# ChatBot with OpenAI

*author: Jael, Yuchen*

<br />

## Description

A chat-bot operator returns answer in text given input text.
This operator is implemented with GPT models from [OpenAI](https://platform.openai.com/docs/guides/embeddings).
Please note you need an [OpenAI API key](https://platform.openai.com/account/api-keys) to access OpenAI.

<br />

## Code Example

Use the default model to answer the question "Who are you?".

*Write a pipeline with explicit inputs/outputs name specifications:*

```python
from towhee.dc2 import pipe, ops, DataCollection

p = (
    pipe.input('question')
        .map('question', 'answer', 
             ops.chatbot.openai(api_key=OPENAI_API_KEY))
        .output('question', 'answer')
)

DataCollection(p('Who are you?')).show()
```

<br />

## Factory Constructor

Create the operator via the following factory method:

***chatbot.openai(model_name: str, api_key: str)***

**Parameters:**

***model_name***: *str*

The model name in string, defaults to 'text-davinci-003'. Supported model names:
- text-davinci-003

***api_key***: *str=None*

The OpenAI API key in string, defaults to None.

<br />

## Interface

The operator takes a piece of text in string as input.
It returns a text emabedding in numpy.ndarray.

***\_\_call\_\_(txt)***

**Parameters:**

***text***: *str*

​	The text in string.

**Returns**:

*answer: str*

​	The answer in string generated by model.

<br />

***supported_model_names()***

Get a list of supported model names.