Hugging Face's logo

Team-ASR-2023
/
wav2vec2-large-xls-r-300m-Arabic

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
2
nrshoudi commited on
Commit
77aa507
•
1 Parent(s): df4c5fd

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +93 -0
README.md ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - wer
7
+ model-index:
8
+ - name: wav2vec2-large-xls-r-300m-Arabic-colab
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-Arabic-colab
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 None dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.0001
20
+ - Wer: 0.0813
21
+ - Cer: 0.0362
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0005
41
+ - train_batch_size: 16
42
+ - eval_batch_size: 6
43
+ - seed: 42
44
+ - gradient_accumulation_steps: 4
45
+ - total_train_batch_size: 64
46
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
+ - lr_scheduler_type: linear
48
+ - lr_scheduler_warmup_steps: 250
49
+ - num_epochs: 30.0
50
+ - mixed_precision_training: Native AMP
51
+
52
+ ### Training results
53
+
54
+ | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
55
+ |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
56
+ | 0.0166 | 1.0 | 51 | 0.0020 | 0.0944 | 0.0431 |
57
+ | 0.015 | 2.0 | 102 | 0.0018 | 0.0939 | 0.0434 |
58
+ | 0.0223 | 3.0 | 153 | 0.0034 | 0.0705 | 0.0311 |
59
+ | 0.0351 | 4.0 | 204 | 0.0089 | 0.1050 | 0.0414 |
60
+ | 0.0473 | 5.0 | 255 | 0.0051 | 0.1224 | 0.0614 |
61
+ | 0.0406 | 6.0 | 306 | 0.0084 | 0.1185 | 0.0547 |
62
+ | 0.0412 | 7.0 | 357 | 0.0030 | 0.0640 | 0.0254 |
63
+ | 0.0301 | 8.0 | 408 | 0.0157 | 0.0708 | 0.0219 |
64
+ | 0.0295 | 9.0 | 459 | 0.0027 | 0.0716 | 0.0298 |
65
+ | 0.0239 | 10.0 | 510 | 0.0077 | 0.0868 | 0.0354 |
66
+ | 0.0266 | 11.0 | 561 | 0.0017 | 0.0733 | 0.0301 |
67
+ | 0.0154 | 12.0 | 612 | 0.0015 | 0.0961 | 0.0385 |
68
+ | 0.0187 | 13.0 | 663 | 0.0006 | 0.1100 | 0.0465 |
69
+ | 0.0156 | 14.0 | 714 | 0.0015 | 0.1030 | 0.0426 |
70
+ | 0.013 | 15.0 | 765 | 0.0014 | 0.1068 | 0.0451 |
71
+ | 0.0136 | 16.0 | 816 | 0.0013 | 0.1066 | 0.0434 |
72
+ | 0.0123 | 17.0 | 867 | 0.0008 | 0.1240 | 0.0587 |
73
+ | 0.0098 | 18.0 | 918 | 0.0006 | 0.1140 | 0.0570 |
74
+ | 0.0108 | 19.0 | 969 | 0.0005 | 0.0843 | 0.0364 |
75
+ | 0.009 | 20.0 | 1020 | 0.0002 | 0.0954 | 0.0438 |
76
+ | 0.0083 | 21.0 | 1071 | 0.0003 | 0.0828 | 0.0377 |
77
+ | 0.0085 | 22.0 | 1122 | 0.0002 | 0.0648 | 0.0267 |
78
+ | 0.0073 | 23.0 | 1173 | 0.0003 | 0.0843 | 0.0373 |
79
+ | 0.0057 | 24.0 | 1224 | 0.0003 | 0.0822 | 0.0367 |
80
+ | 0.0039 | 25.0 | 1275 | 0.0002 | 0.0733 | 0.0329 |
81
+ | 0.0045 | 26.0 | 1326 | 0.0005 | 0.0754 | 0.0335 |
82
+ | 0.008 | 27.0 | 1377 | 0.0008 | 0.0803 | 0.0361 |
83
+ | 0.0045 | 28.0 | 1428 | 0.0001 | 0.0772 | 0.0340 |
84
+ | 0.0043 | 29.0 | 1479 | 0.0001 | 0.0808 | 0.0359 |
85
+ | 0.0042 | 30.0 | 1530 | 0.0001 | 0.0813 | 0.0362 |
86
+
87
+
88
+ ### Framework versions
89
+
90
+ - Transformers 4.26.1
91
+ - Pytorch 1.13.1+cu116
92
+ - Datasets 2.9.0
93
+ - Tokenizers 0.13.2