The model's token ids for text, as its tokenizer splits it — control tokens written as text (<end_of_turn>) included — with no BOS added.
text
<end_of_turn>
Future<List<int>> tokenize(String text);