Hugging Face's logo

Thienpkae
/
wav2vec2-vivos-asr

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
Thienpkae commited on
Commit
b648065
·
verified ·
1 Parent(s): 6b3b690

End of training

Browse files
Files changed (3) hide show
  1. README.md +17 -16
  2. model.safetensors +1 -1
  3. training_args.bin +1 -1
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Wer
24
  type: wer
25
- value: 0.23381058715355313
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the vivos dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.3492
36
- - Wer: 0.2338
37
 
38
  ## Model description
39
 
@@ -52,31 +52,32 @@ More information needed
52
  ### Training hyperparameters
53
 
54
  The following hyperparameters were used during training:
55
- - learning_rate: 0.0001
56
  - train_batch_size: 32
57
  - eval_batch_size: 8
58
  - seed: 42
59
  - gradient_accumulation_steps: 4
60
  - total_train_batch_size: 128
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
- - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_steps: 500
64
  - num_epochs: 20
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
- | Training Loss | Epoch | Step | Validation Loss | Wer |
70
- |:-------------:|:-------:|:----:|:---------------:|:------:|
71
- | 8.4226 | 2.0548 | 150 | 4.9423 | 1.0 |
72
- | 3.59 | 4.1096 | 300 | 3.6898 | 1.0 |
73
- | 3.4271 | 6.1644 | 450 | 3.5183 | 1.0 |
74
- | 2.6948 | 8.2192 | 600 | 1.2770 | 0.8026 |
75
- | 0.7372 | 10.2740 | 750 | 0.5197 | 0.3625 |
76
- | 0.4012 | 12.3288 | 900 | 0.4108 | 0.2911 |
77
- | 0.2974 | 14.3836 | 1050 | 0.3732 | 0.2604 |
78
- | 0.2737 | 16.4384 | 1200 | 0.3550 | 0.2393 |
79
- | 0.2108 | 18.4932 | 1350 | 0.3565 | 0.2434 |
 
80
 
81
 
82
  ### Framework versions
 
22
  metrics:
23
  - name: Wer
24
  type: wer
25
+ value: 0.239169022417987
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
32
 
33
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the vivos dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.3501
36
+ - Wer: 0.2392
37
 
38
  ## Model description
39
 
 
52
  ### Training hyperparameters
53
 
54
  The following hyperparameters were used during training:
55
+ - learning_rate: 8e-05
56
  - train_batch_size: 32
57
  - eval_batch_size: 8
58
  - seed: 42
59
  - gradient_accumulation_steps: 4
60
  - total_train_batch_size: 128
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
+ - lr_scheduler_type: cosine
63
  - lr_scheduler_warmup_steps: 500
64
  - num_epochs: 20
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
70
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
71
+ | 8.7262 | 2.0 | 146 | 5.1186 | 1.0 |
72
+ | 3.6815 | 4.0 | 292 | 3.6847 | 1.0 |
73
+ | 3.4316 | 6.0 | 438 | 3.5415 | 1.0 |
74
+ | 2.8102 | 8.0 | 584 | 1.5866 | 0.9160 |
75
+ | 0.8818 | 10.0 | 730 | 0.5903 | 0.4066 |
76
+ | 0.4305 | 12.0 | 876 | 0.4283 | 0.3104 |
77
+ | 0.3067 | 14.0 | 1022 | 0.3793 | 0.2762 |
78
+ | 0.2819 | 16.0 | 1168 | 0.3620 | 0.2496 |
79
+ | 0.2235 | 18.0 | 1314 | 0.3507 | 0.2405 |
80
+ | 0.211 | 20.0 | 1460 | 0.3501 | 0.2392 |
81
 
82
 
83
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:86896d25cd822dc08de14f4bbe25cd30cd9a49ad97746dab8d61397c6c2b8b51
3
  size 377798672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef938d70673aedb18c43cc805f60cd3a5ba81d86178c9252ce5abc5509f9a8bf
3
  size 377798672
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dd0aac3b25f91179a41c96148fef235fa942ba59c8fd52d4c69b7a28ec10d58f
3
  size 5176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:509b5c205185292663de4c0991d4fa797e00d02428cf8228f04826394b5ef9ef
3
  size 5176