Text Generation
Transformers
Safetensors
cagliostro
causal-lm
small-language-model
from-scratch
custom_code
cagliostro-v2 / tokenizer_config.json
TobiasLogic's picture
tokenizer: restrict model_input_names so generate(**tok(...)) works
c747186 verified
Raw
History Blame Contribute Delete
242 Bytes
{
"tokenizer_class": "PreTrainedTokenizerFast",
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"unk_token": "<|endoftext|>",
"model_max_length": 4096,
"model_input_names": [
"input_ids",
"attention_mask"
]
}