Hugging Face's logo

tomascufaro
/
wav2vec2-large-xls-r-300m-spanish-custom

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
generated_from_trainer
Inference Endpoints
Model card Files Community
1
tomascufaro commited on
Commit
1acdc92
•
1 Parent(s): a5b5c14

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +137 -0
README.md ADDED
@@ -0,0 +1,137 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - common_voice
7
+ model-index:
8
+ - name: wav2vec2-large-xls-r-300m-spanish-custom
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
+ # wav2vec2-large-xls-r-300m-spanish-custom
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: 0.4426
20
+ - Wer: 0.2117
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: 0.0003
40
+ - train_batch_size: 8
41
+ - eval_batch_size: 8
42
+ - seed: 42
43
+ - gradient_accumulation_steps: 2
44
+ - total_train_batch_size: 16
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 500
48
+ - num_epochs: 30
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
54
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
55
+ | 4.2307 | 0.4 | 400 | 1.4431 | 0.9299 |
56
+ | 0.7066 | 0.79 | 800 | 0.5928 | 0.4836 |
57
+ | 0.4397 | 1.19 | 1200 | 0.4341 | 0.3730 |
58
+ | 0.3889 | 1.58 | 1600 | 0.4063 | 0.3499 |
59
+ | 0.3607 | 1.98 | 2000 | 0.3834 | 0.3235 |
60
+ | 0.2866 | 2.37 | 2400 | 0.3885 | 0.3163 |
61
+ | 0.2833 | 2.77 | 2800 | 0.3765 | 0.3140 |
62
+ | 0.2692 | 3.17 | 3200 | 0.3849 | 0.3132 |
63
+ | 0.2435 | 3.56 | 3600 | 0.3779 | 0.2984 |
64
+ | 0.2404 | 3.96 | 4000 | 0.3756 | 0.2934 |
65
+ | 0.2153 | 4.35 | 4400 | 0.3770 | 0.3075 |
66
+ | 0.2087 | 4.75 | 4800 | 0.3819 | 0.3022 |
67
+ | 0.1999 | 5.14 | 5200 | 0.3756 | 0.2959 |
68
+ | 0.1838 | 5.54 | 5600 | 0.3827 | 0.2858 |
69
+ | 0.1892 | 5.93 | 6000 | 0.3714 | 0.2999 |
70
+ | 0.1655 | 6.33 | 6400 | 0.3814 | 0.2812 |
71
+ | 0.1649 | 6.73 | 6800 | 0.3685 | 0.2727 |
72
+ | 0.1668 | 7.12 | 7200 | 0.3832 | 0.2825 |
73
+ | 0.1487 | 7.52 | 7600 | 0.3848 | 0.2788 |
74
+ | 0.152 | 7.91 | 8000 | 0.3810 | 0.2787 |
75
+ | 0.143 | 8.31 | 8400 | 0.3885 | 0.2856 |
76
+ | 0.1353 | 8.7 | 8800 | 0.4103 | 0.2827 |
77
+ | 0.1386 | 9.1 | 9200 | 0.4142 | 0.2874 |
78
+ | 0.1222 | 9.5 | 9600 | 0.3983 | 0.2830 |
79
+ | 0.1288 | 9.89 | 10000 | 0.4179 | 0.2781 |
80
+ | 0.1199 | 10.29 | 10400 | 0.4035 | 0.2789 |
81
+ | 0.1196 | 10.68 | 10800 | 0.4043 | 0.2746 |
82
+ | 0.1169 | 11.08 | 11200 | 0.4105 | 0.2753 |
83
+ | 0.1076 | 11.47 | 11600 | 0.4298 | 0.2686 |
84
+ | 0.1124 | 11.87 | 12000 | 0.4025 | 0.2704 |
85
+ | 0.1043 | 12.26 | 12400 | 0.4209 | 0.2659 |
86
+ | 0.0976 | 12.66 | 12800 | 0.4070 | 0.2672 |
87
+ | 0.1012 | 13.06 | 13200 | 0.4161 | 0.2720 |
88
+ | 0.0872 | 13.45 | 13600 | 0.4245 | 0.2697 |
89
+ | 0.0933 | 13.85 | 14000 | 0.4295 | 0.2684 |
90
+ | 0.0881 | 14.24 | 14400 | 0.4011 | 0.2650 |
91
+ | 0.0848 | 14.64 | 14800 | 0.3991 | 0.2675 |
92
+ | 0.0852 | 15.03 | 15200 | 0.4166 | 0.2617 |
93
+ | 0.0825 | 15.43 | 15600 | 0.4188 | 0.2639 |
94
+ | 0.081 | 15.83 | 16000 | 0.4181 | 0.2547 |
95
+ | 0.0753 | 16.22 | 16400 | 0.4103 | 0.2560 |
96
+ | 0.0747 | 16.62 | 16800 | 0.4017 | 0.2498 |
97
+ | 0.0761 | 17.01 | 17200 | 0.4159 | 0.2563 |
98
+ | 0.0711 | 17.41 | 17600 | 0.4112 | 0.2603 |
99
+ | 0.0698 | 17.8 | 18000 | 0.4335 | 0.2529 |
100
+ | 0.073 | 18.2 | 18400 | 0.4120 | 0.2512 |
101
+ | 0.0665 | 18.6 | 18800 | 0.4335 | 0.2496 |
102
+ | 0.0657 | 18.99 | 19200 | 0.4143 | 0.2468 |
103
+ | 0.0617 | 19.39 | 19600 | 0.4339 | 0.2435 |
104
+ | 0.06 | 19.78 | 20000 | 0.4179 | 0.2438 |
105
+ | 0.0613 | 20.18 | 20400 | 0.4251 | 0.2393 |
106
+ | 0.0583 | 20.57 | 20800 | 0.4347 | 0.2422 |
107
+ | 0.0562 | 20.97 | 21200 | 0.4246 | 0.2377 |
108
+ | 0.053 | 21.36 | 21600 | 0.4198 | 0.2338 |
109
+ | 0.0525 | 21.76 | 22000 | 0.4511 | 0.2427 |
110
+ | 0.0499 | 22.16 | 22400 | 0.4482 | 0.2353 |
111
+ | 0.0475 | 22.55 | 22800 | 0.4449 | 0.2329 |
112
+ | 0.0465 | 22.95 | 23200 | 0.4364 | 0.2320 |
113
+ | 0.0443 | 23.34 | 23600 | 0.4481 | 0.2304 |
114
+ | 0.0458 | 23.74 | 24000 | 0.4442 | 0.2267 |
115
+ | 0.0453 | 24.13 | 24400 | 0.4402 | 0.2261 |
116
+ | 0.0426 | 24.53 | 24800 | 0.4262 | 0.2232 |
117
+ | 0.0431 | 24.93 | 25200 | 0.4251 | 0.2210 |
118
+ | 0.0389 | 25.32 | 25600 | 0.4455 | 0.2232 |
119
+ | 0.039 | 25.72 | 26000 | 0.4372 | 0.2236 |
120
+ | 0.0378 | 26.11 | 26400 | 0.4236 | 0.2212 |
121
+ | 0.0348 | 26.51 | 26800 | 0.4359 | 0.2204 |
122
+ | 0.0361 | 26.9 | 27200 | 0.4248 | 0.2192 |
123
+ | 0.0356 | 27.3 | 27600 | 0.4397 | 0.2184 |
124
+ | 0.0325 | 27.7 | 28000 | 0.4367 | 0.2181 |
125
+ | 0.0313 | 28.09 | 28400 | 0.4477 | 0.2136 |
126
+ | 0.0306 | 28.49 | 28800 | 0.4533 | 0.2135 |
127
+ | 0.0314 | 28.88 | 29200 | 0.4410 | 0.2136 |
128
+ | 0.0307 | 29.28 | 29600 | 0.4457 | 0.2113 |
129
+ | 0.0309 | 29.67 | 30000 | 0.4426 | 0.2117 |
130
+
131
+
132
+ ### Framework versions
133
+
134
+ - Transformers 4.16.0.dev0
135
+ - Pytorch 1.10.1+cu102
136
+ - Datasets 1.17.1.dev0
137
+ - Tokenizers 0.11.0