mlc-chat-config.json

{
    "model_lib": "TinyLlama-1.1b-q4f16_1",
    "local_id": "TinyLlama-1.1B-Chat-v0.3-q4f16_1",
    "estimated_vram_req": 1000000000,
    "conv_config": {
        "seps": ["\n","<|im_end|>"],
        "stop_tokens": [ 2 ],
        "offset": 0,
        "separator_style": 0,
        "messages": [
        ],
        "stop_str": "<|im_end|>",
        "roles": [ "<|im_start|>user", "<|im_start|>assistant" ],
        "role_msg_sep": ": ",
        "role_empty_sep": "\n",
        "system": "<|im_start|>system you are a helpful assistant<|im_end|>",
        "add_bos": true,
        "prefix_tokens": [],
        "name": "chatml2"
    },
    "temperature": 0.7,
    "repetition_penalty": 1.0,
    "top_p": 0.95,
    "mean_gen_len": 128,
    "max_gen_len": 512,
    "max_window_size": 2048,
    "num_shards": 1,
    "shift_fill_factor": 0.3,
    "tokenizer_files": [
        "added_tokens.json",
        "tokenizer.json",
        "tokenizer.model"
    ],
    "model_category": "llama",
    "model_name": "TinyLlama-1.1B-Chat-v0.3",
    "vocab_size": 32003
}