├── README.md
└── infiniteGPT
    └── blastoff.py


/README.md:
--------------------------------------------------------------------------------
 1 | # InfiniteGPT 🚀
 2 | 
 3 | InfiniteGPT is a Python script that lets you input an unlimited size text into the OpenAI API. No more tedious copy & pasting. It's a single python script that can connect to any of the OpenAI API chat models (gpt-3.5-turbo, gpt4, etc.). 
 4 | 
 5 | This eliminates the need for re-prompting when using a large text input or copying and pasting endless chunks of text into chatGPT. 📚
 6 | 
 7 | ## Dependencies 📦
 8 | 
 9 | - python3
10 | - openai
11 | 
12 | ## BYOK (Bring Your Own Keys!) 🔑
13 | 
14 | Go to [OpenAI](https://www.openai.com) to get your personal API keys. 
15 | 
16 | This script does not hide your API keys, so please do so if you plan on integrating it into a public application. ⚠️
17 | 
18 | ## Usage 🛠️
19 | 
20 | 1. Clone the repository
21 | 2. Install the required dependencies
22 | 3. Add your API keys
23 | 4. Run the script
24 | 
25 | ## License 📄
26 | 
27 | MIT License. See [LICENSE](LICENSE) for more information.
28 | 
29 | ## Connect with me 📣
30 | 
31 | I write about using AI tools & share my latest building on Twitter [@ehalm_](https://twitter.com/ehalm_). DM me with any questions. 🐦
32 | 
33 | ## Happy building! 🎉
34 | 


--------------------------------------------------------------------------------
/infiniteGPT/blastoff.py:
--------------------------------------------------------------------------------
 1 | import openai
 2 | from concurrent.futures import ThreadPoolExecutor
 3 | import tiktoken
 4 | 
 5 | # Add your own OpenAI API key
 6 | 
 7 | openai.api_key = "sk-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"
 8 | 
 9 | def load_text(file_path):
10 |     with open(file_path, 'r') as file:
11 |         return file.read()
12 | 
13 | def save_to_file(responses, output_file):
14 |     with open(output_file, 'w') as file:
15 |         for response in responses:
16 |             file.write(response + '\n')
17 | 
18 | # Change your OpenAI chat model accordingly
19 | 
20 | def call_openai_api(chunk):
21 |     response = openai.ChatCompletion.create(
22 |         model="gpt-3.5-turbo",
23 |         messages=[
24 |             {"role": "system", "content": "PASS IN ANY ARBITRARY SYSTEM VALUE TO GIVE THE AI AN IDENITY"},
25 |             {"role": "user", "content": f"YOUR DATA TO PASS IN: {chunk}."},
26 |         ],
27 |         max_tokens=500,
28 |         n=1,
29 |         stop=None,
30 |         temperature=0.5,
31 |     )
32 |     return response.choices[0]['message']['content'].strip()
33 | 
34 | def split_into_chunks(text, tokens=500):
35 |     encoding = tiktoken.encoding_for_model('gpt-3.5-turbo')
36 |     words = encoding.encode(text)
37 |     chunks = []
38 |     for i in range(0, len(words), tokens):
39 |         chunks.append(' '.join(encoding.decode(words[i:i + tokens])))
40 |     return chunks   
41 | 
42 | def process_chunks(input_file, output_file):
43 |     text = load_text(input_file)
44 |     chunks = split_into_chunks(text)
45 |     
46 |     # Processes chunks in parallel
47 |     with ThreadPoolExecutor() as executor:
48 |         responses = list(executor.map(call_openai_api, chunks))
49 | 
50 |     save_to_file(responses, output_file)
51 | 
52 | # Specify your input and output files
53 | 
54 | if __name__ == "__main__":
55 |     input_file = "test_input.txt"
56 |     output_file = "output_og.txt"
57 |     process_chunks(input_file, output_file)
58 | 
59 | # Can take up to a few minutes to run depending on the size of your data input
60 | 


--------------------------------------------------------------------------------