TriDLM-124M-split / tokenizer_config.json
AntonXue's picture
TriDLM split: final checkpoint (20260829-1047-qk1-lr1e-4, step 100000)
6611f0e verified
Raw
History Blame Contribute Delete
340 Bytes
{
"add_prefix_space": false,
"backend": "tokenizers",
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"errors": "replace",
"is_local": false,
"local_files_only": true,
"mask_token": "[MASK]",
"model_max_length": 1024,
"pad_token": null,
"tokenizer_class": "GPT2Tokenizer",
"unk_token": "<|endoftext|>"
}