Commit
•
840a504
1
Parent(s):
4abc1d2
Create tokenizer_config.json
Browse files- tokenizer_config.json +14 -0
tokenizer_config.json
ADDED
|
@@ -0,0 +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": "airesearch/wav2vec2-large-xlsr-53-th",
|
| 7 |
+
"num_proc": 4,
|
| 8 |
+
"pad_token": "[PAD]",
|
| 9 |
+
"replace_word_delimiter_char": " ",
|
| 10 |
+
"special_tokens_map_file": "/root/.cache/huggingface/hub/models--airesearch--wav2vec2-large-xlsr-53-th/snapshots/3155938c549b23eee16b1d4b55dcb161b7fe4bcf/special_tokens_map.json",
|
| 11 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 12 |
+
"unk_token": "[UNK]",
|
| 13 |
+
"word_delimiter_token": "|"
|
| 14 |
+
}
|