mlip-cmu · sayalikandarkar · Nov 22, 2023 · Nov 22, 2023 · Nov 22, 2023 · Nov 22, 2023
diff --git a/.github/workflows/publish.yml b/.github/workflows/publish.yml
@@ -22,7 +22,7 @@ jobs:
     - name: Generate main page
       run: |
         cp .assets/template.html _site/index.html
-        sed 's/\](https:\/\/mlip-cmu.github.io\/s2023\/slides\/\([^\(]*\)\.html) (\[/\](https:\/\/mlip-cmu.github.io\/s2023\/slides\/\1.html) ([md](https:\/\/github.com\/mlip-cmu\/s2023\/blob\/main\/lectures\/\1.md), [pdf](https:\/\/mlip-cmu.github.io\/s2023\/slides\/\1.pdf), \[/' < schedule.md |         sed 's/\](https:\/\/mlip-cmu.github.io\/s2023\/slides\/\([^\(]*\)\.html) *|/\](https:\/\/mlip-cmu.github.io\/s2023\/slides\/\1.html) ([md](https:\/\/github.com\/mlip-cmu\/s2023\/blob\/main\/lectures\/\1.md), [pdf](https:\/\/mlip-cmu.github.io\/s2023\/slides\/\1.pdf)) |/'  > schedule_.md
+        sed 's/\](https:\/\/mlip-cmu.github.io\/s2024\/slides\/\([^\(]*\)\.html) (\[/\](https:\/\/mlip-cmu.github.io\/s2024\/slides\/\1.html) ([md](https:\/\/github.com\/mlip-cmu\/s2024\/blob\/main\/lectures\/\1.md), [pdf](https:\/\/mlip-cmu.github.io\/s2024\/slides\/\1.pdf), \[/' < schedule.md |         sed 's/\](https:\/\/mlip-cmu.github.io\/s2024\/slides\/\([^\(]*\)\.html) *|/\](https:\/\/mlip-cmu.github.io\/s2024\/slides\/\1.html) ([md](https:\/\/github.com\/mlip-cmu\/s2024\/blob\/main\/lectures\/\1.md), [pdf](https:\/\/mlip-cmu.github.io\/s2024\/slides\/\1.pdf)) |/'  > schedule_.md
         sed -i -e '/^\[Schedule\]/r schedule_.md' README.md
         npx marked -i README.md >> _site/index.html
         cat .assets/template_end.html >> _site/index.html

diff --git a/README.md b/README.md
diff --git a/assignments/I1_mlproduct.md b/assignments/I1_mlproduct.md
@@ -49,8 +49,8 @@ Commit all your code changes to your GitHub repository, but *do not commit priva
 
 Additionally upload a short report to Gradescope by [date see Canvas] with the following content: 
 
-* **GitHub link:** Start the document with a link to your last commit on GitHub: On the GitHub webpage, click on the last commit message and copy the URL in the format `https://github.com/[user]/[repo]/commit/[commitid]`. Make sure that the link includes the long ID of the last commit.
-* **Technical description (1 page max):** Briefly describe how you implemented the two features. Provide pointers to the relevant parts of the code, ideally as direct links to files or even to specific lines on GitHub. We prefer readable links in the PDF rather than hyperlinks behind text (e.g., https://github.com/ckaestne/mlip-s23/blob/main/assignments/I1_mlproduct.md rather than [this](https://github.com/ckaestne/mlip-s23/blob/main/assignments/I1_mlproduct.md)).
+* **GitHub link:** Start the document with a link to your last commit on GitHub: On the GitHub webpage, click on the last commit message and copy the URL in the format `https://github.com/cmu-seai/[repo]/commit/[commitid]`. Make sure that the link includes the long ID of the last commit.
+* **Technical description (1 page max):** Briefly describe how you implemented the two features. Provide pointers to the relevant parts of the code, ideally as direct links to files or even to specific lines on GitHub. We prefer readable links in the PDF rather than hyperlinks behind text (e.g., https://github.com/mlip-cmu/s2024/blob/main/assignments/I1_mlproduct.md rather than [this](https://github.com/mlip-cmu/s2024/blob/main/assignments/I1_mlproduct.md)).
 * **User interface design approach (1 page max):** Recommend for each of the two features how the feature should interact with users (automate, prompt, organize, annotate, hybrid) and why. Justify your recommendation, considering forcefulness, frequency, value, and cost. If your implementation differs from the recommended approach, briefly explain how you would change your implementation if you had more time.
 * **Harms (1 page max):** Discuss what possible harms you can anticipate from using machine learning for the features in the applications (e.g., safety, fairness). Identify at least one harm and discuss potential solutions to mitigate the harm. (You do not need to implement the solutions.)
 * **Production challenges (1 page max):** Discuss any technical challenges you anticipate if you want to deploy this feature in production (e.g., scalability, operating costs) and how you would change your implementation if you expected millions of users. Identify at least one problem and discuss corresponding potential solutions. (You do not need to implement the solutions.)

diff --git a/assignments/I3_architecture.md b/assignments/I3_architecture.md
diff --git a/assignments/I4_mlops_tools.md → assignments/I3_mlops_tools.md b/assignments/I4_mlops_tools.md → assignments/I3_mlops_tools.md
diff --git a/assignments/I4_explainability.md b/assignments/I4_explainability.md
@@ -0,0 +1 @@
+tbd
diff --git a/assignments/research_project.md b/assignments/research_project.md
@@ -46,13 +46,13 @@ If you plan to conduct interviews or surveys as part of the project and you plan
 
 ## Deliverables
 
-Submit a draft at the milestone deadlines and a paper and a presentation at the final deadline.
+Email a draft to the instructors at the milestone deadlines and submit the paper to Gradescope for the final deadline. Present the work during the final presentation slot of the class.
 
 The paper should be in a form submittable to a new-idea track, short-paper track, or workshop in the field. It should have at least an introduction motivating the research, one or more clear and motivated research questions, a discussion of the state of the art or related work, and a description of the conducted or planned research, and some results. While we do not enforce a specific page limit or formatting requirements, we would typically expect around 4 pages double-column format, such as for the [ICSE-NIER](https://conf.researchr.org/track/icse-2022/icse-2022-nier---new-ideas-and-emerging-results) track.
 
 The presentation should be no longer than 8 minutes. How you structure the presentation is up to you. You do not need to cover everything, but consider how to make this interesting to the audience. It will be presented in the same time slot as the presentations from the group project.
 
-Send drafts, papers and slides as attachments or links per email to the instructors.
+Send drafts, papers and slides as attachments or links per email to the instructors. Submit the final paper to Gradescope.
 
 ## Grading
 

diff --git a/exams/README.md b/exams/README.md
@@ -10,12 +10,13 @@ Topic-wise, everything covered in class, in the readings, and in recitation is f
 
 Midterms from previous semesters are available as practice. We expect the midterm to have a similar format, though topic coverage differs slightly between semesters.
 
-* [Practice midterm from Fall 2019](https://github.com/ckaestne/seai/blob/F2019/other_material/practice_midterm.pdf) (corresponds quite well to topics covered this semester)
-* [Midterm from Fall 2019](https://github.com/ckaestne/seai/blob/F2019/other_material/midterm.pdf) (we did not cover version control yet)
-* [Final from Fall 2019](https://github.com/ckaestne/seai/blob/F2019/other_material/final_exam.pdf) (covers different topics, but provides yet another scenario)
-* [Midterm from Summer 2020](https://github.com/ckaestne/seai/blob/S2020/exams/midterm.pdf) (slightly fewer topics covered than this semester)
-* [Midterm from Fall 2020](https://github.com/ckaestne/seai/blob/F2020/exams/midterm_f20.pdf) (similar coverage)
-* [Midterm from Spring 2021](https://github.com/ckaestne/seai/blob/S2021/exams/) (similar coverage)
-* [Midterm from Spring 2022](https://github.com/ckaestne/seai/blob/S2022/exams/) (similar coverage, except data quality and infrastructure quality yet)
-* [Midterm from Fall 2022](https://github.com/ckaestne/seai/blob/F2022/exams/) (similar coverage)
+* [Practice midterm from Fall 2019](https://github.com/ckaestne/seai/blob/F2019/other_material/practice_midterm.pdf) 
+* [Midterm from Fall 2019](https://github.com/ckaestne/seai/blob/F2019/other_material/midterm.pdf) 
+* [Final from Fall 2019](https://github.com/ckaestne/seai/blob/F2019/other_material/final_exam.pdf) 
+* [Midterm from Summer 2020](https://github.com/ckaestne/seai/blob/S2020/exams/midterm.pdf) 
+* [Midterm from Fall 2020](https://github.com/ckaestne/seai/blob/F2020/exams/midterm_f20.pdf) 
+* [Midterm from Spring 2021](https://github.com/ckaestne/seai/blob/S2021/exams/) 
+* [Midterm from Spring 2022](https://github.com/ckaestne/seai/blob/S2022/exams/) 
+* [Midterm from Fall 2022](https://github.com/ckaestne/seai/blob/F2022/exams/) 
+* [Midterm from Spring 2023](https://github.com/mlip-cmu/s2023/tree/main/exams) 
 
diff --git a/labs/lab01.md b/labs/lab01.md
@@ -0,0 +1,46 @@
+# Lab 1: Calling, Building, and Securing APIs
+In homework I1 you will use third-party machine learning APIs and in the group project you will develop your own APIs. In this lab, you will experiment with both, connecting to the Azure Vision API and providing your own API endpoint. 
+To receive credit for this lab, show your work to the TA during recitation.
+
+## Deliverables
+- [ ] Create an account and connect to the Azure Vision API
+- [ ] Explain to the TA why hard-coding credentials is a bad idea. Commit your code to GitHub without committing your credentials.
+- [ ] Run the API endpoint with the starter code and demonstrate that it works with an example invocation (e.g., using curl).
+
+## Getting started
+Clone the starter code from this Git repository: https://github.com/eshetty/mlip-api-lab
+
+The code implements a flask web application that receives API requests to analyze an image and return information about the image, including the text contained within. To identify the text, the OCR feature of the Azure Vision API [[documentation](https://westcentralus.dev.cognitive.microsoft.com/docs/services/computer-vision-v3-2/operations/56f91f2e778daf14a499f20d#:~:text=test.jpg%22%7D-,Response%20200,-The%20OCR%20results), [response format](https://westcentralus.dev.cognitive.microsoft.com/docs/services/computer-vision-v3-2/operations/56f91f2e778daf14a499f20d#:~:text=test.jpg%22%7D-,Response%20200,-The%20OCR%20results)] can be used by adjusting the API endpoint and credentials in the code. We use the Azure’s provided libraries to abstract from low-level protocol details.
+
+Install the dependencies in the `requirements.txt` file with pip or similar. To set up the flask server, just run `python3 app.py`. The system should try to analyze an example image and report the results when you go to http://localhost:3000/
+
+## Connecting to the Azure Vision API
+1. Sign up for the a student account for Microsoft Azure: https://azure.microsoft.com/en-us/free/students/ – no credit card required
+
+2. Create an instance of the Computer Vision service and get an API endpoint of your instance of the service.
+
+3. Get a subscription key to authorize your script to call the Computer Vision API.
+
+4. Update the code with the endpoint and key and test it.
+
+## Secure your Credentials
+The starter code hardcodes credentials in the code. This is a bad practice. 
+
+Research and discuss best practices, such as never hard-code credentials, never commit credentials to Git, rotate secrets regularly, encrypt your secrets at rest/in-transit if possible, practice least-access privilege on machines where your credentials are stored as environment variables or within local files.
+
+Rewrite the code to load credentials from a file or an environment variable and commit the code without the credentials to GitHub.
+
+## Calling your own API
+The starter code comes with a flask server that serves the website at http://localhost:3000/ but also exposes an own API at http://localhost:3000/api/v1/analysis/ accepting a GET request with a JSON object with a single field “uri” pointing to an image to analyze.
+
+Identify how to call your own API with a tool like [curl](https://curl.se/docs/manpage.html) or [Postman](https://www.postman.com).
+
+Optionally extend the API or document it with [Swagger](https://swagger.io).
+
+## Additional resources 
+- [Redhat article on API](https://www.redhat.com/en/topics/api/what-are-application-programming-interfaces)
+- [Azure Computer Vision](https://learn.microsoft.com/en-us/python/api/overview/azure/cognitiveservices-vision-computervision-readme?view=azure-python)
+- [API Design Best Practices](https://blog.stoplight.io/crud-api-design?_ga=2.223919515.1813989671.1674077556-1488117179.1674077556)
+- [API Endpoint Best Practices](https://www.telerik.com/blogs/7-tips-building-good-web-api)
+- The file `seai-azure-cv-ocr-api.json` has the structure to test calls to the Azure Vision API with Postman.
+