Instruction Tuning

Why does completion-only loss usually outperform full-sequence loss on the same SFT data?

Tulu-style instruction tuning trades volume for diversity: which axes of diversity matter most?

Walk through the triage order when a customer-support LoRA tanks MMLU by 4 points post-deploy

Which operators are part of the Evol-Instruct pipeline for growing instruction data?

Tulu instruction-tuning family: what is it and who publishes it?

Define supervised fine-tuning (SFT) in one breath.

Roles in a chat training example: pick the three that are standard.

Base model vs instruction-tuned model: what does the user actually feel?

Alpaca-format examples carry three fields: name them

A hand-built chat prompt performs worse than apply_chat_template. Spot the tokenization issues.

Flashcard: what does zero-shot prompting mean and what makes it work at all?

Order these instruction datasets from highest to lowest quality per example

Match instruction dataset to its source / construction method

Why must developers use a model's specific apply_chat_template() rather than building the prompt string manually?