Appendix D of 'Born': References and Online Resources

Forum topic · QianXun · 2026-06-13

Summary

Appendix D of the serialized technical book 'Born' compiles the complete reference list cited throughout the book, covering deep learning fundamentals, Transformers and LLMs, inference optimization, model formats, GPU computing, and related frameworks. Key citations include LeNet-5 (LeCun et al., 1998), the Adam optimizer (Kingma & Ba, 2015), batch normalization (Ioffe & Szegedy, 2015), the Transformer architecture (Vaswani et al., 2017), LLaMA (Touvron et al., 2023), RoPE (Su et al., 2021), FlashAttention (Dao et al., 2022), and KV-Cache scaling analysis (Pope et al., 2023). Model formats covered are ONNX and GGUF from llama.cpp, with WebGPU as the GPU standard. Related frameworks listed include Burn (a Rust deep learning framework that inspired Born's design), Gorgonia (an early Go deep learning project), and llama.cpp. The appendix also provides online resources such as the Born GitHub repository, Go documentation, WebGPU samples, and HuggingFace. This appendix completes the book's 22 chapters and 4 appendices.

This appendix lists the technical papers, standards documents, and open-source projects cited throughout *Born*.

Deep Learning Fundamentals

1. LeCun et al. (1998). Gradient-based learning applied to document recognition. *Proc. IEEE*. — MNIST and LeNet-5 2. Kingma & Ba (2015). Adam: A method for stochastic optimization. *ICLR*. — The Adam optimizer 3. Ioffe & Szegedy (2015). Batch normalization. *ICML*. — Batch normalization

Transformers and LLMs

4. Vaswani et al. (2017). Attention is all you need. *NeurIPS*. — The Transformer 5. Touvron et al. (2023). LLaMA: Open and efficient foundation language models. *Meta AI*. — The LLaMA architecture 6. Su et al. (2021). RoFormer: Enhanced transformer with rotary position embedding. — RoPE

Inference Optimization

7. Dao et al. (2022). FlashAttention: Fast and memory-efficient exact attention. *NeurIPS*. — Flash Attention 8. Pope et al. (2023). Efficiently scaling transformer inference. — KV-Cache analysis

Model Formats

9. ONNX (2017–present). https://onnx.ai/ — Open Neural Network Exchange format 10. Gerganov (2023). GGUF format specification. *llama.cpp*. — The GGUF format

GPU Computing

11. W3C (2023). WebGPU Specification. https://www.w3.org/TR/webgpu/ — The WebGPU standard

Related Frameworks

12. Burn (2023–present). https://burn.dev/ — Rust deep learning framework; a design inspiration for Born 13. Gorgonia (2016–present). https://gorgonia.org/ — An early deep learning attempt in Go 14. llama.cpp (2023–present). https://github.com/ggerganov/llama.cpp — C++ LLM inference engine

Online Resources

| Resource | URL | |------|-----| | Born official repository | https://github.com/born-ml/born | | Go official documentation | https://go.dev/doc/ | | WebGPU tutorials | https://webgpu.github.io/webgpu-samples/ | | HuggingFace | https://huggingface.co/ |

---

📘 *Born* is a serialized technical book. This is Appendix D of 4.

With this appendix, all 22 chapters and 4 appendices of the book have been published. Thank you for reading!