File size: 2,287 Bytes
1fb4368 |
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 |
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
metrics:
- accuracy
model-index:
- name: distilhubert-finetuned-gtzan-v3
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-v3
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: 0.5752
- Accuracy: 0.83
## 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: 15
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.9108 | 1.0 | 113 | 1.9472 | 0.43 |
| 1.3286 | 2.0 | 226 | 1.4173 | 0.65 |
| 1.032 | 3.0 | 339 | 0.9815 | 0.67 |
| 0.726 | 4.0 | 452 | 0.7403 | 0.79 |
| 0.4621 | 5.0 | 565 | 0.6390 | 0.8 |
| 0.3439 | 6.0 | 678 | 0.5248 | 0.85 |
| 0.1592 | 7.0 | 791 | 0.4861 | 0.86 |
| 0.1283 | 8.0 | 904 | 0.4995 | 0.87 |
| 0.1191 | 9.0 | 1017 | 0.4804 | 0.87 |
| 0.0236 | 10.0 | 1130 | 0.6737 | 0.8 |
| 0.0146 | 11.0 | 1243 | 0.6211 | 0.81 |
| 0.0105 | 12.0 | 1356 | 0.5806 | 0.86 |
| 0.008 | 13.0 | 1469 | 0.5645 | 0.84 |
| 0.0082 | 14.0 | 1582 | 0.6033 | 0.83 |
| 0.0072 | 15.0 | 1695 | 0.5752 | 0.83 |
### Framework versions
- Transformers 4.30.0
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
|