TildAlice – Tech Blog for Developers
Python, AI, Deep Learning, Data Analysis, Computer Vision, Reinforcement Learning โ tips and tutorials
Latest Posts
-
Union-Find Basics to Kruskal’s MST Implementation
Union-Find from scratch: path compression, union by rank, and implementing Kruskal's MST. The O(ฮฑ(N)) trick and mistakes I made along the way.
-
Python Memory Profiling: tracemalloc and memray in Prod
Debugged a 47GB memory leak in production using tracemalloc and memray. The exact profiling workflow I use to find Python leaks in under an hour.
-
vLLM vs TensorRT-LLM: RTX 4090 Inference Benchmark
vLLM vs TensorRT-LLM head-to-head on RTX 4090 with Llama 3.1 8B โ throughput, latency, memory usage, and setup complexity compared. TensorRT-LLM wins 2.3x throughput but at what cost?
-
SCADA Data Integration for Multi-Asset PHM Pipelines
MQTT vs direct writes: we tested both on 47 SCADA assets. The hybrid approach cut sync lag 73% and survived 8 months in production.
-
DINOv2 for Medical Image Anomaly Detection
Frozen DINOv2 features beat fine-tuned autoencoders for medical anomaly detection. Here's the preprocessing mistake that tanks accuracy.
-
Spectral Kurtosis for Bearing Faults: Outperforming FFT
Spectral kurtosis caught bearing failure 3 weeks before FFT showed symptoms. When it works, when it fails, and what runs in production.
-
PaddleOCR vs EasyOCR vs Tesseract: Why PaddleOCR Is Slower
Why is PaddleOCR slower than EasyOCR? We benchmark init time, inference speed, model size, and accuracy across PP-OCRv5, EasyOCR, and Tesseract on 10K images.
-
Kalman vs Particle Filter for Bearing State Estimation
Kalman vs particle filter for bearing state estimation: real latency numbers on edge hardware, tuning tricks, and when each one breaks down.