File size: 2,672 Bytes
2acde3e
 
 
 
39c6739
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2acde3e
 
39c6739
 
2acde3e
39c6739
2acde3e
39c6739
 
 
 
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
2acde3e
39c6739
 
 
 
 
 
 
 
 
 
2acde3e
39c6739
2acde3e
39c6739
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2acde3e
 
39c6739
2acde3e
39c6739
 
 
 
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
---
language:
- en
license: mit
base_model: distil-whisper/distil-small.en
tags:
- generated_from_trainer
datasets:
- atc
metrics:
- wer
model-index:
- name: Whisper Large v3 1500 Epochs 2 - nullonesix
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: atc
      type: atc
      args: 'config: en, split: test'
    metrics:
    - name: Wer
      type: wer
      value: 39.23487544483986
---

<!-- 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 Large v3 1500 Epochs 2 - nullonesix

This model is a fine-tuned version of [distil-whisper/distil-small.en](https://huggingface.co/distil-whisper/distil-small.en) on the atc dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4151
- Wer: 39.2349

## 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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 1500
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer     |
|:-------------:|:-------:|:----:|:---------------:|:-------:|
| 2.8313        | 3.5714  | 100  | 2.7177          | 74.1548 |
| 1.1366        | 7.1429  | 200  | 1.6407          | 63.0338 |
| 0.4394        | 10.7143 | 300  | 1.4737          | 47.4644 |
| 0.1686        | 14.2857 | 400  | 1.4481          | 46.3968 |
| 0.0761        | 17.8571 | 500  | 1.3707          | 40.8808 |
| 0.0452        | 21.4286 | 600  | 1.4051          | 38.5231 |
| 0.0188        | 25.0    | 700  | 1.4044          | 36.7883 |
| 0.0167        | 28.5714 | 800  | 1.4217          | 38.8345 |
| 0.0084        | 32.1429 | 900  | 1.4120          | 48.5765 |
| 0.0033        | 35.7143 | 1000 | 1.4151          | 39.2349 |
| 0.0022        | 39.2857 | 1100 | 1.4401          | 39.7242 |
| 0.0008        | 42.8571 | 1200 | 1.4591          | 39.5907 |
| 0.0007        | 46.4286 | 1300 | 1.4679          | 39.5907 |
| 0.0006        | 50.0    | 1400 | 1.4724          | 39.8577 |
| 0.0007        | 53.5714 | 1500 | 1.4737          | 39.7242 |


### Framework versions

- Transformers 4.42.3
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1