anuragshas commited on
Commit
7afc95e
1 Parent(s): 6366826

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +102 -0
README.md ADDED
@@ -0,0 +1,102 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - hi
4
+ license: apache-2.0
5
+ tags:
6
+ - automatic-speech-recognition
7
+ - mozilla-foundation/common_voice_7_0
8
+ - generated_from_trainer
9
+ datasets:
10
+ - common_voice
11
+ model-index:
12
+ - name: ''
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ #
20
+
21
+ This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - HI dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 0.5878
24
+ - Wer: 0.3419
25
+
26
+ ## Model description
27
+
28
+ More information needed
29
+
30
+ ## Intended uses & limitations
31
+
32
+ More information needed
33
+
34
+ ## Training and evaluation data
35
+
36
+ More information needed
37
+
38
+ ## Training procedure
39
+
40
+ ### Training hyperparameters
41
+
42
+ The following hyperparameters were used during training:
43
+ - learning_rate: 7.5e-05
44
+ - train_batch_size: 8
45
+ - eval_batch_size: 16
46
+ - seed: 42
47
+ - gradient_accumulation_steps: 4
48
+ - total_train_batch_size: 32
49
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
+ - lr_scheduler_type: linear
51
+ - lr_scheduler_warmup_steps: 2000
52
+ - num_epochs: 100.0
53
+ - mixed_precision_training: Native AMP
54
+
55
+ ### Training results
56
+
57
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
58
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
59
+ | 1.9859 | 2.72 | 400 | 1.1663 | 0.7948 |
60
+ | 1.2969 | 5.44 | 800 | 0.7725 | 0.6562 |
61
+ | 1.1954 | 8.16 | 1200 | 0.5940 | 0.4904 |
62
+ | 1.164 | 10.88 | 1600 | 0.5338 | 0.4316 |
63
+ | 1.1464 | 13.6 | 2000 | 0.5432 | 0.4226 |
64
+ | 1.1553 | 16.33 | 2400 | 0.5471 | 0.4260 |
65
+ | 1.0985 | 19.05 | 2800 | 0.5290 | 0.4076 |
66
+ | 1.0421 | 21.77 | 3200 | 0.5672 | 0.4181 |
67
+ | 0.9831 | 24.49 | 3600 | 0.5741 | 0.4141 |
68
+ | 0.9827 | 27.21 | 4000 | 0.5754 | 0.4179 |
69
+ | 0.9669 | 29.93 | 4400 | 0.5310 | 0.3889 |
70
+ | 0.9496 | 32.65 | 4800 | 0.5649 | 0.4062 |
71
+ | 0.9112 | 35.37 | 5200 | 0.5738 | 0.3926 |
72
+ | 0.8838 | 38.1 | 5600 | 0.5232 | 0.3768 |
73
+ | 0.8666 | 40.81 | 6000 | 0.5510 | 0.3852 |
74
+ | 0.8366 | 43.54 | 6400 | 0.5436 | 0.3837 |
75
+ | 0.7957 | 46.26 | 6800 | 0.5337 | 0.3775 |
76
+ | 0.7834 | 48.98 | 7200 | 0.5611 | 0.3844 |
77
+ | 0.7685 | 51.7 | 7600 | 0.5710 | 0.4008 |
78
+ | 0.7431 | 54.42 | 8000 | 0.5636 | 0.3726 |
79
+ | 0.7353 | 57.14 | 8400 | 0.5937 | 0.3836 |
80
+ | 0.7001 | 59.86 | 8800 | 0.5815 | 0.3858 |
81
+ | 0.6799 | 62.58 | 9200 | 0.5862 | 0.3696 |
82
+ | 0.6459 | 65.31 | 9600 | 0.6181 | 0.3762 |
83
+ | 0.6121 | 68.03 | 10000 | 0.5637 | 0.3590 |
84
+ | 0.5942 | 70.75 | 10400 | 0.6374 | 0.3882 |
85
+ | 0.5769 | 73.47 | 10800 | 0.6015 | 0.3640 |
86
+ | 0.5689 | 76.19 | 11200 | 0.5669 | 0.3508 |
87
+ | 0.5461 | 78.91 | 11600 | 0.5967 | 0.3621 |
88
+ | 0.5286 | 81.63 | 12000 | 0.5840 | 0.3605 |
89
+ | 0.5057 | 84.35 | 12400 | 0.5848 | 0.3489 |
90
+ | 0.482 | 87.07 | 12800 | 0.5860 | 0.3488 |
91
+ | 0.4655 | 89.79 | 13200 | 0.5780 | 0.3453 |
92
+ | 0.4523 | 92.52 | 13600 | 0.6150 | 0.3532 |
93
+ | 0.4422 | 95.24 | 14000 | 0.5930 | 0.3452 |
94
+ | 0.4436 | 97.96 | 14400 | 0.5867 | 0.3428 |
95
+
96
+
97
+ ### Framework versions
98
+
99
+ - Transformers 4.16.0.dev0
100
+ - Pytorch 1.10.1+cu102
101
+ - Datasets 1.17.1.dev0
102
+ - Tokenizers 0.11.0