File size: 2,365 Bytes
db46acc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- speech_commands
metrics:
- accuracy
model-index:
- name: whisper-tiny-finetuned-no-go-kws
  results:
  - task:
      name: Audio Classification
      type: audio-classification
    dataset:
      name: Speech Commands[no, go]
      type: speech_commands
      config: v0.02
      split: test
      args: v0.02
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.990086741016109
---

<!-- 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-tiny-finetuned-no-go-kws

This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Speech Commands[no, go] dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0842
- Accuracy: 0.9901

## 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: 5e-05
- train_batch_size: 8
- 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_ratio: 0.1
- num_epochs: 10
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.33          | 1.0   | 780  | 0.0272          | 0.9938   |
| 0.0002        | 2.0   | 1560 | 0.0420          | 0.9876   |
| 0.0001        | 3.0   | 2340 | 0.0487          | 0.9913   |
| 0.0011        | 4.0   | 3120 | 0.0789          | 0.9802   |
| 0.0001        | 5.0   | 3900 | 0.0915          | 0.9851   |
| 0.0014        | 6.0   | 4680 | 0.1017          | 0.9839   |
| 0.0           | 7.0   | 5460 | 0.0993          | 0.9888   |
| 0.0           | 8.0   | 6240 | 0.0694          | 0.9913   |
| 0.0           | 9.0   | 7020 | 0.0760          | 0.9926   |
| 0.0           | 10.0  | 7800 | 0.0842          | 0.9901   |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0