- Make a Training Step Fast: Profile, Fix, Prove It
- Run a Small Research Study: Does the Learning Rate Transfer?
- Many Devices: Sharding and Placement for a Tensor Compiler
- Ship a Model: From a Trained PyTorch Network to a Monitored Service
- GPT From Your Own Layers: Tiny Shakespeare
- Train and Evaluate a Tool-Using Agent in PyTorch
- Reward Model and Rejection Sampling from Scratch
- makemore in PyTorch, Part 2: A WaveNet From Your Own Layers
- makemore in PyTorch: From Bigram Counts to an MLP Language Model
- Customer Churn, Start to Finish: A Data Science Capstone
- Sutton and Barto from Scratch 5: Mars Lander from Flight Logs
- Generative Classifiers from Scratch: GDA, Naive Bayes and Fisher LDA vs Logistic Regression
- Classical Computer Vision from Scratch: Filters, Edges, Corners and a Panorama
- Clustering from Scratch in NumPy: HAC, K-Means++, Medoids, GMMs and Spectral Cuts
- Information Theory from Scratch in NumPy: Entropy, KL and Mutual Information
- Choosing the Right Model from Scratch: Evidence, BIC, AIC, Cross-Validation and WAIC
- Train a Mixture-of-Experts LM and Watch It: Balance, Diagnostics, Upcycling and Serving
- Object Detection from Scratch: A Tiny One-Stage Detector
- Exemplar Methods from Scratch: kNN, KDE, Metric Learning and Open-Set Face Matching
- Gradient Boosted Trees from Scratch: AdaBoost, LogitBoost and XGBoost
- Numerical Linear Algebra from Scratch in NumPy
- Optimization from Scratch in NumPy: From Steepest Descent to L-BFGS, SVRG and Adam
- Pretrain a Tiny BERT, Then Fine-Tune Four Heads
- Unsupervised and Semi-Supervised Graph Embeddings: Label Propagation, GAE/VGAE, DGI and Poincare
- Dimensionality Reduction from Scratch in NumPy: PCA, Factor Analysis by EM and the Manifold Zoo
- Beyond One Gaussian: Heteroskedastic MLPs, Mixture Density Networks and MC Dropout from Scratch
- H-Net from Scratch: Dynamic Chunking on Raw Bytes
- Predicting Insurance Claims With GLMs from Scratch
- Learning with 40 Labels: Semi-Supervised and Active Learning from Scratch
- Gaussian Processes from Scratch: Kernels, Marginal Likelihood, Classification and Scaling
- Estimators from Scratch: MLE, MAP, Posteriors and the Bootstrap on One Coin
- Gaussian Inference from Scratch in NumPy: Conditioning, Imputation, Sensor Fusion and Mixtures
- Bayesian and Robust Logistic Regression from Scratch
- Sparse and Bayesian Linear Regression from Scratch
- Train a Neural Network in CUDA from Scratch
- Fix Softmax, Watch Grokking Happen
- GPT-2 to gpt-oss, One Change at a Time
- Why Selection Matters: Mamba's Synthetic Tasks in tinygrad
- Mechanistic Interpretability from Scratch: Find the Induction Circuit in a Tiny GPT
- Serve Like DeepSeek: A Disaggregated Inference Simulator
- Video Diffusion from Scratch: Spatiotemporal DiT with Flow Matching
- Continual Pretraining from Scratch: Teach a Tiny LM a New Domain Without Forgetting the Old
- EmbeddingGemma from Scratch: Train a Text Embedding Model in PyTorch
- RL Fine-Tuning from Scratch: Build the Environment, Train with GRPO
- Score-Based Diffusion from Scratch: SDEs, Samplers and Guidance
- DeepSeek-V4 Hybrid Attention from Scratch
- DeepSeek-V4 Training Stack from Scratch
- Sutton and Barto from Scratch 4: Planning and Policy-Gradient Methods
- Sutton and Barto from Scratch 3: n-Step Methods, Eligibility Traces and Function Approximation
- Sutton and Barto from Scratch 2: Monte Carlo and Temporal-Difference Learning
- Compilers for Machine Learning: Build a Tensor Compiler from Scratch
- CUDA Codegen on a Real GPU: From Loop IR to Tensor Cores
- RLTL;DR in a Crafting World: Teach Yourself the Tech Tree
- Lattice Deduction Transformer from Scratch
- Node Embeddings and Link Prediction from Scratch
- Model Parallelism and Disaggregation from Scratch
- Prefix Caching and Cache-Aware Routing
- LLM Inference Mechanics: Prefill, Decode and the KV Cache
- Speculative Decoding from Scratch in PyTorch
- Serving Operations: Load Testing, Autoscaling and Safe Rollouts
- Trees, Ensembles and SVMs with Scikit-Learn
- Beyond Linearity: Splines and GAMs with Scikit-Learn
- Resampling, Model Selection and Regularization with Scikit-Learn
- Long Short-Term Memory from Scratch in PyTorch
- On-Device Model Compression Pipeline
- Matrix Multiplication Algorithms in CUDA
- Recurrent-Depth Latent Reasoning
- Looped Transformer from Scratch
- Autoencoders and GANs on Fashion-MNIST in PyTorch
- Char-RNN and Seq2Seq with Attention in PyTorch
- Time-Series Forecasting in PyTorch
- Make a Deep Net Converge in PyTorch
- Fashion-MNIST Classifier in PyTorch
- Unsupervised Learning: Segment, Detect, Propagate with Scikit-Learn
- Dimensionality Reduction: Compress MNIST with Scikit-Learn
- Decision Trees: CART, Checked Against Scikit-Learn
- MNIST Classifier Pipeline with Scikit-Learn
- California Housing, End to End with Scikit-Learn
- NextLat from Scratch: Next-Latent Prediction in PyTorch
- Kimi K3 from Scratch: KDA, Attention Residuals, and Stable LatentMoE
- Tiny GPT From Scratch
- Multimodal Autoregressive Image Generator from Scratch in JAX
- RLHF from Scratch on DistilGPT2
- Vision-Language Model from Scratch in PyTorch
- Trainable Mixture of Experts in CUDA
- Build a Mini LLM Inference Server
- Attention Is All You Need: Build the Transformer From Scratch
- Reinforcement Learning for Tic-Tac-Toe: From Minimax to DQN
- Build a Trainable CNN from Scratch in NumPy
- Build Your Own teenygrad: A Tiny Tensor Autograd Engine
- AlphaZero on Connect-4 from Scratch
- VeriLLM: Publicly Verifiable Decentralized LLM Inference from Scratch in PyTorch
- RAG Pipeline
- Full-Bandwidth Transformer from Scratch
- Sim-to-Real RL with Domain Randomization
- Honest Backtesting: A Walk-Forward Signal Evaluator
- Fused LLM Inference Kernels in CUDA
- Denoising Diffusion (DDPM) from Scratch
- Mamba from Scratch: Selective State Spaces
- Flash Attention in CUDA from Scratch
- NumPy Text Classifier from Scratch
- Recirculation from Scratch
- DiLoCo: Distributed Low-Communication Training of Language Models
- Direct Preference Optimization (DPO) from Scratch
- JEPA World Model from Scratch
- Mini Distributed Training and Memory-Constrained Trainer from Scratch in NumPy
- Neural Networks From Scratch: Forward and Backward
- Message-Passing GNNs from Scratch
- Rectified Flow from Scratch
- A/B Testing & Causal Inference Toolkit
- Post-Training Safety Eval Harness
- Sutton and Barto from Scratch 1: Bandits and Dynamic Programming
- NumPy House Price Regression
- Federated Averaging (FedAvg) from Scratch in PyTorch
- NumPy Multiple Linear Regression GD
- Market-Making & Betting-Game Simulator
- Random Forest from Scratch
- Support Vector Machine from Scratch
- Q-Learning on FrozenLake from Scratch
- Build an MLP in JAX from Scratch
- LoRA Fine-Tune a Tiny Chat Model with Unsloth
- From Tokens to Bytes: Compute and Transfer Multipliers from Scratch
- Recommender Systems from Scratch: From Funk SVD to NeuMF and a Feedback Loop
- Quantum Pattern Search
- Quantum Gates, Circuits & Entanglement
- Qubit State Explorer
- Human Image Segmentation with PyTorch and U-Net
- Telecom Customer Churn Prediction with PySpark MLlib
- Character-Level Name Generation with TensorFlow
- Tweet Emotion Recognition with TensorFlow
- Pneumonia Classification with EfficientNet-B4 and PyTorch
- Facial Expression Recognition with EfficientNet-B0 and PyTorch
- Radio Signal Classification with PyTorch and EfficientNet-B0
- Aerial Image Segmentation with PyTorch U-Net
- Object Localization with PyTorch and EfficientNet-B0
- CNN Image Classification with PyTorch and Grad-CAM
- Credit Card Fraud Detection using GAN-Based Data Balancing
- Deep Learning with PyTorch: DCGAN for MNIST Handwritten Digit Generation
- Siamese Network with Triplet Loss for Person Re-Identification using PyTorch
- Fine-Tune BERT for Text Classification with TensorFlow
- Hyperparameter Tuning with Keras Tuner
- Transfer Learning for NLP with TensorFlow Hub
- Basic Image Classification with TensorFlow
- Object Localization with TensorFlow
- Keras Multitask CNN Dual Output
- CNN Filter Visualization using TensorFlow (VGG16)
- Principal Component Analysis (PCA) from scratch
- Linear Regression from scratch
- Logistic Regression from scratch
- Enterprise Knowledge Base Q&A System (RAG Implementation)
- YouTube Summarizer - Generative AI Content Pipeline
- Career Advisor Chatbot Web-App
- Sentiment Analysis of Real-time Flipkart Product Reviews
- Using MLflow for Experiment Tracking & Model Management
Pinned Loading
-
multimodal-autoregressive-image-generator-from-scratch-in-jax
multimodal-autoregressive-image-generator-from-scratch-in-jax PublicBuild a text-conditioned image generator end to end in JAX by training a VQ-VAE to turn tiny images into discrete tokens, then an autoregressive transformer that models text-prefixed token sequence…
Python 1
-
video-diffusion-from-scratch-spatiotemporal-dit-with-flow-matching
video-diffusion-from-scratch-spatiotemporal-dit-with-flow-matching PublicVideo generators such as Sora, Veo and Seedance share one backbone: a diffusion transformer over space-time patches, trained with a flow-matching objective and sampled by integrating an ODE. This p…
Python 1
-
rectified-flow-from-scratch
rectified-flow-from-scratch PublicBuild a complete rectified-flow / linear interpolant flow-matching generative model in pure PyTorch. Instead of a long discrete reverse chain that predicts noise, you learn a velocity field that tr…
Python 1
-
kimi-k3-from-scratch-kda-attention-residuals-and-stable-latentmoe
kimi-k3-from-scratch-kda-attention-residuals-and-stable-latentmoe PublicBuild every architectural innovation from the Kimi K3 technical report in numpy at toy scale: Kimi Delta Attention with lower-bounded decay and an exact chunkwise-parallel form, Gated MLA with NoPE…
Python 3
-
tiny-gpt-from-scratch
tiny-gpt-from-scratch PublicBuild a small character-level GPT end-to-end in pure NumPy, starting from tokenization and array basics and ending with multi-head self-attention, Adam, and sampling. Each step grows the same codeb…
Python 1
-
Kaggle-Notebooks-Avik
Kaggle-Notebooks-Avik PublicA collection of my Kaggle notebooks, featuring end-to-end dataset analyses, competition solutions, hackathons, and research projects across AI, Machine Learning, and Deep Learning.
Jupyter Notebook 12
If the problem persists, check the GitHub status page or contact support.





