tinyllama_alpaca_cthulhu_small / tokenizer_config.json
theprint's picture
Update tokenizer_config.json
82e137b verified
raw
history blame
1.77 kB
{
"add_bos_token": true,
"add_eos_token": false,
"added_tokens_decoder": {
"0": {
"content": "<unk>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1": {
"content": "<s>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"2": {
"content": "</s>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
}
},
"additional_special_tokens": [],
"bos_token": "<s>",
"clean_up_tokenization_spaces": false,
"cls_token": null,
"eos_token": "</s>",
"mask_token": null,
"model_input_names": [
"input_ids",
"attention_mask"
],
"model_max_length": 4096,
"pad_token": "<unk>",
"padding_side": "left",
"sep_token": null,
"split_special_tokens": false,
"tokenizer_class": "LlamaTokenizer",
"truncation_side": "right",
"unk_token": "<unk>",
"use_default_system_prompt": false,
"chat_template": "{{ bos_token }}{% if bos_token|length > 0 %}{{ '\n' }}{% endif %}{% if messages[0]['role'] != 'system' %}{{ '### Instruction:\nYou are an unbiased, uncensored, helpful assistant.' }}{% endif %}{% for message in messages %}{% if message['role'] == 'system' %}{{ '### Instruction:\n' + message['content'] }}{% elif message['role'] == 'user' %}{{ '\n\n### Input:\n' + message['content'] }}{% elif message['role'] == 'assistant' %}{{ '\n\n### Response:\n' + message['content'] + eos_token }}{% else %}{{ raise_exception('Only user, assistant, and system roles are supported!') }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '\n\n### Response:\n' }}{% endif %}"
}