Week 2

Hey, late again, won’t change the published date, so that I can come into the habit of being punctual!

1) Univ Courses Update:

I am currently doing Stanford CS224n: Deep Learning for NLP course

Lectures completed:

1. Pre-training (Lec-9)

2. Post-training (Lec-10)

3. Natural Language Generation (Lec-11)

4. Benchmarking (Lec-11) ## Some ambiguity in the numbering (mentioning the one’s in the playlist)

5. Brain Computer Interfaces (Lec-13)

6. Reasoning & Agents (Lec-14)

7. Hugging Face Tutorial ## This was a rather tedious one, wrote down all the code

Assignments:

Completed the Math part of Assignment 4!

The assignments of the course seem pretty challenging, especially the coding ones.

Learning in a Jupyter notebook was pretty easy, handling the codebase seems a bit tricky as of now. Will hopefully get used to it after a while! 🙂

2) Bootcamps/ Code along courses:

The repo became too big, so wasn’t able to push to GitHub 🥲 with my amateur skills as of now!

Managed to complete:

4. Implementing a GPT model from Scratch To Generate Text

5. Pretraining on Unlabeled Data

6. Finetuning for Classification

7. Instruction Finetuning

Even when I coded along, I don’t think that I got all out of the amazing course, will redo it from the Book, more thoroughly this time!

3) Key Insights someone in the beginning may benefit from:

I came to know of a book Dive into Deep Learning, at a glance it does seem pretty nice. Will do it in the near future hopefully.

4) ?Rain?

Any particular reason the drainage system of Delhi (at least as I witnessed it) is so poor?

Tried to search a bit, most common reasons stated seem logical, maybe something on which someone will build a startup (or already has) about! 🤠

Making a Google doc for the ?—-? part