Hugging Face's logo

badrex
/
xls-r-300-cv17-bulgarian

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
Badr Abdullah commited on
Commit
16c10c6
·
verified ·
1 Parent(s): 712091f

Model save

Browse files
Files changed (1) hide show
  1. README.md +55 -32
README.md CHANGED
@@ -1,8 +1,8 @@
1
  ---
2
  license: apache-2.0
 
3
  tags:
4
  - generated_from_trainer
5
- base_model: facebook/wav2vec2-xls-r-300m
6
  datasets:
7
  - common_voice_17_0
8
  metrics:
@@ -11,8 +11,8 @@ model-index:
11
  - name: xls-r-300-cv17-bulgarian
12
  results:
13
  - task:
14
- type: automatic-speech-recognition
15
  name: Automatic Speech Recognition
 
16
  dataset:
17
  name: common_voice_17_0
18
  type: common_voice_17_0
@@ -20,22 +20,22 @@ model-index:
20
  split: validation
21
  args: bg
22
  metrics:
23
- - type: wer
24
- value: 0.33163942508248323
25
- name: Wer
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
  should probably proofread and complete it, then remove this comment. -->
30
 
31
- [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/badr-nlp/xlsr-continual-finetuning-polish/runs/mul1mun3)
32
  # xls-r-300-cv17-bulgarian
33
 
34
  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_17_0 dataset.
35
  It achieves the following results on the evaluation set:
36
- - Loss: 0.3895
37
- - Wer: 0.3316
38
- - Cer: 0.0811
39
 
40
  ## Model description
41
 
@@ -63,35 +63,58 @@ The following hyperparameters were used during training:
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: linear
65
  - lr_scheduler_warmup_steps: 500
66
- - num_epochs: 15
67
  - mixed_precision_training: Native AMP
68
 
69
  ### Training results
70
 
71
  | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
72
  |:-------------:|:-------:|:----:|:---------------:|:------:|:------:|
73
- | 4.0888 | 0.6579 | 100 | 4.2092 | 1.0 | 1.0 |
74
- | 3.0541 | 1.3158 | 200 | 3.0969 | 1.0 | 1.0 |
75
- | 2.5798 | 1.9737 | 300 | 2.5496 | 0.9998 | 0.8857 |
76
- | 0.5921 | 2.6316 | 400 | 0.7055 | 0.7854 | 0.1992 |
77
- | 0.3964 | 3.2895 | 500 | 0.6162 | 0.6423 | 0.1669 |
78
- | 0.3671 | 3.9474 | 600 | 0.5051 | 0.5392 | 0.1332 |
79
- | 0.2505 | 4.6053 | 700 | 0.4911 | 0.4966 | 0.1265 |
80
- | 0.2061 | 5.2632 | 800 | 0.4461 | 0.4736 | 0.1142 |
81
- | 0.1914 | 5.9211 | 900 | 0.4010 | 0.4242 | 0.1025 |
82
- | 0.1222 | 6.5789 | 1000 | 0.3952 | 0.4053 | 0.0987 |
83
- | 0.2282 | 7.2368 | 1100 | 0.4349 | 0.4146 | 0.1005 |
84
- | 0.1489 | 7.8947 | 1200 | 0.4001 | 0.3886 | 0.0944 |
85
- | 0.1025 | 8.5526 | 1300 | 0.4046 | 0.3702 | 0.0889 |
86
- | 0.0943 | 9.2105 | 1400 | 0.3841 | 0.3621 | 0.0865 |
87
- | 0.1038 | 9.8684 | 1500 | 0.3942 | 0.3450 | 0.0826 |
88
- | 0.0735 | 10.5263 | 1600 | 0.4073 | 0.3820 | 0.0979 |
89
- | 0.0357 | 11.1842 | 1700 | 0.4167 | 0.3626 | 0.0898 |
90
- | 0.0644 | 11.8421 | 1800 | 0.3955 | 0.3451 | 0.0821 |
91
- | 0.0391 | 12.5 | 1900 | 0.3957 | 0.3456 | 0.0841 |
92
- | 0.0473 | 13.1579 | 2000 | 0.3902 | 0.3390 | 0.0826 |
93
- | 0.0199 | 13.8158 | 2100 | 0.3926 | 0.3329 | 0.0803 |
94
- | 0.0513 | 14.4737 | 2200 | 0.3895 | 0.3316 | 0.0811 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95
 
96
 
97
  ### Framework versions
 
1
  ---
2
  license: apache-2.0
3
+ base_model: facebook/wav2vec2-xls-r-300m
4
  tags:
5
  - generated_from_trainer
 
6
  datasets:
7
  - common_voice_17_0
8
  metrics:
 
11
  - name: xls-r-300-cv17-bulgarian
12
  results:
13
  - task:
 
14
  name: Automatic Speech Recognition
15
+ type: automatic-speech-recognition
16
  dataset:
17
  name: common_voice_17_0
18
  type: common_voice_17_0
 
20
  split: validation
21
  args: bg
22
  metrics:
23
+ - name: Wer
24
+ type: wer
25
+ value: 0.2967878948765596
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
  should probably proofread and complete it, then remove this comment. -->
30
 
31
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/badr-nlp/xlsr-continual-finetuning-polish/runs/snulovqw)
32
  # xls-r-300-cv17-bulgarian
33
 
34
  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_17_0 dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.4329
37
+ - Wer: 0.2968
38
+ - Cer: 0.0726
39
 
40
  ## Model description
41
 
 
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: linear
65
  - lr_scheduler_warmup_steps: 500
66
+ - num_epochs: 30
67
  - mixed_precision_training: Native AMP
68
 
69
  ### Training results
70
 
71
  | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
72
  |:-------------:|:-------:|:----:|:---------------:|:------:|:------:|
73
+ | 4.0388 | 0.6579 | 100 | 4.1422 | 1.0 | 1.0 |
74
+ | 3.047 | 1.3158 | 200 | 3.0730 | 1.0 | 1.0 |
75
+ | 2.7349 | 1.9737 | 300 | 2.7601 | 0.9939 | 0.9946 |
76
+ | 0.6047 | 2.6316 | 400 | 0.6984 | 0.7954 | 0.1942 |
77
+ | 0.3868 | 3.2895 | 500 | 0.5550 | 0.5994 | 0.1519 |
78
+ | 0.3423 | 3.9474 | 600 | 0.4548 | 0.4804 | 0.1195 |
79
+ | 0.1942 | 4.6053 | 700 | 0.3973 | 0.4277 | 0.1034 |
80
+ | 0.1754 | 5.2632 | 800 | 0.4166 | 0.4391 | 0.1055 |
81
+ | 0.1734 | 5.9211 | 900 | 0.4146 | 0.4195 | 0.1018 |
82
+ | 0.1089 | 6.5789 | 1000 | 0.3859 | 0.3867 | 0.0937 |
83
+ | 0.233 | 7.2368 | 1100 | 0.4183 | 0.4054 | 0.1005 |
84
+ | 0.1519 | 7.8947 | 1200 | 0.4459 | 0.4151 | 0.1030 |
85
+ | 0.1176 | 8.5526 | 1300 | 0.4026 | 0.3845 | 0.0937 |
86
+ | 0.0997 | 9.2105 | 1400 | 0.3849 | 0.3590 | 0.0869 |
87
+ | 0.1266 | 9.8684 | 1500 | 0.4281 | 0.3781 | 0.0947 |
88
+ | 0.0945 | 10.5263 | 1600 | 0.4471 | 0.3983 | 0.0979 |
89
+ | 0.0575 | 11.1842 | 1700 | 0.4290 | 0.3660 | 0.0897 |
90
+ | 0.0854 | 11.8421 | 1800 | 0.4258 | 0.3749 | 0.0938 |
91
+ | 0.0558 | 12.5 | 1900 | 0.4242 | 0.3644 | 0.0907 |
92
+ | 0.0774 | 13.1579 | 2000 | 0.4339 | 0.3616 | 0.0888 |
93
+ | 0.0397 | 13.8158 | 2100 | 0.4155 | 0.3581 | 0.0882 |
94
+ | 0.0603 | 14.4737 | 2200 | 0.4681 | 0.3737 | 0.0943 |
95
+ | 0.0723 | 15.1316 | 2300 | 0.4446 | 0.3560 | 0.0875 |
96
+ | 0.0746 | 15.7895 | 2400 | 0.4430 | 0.3573 | 0.0889 |
97
+ | 0.0727 | 16.4474 | 2500 | 0.4549 | 0.3470 | 0.0870 |
98
+ | 0.0458 | 17.1053 | 2600 | 0.4581 | 0.3520 | 0.0873 |
99
+ | 0.0694 | 17.7632 | 2700 | 0.4414 | 0.3575 | 0.0896 |
100
+ | 0.0462 | 18.4211 | 2800 | 0.4235 | 0.3261 | 0.0802 |
101
+ | 0.0539 | 19.0789 | 2900 | 0.4496 | 0.3329 | 0.0810 |
102
+ | 0.0368 | 19.7368 | 3000 | 0.4043 | 0.3406 | 0.0846 |
103
+ | 0.0347 | 20.3947 | 3100 | 0.4367 | 0.3225 | 0.0789 |
104
+ | 0.019 | 21.0526 | 3200 | 0.4487 | 0.3272 | 0.0801 |
105
+ | 0.0361 | 21.7105 | 3300 | 0.4272 | 0.3241 | 0.0785 |
106
+ | 0.0475 | 22.3684 | 3400 | 0.4324 | 0.3191 | 0.0781 |
107
+ | 0.0341 | 23.0263 | 3500 | 0.4564 | 0.3398 | 0.0847 |
108
+ | 0.0454 | 23.6842 | 3600 | 0.4415 | 0.3188 | 0.0789 |
109
+ | 0.0346 | 24.3421 | 3700 | 0.4187 | 0.3072 | 0.0751 |
110
+ | 0.1315 | 25.0 | 3800 | 0.4480 | 0.3124 | 0.0765 |
111
+ | 0.0663 | 25.6579 | 3900 | 0.4488 | 0.3151 | 0.0779 |
112
+ | 0.0225 | 26.3158 | 4000 | 0.4372 | 0.3006 | 0.0739 |
113
+ | 0.0382 | 26.9737 | 4100 | 0.4164 | 0.2987 | 0.0730 |
114
+ | 0.0194 | 27.6316 | 4200 | 0.4190 | 0.2942 | 0.0718 |
115
+ | 0.0101 | 28.2895 | 4300 | 0.4328 | 0.2960 | 0.0726 |
116
+ | 0.0224 | 28.9474 | 4400 | 0.4302 | 0.2944 | 0.0720 |
117
+ | 0.0174 | 29.6053 | 4500 | 0.4329 | 0.2968 | 0.0726 |
118
 
119
 
120
  ### Framework versions