arrow_back अभ्यास
Phase 5 • Week 23

Heaps, LLD Structural Patterns, Reproducing GPT-2 + reading "Attention Is All You Need"

verified_user

Weekly Production Deliverable

Target Output
Deliverable: Heaps fully done. Structural Design Patterns done. You've read the Transformer paper after having built one yourself. LangChain + a vector database are set up.
Weeks: W21 W22 W23 Active W24 W25

Monday

0 / 2 done
DSA Heaps (Theory Video); Heapify Algorithm
1.5h
AI/ML Research Paper: "Attention Is All You Need"
Read the foundational paper: "Attention Is All You Need" (Vaswani et al., arxiv.org/abs/1706.03762) — read cold, mapping equations directly to the nanoGPT code you built in Week 22
2.5h

Tuesday

0 / 2 done
DSA Build heap from a given Array; Implement Min Heap
1.5h
CORE Structural Design Patterns 1
Adapter Pattern, Decorator Pattern, Facade Pattern + Real-world use cases, UML diagrams, Java code implementations + Quiz
2.5h

Wednesday

0 / 2 done
DSA Implement Max Heap; Check if an array represents a min heap
1.5h
AI/ML Karpathy: "Let's reproduce GPT-2"
Karpathy Zero to Hero: Architecture alignment (GPT-2 124M weights/biases, GELU activations), DataLoader at scale, distributed data parallel training intuition, mixed precision (fp16/bf16), learning rate warmup
2.5h

Thursday

0 / 2 done
DSA Convert Min Heap to Max Heap; Heap Sort
1.5h
CORE Structural Design Patterns 2
Composite Pattern, Proxy Pattern, Bridge Pattern, Flyweight Pattern + Java code implementations + Quiz
2.5h

Friday

0 / 2 done
DSA K-th Largest element in an array
1.5h
AI/ML Generative AI & LangChain Foundations
Krish Naik — "Complete LangChain Course for Generative AI" (YouTube): Text embeddings, Cosine Similarity, Vector Databases (FAISS, ChromaDB), Document Loaders, Recursive Character Text Splitters
2.5h
Weekend High-Load Execution • 8.0h Daily Deep Focus Lab

Saturday

0 / 3 done
DSA Kth largest element in a stream of running integers
1.5h
AI/ML Build GPT from Scratch (Karpathy "Let's build GPT") (4.0h)
Code nanoGPT transformer decoder blocks, layer normalization, residual stream, and generation loop in PyTorch.
4.0h
APT Quant M23 & Verbal/Logical
Quant M23 Geometry (Basic & Advance) + Verbal/Logical Mixed Practice
2.5h

Sunday

0 / 3 done
DSA Contest + full Heaps revision
1.5h
AI/ML Train nanoGPT on Shakespeare (4.0h)
Train your scratch GPT model; implement temperature and top-k sampling; generate coherent text; push project to GitHub.
4.0h
APT Structural Patterns Mastery & RAG Intro
Implement 2 Structural Patterns in Java + build your first local LangChain RetrievalQA chain over a PDF document
2.5h
edit_note

Week 23 Technical Notes & Journal

#LeetCode #Java #DSA #Week23