Train a tiny causal language model across multiple GPUs using byte-level Tiny Shakespeare data.

Soon you can unlock how this was done.

Behind this: the tool used · the method · what actually resulted.