Skip to content

Latest commit

 

History

History
60 lines (34 loc) · 2.18 KB

README.md

File metadata and controls

60 lines (34 loc) · 2.18 KB

Emogify 😞⚽😇 🍜❤️

we have always struggled to make our text messages more expressive !! such as we text "Let's play football!" and search for the most appropriate emoji to this sentence.

So rather than searching the most appropriate emoji to your sentence "Let's play football" this model can automatically turn this into "Let's play football⚽"

emoji

Datasets

Datasets consisting of different sentences along with the labels. Labels represent a particular emoji representing the sentence.😞⚽😇 🍜❤️

  • test_emoji.csv - for testing our model with shape (315, 4)
  • train_emoji.csv - for training our model with shape (56, 2)

Downloading Link-> Datasets Link

image



GloVe: Global Vectors for Word Representation
-This Project contains English word vectors pre-trained on the combined Wikipedia 2014 + Gigaword
5th Edition corpora (6B tokens, 400K vocab). All tokens are in lowercase. This dataset
contains 50-dimensional, pre trained word vectors.

Downloading Link-> glove.6B.50D.txt


Algorithm

Model
image

LSTM
WorkingOfLSTM

image

Result 🎆

image

Resources 💁🏻‍♂️💁

Keras Sequential model
Understanding LSTM Networks
Guide to LSTM’s