TildAlice – Tech Blog for Developers
Python, AI, Deep Learning, Data Analysis, Computer Vision, Reinforcement Learning โ tips and tutorials
Latest Posts
-
functools.cache vs lru_cache: When 40% Speed Costs Stability
functools.cache runs 40% faster than lru_cache but can crash production. Benchmark data shows when unbounded caching breaks.
-
YOLO Video Inference Memory Leaks: 3 Fixes That Work
Fix YOLO video crashes after 2000 frames: 3 proven solutions for memory leaks, frame drops, and GPU OOM errors in production.
-
MLflow + FastAPI: $2K/Month Model Serving Side Project
MLflow tracking + FastAPI serving generates $2K/month from gradient boosting churn models. Real architecture, deployment costs included.
-
Gearbox Fault Classification: Why Voting Beat Stacking
Soft voting beat stacking by 0.7% on real gearbox fault data. Why simpler ensembles outperform complex ones in multi-fault PHM systems.
-
Gold Price Prediction: Jupyter Notebook to REST API
Build production gold price API with model versioning, drift detection, and monitoring. From Jupyter notebook to deployment in 7 steps.
-
LSTM vs Transformer for Gold Forecasting: Who Wins?
LSTM vs Transformer for gold forecasting: both hit 54% direction accuracy. Benchmark reveals why neither architecture beats the baseline.
-
ARIMA vs LSTM on Gold Prices: 7-Day vs 30-Day Forecast
ARIMA beats LSTM on 7-day gold forecasts but collapses at 30 days. Head-to-head comparison shows when each model fails and exactly why.