2025 05 28 Week 3 Session 3 - TDS May 2025#

2025 05 28 Week 3 Session 3 - TDS May 2025

Duration: 5440.0Here’s an FAQ based on the live tutorial transcript:

Q1: I’m repeatedly getting “Open API key expired” or similar errors, even after trying both AI Pipe and AI Proxy methods. It worked the first time, but not anymore.

A1: This issue was covered in detail during the Week 3, Session 1 live tutorial on May 24th. You can find the recording on YouTube, and it includes specific timestamps. The key is to ensure you’ve correctly set up your API tokens in the environment settings and replaced the URL and token accordingly. If you’re still facing issues after reviewing the tutorial and trying the suggested steps, please raise a specific question in our Discord channel or during a live session, detailing the exact error message. While it’s frustrating, troubleshooting these steps is a crucial part of building your coding background.

Q2: What’s the difference between AI Pipe and AI Proxy, and how do I know which one to use?

A2: Both AI Pipe and AI Proxy serve the same fundamental purpose: to interact with OpenAI models. The main differences lie in the specific models they support and their deployment methods. AI Proxy was used in previous terms. Essentially, instead of directly using an OpenAI API token, these methods act as intermediaries. You can use either; the choice often depends on the specific model version you’re working with. Both serve to bypass direct interaction with OpenAI.

Q3: My GA2 Google Authentication question worked once, but now it’s showing errors, specifically a “key error.” I’ve tried recreating it multiple times.

A3: This common troubleshooting scenario for GA2’s Google Authentication, including handling API keys and proxies, was thoroughly explained in the Week 3, Session 1 live tutorial on May 24th. Please refer to that session for detailed guidance. If the issue persists, it could be related to specific environment settings or needing multiple services running concurrently (as some students have experienced). It’s crucial to identify the exact error message.

Q4: When will Project 1 be discussed, and what are the mandatory steps to prepare for it?

A4: Project 1 details and preparation steps will be covered in dedicated “breakthrough sessions” starting this weekend, following the completion of GA3. These sessions will guide you through the process, including deploying an application with a public URL. There are no specific restrictions on the deployment tools you use (e.g., Olama, online GPT, API, AWS Mechanical Turk). The timeline for Project 1 will be clearly laid out in those sessions.

Q5: The TDSC course only lists MLF as a prerequisite, but I feel it requires a strong coding background. What are the true expectations, and how will I manage if I’m not a coding expert?

A5: The GA1 assessment is designed to help you self-evaluate your readiness for the course. While TDSC is coding-heavy, the prerequisite doesn’t demand you be a coding expert beforehand. The course is structured to help you develop these skills. The expectation is that you can follow along, use code effectively, and actively engage with the learning resources. We provide extensive learning sessions throughout the course, and you’re strongly encouraged to participate and ask questions. Don’t hesitate to use Discord for specific problems.

Q6: In GA3, Question 3, the submission checkbox is missing.

A6: This can often be resolved by using your browser’s developer tools (e.g., Chrome Dev Tools). This technique was covered in a previous session, specifically on “Development Tools Browser Dev Tools CSS Selectors.” You’ll need to explore how to enable or interact with the missing element using these tools.

Q7: I struggled with the GA2 image compression question (reducing to 400KB). I managed it with external tools but couldn’t get a precise result using Python libraries. Is there a specific method for this?

A7: There isn’t just one “proper” way to handle image compression; multiple methods and libraries exist (e.g., Pillow, Scikit-image). This is a common challenge, and sometimes achieving the exact size can be tricky. We will cover this in upcoming learning sessions. In the meantime, feel free to raise specific questions with your code in Discord or during a live session if you’re stuck on a particular library or approach. It’s a prompt engineering question, not purely coding, and often requires creative problem-solving.

Q8: What is the deadline for Project 1?

A8: The current deadline for Project 1 is June 14th (Saturday).

Q9: Will there be any breaks for the Paradox course, or will it run continuously?

A9: At this point, there are no planned breaks. The course structure indicates continuous learning. The initial weeks are designed to be heavily loaded with new concepts and tools. Later sections, like data sourcing and analysis, will likely feel lighter as you become more familiar with the tools and apply existing concepts.

Q10: I find the course content overwhelming, especially with many tools introduced in a short period. I’m struggling with basic concepts like installation and commands. How can I get more in-depth support, particularly for complex topics like Ngrok?

A10: We understand that learning many tools and concepts can be challenging, especially for those new to coding. The course is designed with short, focused videos to maintain engagement, and longer, in-depth discussions happen in live learning sessions.

  • Support options:
    • Live Learning Sessions: These are dedicated times to address specific questions and troubleshoot problems. You’re encouraged to raise any questions you have, even seemingly basic ones.
    • Discord: This is a vital platform for ongoing support. Post your questions, code snippets, and screenshots. Our TAs and instructors are active there and will help you.
    • YouTube Playlists: All live sessions are recorded and available with timed chapters, allowing you to jump to specific topics.
    • Collaboration: Working with peers can also be very beneficial.
  • Course design: The initial weeks are indeed tool-heavy to build a foundational skill set. Later sections focus more on applying these tools. The goal is not to make you an expert in every tool, but to enable you to understand their purpose, boundary conditions, and how to use them effectively for data science deliverables.
  • Regarding Ngrok: This specific topic, along with API keys and proxies, was covered in Week 3, Session 1. If you’re still struggling, raise specific questions in Discord. It’s about understanding the logic and being creative, rather than just memorizing code.
  • Starting late/catching up: It’s never too late. You can go back to previous weeks’ materials, watch recordings, and use Discord to get caught up. Tools learned early on (like Ngrok) are essential and will be used throughout the course.
  • Bottom line: Don’t hesitate to ask questions in all available forums. The more specific your questions, the better we can assist you. Your feedback is valuable for continuous course improvement.

Q11: I’m thinking of buying a new laptop (MacBook with M4 chip). Will it be compatible for the whole BS degree, or should I go with Windows?

A11: You don’t need to purchase a new laptop specifically for this course. TDSC is designed to be compatible with various environments. Whether you have Windows (using WSL), Linux (direct Bash), or Mac (using Homebrew), the tools and concepts will work. There are no restrictions on the specific platform. If you prefer a Mac, you can certainly go with that. Any modern system should support the course requirements.

Q12: My GA2 Question 7 submission didn’t save, and I had to re-submit it. Why did this happen? Also, the output from Ngrok might not be consistent. How do I handle this?

A12: This is an unusual situation as submissions are typically saved in our system once confirmed. We haven’t observed frequent instances of saved answers disappearing. Please raise this specific issue in Discord, detailing what you did and any error messages encountered. This will help us investigate if there’s a technical glitch or an issue with the submission process. We also understand that Ngrok’s output can be inconsistent due to its nature. If you’re consistently getting different results or errors, raise that specific problem in Discord or during a live session. We encourage you to save your work frequently.

Q13: I’m new to coding, and finding the course overwhelming with too many tools and concepts. I also find the videos hard to follow. How can I get more in-depth support, especially for topics like Ngrok?

A13: We understand your concerns. The course is designed with short, focused videos to enhance engagement, and longer, in-depth discussions happen in live learning sessions.

  • Support: Raise all your questions in Discord, use the YouTube playlists with timed chapters, and actively participate in live sessions.
  • Course Design: The initial weeks are tool-heavy by design to build a foundational skill set. Later sections focus more on applying these tools. The goal isn’t to make you an expert in every tool, but to enable you to understand their purpose and how to use them effectively.
  • Ngrok: This specific topic, along with API keys and proxies, was covered in Week 3, Session 1. If you’re still struggling, raise specific questions in Discord. It’s about understanding the logic and being creative, rather than just memorizing code.
  • Timeliness: It’s never too late to learn. If you’re struggling with earlier topics like Ngrok, you can go back to those videos/sessions. The tools learned are essential and will be used throughout the course. Don’t worry about being “behind.”
  • Asking Questions: We encourage you to ask any question, even if it seems basic. The more specific your questions, the better we can assist you. You can use LLMs like ChatGPT to help frame your questions too.

Q14: What is the purpose of the TDSC course? After completing this diploma and having learned all these tools, where would I apply them, and what kind of roles would I be suitable for? I’m struggling to understand the direction and where this will take me.

A14: The TDSC course aims to equip you with the skills to develop and deploy data science applications, rather than just learn individual tools in isolation.

  • Application: You’ll learn to build applications like an LLM (Large Language Model) application that can answer user questions based on pre-loaded content (similar to Project 1). You can deploy these applications publicly.
  • Resume/Job Search: A strong open-source contribution (e.g., building and deploying your own application on GitHub) is highly valued by companies. You can showcase projects built in this course (like the LLM app) on your resume, LinkedIn, and personal website. This demonstrates your ability to not just code, but also deploy and make applications accessible.
  • Professional Environment: The course helps you prepare for a professional environment by teaching collaboration, understanding deliverables, and planning your work. It emphasizes problem-solving processes, like how to approach a task where you don’t know everything, requiring collaboration with other teams/departments and effective time management.
  • Learning Process: The course encourages a proactive, gradual learning approach. You’ll learn by doing, incrementally building your understanding and skills. The goal is to develop the ability to learn and adapt to new challenges, which is crucial in the ever-evolving field of data science.

Q15: When I review the course content, particularly for topics like Ngrok or other tools, I find the explanations generic. I feel like the instructors sometimes go off-topic. Also, the videos are very short. Could we get more in-depth, faculty-specific playlists for TDSC?

A15: We appreciate your feedback.

  • Video Format: The decision for shorter, crisp videos is intentional, aligning with modern learning trends (like short-form social media content) to improve engagement and allow you to quickly grasp specific concepts. Longer sessions have been tried, but shorter ones are generally more effective for focused learning.
  • Content Depth: The course is designed for you to learn by doing and applying concepts. We focus on providing sufficient information to understand a tool’s purpose, its boundary conditions, and how to use it for specific data science deliverables. We don’t aim for mastery of every tool, but rather the ability to effectively utilize them.
  • Support for In-Depth Learning: If you want to go deeper into any topic, we encourage you to:
    1. Ask Specific Questions: Use Discord and live learning sessions. The more precise your question, the better we can assist you.
    2. Utilize All Resources: This includes course materials, YouTube playlists (which are regularly updated with timed chapters), and collaborative learning.
    3. Leverage LLMs: Tools like ChatGPT can be valuable for independent learning and getting quick answers to specific queries.
  • Instructor Focus: The entire team is dedicated to supporting your learning. Any specific questions about instructor style or content delivery can be shared, and we continuously strive to improve.

Q16: Is it too late to start working on GA2 topics like Ngrok? I missed some deadlines. Also, the content feels shallow, and I struggle to understand. What should I do?

A16: It is never too late to start or catch up!

  • Continuous Learning: The tools learned in early GAs (like Ngrok) are foundational and will be used throughout the course. There’s no “closed chapter” for learning.
  • Time & Support: You still have time. Go back to the videos, revisit the live session recordings, and use Discord. If you miss a deadline, focus on learning the material anyway. The priority is understanding.
  • Ask Questions: If you don’t understand, ask! Don’t hesitate to post basic questions in Discord or raise them in live sessions. There’s no such thing as a “dumb question” when you’re trying to learn.
  • Effective Workflow: If you try a solution and it doesn’t work, ensure you save your attempts, note the error, and then raise the question. This structured approach helps us pinpoint the problem.
  • Deadlines: Project 1 has a deadline of June 14th. Ensure you are aware of that.
  • Use All Platforms: Utilize Discord, live learning sessions, and YouTube playlists to your full advantage. Collaboration with peers is also highly encouraged.

Q17: What is the deadline for Project 1?

A17: The current deadline for Project 1 is June 14th.