File size: 4,702 Bytes
24393e2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
license: apache-2.0
base_model: distilbert-base-uncased-distilled-squad
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: distilbert-qasports
  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. -->

# distilbert-qasports

This model is a fine-tuned version of [distilbert-base-uncased-distilled-squad](https://huggingface.co/distilbert-base-uncased-distilled-squad) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4015
- Exact: 76.8499
- F1: 81.2744
- Total: 15041
- Hasans Exact: 76.8499
- Hasans F1: 81.2744
- Hasans Total: 15041
- Best Exact: 76.8499
- Best Exact Thresh: 0.0
- Best F1: 81.2744
- Best F1 Thresh: 0.0

## 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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Exact   | F1      | Total | Hasans Exact | Hasans F1 | Hasans Total | Best Exact | Best Exact Thresh | Best F1 | Best F1 Thresh |
|:-------------:|:------:|:----:|:---------------:|:-------:|:-------:|:-----:|:------------:|:---------:|:------------:|:----------:|:-----------------:|:-------:|:--------------:|
| 0.6883        | 0.1325 | 500  | 0.6008          | 74.7690 | 79.7676 | 15041 | 74.7690      | 79.7676   | 15041        | 74.7690    | 0.0               | 79.7676 | 0.0            |
| 0.5738        | 0.2649 | 1000 | 0.5474          | 75.3407 | 80.2816 | 15041 | 75.3407      | 80.2816   | 15041        | 75.3407    | 0.0               | 80.2816 | 0.0            |
| 0.5853        | 0.3974 | 1500 | 0.5259          | 75.3873 | 80.2217 | 15041 | 75.3873      | 80.2217   | 15041        | 75.3873    | 0.0               | 80.2217 | 0.0            |
| 0.588         | 0.5298 | 2000 | 0.4904          | 76.3978 | 81.0881 | 15041 | 76.3978      | 81.0881   | 15041        | 76.3978    | 0.0               | 81.0881 | 0.0            |
| 0.5214        | 0.6623 | 2500 | 0.4764          | 76.8366 | 81.4327 | 15041 | 76.8366      | 81.4327   | 15041        | 76.8366    | 0.0               | 81.4327 | 0.0            |
| 0.4813        | 0.7947 | 3000 | 0.4586          | 76.9763 | 81.6042 | 15041 | 76.9763      | 81.6042   | 15041        | 76.9763    | 0.0               | 81.6042 | 0.0            |
| 0.5032        | 0.9272 | 3500 | 0.4323          | 76.7835 | 81.4041 | 15041 | 76.7835      | 81.4041   | 15041        | 76.7835    | 0.0               | 81.4041 | 0.0            |
| 0.3549        | 1.0596 | 4000 | 0.4349          | 76.8632 | 81.2899 | 15041 | 76.8632      | 81.2899   | 15041        | 76.8632    | 0.0               | 81.2899 | 0.0            |
| 0.4053        | 1.1921 | 4500 | 0.4199          | 76.9630 | 81.3741 | 15041 | 76.9630      | 81.3741   | 15041        | 76.9630    | 0.0               | 81.3741 | 0.0            |
| 0.3549        | 1.3245 | 5000 | 0.4372          | 77.0427 | 81.6167 | 15041 | 77.0427      | 81.6167   | 15041        | 77.0427    | 0.0               | 81.6167 | 0.0            |
| 0.3707        | 1.4570 | 5500 | 0.4254          | 77.0560 | 81.5058 | 15041 | 77.0560      | 81.5058   | 15041        | 77.0560    | 0.0               | 81.5058 | 0.0            |
| 0.3728        | 1.5894 | 6000 | 0.4086          | 76.9031 | 81.4012 | 15041 | 76.9031      | 81.4012   | 15041        | 76.9031    | 0.0               | 81.4012 | 0.0            |
| 0.4117        | 1.7219 | 6500 | 0.4029          | 76.8233 | 81.4108 | 15041 | 76.8233      | 81.4108   | 15041        | 76.8233    | 0.0               | 81.4108 | 0.0            |
| 0.3785        | 1.8543 | 7000 | 0.3979          | 77.0427 | 81.4664 | 15041 | 77.0427      | 81.4664   | 15041        | 77.0427    | 0.0               | 81.4664 | 0.0            |
| 0.3564        | 1.9868 | 7500 | 0.4015          | 76.8499 | 81.2744 | 15041 | 76.8499      | 81.2744   | 15041        | 76.8499    | 0.0               | 81.2744 | 0.0            |


### Framework versions

- Transformers 4.48.2
- Pytorch 2.0.0+cu117
- Datasets 3.2.0
- Tokenizers 0.21.0