Hugging Face's logo

tonyalves
/
output

Automatic Speech Recognition
Transformers
PyTorch
Portuguese
wav2vec2
mozilla-foundation/common_voice_8_0
Generated from Trainer
Inference Endpoints
Model card Files Community
1
tonyalves commited on
Commit
c7a89de
·
1 Parent(s): 412e7c9

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +139 -0
README.md ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - common_voice
7
+ model-index:
8
+ - name: output
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # output
16
+
17
+ This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.1506
20
+ - Wer: 0.1353
21
+
22
+ ## Model description
23
+
24
+ More information needed
25
+
26
+ ## Intended uses & limitations
27
+
28
+ More information needed
29
+
30
+ ## Training and evaluation data
31
+
32
+ More information needed
33
+
34
+ ## Training procedure
35
+
36
+ ### Training hyperparameters
37
+
38
+ The following hyperparameters were used during training:
39
+ - learning_rate: 7.5e-05
40
+ - train_batch_size: 8
41
+ - eval_batch_size: 8
42
+ - seed: 42
43
+ - gradient_accumulation_steps: 4
44
+ - total_train_batch_size: 32
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 2000
48
+ - num_epochs: 50.0
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
54
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
55
+ | 4.1367 | 0.64 | 500 | 3.8825 | 1.0 |
56
+ | 2.9677 | 1.29 | 1000 | 2.9498 | 1.0 |
57
+ | 1.5884 | 1.93 | 1500 | 0.6722 | 0.6493 |
58
+ | 1.2292 | 2.57 | 2000 | 0.3635 | 0.3202 |
59
+ | 1.1314 | 3.22 | 2500 | 0.2970 | 0.2680 |
60
+ | 1.0879 | 3.86 | 3000 | 0.2671 | 0.2486 |
61
+ | 1.0344 | 4.5 | 3500 | 0.2625 | 0.2239 |
62
+ | 1.0109 | 5.15 | 4000 | 0.2520 | 0.2230 |
63
+ | 0.9966 | 5.79 | 4500 | 0.2280 | 0.2105 |
64
+ | 0.9815 | 6.43 | 5000 | 0.2254 | 0.2179 |
65
+ | 0.9744 | 7.08 | 5500 | 0.2301 | 0.2137 |
66
+ | 0.9487 | 7.72 | 6000 | 0.2224 | 0.2051 |
67
+ | 0.9431 | 8.37 | 6500 | 0.2105 | 0.1992 |
68
+ | 0.9365 | 9.01 | 7000 | 0.2114 | 0.2019 |
69
+ | 0.9268 | 9.65 | 7500 | 0.2097 | 0.1988 |
70
+ | 0.9292 | 10.3 | 8000 | 0.2120 | 0.1986 |
71
+ | 0.929 | 10.94 | 8500 | 0.2048 | 0.1998 |
72
+ | 0.9017 | 11.58 | 9000 | 0.2035 | 0.1999 |
73
+ | 0.8898 | 12.23 | 9500 | 0.1961 | 0.1908 |
74
+ | 0.8799 | 12.87 | 10000 | 0.1945 | 0.1817 |
75
+ | 0.869 | 13.51 | 10500 | 0.1929 | 0.1844 |
76
+ | 0.8572 | 14.16 | 11000 | 0.1941 | 0.1888 |
77
+ | 0.8691 | 14.8 | 11500 | 0.1912 | 0.1804 |
78
+ | 0.8645 | 15.44 | 12000 | 0.1950 | 0.1851 |
79
+ | 0.8468 | 16.09 | 12500 | 0.1879 | 0.1770 |
80
+ | 0.8405 | 16.73 | 13000 | 0.1881 | 0.1759 |
81
+ | 0.8647 | 17.37 | 13500 | 0.1861 | 0.1740 |
82
+ | 0.8477 | 18.02 | 14000 | 0.1782 | 0.1702 |
83
+ | 0.811 | 18.66 | 14500 | 0.1915 | 0.1757 |
84
+ | 0.8165 | 19.3 | 15000 | 0.1820 | 0.1724 |
85
+ | 0.8166 | 19.95 | 15500 | 0.1798 | 0.1697 |
86
+ | 0.8167 | 20.59 | 16000 | 0.1805 | 0.1752 |
87
+ | 0.7908 | 21.24 | 16500 | 0.1761 | 0.1699 |
88
+ | 0.7925 | 21.88 | 17000 | 0.1740 | 0.1709 |
89
+ | 0.7803 | 22.52 | 17500 | 0.1815 | 0.1727 |
90
+ | 0.7839 | 23.17 | 18000 | 0.1737 | 0.1694 |
91
+ | 0.7815 | 23.81 | 18500 | 0.1732 | 0.1630 |
92
+ | 0.767 | 24.45 | 19000 | 0.1724 | 0.1648 |
93
+ | 0.7672 | 25.1 | 19500 | 0.1706 | 0.1596 |
94
+ | 0.7691 | 25.74 | 20000 | 0.1718 | 0.1618 |
95
+ | 0.7547 | 26.38 | 20500 | 0.1694 | 0.1565 |
96
+ | 0.7498 | 27.03 | 21000 | 0.1706 | 0.1582 |
97
+ | 0.7459 | 27.67 | 21500 | 0.1663 | 0.1586 |
98
+ | 0.7374 | 28.31 | 22000 | 0.1651 | 0.1567 |
99
+ | 0.7499 | 28.96 | 22500 | 0.1668 | 0.1549 |
100
+ | 0.7471 | 29.6 | 23000 | 0.1667 | 0.1553 |
101
+ | 0.7369 | 30.24 | 23500 | 0.1659 | 0.1556 |
102
+ | 0.7389 | 30.89 | 24000 | 0.1668 | 0.1538 |
103
+ | 0.7197 | 31.53 | 24500 | 0.1687 | 0.1561 |
104
+ | 0.71 | 32.17 | 25000 | 0.1666 | 0.1516 |
105
+ | 0.7199 | 32.82 | 25500 | 0.1640 | 0.1523 |
106
+ | 0.7194 | 33.46 | 26000 | 0.1659 | 0.1528 |
107
+ | 0.6923 | 34.11 | 26500 | 0.1662 | 0.1507 |
108
+ | 0.7054 | 34.75 | 27000 | 0.1641 | 0.1486 |
109
+ | 0.6955 | 35.39 | 27500 | 0.1634 | 0.1497 |
110
+ | 0.7084 | 36.04 | 28000 | 0.1618 | 0.1478 |
111
+ | 0.6917 | 36.68 | 28500 | 0.1589 | 0.1471 |
112
+ | 0.687 | 37.32 | 29000 | 0.1589 | 0.1450 |
113
+ | 0.6914 | 37.97 | 29500 | 0.1588 | 0.1465 |
114
+ | 0.6646 | 38.61 | 30000 | 0.1602 | 0.1468 |
115
+ | 0.6667 | 39.25 | 30500 | 0.1588 | 0.1444 |
116
+ | 0.6754 | 39.9 | 31000 | 0.1587 | 0.1455 |
117
+ | 0.6632 | 40.54 | 31500 | 0.1586 | 0.1461 |
118
+ | 0.6619 | 41.18 | 32000 | 0.1571 | 0.1441 |
119
+ | 0.6561 | 41.83 | 32500 | 0.1564 | 0.1420 |
120
+ | 0.6492 | 42.47 | 33000 | 0.1539 | 0.1437 |
121
+ | 0.6649 | 43.11 | 33500 | 0.1512 | 0.1406 |
122
+ | 0.6511 | 43.76 | 34000 | 0.1539 | 0.1384 |
123
+ | 0.6551 | 44.4 | 34500 | 0.1520 | 0.1384 |
124
+ | 0.6452 | 45.05 | 35000 | 0.1510 | 0.1368 |
125
+ | 0.6155 | 45.69 | 35500 | 0.1522 | 0.1375 |
126
+ | 0.628 | 46.33 | 36000 | 0.1522 | 0.1366 |
127
+ | 0.6389 | 46.97 | 36500 | 0.1513 | 0.1377 |
128
+ | 0.6265 | 47.62 | 37000 | 0.1512 | 0.1369 |
129
+ | 0.6197 | 48.26 | 37500 | 0.1511 | 0.1362 |
130
+ | 0.621 | 48.91 | 38000 | 0.1510 | 0.1357 |
131
+ | 0.6259 | 49.55 | 38500 | 0.1506 | 0.1353 |
132
+
133
+
134
+ ### Framework versions
135
+
136
+ - Transformers 4.17.0
137
+ - Pytorch 1.9.1+cu102
138
+ - Datasets 2.0.0
139
+ - Tokenizers 0.11.6