Hugging Face's logo

nithiroj
/
distilhubert-finetuned-gtzan

Audio Classification
Transformers
PyTorch
hubert
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
nithiroj commited on
Commit
c519480
•
1 Parent(s): 60a8335

End of training

Browse files
Files changed (2) hide show
  1. README.md +21 -19
  2. pytorch_model.bin +1 -1
README.md CHANGED
@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.8069
36
  - Accuracy: 0.83
37
 
38
  ## Model description
@@ -53,9 +53,11 @@ More information needed
53
 
54
  The following hyperparameters were used during training:
55
  - learning_rate: 5e-05
56
- - train_batch_size: 8
57
- - eval_batch_size: 8
58
  - seed: 42
 
 
59
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
  - lr_scheduler_type: linear
61
  - lr_scheduler_warmup_ratio: 0.1
@@ -65,26 +67,26 @@ The following hyperparameters were used during training:
65
 
66
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
67
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
68
- | 2.1133 | 1.0 | 113 | 1.9482 | 0.42 |
69
- | 1.3938 | 2.0 | 226 | 1.2709 | 0.72 |
70
- | 1.098 | 3.0 | 339 | 0.9370 | 0.76 |
71
- | 0.849 | 4.0 | 452 | 0.9364 | 0.74 |
72
- | 0.5107 | 5.0 | 565 | 0.7350 | 0.73 |
73
- | 0.364 | 6.0 | 678 | 0.7841 | 0.75 |
74
- | 0.414 | 7.0 | 791 | 0.6663 | 0.82 |
75
- | 0.2109 | 8.0 | 904 | 0.6041 | 0.84 |
76
- | 0.1428 | 9.0 | 1017 | 0.6227 | 0.83 |
77
- | 0.1046 | 10.0 | 1130 | 0.6971 | 0.84 |
78
- | 0.0251 | 11.0 | 1243 | 0.7502 | 0.82 |
79
- | 0.0522 | 12.0 | 1356 | 0.7742 | 0.82 |
80
- | 0.0116 | 13.0 | 1469 | 0.7442 | 0.85 |
81
- | 0.0081 | 14.0 | 1582 | 0.7850 | 0.84 |
82
- | 0.0094 | 15.0 | 1695 | 0.8069 | 0.83 |
83
 
84
 
85
  ### Framework versions
86
 
87
- - Transformers 4.32.0
88
  - Pytorch 2.0.1+cu118
89
  - Datasets 2.14.4
90
  - Tokenizers 0.13.3
 
32
 
33
  This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.7984
36
  - Accuracy: 0.83
37
 
38
  ## Model description
 
53
 
54
  The following hyperparameters were used during training:
55
  - learning_rate: 5e-05
56
+ - train_batch_size: 4
57
+ - eval_batch_size: 4
58
  - seed: 42
59
+ - gradient_accumulation_steps: 2
60
+ - total_train_batch_size: 8
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_ratio: 0.1
 
67
 
68
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
69
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
70
+ | 2.1366 | 1.0 | 112 | 2.0054 | 0.29 |
71
+ | 1.4511 | 2.0 | 225 | 1.3306 | 0.63 |
72
+ | 0.941 | 3.0 | 337 | 0.9689 | 0.7 |
73
+ | 0.7428 | 4.0 | 450 | 0.9399 | 0.69 |
74
+ | 0.4607 | 5.0 | 562 | 0.8357 | 0.73 |
75
+ | 0.3398 | 6.0 | 675 | 0.7089 | 0.81 |
76
+ | 0.2055 | 7.0 | 787 | 0.7338 | 0.81 |
77
+ | 0.1662 | 8.0 | 900 | 0.7718 | 0.81 |
78
+ | 0.0939 | 9.0 | 1012 | 0.6441 | 0.85 |
79
+ | 0.028 | 10.0 | 1125 | 0.7501 | 0.83 |
80
+ | 0.0144 | 11.0 | 1237 | 0.8078 | 0.82 |
81
+ | 0.01 | 12.0 | 1350 | 0.7597 | 0.83 |
82
+ | 0.0095 | 13.0 | 1462 | 0.7818 | 0.83 |
83
+ | 0.0092 | 14.0 | 1575 | 0.7901 | 0.83 |
84
+ | 0.0078 | 14.93 | 1680 | 0.7984 | 0.83 |
85
 
86
 
87
  ### Framework versions
88
 
89
+ - Transformers 4.32.1
90
  - Pytorch 2.0.1+cu118
91
  - Datasets 2.14.4
92
  - Tokenizers 0.13.3
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3c1042122a42cbf9d24376ffa1fc90ca9d8e5d32b3f7b65d85356aab74074184
3
  size 94783376
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf9e2c21d7b9fb996b6c027ba1cb97ca8590fcfdec7f6359ab8bc0c39a940902
3
  size 94783376