Instructions to use Xenova/Kimi-K3-tokenizer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Xenova/Kimi-K3-tokenizer with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Xenova/Kimi-K3-tokenizer", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "additional_special_tokens": [ | |
| "<|end_of_msg|>", | |
| "[start_header_id]", | |
| "[end_header_id]", | |
| "[EOT]", | |
| "<|media_begin|>", | |
| "<|media_content|>", | |
| "<|media_end|>", | |
| "<|media_pad|>", | |
| "<osagent_mode>" | |
| ], | |
| "backend": "tokenizers", | |
| "bos_token": "[BOS]", | |
| "clean_up_tokenization_spaces": false, | |
| "eos_token": "[EOS]", | |
| "model_max_length": 1000000000000000019884624838656, | |
| "pad_token": "[PAD]", | |
| "tokenizer_class": "PreTrainedTokenizerFast", | |
| "unk_token": "[UNK]" | |
| } | |