UmarRamzan
commited on
Commit
•
34e478f
1
Parent(s):
f535954
Upload processor
Browse files- preprocessor_config.json +11 -0
- tokenizer_config.json +1 -0
preprocessor_config.json
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"feature_extractor_type": "SeamlessM4TFeatureExtractor",
|
3 |
+
"feature_size": 80,
|
4 |
+
"num_mel_bins": 80,
|
5 |
+
"padding_side": "right",
|
6 |
+
"padding_value": 1,
|
7 |
+
"processor_class": "Wav2Vec2BertProcessor",
|
8 |
+
"return_attention_mask": true,
|
9 |
+
"sampling_rate": 16000,
|
10 |
+
"stride": 2
|
11 |
+
}
|
tokenizer_config.json
CHANGED
@@ -39,6 +39,7 @@
|
|
39 |
"eos_token": "</s>",
|
40 |
"model_max_length": 1000000000000000019884624838656,
|
41 |
"pad_token": "[PAD]",
|
|
|
42 |
"replace_word_delimiter_char": " ",
|
43 |
"target_lang": null,
|
44 |
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
|
|
39 |
"eos_token": "</s>",
|
40 |
"model_max_length": 1000000000000000019884624838656,
|
41 |
"pad_token": "[PAD]",
|
42 |
+
"processor_class": "Wav2Vec2BertProcessor",
|
43 |
"replace_word_delimiter_char": " ",
|
44 |
"target_lang": null,
|
45 |
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|