Category
3 articles in this category.

A comprehensive explanation of transformer architecture, self-attention mechanism, and how models like GPT and BERT work under the hood.

Forecasting that survives review: start with naive baselines, build leak-free lag features, train an LSTM in PyTorch, and judge it with walk-forward validation.

When fine-tuning is worth it, how to prepare chat-format data, and a minimal QLoRA training script with transformers, peft and trl, plus how to evaluate and merge the adapter.