Natural Language Processing with Deep Learning and TensorFlow

on YouTube ↗54m

" summary

Walks through a practical pipeline: use a 128-dim TF Hub sentence embeddings as input to a DNN (512, 128 hidden layers) trained with TensorFlow Estimators for binary sentiment on the IMDb dataset, reaching roughly ≈79% accuracy on IMDb test set after 1,000 steps; builds a character-level generator with 512-unit LSTM with 128-dim embeddings using Keras over TensorFlow to mimic Shakespeare (training needs many epochs to improve); shows Cloud ML Engine for GPU/TPU-backed training, distributed training and hyperparameter tuning.

AI-written summary · about

" chapters (17)

1. 0:00Introduction
2. 1:02Speaker background
3. 3:11NLP overview
4. 5:57Complex NLP tasks
5. 7:16Sentiment analysis challenges
6. 8:45AI in Sci-Fi vs Reality
7. 10:21Chatbots and QA
8. 13:16Sentiment analysis demo
9. 16:58Text feature engineering
10. 22:52TensorFlow Hub
11. 25:41Estimator API
12. 26:26Demo: Sentiment analysis
13. 33:24Text generation
14. 34:16Shakespeare demo
15. 40:06Model training and tuning
16. 45:24Scaling with Cloud ML
17. 52:08Conclusion and Q&A

" related talks

• 50mDeep Learning and NLP with SparkAndy Petrella · Scala Days Berlin 2016
• 01h 59mDeep Learning and Modern NLPZachary S. Brown · PyData Miami 2019
• 01h 30mIntroduction to Deep Learning & Natural Language ProcessingRaghotham Sripadraj, Nischal HP · PyData London 2016
• 45mDeep Learning and NLP with SparkMelanie Warrick, Andy Petrella · Scala Days Berlin 2016
• 49mBen Hall & Barbara FusinskaBen Hall, Barbara Fusinska · Copenhagen Developers Festival 2023
• 45mDeep Learning for Natural Language ProcessingDevashish Shankar · The Fifth Elephant 2015
• 01h 26mBagging to BERT- A tour of applied NLPBenjamin Batorsky · PyData NYC 2022
• 01h 07mDeep Learning with TensorFlowSeth Juarez · NDC Oslo 2018