Fine-Tuning & Model Alignment
3 Topics
1
QLoRA & Parameter-Efficient Fine-Tuning
Low-rank adaptation, 4-bit quantization training, and Hugging Face PEFT.
High - Core skill for adapting open-source models to domain tasks.
2
DPO & RLHF: Aligning LLMs to Human Preferences
Direct Preference Optimization, reward modeling, and rejection sampling.
High - Core alignment technique used by OpenAI, Anthropic, and Meta.
3
Synthetic Data Generation at Scale
Teacher-student distillation, Evol-Instruct, and building proprietary training datasets.
High - The secret weapon behind WizardLM, Phi-3, and most top open-source models.