visual-learning.github.io/hw3.html at master · visual-learning/visual-learning.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<link rel="icon" href="https://www.ri.cmu.edu/wp-content/uploads/2017/04/ri-favicon.ico">

<title>16824 Fall 2025</title>

  <link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.2.0/css/bootstrap.min.css">
  <link href="css/style.css" rel="stylesheet" type="text/css" />
</head>

    <body>
        <div class="container">
                <table border="0" align="center">
                    <tr>
                        <td width="900" align="center" valign="middle">
                        <span class="title"> <h1>16-824: Visual Learning and Recognition</h1></span></td>
                    </tr>
                        <td colspan="3" align="center"><h3>Fall 2025</h3> </td>
                    </tr>
                    <tr>

                    <td colspan="3" align="center"><span class="menubar">[ <a href="index.html">Home</a> | <a href="schedule.html">Schedule</a> | <a href="resources.html">Assignments and Resources</a> | <a href="https://piazza.com/cmu/fall2025/16824/">Piazza</a> | <a href="previous.html">Previous Offerings</a>]</span></td>
                    </tr>
                    </table>
                    <br>

            <h2>Homework 3: Visual Question Answering with PyTorch </h2>
            <div class="col-md-12">
                <p> This homework is due on <b>11:59 PM ET November 13th, 2025</b>. We will solve two tasks with Transformers.</p>

                <br>
                <p> <b>Instructions:</b> </p>
                <ul>
                    <li> Starter code and full assignment can be found <a href="https://github.com/visual-learning/transformers">here</a>. </li>
                    <li> Please submit your solutions on Gradescope.</li>
                    <li> Your hand-in should consist of two parts: 1) a single PDF report and 2) all your code in a compressed zip file. Please name the report "AndrewID.pdf" and the code "AndrewID.zip".</li>
                    <li> It is okay to ask Google for help, but it is not okay to plagiarize. Feel free to search “how to add cross entropy loss in pytorch,” but do <b>not</b> copy and paste the code from online repositories. </li>
                    <li> The solution will be graded by the final result as well as the correctness of the code. You will get partial credits even if the final result is not perfect.
                    <li> <b>Late Days:</b> Each student has 5 late days for the whole semester, tracked by Gradescope. Each additional late day will incur a 10% penalty on that homework.</li>
                    <li> <b>Start early!</b></li>
                </ul>

                <br>
                <p> <b> Acknowledgements:</b> This assignment is the result of the collective effort of all TAs that served this course over the last years. We thank all of them!</p>
            </div>
        </div>
    </body>
</html>