Hugging Face's logo

lgris
/
wav2vec2-xls-r-300m-gn-cv8-4

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Guaraní
wav2vec2
Generated from Trainer
robust-speech-event
hf-asr-leaderboard
Eval Results
Inference Endpoints
Model card Files Metrics Community
lgris commited on
Commit
359f2e4
•
1 Parent(s): ddbb544

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +105 -0
README.md ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ datasets:
6
+ - common_voice
7
+ model-index:
8
+ - name: wav2vec2-xls-r-300m-gn-cv8-4
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-xls-r-300m-gn-cv8-4
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: 1.5805
20
+ - Wer: 0.7545
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: 0.0001
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: 100
48
+ - training_steps: 13000
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
54
+ |:-------------:|:------:|:-----:|:---------------:|:------:|
55
+ | 9.2216 | 16.65 | 300 | 3.2771 | 1.0 |
56
+ | 3.1804 | 33.32 | 600 | 2.2869 | 1.0 |
57
+ | 1.5856 | 49.97 | 900 | 0.9573 | 0.8772 |
58
+ | 1.0299 | 66.65 | 1200 | 0.9044 | 0.8082 |
59
+ | 0.8916 | 83.32 | 1500 | 0.9478 | 0.8056 |
60
+ | 0.8451 | 99.97 | 1800 | 0.8814 | 0.8107 |
61
+ | 0.7649 | 116.65 | 2100 | 0.9897 | 0.7826 |
62
+ | 0.7185 | 133.32 | 2400 | 0.9988 | 0.7621 |
63
+ | 0.6595 | 149.97 | 2700 | 1.0607 | 0.7749 |
64
+ | 0.6211 | 166.65 | 3000 | 1.1826 | 0.7877 |
65
+ | 0.59 | 183.32 | 3300 | 1.1060 | 0.7826 |
66
+ | 0.5383 | 199.97 | 3600 | 1.1826 | 0.7852 |
67
+ | 0.5205 | 216.65 | 3900 | 1.2148 | 0.8261 |
68
+ | 0.4786 | 233.32 | 4200 | 1.2710 | 0.7928 |
69
+ | 0.4482 | 249.97 | 4500 | 1.1943 | 0.7980 |
70
+ | 0.4149 | 266.65 | 4800 | 1.2449 | 0.8031 |
71
+ | 0.3904 | 283.32 | 5100 | 1.3100 | 0.7928 |
72
+ | 0.3619 | 299.97 | 5400 | 1.3125 | 0.7596 |
73
+ | 0.3496 | 316.65 | 5700 | 1.3699 | 0.7877 |
74
+ | 0.3277 | 333.32 | 6000 | 1.4344 | 0.8031 |
75
+ | 0.2958 | 349.97 | 6300 | 1.4093 | 0.7980 |
76
+ | 0.2883 | 366.65 | 6600 | 1.3296 | 0.7570 |
77
+ | 0.2598 | 383.32 | 6900 | 1.4026 | 0.7980 |
78
+ | 0.2564 | 399.97 | 7200 | 1.4847 | 0.8031 |
79
+ | 0.2408 | 416.65 | 7500 | 1.4896 | 0.8107 |
80
+ | 0.2266 | 433.32 | 7800 | 1.4232 | 0.7698 |
81
+ | 0.224 | 449.97 | 8100 | 1.5560 | 0.7903 |
82
+ | 0.2038 | 466.65 | 8400 | 1.5355 | 0.7724 |
83
+ | 0.1948 | 483.32 | 8700 | 1.4624 | 0.7621 |
84
+ | 0.1995 | 499.97 | 9000 | 1.5808 | 0.7724 |
85
+ | 0.1864 | 516.65 | 9300 | 1.5653 | 0.7698 |
86
+ | 0.18 | 533.32 | 9600 | 1.4868 | 0.7494 |
87
+ | 0.1689 | 549.97 | 9900 | 1.5379 | 0.7749 |
88
+ | 0.1624 | 566.65 | 10200 | 1.5936 | 0.7749 |
89
+ | 0.1537 | 583.32 | 10500 | 1.6436 | 0.7801 |
90
+ | 0.1455 | 599.97 | 10800 | 1.6401 | 0.7673 |
91
+ | 0.1437 | 616.65 | 11100 | 1.6069 | 0.7673 |
92
+ | 0.1452 | 633.32 | 11400 | 1.6041 | 0.7519 |
93
+ | 0.139 | 649.97 | 11700 | 1.5758 | 0.7545 |
94
+ | 0.1299 | 666.65 | 12000 | 1.5559 | 0.7545 |
95
+ | 0.127 | 683.32 | 12300 | 1.5776 | 0.7596 |
96
+ | 0.1264 | 699.97 | 12600 | 1.5790 | 0.7519 |
97
+ | 0.1209 | 716.65 | 12900 | 1.5805 | 0.7545 |
98
+
99
+
100
+ ### Framework versions
101
+
102
+ - Transformers 4.16.1
103
+ - Pytorch 1.10.0+cu111
104
+ - Datasets 1.18.2
105
+ - Tokenizers 0.11.0