References & Further Reading

References

  1. E. J. Hu, Y. Shen, P. Wallis, Z. Allen-Zhu, Y. Li, S. Wang, L. Wang, and W. Chen, LoRA: Low-Rank Adaptation of Large Language Models, ICLR 2022, 2021

    Introduces LoRA for parameter-efficient fine-tuning.

  2. T. Dettmers, A. Pagnoni, A. Holtzman, and L. Zettlemoyer, QLoRA: Efficient Finetuning of Quantized Language Models, NeurIPS, 2023

    QLoRA enables fine-tuning 65B models on a single 48GB GPU.

  3. A. Karpathy, nanoGPT, 2023

    Minimal GPT implementation for educational purposes.

  4. R. Taori et al., Stanford Alpaca: An Instruction-following LLaMA Model, 2023

    Demonstrates instruction tuning on 52K examples.

  5. H. Liu, C. Li, Q. Wu, and Y. J. Lee, Visual Instruction Tuning, NeurIPS, 2023

    LLaVA: vision-language model via visual instruction tuning.

Further Reading

  • PEFT library

    HuggingFace PEFT (https://github.com/huggingface/peft)

    Production-ready implementation of LoRA, QLoRA, and other PEFT methods.

  • Fine-tuning best practices

    S. Gunasekar et al., Textbooks Are All You Need, 2023

    Shows that data quality dramatically impacts fine-tuning effectiveness.

  • Model merging

    mergekit (https://github.com/cg123/mergekit)

    Tools for merging multiple fine-tuned model variants.

ExercisesCh 38 →