TinyTalk-2 / tokenizer_config.json
TheREZOR's picture
TinyTalk 2: 8M chat fine-tune (cardputer_ai v2.0)
761e44e verified
Raw
History Blame Contribute Delete
488 Bytes
{
"add_prefix_space": false,
"backend": "tokenizers",
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"errors": "replace",
"is_local": true,
"local_files_only": false,
"model_max_length": 2048,
"pad_token": null,
"tokenizer_class": "GPT2Tokenizer",
"unk_token": "<|endoftext|>",
"chat_template": "{% for m in messages %}{% if m['role'] == 'user' %}User: {{ m['content'] }}\n{% else %}Bot: {{ m['content'] }}<|endoftext|>\n{% endif %}{% endfor %}Bot:"
}