TinyStoriesV2-656K code

What's Inside

• train_.ipynb: This one's all about training a language model on the TinyStoriesV2 dataset. We're talking data prep, model setup, training details, and even a bit of model inference.

• train_tokenizer.ipynb: Here, we're crafting a brand-new tokenizer from scratch, just for our TinyStoriesV2 needs. It's got dataset loading, corpus prep for tokenizing, and the actual tokenizer training.