Re-add my commit with my best results (9d65f1419f0e87aa21be9fdf72ca8bcb9a395367) after adding a task tag
Browse files
README.md
CHANGED
@@ -3,86 +3,35 @@ license: apache-2.0
|
|
3 |
base_model: ntu-spml/distilhubert
|
4 |
tags:
|
5 |
- generated_from_trainer
|
6 |
-
- audio-classification
|
7 |
datasets:
|
8 |
- marsyas/gtzan
|
9 |
-
|
10 |
-
|
11 |
model-index:
|
12 |
- name: distilhubert-finetuned-gtzan
|
13 |
-
results:
|
14 |
-
- task:
|
15 |
-
name: Audio Classification
|
16 |
-
type: audio-classification
|
17 |
-
dataset:
|
18 |
-
name: GTZAN
|
19 |
-
type: marsyas/gtzan
|
20 |
-
config: all
|
21 |
-
split: train
|
22 |
-
args: all
|
23 |
-
metrics:
|
24 |
-
- name: Accuracy
|
25 |
-
type: accuracy
|
26 |
-
value: 0.85
|
27 |
---
|
28 |
|
29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
30 |
-
should probably proofread and complete it, then remove this comment. -->
|
31 |
-
|
32 |
-
# distilhubert-finetuned-gtzan
|
33 |
|
34 |
This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
|
35 |
It achieves the following results on the evaluation set:
|
36 |
-
-
|
37 |
-
-
|
|
|
|
|
|
|
|
|
|
|
38 |
|
39 |
## Model description
|
40 |
|
41 |
-
|
42 |
-
|
43 |
-
## Intended uses & limitations
|
44 |
-
|
45 |
-
More information needed
|
46 |
-
|
47 |
-
## Training and evaluation data
|
48 |
-
|
49 |
-
More information needed
|
50 |
-
|
51 |
-
## Training procedure
|
52 |
-
|
53 |
-
### Training hyperparameters
|
54 |
-
|
55 |
-
The following hyperparameters were used during training:
|
56 |
-
- learning_rate: 5e-05
|
57 |
-
- train_batch_size: 8
|
58 |
-
- eval_batch_size: 8
|
59 |
-
- seed: 42
|
60 |
-
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
61 |
-
- lr_scheduler_type: linear
|
62 |
-
- lr_scheduler_warmup_ratio: 0.1
|
63 |
-
- num_epochs: 11
|
64 |
- mixed_precision_training: Native AMP
|
65 |
|
66 |
-
### Training results
|
67 |
-
|
68 |
-
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
69 |
-
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
70 |
-
| 1.9767 | 1.0 | 113 | 1.8825 | 0.53 |
|
71 |
-
| 1.1647 | 2.0 | 226 | 1.3428 | 0.62 |
|
72 |
-
| 1.0652 | 3.0 | 339 | 1.0057 | 0.7 |
|
73 |
-
| 0.6664 | 4.0 | 452 | 0.8889 | 0.72 |
|
74 |
-
| 0.5592 | 5.0 | 565 | 0.6754 | 0.81 |
|
75 |
-
| 0.3939 | 6.0 | 678 | 0.6415 | 0.84 |
|
76 |
-
| 0.2753 | 7.0 | 791 | 0.6098 | 0.84 |
|
77 |
-
| 0.1135 | 8.0 | 904 | 0.6292 | 0.83 |
|
78 |
-
| 0.1878 | 9.0 | 1017 | 0.5693 | 0.86 |
|
79 |
-
| 0.0864 | 10.0 | 1130 | 0.6220 | 0.83 |
|
80 |
-
| 0.0685 | 11.0 | 1243 | 0.5797 | 0.85 |
|
81 |
-
|
82 |
-
|
83 |
### Framework versions
|
84 |
|
85 |
- Transformers 4.44.0
|
86 |
- Pytorch 2.4.0
|
87 |
-
- Datasets
|
88 |
- Tokenizers 0.19.1
|
|
|
3 |
base_model: ntu-spml/distilhubert
|
4 |
tags:
|
5 |
- generated_from_trainer
|
|
|
6 |
datasets:
|
7 |
- marsyas/gtzan
|
8 |
+
|
9 |
+
|
10 |
model-index:
|
11 |
- name: distilhubert-finetuned-gtzan
|
12 |
+
results: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
---
|
14 |
|
15 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
|
|
|
16 |
|
17 |
This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
|
18 |
It achieves the following results on the evaluation set:
|
19 |
+
- eval_loss: 0.5534
|
20 |
+
- eval_accuracy: 0.88
|
21 |
+
- eval_runtime: 66.8648
|
22 |
+
- eval_samples_per_second: 1.496
|
23 |
+
- eval_steps_per_second: 0.194
|
24 |
+
- epoch: 1.0
|
25 |
+
- step: 113
|
26 |
|
27 |
## Model description
|
28 |
|
29 |
+
- num_epochs: 9
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
- mixed_precision_training: Native AMP
|
31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
### Framework versions
|
33 |
|
34 |
- Transformers 4.44.0
|
35 |
- Pytorch 2.4.0
|
36 |
+
- Datasets 2.21.0
|
37 |
- Tokenizers 0.19.1
|