__author__ = "Vinay Kumar"
__year__ = "2019"
__project__ = "flixstockJam"

Hi,

I have provided 3 methods to solve the problem sattement assigned.

Method-1: Using Jupyter notebook and fastai library. I trained the models in Kaggle servers src/flixstockJam.ipynb You can also find my notebook on Kaggle at https://www.kaggle.com/vinaykumar2491/flixstockjam/

Method-2: A Python3.6+ version of the same Method-1 as a python script. src/run_fast_flixNet.py

Method-3: A pure Pytorch 1+ implementation of the solution without using fastai library. All the datawrangling and cleaning was done by me and scripts are present in this repo. src/run_torch_flixNet.py All the written support modules are in src/modules

I have written the documention of the methods and reasons in the notebooks and scripts itself.
Please read the corrsponding comments and documentation.

If given more time to experiment and test the problem with adequate hardware support,
I am confdent of finding better models with better performance.

Thank you Vinay Kumar