File size: 1,801 Bytes
d79e38c
 
 
 
 
 
 
 
 
1d9adf4
d79e38c
 
 
 
 
 
 
 
 
1d9adf4
d79e38c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1d9adf4
 
d79e38c
 
 
 
 
1d9adf4
d79e38c
 
 
1d9adf4
 
 
 
 
 
 
 
 
 
 
d79e38c
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
model-index:
- name: distilhubert-finetuned-gtzan
  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. -->

# 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: 0.4454

## 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: 9.349509030319398e-05
- train_batch_size: 12
- 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: 9

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.904         | 1.0   | 75   | 1.7595          |
| 1.2214        | 2.0   | 150  | 1.1147          |
| 0.862         | 3.0   | 225  | 0.7765          |
| 0.6679        | 4.0   | 300  | 0.6600          |
| 0.4188        | 5.0   | 375  | 0.4797          |
| 0.3369        | 6.0   | 450  | 0.5607          |
| 0.1591        | 7.0   | 525  | 0.4668          |
| 0.0591        | 8.0   | 600  | 0.4493          |
| 0.0718        | 9.0   | 675  | 0.4454          |


### Framework versions

- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3