Andrey Karpathy. Neural Networks: Zero to Hero

These are transcripts for Andrej Karpathy episodes.

• Download the audio of a youTube video (ty yt-dlp): yt-dlp -x --audio-format mp3 -o {mp3_file} -- {youtube_video_id}
• Transcribe it (ty OpenAI Whisper): whisper --language en --model large -o {out_dir} -- {mp3_file}

Send comments to @averkij on Twitter. GitHub repo is here.

1 The spelled-out intro to neural networks and backpropagation: building micrograd #1

2 The spelled-out intro to language modeling: building makemore #2

3 Building makemore Part 2: MLP #3

4 Building makemore Part 3: Activations & Gradients, BatchNorm #4