https://zhiyunrichardpeng.github.io/nlp-class/

1<!DOCTYPE html>
2<html lang="en">
3  <head>
4    <meta charset="utf-8">
5    <meta http-equiv="X-UA-Compatible" content="IE=edge">
6    <meta name="viewport" content="width=device-width, initial-scale=1.0">
7    <link rel="shortcut icon" type="image/x-icon" href="/nlp-classfavicon.ico">
8    <title>SFU NLP class: Course Information</title>
9
10    <!-- CSS -->
11    <link href="/nlp-class/dist/css/bootstrap.min.css" rel="stylesheet">
12    <link href="/nlp-class/dist/css/bootstrap-glyphicons.css" rel="stylesheet">
13    <link href="/nlp-class/assets/css/nlp-class.css" rel="stylesheet">   
14    <link href="//netdna.bootstrapcdn.com/font-awesome/4.0.3/css/font-awesome.css" rel="stylesheet">
15
16    <!-- HTML5 shim and Respond.js IE8 support of HTML5 elements and media queries -->
17    <!--[if lt IE 9]>
18      

18<script src="https://oss.maxcdn.com/libs/html5shiv/3.7.0/html5shiv.js"></script>

18
19      

19<script src="https://oss.maxcdn.com/libs/respond.js/1.3.0/respond.min.js"></script>

19
20    <![endif]-->
21
22    

22<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.10.2/jquery.min.js"></script>

22
23
24    <!-- MathJax
25    

25<script type="text/javascript" async
26      src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.4/MathJax.js?config=TeX-MML-AM_CHTML">
27    </script>

27
28    -->
29    

29<script type="text/x-mathjax-config">
30  MathJax.Hub.Config({
31    TeX: {
32      equationNumbers: {
33        autoNumber: "AMS"
34      }
35    },
36    tex2jax: {
37      inlineMath: [ ['$','$'], ['\\(', '\\)'] ],
38      displayMath: [ ['$$','$$'], ['\\[', '\\]'] ],
39      processEscapes: true,
40    }
41  });
42</script>

42

43<script type="text/javascript"
44	src="//cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML">
45</script>

45
46
47  </head>
48  <body>
49    <a class="sr-only" href="#content">Skip to main content</a>
50
51    <!-- Docs master nav -->
52    <header class="navbar navbar-fixed-top navbar-default" role="banner">
53      <div class="container">
54        <div class="row">
55        <div class="navbar-header">
56          <button class="navbar-toggle" type="button" data-toggle="collapse" data-target=".navbar-collapse">
57            <span class="sr-only">Toggle navigation</span>
58            <span class="icon-bar"></span>
59            <span class="icon-bar"></span>
60            <span class="icon-bar"></span>
61          </button>
62          <ul class="nav navbar-nav">
63            <li id="main_page"><a href="/nlp-class/index.html" class="navbar-brand">Natural Language Processing</a></li>
64          </ul>
65        </div>
66        <nav class="collapse navbar-collapse" role="navigation">
67          <ul class="nav navbar-nav">
68            <li id="syllabus">
69              <a class="dropdown-toggle" data-toggle="dropdown" href="#">Syllabus<span class="caret"></span></a>
70              <ol class="dropdown-menu">
71                <li><a href="/nlp-class/schedule.html">Schedule</a></li>
72                <li><a href="/nlp-class/syllabus.html">Syllabus (not maintained in 2026 fall)</a></li>
73              </ol>
74            </li>
75             <li id="homework">
76              <a class="dropdown-toggle" data-toggle="dropdown" href="#">Homework <span class="caret"></span></a>
77              <ol class="dropdown-menu">
78                <li><a href="/nlp-class/hw0.html">0. Setup</a></li>
79                <li><a href="/nlp-class/hw1.html">1. Contextual Spell Checking</a></li>
80                <li><a href="/nlp-class/hw2.html">2. BERT Finetuning</a></li>
81                <li><a href="/nlp-class/hw3.html">3. Cross Attention</a></li>
82                <li><a href="/nlp-class/hw4.html">4. Prompt Tuning for Text Generation</a></li>
83              </ol>
84            </li>
85             
86            <li id="project"><a href="project.html">Project</a></li>
87            <li id="resources"><a href="resources.html">Resources</a></li>
88            <li id="faq"><a href="/nlp-class/faq.html">FAQ</a></li>
89          </ul>
90        </nav>
91      </div>
92      </div>
93    </header>
94
95    <div class="container">
96      <div class="row">
97        <div class="col-sm-2 hidden-sm hidden-xs">
98        
99          <a href="/nlp-class/http://en.wikipedia.org/wiki/Languages_in_Star_Wars"> 
100          <img src="/nlp-class/assets/img/C-3PO.jpg" class="img-responsive img-rounded" alt=""/>
101          </a> 
102          <span class="text-muted"><i>In Star Wars, C-3PO is fluent in over six million forms of communication.</i></span>
103        
104        </div>
105        <div class="col-sm-10">
106          <h2 id="natural-language-processing-fall-2026">Natural Language Processing <span class="text-muted">Fall 2026</span></h2>
107
108<h4 id="announcement-updated-sept-29">Announcement (Updated Sept 29)</h4>
109
110<p>Hi class,</p>
111
112<p>The reformation, taking the inputs from students and previous instructors, after the meeting with our teaching team and majority vote, is as below:</p>
113
114<ol>
115  <li>The quiz will be discontinued. Its 10% grade moves to 5% each for midterm and final. Please use the conceptual assignments to review the course materials.</li>
116  <li>quiz 0: in the first class, no grading. quiz 1 (1 score) and quiz 2 (1 score), the grades will be bonus grades. The bonus scores are added AFTER the curve. That means, if you lost these two bonus scores, but you obtained 100 scores for the rest of the tasks, you will have 100 scores after the curve.</li>
117  <li>10% engagement changes to 10% in-person attendance.</li>
118  <li>We only use main website and Coursys (all submissions go to here) from now on. I.e., we discontinue the usage of Gradescope and Canvas. The Deadline and course information in Coursys will synchronize with the main website.</li>
119  <li>We will try to provide sample exams for midterm and final exam. A calculator, one page (double sided) cheat sheet, are allowed. Final exam date is not determined.</li>
120  <li>Regarding “weekly online quizzes that allow unlimited attempts”, I like this idea, but unfortunately the teaching team has voted against it. I may bring this to the next NLP course design.
121-In the meantime, you can still make use of the AI to design the “quizzes that allow unlimited attempts”, to achieve the learning effect. Also, the small questions I asked in class, and shown in the slides, can also be the self-evaluation quizzes.</li>
122</ol>
123
124<p>

124The basic idea behind these changes is: a) Enforce the fairness; b) Keep the grading rubric stable after this reformation. The reformation may not satisfy everyone, but we have tried our best to make the decision.</p>
125
126<p>Take a good rest on Sept 30th’s holiday and see you on Oct 2nd,<br />
127Richard</p>
128
129<hr />
130
131<p>Imagine a world where you can pick up a phone and talk in English,
132while at the other end of the line your words are <a href="https://www.youtube.com/watch?v=Nu-nlQqFCKg">spoken in
133Chinese</a>.  Imagine a
134<a href="http://mitpress.mit.edu/books/embodied-conversational-agents">computer animated representation of
135yourself</a>
136speaking fluently what you have written in an email. Imagine a computer <a href="https://www.gwern.net/GPT-3">writing new poetry and stories</a> from a prompt or <a href="https://huggingface.co/spaces/stabilityai/stable-diffusion">generating art 
137based on descriptions</a>. Imagine
138automatically uncovering protein/drug interactions in <a href="http://fable.chop.edu/">petabytes
139of medical abstracts</a>. Imagine feeding a
140computer an ancient script that no living person can read, then
141listening as <a href="https://isi.edu/natural-language/mt/decipher.html">the computer reads aloud in this dead
142language</a>.
143Imagine a computer that can <a href="https://www.youtube.com/watch?v=lI-M7O_bRNg">do better than humans at answering
144questions</a>.</p>
145
146<p>Natural Language Processing is the automatic analysis of human
147languages such as English, Korean, and thousands of others analyzed
148by computer algorithms. Unlike artificially created programming
149languages where the structure and meaning of programs is easy to
150encode, human languages provide an interesting challenge, both in
151terms of its analysis and the learning of language from observations.</p>
152
153<h4 id="instructor">Instructor</h4>
154<ul>
155  <li><a href="https://zhiyunrichardpeng.github.io/">Zhiyun (Richard) Peng</a></li>
156</ul>
157
158<h4 id="teaching-assistants">Teaching Assistants</h4>
159<ul>
160
161<li>Nima Forouzi, <code>nfa23</code>, Office hour: Thursdays 6 to 7 PM on [Zoom](https://sfu.zoom.us/j/83421743065).</li>
162
163<li>Borui Li, <code>bla127</code>, Office hour: Thursdays 6 to 7 PM on [Zoom](https://sfu.zoom.us/j/83421743065).</li>
164
165<li>Qinchan Li, <code>qla126</code>, Office hour: Thursdays 6 to 7 PM on [Zoom](https://sfu.zoom.us/j/83421743065).</li>
166
167</ul>
168
169<h4 id="asking-for-help">Asking for help</h4>
170<ul>
171  <li>Ask for help on <a href="https://coursys.sfu.ca/2026fa-cmpt-713-x1/forum/">Coursys Discussion Forum</a></li>
172  <li>Instructor office hours: On Wednesdays from sept 16 to Nov 29, at 11 - 12 PM, at TASC1 9409, except national holidays.</li>
173  <li><b>No emails</b> to the TAs and strictly emails about personal matters to the instructor</li>
174  <li>Use only SFU email address and use either <code class="language-plaintext highlighter-rouge">cmpt413:</code> or<code class="language-plaintext highlighter-rouge">cmpt713:</code> as subject prefix</li>
175  <li>Always post to the <a href="https://coursys.sfu.ca/2026fa-cmpt-713-x1/forum/">Coursys Discussion Forum</a> instead of email. If you have to email use your SFU email address only.</li>
176</ul>
177
178<h4 id="time-and-place">Time and place</h4>
179<ul>
180  <li>Sep 9 – Dec 6, 2026: Wed, 1:30–2:20 p.m.</li>
181  <li>Sep 9 – Dec 6, 2026: Fri, 12:30–2:20 p.m.</li>
182  <li>location : TBD, Burnaby campus.</li>
183  <li>Last day of classes: Dec 6, 2026</li>
184</ul>
185
186<p>Links to course material will be made available on <a href="https://coursys.sfu.ca/2026fa-cmpt-713-x1/">Coursys</a>.</p>
187
188<!-- #### Calendar
189* [Subscribe](https://calendar.google.com/calendar/ical/qgs97od9ef8t1582k2rqomv5ho%40group.calendar.google.com/public/basic.ics)
190 -->
191
192<h4 id="prerequisites">Prerequisites</h4>
193<p>There are no formal prerequisites for this class.  However, you are expected to be familiar with the following:</p>
194<ul>
195  <li>Proficiency in Python - Programming assignments will be in python (numpy and pytorch will be used).</li>
196  <li>Calculus and Linear Algebra (MATH 151, MATH 232/240) - You will need to be comfortable with taking multivariable derivatives</li>
197  <li>Basic Probability and Statistics (CMPT 210 or STAT 270)</li>
198  <li>Basic Machine Learning (CMPT 410/726) is strongly recommended (Note: C

198MPT 410 was previously offered as CMPT 419 under the title “Machine Learning”)</li>
199</ul>
200
201<p>There will be optional TA led tutorials that will help review these topics.</p>
202
203<h4 id="textbook">Textbook</h4>
204<ul>
205  <li>No required textbook. Online readings provided in Syllabus.</li>
206  <li>Many of the readings will be from the following:
207    <ul>
208      <li><a href="https://web.stanford.edu/~jurafsky/slp3/">Speech and Language Processing</a> by <a href="http://www.stanford.edu/~jurafsky">Dan Jurafsky</a> and <a href="http://www.cs.colorado.edu/~martin">James Martin</a>.</li>
209      <li><a href="https://github.com/jacobeisenstein/gt-nlp-class/blob/master/notes/eisenstein-nlp-notes.pdf">Natural Language Processing</a> by <a href="https://jacobeisenstein.github.io/">Jacob Eisenstein</a></li>
210      <li><a href="http://u.cs.biu.ac.il/~yogo/nnlp.pdf">A Primer on Neural Network Models for Natural Language Processing</a> by Yoav Goldberg (see also <a href="http://www.morganclaypool.com/doi/10.2200/S00762ED1V01Y201703HLT037">Neural Network methods for Natural Language Processing</a>).</li>
211    </ul>
212  </li>
213</ul>
214
215<h4 id="grading">Grading</h4>
216<ul>
217  <li>Submit all homework assignments and check your grades on <a href="https://coursys.sfu.ca/2026fa-cmpt-713-x1/">Coursys</a>. We have discontinued the usage of Gradescope and Canvas. Deadlines and course information in Coursys will synchronize with the main website.</li>
218  <li>Programming setup and diagnostic homework (4%)
219    <ul>
220      <li>HW0 due on Sep 18, 2026</li>
221    </ul>
222  </li>
223  <li>Four homeworks (16% total - 4% each). Due dates:
224    <ul>
225      <li>HW1 on Oct 7, 2026</li>
226      <li>HW2 on Oct 23, 2026</li>
227      <li>HW3 on Oct 30, 2026</li>
228      <li>HW4 on Nov 20, 2026</li>
229    </ul>
230  </li>
231  <li>Midterm Exam (20%)</li>
232  <li>Final Exam (25%)</li>
233  <li>Final Project (25% total)
234    <ul>
235      <li>Project Proposal: Due on Oct 28, 2026 (3%)</li>
236      <li>Project Milestone: Due on Nov 25, 2026 (5%)</li>
237      <li>Project “Poster” Presentation: Poster due on Dec 4, 2026 (5%)</li>
238      <li>Project Report and Code: Due on TBD (12%)</li>
239    </ul>
240  </li>
241  <li>Participation (10%): In-person attendance.</li>
242  <li>Bonus: Quizzes 1 and 2 (1 point each) will be added to your final score as bonus points outside the standard grading scheme. The bonus scores are added AFTER the curve. Please use the conceptual assignments to review the course materials.</li>
243</ul>
244
245<h4 id="exams">Exams</h4>
246<ul>
247  <li>We will try to provide sample exams for midterm and final exam.</li>
248  <li>A calculator and one page (double sided) cheat sheet are allowed.</li>
249  <li>Final exam date is not determined.</li>
250</ul>
251
252
253        </div>
254      </div>
255
256      <footer class="text-center text-muted">
257        <hr/>
258        Last updated October 02, 2026.<br/>
259        Forked and adapted from Anoop Sarkar's <a href="http://anoopsarkar.github.io/nlp-class/">SFU NLP class</a> which was forked from the JHU MT class code on <a href="https://github.com/mt-class/jhu">github <i class="fa fa-github-alt"></i></a> by <a href="https://github.com/mjpost">Matt Post</a> and <a href="https://github.com/alopez">Adam Lopez</a>.<br/>
260        <br/><br/>
261      </footer>
262    </div>
263
264    <!-- Page content of course! -->
265    <!-- JS and analytics only. -->
266    <!-- Bootstrap core JavaScript
267    ================================================== -->
268    <!-- Placed at the end of the document so the pages load faster -->
269    

269<script src="/nlp-class/dist/js/bootstrap.js"></script>

269
270    

270<script type="text/javascript">
271      $(document).ready(function(){
272        $("#main_page").addClass("active");
273      });
274    </script>

274
275  </body>
276</html>