Spolling Bree

Spolling Bree is a version of the NYT Spelling Bee game that only accepts fake words that are plausible.

Model

Originally plausibility was rated by a simple statistical model trained on trigrams from every English word.

The current model is machine learning based, and uses a character based RNN with GRUs trained locally on English words.

Play the live demo at bee.jminjie.com.

To deploy at localhost:8998, run python3 server.py debug

Name		Name	Last commit message	Last commit date
Latest commit History 44 Commits
chars_from_ids		chars_from_ids
ids_from_chars		ids_from_chars
static		static
templates		templates
training_checkpoints		training_checkpoints
.gitignore		.gitignore
20k.txt		20k.txt
README.md		README.md
base_model_saved_weights.data-00000-of-00001		base_model_saved_weights.data-00000-of-00001
base_model_saved_weights.index		base_model_saved_weights.index
checkpoint		checkpoint
dailyletters.py		dailyletters.py
load_model.ipynb		load_model.ipynb
plausiblewords.py		plausiblewords.py
records.py		records.py
rnn_gru_model.py		rnn_gru_model.py
rnn_plausiblewords.py		rnn_plausiblewords.py
server.py		server.py
test_dailyletters.py		test_dailyletters.py
test_plausiblewords.py		test_plausiblewords.py
test_rnnplausiblewords.py		test_rnnplausiblewords.py
train_and_save_model-20k.ipynb		train_and_save_model-20k.ipynb
train_and_save_model.ipynb		train_and_save_model.ipynb
words_alpha.txt		words_alpha.txt