File size: 8,857 Bytes
160c5c1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f03e49c
 
 
 
 
 
160c5c1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f03e49c
160c5c1
 
 
 
 
 
 
f03e49c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
160c5c1
 
 
 
 
 
 
 
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
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
---

license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: hubert-classifier-aug-fold-0
  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. -->

# hubert-classifier-aug-fold-0

This model is a fine-tuned version of [facebook/hubert-base-ls960](https://huggingface.co/facebook/hubert-base-ls960) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7030
- Accuracy: 0.8571
- Precision: 0.8685
- Recall: 0.8571
- F1: 0.8514
- Binary: 0.9007

## 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: 0.0001

- train_batch_size: 32

- eval_batch_size: 32

- seed: 42

- gradient_accumulation_steps: 4

- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500

- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1     | Binary |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:------:|
| No log        | 0.13  | 50   | 4.4206          | 0.0243   | 0.0145    | 0.0243 | 0.0115 | 0.1888 |
| No log        | 0.27  | 100  | 4.3046          | 0.0458   | 0.0194    | 0.0458 | 0.0171 | 0.3013 |
| No log        | 0.4   | 150  | 3.9493          | 0.0714   | 0.0250    | 0.0714 | 0.0284 | 0.3407 |
| No log        | 0.54  | 200  | 3.7625          | 0.0768   | 0.0518    | 0.0768 | 0.0328 | 0.3255 |
| No log        | 0.67  | 250  | 3.3324          | 0.1644   | 0.0956    | 0.1644 | 0.0982 | 0.4112 |
| No log        | 0.81  | 300  | 3.0658          | 0.2466   | 0.1774    | 0.2466 | 0.1719 | 0.4704 |
| No log        | 0.94  | 350  | 2.7642          | 0.3450   | 0.2437    | 0.3450 | 0.2520 | 0.5375 |
| 3.7949        | 1.08  | 400  | 2.4739          | 0.3760   | 0.3261    | 0.3760 | 0.3038 | 0.5586 |
| 3.7949        | 1.21  | 450  | 2.1757          | 0.4757   | 0.3860    | 0.4757 | 0.4022 | 0.6303 |
| 3.7949        | 1.35  | 500  | 1.8422          | 0.5323   | 0.4744    | 0.5323 | 0.4677 | 0.6717 |
| 3.7949        | 1.48  | 550  | 1.6818          | 0.5889   | 0.5464    | 0.5889 | 0.5402 | 0.7100 |
| 3.7949        | 1.62  | 600  | 1.4944          | 0.6173   | 0.6054    | 0.6173 | 0.5761 | 0.7319 |
| 3.7949        | 1.75  | 650  | 1.3503          | 0.6429   | 0.6457    | 0.6429 | 0.6069 | 0.7481 |
| 3.7949        | 1.89  | 700  | 1.2159          | 0.6752   | 0.6724    | 0.6752 | 0.6372 | 0.7732 |
| 1.9673        | 2.02  | 750  | 1.0849          | 0.7049   | 0.7117    | 0.7049 | 0.6748 | 0.7930 |
| 1.9673        | 2.16  | 800  | 1.0424          | 0.7183   | 0.7257    | 0.7183 | 0.6987 | 0.8011 |
| 1.9673        | 2.29  | 850  | 0.8607          | 0.7830   | 0.7911    | 0.7830 | 0.7694 | 0.8470 |
| 1.9673        | 2.43  | 900  | 0.8606          | 0.7668   | 0.7813    | 0.7668 | 0.7506 | 0.8372 |
| 1.9673        | 2.56  | 950  | 0.7939          | 0.7803   | 0.7739    | 0.7803 | 0.7606 | 0.8487 |
| 1.9673        | 2.7   | 1000 | 0.7883          | 0.8059   | 0.8257    | 0.8059 | 0.7976 | 0.8656 |
| 1.9673        | 2.83  | 1050 | 0.7567          | 0.8032   | 0.8214    | 0.8032 | 0.7947 | 0.8640 |
| 1.9673        | 2.97  | 1100 | 0.6989          | 0.8181   | 0.8399    | 0.8181 | 0.8063 | 0.8745 |
| 1.0987        | 3.1   | 1150 | 0.7500          | 0.8100   | 0.8223    | 0.8100 | 0.8043 | 0.8660 |
| 1.0987        | 3.24  | 1200 | 0.6802          | 0.8261   | 0.8381    | 0.8261 | 0.8184 | 0.8794 |
| 1.0987        | 3.37  | 1250 | 0.6614          | 0.8396   | 0.8558    | 0.8396 | 0.8359 | 0.8877 |
| 1.0987        | 3.51  | 1300 | 0.6928          | 0.8261   | 0.8511    | 0.8261 | 0.8236 | 0.8791 |
| 1.0987        | 3.64  | 1350 | 0.6146          | 0.8410   | 0.8588    | 0.8410 | 0.8401 | 0.8896 |
| 1.0987        | 3.78  | 1400 | 0.6958          | 0.8248   | 0.8412    | 0.8248 | 0.8191 | 0.8796 |
| 1.0987        | 3.91  | 1450 | 0.6785          | 0.8342   | 0.8556    | 0.8342 | 0.8309 | 0.8857 |
| 0.7483        | 4.05  | 1500 | 0.7412          | 0.8261   | 0.8461    | 0.8261 | 0.8244 | 0.8784 |
| 0.7483        | 4.18  | 1550 | 0.6778          | 0.8356   | 0.8538    | 0.8356 | 0.8317 | 0.8868 |
| 0.7483        | 4.32  | 1600 | 0.7032          | 0.8437   | 0.8657    | 0.8437 | 0.8405 | 0.8946 |
| 0.7483        | 4.45  | 1650 | 0.7373          | 0.8329   | 0.8564    | 0.8329 | 0.8299 | 0.8850 |
| 0.7483        | 4.59  | 1700 | 0.6958          | 0.8423   | 0.8593    | 0.8423 | 0.8401 | 0.8915 |
| 0.7483        | 4.72  | 1750 | 0.7395          | 0.8329   | 0.8513    | 0.8329 | 0.8327 | 0.8865 |
| 0.7483        | 4.86  | 1800 | 0.7017          | 0.8477   | 0.8651    | 0.8477 | 0.8453 | 0.8953 |
| 0.7483        | 4.99  | 1850 | 0.7240          | 0.8423   | 0.8582    | 0.8423 | 0.8410 | 0.8922 |
| 0.5887        | 5.12  | 1900 | 0.6810          | 0.8464   | 0.8694    | 0.8464 | 0.8451 | 0.8943 |
| 0.5887        | 5.26  | 1950 | 0.6091          | 0.8706   | 0.8828    | 0.8706 | 0.8688 | 0.9111 |
| 0.5887        | 5.39  | 2000 | 0.6617          | 0.8491   | 0.8669    | 0.8491 | 0.8474 | 0.8974 |
| 0.5887        | 5.53  | 2050 | 0.6712          | 0.8477   | 0.8662    | 0.8477 | 0.8458 | 0.8966 |
| 0.5887        | 5.66  | 2100 | 0.6988          | 0.8437   | 0.8570    | 0.8437 | 0.8413 | 0.8915 |
| 0.5887        | 5.8   | 2150 | 0.6644          | 0.8477   | 0.8624    | 0.8477 | 0.8455 | 0.8953 |
| 0.5887        | 5.93  | 2200 | 0.6416          | 0.8652   | 0.8790    | 0.8652 | 0.8622 | 0.9073 |
| 0.485         | 6.07  | 2250 | 0.6484          | 0.8585   | 0.8705    | 0.8585 | 0.8568 | 0.9030 |
| 0.485         | 6.2   | 2300 | 0.6690          | 0.8585   | 0.8736    | 0.8585 | 0.8564 | 0.9019 |
| 0.485         | 6.34  | 2350 | 0.6469          | 0.8639   | 0.8790    | 0.8639 | 0.8616 | 0.9071 |
| 0.485         | 6.47  | 2400 | 0.7418          | 0.8518   | 0.8684    | 0.8518 | 0.8515 | 0.8968 |
| 0.485         | 6.61  | 2450 | 0.6821          | 0.8625   | 0.8788    | 0.8625 | 0.8615 | 0.9075 |
| 0.485         | 6.74  | 2500 | 0.7012          | 0.8652   | 0.8837    | 0.8652 | 0.8636 | 0.9090 |
| 0.485         | 6.88  | 2550 | 0.6546          | 0.8652   | 0.8819    | 0.8652 | 0.8658 | 0.9085 |
| 0.4243        | 7.01  | 2600 | 0.6619          | 0.8639   | 0.8769    | 0.8639 | 0.8631 | 0.9070 |
| 0.4243        | 7.15  | 2650 | 0.7000          | 0.8531   | 0.8716    | 0.8531 | 0.8518 | 0.8995 |
| 0.4243        | 7.28  | 2700 | 0.6560          | 0.8720   | 0.8864    | 0.8720 | 0.8715 | 0.9112 |
| 0.4243        | 7.42  | 2750 | 0.6458          | 0.8639   | 0.8786    | 0.8639 | 0.8634 | 0.9074 |
| 0.4243        | 7.55  | 2800 | 0.6701          | 0.8747   | 0.8896    | 0.8747 | 0.8742 | 0.9155 |
| 0.4243        | 7.69  | 2850 | 0.7282          | 0.8477   | 0.8706    | 0.8477 | 0.8477 | 0.8970 |
| 0.4243        | 7.82  | 2900 | 0.6578          | 0.8612   | 0.8726    | 0.8612 | 0.8597 | 0.9061 |
| 0.4243        | 7.96  | 2950 | 0.6244          | 0.8720   | 0.8829    | 0.8720 | 0.8704 | 0.9142 |
| 0.378         | 8.09  | 3000 | 0.6445          | 0.8733   | 0.8896    | 0.8733 | 0.8726 | 0.9140 |
| 0.378         | 8.23  | 3050 | 0.6983          | 0.8612   | 0.8766    | 0.8612 | 0.8606 | 0.9055 |
| 0.378         | 8.36  | 3100 | 0.6355          | 0.8760   | 0.8922    | 0.8760 | 0.8750 | 0.9154 |
| 0.378         | 8.5   | 3150 | 0.6770          | 0.8747   | 0.8883    | 0.8747 | 0.8726 | 0.9135 |
| 0.378         | 8.63  | 3200 | 0.6472          | 0.8706   | 0.8798    | 0.8706 | 0.8680 | 0.9097 |
| 0.378         | 8.77  | 3250 | 0.7739          | 0.8544   | 0.8691    | 0.8544 | 0.8512 | 0.8970 |
| 0.378         | 8.9   | 3300 | 0.6805          | 0.8612   | 0.8766    | 0.8612 | 0.8587 | 0.9046 |
| 0.3491        | 9.04  | 3350 | 0.6382          | 0.8733   | 0.8829    | 0.8733 | 0.8717 | 0.9135 |
| 0.3491        | 9.17  | 3400 | 0.6927          | 0.8652   | 0.8793    | 0.8652 | 0.8642 | 0.9080 |
| 0.3491        | 9.31  | 3450 | 0.8407          | 0.8518   | 0.8711    | 0.8518 | 0.8488 | 0.8996 |
| 0.3491        | 9.44  | 3500 | 0.6628          | 0.8733   | 0.8823    | 0.8733 | 0.8714 | 0.9121 |


### Framework versions

- Transformers 4.38.2
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.15.1