___notes___preprocessing_part_2.txt

cd to 'preprocessing' folder


run the following command:
python3 pretrained_vectors_simple.py -d amazon_instant_video



NOTE:
    you should have the pretrained word embeddings stored somewhere
    to specify the location of your pretrained word embeddings, go to 'pretrained_vectors_simple.py'
    change the following line:
    input_embeddings    = "../../GoogleNews-vectors-negative300.bin"



There should be 2 files created in the corresponding folder, e.g. datasets/amazon_instant_video/

The log file:
(1) E.g. datasets/amazon_instant_video/amazon_instant_video___pretrained_vectors_log.txt

Files required to run the model:
(2) E.g. datasets/amazon_instant_video/amazon_instant_video_wid_wordEmbed.npy



Purpose of files:
(2) contains the mapping for wid to wordEmbed, and each wordEmbed is a vector
    each wordEmbed either comes from the pretrained word embeddings file, or randomly initialized if not found in the set of pretrained words