PyTorch
Tensor operations, model definition, optimization, and distributed execution. See the official PyTorch documentation ↗.
05 / Credits
A new identity, with its technical roots left visible.
Jarvisn't is Temuujin's project. It also stands on thoughtful open-source work that deserves direct credit.
Technical foundation
nanochat
Jarvisn't is built on top of nanochat, an open-source language-model training project by Andrej Karpathy. Its compact end-to-end training stack remains the technical foundation of this repository.
Visit upstream repository ↗Optimization reference
modded-nanogpt
The Muon/AdamW optimizer implementation in Jarvisn't is adapted from Keller Jordan's modded-nanogpt work. The repository and its optimization experiments are credited here as a direct technical source.
Visit Keller Jordan's repository ↗RustBPE
The tokenizer implementation used by the training pipeline.
FlashAttention
Hardware-aware attention acceleration, with a PyTorch SDPA fallback.
Next.js
The framework behind this documentation website.
Keller Jordan
Creator of modded-nanogpt, whose Muon optimizer work informed this training stack.
Implementation citations, dataset-source URLs, and historical comments remain in the code where they provide useful provenance.