File size: 2,540 Bytes
849d901
 
79b7034
 
849d901
 
 
79b7034
 
849d901
 
 
 
 
79b7034
 
 
 
 
 
 
 
 
 
 
 
 
 
 
849d901
 
 
 
 
 
 
79b7034
849d901
79b7034
 
 
 
849d901
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
language:
- en
base_model: Hartunka/tiny_bert_km_50_v2
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
- f1
model-index:
- name: tiny_bert_km_50_v2_qqp
  results:
  - task:
      name: Text Classification
      type: text-classification
    dataset:
      name: GLUE QQP
      type: glue
      args: qqp
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.8069997526589167
    - name: F1
      type: f1
      value: 0.7304197616168595
---

<!-- 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. -->

# tiny_bert_km_50_v2_qqp

This model is a fine-tuned version of [Hartunka/tiny_bert_km_50_v2](https://huggingface.co/Hartunka/tiny_bert_km_50_v2) on the GLUE QQP dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4256
- Accuracy: 0.8070
- F1: 0.7304
- Combined Score: 0.7687

## 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: 256
- eval_batch_size: 256
- seed: 10
- 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

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Accuracy | F1     | Combined Score |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:------:|:--------------:|
| 0.4993        | 1.0   | 1422  | 0.4612          | 0.7790   | 0.6616 | 0.7203         |
| 0.4141        | 2.0   | 2844  | 0.4278          | 0.7986   | 0.7121 | 0.7554         |
| 0.3569        | 3.0   | 4266  | 0.4256          | 0.8070   | 0.7304 | 0.7687         |
| 0.3108        | 4.0   | 5688  | 0.4395          | 0.8137   | 0.7326 | 0.7732         |
| 0.273         | 5.0   | 7110  | 0.4447          | 0.8171   | 0.7535 | 0.7853         |
| 0.2403        | 6.0   | 8532  | 0.4629          | 0.8187   | 0.7591 | 0.7889         |
| 0.2117        | 7.0   | 9954  | 0.4907          | 0.8172   | 0.7564 | 0.7868         |
| 0.1871        | 8.0   | 11376 | 0.5383          | 0.8219   | 0.7543 | 0.7881         |


### Framework versions

- Transformers 4.50.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.21.1