TildAlice – Tech Blog for Developers
Python, AI, Deep Learning, Data Analysis, Computer Vision, Reinforcement Learning โ tips and tutorials
Latest Posts
-
DreamerV3 World Model RL: 200-Line PyTorch Implementation
Build DreamerV3's world model RL from scratch in 200 lines โ RSSM dynamics, latent-space dreaming, and KL balancing tricks that prevent collapse.
-
Diffusion Models from Scratch: DDPM Training in PyTorch
Build DDPM from scratch in PyTorch: forward diffusion, U-Net denoising, training loop. Real gradient explosion fixes and NaN debugging tips.
-
Autograd Engine from Scratch: 200 Lines to Beat PyTorch
Built a 200-line autograd engine that beats PyTorch on tiny models. Here's the reverse-mode autodiff math and training loop that actually work.
-
PyTorch Neural Network Guide: 5 Mistakes That Break Training
Fix 5 silent PyTorch mistakes that break training: gradient traps, init failures, device mismatches. Real debugging code included.
-
Haar Cascades to YOLO: Face Detection Migration Guide
Migrate from Haar Cascades to YOLO for face detection. Compare accuracy, speed, real-world performance with code examples and benchmarks.
-
Mamba-2 vs Griffin vs RWKV-6: SSM Architecture Benchmark
Compare Mamba-2, Griffin, and RWKV-6 SSM architectures head-to-head. Benchmark results reveal surprising performance gaps you need to know.
-
yfinance vs Alpha Vantage vs Polygon: Real Cost Per 1M Calls
yfinance is free but breaks twice a year. Alpha Vantage costs $1,389 per million calls. Polygon offers $0-$299 tiers that actually scale. Real numbers.
-
You Don’t Need Docker for Local Dev: 4 Faster Alternatives
Skip Docker overhead in local development with these 4 lightweight alternatives. Python venv, direnv, and moreโcompare speed and simplicity.