Let's Talk SSL
My rambling on state of art model comparison in self supervised learning and everything in between: No math just thoughts, opinions and vibes
Aug 4, 20266 min read210
Command Palette
Search for a command to run...
Search for a command to run...
My rambling on state of art model comparison in self supervised learning and everything in between: No math just thoughts, opinions and vibes
Over the last three weeks I built a decoder-only GPT, a Vision Transformer, RoPE, and GQA, all from raw PyTorch, no copy-pasting a reference implementation. The models train and the loss curves look r