tokenizer.json

{
    "module": "keras_nlp.src.models.llama3.llama3_tokenizer",
    "class_name": "Llama3Tokenizer",
    "config": {
        "name": "llama3_tokenizer",
        "trainable": true,
        "dtype": {
            "module": "keras",
            "class_name": "DTypePolicy",
            "config": {
                "name": "int32"
            },
            "registered_name": null
        },
        "sequence_length": null,
        "add_prefix_space": false
    },
    "registered_name": "keras_nlp>Llama3Tokenizer"
}