Upload WhisperForConditionalGeneration
Browse files- config.json +65 -0
- generation_config.json +176 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./models/whisper-large-v3-no-numbers",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"apply_spec_augment": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"WhisperForConditionalGeneration"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"begin_suppress_tokens": [
|
| 11 |
+
210,
|
| 12 |
+
49839
|
| 13 |
+
],
|
| 14 |
+
"bos_token_id": 49839,
|
| 15 |
+
"classifier_proj_size": 256,
|
| 16 |
+
"d_model": 1280,
|
| 17 |
+
"decoder_attention_heads": 20,
|
| 18 |
+
"decoder_ffn_dim": 5120,
|
| 19 |
+
"decoder_layerdrop": 0.0,
|
| 20 |
+
"decoder_layers": 32,
|
| 21 |
+
"decoder_start_token_id": 49840,
|
| 22 |
+
"dropout": 0.0,
|
| 23 |
+
"encoder_attention_heads": 20,
|
| 24 |
+
"encoder_ffn_dim": 5120,
|
| 25 |
+
"encoder_layerdrop": 0.0,
|
| 26 |
+
"encoder_layers": 32,
|
| 27 |
+
"eos_token_id": 49839,
|
| 28 |
+
"forced_decoder_ids": [
|
| 29 |
+
[
|
| 30 |
+
1,
|
| 31 |
+
49841
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
2,
|
| 35 |
+
49941
|
| 36 |
+
],
|
| 37 |
+
[
|
| 38 |
+
3,
|
| 39 |
+
49945
|
| 40 |
+
]
|
| 41 |
+
],
|
| 42 |
+
"init_std": 0.02,
|
| 43 |
+
"is_encoder_decoder": true,
|
| 44 |
+
"mask_feature_length": 10,
|
| 45 |
+
"mask_feature_min_masks": 0,
|
| 46 |
+
"mask_feature_prob": 0.0,
|
| 47 |
+
"mask_time_length": 10,
|
| 48 |
+
"mask_time_min_masks": 2,
|
| 49 |
+
"mask_time_prob": 0.05,
|
| 50 |
+
"max_length": 448,
|
| 51 |
+
"max_source_positions": 1500,
|
| 52 |
+
"max_target_positions": 448,
|
| 53 |
+
"median_filter_width": 7,
|
| 54 |
+
"model_type": "whisper",
|
| 55 |
+
"num_hidden_layers": 32,
|
| 56 |
+
"num_mel_bins": 128,
|
| 57 |
+
"pad_token_id": 49838,
|
| 58 |
+
"scale_embedding": false,
|
| 59 |
+
"suppress_tokens": [],
|
| 60 |
+
"torch_dtype": "float32",
|
| 61 |
+
"transformers_version": "4.44.0",
|
| 62 |
+
"use_cache": true,
|
| 63 |
+
"use_weighted_layer_sum": false,
|
| 64 |
+
"vocab_size": 51447
|
| 65 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,176 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alignment_heads": [
|
| 3 |
+
[
|
| 4 |
+
7,
|
| 5 |
+
0
|
| 6 |
+
],
|
| 7 |
+
[
|
| 8 |
+
10,
|
| 9 |
+
17
|
| 10 |
+
],
|
| 11 |
+
[
|
| 12 |
+
12,
|
| 13 |
+
18
|
| 14 |
+
],
|
| 15 |
+
[
|
| 16 |
+
13,
|
| 17 |
+
12
|
| 18 |
+
],
|
| 19 |
+
[
|
| 20 |
+
16,
|
| 21 |
+
1
|
| 22 |
+
],
|
| 23 |
+
[
|
| 24 |
+
17,
|
| 25 |
+
14
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
19,
|
| 29 |
+
11
|
| 30 |
+
],
|
| 31 |
+
[
|
| 32 |
+
21,
|
| 33 |
+
4
|
| 34 |
+
],
|
| 35 |
+
[
|
| 36 |
+
24,
|
| 37 |
+
1
|
| 38 |
+
],
|
| 39 |
+
[
|
| 40 |
+
25,
|
| 41 |
+
6
|
| 42 |
+
]
|
| 43 |
+
],
|
| 44 |
+
"begin_suppress_tokens": [
|
| 45 |
+
210,
|
| 46 |
+
49839
|
| 47 |
+
],
|
| 48 |
+
"bos_token_id": 49839,
|
| 49 |
+
"decoder_start_token_id": 49840,
|
| 50 |
+
"eos_token_id": 49839,
|
| 51 |
+
"forced_decoder_ids": [
|
| 52 |
+
[
|
| 53 |
+
1,
|
| 54 |
+
null
|
| 55 |
+
],
|
| 56 |
+
[
|
| 57 |
+
2,
|
| 58 |
+
49941
|
| 59 |
+
]
|
| 60 |
+
],
|
| 61 |
+
"is_multilingual": true,
|
| 62 |
+
"lang_to_id": {
|
| 63 |
+
"<|af|>": 49909,
|
| 64 |
+
"<|am|>": 49916,
|
| 65 |
+
"<|ar|>": 49854,
|
| 66 |
+
"<|as|>": 49932,
|
| 67 |
+
"<|az|>": 49886,
|
| 68 |
+
"<|ba|>": 49937,
|
| 69 |
+
"<|be|>": 49912,
|
| 70 |
+
"<|bg|>": 49874,
|
| 71 |
+
"<|bn|>": 49884,
|
| 72 |
+
"<|bo|>": 49929,
|
| 73 |
+
"<|br|>": 49891,
|
| 74 |
+
"<|bs|>": 49897,
|
| 75 |
+
"<|ca|>": 49852,
|
| 76 |
+
"<|cs|>": 49865,
|
| 77 |
+
"<|cy|>": 49879,
|
| 78 |
+
"<|da|>": 49867,
|
| 79 |
+
"<|de|>": 49843,
|
| 80 |
+
"<|el|>": 49863,
|
| 81 |
+
"<|en|>": 49841,
|
| 82 |
+
"<|es|>": 49844,
|
| 83 |
+
"<|et|>": 49889,
|
| 84 |
+
"<|eu|>": 49892,
|
| 85 |
+
"<|fa|>": 49882,
|
| 86 |
+
"<|fi|>": 49859,
|
| 87 |
+
"<|fo|>": 49920,
|
| 88 |
+
"<|fr|>": 49847,
|
| 89 |
+
"<|gl|>": 49901,
|
| 90 |
+
"<|gu|>": 49915,
|
| 91 |
+
"<|haw|>": 49934,
|
| 92 |
+
"<|ha|>": 49936,
|
| 93 |
+
"<|he|>": 49861,
|
| 94 |
+
"<|hi|>": 49858,
|
| 95 |
+
"<|hr|>": 49873,
|
| 96 |
+
"<|ht|>": 49921,
|
| 97 |
+
"<|hu|>": 49868,
|
| 98 |
+
"<|hy|>": 49894,
|
| 99 |
+
"<|id|>": 49857,
|
| 100 |
+
"<|is|>": 49893,
|
| 101 |
+
"<|it|>": 49856,
|
| 102 |
+
"<|ja|>": 49848,
|
| 103 |
+
"<|jw|>": 49938,
|
| 104 |
+
"<|ka|>": 49911,
|
| 105 |
+
"<|kk|>": 49898,
|
| 106 |
+
"<|km|>": 49905,
|
| 107 |
+
"<|kn|>": 49888,
|
| 108 |
+
"<|ko|>": 49846,
|
| 109 |
+
"<|la|>": 49876,
|
| 110 |
+
"<|lb|>": 49927,
|
| 111 |
+
"<|ln|>": 49935,
|
| 112 |
+
"<|lo|>": 49918,
|
| 113 |
+
"<|lt|>": 49875,
|
| 114 |
+
"<|lv|>": 49883,
|
| 115 |
+
"<|mg|>": 49931,
|
| 116 |
+
"<|mi|>": 49877,
|
| 117 |
+
"<|mk|>": 49890,
|
| 118 |
+
"<|ml|>": 49878,
|
| 119 |
+
"<|mn|>": 49896,
|
| 120 |
+
"<|mr|>": 49902,
|
| 121 |
+
"<|ms|>": 49864,
|
| 122 |
+
"<|mt|>": 49925,
|
| 123 |
+
"<|my|>": 49928,
|
| 124 |
+
"<|ne|>": 49895,
|
| 125 |
+
"<|nl|>": 49853,
|
| 126 |
+
"<|nn|>": 49924,
|
| 127 |
+
"<|no|>": 49870,
|
| 128 |
+
"<|oc|>": 49910,
|
| 129 |
+
"<|pa|>": 49903,
|
| 130 |
+
"<|pl|>": 49851,
|
| 131 |
+
"<|ps|>": 49922,
|
| 132 |
+
"<|pt|>": 49849,
|
| 133 |
+
"<|ro|>": 49866,
|
| 134 |
+
"<|ru|>": 49845,
|
| 135 |
+
"<|sa|>": 49926,
|
| 136 |
+
"<|sd|>": 49914,
|
| 137 |
+
"<|si|>": 49904,
|
| 138 |
+
"<|sk|>": 49880,
|
| 139 |
+
"<|sl|>": 49887,
|
| 140 |
+
"<|sn|>": 49906,
|
| 141 |
+
"<|so|>": 49908,
|
| 142 |
+
"<|sq|>": 49899,
|
| 143 |
+
"<|sr|>": 49885,
|
| 144 |
+
"<|su|>": 49939,
|
| 145 |
+
"<|sv|>": 49855,
|
| 146 |
+
"<|sw|>": 49900,
|
| 147 |
+
"<|ta|>": 49869,
|
| 148 |
+
"<|te|>": 49881,
|
| 149 |
+
"<|tg|>": 49913,
|
| 150 |
+
"<|th|>": 49871,
|
| 151 |
+
"<|tk|>": 49923,
|
| 152 |
+
"<|tl|>": 49930,
|
| 153 |
+
"<|tr|>": 49850,
|
| 154 |
+
"<|tt|>": 49933,
|
| 155 |
+
"<|uk|>": 49862,
|
| 156 |
+
"<|ur|>": 49872,
|
| 157 |
+
"<|uz|>": 49919,
|
| 158 |
+
"<|vi|>": 49860,
|
| 159 |
+
"<|yi|>": 49917,
|
| 160 |
+
"<|yo|>": 49907,
|
| 161 |
+
"<|yue|>": 49940,
|
| 162 |
+
"<|zh|>": 49842
|
| 163 |
+
},
|
| 164 |
+
"max_initial_timestamp_index": 50,
|
| 165 |
+
"max_length": 448,
|
| 166 |
+
"no_timestamps_token_id": 49946,
|
| 167 |
+
"pad_token_id": 49839,
|
| 168 |
+
"prev_sot_token_id": 49944,
|
| 169 |
+
"return_timestamps": false,
|
| 170 |
+
"suppress_tokens": [],
|
| 171 |
+
"task_to_id": {
|
| 172 |
+
"transcribe": 49942,
|
| 173 |
+
"translate": 49941
|
| 174 |
+
},
|
| 175 |
+
"transformers_version": "4.44.0"
|
| 176 |
+
}
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6269c5b070d7b142a9ad236f3aa540d51e3513de94aa4ce2fb4cc3cabbb2e147
|
| 3 |
+
size 4991303600
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:630ca774672856d2e0e39a702e590f635a1cfc5726a64b6578ab46dd367369a9
|
| 3 |
+
size 1180663192
|
model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|