Hugging Face's logo

peter2000
/
wav2vec2-large-xls-r-300m-kinyarwanda

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
1
peter2000 commited on
Commit
9a4464f
•
1 Parent(s): 01dd782

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +78 -11
README.md CHANGED
@@ -16,13 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
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
- - eval_loss: 0.5200
20
- - eval_wer: 1.0508
21
- - eval_runtime: 291.5311
22
- - eval_samples_per_second: 10.788
23
- - eval_steps_per_second: 1.351
24
- - epoch: 4.65
25
- - step: 24400
26
 
27
  ## Model description
28
 
@@ -41,18 +36,90 @@ More information needed
41
  ### Training hyperparameters
42
 
43
  The following hyperparameters were used during training:
44
- - learning_rate: 0.0001
45
- - train_batch_size: 8
46
  - eval_batch_size: 8
47
  - seed: 42
48
  - gradient_accumulation_steps: 2
49
- - total_train_batch_size: 16
50
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
51
  - lr_scheduler_type: linear
52
- - lr_scheduler_warmup_steps: 500
53
  - num_epochs: 8
54
  - mixed_precision_training: Native AMP
55
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
56
  ### Framework versions
57
 
58
  - Transformers 4.11.3
 
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: 0.3917
20
+ - Wer: 0.3246
 
 
 
 
 
21
 
22
  ## Model description
23
 
 
36
  ### Training hyperparameters
37
 
38
  The following hyperparameters were used during training:
39
+ - learning_rate: 7e-05
40
+ - train_batch_size: 12
41
  - eval_batch_size: 8
42
  - seed: 42
43
  - gradient_accumulation_steps: 2
44
+ - total_train_batch_size: 24
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 400
48
  - num_epochs: 8
49
  - mixed_precision_training: Native AMP
50
 
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
54
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
55
+ | 9.0634 | 0.12 | 400 | 3.0554 | 1.0 |
56
+ | 2.8009 | 0.24 | 800 | 1.5927 | 0.9554 |
57
+ | 0.9022 | 0.36 | 1200 | 0.7328 | 0.6445 |
58
+ | 0.6213 | 0.48 | 1600 | 0.6138 | 0.5510 |
59
+ | 0.5299 | 0.6 | 2000 | 0.6072 | 0.5223 |
60
+ | 0.4999 | 0.72 | 2400 | 0.5449 | 0.4969 |
61
+ | 0.4731 | 0.84 | 2800 | 0.5261 | 0.4828 |
62
+ | 0.458 | 0.96 | 3200 | 0.5058 | 0.4607 |
63
+ | 0.4158 | 1.09 | 3600 | 0.4892 | 0.4463 |
64
+ | 0.4037 | 1.21 | 4000 | 0.4759 | 0.4429 |
65
+ | 0.4021 | 1.33 | 4400 | 0.4615 | 0.4330 |
66
+ | 0.3934 | 1.45 | 4800 | 0.4593 | 0.4315 |
67
+ | 0.3808 | 1.57 | 5200 | 0.4736 | 0.4344 |
68
+ | 0.3838 | 1.69 | 5600 | 0.4569 | 0.4249 |
69
+ | 0.3726 | 1.81 | 6000 | 0.4473 | 0.4140 |
70
+ | 0.3623 | 1.93 | 6400 | 0.4403 | 0.4097 |
71
+ | 0.3517 | 2.05 | 6800 | 0.4389 | 0.4061 |
72
+ | 0.333 | 2.17 | 7200 | 0.4383 | 0.4104 |
73
+ | 0.3354 | 2.29 | 7600 | 0.4360 | 0.3955 |
74
+ | 0.3257 | 2.41 | 8000 | 0.4226 | 0.3942 |
75
+ | 0.3275 | 2.53 | 8400 | 0.4206 | 0.4040 |
76
+ | 0.3262 | 2.65 | 8800 | 0.4172 | 0.3875 |
77
+ | 0.3206 | 2.77 | 9200 | 0.4209 | 0.3877 |
78
+ | 0.323 | 2.89 | 9600 | 0.4177 | 0.3825 |
79
+ | 0.3099 | 3.01 | 10000 | 0.4101 | 0.3691 |
80
+ | 0.3008 | 3.14 | 10400 | 0.4055 | 0.3709 |
81
+ | 0.2918 | 3.26 | 10800 | 0.4085 | 0.3800 |
82
+ | 0.292 | 3.38 | 11200 | 0.4089 | 0.3713 |
83
+ | 0.292 | 3.5 | 11600 | 0.4092 | 0.3730 |
84
+ | 0.2785 | 3.62 | 12000 | 0.4151 | 0.3687 |
85
+ | 0.2941 | 3.74 | 12400 | 0.4004 | 0.3639 |
86
+ | 0.2838 | 3.86 | 12800 | 0.4108 | 0.3703 |
87
+ | 0.2854 | 3.98 | 13200 | 0.3911 | 0.3596 |
88
+ | 0.2683 | 4.1 | 13600 | 0.3944 | 0.3575 |
89
+ | 0.2647 | 4.22 | 14000 | 0.3836 | 0.3538 |
90
+ | 0.2704 | 4.34 | 14400 | 0.4006 | 0.3540 |
91
+ | 0.2664 | 4.46 | 14800 | 0.3974 | 0.3553 |
92
+ | 0.2662 | 4.58 | 15200 | 0.3890 | 0.3470 |
93
+ | 0.2615 | 4.7 | 15600 | 0.3856 | 0.3507 |
94
+ | 0.2553 | 4.82 | 16000 | 0.3814 | 0.3497 |
95
+ | 0.2587 | 4.94 | 16400 | 0.3837 | 0.3440 |
96
+ | 0.2522 | 5.06 | 16800 | 0.3834 | 0.3486 |
97
+ | 0.2451 | 5.19 | 17200 | 0.3897 | 0.3414 |
98
+ | 0.2423 | 5.31 | 17600 | 0.3864 | 0.3481 |
99
+ | 0.2434 | 5.43 | 18000 | 0.3808 | 0.3416 |
100
+ | 0.2525 | 5.55 | 18400 | 0.3795 | 0.3408 |
101
+ | 0.2427 | 5.67 | 18800 | 0.3841 | 0.3411 |
102
+ | 0.2411 | 5.79 | 19200 | 0.3804 | 0.3366 |
103
+ | 0.2404 | 5.91 | 19600 | 0.3800 | 0.3328 |
104
+ | 0.2372 | 6.03 | 20000 | 0.3749 | 0.3335 |
105
+ | 0.2244 | 6.15 | 20400 | 0.3820 | 0.3327 |
106
+ | 0.2381 | 6.27 | 20800 | 0.3789 | 0.3325 |
107
+ | 0.2294 | 6.39 | 21200 | 0.3867 | 0.3298 |
108
+ | 0.2378 | 6.51 | 21600 | 0.3843 | 0.3281 |
109
+ | 0.2312 | 6.63 | 22000 | 0.3813 | 0.3277 |
110
+ | 0.2411 | 6.75 | 22400 | 0.3780 | 0.3268 |
111
+ | 0.2315 | 6.87 | 22800 | 0.3790 | 0.3280 |
112
+ | 0.241 | 6.99 | 23200 | 0.3776 | 0.3281 |
113
+ | 0.2313 | 7.11 | 23600 | 0.3929 | 0.3283 |
114
+ | 0.2423 | 7.24 | 24000 | 0.3905 | 0.3280 |
115
+ | 0.2337 | 7.36 | 24400 | 0.3979 | 0.3249 |
116
+ | 0.2368 | 7.48 | 24800 | 0.3980 | 0.3257 |
117
+ | 0.2409 | 7.6 | 25200 | 0.3937 | 0.3229 |
118
+ | 0.2416 | 7.72 | 25600 | 0.3867 | 0.3237 |
119
+ | 0.2364 | 7.84 | 26000 | 0.3912 | 0.3253 |
120
+ | 0.234 | 7.96 | 26400 | 0.3917 | 0.3246 |
121
+
122
+
123
  ### Framework versions
124
 
125
  - Transformers 4.11.3