Hugging Face's logo

abdusah
/
aradia-ctc-distilhubert-ft

Automatic Speech Recognition
Transformers
PyTorch
hubert
abdusahmbzuai/arabic_speech_massive_sm
generated_from_trainer
Inference Endpoints
Model card Files Community
Abdulwahab Sahyoun commited on
Commit
38762f9
•
1 Parent(s): 29ca627

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +129 -0
README.md ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ model-index:
6
+ - name: aradia-ctc-distilhubert-ft
7
+ results: []
8
+ ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+ # aradia-ctc-distilhubert-ft
14
+
15
+ This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 2.7114
18
+ - Wer: 0.8908
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.0003
38
+ - train_batch_size: 32
39
+ - eval_batch_size: 32
40
+ - seed: 42
41
+ - gradient_accumulation_steps: 2
42
+ - total_train_batch_size: 64
43
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
+ - lr_scheduler_type: linear
45
+ - lr_scheduler_warmup_steps: 500
46
+ - num_epochs: 30.0
47
+ - mixed_precision_training: Native AMP
48
+
49
+ ### Training results
50
+
51
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
52
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
53
+ | No log | 0.43 | 100 | 4.4129 | 1.0 |
54
+ | No log | 0.87 | 200 | 3.5927 | 1.0 |
55
+ | No log | 1.3 | 300 | 3.3780 | 1.0 |
56
+ | No log | 1.74 | 400 | 3.0830 | 1.0 |
57
+ | 5.3551 | 2.17 | 500 | 2.6278 | 0.9999 |
58
+ | 5.3551 | 2.61 | 600 | 1.8359 | 1.0000 |
59
+ | 5.3551 | 3.04 | 700 | 1.7878 | 0.9914 |
60
+ | 5.3551 | 3.48 | 800 | 1.5219 | 0.9875 |
61
+ | 5.3551 | 3.91 | 900 | 1.4348 | 0.9879 |
62
+ | 1.7199 | 4.35 | 1000 | 1.4354 | 0.9644 |
63
+ | 1.7199 | 4.78 | 1100 | 1.5210 | 0.9519 |
64
+ | 1.7199 | 5.22 | 1200 | 1.3607 | 0.9475 |
65
+ | 1.7199 | 5.65 | 1300 | 1.3839 | 0.9343 |
66
+ | 1.7199 | 6.09 | 1400 | 1.2806 | 0.8944 |
67
+ | 1.2342 | 6.52 | 1500 | 1.3036 | 0.9011 |
68
+ | 1.2342 | 6.95 | 1600 | 1.3704 | 0.9072 |
69
+ | 1.2342 | 7.39 | 1700 | 1.2981 | 0.8891 |
70
+ | 1.2342 | 7.82 | 1800 | 1.2786 | 0.8733 |
71
+ | 1.2342 | 8.26 | 1900 | 1.2897 | 0.8867 |
72
+ | 0.9831 | 8.69 | 2000 | 1.4436 | 0.8780 |
73
+ | 0.9831 | 9.13 | 2100 | 1.3680 | 0.8873 |
74
+ | 0.9831 | 9.56 | 2200 | 1.3471 | 0.8692 |
75
+ | 0.9831 | 10.0 | 2300 | 1.3725 | 0.8729 |
76
+ | 0.9831 | 10.43 | 2400 | 1.4439 | 0.8771 |
77
+ | 0.8071 | 10.87 | 2500 | 1.5114 | 0.8928 |
78
+ | 0.8071 | 11.3 | 2600 | 1.6156 | 0.8958 |
79
+ | 0.8071 | 11.74 | 2700 | 1.4381 | 0.8749 |
80
+ | 0.8071 | 12.17 | 2800 | 1.5088 | 0.8717 |
81
+ | 0.8071 | 12.61 | 2900 | 1.5486 | 0.8813 |
82
+ | 0.6321 | 13.04 | 3000 | 1.4536 | 0.8884 |
83
+ | 0.6321 | 13.48 | 3100 | 1.4679 | 0.8947 |
84
+ | 0.6321 | 13.91 | 3200 | 1.5628 | 0.9117 |
85
+ | 0.6321 | 14.35 | 3300 | 1.5831 | 0.8716 |
86
+ | 0.6321 | 14.78 | 3400 | 1.6733 | 0.8702 |
87
+ | 0.4998 | 15.22 | 3500 | 1.8225 | 0.8665 |
88
+ | 0.4998 | 15.65 | 3600 | 1.8558 | 0.8732 |
89
+ | 0.4998 | 16.09 | 3700 | 1.7513 | 0.8766 |
90
+ | 0.4998 | 16.52 | 3800 | 1.8562 | 0.8753 |
91
+ | 0.4998 | 16.95 | 3900 | 1.9018 | 0.8704 |
92
+ | 0.4421 | 17.39 | 4000 | 1.9341 | 0.8789 |
93
+ | 0.4421 | 17.82 | 4100 | 1.9582 | 0.8781 |
94
+ | 0.4421 | 18.26 | 4200 | 1.8863 | 0.8821 |
95
+ | 0.4421 | 18.69 | 4300 | 1.9366 | 0.8847 |
96
+ | 0.4421 | 19.13 | 4400 | 2.1902 | 0.8721 |
97
+ | 0.3712 | 19.56 | 4500 | 2.1641 | 0.8670 |
98
+ | 0.3712 | 20.0 | 4600 | 2.1639 | 0.8776 |
99
+ | 0.3712 | 20.43 | 4700 | 2.2695 | 0.9030 |
100
+ | 0.3712 | 20.87 | 4800 | 2.1909 | 0.8937 |
101
+ | 0.3712 | 21.3 | 4900 | 2.1606 | 0.8959 |
102
+ | 0.3067 | 21.74 | 5000 | 2.1756 | 0.8943 |
103
+ | 0.3067 | 22.17 | 5100 | 2.4092 | 0.8773 |
104
+ | 0.3067 | 22.61 | 5200 | 2.4991 | 0.8721 |
105
+ | 0.3067 | 23.04 | 5300 | 2.3340 | 0.8910 |
106
+ | 0.3067 | 23.48 | 5400 | 2.3567 | 0.8946 |
107
+ | 0.2764 | 23.91 | 5500 | 2.3215 | 0.8897 |
108
+ | 0.2764 | 24.35 | 5600 | 2.4824 | 0.9002 |
109
+ | 0.2764 | 24.78 | 5700 | 2.4585 | 0.8963 |
110
+ | 0.2764 | 25.22 | 5800 | 2.5804 | 0.8879 |
111
+ | 0.2764 | 25.65 | 5900 | 2.5814 | 0.8903 |
112
+ | 0.2593 | 26.09 | 6000 | 2.5374 | 0.8868 |
113
+ | 0.2593 | 26.52 | 6100 | 2.5346 | 0.8922 |
114
+ | 0.2593 | 26.95 | 6200 | 2.5465 | 0.8873 |
115
+ | 0.2593 | 27.39 | 6300 | 2.6002 | 0.8919 |
116
+ | 0.2593 | 27.82 | 6400 | 2.6102 | 0.8928 |
117
+ | 0.227 | 28.26 | 6500 | 2.6925 | 0.8914 |
118
+ | 0.227 | 28.69 | 6600 | 2.6981 | 0.8913 |
119
+ | 0.227 | 29.13 | 6700 | 2.6872 | 0.8891 |
120
+ | 0.227 | 29.56 | 6800 | 2.7015 | 0.8897 |
121
+ | 0.227 | 30.0 | 6900 | 2.7114 | 0.8908 |
122
+
123
+
124
+ ### Framework versions
125
+
126
+ - Transformers 4.18.0.dev0
127
+ - Pytorch 1.10.2+cu113
128
+ - Datasets 1.18.4
129
+ - Tokenizers 0.11.6