Movie recommendation system written in Python with PyTorch.
1.1 Create a validation data with the index of the row:
python3 generate_validation_data.py 33
2.1 Make sure to create the trained_models
directory before training a model:
mkdir trained_models
2.2 Start TensorBoard:
tensorboard --logdir=runs
2.3 Train a model:
python3 train_model.py
3.1 Run the program with the file path of the model:
python3 main.py trained_models/model_20241017_124650_93.pt
Note: It is a good idea to pick a model with the highest number of epochs.
A user chooses this as their favorite movie:
Here is the result after running the movie recommendation program 100 times:
Note: The measure_accuracy.py
script runs the program 100 times for me.
The accuracy of the model is 81%. There is certainly more work that has to be done to get it to a 90% accuracy. An accuracy above 75% is pretty good for a demo project.