marian-finetuned-kde4-en-to-ro / tokenizer_config.json
Ioana23's picture
Training in progress epoch 0
aa8a306
raw
history blame contribute delete
306 Bytes
{
"clean_up_tokenization_spaces": true,
"eos_token": "</s>",
"model_max_length": 512,
"pad_token": "<pad>",
"return_tensors": "tf",
"separate_vocabs": false,
"source_lang": "en",
"sp_model_kwargs": {},
"target_lang": "ro",
"tokenizer_class": "MarianTokenizer",
"unk_token": "<unk>"
}