File size: 2,915 Bytes
4b0a79a
208fbea
cc245c2
208fbea
 
 
 
 
 
 
4b0a79a
 
208fbea
 
 
 
 
 
 
cc245c2
 
 
208fbea
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc245c2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
208fbea
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: facebook/hubert-large-ll60k
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: speech_ocean_hubert_mdd
  results: []
---

<!-- 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. -->

# speech_ocean_hubert_mdd

This model is a fine-tuned version of [facebook/hubert-large-ll60k](https://huggingface.co/facebook/hubert-large-ll60k) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2060
- Wer: 0.0531
- Cer: 0.0507

## 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: 0.0003
- 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: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 20
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer    | Cer    |
|:-------------:|:-------:|:----:|:---------------:|:------:|:------:|
| 42.5933       | 0.9873  | 39   | 36.3096         | 0.9995 | 0.9985 |
| 15.8463       | 2.0     | 79   | 7.9376          | 1.0    | 1.0    |
| 6.5393        | 2.9873  | 118  | 4.5528          | 1.0    | 1.0    |
| 3.9947        | 4.0     | 158  | 3.8592          | 1.0    | 1.0    |
| 3.8025        | 4.9873  | 197  | 3.8002          | 1.0    | 1.0    |
| 3.7706        | 6.0     | 237  | 3.7435          | 1.0    | 1.0    |
| 3.7676        | 6.9873  | 276  | 3.7276          | 1.0    | 1.0    |
| 3.7353        | 8.0     | 316  | 3.7150          | 1.0    | 1.0    |
| 3.7126        | 8.9873  | 355  | 3.6717          | 1.0    | 1.0    |
| 3.5628        | 10.0    | 395  | 3.3098          | 1.0    | 1.0    |
| 2.837         | 10.9873 | 434  | 2.3304          | 0.8277 | 0.8915 |
| 2.1018        | 12.0    | 474  | 1.5575          | 0.5441 | 0.6213 |
| 1.6164        | 12.9873 | 513  | 1.0106          | 0.2678 | 0.2596 |
| 1.1823        | 14.0    | 553  | 0.6938          | 0.1788 | 0.1509 |
| 0.9451        | 14.9873 | 592  | 0.4673          | 0.1154 | 0.0925 |
| 0.7055        | 16.0    | 632  | 0.3455          | 0.0893 | 0.0767 |
| 0.5434        | 16.9873 | 671  | 0.2803          | 0.0718 | 0.0637 |
| 0.4867        | 18.0    | 711  | 0.2362          | 0.0608 | 0.0566 |
| 0.4172        | 18.9873 | 750  | 0.2125          | 0.0551 | 0.0522 |
| 0.4406        | 19.7468 | 780  | 0.2060          | 0.0531 | 0.0507 |


### Framework versions

- Transformers 4.40.0
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1