File size: 2,859 Bytes
68eebd9
3655884
 
68eebd9
 
 
 
 
 
 
 
 
 
 
 
 
 
3655884
68eebd9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3655884
68eebd9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- en
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
- f1
model-index:
- name: baseline-ft-mrpc-IRoberta-b-8bit
  results:
  - task:
      name: Text Classification
      type: text-classification
    dataset:
      name: GLUE MRPC
      type: glue
      config: mrpc
      split: validation
      args: mrpc
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.8970588235294118
    - name: F1
      type: f1
      value: 0.9257950530035336
---

<!-- 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. -->

# baseline-ft-mrpc-IRoberta-b-8bit

This model is a fine-tuned version of [vuiseng9/baseline-ft-mrpc-IRoberta-b-unquantized](https://huggingface.co/vuiseng9/baseline-ft-mrpc-IRoberta-b-unquantized) on the GLUE MRPC dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3871
- Accuracy: 0.8971
- F1: 0.9258
- Combined Score: 0.9114

## 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-07
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 12.0

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     | Combined Score |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:--------------:|
| 0.0021        | 1.0   | 230  | 0.4017          | 0.8848   | 0.9147 | 0.8998         |
| 0.0026        | 2.0   | 460  | 0.4105          | 0.8873   | 0.9173 | 0.9023         |
| 0.0026        | 3.0   | 690  | 0.3707          | 0.8946   | 0.9236 | 0.9091         |
| 0.0037        | 4.0   | 920  | 0.3893          | 0.8946   | 0.9228 | 0.9087         |
| 1.324         | 5.0   | 1150 | 0.3871          | 0.8897   | 0.9204 | 0.9050         |
| 0.0227        | 6.0   | 1380 | 0.3951          | 0.8897   | 0.9201 | 0.9049         |
| 0.0081        | 7.0   | 1610 | 0.3818          | 0.8824   | 0.9155 | 0.8989         |
| 0.0054        | 8.0   | 1840 | 0.3902          | 0.8873   | 0.9181 | 0.9027         |
| 0.0383        | 9.0   | 2070 | 0.3659          | 0.8922   | 0.9225 | 0.9073         |
| 0.3861        | 10.0  | 2300 | 0.4260          | 0.8652   | 0.9030 | 0.8841         |
| 0.0028        | 11.0  | 2530 | 0.3619          | 0.8946   | 0.9234 | 0.9090         |
| 0.0957        | 12.0  | 2760 | 0.3871          | 0.8971   | 0.9258 | 0.9114         |


### Framework versions

- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3