File size: 5,387 Bytes
81917c5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2a4fc12
 
 
 
 
 
 
 
 
 
 
 
 
81917c5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
74eaf05
a889bca
aebd529
698f30a
36d305a
541b7ba
2a4fc12
 
81917c5
 
 
 
 
 
 
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
---
language:
- 'no'
license: apache-2.0
base_model: NbAiLabBeta/nb-whisper-medium
tags:
- audio
- asr
- automatic-speech-recognition
- hf-asr-leaderboard
model-index:
- name: nb-whisper-medium-v0.7-semantic
  results: []
---

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

# nb-whisper-medium-v0.7-semantic

This model is a fine-tuned version of [NbAiLabBeta/nb-whisper-medium](https://huggingface.co/NbAiLabBeta/nb-whisper-medium) on the NbAiLab/ncc_speech_styling_v4 dataset.
It achieves the following results on the evaluation set:
- step: 249
- validation_nst_loss: 0.6407
- train_loss: 0.6265
- validation_nst_wer: 2.4389
- validation_nst_cer: 0.7737
- validation_nst_exact_wer: 3.1412
- validation_nst_exact_cer: 0.8810
- validation_clean_stortinget_no_loss: 0.7441
- validation_clean_stortinget_no_wer: 9.1639
- validation_clean_stortinget_no_cer: 5.9596
- validation_clean_stortinget_no_exact_wer: 12.2050
- validation_clean_stortinget_no_exact_cer: 6.4346

## 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: 2.5e-05
- lr_scheduler_type: linear
- per_device_train_batch_size: 32
- total_train_batch_size_per_node: 128
- total_train_batch_size: 1024
- total_optimization_steps: 250
- starting_optimization_step: None
- finishing_optimization_step: 250
- num_train_dataset_workers: 32
- num_hosts: 8
- total_num_training_examples: 256,000
- steps_per_epoch: _To be computed after first epoch_
- num_beams: None
- weight_decay: 0.01
- adam_beta1: 0.9
- adam_beta2: 0.98
- adam_epsilon: 1e-06
- dropout: True
- bpe_dropout_probability: 0.2
- activation_dropout_probability: 0.1

### Training results

| step | validation_nst_loss | train_loss | validation_nst_wer | validation_nst_cer | validation_nst_exact_wer | validation_nst_exact_cer | validation_clean_stortinget_no_loss | validation_clean_stortinget_no_wer | validation_clean_stortinget_no_cer | validation_clean_stortinget_no_exact_wer | validation_clean_stortinget_no_exact_cer |
|:----:|:-------------------:|:----------:|:------------------:|:------------------:|:------------------------:|:------------------------:|:-----------------------------------:|:----------------------------------:|:----------------------------------:|:----------------------------------------:|:----------------------------------------:|
| 0    | 0.4372              | 1.1581     | 2.2865             | 0.7038             | 2.9452                   | 0.8059                   | 0.7951                              | 8.8939                             | 5.6567                             | 11.9393                                  | 6.1348                                   |
| 40   | 0.7551              | 0.7309     | 2.7546             | 0.9443             | 3.4950                   | 1.0568                   | 0.7489                              | 9.8437                             | 6.4691                             | 12.9830                                  | 6.9559                                   |
| 80   | 0.7655              | 0.6341     | 2.5695             | 0.8250             | 3.3317                   | 0.9414                   | 0.7355                              | 9.3794                             | 6.1000                             | 12.4090                                  | 6.5664                                   |
| 120  | 0.7493              | 0.6262     | 2.4716             | 0.8129             | 3.2283                   | 0.9286                   | 0.7432                              | 9.3344                             | 6.0544                             | 12.3592                                  | 6.5247                                   |
| 160  | 0.7165              | 0.6353     | 2.5859             | 0.8259             | 3.3154                   | 0.9378                   | 0.7455                              | 9.3179                             | 6.0647                             | 12.3852                                  | 6.5418                                   |
| 200  | 0.6678              | 0.6169     | 2.4661             | 0.7812             | 3.1575                   | 0.8892                   | 0.7430                              | 9.2752                             | 6.0298                             | 12.3378                                  | 6.5079                                   |
| 240  | 0.6409              | 0.6385     | 2.4498             | 0.7849             | 3.1520                   | 0.8911                   | 0.7434                              | 9.1900                             | 5.9941                             | 12.1931                                  | 6.4639                                   |
| 249  | 0.6407              | 0.6265     | 2.4389             | 0.7737             | 3.1412                   | 0.8810                   |
| 249  | 0.7441              | 0.6265     | 9.1639             | 5.9596             | 12.2050                  | 6.4346                   |


### Framework versions

- Transformers 4.34.1
- Datasets 2.15.0
- Tokenizers 0.14.1