UNIVERSITY BREAKS AND IMPORTANT DATES
Please visit the MIT Registrar Calendar for the latest dates:
- Sept 4 (Wed): Semester begins
- Sept 13 (Fri): Last day for seniors and grad students to change H1 subjects to/from P/D/F
- Sept 20 (Fri): Student holiday (no classes)
- Oct 14 (Mon): Indigenous People’s Day (no classes)
- Nov 11 (Mon): Veteran’s Day (no classes)
- Nov 20 (Wed): Drop date. Last day to cancel full-term subjects.
- Nov 28 (Thurs) - Nov 29 (Fri): Thanksgiving Break
- Dec 11 (Wed): Last day of classes
- Dec 16 (Mon) - Dec 20 (Fri): Exam Period (no classes)
OUR CLASS
- Lectures are on Tuesdays and Thursdays at 11am-12:30am in 32.123
- First lecture is on Sept 5 (Thurs)
- Earlier lectures will mostly concern representation and Deep Learning models
- Later lectures will include guest lectures about particular NLP Problems/Tasks
- The last content-based lecture will be Dec 3 (Tues)
- We will not have class on Oct 15 (Tues) and Nov 28 (Thurs), due to university breaks
- Final Presentations will be on Dec 5 (Thurs) and Dec 10 (Tues)
- Homeworks are due Thursdays at 11:59pm EST and you will have roughly 2 weeks to complete them.
- Research Projects will span 12 weeks of the semester, with several deliverables due throughout
- Project proposals are due on Oct 15 (Tues)
- Project progress reports are due on Nov 12 (Tues)
- All posters must be uploaded on Dec 5 (Thurs)
- Poster sessions will take place on Dec 5 (Thurs) and Dec 10 (Tues)
- Final projects are due on Dec 10 (Tues)
LECTURES
- Lecture 1: Introduction + ML Basics
- Lecture 2: Classification (linear models, neural nets)
- Lecture 3: Sequence models 1 (ngrams, log-linear LMs, word2vec)
- Lecture 4: Sequence models 2 (RNNs)
- Lecture 5: Sequence models 3 (seq2seq + attention)
- Lecture 6: Transformers
- Lecture 7: Pretraining 1 (BERT and GPT)
- Lecture 8: Pretraining 2 (SFT and RLHF)
- Lecture 9: Efficient training (MoE, quantization, LoRA)
- Lecture 10: Doing research
- Lecture 11: Decoding 1 (prompting, CoT, and agents)
- Lecture 12: Decoding 2 (search and sampling)
- Lecture 13: Multimodality
- Lecture 14: Midterm review
- Lecture 15: Midterm
- Lecture 16: Beyond transformers (SSMs, Mamba, …)
- Lecture 17: NLP Engineering
- Lecture 18: Interpretability
- Lecture 19: Guest Lecture: Speech
- Lecture 20: Struct pred
- Lecture 21: Guest Lecture: Intellectual property (SERC)
- Lecture 22: Bias & fairness
- Lecture 23: Human language processing
- Lecture 24: Conclusion
Quick Due Dates
- 9/19 HW1
- 10/3 HW2
- 10/15 Project Proposal
- 10/17 HW3
- 10/29 Midterm
- 11/12 Project Progress Report
- 12/5 Poster upload
- 12/5 & 12/10 Poster sessions
- 12/10 Final project