Hugging Face's logo

dennohpeter
/
wav2vec2-large-xlsr-53-sw-tokenizer

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Eval Results
Model card Files
xet
Metrics Community
dennohpeter commited on
Commit
9a0c124
·
verified ·
1 Parent(s): 9c06fa0

End of training

Browse files
Files changed (1) hide show
  1. README.md +72 -19
README.md CHANGED
@@ -12,18 +12,18 @@ model-index:
12
  - name: wav2vec2-large-xlsr-53-sw-tokenizer
13
  results:
14
  - task:
15
- type: automatic-speech-recognition
16
  name: Automatic Speech Recognition
 
17
  dataset:
18
  name: common_voice_17_0
19
  type: common_voice_17_0
20
  config: sw
21
- split: train
22
  args: sw
23
  metrics:
24
- - type: wer
25
- value: 0.28685958074769186
26
- name: Wer
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,8 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
33
 
34
  This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the common_voice_17_0 dataset.
35
  It achieves the following results on the evaluation set:
36
- - Loss: 0.2825
37
- - Wer: 0.2869
38
 
39
  ## Model description
40
 
@@ -57,26 +57,79 @@ The following hyperparameters were used during training:
57
  - train_batch_size: 8
58
  - eval_batch_size: 8
59
  - seed: 42
60
- - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
61
  - lr_scheduler_type: linear
62
  - lr_scheduler_warmup_steps: 1000
63
- - num_epochs: 1
64
  - mixed_precision_training: Native AMP
65
 
66
  ### Training results
67
 
68
- | Training Loss | Epoch | Step | Validation Loss | Wer |
69
- |:-------------:|:------:|:----:|:---------------:|:------:|
70
- | No log | 0.1721 | 1000 | 0.6533 | 0.6980 |
71
- | No log | 0.3441 | 2000 | 0.4034 | 0.4430 |
72
- | 2.231 | 0.5162 | 3000 | 0.3369 | 0.3767 |
73
- | 2.231 | 0.6882 | 4000 | 0.3075 | 0.3246 |
74
- | 2.231 | 0.8603 | 5000 | 0.2825 | 0.2869 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
75
 
76
 
77
  ### Framework versions
78
 
79
- - Transformers 4.53.3
80
- - Pytorch 2.7.1+cu126
81
  - Datasets 3.6.0
82
- - Tokenizers 0.21.2
 
12
  - name: wav2vec2-large-xlsr-53-sw-tokenizer
13
  results:
14
  - task:
 
15
  name: Automatic Speech Recognition
16
+ type: automatic-speech-recognition
17
  dataset:
18
  name: common_voice_17_0
19
  type: common_voice_17_0
20
  config: sw
21
+ split: test
22
  args: sw
23
  metrics:
24
+ - name: Wer
25
+ type: wer
26
+ value: 0.3240375929734197
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
33
 
34
  This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the common_voice_17_0 dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.4306
37
+ - Wer: 0.3240
38
 
39
  ## Model description
40
 
 
57
  - train_batch_size: 8
58
  - eval_batch_size: 8
59
  - seed: 42
60
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
61
  - lr_scheduler_type: linear
62
  - lr_scheduler_warmup_steps: 1000
63
+ - num_epochs: 10
64
  - mixed_precision_training: Native AMP
65
 
66
  ### Training results
67
 
68
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
69
+ |:-------------:|:------:|:-----:|:---------------:|:------:|
70
+ | No log | 0.1721 | 1000 | 0.7966 | 0.7685 |
71
+ | No log | 0.3441 | 2000 | 0.5178 | 0.5562 |
72
+ | 2.0511 | 0.5162 | 3000 | 0.4524 | 0.5039 |
73
+ | 2.0511 | 0.6882 | 4000 | 0.4207 | 0.4615 |
74
+ | 2.0511 | 0.8603 | 5000 | 0.4031 | 0.4437 |
75
+ | 0.2699 | 1.0323 | 6000 | 0.3875 | 0.4224 |
76
+ | 0.2699 | 1.2044 | 7000 | 0.3870 | 0.4141 |
77
+ | 0.2699 | 1.3765 | 8000 | 0.3811 | 0.4143 |
78
+ | 0.1994 | 1.5485 | 9000 | 0.3689 | 0.4026 |
79
+ | 0.1994 | 1.7206 | 10000 | 0.3603 | 0.3915 |
80
+ | 0.1994 | 1.8926 | 11000 | 0.3561 | 0.3862 |
81
+ | 0.1838 | 2.0647 | 12000 | 0.3502 | 0.3809 |
82
+ | 0.1838 | 2.2368 | 13000 | 0.3580 | 0.3763 |
83
+ | 0.1838 | 2.4088 | 14000 | 0.3445 | 0.3747 |
84
+ | 0.1472 | 2.5809 | 15000 | 0.3416 | 0.3720 |
85
+ | 0.1472 | 2.7529 | 16000 | 0.3599 | 0.3709 |
86
+ | 0.1472 | 2.9250 | 17000 | 0.3503 | 0.3666 |
87
+ | 0.1405 | 3.0970 | 18000 | 0.3549 | 0.3624 |
88
+ | 0.1405 | 3.2691 | 19000 | 0.3476 | 0.3582 |
89
+ | 0.1405 | 3.4412 | 20000 | 0.3359 | 0.3574 |
90
+ | 0.116 | 3.6132 | 21000 | 0.3487 | 0.3600 |
91
+ | 0.116 | 3.7853 | 22000 | 0.3439 | 0.3552 |
92
+ | 0.116 | 3.9573 | 23000 | 0.3502 | 0.3579 |
93
+ | 0.1103 | 4.1294 | 24000 | 0.3436 | 0.3513 |
94
+ | 0.1103 | 4.3014 | 25000 | 0.3502 | 0.3502 |
95
+ | 0.1103 | 4.4735 | 26000 | 0.3381 | 0.3534 |
96
+ | 0.0957 | 4.6456 | 27000 | 0.3411 | 0.3482 |
97
+ | 0.0957 | 4.8176 | 28000 | 0.3425 | 0.3456 |
98
+ | 0.0957 | 4.9897 | 29000 | 0.3331 | 0.3425 |
99
+ | 0.0883 | 5.1617 | 30000 | 0.3620 | 0.3449 |
100
+ | 0.0883 | 5.3338 | 31000 | 0.3403 | 0.3430 |
101
+ | 0.0883 | 5.5058 | 32000 | 0.3590 | 0.3429 |
102
+ | 0.0757 | 5.6779 | 33000 | 0.3474 | 0.3402 |
103
+ | 0.0757 | 5.8500 | 34000 | 0.3395 | 0.3378 |
104
+ | 0.0757 | 6.0220 | 35000 | 0.3565 | 0.3395 |
105
+ | 0.0695 | 6.1941 | 36000 | 0.3729 | 0.3397 |
106
+ | 0.0695 | 6.3661 | 37000 | 0.3676 | 0.3368 |
107
+ | 0.0695 | 6.5382 | 38000 | 0.3748 | 0.3364 |
108
+ | 0.0601 | 6.7103 | 39000 | 0.3783 | 0.3360 |
109
+ | 0.0601 | 6.8823 | 40000 | 0.3657 | 0.3363 |
110
+ | 0.0601 | 7.0544 | 41000 | 0.3808 | 0.3343 |
111
+ | 0.0542 | 7.2264 | 42000 | 0.3934 | 0.3361 |
112
+ | 0.0542 | 7.3985 | 43000 | 0.3787 | 0.3369 |
113
+ | 0.0542 | 7.5705 | 44000 | 0.3920 | 0.3310 |
114
+ | 0.0487 | 7.7426 | 45000 | 0.3906 | 0.3321 |
115
+ | 0.0487 | 7.9147 | 46000 | 0.3934 | 0.3323 |
116
+ | 0.0487 | 8.0867 | 47000 | 0.4060 | 0.3305 |
117
+ | 0.0412 | 8.2588 | 48000 | 0.4145 | 0.3301 |
118
+ | 0.0412 | 8.4308 | 49000 | 0.4125 | 0.3282 |
119
+ | 0.0412 | 8.6029 | 50000 | 0.4111 | 0.3286 |
120
+ | 0.0381 | 8.7749 | 51000 | 0.4113 | 0.3265 |
121
+ | 0.0381 | 8.9470 | 52000 | 0.4147 | 0.3268 |
122
+ | 0.0381 | 9.1191 | 53000 | 0.4221 | 0.3271 |
123
+ | 0.0338 | 9.2911 | 54000 | 0.4299 | 0.3268 |
124
+ | 0.0338 | 9.4632 | 55000 | 0.4221 | 0.3250 |
125
+ | 0.0338 | 9.6352 | 56000 | 0.4314 | 0.3245 |
126
+ | 0.0318 | 9.8073 | 57000 | 0.4307 | 0.3243 |
127
+ | 0.0318 | 9.9794 | 58000 | 0.4306 | 0.3240 |
128
 
129
 
130
  ### Framework versions
131
 
132
+ - Transformers 4.55.4
133
+ - Pytorch 2.8.0+cu126
134
  - Datasets 3.6.0
135
+ - Tokenizers 0.21.4