Update Get Started Documentation

docs/Get_started/config.rst

@@ -0,0 +1,123 @@
+OpenAI API Calls Configurations
+===============================
+
+LionAGI natively supports OpenAI API calls. You have the flexibility to customize models, adjust model parameters, and
+tailor API service settings.
+
+.. warning::
+   ❗❗Calling API with maximum throughput over large set of data with advanced models i.e. gpt-4 can
+   get **EXPENSIVE IN JUST SECONDS**
+
+
+Most usefully you can customize:
+
+- ``oai_llmconfig``: the default model parameters for every API call
+- ``OpenAIService``: the default settings for the API service
+
+``oai_llmconfig``
+-----------------
+
+Currently, our default llmconfig for Open AI API calls is
+
+.. code-block:: python
+
+   {'model': 'gpt-4-1106-preview',
+   'frequency_penalty': 0,
+   'max_tokens': None,
+   'n': 1,
+   'presence_penalty': 0,
+   'response_format': {'type': 'text'},
+   'seed': None,
+   'stop': None,
+   'stream': False,
+   'temperature': 0.7,
+   'top_p': 1,
+   'tools': None,
+   'tool_choice': 'none',
+   'user': None}
+
+If you wish to change the default behavior, you can update the configuration using ``lionagi.oai_llmconfig``.
+For example:
+
+.. code-block:: python
+
+   import lionagi as li
+
+   li.oai_llmconfig['model'] = 'gpt-3.5-turbo'
+
+or
+
+.. code-block:: python
+
+   li.oai_llmconfig.update({'model': 'gpt-3.5-turbo'})
+
+Alternatively, you can make changes in the ``Session``.
+
+- Pass in a new llmconfig into the ``Session``. (Recommended for completely different config)
+
+For example:
+
+.. code-block:: python
+
+   import lionagi as li
+
+   system = 'you are a helpful assistant'
+   customized_llmconfig = {...} # your customized llmconfig setting
+
+   session = li.Session(system=system, llmconfig=customized_llmconfig)
+
+Or
+
+- Update directly in the session.
+
+For example:
+
+.. code-block:: python
+
+   import lionagi as li
+
+   system = 'you are a helpful assistant'
+
+   session = li.Session(system)
+   session.llmconfig.update({"model": "gpt-3.5-turbo", "temperature": 0.5})
+
+``OpenAIService``
+-----------
+
+``OpenAIService`` provides a foundation for seamless integration and utilization of the API service. By default, the
+rate limits are set to be **tier 1** of OpenAI model `gpt-4-1104-preview`.
+
+
+You can modify rate limits to fit different cases. For example:
+
+.. code-block:: python
+
+   import lionagi as li
+
+   system = 'you are a helpful assistant'
+
+   api_service = li.OpenAIService(max_requests_per_minute=10, max_tokens_per_minute=10000)
+   session = li.Session(system, api_service)
+
+.. note::
+
+   For more information about rate limits, please check the `OpenAI usage limit documentation <https://platform.openai.com/docs/guides/rate-limits?context=tier-free)>`_
+
+If you have more than one API key, please add them to the `.env` file. To use an API key other than the default
+OPENAI_API_KEY, ensure it is appropriately specified in the configuration.
+
+.. code-block:: python
+
+   import os
+   from dotenv import load_dotenv
+   load_dotenv()
+
+   # let's say you added the second API key OPENAI_API_KEY2
+   api_key2 = os.getenv("OPENAI_API_KEY2")
+
+   api_service = li.OpenAIService(api_key=api_key2)
+   session = li.Session(system, api_service=api_service)
+
+.. note::
+
+   If you wish to apply the same ``api_service`` setting across multiple sessions, make sure to pass it to each of these sessions.

docs/Get_started/quick_start.rst

@@ -0,0 +1,26 @@
+Quick Start
+===========
+
+Start your LionAGI experience with our demo example.
+
+.. code-block:: python
+
+  import lionagi as li
+
+  # define system messages, context and user instruction
+  system = "You are a helpful assistant designed to perform calculations."
+  instruction = {"Addition":"Add the two numbers together i.e. x+y"}
+  context = {"x": 10, "y": 5}
+
+  # Initialize a session with a system message
+  calculator = li.Session(system=system)
+
+  # run a LLM API call
+  result = await calculator.initiate(instruction=instruction,
+                                     context=context)
+
+  print(f"Calculation Result: {result}")
+
+Define your own system message, context, and instruction to initiate interactions with the language model.
+
+Check out examples in the following sections to explore more diverse applications of LionAGI.

docs/index.rst

@@ -87,6 +87,8 @@ Python 3.9 or higher.
    :hidden:
 
    Get_started/installation.rst
+   Get_started/quick_start.rst
+   Get_started/config.rst
 
 .. toctree::
    :maxdepth: 1