Topic modelling using RAPIDS and BERT #41

mayankanand007 · 2021-11-19T17:11:23Z

Workflow to read a set of documents and extract topics from it leveraging BERT, TF-IDF and NVIDIA RAPIDS.

review-notebook-app · 2021-11-19T17:11:27Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

VibhuJawa

Thanks for all the work on this @mayankanand007 . Work looks great.

I have done an inital review and requested changes.

cuBERT_topic_modelling/README.md

cuBERT_topic_modelling/cuBERTopic.py

cuBERT_topic_modelling/tests/test_umap_dr.py

cuBERT_topic_modelling/cuBERTopic.py

…apids-examples into topic_modelling

cuBERT_topic_modelling/tests/test_subwordtokenizer.py

cuBERT_topic_modelling/cuBERTopic.py

cuBERT_topic_modelling/utils/sparse_matrix_utils.py

mayankanand007 · 2021-11-25T22:47:27Z

couple of things left, but I think this is good for another round of review 😄 as a lot of code has been refactored.

VibhuJawa

Thanks for working through this. Have requested additional changes.

cuBERT_topic_modelling/README.md

cuBERT_topic_modelling/cuBERTopic.py

cuBERT_topic_modelling/tests/test_ctfidf.py

cuBERT_topic_modelling/tests/test_subwordtokenizer.py

cuBERT_topic_modelling/tests/test_umap_dr.py

cuBERT_topic_modelling/README.md

cuBERT_topic_modelling/ctfidf.py

cuBERT_topic_modelling/tests/test_data_preprocess.py

cuBERT_topic_modelling/cuBERTopic.py

VibhuJawa

Thanks for working through the reviews The PR looks close, have requested some more changes.

Also, ensure that you have run flake8/black on the test python files aswell.

cuBERT_topic_modelling/cuBERTopic.py

cuBERT_topic_modelling/tests/test_ctfidf.py

cuBERT_topic_modelling/embedding_extraction.py

VibhuJawa

Thanks a lot for pushing all these changes through @mayankanand007 as well as testing it on the whole dataset.

I think we are close to completion, have requested very minor changes but we should be good to merge soon.

cuBERT_topic_modelling/ctfidf.py

cuBERT_topic_modelling/benchmarks/benchmark_berttopic.ipynb

cuBERT_topic_modelling/benchmarks/benchmark_cuBERTopic.ipynb

cuBERT_topic_modelling/benchmarks/benchmark_fit_transform_default.ipynb

cuBERT_topic_modelling/berttopic_example.ipynb

VibhuJawa

Thanks for working on this Mayank. LGTM

mayankanand007 and others added 8 commits November 18, 2021 20:37

added topic modelling code

efb7224

squashing bugs

eecc25e

remove prints

5761738

fix prints

43b7522

updated readme with relevant information

afed317

Update README.md

33a81f9

Update README.md

f753078

added topic modelling example link

6ee8924

mayankanand007 requested a review from VibhuJawa November 19, 2021 17:11

VibhuJawa requested changes Nov 19, 2021

View reviewed changes

VibhuJawa added the Waiting on Author Waiting on Author label Nov 19, 2021

mayankanand007 added 8 commits November 22, 2021 14:55

fix bugs

2f7fe75

Merge branch 'topic_modelling' of https://github.com/mayankanand007/r…

2f2727d

…apids-examples into topic_modelling

addressing reviews part 1

678da80

updated readme with installation steps

a3466f5

clean up commented code

747b8cf

updated example notebook

aecd7de

fix issues with matrix manipulation

158f990

integrated subwordtokenizer and code refactoring

ad652e9

mayankanand007 commented Nov 25, 2021

View reviewed changes

cuBERT_topic_modelling/tests/test_subwordtokenizer.py Show resolved Hide resolved

mayankanand007 commented Nov 25, 2021

View reviewed changes

cuBERT_topic_modelling/cuBERTopic.py Outdated Show resolved Hide resolved

mayankanand007 commented Nov 25, 2021

View reviewed changes

cuBERT_topic_modelling/utils/sparse_matrix_utils.py Outdated Show resolved Hide resolved

style fixes

e9c090c

mayankanand007 removed the Waiting on Author Waiting on Author label Nov 26, 2021

mayankanand007 added 3 commits November 26, 2021 20:27

fixed embedding extraction base case

ea8455b

fixed style, added docstrings

ef9668a

dlpack optimizations

217fe06

VibhuJawa requested changes Nov 28, 2021

View reviewed changes

VibhuJawa reviewed Dec 2, 2021

View reviewed changes

cuBERT_topic_modelling/README.md Outdated Show resolved Hide resolved

VibhuJawa reviewed Dec 2, 2021

View reviewed changes

cuBERT_topic_modelling/ctfidf.py Outdated Show resolved Hide resolved

VibhuJawa reviewed Dec 2, 2021

View reviewed changes

cuBERT_topic_modelling/tests/test_data_preprocess.py Outdated Show resolved Hide resolved

VibhuJawa reviewed Dec 2, 2021

View reviewed changes

cuBERT_topic_modelling/cuBERTopic.py Outdated Show resolved Hide resolved

VibhuJawa reviewed Dec 2, 2021

View reviewed changes

cuBERT_topic_modelling/cuBERTopic.py Outdated Show resolved Hide resolved

VibhuJawa requested changes Dec 2, 2021

View reviewed changes

cuBERT_topic_modelling/cuBERTopic.py Outdated Show resolved Hide resolved

cuBERT_topic_modelling/cuBERTopic.py Outdated Show resolved Hide resolved

cuBERT_topic_modelling/tests/test_ctfidf.py Show resolved Hide resolved

mayankanand007 added 12 commits December 8, 2021 23:00

addressed some code reviews

55396e9

addressed reviews part 2

b4b71a5

moved vocabulary files

d7a6700

fixed spacing issue in README

2229aa9

updated diagram with better clarity

c5054f2

updated directory tree

6b723b6

created conda-env files for dev, prod

10937c0

added clarification regarding CUDA version

7428484

fixed spelling error

b6e21d5

removed redundant code to sort by frequency

ff177fb

refactored fit_transform test based on the new logic

7b3e2ee

some del statements to free up GPU memory

96e3cd7

mayankanand007 commented Dec 13, 2021

View reviewed changes

cuBERT_topic_modelling/embedding_extraction.py Outdated Show resolved Hide resolved

mayankanand007 added 7 commits December 15, 2021 01:25

updated example with correct runtimes

89d6f4b

updated example notebook with time

c24b9a2

created scripts to compare and benchmark implementations

2b26f14

resolved CUDA OOM issue in embedding extraction

13fc7e7

updated end to end example with RMM enabled

e1f9ae5

post updated example

d36a7ef

fixed emptydoc issue in preprocessing and updated test

bf60566

VibhuJawa requested changes Dec 21, 2021

View reviewed changes

updated tokenizer parallelism flag and module imports

f0ce6a1

VibhuJawa approved these changes Dec 21, 2021

View reviewed changes

VibhuJawa merged commit ce24e3d into rapidsai:main Dec 21, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Topic modelling using RAPIDS and BERT #41

Topic modelling using RAPIDS and BERT #41

mayankanand007 commented Nov 19, 2021

review-notebook-app bot commented Nov 19, 2021

VibhuJawa left a comment

mayankanand007 commented Nov 25, 2021

VibhuJawa left a comment

VibhuJawa left a comment

VibhuJawa left a comment

VibhuJawa left a comment

Topic modelling using RAPIDS and BERT #41

Topic modelling using RAPIDS and BERT #41

Conversation

mayankanand007 commented Nov 19, 2021

review-notebook-app bot commented Nov 19, 2021

VibhuJawa left a comment

Choose a reason for hiding this comment

mayankanand007 commented Nov 25, 2021

VibhuJawa left a comment

Choose a reason for hiding this comment

VibhuJawa left a comment

Choose a reason for hiding this comment

VibhuJawa left a comment

Choose a reason for hiding this comment

VibhuJawa left a comment

Choose a reason for hiding this comment