From 3e6de2c92d8f86712d7964c48397f8816205a916 Mon Sep 17 00:00:00 2001 From: tkoooo <100322984+tkoooo@users.noreply.github.com> Date: Tue, 30 Apr 2024 23:08:27 -0500 Subject: [PATCH] Update README.md --- README.md | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index a3bb7e9..74092aa 100644 --- a/README.md +++ b/README.md @@ -1,12 +1,14 @@ # CMPS 6730 Project Pickup # Mackenzie Bookamer and Thalia Koutsougeras -**Are you a graph? -Because you’ve got my heart all connected.** +***Are you a graph? +Because you’ve got my heart all connected.*** That's the power of the *PickupMaster 9000*. This CS/math-themed pickup line generator is here to meet your needs, whether it's to complete the end of your pickup line or to practice honing your skills with you. By fine-tuning the pre-trained language model (PLM) GPT-2 from OpenAI with a wide variety of math and computer science pickup lines, we were able to generate coherent and subject-relevant pickup lines. We pull the PLM from the HuggingFace model hub and use the GPT-2 Tokenizer as well as Transformers Trainer class to process and train our data, then subsequently feed user input back through our fine-tuned model to generate a response. To facilitate user interaction with our model, we created a Flask app that takes in user input and generates 2 pickup lines from our fine tuned model - better than the PLM without fine-tuning! Below is a look at our Flask app in action! +flask pg1 +flask pg2 ### Contents