Say goodbye to sleepless nights reading endless scientific papers for your research. Meet P2P_GPT (Papers-to-Paper GPT), the ultimate research tool that magically creates a new paper from your stack of research. Less coffee breaks than your grad student research assistant, but with all the write skills.

---

Get Started Locally 🧙‍♂️

Can't wait to dive into the academic ocean with P2P_GPT? Don your wizard's hat and follow these magical spells to run the program locally:

1. Clone the Ancient Scroll: Don't wait for the owl, get the magic right away!

   git clone https://github.com/tsvant/P2P_GPT
   

2. Brew the Potion: Mix all the required ingredients by conjuring:

   pip install -r requirements.txt
   

3. Whisper the Secret Key: Modify the config.py with your own OpenAI API key. Be mindful of your treasure chest; set up API usage hard limits to avoid going bankrupt. Handle with care, young sorcerer!

4. Summon the Research Spirit: Open main.py and ignite the fire. Behold the message:

   In order to start creating your own research paper,
    based on the sources you provide, you should do two things:
    1. Upload .txt files of your desired sources in the 'Upload here' folder of the project.
    2. When you finish step 1, fill in the research goal you have in mind in the blank below:
    Please enter the goal of your research paper:
   

5. Embark on the Quest: Follow the instructions of that mystical message, and let the script be your guiding star. At out journey's end, or should I say at the edge of discovery, your new scroll of enlightenment resides in:

   ...\P2P Output\0. Brand new research paper\Final research paper.docx
   

6. Brace for Adventure: Our spellbook is still being penned, so errors might appear. Fear not! 90% of the time, a simple retry of main.py fixes it. We've enchanted it with some occasional checkmarks, but the save system is more of an ancient scroll than a modern spell.

Voilà! Your Academic Parchment is Ready! 🎉 Start brewing insights with P2P_GPT!

---

How It Works 🛠

P2P_GPT performs like a Swiss Army knife in a jungle of academia:

1. File Processing: Carves your papers into bite-sized chunks. Akin to pre-chewed food.
2. Metadata Extraction: Even the introverted abstract and shy metadata can't hide. We find them. We extract them. We categorize.
3. Goal Alignment: Checks paper alignment with your goals better than a chiropractor with scoliosis.
4. Chunk Analysis: Evaluates each chunk, like a cheese critic.
5. Research Findings Generation: Extracts research findings faster than a pioneer on a gold rush.
6. Literature Review & Main Body Generation: Cooks up main sections and literature review, like a well-trained digital Gordon Ramsay. Always tasty, minus the 'F-words'.
7. Conclusion & Title Generation: Spits out a concise conclusion and a catchy title. Silent, but deadly.
8. List of Literature & Abstract Generation: Rolls out references smoother than a yoga teacher navigating a complex pose. The abstract? You'll find it just breathtaking.
9. Paper Generation: In the end, stitches everything into a word-quilt that's academically warm and fluffy.

---

What's Next 🚀

1. Multilingual Support: More languages! Adiós, auf wiedersehen, au revoir to language barriers.
2. More File Formats: PDFs aren't books' yet-to-evolve cousin. We get it and we're on it.
3. Customization and Peer-Review Simulation: You comment on a particular section for it to be regenerated accordingly. Papa knows best.
4. Parallelization: The more you have sources, the more excited you should be about it.
5. Integration with Digital Libraries: My lawyer said not to mention specifics.

We've got your writing covered! Happy researching! 🎉