Upload production-ready Nepali tokenizer with full config
Browse files- tokenizer_config.json +5 -1
tokenizer_config.json
CHANGED
|
@@ -1,6 +1,10 @@
|
|
| 1 |
{
|
| 2 |
"backend": "tokenizers",
|
|
|
|
| 3 |
"clean_up_tokenization_spaces": true,
|
|
|
|
| 4 |
"model_max_length": 1000000000000000019884624838656,
|
| 5 |
-
"
|
|
|
|
|
|
|
| 6 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "[BOS]",
|
| 4 |
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "[EOS]",
|
| 6 |
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"pad_token": "[PAD]",
|
| 8 |
+
"tokenizer_class": "TokenizersBackend",
|
| 9 |
+
"unk_token": "[UNK]"
|
| 10 |
}
|