diff --git a/README.md b/README.md index 61ea6ab..0fa50cc 100644 --- a/README.md +++ b/README.md @@ -157,7 +157,7 @@ Roadmap_ZH -### 🎯框架图 +### 🔗框架图

@@ -170,10 +170,11 @@ - [🎇最近更新](#最近更新) - [🏆荣誉栏](#荣誉栏) - [🎯路线图](#路线图) - - [🎯框架图](#框架图) + - [🔗框架图](#框架图) - [目录](#目录) - [开发前的配置要求](#开发前的配置要求) - [**使用指南**](#使用指南) + - [快速体验](#快速体验) - [数据构建](#数据构建) - [微调指南](#微调指南) - [部署指南](#部署指南) @@ -201,12 +202,19 @@ git clone https://github.com/SmartFlowAI/EmoLLM.git ``` 2. 依次阅读或者选择感兴趣的部分阅读： + - [快速体验](#快速体验) - [数据构建](#数据构建) - [微调指南](#微调指南) - [部署指南](#部署指南) - [RAG](#rag检索增强生成pipeline) - 查看更多详情 + +### 快速体验 + +- 请阅读[快速体验](docs/quick_start.md)查阅 + + ### 数据构建 - 请阅读[数据构建指南](generate_data/tutorial.md)查阅 diff --git a/README_EN.md b/README_EN.md index 345bcc1..31a426e 100644 --- a/README_EN.md +++ b/README_EN.md @@ -199,24 +199,16 @@ git clone https://github.com/SmartFlowAI/EmoLLM.git ``` 1. Read in sequence or read sections you're interested in： - - [File Directory Explanation](#file-directory-explanation) + - [Quick Start](#quick-start) - [Data Construction](#data-construction) - [Fine-tuning Guide](#fine-tuning-guide) - [Deployment Guide](#deployment-guide) + - [RAG][] - View More Details -### File Directory Explanation -``` -├─assets: Image Resources -├─datasets: Dataset -├─demo: demo scripts -├─generate_data: Data Generation Guide -│ └─xinghuo -├─scripts: Some Available Tools -└─xtuner_config：Fine-tuning Guide - └─images -``` +### Quick start +- Please read [Quick Start](docs/quick_start_EN.md) to see. ### Data Construction diff --git a/docs/quick_start.md b/docs/quick_start.md new file mode 100644 index 0000000..5d2473b --- /dev/null +++ b/docs/quick_start.md @@ -0,0 +1,34 @@ +### 1、部署环境 +- 操作系统：Ubuntu 22.04.4 LTS +- CPU：Intel (R) Xeon (R) CPU E 5-2650，32G（在线 GPU 服务器） +- 显卡：NVIDIA RTX 4060Ti 16G，NVIDIA-SMI 535.104.05 Driver Version: 535.104.05 CUDA Version: 12.2 +- Python 3.11.5 + +### 2、默认部署步骤 +- 1、Clone 代码或者手动下载代码放置服务器： +``` +git clone https://github.com/SmartFlowAI/EmoLLM.git +``` +- 2、安装 Python 依赖库： +``` +# cd EmoLLM +# pip install -r requirements.txt +``` +- 3、安装 oss2 库用于自动下载模型（requirements.txt 文件缺少该库） +``` +# pip install oss2 +``` +- 4、【重要】下载模型目录，选择你想要使用的模型，手动下载去 openxlab（如 [EmoLLM_Model](https://openxlab.org.cn/models/detail/jujimeizuo/EmoLLM_Model)） 或者 Huggingface 等其他平台下载模型目录，将全部文件放在 `EmoLLM/model` 目录下。注意，这一步必须要有，否则自动下载会报错，打包下载时并不会下载 LFS（如 pytorch_model-00001-of-00008.bin）文件的完整文件只是一个引用而已。 +- 5、下载模型，特指模型中的 pytorch_model-XXX.bin 文件，可通过手动或者自动方式下载模型。 +- 5.1、模型手动下载则去 openxlab（如 [EmoLLM_Model](https://openxlab.org.cn/models/detail/jujimeizuo/EmoLLM_Model)） 或者其他平台下载模型，将全部文件放在 `EmoLLM/model` 目录下，然后修改 web_demo-aiwei.py 或者 web_internlm2.py 文件（app. Py 调用哪一个就修改哪一个文件）注释掉开头的下面代码，防止脚本自动重新下载模型： +``` +# download(model_repo='ajupyter/EmoLLM_aiwei', +# output='model') +``` +- 5.2、模型自动下载，进入 EmoLLM 目录后有三个文件，分别是 app.py、web_demo-aiwei.py、web_internlm2.py。app.py 是调用后面两个脚本的。想要调用哪一个脚本就注释掉另外一个脚本。并运行如下 +``` +# python ./app.py +``` +- 注意：默认情况下 web_demo-aiwei.py 自动下载的是 openxlab 平台的 ajupyter/EmoLLM_aiwei 模型，web_internlm2.py 下载的是 jujimeizuo/EmoLLM_Model 模型。脚本会自动将模型下载到 EmoLLM/model 目录，下载完成成功运行后，按照 5.1 的手动下载的操作修改 web_demo-aiwei.py 或者 web_internlm2.py 文件注释掉下载代码，防止下次运行重新下载。 +6、运行 app.py 后等待模型下载并且加载完毕，可通过浏览器访问： http://0.0.0.0:7860 地址访问模型 web 页面。可修改 app.py 文件修改 web 页面访问端口。 +7、替换模型，EmoLLM 提供了多种开源模型，分别上传至 openxlab、Huggingface 平台，有[爹系男友心理咨询师 ](https://openxlab.org.cn/models/detail/chg0901/EmoLLM_Daddy-like_BF)、[老母亲心理咨询师](https://huggingface.co/brycewang2018/EmoLLM-mother/tree/main)、[温柔御姐心理医生艾薇](https://openxlab.org.cn/models/detail/ajupyter/EmoLLM_aiwei)等角色，有 EmoLLM_internlm2_7b_full、EmoLLM-InternLM7B-base-10e 等多个模型可选择，目前评测 [EmoLLM_internlm2_7b_ful](https://openxlab.org.cn/models/detail/ajupyter/EmoLLM_internlm2_7b_full) 模型效果较好。可重复步骤 4、5手动或自动下载相关模型放在 `EmoLLM/model` 目录下或者修改 web_demo-aiwei.py 或者 web_internlm2.py 文件的 download 函数仓库地址自动下载。 \ No newline at end of file diff --git a/docs/quick_start_EN.md b/docs/quick_start_EN.md new file mode 100644 index 0000000..29a4788 --- /dev/null +++ b/docs/quick_start_EN.md @@ -0,0 +1,36 @@ +### 1. Deployment Environment +- Operating system: Ubuntu 22.04.4 LTS +- CPU: Intel (R) Xeon (R) CPU E 5-2650, 32G +- Graphics card: NVIDIA RTX 4060Ti 16G, NVIDIA-SMI 535.104.05 Driver Version: 535.104.05 CUDA Version: 12.2 +- Python 3.11.5 + +### 2. Default Deployment Steps +- 1. Clone the code or manually download the code and place it on the server: +``` +git clone https://github.com/SmartFlowAI/EmoLLM.git +``` +- 2. Install Python dependencies: +``` +# cd EmoLLM +# pip install -r requirements.txt +``` +- 3. Install the oss2 library for automatic model download (missing from the requirements.txt file): +``` +# pip install oss2 +``` +- 4. Important:Download the model catalog, select the model you want to use, to download the model catalog manually, go to openxlab (like [EmoLLM_Model](https://openxlab.org.cn/models/detail/jujimeizuo/EmoLLM_Model)) or Huggingface or other platforms, Place all files in the `EmoLLM/model` directory. Note that this step is mandatory, otherwise the automatic download will report an error, The package download does not download the full LFS (like pytorch_model-00001-of-00008.bin) file, just a reference to it. +- 5. Download the model, specifically the pytorch_model-XXX.bin file in the model, +models can be downloaded manually or automatically. +- 5.1. Models can be downloaded manually from openxlab or other platforms, put all the files in the `EmoLLM/model` directory, and then modify the web_demo-aiwei.py or web_internlm2.py files (whichever one is called by app.py) by commenting out the following code at the beginning, to prevent the script from automatically re-downloading the model： +``` +# download(model_repo='ajupyter/EmoLLM_aiwei', +# output='model') +``` +- 5.2. Automatic model download, There are three files in the EmoLLM directory, app.py、web_demo-aiwei.py、web_internlm2.py, app.py calls the last two scripts. Comment out the other script if you want to call it. Run: +``` +# python ./app.py +``` +- Note: By default, web_demo-aiwei.py automatically downloads the ajupyter/EmoLLM_aiwei model of openxlab platform, and web_internlm2.py downloads the jujimeizuo/EmoLLM_Model model. The script will automatically download the model to EmoLLM/model directory, after the download is completed and run successfully, follow the manual download operation in 5.1 to modify the web_demo-aiwei.py or web_internlm2.py file to comment out the download code, so as to prevent re-downloading in the next run. +6、 Run app.py and wait for the model to download and load. Then access the model web page through your browser by going to: http://0.0.0.0:7860 address. You can modify the app.py file to change the web page access port. +7. Use of other models, EmoLLM offers several versions of the open source model, uploaded to openxlab、Huggingface, and other platforms. There are roles such as the [father's boyfriend counselor](https://openxlab.org.cn/models/detail/chg0901/EmoLLM_Daddy-like_BF), [the old mother's counselor](https://huggingface.co/brycewang2018/EmoLLM-mother/tree/main), and [the gentle royal psychiatrist](https://openxlab.org.cn/models/detail/ajupyter/EmoLLM_aiwei). There are several models to choose from such as EmoLLM_internlm2_7b_full, EmoLLM-InternLM7B-base-10e and so on. +Current review [EmoLLM_internlm2_7b_ful] (https://openxlab.org.cn/models/detail/ajupyter/EmoLLM_internlm2_7b_full) model works better. You can repeat steps 4 and 5 to download the model manually or automatically in the `EmoLLM/model` directory or modify the download function in web_demo-aiwei.py or web_internlm2.py to download the model automatically. \ No newline at end of file