Skip to content

Latest commit

 

History

History
44 lines (30 loc) · 1.28 KB

README.md

File metadata and controls

44 lines (30 loc) · 1.28 KB

Stanford CS224n Winter 2019

Natural Language Processing with Deep Learning - Assignment Solutions

Read the handout PDFs in each assignment folder for instructions and details.

Results Showcase

Assignment 1 (Word Vector Visualization & Interpretation)

Please see the Notebook.

Assignment 2 (Word2Vec)

Plot of Word Vector Space.

Plot

Assignment 3 (Dependency Parsing)

  • DEBUG = True Output Log.

    • Best Avg Train Loss - 0.116
    • Best Dev UAS - 73.76

  • DEBUG = False Output Log.

    • Best Avg Train Loss - 0.058
    • Best Dev UAS - 88.78
    • Best Test UAS - 88.91

Assignment 4 (Neural Machine Translation)

Click here to see the Test Output.

Console Output:

(local_nmt) $ sh run.sh test
load test source sentences from [./en_es_data/test.es]
load test target sentences from [./en_es_data/test.en]
load model from /media/storage2/nmt/model.bin
Decoding: 100%|███████████████████████████████████████████████████| 8064/8064 [04:20<00:00, 30.90it/s]
Corpus BLEU: 22.768024049390718

Assignment 5

Coming Soon!