Upload WhisperForConditionalGeneration
Browse files- config.json +13 -3
- generation_config.json +4 -6
- pytorch_model.bin +2 -2
config.json
CHANGED
@@ -1,7 +1,8 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"activation_dropout": 0.0,
|
4 |
"activation_function": "gelu",
|
|
|
5 |
"architectures": [
|
6 |
"WhisperForConditionalGeneration"
|
7 |
],
|
@@ -11,6 +12,7 @@
|
|
11 |
50257
|
12 |
],
|
13 |
"bos_token_id": 50257,
|
|
|
14 |
"d_model": 1024,
|
15 |
"decoder_attention_heads": 16,
|
16 |
"decoder_ffn_dim": 4096,
|
@@ -26,16 +28,24 @@
|
|
26 |
"forced_decoder_ids": null,
|
27 |
"init_std": 0.02,
|
28 |
"is_encoder_decoder": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
"max_length": 448,
|
30 |
"max_source_positions": 1500,
|
31 |
"max_target_positions": 448,
|
|
|
32 |
"model_type": "whisper",
|
33 |
"num_hidden_layers": 24,
|
34 |
"num_mel_bins": 80,
|
35 |
"pad_token_id": 50257,
|
36 |
"scale_embedding": false,
|
37 |
"torch_dtype": "float32",
|
38 |
-
"transformers_version": "4.
|
39 |
-
"use_cache":
|
|
|
40 |
"vocab_size": 51865
|
41 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "output",
|
3 |
"activation_dropout": 0.0,
|
4 |
"activation_function": "gelu",
|
5 |
+
"apply_spec_augment": true,
|
6 |
"architectures": [
|
7 |
"WhisperForConditionalGeneration"
|
8 |
],
|
|
|
12 |
50257
|
13 |
],
|
14 |
"bos_token_id": 50257,
|
15 |
+
"classifier_proj_size": 256,
|
16 |
"d_model": 1024,
|
17 |
"decoder_attention_heads": 16,
|
18 |
"decoder_ffn_dim": 4096,
|
|
|
28 |
"forced_decoder_ids": null,
|
29 |
"init_std": 0.02,
|
30 |
"is_encoder_decoder": true,
|
31 |
+
"mask_feature_length": 64,
|
32 |
+
"mask_feature_min_masks": 0,
|
33 |
+
"mask_feature_prob": 0.1,
|
34 |
+
"mask_time_length": 10,
|
35 |
+
"mask_time_min_masks": 2,
|
36 |
+
"mask_time_prob": 0.3,
|
37 |
"max_length": 448,
|
38 |
"max_source_positions": 1500,
|
39 |
"max_target_positions": 448,
|
40 |
+
"median_filter_width": 7,
|
41 |
"model_type": "whisper",
|
42 |
"num_hidden_layers": 24,
|
43 |
"num_mel_bins": 80,
|
44 |
"pad_token_id": 50257,
|
45 |
"scale_embedding": false,
|
46 |
"torch_dtype": "float32",
|
47 |
+
"transformers_version": "4.31.0.dev0",
|
48 |
+
"use_cache": true,
|
49 |
+
"use_weighted_layer_sum": false,
|
50 |
"vocab_size": 51865
|
51 |
}
|
generation_config.json
CHANGED
@@ -14,10 +14,6 @@
|
|
14 |
[
|
15 |
2,
|
16 |
50359
|
17 |
-
],
|
18 |
-
[
|
19 |
-
3,
|
20 |
-
50363
|
21 |
]
|
22 |
],
|
23 |
"is_multilingual": true,
|
@@ -125,7 +121,7 @@
|
|
125 |
"max_initial_timestamp_index": 1,
|
126 |
"max_length": 448,
|
127 |
"no_timestamps_token_id": 50363,
|
128 |
-
"pad_token_id":
|
129 |
"suppress_tokens": [
|
130 |
1,
|
131 |
2,
|
@@ -210,6 +206,8 @@
|
|
210 |
49870,
|
211 |
50254,
|
212 |
50258,
|
|
|
|
|
213 |
50360,
|
214 |
50361,
|
215 |
50362
|
@@ -218,5 +216,5 @@
|
|
218 |
"transcribe": 50359,
|
219 |
"translate": 50358
|
220 |
},
|
221 |
-
"transformers_version": "4.
|
222 |
}
|
|
|
14 |
[
|
15 |
2,
|
16 |
50359
|
|
|
|
|
|
|
|
|
17 |
]
|
18 |
],
|
19 |
"is_multilingual": true,
|
|
|
121 |
"max_initial_timestamp_index": 1,
|
122 |
"max_length": 448,
|
123 |
"no_timestamps_token_id": 50363,
|
124 |
+
"pad_token_id": 50257,
|
125 |
"suppress_tokens": [
|
126 |
1,
|
127 |
2,
|
|
|
206 |
49870,
|
207 |
50254,
|
208 |
50258,
|
209 |
+
50358,
|
210 |
+
50359,
|
211 |
50360,
|
212 |
50361,
|
213 |
50362
|
|
|
216 |
"transcribe": 50359,
|
217 |
"translate": 50358
|
218 |
},
|
219 |
+
"transformers_version": "4.31.0.dev0"
|
220 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cad5840b5ac1aec0398192fab7f06e682471d5636d33e4a8cee42517d8e5a151
|
3 |
+
size 3055741918
|