* Clean-up the llama2.c wasm example. * Use a proper tokenizer.
* Move the whisper example. * More renaming. * Add llama2 as a new wasm example. * Live generation. * More of the llama wasm example. * Formatting.