This language model is similar to CharLLama-2.6B in terms of training process and usage, but with a smaller capacity. It has 1,369,634,304 parameters.

For detailed information, including pretraining data, tokenization, and usage examples, please refer to the CharLLama-2.6B documentation.