English
← Back to AI Technology

Training and post-training

LoRA / QLoRA

Parameter-efficient fine-tuning methods that train small low-rank adapters while keeping base weights frozen. QLoRA combines adapters with a quantized base model to reduce memory use further.

Reference period
2021 / 2023
Last reviewed

Key terms

  • low-rank adapters
  • frozen base model
  • 4-bit training
  • PEFT

Connected across the project

Primary sources

  1. LoRA
  2. QLoRA

This is a curated technical map, not a claim of comprehensive coverage.