bigmorning commited on
Commit
f257d54
1 Parent(s): 8c76755

Upload TFWhisperForConditionalGeneration

Browse files
Files changed (4) hide show
  1. README.md +73 -0
  2. config.json +64 -0
  3. generation_config.json +117 -0
  4. tf_model.h5 +3 -0
README.md ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: openai/whisper-tiny
4
+ tags:
5
+ - generated_from_keras_callback
6
+ model-index:
7
+ - name: wh_4_sun_syl_w_0_lr_8en5_b32_0015
8
+ results: []
9
+ ---
10
+
11
+ <!-- This model card has been generated automatically according to the information Keras had access to. You should
12
+ probably proofread and complete it, then remove this comment. -->
13
+
14
+ # wh_4_sun_syl_w_0_lr_8en5_b32_0015
15
+
16
+ This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on an unknown dataset.
17
+ It achieves the following results on the evaluation set:
18
+ - Train Loss: 1.7258
19
+ - Train Accuracy: 0.0223
20
+ - Train Wermet: 0.5356
21
+ - Train Wermet Syl: 0.7426
22
+ - Validation Loss: 1.4314
23
+ - Validation Accuracy: 0.0186
24
+ - Validation Wermet: 0.4270
25
+ - Validation Wermet Syl: 0.4029
26
+ - Epoch: 14
27
+
28
+ ## Model description
29
+
30
+ More information needed
31
+
32
+ ## Intended uses & limitations
33
+
34
+ More information needed
35
+
36
+ ## Training and evaluation data
37
+
38
+ More information needed
39
+
40
+ ## Training procedure
41
+
42
+ ### Training hyperparameters
43
+
44
+ The following hyperparameters were used during training:
45
+ - optimizer: {'name': 'AdamWeightDecay', 'learning_rate': 8e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0}
46
+ - training_precision: float32
47
+
48
+ ### Training results
49
+
50
+ | Train Loss | Train Accuracy | Train Wermet | Train Wermet Syl | Validation Loss | Validation Accuracy | Validation Wermet | Validation Wermet Syl | Epoch |
51
+ |:----------:|:--------------:|:------------:|:----------------:|:---------------:|:-------------------:|:-----------------:|:---------------------:|:-----:|
52
+ | 5.2081 | 0.0105 | 1.5151 | 1.1936 | 4.0143 | 0.0113 | 0.9817 | 0.9765 | 0 |
53
+ | 4.7532 | 0.0116 | 0.8756 | 0.8371 | 3.9494 | 0.0114 | 0.9457 | 0.9171 | 1 |
54
+ | 4.7031 | 0.0117 | 0.8615 | 0.8175 | 3.9244 | 0.0114 | 0.9383 | 0.8993 | 2 |
55
+ | 4.6799 | 0.0117 | 0.8569 | 0.8117 | 3.9091 | 0.0114 | 0.9659 | 0.9487 | 3 |
56
+ | 4.6415 | 0.0117 | 0.8450 | 0.7985 | 3.8828 | 0.0115 | 0.9227 | 0.8867 | 4 |
57
+ | 4.6194 | 0.0118 | 0.8366 | 0.7901 | 3.8588 | 0.0115 | 0.9314 | 0.8973 | 5 |
58
+ | 4.5993 | 0.0118 | 0.8253 | 0.7803 | 3.9068 | 0.0116 | 0.9000 | 0.8526 | 6 |
59
+ | 4.5584 | 0.0120 | 0.7864 | 0.7455 | 3.7567 | 0.0118 | 0.8407 | 0.7907 | 7 |
60
+ | 4.4423 | 0.0123 | 0.7655 | 0.7209 | 3.5099 | 0.0123 | 0.8256 | 0.7786 | 8 |
61
+ | 4.0719 | 0.0133 | 0.7479 | 0.7112 | 2.9385 | 0.0135 | 0.7636 | 0.7193 | 9 |
62
+ | 3.5177 | 0.0148 | 0.7049 | 0.6917 | 2.4295 | 0.0148 | 0.6907 | 0.6559 | 10 |
63
+ | 2.8931 | 0.0170 | 0.6532 | 0.6928 | 2.0341 | 0.0161 | 0.6031 | 0.5778 | 11 |
64
+ | 2.4018 | 0.0190 | 0.6050 | 0.7024 | 1.7053 | 0.0174 | 0.5373 | 0.5223 | 12 |
65
+ | 1.9940 | 0.0210 | 0.5579 | 0.7168 | 1.5422 | 0.0180 | 0.4803 | 0.4793 | 13 |
66
+ | 1.7258 | 0.0223 | 0.5356 | 0.7426 | 1.4314 | 0.0186 | 0.4270 | 0.4029 | 14 |
67
+
68
+
69
+ ### Framework versions
70
+
71
+ - Transformers 4.34.0.dev0
72
+ - TensorFlow 2.13.0
73
+ - Tokenizers 0.13.3
config.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-tiny",
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
+ 50257
12
+ ],
13
+ "bos_token_id": 50257,
14
+ "classifier_proj_size": 256,
15
+ "d_model": 384,
16
+ "decoder_attention_heads": 6,
17
+ "decoder_ffn_dim": 1536,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 4,
20
+ "decoder_start_token_id": 50258,
21
+ "dropout": 0.0,
22
+ "encoder_attention_heads": 6,
23
+ "encoder_ffn_dim": 1536,
24
+ "encoder_layerdrop": 0.0,
25
+ "encoder_layers": 4,
26
+ "eos_token_id": 50257,
27
+ "forced_decoder_ids": [
28
+ [
29
+ 1,
30
+ 50289
31
+ ],
32
+ [
33
+ 2,
34
+ 50359
35
+ ],
36
+ [
37
+ 3,
38
+ 50363
39
+ ]
40
+ ],
41
+ "init_std": 0.02,
42
+ "is_encoder_decoder": true,
43
+ "mask_feature_length": 10,
44
+ "mask_feature_min_masks": 0,
45
+ "mask_feature_prob": 0.0,
46
+ "mask_time_length": 10,
47
+ "mask_time_min_masks": 2,
48
+ "mask_time_prob": 0.05,
49
+ "max_length": 448,
50
+ "max_source_positions": 1500,
51
+ "max_target_positions": 448,
52
+ "median_filter_width": 7,
53
+ "model_type": "whisper",
54
+ "num_hidden_layers": 4,
55
+ "num_mel_bins": 80,
56
+ "pad_token_id": -100,
57
+ "scale_embedding": false,
58
+ "suppress_tokens": [],
59
+ "torch_dtype": "float32",
60
+ "transformers_version": "4.34.0.dev0",
61
+ "use_cache": true,
62
+ "use_weighted_layer_sum": false,
63
+ "vocab_size": 51865
64
+ }
generation_config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "begin_suppress_tokens": [
4
+ 220,
5
+ 50257
6
+ ],
7
+ "bos_token_id": 50257,
8
+ "decoder_start_token_id": 50258,
9
+ "eos_token_id": 50257,
10
+ "forced_decoder_ids": [
11
+ [
12
+ 1,
13
+ 50259
14
+ ],
15
+ [
16
+ 2,
17
+ 50359
18
+ ],
19
+ [
20
+ 3,
21
+ 50363
22
+ ]
23
+ ],
24
+ "max_length": 448,
25
+ "pad_token_id": 50257,
26
+ "suppress_tokens": [
27
+ 1,
28
+ 2,
29
+ 7,
30
+ 8,
31
+ 9,
32
+ 10,
33
+ 14,
34
+ 25,
35
+ 26,
36
+ 27,
37
+ 28,
38
+ 29,
39
+ 31,
40
+ 58,
41
+ 59,
42
+ 60,
43
+ 61,
44
+ 62,
45
+ 63,
46
+ 90,
47
+ 91,
48
+ 92,
49
+ 93,
50
+ 359,
51
+ 503,
52
+ 522,
53
+ 542,
54
+ 873,
55
+ 893,
56
+ 902,
57
+ 918,
58
+ 922,
59
+ 931,
60
+ 1350,
61
+ 1853,
62
+ 1982,
63
+ 2460,
64
+ 2627,
65
+ 3246,
66
+ 3253,
67
+ 3268,
68
+ 3536,
69
+ 3846,
70
+ 3961,
71
+ 4183,
72
+ 4667,
73
+ 6585,
74
+ 6647,
75
+ 7273,
76
+ 9061,
77
+ 9383,
78
+ 10428,
79
+ 10929,
80
+ 11938,
81
+ 12033,
82
+ 12331,
83
+ 12562,
84
+ 13793,
85
+ 14157,
86
+ 14635,
87
+ 15265,
88
+ 15618,
89
+ 16553,
90
+ 16604,
91
+ 18362,
92
+ 18956,
93
+ 20075,
94
+ 21675,
95
+ 22520,
96
+ 26130,
97
+ 26161,
98
+ 26435,
99
+ 28279,
100
+ 29464,
101
+ 31650,
102
+ 32302,
103
+ 32470,
104
+ 36865,
105
+ 42863,
106
+ 47425,
107
+ 49870,
108
+ 50254,
109
+ 50258,
110
+ 50358,
111
+ 50359,
112
+ 50360,
113
+ 50361,
114
+ 50362
115
+ ],
116
+ "transformers_version": "4.34.0.dev0"
117
+ }
tf_model.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c53642871773e248e23612905fe619c5dbc7bc83d22872c55dbe997037e14cf
3
+ size 151253960