GPT-2-wikitext-chunks-epsroot-0 / tokenizer_config.json
luciaquirke's picture
Add base GPT-2 trained on wikitext 512-chunks (eps_root=0), before leave-1%-out retrains (checkpoints/step_287)
024b394 verified
Raw
History Blame Contribute Delete
315 Bytes
{
"add_prefix_space": false,
"backend": "tokenizers",
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"errors": "replace",
"is_local": false,
"local_files_only": false,
"model_max_length": 1024,
"pad_token": null,
"tokenizer_class": "GPT2Tokenizer",
"unk_token": "<|endoftext|>"
}