Skip to content

kuroazai/ML_Tinder

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Tinder Machine Learning

WORK IN PROGRESS STATUS Completion MADE WITH MADE WITH MADE WITH MADE WITH MADE WITH

Currently Absolutely broken as I overhual it with my new found experience with machine learning and overall programming this wonderful "research" tool will be rolled out SOON. This is soley for "research" purposes for ultilizing image classification with keras base models, google resnet and inception models. current highest accuracy model has achieved is 85% and the model/weights will not be uploaded.

Please read tinder tos before using this script, use at your own discretion.

Installation

Below are the libraries required to run the script

  pip install -r requirements.txt

Model implementation Roadmap

  • Implement Google Resnet [X]

  • Implement Inception v3 [X]

  • Implement GPT-3 [O] - due to the nature of the model it is not suitable for this task

Running Instructions

STEP 1 - Get your data

"python tinder.py mode=data", this will launch selenium firefox/whichever browser you set it to run with and continously store your likes/dislikes. This requires your browser to be already logged into tinder in the browser profile you provide.

STEP 2 - Train on your gathered data

"python tinder_model.py", Once you have gathered n number of images you can start attempting to train your model. my epoch count is quite ridculous but I am training on 6 GPUs.

STEP 3 - Test Validate

"python tinder.py mode=auto" , This will start "research" swiping based on the models prediction on what you find attractive.

STEP 4 - Repeat entire process until you're happy or give up and embrace the cold embrace of loneliness.

About

"Research" combing image classification and natural language processing to find your ideal lover whilst you sleep. Use at your own risk :3

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages