Hugging Face's logo

BALAKA
/
wav2vec2-large-xlsr-53-thai

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
1
BALAKA commited on
Commit
5078890
·
1 Parent(s): 09cf5ab

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +33 -30
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  metrics:
22
  - name: Wer
23
  type: wer
24
- value: 0.7535428219346888
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model is a fine-tuned version of [airesearch/wav2vec2-large-xlsr-53-th](https://huggingface.co/airesearch/wav2vec2-large-xlsr-53-th) on the common_voice dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 3.4777
35
- - Wer: 0.7535
36
 
37
  ## Model description
38
 
@@ -65,33 +65,36 @@ The following hyperparameters were used during training:
65
 
66
  | Training Loss | Epoch | Step | Validation Loss | Wer |
67
  |:-------------:|:-----:|:----:|:---------------:|:------:|
68
- | 6.6911 | 3.7 | 100 | 3.3086 | 1.0 |
69
- | 3.3573 | 7.41 | 200 | 3.2422 | 1.0 |
70
- | 2.7807 | 11.11 | 300 | 2.4797 | 0.9057 |
71
- | 1.0981 | 14.81 | 400 | 2.3788 | 0.7677 |
72
- | 0.766 | 18.52 | 500 | 2.3853 | 0.7702 |
73
- | 0.6174 | 22.22 | 600 | 2.4137 | 0.7511 |
74
- | 0.5404 | 25.93 | 700 | 2.5811 | 0.7455 |
75
- | 0.4758 | 29.63 | 800 | 2.7067 | 0.7671 |
76
- | 0.4322 | 33.33 | 900 | 2.7301 | 0.7437 |
77
- | 0.3746 | 37.04 | 1000 | 2.9974 | 0.7591 |
78
- | 0.3288 | 40.74 | 1100 | 2.9867 | 0.7517 |
79
- | 0.3181 | 44.44 | 1200 | 3.0836 | 0.7683 |
80
- | 0.2996 | 48.15 | 1300 | 3.1538 | 0.7652 |
81
- | 0.2697 | 51.85 | 1400 | 3.1420 | 0.7659 |
82
- | 0.2608 | 55.56 | 1500 | 3.2246 | 0.7566 |
83
- | 0.239 | 59.26 | 1600 | 3.2098 | 0.7560 |
84
- | 0.2399 | 62.96 | 1700 | 3.3133 | 0.7616 |
85
- | 0.2183 | 66.67 | 1800 | 3.2815 | 0.7554 |
86
- | 0.2231 | 70.37 | 1900 | 3.3064 | 0.7603 |
87
- | 0.2304 | 74.07 | 2000 | 3.3705 | 0.7603 |
88
- | 0.1854 | 77.78 | 2100 | 3.3617 | 0.7591 |
89
- | 0.1986 | 81.48 | 2200 | 3.4577 | 0.7579 |
90
- | 0.1952 | 85.19 | 2300 | 3.5122 | 0.7548 |
91
- | 0.1883 | 88.89 | 2400 | 3.4305 | 0.7548 |
92
- | 0.1872 | 92.59 | 2500 | 3.4533 | 0.7511 |
93
- | 0.1857 | 96.3 | 2600 | 3.4653 | 0.7542 |
94
- | 0.222 | 100.0 | 2700 | 3.4777 | 0.7535 |
 
 
 
95
 
96
 
97
  ### Framework versions
 
21
  metrics:
22
  - name: Wer
23
  type: wer
24
+ value: 0.7430683918669131
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
31
 
32
  This model is a fine-tuned version of [airesearch/wav2vec2-large-xlsr-53-th](https://huggingface.co/airesearch/wav2vec2-large-xlsr-53-th) on the common_voice dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 3.3576
35
+ - Wer: 0.7431
36
 
37
  ## Model description
38
 
 
65
 
66
  | Training Loss | Epoch | Step | Validation Loss | Wer |
67
  |:-------------:|:-----:|:----:|:---------------:|:------:|
68
+ | 6.7312 | 3.33 | 100 | 3.3592 | 1.0 |
69
+ | 3.3687 | 6.67 | 200 | 3.2175 | 1.0 |
70
+ | 2.4527 | 10.0 | 300 | 2.2648 | 0.7911 |
71
+ | 1.0505 | 13.33 | 400 | 2.2322 | 0.7659 |
72
+ | 0.7725 | 16.67 | 500 | 2.2775 | 0.7505 |
73
+ | 0.6289 | 20.0 | 600 | 2.3209 | 0.7498 |
74
+ | 0.543 | 23.33 | 700 | 2.4494 | 0.7572 |
75
+ | 0.4991 | 26.67 | 800 | 2.5798 | 0.7597 |
76
+ | 0.4492 | 30.0 | 900 | 2.5685 | 0.7461 |
77
+ | 0.3737 | 33.33 | 1000 | 2.6186 | 0.7486 |
78
+ | 0.3358 | 36.67 | 1100 | 2.7781 | 0.7480 |
79
+ | 0.3247 | 40.0 | 1200 | 2.8999 | 0.7535 |
80
+ | 0.2963 | 43.33 | 1300 | 2.8668 | 0.7388 |
81
+ | 0.2825 | 46.67 | 1400 | 2.8983 | 0.7449 |
82
+ | 0.2651 | 50.0 | 1500 | 2.9699 | 0.7461 |
83
+ | 0.2597 | 53.33 | 1600 | 2.9930 | 0.7314 |
84
+ | 0.2629 | 56.67 | 1700 | 2.9852 | 0.7406 |
85
+ | 0.2406 | 60.0 | 1800 | 3.0552 | 0.7474 |
86
+ | 0.2293 | 63.33 | 1900 | 3.1058 | 0.7344 |
87
+ | 0.2193 | 66.67 | 2000 | 3.1594 | 0.7406 |
88
+ | 0.2174 | 70.0 | 2100 | 3.2351 | 0.7369 |
89
+ | 0.2127 | 73.33 | 2200 | 3.2696 | 0.7388 |
90
+ | 0.2061 | 76.67 | 2300 | 3.2954 | 0.7566 |
91
+ | 0.1947 | 80.0 | 2400 | 3.2878 | 0.7529 |
92
+ | 0.199 | 83.33 | 2500 | 3.3233 | 0.7486 |
93
+ | 0.1961 | 86.67 | 2600 | 3.3136 | 0.7437 |
94
+ | 0.1928 | 90.0 | 2700 | 3.3240 | 0.7406 |
95
+ | 0.1875 | 93.33 | 2800 | 3.3479 | 0.7425 |
96
+ | 0.1852 | 96.67 | 2900 | 3.3681 | 0.7425 |
97
+ | 0.1814 | 100.0 | 3000 | 3.3576 | 0.7431 |
98
 
99
 
100
  ### Framework versions