Chunkgpt

- Added the `final_step` parameter to `Chunker.summarize()`, allowing users to determine whether ChunkGPT should summarize the intermediate summaries (`final_step='summarize'`), or return a concatenation of the intermediate summaries (`final_step='combine'`)
- Addresses [this community suggestion](https://github.com/dimays/chunkgpt/issues/9)

- Updated documentation to reflect actual import syntax

From:
`from chunkgpt import Chunker`

To:
`from chunkgpt.chunkgpt import Chunker`

- Moved requirements.txt back to root
- Added setuptools config to include requirements.txt in pip install
- Re-built package

- Moved requirements.txt into src so they can be included with pip install
- Incremented version number

🚀 Announcing ChunkGPT!

A new way to summarize large text files using OpenAI's GPT-based models.

Features include:
- Ability to easily summarize giant bodies of text without ever worrying about token limits
- Gracefully retry generating chat completions when the server is unavailable or a rate limit is reached
- Easily track your tokens used and estimated cost (based on August 2023 pricing)
- Flexible parameters to customize advanced behavior and tweak settings to suit your project

Check out the [README](https://github.com/dimays/chunkgpt#readme) for more usage details and examples!