Upload tokenizer
Browse files- tokenizer_config.json +3 -3
tokenizer_config.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
| 1 |
{
|
| 2 |
"bos_token": "<s>",
|
| 3 |
"do_lower_case": false,
|
|
|
|
| 4 |
"eos_token": "</s>",
|
| 5 |
"model_max_length": 1000000000000000019884624838656,
|
| 6 |
"name_or_path": "jonatasgrosman/wav2vec2-large-xlsr-53-russian",
|
| 7 |
"pad_token": "<pad>",
|
| 8 |
-
"
|
| 9 |
-
"replace_word_delimiter_char": " ",
|
| 10 |
"special_tokens_map_file": "/home/ar4ikov/.cache/huggingface/hub/models--jonatasgrosman--wav2vec2-large-xlsr-53-russian/snapshots/2329100508896c6d9b157019803ab5601e6f3406/special_tokens_map.json",
|
| 11 |
-
"tokenizer_class": "
|
| 12 |
"unk_token": "<unk>",
|
| 13 |
"word_delimiter_token": "|"
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"bos_token": "<s>",
|
| 3 |
"do_lower_case": false,
|
| 4 |
+
"do_normalize": false,
|
| 5 |
"eos_token": "</s>",
|
| 6 |
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
"name_or_path": "jonatasgrosman/wav2vec2-large-xlsr-53-russian",
|
| 8 |
"pad_token": "<pad>",
|
| 9 |
+
"return_attention_mask": false,
|
|
|
|
| 10 |
"special_tokens_map_file": "/home/ar4ikov/.cache/huggingface/hub/models--jonatasgrosman--wav2vec2-large-xlsr-53-russian/snapshots/2329100508896c6d9b157019803ab5601e6f3406/special_tokens_map.json",
|
| 11 |
+
"tokenizer_class": "Wav2Vec2Tokenizer",
|
| 12 |
"unk_token": "<unk>",
|
| 13 |
"word_delimiter_token": "|"
|
| 14 |
}
|