File size: 2,547 Bytes
3369529
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
language:
- hi
license: apache-2.0
base_model: openai/whisper-small
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- aihub_adult_baseline
model-index:
- name: whisper-small-Yfreq_speed_pause
  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. -->

# whisper-small-Yfreq_speed_pause

This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the aihub old10 adult speed pause changed dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2307
- Cer: 7.4836

## 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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 2
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Cer    |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 0.4388        | 0.1289 | 100  | 0.3025          | 7.1957 |
| 0.2656        | 0.2579 | 200  | 0.2567          | 6.9725 |
| 0.2193        | 0.3868 | 300  | 0.2541          | 7.3073 |
| 0.2021        | 0.5158 | 400  | 0.2426          | 7.2075 |
| 0.1913        | 0.6447 | 500  | 0.2465          | 7.6539 |
| 0.1617        | 0.7737 | 600  | 0.2449          | 7.0371 |
| 0.1561        | 0.9026 | 700  | 0.2403          | 7.1546 |
| 0.0734        | 1.0316 | 800  | 0.2329          | 7.5070 |
| 0.0571        | 1.1605 | 900  | 0.2329          | 7.4424 |
| 0.0631        | 1.2895 | 1000 | 0.2330          | 7.5952 |
| 0.06          | 1.4184 | 1100 | 0.2358          | 7.5129 |
| 0.0655        | 1.5474 | 1200 | 0.2305          | 7.4659 |
| 0.0611        | 1.6763 | 1300 | 0.2303          | 7.3308 |
| 0.0574        | 1.8053 | 1400 | 0.2308          | 7.5482 |
| 0.0615        | 1.9342 | 1500 | 0.2307          | 7.4836 |


### Framework versions

- Transformers 4.46.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.1