# OpenAI

### How to configure and use it?

#### Pre-Requisite(s)

* `openai_api_key` (required) - Set an OpenAI key for running the OpenAI Model
* `model_name` (optional) - Set which model of the OpenAI model you want to use.\
  Defaults to `gpt-3.5-turbo-16k`

#### Running in a Colab/Kaggle/Python scripts(s)

```python
from genai_stack.model import OpenAIGpt35Model

llm  = OpenAIGpt35Model.from_kwargs(fields={"openai_api_key": "sk-xxxx"})  # Update with your OpenAI Key
model_response = llm.predict("How long AI has been around.")
print(model_response["result"])
```

1. Import the model from genai-stack
2. Instantiate the class with `openai_api_key`
3. call `.predict()` method and pass the query you want the model to answer to.
4. Print the response. As the response is a dictionary, get the result only.
   * The response on predict() from the model includes *result* and *source\_documents*.

#### Running the model in a webserver

If you want to run the model in a webserver and interact with it with HTTP requests, the model provides a way to run it.

1. As a Python script

We use FastAPI + Uvicorn to run a model in a webserver.

Set the response class. Default response class is `fastapi.responses.Response`. It can be customized as done in the below code snippet.

```python
from genai_stack.model import OpenAIGpt35Model
from fastapi.responses import JSONResponse

llm  = OpenAIGpt35Model.from_kwargs(fields={"openai_api_key": "sk-xxxx"})
llm.run_http_server(response_class=JSONResponse)
```

A uvicorn server should start as below.

```bash
INFO:     Started server process [137717]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8082 (Press CTRL+C to quit)
```

Making HTTP requests.\
URL - <http://localhost:8082/predict/>

```python
import requests
response = requests.post("http://localhost:8082/predict/",data="How long AI has been around.")
print(response.text)
```

2. As a CLI

Create a `model.json` file with the following contents:

{% code fullWidth="false" %}

```json
{
    "model": {       
        "name": "gpt3.5",
        "fields": {
            "openai_api_key": "sk-***"
        }

    }
}
```

{% endcode %}

Run the below CLI

```bash
genai-stack start --config_file model.json
```

```bash
 ██████╗ ███████╗███╗   ██╗ █████╗ ██╗    ███████╗████████╗ █████╗  ██████╗██╗  ██╗    
██╔════╝ ██╔════╝████╗  ██║██╔══██╗██║    ██╔════╝╚══██╔══╝██╔══██╗██╔════╝██║ ██╔╝    
██║  ███╗█████╗  ██╔██╗ ██║███████║██║    ███████╗   ██║   ███████║██║     █████╔╝     
██║   ██║██╔══╝  ██║╚██╗██║██╔══██║██║    ╚════██║   ██║   ██╔══██║██║     ██╔═██╗     
╚██████╔╝███████╗██║ ╚████║██║  ██║██║    ███████║   ██║   ██║  ██║╚██████╗██║  ██╗    
 ╚═════╝ ╚══════╝╚═╝  ╚═══╝╚═╝  ╚═╝╚═╝    ╚══════╝   ╚═╝   ╚═╝  ╚═╝ ╚═════╝╚═╝  ╚═╝
INFO:     Started server process [641734]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8082 (Press CTRL+C to quit)
```


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://genaistack.aiplanet.com/v0.1.0/components/models-llms/openai.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.