Dev372 commited on
Commit
1cc38b5
1 Parent(s): 2911cb9

End of training

Browse files
Files changed (2) hide show
  1. README.md +87 -0
  2. generation_config.json +137 -0
README.md ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ base_model: openai/whisper-tiny.en
6
+ tags:
7
+ - generated_from_trainer
8
+ datasets:
9
+ - Dev372/Medical_STT_Dataset_1.1
10
+ metrics:
11
+ - wer
12
+ model-index:
13
+ - name: English Whisper Model
14
+ results:
15
+ - task:
16
+ name: Automatic Speech Recognition
17
+ type: automatic-speech-recognition
18
+ dataset:
19
+ name: Medical
20
+ type: Dev372/Medical_STT_Dataset_1.1
21
+ args: 'split: test'
22
+ metrics:
23
+ - name: Wer
24
+ type: wer
25
+ value: 6.286946013912929
26
+ ---
27
+
28
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
+ should probably proofread and complete it, then remove this comment. -->
30
+
31
+ # English Whisper Model
32
+
33
+ This model is a fine-tuned version of [openai/whisper-tiny.en](https://huggingface.co/openai/whisper-tiny.en) on the Medical dataset.
34
+ It achieves the following results on the evaluation set:
35
+ - Loss: 0.1269
36
+ - Wer: 6.2869
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 1e-05
56
+ - train_batch_size: 18
57
+ - eval_batch_size: 8
58
+ - seed: 42
59
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
+ - lr_scheduler_type: linear
61
+ - lr_scheduler_warmup_steps: 500
62
+ - training_steps: 1100
63
+ - mixed_precision_training: Native AMP
64
+
65
+ ### Training results
66
+
67
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
68
+ |:-------------:|:------:|:----:|:---------------:|:-------:|
69
+ | 1.2361 | 0.2825 | 100 | 1.0425 | 10.4870 |
70
+ | 0.6631 | 0.5650 | 200 | 0.6451 | 9.4908 |
71
+ | 0.419 | 0.8475 | 300 | 0.3854 | 8.5535 |
72
+ | 0.1538 | 1.1299 | 400 | 0.1895 | 7.2635 |
73
+ | 0.1234 | 1.4124 | 500 | 0.1644 | 6.8454 |
74
+ | 0.1134 | 1.6949 | 600 | 0.1470 | 6.6201 |
75
+ | 0.1071 | 1.9774 | 700 | 0.1358 | 6.0289 |
76
+ | 0.0721 | 2.2599 | 800 | 0.1329 | 6.1302 |
77
+ | 0.0693 | 2.5424 | 900 | 0.1299 | 6.3065 |
78
+ | 0.0635 | 2.8249 | 1000 | 0.1275 | 6.5025 |
79
+ | 0.0441 | 3.1073 | 1100 | 0.1269 | 6.2869 |
80
+
81
+
82
+ ### Framework versions
83
+
84
+ - Transformers 4.43.3
85
+ - Pytorch 2.1.2
86
+ - Datasets 2.20.0
87
+ - Tokenizers 0.19.1
generation_config.json ADDED
@@ -0,0 +1,137 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 3,
5
+ 3
6
+ ],
7
+ [
8
+ 4,
9
+ 7
10
+ ],
11
+ [
12
+ 5,
13
+ 1
14
+ ],
15
+ [
16
+ 5,
17
+ 5
18
+ ],
19
+ [
20
+ 5,
21
+ 7
22
+ ]
23
+ ],
24
+ "begin_suppress_tokens": [
25
+ 220,
26
+ 50256
27
+ ],
28
+ "bos_token_id": 50257,
29
+ "decoder_start_token_id": 50257,
30
+ "eos_token_id": 50256,
31
+ "forced_decoder_ids": [
32
+ [
33
+ 1,
34
+ 50362
35
+ ]
36
+ ],
37
+ "is_multilingual": false,
38
+ "max_initial_timestamp_index": 50,
39
+ "max_length": 448,
40
+ "no_timestamps_token_id": 50362,
41
+ "pad_token_id": 50256,
42
+ "prev_sot_token_id": 50360,
43
+ "return_timestamps": false,
44
+ "suppress_tokens": [
45
+ 1,
46
+ 2,
47
+ 7,
48
+ 8,
49
+ 9,
50
+ 10,
51
+ 14,
52
+ 25,
53
+ 26,
54
+ 27,
55
+ 28,
56
+ 29,
57
+ 31,
58
+ 58,
59
+ 59,
60
+ 60,
61
+ 61,
62
+ 62,
63
+ 63,
64
+ 90,
65
+ 91,
66
+ 92,
67
+ 93,
68
+ 357,
69
+ 366,
70
+ 438,
71
+ 532,
72
+ 685,
73
+ 705,
74
+ 796,
75
+ 930,
76
+ 1058,
77
+ 1220,
78
+ 1267,
79
+ 1279,
80
+ 1303,
81
+ 1343,
82
+ 1377,
83
+ 1391,
84
+ 1635,
85
+ 1782,
86
+ 1875,
87
+ 2162,
88
+ 2361,
89
+ 2488,
90
+ 3467,
91
+ 4008,
92
+ 4211,
93
+ 4600,
94
+ 4808,
95
+ 5299,
96
+ 5855,
97
+ 6329,
98
+ 7203,
99
+ 9609,
100
+ 9959,
101
+ 10563,
102
+ 10786,
103
+ 11420,
104
+ 11709,
105
+ 11907,
106
+ 13163,
107
+ 13697,
108
+ 13700,
109
+ 14808,
110
+ 15306,
111
+ 16410,
112
+ 16791,
113
+ 17992,
114
+ 19203,
115
+ 19510,
116
+ 20724,
117
+ 22305,
118
+ 22935,
119
+ 27007,
120
+ 30109,
121
+ 30420,
122
+ 33409,
123
+ 34949,
124
+ 40283,
125
+ 40493,
126
+ 40549,
127
+ 47282,
128
+ 49146,
129
+ 50257,
130
+ 50357,
131
+ 50358,
132
+ 50359,
133
+ 50360,
134
+ 50361
135
+ ],
136
+ "transformers_version": "4.43.3"
137
+ }