# nice-coding-ai

**Repository Path**: mmmm96_admin/nice-coding-ai

## Basic Information

- **Project Name**: nice-coding-ai
- **Description**: 分别基于LlamaIndex和LangChain结合Qdrant实现RAG（Retrieval-Augmented Generation，检索增强生成）
- **Primary Language**: Python
- **License**: GPL-2.0
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2025-04-15
- **Last Updated**: 2025-04-15

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

## How to config it

### Use poetry

Installation Poetry:

guide <https://python-poetry.org/docs/#installing-with-the-official-installer>

### Create .venv

```bash
python3.11 -m venv .venv
```

### Start venv

```bash
poetry shell
```

### Exit

```bash
deactivate
```

### Install package

```bash
poetry install
```

### Config openai key

```bash
export OPENAI_API_KEY="your_api_key_here"
 ```

## What is langchain

![](assert/readme/q-and-a-flow.png)

## What is Llamaindex

![](assert/readme/llamaindex.png)

## Use Qdrant

进入docker目录

```bash
docker-compose up -d
```

Document: <https://qdrant.tech/documentation/concepts/collections/>
![](assert/readme/qdrant.png)
粗略的理解成:

- collection -> table
- point -> row
- payload -> column

start with docker-compose

- REST API: localhost:6333
- Web UI: <localhost:6333/dashboard>
- GRPC API: localhost:6334

## Start up

```bash
uvicorn src.main:app --reload --port 5000
```

swagger: <http://127.0.0.1:5000/docs>

如果要使用离线的llm，需要安装ollama以及模型 <https://ollama.com/>
```bash
ollama run llama2-chinese:7b-chat-q2_K
```

rest接口分别对应LangChain、LlamaIndex录入Qdrant以及问答

![](assert/readme/fastapi.png)
LangChain: 

+ /langchain-qdrant/openai-load-data (使用openai模型入qdrant)
+ /langchain-qdrant/huggingface-load-data (使用huggingface中的模型入qdrant)
+ /langchain/llm-openai/embedding-openai/ask-question (以openai模型做Embedding，使用openai llm问答)
+ /langchain/llm-openai/embedding-huggingface/ask-question (以huggingface模型做Embedding，使用openai llm问答)
+ /langchain/llm-ollama/embedding-huggingface/ask-question (以huggingface模型做Embedding，使用ollama llm问答)

LlamaIndex:
+ /llamaindex-qdrant/openai-load-data (使用openai模型入qdrant)
+ /llamaindex-qdrant/huggingface-load-data (使用huggingface中的模型入qdrant)
+ /llamaindex/llm-openai/embedding-openai/ask-question (以openai模型做Embedding，使用openai llm问答)
+ /llamaindex/llm-openai/embedding-huggingface/ask-question (以huggingface模型做Embedding，使用openai llm问答)
+ /llamaindex/llm-ollama/embedding-huggingface/ask-question (以huggingface模型做Embedding，使用ollama llm问答)

欲知后事如何，可关注公众号保持联络😘​

![](assert/readme/qrcode_for_nice_coding.jpg)