nanochat Guide

Auto-generated documentation for karpathy/nanochat.

1

• Overview

2

• Getting Started

2.1

• Installation and Environment Setup

2.2

• Reproducing GPT-2 Capability Model

2.3

• Running on CPU or Single GPU

3

• Training Base Models

3.1

• Configuring Model Size and Training Horizon

3.2

• Monitoring and Checkpoints

4

• Tokenizer Training and Evaluation

5

• Training Chat Models

5.1

• Supervised Finetuning (SFT)

6

• Model Evaluation

6.1

• Base Model Evaluation

6.2

• Chat Model Evaluation

7

• Chatting with Models

7.1

• Web Chat UI

7.2

• CLI Chat

8

• Advanced Workflows

9

• Configuration Reference

9.1

• Hardware and Precision Options

10

• Leaderboard and Optimization