File size: 3,125 Bytes
1ebfc03
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
95
96
97
98
99
100
---

library_name: transformers
license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
metrics:
- accuracy
model-index:
- name: distilhubert-finetuned-gtzan
  results:
  - task:
      name: Audio Classification
      type: audio-classification
    dataset:
      name: GTZAN
      type: marsyas/gtzan
      config: all
      split: train
      args: all
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.8333333333333334
---


<!-- 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. -->

# distilhubert-finetuned-gtzan

This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7729
- Accuracy: 0.8333

## 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: 10

- eval_batch_size: 10

- seed: 42

- gradient_accumulation_steps: 2

- total_train_batch_size: 20
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1

- num_epochs: 20
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 2.188         | 1.0   | 35   | 2.1681          | 0.2692   |
| 1.887         | 2.0   | 70   | 1.8252          | 0.5769   |
| 1.5321        | 3.0   | 105  | 1.4375          | 0.5385   |
| 1.0946        | 4.0   | 140  | 1.2295          | 0.6282   |
| 0.9091        | 5.0   | 175  | 1.0390          | 0.6923   |
| 0.6839        | 6.0   | 210  | 0.9047          | 0.7821   |
| 0.5769        | 7.0   | 245  | 0.8309          | 0.7308   |
| 0.4118        | 8.0   | 280  | 0.9522          | 0.6538   |
| 0.3767        | 9.0   | 315  | 0.8164          | 0.7308   |
| 0.2247        | 10.0  | 350  | 0.6987          | 0.8205   |
| 0.1392        | 11.0  | 385  | 0.7565          | 0.7692   |
| 0.0886        | 12.0  | 420  | 0.7082          | 0.8205   |
| 0.0583        | 13.0  | 455  | 0.7529          | 0.8205   |
| 0.0383        | 14.0  | 490  | 0.7678          | 0.7949   |
| 0.0345        | 15.0  | 525  | 0.7480          | 0.8333   |
| 0.0269        | 16.0  | 560  | 0.7542          | 0.8333   |
| 0.0246        | 17.0  | 595  | 0.7550          | 0.8205   |
| 0.0233        | 18.0  | 630  | 0.7725          | 0.8333   |
| 0.0225        | 19.0  | 665  | 0.7701          | 0.8333   |
| 0.0225        | 20.0  | 700  | 0.7729          | 0.8333   |


### Framework versions

- Transformers 4.45.1
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.0