skip to content
luminary.blog
by Oz Akan

Technical

RSS feed

Posts in 2025

  • $Which Loss Function Do LLMs use?

    Exploring Cross-Entropy Loss in Large Language Models.

    fountain sketch
  • $What is Matryoshka Representation Learning (MRL)?

    Nesting Power and Flexibility into ML Embeddings

    Matryoshka Babies
  • $2-Hour Streamlit Workshop

    Beginners with a basic understanding of Python. No web development experience is required.

    graph sketch
  • $UE8M0 FP8 Number Format

    Training LLMs without H100 using UE8M0 FP8 number format.

    fp8 sketch
  • $Understanding ML Numerical Formats

    Understanding INT4, INT8, FP16, BF16, and TF32 formats in machine learning - their precision, speed, and memory trade-offs for training and inference.

    number one sketched
  • $What do GPT-OSS and Gemma 3 really offer?

    GPT-OSS and Gemma 3: two new small-but-powerful language models pushing the boundaries.

    baby robot
  • $What are Positional Embeddings?

    The mathematical technique that teaches AI models where each word sits in a sequence.

    suprised robot
  • $Words, Tokens and Embeddings

    How language models convert token IDs into meaningful vector representations that capture semantic relationships.

    happy robot
  • $Subword Tokenization Algorithms

    Understanding the algorithms behind tokenization in Large Language Models.

    cute robot sketch
  • $What is LLM Inference?

    Understanding how Large Language Models generate text through the inference process.

    cute robot sketch