File size: 7,948 Bytes
7805f07
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: mit
base_model: SCUT-DLVCLab/lilt-roberta-en-base
tags:
- generated_from_trainer
model-index:
- name: plus2_model
  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. -->

# plus2_model

This model is a fine-tuned version of [SCUT-DLVCLab/lilt-roberta-en-base](https://huggingface.co/SCUT-DLVCLab/lilt-roberta-en-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7698
- Eader: {'precision': 0.3232876712328767, 'recall': 0.2532188841201717, 'f1': 0.28399518652226236, 'number': 466}
- Nswer: {'precision': 0.43698252069917204, 'recall': 0.38183279742765275, 'f1': 0.4075504075504076, 'number': 1244}
- Uestion: {'precision': 0.392904953145917, 'recall': 0.425670775924583, 'f1': 0.40863209189001043, 'number': 1379}
- Overall Precision: 0.4005
- Overall Recall: 0.3820
- Overall F1: 0.3911
- Overall Accuracy: 0.5776

## 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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2500

### Training results

| Training Loss | Epoch | Step | Validation Loss | Eader                                                                                                       | Nswer                                                                                                        | Uestion                                                                                                      | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:-----------------------------------------------------------------------------------------------------------:|:------------------------------------------------------------------------------------------------------------:|:------------------------------------------------------------------------------------------------------------:|:-----------------:|:--------------:|:----------:|:----------------:|
| 0.1902        | 0.7   | 200  | 0.2840          | {'precision': 0.20205479452054795, 'recall': 0.12660944206008584, 'f1': 0.15567282321899734, 'number': 466} | {'precision': 0.2560738581146744, 'recall': 0.42363344051446944, 'f1': 0.3192004845548152, 'number': 1244}   | {'precision': 0.3292753623188406, 'recall': 0.41189267585206674, 'f1': 0.36597938144329895, 'number': 1379}  | 0.2832            | 0.3736         | 0.3222     | 0.6408           |
| 0.1315        | 1.4   | 400  | 0.3785          | {'precision': 0.2557544757033248, 'recall': 0.2145922746781116, 'f1': 0.23337222870478413, 'number': 466}   | {'precision': 0.3143491124260355, 'recall': 0.34163987138263663, 'f1': 0.327426810477658, 'number': 1244}    | {'precision': 0.31990924560408396, 'recall': 0.40899202320522116, 'f1': 0.35900700190961177, 'number': 1379} | 0.3106            | 0.3525         | 0.3303     | 0.5741           |
| 0.1117        | 2.11  | 600  | 0.5260          | {'precision': 0.33003300330033003, 'recall': 0.2145922746781116, 'f1': 0.2600780234070221, 'number': 466}   | {'precision': 0.37556973564266183, 'recall': 0.3311897106109325, 'f1': 0.35198633062793677, 'number': 1244}  | {'precision': 0.37993311036789296, 'recall': 0.41189267585206674, 'f1': 0.39526791927627, 'number': 1379}    | 0.3731            | 0.3496         | 0.3610     | 0.5467           |
| 0.0824        | 2.81  | 800  | 0.4716          | {'precision': 0.29289940828402367, 'recall': 0.21244635193133046, 'f1': 0.2462686567164179, 'number': 466}  | {'precision': 0.38620689655172413, 'recall': 0.36012861736334406, 'f1': 0.3727121464226289, 'number': 1244}  | {'precision': 0.36801541425818884, 'recall': 0.41551849166062366, 'f1': 0.3903269754768393, 'number': 1379}  | 0.3666            | 0.3626         | 0.3646     | 0.5599           |
| 0.0654        | 3.51  | 1000 | 0.3955          | {'precision': 0.3408360128617363, 'recall': 0.22746781115879827, 'f1': 0.2728442728442728, 'number': 466}   | {'precision': 0.3290094339622642, 'recall': 0.44855305466237944, 'f1': 0.37959183673469393, 'number': 1244}  | {'precision': 0.3595634692705342, 'recall': 0.45395213923132705, 'f1': 0.4012820512820513, 'number': 1379}   | 0.3442            | 0.4176         | 0.3774     | 0.6413           |
| 0.0701        | 4.21  | 1200 | 0.5372          | {'precision': 0.3501577287066246, 'recall': 0.23819742489270387, 'f1': 0.2835249042145594, 'number': 466}   | {'precision': 0.39559543230016314, 'recall': 0.38987138263665594, 'f1': 0.3927125506072874, 'number': 1244}  | {'precision': 0.3765156349712827, 'recall': 0.4278462654097172, 'f1': 0.40054310930074677, 'number': 1379}   | 0.3814            | 0.3839         | 0.3826     | 0.5792           |
| 0.048         | 4.91  | 1400 | 0.7393          | {'precision': 0.37168141592920356, 'recall': 0.18025751072961374, 'f1': 0.24277456647398846, 'number': 466} | {'precision': 0.4155972359328727, 'recall': 0.3384244372990354, 'f1': 0.3730615861763403, 'number': 1244}    | {'precision': 0.39614855570839064, 'recall': 0.4176939811457578, 'f1': 0.4066360748323332, 'number': 1379}   | 0.4014            | 0.3500         | 0.3739     | 0.5258           |
| 0.0372        | 5.61  | 1600 | 0.6617          | {'precision': 0.34726688102893893, 'recall': 0.2317596566523605, 'f1': 0.277992277992278, 'number': 466}    | {'precision': 0.390325271059216, 'recall': 0.3762057877813505, 'f1': 0.38313548915268114, 'number': 1244}    | {'precision': 0.39173228346456695, 'recall': 0.43292240754169686, 'f1': 0.4112986565621771, 'number': 1379}  | 0.3866            | 0.3797         | 0.3831     | 0.5713           |
| 0.0347        | 6.32  | 1800 | 0.7866          | {'precision': 0.3680297397769517, 'recall': 0.21244635193133046, 'f1': 0.2693877551020408, 'number': 466}   | {'precision': 0.42815814850530376, 'recall': 0.35691318327974275, 'f1': 0.38930293730819815, 'number': 1244} | {'precision': 0.39893617021276595, 'recall': 0.43509789702683105, 'f1': 0.4162330905306972, 'number': 1379}  | 0.4068            | 0.3700         | 0.3875     | 0.5454           |
| 0.0278        | 7.02  | 2000 | 0.6686          | {'precision': 0.34110787172011664, 'recall': 0.2510729613733906, 'f1': 0.2892459826946848, 'number': 466}   | {'precision': 0.4066115702479339, 'recall': 0.3954983922829582, 'f1': 0.40097799511002447, 'number': 1244}   | {'precision': 0.39158163265306123, 'recall': 0.4452501812907904, 'f1': 0.4166949440108585, 'number': 1379}   | 0.3919            | 0.3959         | 0.3939     | 0.6009           |
| 0.0233        | 7.72  | 2200 | 0.7673          | {'precision': 0.2876712328767123, 'recall': 0.2703862660944206, 'f1': 0.2787610619469027, 'number': 466}    | {'precision': 0.4287020109689214, 'recall': 0.3770096463022508, 'f1': 0.4011976047904192, 'number': 1244}    | {'precision': 0.38976109215017063, 'recall': 0.41406816533720087, 'f1': 0.4015471167369901, 'number': 1379}  | 0.3891            | 0.3775         | 0.3832     | 0.5676           |
| 0.0205        | 8.42  | 2400 | 0.7698          | {'precision': 0.3232876712328767, 'recall': 0.2532188841201717, 'f1': 0.28399518652226236, 'number': 466}   | {'precision': 0.43698252069917204, 'recall': 0.38183279742765275, 'f1': 0.4075504075504076, 'number': 1244}  | {'precision': 0.392904953145917, 'recall': 0.425670775924583, 'f1': 0.40863209189001043, 'number': 1379}     | 0.4005            | 0.3820         | 0.3911     | 0.5776           |


### Framework versions

- Transformers 4.34.0
- Pytorch 2.1.0.dev20230810
- Datasets 2.14.4
- Tokenizers 0.14.1