Tag: Benchmarking
-
Stop Using Temperature 0 for LLM Evals: Why It Breaks Benchmarks
Temperature 0 breaks LLM evals by hiding variance and selecting for memorization. Here's why you should sample at 0.5 instead โ with real accuracy gaps.
-
list.append vs Comprehension vs deque: 100k Benchmark
Benchmark reveals why list comprehensions beat append() by 15% in Pythonโbut there's an even faster method for large datasets you should know.
-
FastAPI vs Flask ML Serving: 5 Benchmarks Beginners Miss
5 FastAPI vs Flask benchmarks with real models: async preprocessing wins 2.3x, streaming shines, but Flask still beats memory usage. See where each wins.