update model card README.md
Browse files
README.md
CHANGED
@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
18 |
|
19 |
This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
|
20 |
It achieves the following results on the evaluation set:
|
21 |
-
- Loss:
|
22 |
-
- Accuracy: 0.
|
23 |
|
24 |
## Model description
|
25 |
|
@@ -39,40 +39,25 @@ More information needed
|
|
39 |
|
40 |
The following hyperparameters were used during training:
|
41 |
- learning_rate: 5e-05
|
42 |
-
- train_batch_size:
|
43 |
-
- eval_batch_size:
|
44 |
- seed: 42
|
45 |
- distributed_type: multi-GPU
|
46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
- lr_scheduler_type: linear
|
48 |
- lr_scheduler_warmup_ratio: 0.1
|
49 |
-
- num_epochs:
|
50 |
- mixed_precision_training: Native AMP
|
51 |
|
52 |
### Training results
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
55 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
56 |
-
|
|
57 |
-
| 1.
|
58 |
-
|
|
59 |
-
| 0.
|
60 |
-
| 0.
|
61 |
-
| 0.4422 | 6.0 | 1350 | 0.7370 | 0.76 |
|
62 |
-
| 0.1283 | 7.0 | 1575 | 0.7234 | 0.84 |
|
63 |
-
| 0.0076 | 8.0 | 1800 | 0.8727 | 0.85 |
|
64 |
-
| 0.0037 | 9.0 | 2025 | 0.9373 | 0.84 |
|
65 |
-
| 0.1723 | 10.0 | 2250 | 0.9524 | 0.86 |
|
66 |
-
| 0.0016 | 11.0 | 2475 | 1.0349 | 0.84 |
|
67 |
-
| 0.0016 | 12.0 | 2700 | 1.0471 | 0.85 |
|
68 |
-
| 0.0011 | 13.0 | 2925 | 1.0802 | 0.85 |
|
69 |
-
| 0.0009 | 14.0 | 3150 | 1.0722 | 0.85 |
|
70 |
-
| 0.0007 | 15.0 | 3375 | 1.0931 | 0.85 |
|
71 |
-
| 0.0007 | 16.0 | 3600 | 1.1442 | 0.85 |
|
72 |
-
| 0.0007 | 17.0 | 3825 | 1.1239 | 0.85 |
|
73 |
-
| 0.0005 | 18.0 | 4050 | 1.1810 | 0.85 |
|
74 |
-
| 0.0006 | 19.0 | 4275 | 1.1560 | 0.85 |
|
75 |
-
| 0.0005 | 20.0 | 4500 | 1.1326 | 0.86 |
|
76 |
|
77 |
|
78 |
### Framework versions
|
|
|
18 |
|
19 |
This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
|
20 |
It achieves the following results on the evaluation set:
|
21 |
+
- Loss: 0.7824
|
22 |
+
- Accuracy: 0.8
|
23 |
|
24 |
## Model description
|
25 |
|
|
|
39 |
|
40 |
The following hyperparameters were used during training:
|
41 |
- learning_rate: 5e-05
|
42 |
+
- train_batch_size: 8
|
43 |
+
- eval_batch_size: 8
|
44 |
- seed: 42
|
45 |
- distributed_type: multi-GPU
|
46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
- lr_scheduler_type: linear
|
48 |
- lr_scheduler_warmup_ratio: 0.1
|
49 |
+
- num_epochs: 5
|
50 |
- mixed_precision_training: Native AMP
|
51 |
|
52 |
### Training results
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
55 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
56 |
+
| 1.6129 | 1.0 | 113 | 1.7019 | 0.52 |
|
57 |
+
| 1.1591 | 2.0 | 226 | 1.1881 | 0.73 |
|
58 |
+
| 1.0244 | 3.0 | 339 | 0.9023 | 0.8 |
|
59 |
+
| 0.7501 | 4.0 | 452 | 0.8744 | 0.72 |
|
60 |
+
| 0.6153 | 5.0 | 565 | 0.7824 | 0.8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
61 |
|
62 |
|
63 |
### Framework versions
|