File size: 2,905 Bytes
a7d817e
7c6cf8d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a7d817e
 
7c6cf8d
 
a7d817e
7c6cf8d
a7d817e
7c6cf8d
 
 
 
 
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
a7d817e
7c6cf8d
 
 
 
 
 
 
 
 
 
 
a7d817e
7c6cf8d
a7d817e
7c6cf8d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a7d817e
 
7c6cf8d
a7d817e
7c6cf8d
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
---
license: mit
base_model: facebook/w2v-bert-2.0
tags:
- generated_from_trainer
datasets:
- common_voice_17_0
metrics:
- wer
model-index:
- name: adapter_freezed_base_const_lr
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: common_voice_17_0
      type: common_voice_17_0
      config: hy-AM
      split: test
      args: hy-AM
    metrics:
    - name: Wer
      type: wer
      value: 0.9281584969288209
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# adapter_freezed_base_const_lr

This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) on the common_voice_17_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9200
- Wer: 0.9282
- Cer: 0.2562

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- num_epochs: 10
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    | Cer    |
|:-------------:|:------:|:----:|:---------------:|:------:|:------:|
| 1.3224        | 0.6154 | 200  | 1.3171          | 0.9949 | 0.3890 |
| 1.02          | 1.2308 | 400  | 1.0780          | 0.9728 | 0.3233 |
| 0.9256        | 1.8462 | 600  | 0.9799          | 0.9738 | 0.2955 |
| 0.8377        | 2.4615 | 800  | 0.9756          | 0.9663 | 0.2919 |
| 0.7836        | 3.0769 | 1000 | 0.9143          | 0.9535 | 0.2730 |
| 0.7516        | 3.6923 | 1200 | 0.8908          | 0.9373 | 0.2671 |
| 0.6714        | 4.3077 | 1400 | 0.9088          | 0.9497 | 0.2692 |
| 0.6749        | 4.9231 | 1600 | 0.9006          | 0.9566 | 0.2681 |
| 0.6223        | 5.5385 | 1800 | 0.8686          | 0.9322 | 0.2587 |
| 0.5643        | 6.1538 | 2000 | 0.8846          | 0.9422 | 0.2580 |
| 0.5773        | 6.7692 | 2200 | 0.8960          | 0.9396 | 0.2644 |
| 0.5067        | 7.3846 | 2400 | 0.8778          | 0.9273 | 0.2545 |
| 0.5123        | 8.0    | 2600 | 0.8919          | 0.9379 | 0.2601 |
| 0.4729        | 8.6154 | 2800 | 0.9131          | 0.9597 | 0.2587 |
| 0.406         | 9.2308 | 3000 | 0.9032          | 0.9389 | 0.2564 |
| 0.4286        | 9.8462 | 3200 | 0.9200          | 0.9282 | 0.2562 |


### Framework versions

- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1