ptah23 commited on
Commit
2d24e43
1 Parent(s): d7b2f4a

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +96 -0
README.md ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: openai/whisper-small
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - google/fleurs
8
+ metrics:
9
+ - wer
10
+ model-index:
11
+ - name: whisper-small-af-ZA
12
+ results:
13
+ - task:
14
+ name: Automatic Speech Recognition
15
+ type: automatic-speech-recognition
16
+ dataset:
17
+ name: google/fleurs
18
+ type: google/fleurs
19
+ config: af_za
20
+ split: train+validation
21
+ args: af_za
22
+ metrics:
23
+ - name: Wer
24
+ type: wer
25
+ value: 0.36644093303235514
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
+ # whisper-small-af-ZA
32
+
33
+ This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the google/fleurs dataset.
34
+ It achieves the following results on the evaluation set:
35
+ - Loss: 0.5728
36
+ - Wer Ortho: 0.3943
37
+ - Wer: 0.3664
38
+
39
+ ## Model description
40
+
41
+ More information needed
42
+
43
+ ## Intended uses & limitations
44
+
45
+ More information needed
46
+
47
+ ## Training and evaluation data
48
+
49
+ More information needed
50
+
51
+ ## Training procedure
52
+
53
+ ### Training hyperparameters
54
+
55
+ The following hyperparameters were used during training:
56
+ - learning_rate: 1e-05
57
+ - train_batch_size: 16
58
+ - eval_batch_size: 16
59
+ - seed: 42
60
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
61
+ - lr_scheduler_type: constant_with_warmup
62
+ - lr_scheduler_warmup_steps: 5
63
+ - training_steps: 2000
64
+
65
+ ### Training results
66
+
67
+ | Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
68
+ |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|
69
+ | 0.7731 | 1.45 | 100 | 0.7280 | 0.3863 | 0.3740 |
70
+ | 0.2103 | 2.9 | 200 | 0.5116 | 0.3859 | 0.3661 |
71
+ | 0.0633 | 4.35 | 300 | 0.4967 | 0.3008 | 0.2810 |
72
+ | 0.0249 | 5.8 | 400 | 0.5003 | 0.3477 | 0.3299 |
73
+ | 0.0143 | 7.25 | 500 | 0.5191 | 0.3660 | 0.3510 |
74
+ | 0.0053 | 8.7 | 600 | 0.5149 | 0.3221 | 0.3070 |
75
+ | 0.0035 | 10.14 | 700 | 0.5345 | 0.3443 | 0.3266 |
76
+ | 0.0027 | 11.59 | 800 | 0.5339 | 0.3344 | 0.3175 |
77
+ | 0.0026 | 13.04 | 900 | 0.5435 | 0.3328 | 0.3134 |
78
+ | 0.0037 | 14.49 | 1000 | 0.5346 | 0.2714 | 0.2506 |
79
+ | 0.0045 | 15.94 | 1100 | 0.5438 | 0.3389 | 0.3220 |
80
+ | 0.0028 | 17.39 | 1200 | 0.5588 | 0.2740 | 0.2551 |
81
+ | 0.0036 | 18.84 | 1300 | 0.5466 | 0.2702 | 0.2728 |
82
+ | 0.0035 | 20.29 | 1400 | 0.5364 | 0.3332 | 0.3119 |
83
+ | 0.0056 | 21.74 | 1500 | 0.5608 | 0.2721 | 0.2506 |
84
+ | 0.0037 | 23.19 | 1600 | 0.5443 | 0.3027 | 0.2833 |
85
+ | 0.0035 | 24.64 | 1700 | 0.5466 | 0.3866 | 0.3631 |
86
+ | 0.0024 | 26.09 | 1800 | 0.5628 | 0.3416 | 0.3198 |
87
+ | 0.0036 | 27.54 | 1900 | 0.5495 | 0.3122 | 0.2946 |
88
+ | 0.0016 | 28.99 | 2000 | 0.5728 | 0.3943 | 0.3664 |
89
+
90
+
91
+ ### Framework versions
92
+
93
+ - Transformers 4.31.0.dev0
94
+ - Pytorch 1.12.1+cu116
95
+ - Datasets 2.4.0
96
+ - Tokenizers 0.12.1