From 7369a6a550649056b5548bae937712b899f11743 Mon Sep 17 00:00:00 2001 From: Pavlos Protopapas Date: Mon, 14 Aug 2023 17:01:23 -0400 Subject: [PATCH] added project policy description --- .DS_Store | Bin 0 -> 6148 bytes README.md | 83 ++++++++++++++++++++++++++++++++++++------------------ 2 files changed, 56 insertions(+), 27 deletions(-) create mode 100644 .DS_Store diff --git a/.DS_Store b/.DS_Store new file mode 100644 index 0000000000000000000000000000000000000000..fde6fd1fa0015f2dd3f366ebbc465f58f9195933 GIT binary patch literal 6148 zcmeHKOHRWu5S@Wi1tFm>S^5gxz^cj#H~^m}3naH$q+N99qx0qiYC_hipqa{kw&&yV zlWJ#(h}XyVk!VFk4Voa!U`AwlR(0jZOCW2Eo{qGqEp@b;33Q1|lKVZq&>QmJ(|P_K z<_8+uVO!q&zDV-7F#OO1796z2F-PhK>VilojWh8T&K<&fl1V>L8)VlO_}cm66~Sa(PKq~^rgF#2F17#K5f zZPS&!|4;bK%oh1$mzV_u!N5ObfESzR%^HXDxAn*IHqHIyi_uI|A25Kuy*3kH6Hflv16HUj_v literal 0 HcmV?d00001 diff --git a/README.md b/README.md index 470680e..20fb870 100644 --- a/README.md +++ b/README.md @@ -31,31 +31,31 @@ We have designed an in-depth curriculum to ensure a comprehensive understanding 1. **Introduction:** - Begin with an understanding of the importance of MLOps and how it fits in the broader AI and software development ecosystem. -​ + ​ 2. **Virtual Environments and Virtual Machines:** - Delve into the foundations of isolated software environments, their importance in AI development, and how virtual machines offer a layer of abstraction over physical hardware. -​ + ​ 3. **Containers:** - Understand the concept of containerization using tools like Docker, and how they differ from virtual machines. -​ + ​ 4. **Cloud Functions:** - Explore serverless architectures, focusing on platforms like GCP Cloud functions (AWS Lambda's equivalent), ensuring efficient model deployment without managing server infrastructures. -​ + ​ 5. **Data Pipelines, Dask, & Cloud Storage:** - Learn to create efficient data workflows, use Dask for parallel computing, and understand how cloud storage solutions fit into the MLOps ecosystem. -​ + ​ 6. **TF Data and TF Records:** - Dive into TensorFlow-specific methods for data ingestion and management, ensuring efficient data preprocessing and storage for your models. -​ + ​ 7. **Data Parallelization:** - Grasp techniques for distributing data processing tasks across multiple processors or nodes. -​ + ​ 8. **Data Versioning:** - Explore tools like Pachyderm, and understand the significance of maintaining different versions of datasets for reproducibility and model training. -​ + ​ 9. **Advanced Training Workflows:** - Deep dive into experiment tracking using tools like Weights & Biases, and harness the power of multi-GPU setups for faster model training. -​ + ​ 10. **Advanced Inference Workflows:** - Understand the nuances of model optimization techniques like Distillation, Quantization, and Compression. Explore TensorFlow Lite, monitor your models post-deployment, and be prepared for challenges like data drift. ​ @@ -82,29 +82,29 @@ In today's AI-driven world, building a robust deep learning model is only half t ​ ## Prerequisites -​ + To ensure a seamless learning experience and to make the most of this course, participants are expected to come with a foundational knowledge in the following areas: ​ 1. **Programming Proficiency in Python:** - A strong command over Python's basic constructs, including functions, classes, and modules. Familiarity with libraries like NumPy, Pandas, Matplotlib is essential, as they form the backbone of many data manipulation tasks in AI. -​ + ​ 2. **Deep Learning Framework - Tensorflow:** - A working knowledge of the TensorFlow (or PyTorch) framework is crucial, as many topics will delve into its functionalities and methods. Understanding TensorFlow's basic operations, data handling, and model building mechanisms will be invaluable. -​ + ​ 3. **Basic Shell Commands:** - Comfortability in navigating the command-line interface (CLI), executing shell commands, and performing basic file operations are foundational for many MLOps tasks. -​ + ​ 4. **Basic Data Structures:** - A good grasp of Python's primary data structures, especially dictionaries and lists, will be instrumental in understanding and manipulating data. -​ + ​ 5. **File I/O:** - Knowledge of basic file input/output operations in Python, including reading from and writing to files, is vital for tasks involving data storage and manipulation. -​ + ​ 6. **General AI and ML Concepts:** - While this course is centered around MLOps, a basic understanding of AI and machine learning concepts, including what models are and how they are trained, will set the context for many advanced topics. -​ -It's important to note that while prior knowledge in these areas will provide a solid foundation, the course has been structured to ensure gradual progression. Even if you're not an expert in all of the prerequisites, a willingness to learn and engage actively in the course's hands-on components will be crucial for success. If you find yourself struggling with some concepts, we encourage leveraging the course resources, attending office hours, and participating in peer discussions to reinforce your understanding. -​ + ​ + It's important to note that while prior knowledge in these areas will provide a solid foundation, the course has been structured to ensure gradual progression. Even if you're not an expert in all of the prerequisites, a willingness to learn and engage actively in the course's hands-on components will be crucial for success. If you find yourself struggling with some concepts, we encourage leveraging the course resources, attending office hours, and participating in peer discussions to reinforce your understanding. + ​ ## Lecture @@ -115,7 +115,7 @@ Date Time ## Course Components -​ + - **Weekly Sessions:** Structured lectures focusing on the core topics. - **Office Hours:** Dedicated time with the instructors and TAs for queries and clarifications. @@ -129,8 +129,37 @@ Date Time - **Team Projects:** Collaborative assignments that culminate in the creation of a fully functional AI app. - **Discussion Forums:** Platforms for peer-to-peer learning, discussions, and knowledge sharing. -​ -Remember, these components might undergo changes to ensure the best learning experience, so always stay updated with the latest course schedule and announcements. + ​ + Remember, these components might undergo changes to ensure the best learning experience, so always stay updated with the latest course schedule and announcements. + +**Team Projects: Project-Based Learning: Crafting Your Own AI Solutions** + +In the dynamic realm of AI and MLOps, hands-on experience is paramount. This course encourages each student to bring a unique perspective by working on self-conceived projects. Here's what you need to know: + +**1. Crafting Your AI Project:** +- Students are expected to conceptualize and develop their own projects. While our teaching staff is here to provide ideas and guidance, the core objective is for each student to nurture and shape their original initiative. + +- By the end of the semester, the aim is to transform your idea into a fully functional web-app or mobile application. + +- Project Scope: Your project should incorporate some element of modeling, ensuring it aligns with the learning objectives of the course. Moreover, it is essential that every component of the project CAN be evaluable by our teaching staff. + +- **Unleash Your Creativity:** Whether you're driven by a start-up vision, by research lab innovations, or inspired by a personal hobby, this is your platform to bring that idea to life. + +**2. A Guided Demonstration by Pavlos:** +- We, the teaching team,, will undertake a project that Pavlos proposes throughout the semester. This serves as a demonstration and reference point. + +- Each week will spotlight a different facet of Pavlos' project development. This structured showcase offers students a practical insight of course concepts. + +- Parallelly, students will be prompted to integrate the week's learnings into their projects, ensuring a steady progression towards their end goals. + +**3. Milestones and Assessment:** +- The course will be punctuated with key milestones, designed to assess your project's evolution and your grasp of the MLOps concepts. Details of these milestones will be shared in due course. + +- It's imperative to understand that a significant portion of your grade hinges on these milestones. They are not just checkpoints but pivotal phases that contribute to your project's holistic development and your learning journey. + +**In Summation:** + +The heart of this course is experiential learning. We fervently believe that your ideas and paralleling them with structured guidance, we can equip you with the tangible skills essential in today's AI-driven world. ## Course Policies @@ -140,23 +169,23 @@ Remember, these components might undergo changes to ensure the best learning exp - **ED Forum:** For questions related to exercises, course content, or package installations, post on the ED forum first. This promotes peer learning, as your peers might have faced similar issues. All posts are visible to everyone, and the teaching staff monitors them regularly. - **Office Hours:** Should you need more personalized assistance or deeper clarifications, attending office hours is recommended. - **Email:** For private matters or specific concerns, feel free to directly email the instructor. -​ + ​ 2. **Academic Honesty:** - This course places a strong emphasis on ethical behavior. Whether it's ethically handling data or attributing the work of others, students are expected to maintain high standards of integrity. - **Acceptable Behaviors:** Discussing course materials, engaging in office hours, debugging with peers, using and citing small portions of code found online, seeking online knowledge, and seeking guidance from tutors. - **Unacceptable Behaviors:** Accessing or sharing solutions before submission, plagiarizing, not citing sources of external code or techniques, paying or offering payment for coursework, and sharing course material with future potential students. - Engaging in unacceptable behaviors will lead to disciplinary action. When in doubt, always consult the course instructors. -​ + ​ 3. **Deadlines:** - Adhering to deadlines is crucial for a consistent learning experience. All quizzes and exercises are due before the day of the next session. No extensions will be provided, so plan your time effectively. -​ + ​ 4. **Collaboration & Teamwork:** - Collaboration is encouraged, especially for projects. However, ensure you contribute equally and do not divide tasks in a way that prevents you from understanding all parts of the assignment. 5. **Feedback & Evaluation:** - Continuous feedback is vital for the learning process. While the course has several grading components, always focus on understanding rather than just marks. Do provide feedback on the course structure, content, and delivery, so we can continually improve. -​ + ​ 6. **Accessibility & Inclusion:** - We strive to make this course accessible to all. If you have any special needs or require specific accommodations, please reach out to the course administrators at the earliest. -​ -Students are encouraged to thoroughly understand and abide by these policies. The guiding principle should always be fairness, respect for one's own learning journey, and respect for the learning journey of peers. \ No newline at end of file + ​ + Students are encouraged to thoroughly understand and abide by these policies. The guiding principle should always be fairness, respect for one's own learning journey, and respect for the learning journey of peers. \ No newline at end of file