TildAlice – Tech Blog for Developers
Python, AI, Deep Learning, Data Analysis, Computer Vision, Reinforcement Learning โ tips and tutorials
Latest Posts
-
f-strings vs % Formatting: When Old Syntax Wins 2.1x Speed
Compare f-strings vs % formatting performance in Python. Discover why legacy syntax achieves 2.1x speed gains in specific scenarios you use daily.
-
Pandas Interpolation Methods: 5 Techniques Benchmarked on Sensor Data
Compare Pandas interpolation methods on real sensor data. Benchmark 5 techniques to find which fills missing values fastest with lowest error.
-
Recursion vs Iteration: Linked List Speed & Stack Limit
Compare recursion vs iteration for linked lists: speed benchmarks, Python stack limits, and when to choose each approach in practice.
-
PagedAttention in vLLM: KV Cache Paging for 24x Throughput
vLLM's PagedAttention cuts KV cache waste from 60-80% to near zero. Real benchmarks show 2-24x throughput gains over HuggingFaceโhere's how paging works.
-
LSTM vs GRU vs Transformer RUL: NASA CMAPSS Memory Test
Compare LSTM, GRU, and Transformer models for RUL prediction on NASA CMAPSS dataset. Which architecture wins the turbofan memory test?
-
Git Rebase –onto: 4 Interview Scenarios That Break Merge
Drop commits mid-branch, fix orphaned dependencies, rebase onto merge pointsโ4 production scenarios where --onto beats cherry-pick every time.
-
LoRA vs DoRA: 7B Model Training Speed Cuts 34% Cost
DoRA cuts LLM fine-tuning cost 44% vs LoRA but delivers 5% better multi-turn reasoning. Real A100 benchmarks, NaN debugging, and when to pick each.
-
Excel Pivot Tables to Pandas GroupBy: Migration Guide
Excel pivot tables break at 500MB. Pandas groupby handles 1M rows in 18ms. See the actual migration steps with messy data and real pitfalls.