mattdangerw
commited on
Commit
•
26a0f1c
1
Parent(s):
5e2b8ff
Upload folder using huggingface_hub
Browse files- audio_converter.json +2 -2
- config.json +2 -4
- metadata.json +4 -3
- tokenizer.json +18 -8
audio_converter.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"module": "
|
3 |
"class_name": "WhisperAudioConverter",
|
4 |
"config": {
|
5 |
"name": "whisper_audio_converter",
|
@@ -18,5 +18,5 @@
|
|
18 |
"sampling_rate": 16000,
|
19 |
"max_audio_length": 30
|
20 |
},
|
21 |
-
"registered_name": "
|
22 |
}
|
|
|
1 |
{
|
2 |
+
"module": "keras_hub.src.models.whisper.whisper_audio_converter",
|
3 |
"class_name": "WhisperAudioConverter",
|
4 |
"config": {
|
5 |
"name": "whisper_audio_converter",
|
|
|
18 |
"sampling_rate": 16000,
|
19 |
"max_audio_length": 30
|
20 |
},
|
21 |
+
"registered_name": "keras_hub>WhisperAudioConverter"
|
22 |
}
|
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"module": "
|
3 |
"class_name": "WhisperBackbone",
|
4 |
"config": {
|
5 |
"name": "whisper_backbone",
|
@@ -14,7 +14,5 @@
|
|
14 |
"max_encoder_sequence_length": 3000,
|
15 |
"max_decoder_sequence_length": 448
|
16 |
},
|
17 |
-
"registered_name": "
|
18 |
-
"assets": [],
|
19 |
-
"weights": "model.weights.h5"
|
20 |
}
|
|
|
1 |
{
|
2 |
+
"module": "keras_hub.src.models.whisper.whisper_backbone",
|
3 |
"class_name": "WhisperBackbone",
|
4 |
"config": {
|
5 |
"name": "whisper_backbone",
|
|
|
14 |
"max_encoder_sequence_length": 3000,
|
15 |
"max_decoder_sequence_length": 448
|
16 |
},
|
17 |
+
"registered_name": "keras_hub>WhisperBackbone"
|
|
|
|
|
18 |
}
|
metadata.json
CHANGED
@@ -1,6 +1,7 @@
|
|
1 |
{
|
2 |
-
"keras_version": "3.0
|
3 |
-
"
|
4 |
"parameter_count": 763857920,
|
5 |
-
"date_saved": "
|
|
|
6 |
}
|
|
|
1 |
{
|
2 |
+
"keras_version": "3.7.0",
|
3 |
+
"keras_hub_version": "0.19.0",
|
4 |
"parameter_count": 763857920,
|
5 |
+
"date_saved": "2024-12-21@23:19:34",
|
6 |
+
"tasks": []
|
7 |
}
|
tokenizer.json
CHANGED
@@ -1,12 +1,27 @@
|
|
1 |
{
|
2 |
-
"module": "
|
3 |
"class_name": "WhisperTokenizer",
|
4 |
"config": {
|
5 |
"name": "whisper_tokenizer",
|
6 |
"trainable": true,
|
7 |
-
"dtype":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
"sequence_length": null,
|
9 |
"add_prefix_space": false,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
"special_tokens": {
|
11 |
"<|startoftranscript|>": 50258,
|
12 |
"<|endoftext|>": 50257,
|
@@ -116,10 +131,5 @@
|
|
116 |
"<|zh|>": 50260
|
117 |
}
|
118 |
},
|
119 |
-
"registered_name": "
|
120 |
-
"assets": [
|
121 |
-
"assets/tokenizer/merges.txt",
|
122 |
-
"assets/tokenizer/vocabulary.json"
|
123 |
-
],
|
124 |
-
"weights": null
|
125 |
}
|
|
|
1 |
{
|
2 |
+
"module": "keras_hub.src.models.whisper.whisper_tokenizer",
|
3 |
"class_name": "WhisperTokenizer",
|
4 |
"config": {
|
5 |
"name": "whisper_tokenizer",
|
6 |
"trainable": true,
|
7 |
+
"dtype": {
|
8 |
+
"module": "keras",
|
9 |
+
"class_name": "DTypePolicy",
|
10 |
+
"config": {
|
11 |
+
"name": "int32"
|
12 |
+
},
|
13 |
+
"registered_name": null
|
14 |
+
},
|
15 |
+
"config_file": "tokenizer.json",
|
16 |
"sequence_length": null,
|
17 |
"add_prefix_space": false,
|
18 |
+
"unsplittable_tokens": [
|
19 |
+
"<|startoftranscript|>",
|
20 |
+
"<|endoftext|>",
|
21 |
+
"<|notimestamps|>",
|
22 |
+
"<|translate|>",
|
23 |
+
"<|transcribe|>"
|
24 |
+
],
|
25 |
"special_tokens": {
|
26 |
"<|startoftranscript|>": 50258,
|
27 |
"<|endoftext|>": 50257,
|
|
|
131 |
"<|zh|>": 50260
|
132 |
}
|
133 |
},
|
134 |
+
"registered_name": "keras_hub>WhisperTokenizer"
|
|
|
|
|
|
|
|
|
|
|
135 |
}
|