yum install git-lfs"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:67
#: 3bc042bd5cac4007afc9f68e7b5044fe
msgid "ubuntu:app-get install git-lfs"
#: ../../getting_started/install/deploy/deploy.md:78
#: bab08604a3ba45b9b827ff5a4b931601
msgid "ubuntu:apt-get install git-lfs"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:69
#: 5915ed1290e84ed9b6782c6733d88891
#: ../../getting_started/install/deploy/deploy.md:80
#: b4a107e5f8524acc9aed74318880f9f3
msgid "macos:brew install git-lfs"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:86
#: 104f1e75b0a54300af440ca3b64217a3
#: ../../getting_started/install/deploy/deploy.md:97
#: ecb5fa1f18154685bb4336d04ac3a386
msgid ""
"The model files are large and will take a long time to download. During "
"the download, let's configure the .env file, which needs to be copied and"
" created from the .env.template"
msgstr "模型文件很大，需要很长时间才能下载。在下载过程中，让我们配置.env文件，它需要从。env.template中复制和创建。"
#: ../../getting_started/install/deploy/deploy.md:88
#: 228c6729c23f4e17b0475b834d7edb01
#: ../../getting_started/install/deploy/deploy.md:99
#: 0f08b0ecbea14cbdba29ea8d87cf24b4
msgid ""
"if you want to use openai llm service, see [LLM Use FAQ](https://db-"
"gpt.readthedocs.io/en/latest/getting_started/faq/llm/llm_faq.html)"
@ -289,20 +326,20 @@ msgstr ""
"如果想使用openai大模型服务, 可以参考[LLM Use FAQ](https://db-"
"gpt.readthedocs.io/en/latest/getting_started/faq/llm/llm_faq.html)"
#: ../../getting_started/install/deploy/deploy.md:91
#: c444514ba77b46468721888fe7df9e74
#: ../../getting_started/install/deploy/deploy.md:102
#: 6efb9a45ab2c45c7b4770f987b639c52
msgid "cp .env.template .env"
msgstr "cp .env.template .env"
#: ../../getting_started/install/deploy/deploy.md:94
#: 1514e937757e461189b369da73884a6c
#: ../../getting_started/install/deploy/deploy.md:105
#: b9d2b81a2cf440c3b49a5c06759eb2ba
msgid ""
"You can configure basic parameters in the .env file, for example setting "
"LLM_MODEL to the model to be used"
msgstr "您可以在.env文件中配置基本参数，例如将LLM_MODEL设置为要使用的模型。"
#: ../../getting_started/install/deploy/deploy.md:96
#: 4643cdf76bd947fdb86fc4691b98935c
#: ../../getting_started/install/deploy/deploy.md:107
#: 2f6afa40ca994115b16ba28baaf65bde
msgid ""
"([Vicuna-v1.5](https://huggingface.co/lmsys/vicuna-13b-v1.5) based on "
"llama-2 has been released, we recommend you set `LLM_MODEL=vicuna-"
@ -312,51 +349,51 @@ msgstr ""
"/vicuna-13b-v1.5)， "
"目前Vicuna-v1.5模型(基于llama2)已经开源了，我们推荐你使用这个模型通过设置LLM_MODEL=vicuna-13b-v1.5"
#: ../../getting_started/install/deploy/deploy.md:98
#: acf91810f12b4ad0bd830299eb24850f
#: ../../getting_started/install/deploy/deploy.md:109
#: 7c5883f9594646198f464e6dafb2f0ff
msgid "3. Run"
msgstr "3. Run"
#: ../../getting_started/install/deploy/deploy.md:100
#: ea82d67451724c2399f8903ea3c52dff
#: ../../getting_started/install/deploy/deploy.md:111
#: 0e3719a238eb4332b7c15efa3f16e3e2
msgid "**(Optional) load examples into SQLlite**"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:105
#: a00987ec21364389b7feec58b878c2a1
#: ../../getting_started/install/deploy/deploy.md:116
#: c901055131ce4688b1c602393913b675
msgid "On windows platform:"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:110
#: db5c000e6abe4e1cb94e6f4f14247eb7
#: ../../getting_started/install/deploy/deploy.md:121
#: 777a50f9167c4b8f9c2a96682ccc4c4a
msgid "1.Run db-gpt server"
msgstr "1.Run db-gpt server"
#: ../../getting_started/install/deploy/deploy.md:116
#: dbeecff230174132b85d1d4549d3c07e
#: ../../getting_started/install/deploy/deploy.md:127
#: 62aafb652df8478281ab633d8d082e7f
msgid "Open http://localhost:5000 with your browser to see the product."
msgstr "打开浏览器访问http://localhost:5000"
#: ../../getting_started/install/deploy/deploy.md:119
#: 22d6321e6226472e878a95d3c8a9aad8
#: ../../getting_started/install/deploy/deploy.md:130
#: cff18fc20ffd4716bc7cf377730dd5ec
msgid "If you want to access an external LLM service, you need to"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:121
#: 561dfe9a864540d6ac582f0977b2c9ad
#: ../../getting_started/install/deploy/deploy.md:132
#: f27c3aa9e627480a96cd04fcd4bfdaec
msgid ""
"1.set the variables LLM_MODEL=YOUR_MODEL_NAME, "
"MODEL_SERVER=YOUR_MODEL_SERVER（eg:http://localhost:5000） in the .env "
"file."
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:123
#: 55ceca48e40147a99ab4d23392349156
#: ../../getting_started/install/deploy/deploy.md:134
#: e05a395f67924514929cd025fab67e44
msgid "2.execute dbgpt_server.py in light mode"
msgstr ""
#: ../../getting_started/install/deploy/deploy.md:126
#: 02d42956a2734c739ad1cb9ce59142ce
#: ../../getting_started/install/deploy/deploy.md:137
#: a5d7fcb46ba446bf9913646b28b036ed
msgid ""
"If you want to learn about dbgpt-webui, read https://github./csunny/DB-"
"GPT/tree/new-page-framework/datacenter"
@ -364,55 +401,55 @@ msgstr ""
"如果你想了解web-ui, 请访问https://github./csunny/DB-GPT/tree/new-page-"
"framework/datacenter"
#: ../../getting_started/install/deploy/deploy.md:132
#: d813eb43b97445a08e058d336249e6f6
#: ../../getting_started/install/deploy/deploy.md:143
#: 90c614e7744c4a7f843adb8968b58c78
#, fuzzy
msgid "Multiple GPUs"
msgstr "4. Multiple GPUs"
#: ../../getting_started/install/deploy/deploy.md:134
#: 0ac795f274d24de7b37f9584763e113d
#: ../../getting_started/install/deploy/deploy.md:145
#: 7b72e7cbd9d246299de5986772df4825
msgid ""
"DB-GPT will use all available gpu by default. And you can modify the "
"setting `CUDA_VISIBLE_DEVICES=0,1` in `.env` file to use the specific gpu"
" IDs."
msgstr "DB-GPT默认加载可利用的gpu，你也可以通过修改 在`.env`文件 `CUDA_VISIBLE_DEVICES=0,1`来指定gpu IDs"
#: ../../getting_started/install/deploy/deploy.md:136
#: 2be557e2b5414d478d375bce0474558d
#: ../../getting_started/install/deploy/deploy.md:147
#: b7e2f7bbf625464489b3fd9aedb0ed59
msgid ""
"Optionally, you can also specify the gpu ID to use before the starting "
"command, as shown below:"
msgstr "你也可以指定gpu ID启动"
#: ../../getting_started/install/deploy/deploy.md:146
#: 222f1ebb5cb64675a0c319552d14303e
#: ../../getting_started/install/deploy/deploy.md:157
#: 69fd2183a143428fb77949f58381d455
msgid ""
"You can modify the setting `MAX_GPU_MEMORY=xxGib` in `.env` file to "
"configure the maximum memory used by each GPU."
msgstr "同时你可以通过在.env文件设置`MAX_GPU_MEMORY=xxGib`修改每个GPU的最大使用内存"
#: ../../getting_started/install/deploy/deploy.md:148
#: fb92349f9fe049d5b23b9ead17caf895
#: ../../getting_started/install/deploy/deploy.md:159
#: 6cd03b9728f943a4a632aa9b061931f0
#, fuzzy
msgid "Not Enough Memory"
msgstr "5. Not Enough Memory"
#: ../../getting_started/install/deploy/deploy.md:150
#: 30a1105d728a474c9cd14638feab4b59
#: ../../getting_started/install/deploy/deploy.md:161
#: 4837aba4c80b42819c1a6345de0aa820
msgid "DB-GPT supported 8-bit quantization and 4-bit quantization."
msgstr "DB-GPT 支持 8-bit quantization 和 4-bit quantization."
#: ../../getting_started/install/deploy/deploy.md:152
#: eb2e576379434bfa828c98ee374149f5
#: ../../getting_started/install/deploy/deploy.md:163
#: c1a701e9bc4c4439adfb930d0e953cec
msgid ""
"You can modify the setting `QUANTIZE_8bit=True` or `QUANTIZE_4bit=True` "
"in `.env` file to use quantization(8-bit quantization is enabled by "
"default)."
msgstr "你可以通过在.env文件设置`QUANTIZE_8bit=True` or `QUANTIZE_4bit=True`"
#: ../../getting_started/install/deploy/deploy.md:154
#: eeaecfd77d8546a6afc1357f9f1684bf
#: ../../getting_started/install/deploy/deploy.md:165
#: 205c101f1f774130a5853dd9b7373d36
msgid ""
"Llama-2-70b with 8-bit quantization can run with 80 GB of VRAM, and 4-bit"
" quantization can run with 48 GB of VRAM."
@ -468,3 +505,6 @@ msgstr ""
#~ "注意，需要安装[requirements.txt](https://github.com/eosphoros-ai/DB-"
#~ "GPT/blob/main/requirements.txt)涉及的所有的依赖"
#~ msgid "ubuntu:app-get install git-lfs"
#~ msgstr ""

101
docs/locales/zh_CN/LC_MESSAGES/use_cases/tool_use_with_plugin.po
View File

@ -8,7 +8,7 @@ msgid ""
msgstr ""
"Project-Id-Version: DB-GPT 0.3.0\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2023-06-14 21:47+0800\n"
"POT-Creation-Date: 2023-10-17 17:24+0800\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language: zh_CN\n"
@ -19,86 +19,46 @@ msgstr ""
"Content-Transfer-Encoding: 8bit\n"
"Generated-By: Babel 2.12.1\n"
#: ../../use_cases/tool_use_with_plugin.md:1 584817fdb00047de8f8d7ae02ce86783
#: ../../use_cases/tool_use_with_plugin.md:1 1fb6c590034347ff9bf374dcf0a63fd3
msgid "Tool use with plugin"
msgstr "插件工具"
#: ../../use_cases/tool_use_with_plugin.md:3 74d688e857ee4afe9237aa959238d3df
#: ../../use_cases/tool_use_with_plugin.md:3 b48206ede79641fdabb3afd4c5f7fa7e
msgid ""
"DB-GPT supports a variety of plug-ins, such as MySQL, MongoDB, ClickHouse"
" and other database tool plug-ins. In addition, some database management "
"platforms can also package their interfaces and package them into plug-"
"ins, and use the model to realize the ability of \"single-sentence "
"requirements\""
"DB-GPT supports a variety of plug-ins, such as BaiduSearch, SendEmail. In"
" addition, some database management platforms can also package their "
"interfaces and package them into plug-ins, and use the model to realize "
"the ability of \"single-sentence requirements\""
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:6 55754e6c89d149cd9eb5f935fd9dc761
msgid "DB-GPT-DASHBOARD-PLUGIN"
#: ../../use_cases/tool_use_with_plugin.md:6 e90be2eb88c140b5b0ac3e6b6fac76bc
msgid "Baidu-Search-Plugin"
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:8 d3c0287afa81409f9bda6fc495d63917
#: ../../use_cases/tool_use_with_plugin.md:8 0b98dfd78d49426098974d3d9c2d962b
msgid ""
"[](https://github.com/csunny/DB-GPT-"
"[Db-GPT Plugins](https://github.com/eosphoros-ai/DB-GPT-"
"Plugins/blob/main/src/dbgpt_plugins/Readme.md)"
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:10 a65c05f21ee94e8da1f14076dbed8123
#: ../../use_cases/tool_use_with_plugin.md:10 c34f612a2ec449548ec90d3bcbf5d9a0
msgid ""
"This is a DB-GPT plugin to generate data analysis charts, if you want to "
"use the test sample data, please first pull the code of [DB-GPT-"
"Plugins](https://github.com/csunny/DB-GPT-Plugins), run the command to "
"generate test DuckDB data, and then copy the generated data file to the "
"`/pilot/mock_datas` directory of the DB-GPT project."
"Perform search queries using the Baidu search engine [DB-GPT-"
"Plugins](https://github.com/eosphoros-ai/DB-GPT-Plugins)."
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:21 c25ef922010442f5be632f6d8f2e730c
#: ../../use_cases/tool_use_with_plugin.md:21 81638eb1f4f34b39a1ce383f6ef5720f
msgid ""
"Test Case: Use a histogram to analyze the total order amount of users in "
"different cities."
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:26 3f07d6e71ced4011998b1f1fda640394
#: ../../use_cases/tool_use_with_plugin.md:26 8cf2cbbff8cc408c9e885a406d34dbcb
msgid ""
"More detail see: [DB-DASHBOARD](https://github.com/csunny/DB-GPT-"
"More detail see: [DB-DASHBOARD](https://github.com/eosphoros-ai/DB-GPT-"
"Plugins/blob/main/src/dbgpt_plugins/Readme.md)"
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:29 20e5d3aed30847ccac905d0d5268824f
msgid "DB-GPT-SQL-Execution-Plugin"
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:32 4ebfd33a77e547edb1de9d3159745cb6
msgid "This is an DbGPT plugin to connect Generic Db And Execute SQL."
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:35 8c11ec372d9346e79e5ebba390b15919
msgid "DB-GPT-Bytebase-Plugin"
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:37 b01eb72df51648a293613dbab2bbe4f0
msgid ""
"To use a tool or platform plugin, you should first deploy a plugin. "
"Taking the open-source database management platform Bytebase as an "
"example, you can deploy your Bytebase service with one click using Docker"
" and access it at http://127.0.0.1:5678. More details can be found at "
"https://github.com/bytebase/bytebase."
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:53 1cdcd5fc42b6433ba5573fc157328c5c
msgid ""
"Note: If your machine's CPU architecture is `ARM`, please use `--platform"
" linux/arm64` instead."
msgstr ""
#: ../../use_cases/tool_use_with_plugin.md:55 179dc86ad25f4498af7c90f570f1a556
msgid ""
"Select the plugin on DB-GPT（All built-in plugins are from our repository:"
" https://github.com/csunny/DB-GPT-Plugins），choose DB-GPT-Bytebase-Plugin."
" Supporting functions include creating projects, creating environments, "
"creating database instances, creating databases, database DDL/DML "
"operations, and ticket approval process, etc."
msgstr ""
#~ msgid ""
#~ "DB-GPT supports a variety of "
#~ "plug-ins, such as MySQL, MongoDB, "
@ -110,11 +70,9 @@ msgstr ""
#~ " realize the ability of \"single-"
#~ "sentence requirements\""
#~ msgstr ""
#~ "DB-"
#~ "GPT支持各种插件，例如MySQL、MongoDB、ClickHouse等数据库工具插件。此外，一些数据库管理平台也可以将它们的接口打包成插件，使用该模型实现\"一句话需求\"的能力。"
#~ msgid "DB-GPT-DASHBOARD-PLUGIN"
#~ msgstr "DB-GPT-DASHBOARD-PLUGIN"
#~ msgstr ""
#~ msgid ""
#~ "[Db-GPT Chart Plugin](https://github.com/csunny"
@ -135,9 +93,6 @@ msgstr ""
#~ "the `/pilot/mock_datas` directory of the "
#~ "DB-GPT project."
#~ msgstr ""
#~ "这是一个DB-GPT插件，用于生成数据分析图表。如果您想使用测试样本数据，请先拉取 DB-GPT-"
#~ "Plugins 的代码，运行命令以生成测试 DuckDB 数据，然后将生成的数据文件复制到 "
#~ "DB-GPT 项目的 /pilot/mock_datas 目录中。"
#~ msgid ""
#~ "Test Case: Use a histogram to "
@ -150,17 +105,15 @@ msgstr ""
#~ "DASHBOARD](https://github.com/csunny/DB-GPT-"
#~ "Plugins/blob/main/src/dbgpt_plugins/Readme.md)"
#~ msgstr ""
#~ "更多详情请看：[DB-DASHBOARD](https://github.com/csunny/DB-GPT-"
#~ "Plugins/blob/main/src/dbgpt_plugins/Readme.md)"
#~ msgid "DB-GPT-SQL-Execution-Plugin"
#~ msgstr "DB-GPT-SQL-Execution-Plugin"
#~ msgstr ""
#~ msgid "This is an DbGPT plugin to connect Generic Db And Execute SQL."
#~ msgstr "这是一个 DbGPT 插件，用于连接通用数据库并执行 SQL。"
#~ msgstr ""
#~ msgid "DB-GPT-Bytebase-Plugin"
#~ msgstr "DB-GPT-Bytebase-Plugin"
#~ msgstr ""
#~ msgid ""
#~ "To use a tool or platform plugin,"
@ -173,14 +126,12 @@ msgstr ""
#~ " More details can be found at "
#~ "https://github.com/bytebase/bytebase."
#~ msgstr ""
#~ "要使用一个工具或平台插件，您应该首先部署一个插件。以开源数据库管理平台Bytebase为例，您可以使用Docker一键部署Bytebase服务，并通过http://127.0.0.1:5678进行访问。更多细节可以在"
#~ " https://github.com/bytebase/bytebase 找到。"
#~ msgid ""
#~ "Note: If your machine's CPU architecture"
#~ " is `ARM`, please use `--platform "
#~ "linux/arm64` instead."
#~ msgstr "备注：如果你的机器CPU架构是ARM,请使用--platform linux/arm64 代替"
#~ msgstr ""
#~ msgid ""
#~ "Select the plugin on DB-GPT（All "
@ -193,7 +144,9 @@ msgstr ""
#~ "database DDL/DML operations, and ticket "
#~ "approval process, etc."
#~ msgstr ""
#~ "在DB-GPT上选择插件（所有内置插件均来自我们的仓库：https://github.com/csunny/DB-"
#~ "GPT-Plugins），选择DB-GPT-Bytebase-"
#~ "Plugin。支持的功能包括创建项目、创建环境、创建数据库实例、创建数据库、数据库DDL/DML操作和审批流程等。"
#~ msgid ""
#~ "[](https://github.com/csunny/DB-GPT-"
#~ "Plugins/blob/main/src/dbgpt_plugins/Readme.md)"
#~ msgstr ""