PROJECT 01
YOU ARE HERERandom Forest from Scratch
Build a complete Random Forest classifier from impurity math to a bagged ensemble with bootstrapping and feature subsampling.
MediumNumPy1.3h
Start anywhere →LVL 01SK
YOUR MACHINE LEARNING ODYSSEY
Master Machine Learning.
One Project at a Time.
A project-first path from classical algorithms to production AI systems. Build, learn, and leave a trail worth sharing.
CURRENT LEVEL1
TOTAL XP0
DAY STREAK0 days
GLOBAL RANK#12,847
THE PATH AHEAD
Your Learning Journey
Complete each project to unlock the next challenge.
0 of 26 completed
0%
PROJECT 02
Support Vector Machine from Scratch
Build a linear SVM using hinge loss, a regularized objective, and gradient-based optimization.
MediumNumPy0.9h
Start anywhere →PROJECT 03
Build a Trainable CNN from Scratch
Assemble a LeNet-style convolutional network with im2col convolutions, gradients, Adam, and a complete training loop.
HardNumPy4.9h
Start anywhere →PROJECT 04
Transformer from Scratch
Reimplement the original encoder-decoder Transformer with multi-head attention, scheduling, and beam search.
HardPyTorch6.6h
Start anywhere →PROJECT 05
Tiny GPT From Scratch
Grow a character-level language model from a bigram baseline into a working GPT with multi-head attention and sampling.
HardNumPy · PyTorch13.8h
Start anywhere →PROJECT 06
Q-Learning on FrozenLake
Train a tabular Q-learning agent with epsilon-greedy exploration and greedy evaluation.
EasyNumPy1.3h
Start anywhere →PROJECT 07
AlphaZero on Connect-4
Build the game engine, policy-value network, PUCT MCTS, self-play generation, training, and baseline evaluation.
HardPyTorch4.8h
Start anywhere →PROJECT 08
RAG Pipeline
Construct ingestion, chunking, embeddings, hybrid retrieval, grounded generation, evaluation, and conversational memory.
HardPyTorch4.3h
Start anywhere →PROJECT 09
DPO from Scratch
Implement log-prob utilities, Bradley–Terry preferences, DPO gradients, IPO variants, and a full evaluation pipeline.
HardNumPy2.3h
Start anywhere →PROJECT 10
RLHF from Scratch
Build decoding, SFT, LoRA, reward modeling, PPO, preference optimization, evaluation, and a model comparison interface.
HardPyTorch5.4h
Start anywhere →PROJECT 11
Vision-Language Model
Build a ViT encoder, multimodal projector, causal decoder, training loop, and caption generation from raw tensor operations.
HardPyTorch5.2h
Start anywhere →PROJECT 12
Flash Attention in CUDA
Build a tiled, IO-aware Flash Attention kernel with online softmax and causal masking.
HardCUDA2.2h
Start anywhere →PROJECT 13
Fused LLM Inference Kernels
Implement reductions, activations, fused RMSNorm, Softmax, RoPE, and SwiGLU kernels for efficient inference.
HardCUDA1.7h
Start anywhere →PROJECT 14
Mini LLM Inference Server
Construct sampling, tokenization, KV caching, paged allocation, continuous batching, streaming, and benchmarking.
HardPython4.3h
Start anywhere →PROJECT 15
VeriLLM
Wrap cached GPT inference in Merkle commitments and spot checks, then simulate a decentralized committee with rewards and slashing.
HardPyTorch4.8h
Start anywhere →PROJECT 16
MLP in JAX from Scratch
Implement initialization, forward passes, loss, autodiff, and pure functional SGD updates in JAX.
EasyJAX1.8h
Start anywhere →PROJECT 17
Build Your Own teenygrad
Construct a lazy buffer, reverse-mode autodiff engine, tensor API, neural primitives, and train a small MLP.
HardNumPy4.8h
Start anywhere →PROJECT 18
Federated Averaging
Implement client partitioning, local SGD, weighted aggregation, partial participation, and non-IID experiments.
MediumPyTorch2.2h
Start anywhere →PROJECT 19
DiLoCo Distributed Training
Train workers locally, aggregate pseudo-gradients with an outer optimizer, and quantify communication savings.
HardPyTorch2.5h
Start anywhere →PROJECT 20
Memory-Constrained Trainer
Implement accumulation, checkpointing, mixed precision, all-reduce, and ZeRO-style optimizer sharding.
HardNumPy3.3h
Start anywhere →PROJECT 21
Trainable MoE in CUDA
Build CUDA kernels for a complete sparse mixture-of-experts forward, backward, and training loop.
HardCUDA4.3h
Start anywhere →PROJECT 22
Multimodal Image Generator
Train a VQ-VAE image tokenizer and an autoregressive text-conditioned transformer with guided sampling.
HardJAX5.3h
Start anywhere →PROJECT 23
LoRA Fine-Tune a Chat Model
Load a 4-bit Qwen model, attach LoRA adapters, format instructions, run SFT, and generate with the tuned model.
EasyUnsloth1.7h
Start anywhere →PROJECT 24
A/B Testing & Causal Inference
Build proportion tests, sample sizing, multiple-testing corrections, difference-in-differences, and synthetic control.
MediumPython1.8h
Start anywhere →PROJECT 25
Market-Making Simulator
Build expected-value games and a quoting engine that manages inventory, adverse selection, and P&L.
MediumPython1.2h
Start anywhere →PROJECT 26
Tic-Tac-Toe: Minimax to DQN
Build the game engine, minimax, tabular self-play, and a DQN agent, then compare value and policy learners.
HardPyTorch7.7h
Start anywhere →Mastery awaitsComplete the journey
LEARNING SCORE0
Top 18% of learners
0%
TIME REMAINING101 hours
At your current pace
NEXT MILESTONE5 Projects
5 projects to go