Hugging Face's logo

DrishtiSharma
/
wav2vec2-large-xls-r-300m-bg-v1

Automatic Speech Recognition
Transformers
PyTorch
Bulgarian
wav2vec2
Generated from Trainer
hf-asr-leaderboard
model_for_talk
mozilla-foundation/common_voice_8_0
robust-speech-event
Eval Results
Inference Endpoints
Model card Files Community
DrishtiSharma commited on
Commit
fcedf11
•
1 Parent(s): 885bdf1

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +23 -27
README.md CHANGED
@@ -1,10 +1,6 @@
1
  ---
2
- language:
3
- - bg
4
  license: apache-2.0
5
  tags:
6
- - automatic-speech-recognition
7
- - mozilla-foundation/common_voice_8_0
8
  - generated_from_trainer
9
  datasets:
10
  - common_voice
@@ -18,10 +14,10 @@ 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-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - BG dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.7063
24
- - Wer: 0.5557
25
 
26
  ## Model description
27
 
@@ -46,7 +42,7 @@ The following hyperparameters were used during training:
46
  - seed: 42
47
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
  - lr_scheduler_type: linear
49
- - lr_scheduler_warmup_steps: 1200
50
  - num_epochs: 50.0
51
  - mixed_precision_training: Native AMP
52
 
@@ -54,25 +50,25 @@ The following hyperparameters were used during training:
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Wer |
56
  |:-------------:|:-----:|:----:|:---------------:|:------:|
57
- | 3.9554 | 2.61 | 300 | 3.8166 | 1.0 |
58
- | 2.9853 | 5.22 | 600 | 2.9793 | 1.0 |
59
- | 0.6939 | 7.83 | 900 | 0.9723 | 0.9149 |
60
- | 0.352 | 10.43 | 1200 | 0.6556 | 0.7296 |
61
- | 0.2369 | 13.04 | 1500 | 0.6315 | 0.6928 |
62
- | 0.1757 | 15.65 | 1800 | 0.6387 | 0.6724 |
63
- | 0.1517 | 18.26 | 2100 | 0.6607 | 0.6265 |
64
- | 0.1229 | 20.87 | 2400 | 0.6520 | 0.6183 |
65
- | 0.1201 | 23.48 | 2700 | 0.6750 | 0.6116 |
66
- | 0.1076 | 26.09 | 3000 | 0.6698 | 0.6006 |
67
- | 0.1006 | 28.7 | 3300 | 0.6804 | 0.5872 |
68
- | 0.0952 | 31.3 | 3600 | 0.7010 | 0.6003 |
69
- | 0.0894 | 33.91 | 3900 | 0.7069 | 0.5840 |
70
- | 0.0873 | 36.52 | 4200 | 0.6765 | 0.5781 |
71
- | 0.0798 | 39.13 | 4500 | 0.6822 | 0.5617 |
72
- | 0.0767 | 41.74 | 4800 | 0.7003 | 0.5638 |
73
- | 0.0717 | 44.35 | 5100 | 0.6970 | 0.5591 |
74
- | 0.0687 | 46.96 | 5400 | 0.7073 | 0.5603 |
75
- | 0.0669 | 49.57 | 5700 | 0.7060 | 0.5559 |
76
 
77
 
78
  ### Framework versions
 
1
  ---
 
 
2
  license: apache-2.0
3
  tags:
 
 
4
  - generated_from_trainer
5
  datasets:
6
  - common_voice
 
14
 
15
  #
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: 0.5217
20
+ - Wer: 0.4698
21
 
22
  ## Model description
23
 
 
42
  - seed: 42
43
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
  - lr_scheduler_type: linear
45
+ - lr_scheduler_warmup_steps: 2000
46
  - num_epochs: 50.0
47
  - mixed_precision_training: Native AMP
48
 
 
50
 
51
  | Training Loss | Epoch | Step | Validation Loss | Wer |
52
  |:-------------:|:-----:|:----:|:---------------:|:------:|
53
+ | 4.3711 | 2.61 | 300 | 4.3122 | 1.0 |
54
+ | 3.1653 | 5.22 | 600 | 3.1156 | 1.0 |
55
+ | 2.8904 | 7.83 | 900 | 2.8421 | 0.9918 |
56
+ | 0.9207 | 10.43 | 1200 | 0.9895 | 0.8689 |
57
+ | 0.6384 | 13.04 | 1500 | 0.6994 | 0.7700 |
58
+ | 0.5215 | 15.65 | 1800 | 0.5628 | 0.6443 |
59
+ | 0.4573 | 18.26 | 2100 | 0.5316 | 0.6174 |
60
+ | 0.3875 | 20.87 | 2400 | 0.4932 | 0.5779 |
61
+ | 0.3562 | 23.48 | 2700 | 0.4972 | 0.5475 |
62
+ | 0.3218 | 26.09 | 3000 | 0.4895 | 0.5219 |
63
+ | 0.2954 | 28.7 | 3300 | 0.5226 | 0.5192 |
64
+ | 0.287 | 31.3 | 3600 | 0.4957 | 0.5146 |
65
+ | 0.2587 | 33.91 | 3900 | 0.4944 | 0.4893 |
66
+ | 0.2496 | 36.52 | 4200 | 0.4976 | 0.4895 |
67
+ | 0.2365 | 39.13 | 4500 | 0.5185 | 0.4819 |
68
+ | 0.2264 | 41.74 | 4800 | 0.5152 | 0.4776 |
69
+ | 0.2224 | 44.35 | 5100 | 0.5031 | 0.4746 |
70
+ | 0.2096 | 46.96 | 5400 | 0.5062 | 0.4708 |
71
+ | 0.2038 | 49.57 | 5700 | 0.5217 | 0.4698 |
72
 
73
 
74
  ### Framework versions