File size: 4,948 Bytes
59081f4
 
 
 
52d9354
 
59081f4
 
 
 
52d9354
 
 
 
 
 
 
 
 
 
 
59081f4
 
 
 
 
 
 
52d9354
59081f4
52d9354
 
59081f4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: facebook/bart-base
tags:
- generated_from_trainer
datasets:
- datasets/all_binary_and_xe_ey_fae_counterfactual
metrics:
- accuracy
model-index:
- name: bart-base-finetuned-xe_ey_fae
  results:
  - task:
      name: Masked Language Modeling
      type: fill-mask
    dataset:
      name: datasets/all_binary_and_xe_ey_fae_counterfactual
      type: datasets/all_binary_and_xe_ey_fae_counterfactual
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.7180178883360112
---

<!-- 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. -->

# bart-base-finetuned-xe_ey_fae

This model is a fine-tuned version of [facebook/bart-base](https://huggingface.co/facebook/bart-base) on the datasets/all_binary_and_xe_ey_fae_counterfactual dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3945
- Accuracy: 0.7180

## 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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 100
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 5.4226        | 0.06  | 500   | 3.8138          | 0.3628   |
| 4.0408        | 0.12  | 1000  | 3.0576          | 0.4630   |
| 3.4979        | 0.18  | 1500  | 2.7016          | 0.5133   |
| 3.1691        | 0.24  | 2000  | 2.4880          | 0.5431   |
| 2.9564        | 0.3   | 2500  | 2.3309          | 0.5644   |
| 2.8078        | 0.35  | 3000  | 2.2320          | 0.5792   |
| 2.6741        | 0.41  | 3500  | 2.1506          | 0.5924   |
| 2.5323        | 0.47  | 4000  | 1.9846          | 0.6176   |
| 2.3678        | 0.53  | 4500  | 1.8813          | 0.6375   |
| 2.25          | 0.59  | 5000  | 1.8100          | 0.6497   |
| 2.1795        | 0.65  | 5500  | 1.7632          | 0.6579   |
| 2.1203        | 0.71  | 6000  | 1.7238          | 0.6646   |
| 2.0764        | 0.77  | 6500  | 1.6856          | 0.6713   |
| 2.026         | 0.83  | 7000  | 1.6569          | 0.6760   |
| 1.9942        | 0.89  | 7500  | 1.6309          | 0.6803   |
| 1.9665        | 0.95  | 8000  | 1.6122          | 0.6836   |
| 1.9395        | 1.0   | 8500  | 1.5913          | 0.6866   |
| 1.9155        | 1.06  | 9000  | 1.5758          | 0.6895   |
| 1.8828        | 1.12  | 9500  | 1.5607          | 0.6918   |
| 1.8721        | 1.18  | 10000 | 1.5422          | 0.6948   |
| 1.8474        | 1.24  | 10500 | 1.5320          | 0.6964   |
| 1.8293        | 1.3   | 11000 | 1.5214          | 0.6978   |
| 1.8129        | 1.36  | 11500 | 1.5102          | 0.6998   |
| 1.8148        | 1.42  | 12000 | 1.5010          | 0.7013   |
| 1.7903        | 1.48  | 12500 | 1.4844          | 0.7038   |
| 1.7815        | 1.54  | 13000 | 1.4823          | 0.7039   |
| 1.7637        | 1.6   | 13500 | 1.4746          | 0.7052   |
| 1.7623        | 1.66  | 14000 | 1.4701          | 0.7061   |
| 1.7402        | 1.71  | 14500 | 1.4598          | 0.7076   |
| 1.7376        | 1.77  | 15000 | 1.4519          | 0.7090   |
| 1.7287        | 1.83  | 15500 | 1.4501          | 0.7101   |
| 1.7273        | 1.89  | 16000 | 1.4409          | 0.7107   |
| 1.7119        | 1.95  | 16500 | 1.4314          | 0.7125   |
| 1.7098        | 2.01  | 17000 | 1.4269          | 0.7129   |
| 1.6978        | 2.07  | 17500 | 1.4275          | 0.7132   |
| 1.698         | 2.13  | 18000 | 1.4218          | 0.7140   |
| 1.6837        | 2.19  | 18500 | 1.4151          | 0.7147   |
| 1.6908        | 2.25  | 19000 | 1.4137          | 0.7149   |
| 1.6902        | 2.31  | 19500 | 1.4085          | 0.7161   |
| 1.6741        | 2.36  | 20000 | 1.4121          | 0.7154   |
| 1.6823        | 2.42  | 20500 | 1.4037          | 0.7165   |
| 1.6692        | 2.48  | 21000 | 1.4039          | 0.7164   |
| 1.6669        | 2.54  | 21500 | 1.4015          | 0.7172   |
| 1.6613        | 2.6   | 22000 | 1.3979          | 0.7179   |
| 1.664         | 2.66  | 22500 | 1.3960          | 0.7180   |
| 1.6615        | 2.72  | 23000 | 1.4012          | 0.7172   |
| 1.6627        | 2.78  | 23500 | 1.3974          | 0.7178   |
| 1.6489        | 2.84  | 24000 | 1.3948          | 0.7182   |
| 1.6429        | 2.9   | 24500 | 1.3921          | 0.7184   |
| 1.6477        | 2.96  | 25000 | 1.3910          | 0.7182   |


### Framework versions

- Transformers 4.36.2
- Pytorch 2.2.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2