This workshop is co-hosted with San Francisco Bay ACM Chapter (https://www.meetup.com/SF-Bay-ACM)
Natural Language Processing (NLP) is the fastest-growing field of deep learning with interest and funding from top AI companies to solve problems of language, text, and unstructured information. This has resulted in a tremendous focus on model building that combines language, mathematics, and computer science.
This training will focus on problems of text summarization, question answering, and sentiment classification using modern approaches to model-building (GNMT, BERT). We will apply this to real-world problems to create an NLP pipeline on top of the PyTorch framework and spaCy.
This four-part training offers both theoretical and practical, lab-heavy modules. By completing you would be able to:
Have working knowledge of PyTorch to train your own deep learning models.
Use OpenVINO to run model optimizer
Use less compute and memory for deploying model inference in production.
Build end to end NLP pipeline with everything you learn
This training is packed with practical exercises and code labs.
We meet twice on the weekend at AICamp online classroom (powered by zoom)
Practical walkthroughs that present solutions to actual, real-world problems and challenges
A no-nonsense teaching style that cuts through all the cruft and help you master NLP and Pytorch
- Session 1: Dec 12, 10am-11:30am PST (US Pacific Time, GMT-8)
- Session 2: Dec 12, 11:30am-1:00pm PST
- Session 3: Dec 13, 10am-11:30am PST
- Session 4: Dec 13, 11:30am-1:00pm PST
- 4 sessions / 6 hours
- Live session (with zoom) and real time interaction
- Slack support during/after class
Check the content tab for full course outlines.
WHO SHOULD LEARN:
Developers, engineers, data scientists
Python coding skills, intro to PyTorch framework is helpful, ML basics, Use Jupyter notebook on a chrome browser
Est. time spend per week: 6 hours live class (required) + 2 hours homework (optional)
Full refund upon request one day before the event starts (Dec 11th, 2020 10:00am PST). 6% transaction fee is not refundable.
If missed live sessions, you can watch recordings any time, along with interactive learning tools, slides, course notes
Students have one year access to course materials
Earn Certificate of Completion
Free one year ACM membership ($25 value)
You will have access to all the notebooks, training material to build your own apps. You should be able to directly work on these using Google Colab. For the Workshop itself, we will have AWS instances available for use.
Module 1: Foundations (20mins)
- Fundamentals and application of Language Modeling
- Classical vs Deep Learning NLP
- NLP Pipeline
Lab: NLK (20mins)
- Setting up your environment
- NLTK (tokenization)
Module 2: NLP for documents (30mins)
- Use NLP pipeline to process documents
- POS, Word embedding
Module 3: NLP with SpaCy (30mins)
- Key packages & librariess
Lab: SpaCy (20mins)
Module 4: TFIDF & Logistic Regression (30mins)
- Logistic Regression
Lab: Disaster Detection (20mins)
Day 1 recap
Lab: Quora, using LSTM (30mins)
Module 5: Pre-trained models (30mins)
Lab: Disaster Detection using BERT (20mins)
Module 6: Sentiment analysis (30mins)
Lab: Headline Classifier using BERT (20mins)
Lab: LSTM based sequence classifier (20mins)
Module 7: Text summarization (30mins)
Lab: Text summarization (20mins)
Module 8: NLP in production (20mins)
- Scheduler Overview
- Implementation for AirFlow
Sr. Data Scientist working on a variety of revenue-generating projects for clients involving machine learning and deep learning. He worked as Sr Data Scientist at Apple for 10 years, and a Sr Program Manager at Applied Materials
Lead AI of Intel, where he focuses on enabling the AI ecosystem on heterogeneous compute. He has over 15 years of technical and enabling experience, spanning optical modeling, statistical analysis, and capital equipment supply chain at Intel. He has over 20 published papers and 4 patents. He has a Ph.D. in EECS from UC Berkeley