makemore
A small, hackable character-level language model that learns patterns from line-based text files and generates similar items, such as name ideas.
Share on XLicense: MIT
Overview
makemore is a small Python program that learns character patterns from text with one item per line and generates new items resembling the examples. It helps create more names, company names, or word-like strings from a list. Run it with an input file and output directory to train a model; training logs and the model are saved there. Model choices range from bigrams to a Transformer, and you can generate more samples from the best saved model.
Key features
- Train on a text file with one item per line and save logs and the model to an output directory.
- Generate new character sequences that resemble the training text.
- Choose among bigram, MLP, RNN, LSTM, GRU, and Transformer models; CNN is in progress.
- Use --sample-only to generate samples from the best saved model.
Best for
Learners exploring character-level language models and people who want to generate name-like or word-like examples from a text list.
- Upstream
- karpathy/makemore
- Fork on GitHub
- Guo-astro/makemore
- Upstream stars
- 4.4k
- Category
- AI agents and LLM tools
- Language
- Python
- License
- MIT
- Forked
- 2026-10-09
- Sync status
- In syncLast synced 2026-10-10
More in AI agents and LLM tools
An open-source curriculum for learning AI engineering by implementing model internals, retrieval pipelines, and agent runtimes. Study through lessons, interactive labs, or staged coding projects, and inspect the code and evaluation results.
Forked 2026-10-09Last synced 2026-10-10License: MITAI agents and LLM toolsGitHub
uzu is an inference engine for running AI models in apps, helping developers keep inference on-device instead of relying on remote services. Use it to download a model and build chat features.
Forked 2026-10-09Last synced 2026-10-10License: MITAI agents and LLM toolsGitHub
Anti Slop gives AI coding agents rules to avoid generic UI, filler text, and AI-shaped code. Use it as a filter alongside your own design direction.
Forked 2026-10-08Last synced 2026-10-10License: MITAI agents and LLM toolsGitHub
Apache Maka is an agent workspace that records every run, tool action, and model decision so work can be reviewed, recovered, and repeated with a clear trace.
Forked 2026-10-08Last synced 2026-10-10License: Apache-2.0AI agents and LLM toolsGitHub