Skip to content
No description, website, or topics provided.
Python Shell
Branch: master
Clone or download
Latest commit 6801487 Apr 17, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
benchmark_outputs gcp v100 results Apr 17, 2019
cache add benchmark files Apr 16, 2019
data re run on three prods Nov 17, 2018
hb Initial DBSCAN implementation Nov 14, 2018
jeff integrate jeffs count code, rebuild db with full dataset Nov 9, 2018
papers modified database and data analysis, ready for big run Nov 29, 2018
results ABS output files for 1 product per batch training Dec 13, 2018
sentiment_analysis
.gitignore Updates data.zip Dec 14, 2018
README.md Update README.md Dec 14, 2018
__init__.py
abs_test_set_8.csv working ABS code; TODO: insert EOS token for periods Dec 10, 2018
abs_test_set_8_all.csv working ABS code; TODO: insert EOS token for periods Dec 10, 2018
abs_train_set_8.csv
abs_train_set_8_all.csv working ABS code; TODO: insert EOS token for periods Dec 10, 2018
bench_sentiment.sh bench Apr 17, 2019
bench_sentiment_all.sh bench Apr 17, 2019
config.py default settings ammended Dec 14, 2018
create_database.py
create_database_server.py tensorflow advanced hub encoders Dec 1, 2018
data.json move hb files to outer, move jeffs file inside Nov 8, 2018
data.zip Updates data.zip Dec 14, 2018
data_stats.py modified database and data analysis, ready for big run Nov 29, 2018
data_utils.py tf garbage collection for faster preprocessing Dec 2, 2018
df2use_test.csv integrate jeffs count code, rebuild db with full dataset Nov 9, 2018
df2use_train.csv integrate jeffs count code, rebuild db with full dataset Nov 9, 2018
evaluate_abs.py abs eval for sentiment complete, uploading Dec 13, 2018
extractive_summ_modules.py preprocessed tf hub embeddings working Dec 2, 2018
final.pdf final reportv2 Dec 15, 2018
frequency.png new freq chart Nov 30, 2018
main_abs.py submission updates Dec 14, 2018
main_cluster.py submission updates Dec 14, 2018
main_encode.py move hb files to outer, move jeffs file inside Nov 8, 2018
model_data.py submission updates Dec 14, 2018
modules.py final ABS with eval pipeline Dec 13, 2018
num_reviews.csv new freq chart Nov 30, 2018
num_reviews_filt.csv move hb files to outer, move jeffs file inside Nov 8, 2018
requirements.txt Update requirements.txt Dec 13, 2018
sequence_modules.py
start_exp.sh start script mods Dec 1, 2018
summary_dict.json move hb files to outer, move jeffs file inside Nov 8, 2018
summary_dict_proposal.json integrate jeffs count code, rebuild db with full dataset Nov 9, 2018
text_encoders.py use batching bug fix Dec 1, 2018
tf_bench_template.sh bench Apr 16, 2019

README.md

Recognizing themes in Amazon reviews through Unsupervised Multi-Document Summarization

Installing

conda create --name cs221_project python=3.6
source activate cs221_project
pip install -r requirements.txt

Unzip data

unzip data.zip

Running extractive

mkdir tmp; 
export TFHUB_CACHE_DIR=tmp; 
python main_cluster.py --prepare_embeddings=True --embeddings_preprocessed=False

Running abstractive

python model_data.py
python main_abs.py --train_abs=True --debug=False --test_abs=False --cold_start=True
python main_abs.py --train_abs=False --debug=False --test_abs=True --cold_start=False
You can’t perform that action at this time.