优化项目结构，增加 Examples ToDo

2B-dada · Jul 21, 2024 · 5877a41 · 5877a41
1 parent 21d7f48
commit 5877a41
Show file tree

Hide file tree

Showing 518 changed files with 2,632 additions and 2,619 deletions.
diff --git a/README.md b/README.md
diff --git a/examples/readme.md b/examples/readme.md
@@ -0,0 +1,13 @@
+# self-llm Examples
+
+在学习者完成了基础部分的学习之后，我们将会提供一些例子来帮助学习者更好的理解和掌握大模型应用开发。我们会以应用类型为向导，提供一些优秀的大模型应用项目案例 Demo ，使得学习者完成我们的 examples 后，能够更好的理解和掌握大模型应用开发的技术要点和掌握大模型应用二次开发或单独开发的能力。
+
+## Examples 目录
+
+- 角色扮演
+  - [ ] Chat-嬛嬛
+  - [ ] Chat-悟空
+- 办公效率
+  - [ ] 搭建 RAG 对话系统
+- 学习教育
+  - [ ] ChatTest
diff --git a/Atom/01-Atom-7B-chat-WebDemo.md → models/Atom/01-Atom-7B-chat-WebDemo.md b/Atom/01-Atom-7B-chat-WebDemo.md → models/Atom/01-Atom-7B-chat-WebDemo.md
diff --git a/Atom/02-Atom-7B-Chat Lora 微调.md → models/Atom/02-Atom-7B-Chat Lora 微调.md b/Atom/02-Atom-7B-Chat Lora 微调.md → models/Atom/02-Atom-7B-Chat Lora 微调.md
diff --git a/Atom/02-Atom-7B-Chat-Lora/train.py → models/Atom/02-Atom-7B-Chat-Lora/train.py b/Atom/02-Atom-7B-Chat-Lora/train.py → models/Atom/02-Atom-7B-Chat-Lora/train.py
diff --git a/Atom/02-Atom-7B-Chat-Lora/train.sh → models/Atom/02-Atom-7B-Chat-Lora/train.sh b/Atom/02-Atom-7B-Chat-Lora/train.sh → models/Atom/02-Atom-7B-Chat-Lora/train.sh
diff --git a/Atom/03-Atom-7B-Chat 接入langchain搭建知识库助手.md → ...tom/03-Atom-7B-Chat 接入langchain搭建知识库助手.md b/Atom/03-Atom-7B-Chat 接入langchain搭建知识库助手.md → ...tom/03-Atom-7B-Chat 接入langchain搭建知识库助手.md
diff --git a/...03-Atom-7B-Chat 接入langchain搭建知识库助手/LLM.py → ...03-Atom-7B-Chat 接入langchain搭建知识库助手/LLM.py b/...03-Atom-7B-Chat 接入langchain搭建知识库助手/LLM.py → ...03-Atom-7B-Chat 接入langchain搭建知识库助手/LLM.py
diff --git a/...om-7B-Chat 接入langchain搭建知识库助手/creat_db.py → ...om-7B-Chat 接入langchain搭建知识库助手/creat_db.py b/...om-7B-Chat 接入langchain搭建知识库助手/creat_db.py → ...om-7B-Chat 接入langchain搭建知识库助手/creat_db.py
diff --git a/...Atom-7B-Chat 接入langchain搭建知识库助手/readme.md → ...Atom-7B-Chat 接入langchain搭建知识库助手/readme.md b/...Atom-7B-Chat 接入langchain搭建知识库助手/readme.md → ...Atom-7B-Chat 接入langchain搭建知识库助手/readme.md
diff --git a/...-7B-Chat 接入langchain搭建知识库助手/run_gradio.py → ...-7B-Chat 接入langchain搭建知识库助手/run_gradio.py b/...-7B-Chat 接入langchain搭建知识库助手/run_gradio.py → ...-7B-Chat 接入langchain搭建知识库助手/run_gradio.py
diff --git a/Atom/04-Atom-7B-chat 全量微调.md → models/Atom/04-Atom-7B-chat 全量微调.md b/Atom/04-Atom-7B-chat 全量微调.md → models/Atom/04-Atom-7B-chat 全量微调.md
diff --git a/Atom/images/image-1.png → models/Atom/images/image-1.png b/Atom/images/image-1.png → models/Atom/images/image-1.png
diff --git a/Atom/images/image-2.png → models/Atom/images/image-2.png b/Atom/images/image-2.png → models/Atom/images/image-2.png
diff --git a/Atom/images/image-3.png → models/Atom/images/image-3.png b/Atom/images/image-3.png → models/Atom/images/image-3.png
diff --git a/Atom/images/image-4.png → models/Atom/images/image-4.png b/Atom/images/image-4.png → models/Atom/images/image-4.png
diff --git a/Atom/images/image-5.png → models/Atom/images/image-5.png b/Atom/images/image-5.png → models/Atom/images/image-5.png
diff --git a/Atom/images/image-6.png → models/Atom/images/image-6.png b/Atom/images/image-6.png → models/Atom/images/image-6.png
diff --git a/Atom/images/image-7.png → models/Atom/images/image-7.png b/Atom/images/image-7.png → models/Atom/images/image-7.png
diff --git a/Atom/images/image-8.png → models/Atom/images/image-8.png b/Atom/images/image-8.png → models/Atom/images/image-8.png
diff --git a/Atom/images/image-9.png → models/Atom/images/image-9.png b/Atom/images/image-9.png → models/Atom/images/image-9.png
diff --git a/...huan/01-Baichuan2-7B-chat+FastApi+部署调用.md → ...huan/01-Baichuan2-7B-chat+FastApi+部署调用.md b/...huan/01-Baichuan2-7B-chat+FastApi+部署调用.md → ...huan/01-Baichuan2-7B-chat+FastApi+部署调用.md
diff --git a/BaiChuan/02-Baichuan-7B-chat+WebDemo.md → ...s/BaiChuan/02-Baichuan-7B-chat+WebDemo.md b/BaiChuan/02-Baichuan-7B-chat+WebDemo.md → ...s/BaiChuan/02-Baichuan-7B-chat+WebDemo.md
diff --git a/...huan/03-Baichuan2-7B-chat接入LangChain框架.md → ...huan/03-Baichuan2-7B-chat接入LangChain框架.md b/...huan/03-Baichuan2-7B-chat接入LangChain框架.md → ...huan/03-Baichuan2-7B-chat接入LangChain框架.md
diff --git a/BaiChuan/04-Baichuan2-7B-chat Lora 微调.ipynb → ...iChuan/04-Baichuan2-7B-chat Lora 微调.ipynb b/BaiChuan/04-Baichuan2-7B-chat Lora 微调.ipynb → ...iChuan/04-Baichuan2-7B-chat Lora 微调.ipynb
diff --git a/BaiChuan/04-Baichuan2-7B-chat+lora+微调.md → .../BaiChuan/04-Baichuan2-7B-chat+lora+微调.md b/BaiChuan/04-Baichuan2-7B-chat+lora+微调.md → .../BaiChuan/04-Baichuan2-7B-chat+lora+微调.md
diff --git a/BaiChuan/images/image1.png → models/BaiChuan/images/image1.png b/BaiChuan/images/image1.png → models/BaiChuan/images/image1.png
diff --git a/BaiChuan/images/image10.png → models/BaiChuan/images/image10.png b/BaiChuan/images/image10.png → models/BaiChuan/images/image10.png
diff --git a/BaiChuan/images/image11.png → models/BaiChuan/images/image11.png b/BaiChuan/images/image11.png → models/BaiChuan/images/image11.png
diff --git a/BaiChuan/images/image12.png → models/BaiChuan/images/image12.png b/BaiChuan/images/image12.png → models/BaiChuan/images/image12.png
diff --git a/BaiChuan/images/image13.png → models/BaiChuan/images/image13.png b/BaiChuan/images/image13.png → models/BaiChuan/images/image13.png
diff --git a/BaiChuan/images/image14.png → models/BaiChuan/images/image14.png b/BaiChuan/images/image14.png → models/BaiChuan/images/image14.png
diff --git a/BaiChuan/images/image15.png → models/BaiChuan/images/image15.png b/BaiChuan/images/image15.png → models/BaiChuan/images/image15.png
diff --git a/BaiChuan/images/image16.png → models/BaiChuan/images/image16.png b/BaiChuan/images/image16.png → models/BaiChuan/images/image16.png
diff --git a/BaiChuan/images/image17.png → models/BaiChuan/images/image17.png b/BaiChuan/images/image17.png → models/BaiChuan/images/image17.png
diff --git a/BaiChuan/images/image18.png → models/BaiChuan/images/image18.png b/BaiChuan/images/image18.png → models/BaiChuan/images/image18.png
diff --git a/BaiChuan/images/image2.png → models/BaiChuan/images/image2.png b/BaiChuan/images/image2.png → models/BaiChuan/images/image2.png
diff --git a/BaiChuan/images/image20.png → models/BaiChuan/images/image20.png b/BaiChuan/images/image20.png → models/BaiChuan/images/image20.png
diff --git a/BaiChuan/images/image23.png → models/BaiChuan/images/image23.png b/BaiChuan/images/image23.png → models/BaiChuan/images/image23.png
diff --git a/BaiChuan/images/image25.png → models/BaiChuan/images/image25.png b/BaiChuan/images/image25.png → models/BaiChuan/images/image25.png
diff --git a/BaiChuan/images/image26.png → models/BaiChuan/images/image26.png b/BaiChuan/images/image26.png → models/BaiChuan/images/image26.png
diff --git a/BaiChuan/images/image27.png → models/BaiChuan/images/image27.png b/BaiChuan/images/image27.png → models/BaiChuan/images/image27.png
diff --git a/BaiChuan/images/image3.png → models/BaiChuan/images/image3.png b/BaiChuan/images/image3.png → models/BaiChuan/images/image3.png
diff --git a/BaiChuan/images/image4.png → models/BaiChuan/images/image4.png b/BaiChuan/images/image4.png → models/BaiChuan/images/image4.png
diff --git a/BaiChuan/images/image6.png → models/BaiChuan/images/image6.png b/BaiChuan/images/image6.png → models/BaiChuan/images/image6.png
diff --git a/BaiChuan/images/image7.png → models/BaiChuan/images/image7.png b/BaiChuan/images/image7.png → models/BaiChuan/images/image7.png
diff --git a/BaiChuan/images/image8.png → models/BaiChuan/images/image8.png b/BaiChuan/images/image8.png → models/BaiChuan/images/image8.png
diff --git a/BaiChuan/images/image9.png → models/BaiChuan/images/image9.png b/BaiChuan/images/image9.png → models/BaiChuan/images/image9.png
diff --git a/BlueLM/01-BlueLM-7B-Chat FastApi 部署.md → ...ls/BlueLM/01-BlueLM-7B-Chat FastApi 部署.md b/BlueLM/01-BlueLM-7B-Chat FastApi 部署.md → ...ls/BlueLM/01-BlueLM-7B-Chat FastApi 部署.md
diff --git a/BlueLM/02-BlueLM-7B-Chat langchain 接入.md → .../BlueLM/02-BlueLM-7B-Chat langchain 接入.md b/BlueLM/02-BlueLM-7B-Chat langchain 接入.md → .../BlueLM/02-BlueLM-7B-Chat langchain 接入.md
diff --git a/BlueLM/03-BlueLM-7B-Chat WebDemo 部署.md → ...ls/BlueLM/03-BlueLM-7B-Chat WebDemo 部署.md b/BlueLM/03-BlueLM-7B-Chat WebDemo 部署.md → ...ls/BlueLM/03-BlueLM-7B-Chat WebDemo 部署.md
diff --git a/BlueLM/04-BlueLM-7B-Chat Lora 微调.ipynb → ...ls/BlueLM/04-BlueLM-7B-Chat Lora 微调.ipynb b/BlueLM/04-BlueLM-7B-Chat Lora 微调.ipynb → ...ls/BlueLM/04-BlueLM-7B-Chat Lora 微调.ipynb
diff --git a/BlueLM/04-BlueLM-7B-Chat Lora 微调.md → models/BlueLM/04-BlueLM-7B-Chat Lora 微调.md b/BlueLM/04-BlueLM-7B-Chat Lora 微调.md → models/BlueLM/04-BlueLM-7B-Chat Lora 微调.md
diff --git a/BlueLM/04-BlueLM-7B-Chat Lora 微调.py → models/BlueLM/04-BlueLM-7B-Chat Lora 微调.py b/BlueLM/04-BlueLM-7B-Chat Lora 微调.py → models/BlueLM/04-BlueLM-7B-Chat Lora 微调.py
diff --git a/BlueLM/images/202403191628941.png → models/BlueLM/images/202403191628941.png b/BlueLM/images/202403191628941.png → models/BlueLM/images/202403191628941.png
diff --git a/BlueLM/images/202403191813385.png → models/BlueLM/images/202403191813385.png b/BlueLM/images/202403191813385.png → models/BlueLM/images/202403191813385.png
diff --git a/BlueLM/images/202403201210690.png → models/BlueLM/images/202403201210690.png b/BlueLM/images/202403201210690.png → models/BlueLM/images/202403201210690.png
diff --git a/BlueLM/images/202403201229542.png → models/BlueLM/images/202403201229542.png b/BlueLM/images/202403201229542.png → models/BlueLM/images/202403201229542.png
diff --git a/BlueLM/images/202403202153465.png → models/BlueLM/images/202403202153465.png b/BlueLM/images/202403202153465.png → models/BlueLM/images/202403202153465.png
diff --git a/...GLM/01-CharacterGLM-6B Transformer部署调用.md → ...GLM/01-CharacterGLM-6B Transformer部署调用.md b/...GLM/01-CharacterGLM-6B Transformer部署调用.md → ...GLM/01-CharacterGLM-6B Transformer部署调用.md
@@ -1,77 +1,77 @@
-# CharacterGLM-6B Transformers部署调用
-
-## 环境准备
-
-在autodl平台中租一个3090等24G显存的显卡机器，如下图所示镜像选择PyTorch-->2.0.0-->3.8(ubuntu20.04)-->11.8
-
-![image](https://github.com/suncaleb1/self-llm/assets/155936975/fc4c6323-d338-4d66-a244-bbefe7da3746)
-
-接下来打开刚刚租用服务器的JupyterLab，并且打开其中的终端开始环境配置、模型下载和运行demo。
-
-pip换源和安装依赖包
-
-```python
-#升级pip
-python -m pip install --upgrade pip
-#更换 pypi 源加速库的安装
-pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
-
-pip install modelscope
-pip install transformers
-pip install sentencepiece
-```
-
-## 模型下载
-
-使用 modelscope 中的snapshot_download函数下载模型，第一个参数为模型名称，参数cache_dir为模型的下载路径。
-
-在 /root/autodl-tmp 路径下新建 download.py 文件并在其中输入以下内容，粘贴代码后记得保存文件，如下图所示。并运行 python /root/autodl-tmp/download.py执行下载，模型大小为 12 GB，下载模型大概需要 10~15 分钟
-
-```python
-import torch
-from modelscope import snapshot_download, AutoModel, AutoTokenizer
-import os
-model_dir = snapshot_download('THUCoAI/CharacterGLM-6B', cache_dir='/root/autodl-tmp', revision='master')
-```
-
-## 代码准备
-
-```python
-from transformers import AutoTokenizer,AutoModelForCausalLM
-import torch
-# 使用模型下载到的本地路径以加载
-model_dir = '/root/autodl-tmp/THUCoAI/CharacterGLM-6B'
-# 分词器的加载，本地加载，trust_remote_code=True设置允许从网络上下载模型权重和相关的代码
-tokenizer = AutoTokenizer.from_pretrained(model_dir, trust_remote_code=True)
-# 模型加载，本地加载，使用AutoModelForCausalLM类
-model = AutoModelForCausalLM.from_pretrained(model_dir, trust_remote_code=True)
-# 将模型移动到GPU上进行加速（如果有GPU的话）
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-model.to(device)
-# 使用模型的评估模式来产生对话
-model.eval()
-session_meta = {'user_info': '我是陆星辰，是一个男性，是一位知名导演，也是苏梦远的合作导演。我擅长拍摄音乐题材的电影。苏梦远对我的态度是尊敬的，并视我为良师益友。', 'bot_info': '苏梦远，本名苏远心，是一位当红的国内女歌手及演员。在参加选秀节目后，凭借独特的嗓音及出众的舞台魅力迅速成名，进入娱乐圈。她外表美丽动人，但真正的魅力在于她的才华和勤奋。苏梦远是音乐学院毕业的优秀生，善于创作，拥有多首热门原创歌曲。除了音乐方面的成就，她还热衷于慈善事业，积极参加公益活动，用实际行动传递正能量。在工作中，她对待工作非常敬业，拍戏时总是全身心投入角色，赢得了业内人士的赞誉和粉丝的喜爱。虽然在娱乐圈，但她始终保持低调、谦逊的态度，深得同行尊重。在表达时，苏梦远喜欢使用“我们”和“一起”，强调团队精神。', 'bot_name': '苏梦远', 'user_name': '陆星辰'}
-# 第一轮对话
-response, history = model.chat(tokenizer, session_meta,"你好呀，小苏", history=[])
-print(response)
-# 第二轮对话
-response, history = model.chat(tokenizer, session_meta,"最近对音乐有什么新的想法吗", history=history)
-print(response)
-# 第三轮对话
-response, history = model.chat(tokenizer,session_meta, "那我们商量一下下一部音乐电影的拍摄，好嘛？", history=history)
-print(response)
-```
-
-## 部署
-
-在终端输入以下命令运行trans.py，即实现CharacterGLM-6B的Transformers部署调用
-
-```python
-cd /root/autodl-tmp
-python trans.py
-```
-
-观察命令行中loading checkpoint表示模型正在加载，等待模型加载完成产生对话，如下图所示
-
-![image](https://github.com/suncaleb1/self-llm/assets/155936975/f9d65275-fa89-4039-95c5-7cc0615753e2)
-
+# CharacterGLM-6B Transformers部署调用
+
+## 环境准备
+
+在autodl平台中租一个3090等24G显存的显卡机器，如下图所示镜像选择PyTorch-->2.0.0-->3.8(ubuntu20.04)-->11.8
+
+![image](https://github.com/suncaleb1/self-llm/assets/155936975/fc4c6323-d338-4d66-a244-bbefe7da3746)
+
+接下来打开刚刚租用服务器的JupyterLab，并且打开其中的终端开始环境配置、模型下载和运行demo。
+
+pip换源和安装依赖包
+
+```python
+#升级pip
+python -m pip install --upgrade pip
+#更换 pypi 源加速库的安装
+pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
+
+pip install modelscope
+pip install transformers
+pip install sentencepiece
+```
+
+## 模型下载
+
+使用 modelscope 中的snapshot_download函数下载模型，第一个参数为模型名称，参数cache_dir为模型的下载路径。
+
+在 /root/autodl-tmp 路径下新建 download.py 文件并在其中输入以下内容，粘贴代码后记得保存文件，如下图所示。并运行 python /root/autodl-tmp/download.py执行下载，模型大小为 12 GB，下载模型大概需要 10~15 分钟
+
+```python
+import torch
+from modelscope import snapshot_download, AutoModel, AutoTokenizer
+import os
+model_dir = snapshot_download('THUCoAI/CharacterGLM-6B', cache_dir='/root/autodl-tmp', revision='master')
+```
+
+## 代码准备
+
+```python
+from transformers import AutoTokenizer,AutoModelForCausalLM
+import torch
+# 使用模型下载到的本地路径以加载
+model_dir = '/root/autodl-tmp/THUCoAI/CharacterGLM-6B'
+# 分词器的加载，本地加载，trust_remote_code=True设置允许从网络上下载模型权重和相关的代码
+tokenizer = AutoTokenizer.from_pretrained(model_dir, trust_remote_code=True)
+# 模型加载，本地加载，使用AutoModelForCausalLM类
+model = AutoModelForCausalLM.from_pretrained(model_dir, trust_remote_code=True)
+# 将模型移动到GPU上进行加速（如果有GPU的话）
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+# 使用模型的评估模式来产生对话
+model.eval()
+session_meta = {'user_info': '我是陆星辰，是一个男性，是一位知名导演，也是苏梦远的合作导演。我擅长拍摄音乐题材的电影。苏梦远对我的态度是尊敬的，并视我为良师益友。', 'bot_info': '苏梦远，本名苏远心，是一位当红的国内女歌手及演员。在参加选秀节目后，凭借独特的嗓音及出众的舞台魅力迅速成名，进入娱乐圈。她外表美丽动人，但真正的魅力在于她的才华和勤奋。苏梦远是音乐学院毕业的优秀生，善于创作，拥有多首热门原创歌曲。除了音乐方面的成就，她还热衷于慈善事业，积极参加公益活动，用实际行动传递正能量。在工作中，她对待工作非常敬业，拍戏时总是全身心投入角色，赢得了业内人士的赞誉和粉丝的喜爱。虽然在娱乐圈，但她始终保持低调、谦逊的态度，深得同行尊重。在表达时，苏梦远喜欢使用“我们”和“一起”，强调团队精神。', 'bot_name': '苏梦远', 'user_name': '陆星辰'}
+# 第一轮对话
+response, history = model.chat(tokenizer, session_meta,"你好呀，小苏", history=[])
+print(response)
+# 第二轮对话
+response, history = model.chat(tokenizer, session_meta,"最近对音乐有什么新的想法吗", history=history)
+print(response)
+# 第三轮对话
+response, history = model.chat(tokenizer,session_meta, "那我们商量一下下一部音乐电影的拍摄，好嘛？", history=history)
+print(response)
+```
+
+## 部署
+
+在终端输入以下命令运行trans.py，即实现CharacterGLM-6B的Transformers部署调用
+
+```python
+cd /root/autodl-tmp
+python trans.py
+```
+
+观察命令行中loading checkpoint表示模型正在加载，等待模型加载完成产生对话，如下图所示
+
+![image](https://github.com/suncaleb1/self-llm/assets/155936975/f9d65275-fa89-4039-95c5-7cc0615753e2)
+