Skip to content
#

tranformer

Here are 22 public repositories matching this topic...

A transformer language model built entirely from scratch — hand-written tokenizer, attention mechanism, and training loop in PyTorch. Pretrained on Wikipedia, fine-tuned on real SEC filings. Every loss curve and generated sample in this repo is real, captured output.

  • Updated Aug 22, 2026
  • Jupyter Notebook

Add this topic to your repo

To associate your repository with the tranformer topic, visit your repo's landing page and select "manage topics."

Learn more