File size: 2,133 Bytes
3c5983e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ec9aef6
 
 
 
 
3c5983e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ae775ba
ec9aef6
3c5983e
 
 
 
 
 
ec9aef6
 
 
 
 
 
3c5983e
 
 
 
ae775ba
 
3c5983e
 
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: microsoft/deberta-v3-base
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: fine_tuned_deberta
  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. -->

# fine_tuned_deberta

This model is a fine-tuned version of [microsoft/deberta-v3-base](https://huggingface.co/microsoft/deberta-v3-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2283
- Accuracy: 0.9331
- F1: 0.9272
- Precision: 1.0
- Recall: 0.8643

## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 10
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 0.7017        | 0.96  | 17   | 0.6835          | 0.5352   | 0.1081 | 1.0       | 0.0571 |
| 0.6085        | 1.97  | 35   | 0.5872          | 0.6866   | 0.5822 | 0.8493    | 0.4429 |
| 0.518         | 2.99  | 53   | 0.4436          | 0.7958   | 0.8141 | 0.7384    | 0.9071 |
| 0.2366        | 4.0   | 71   | 0.2283          | 0.9331   | 0.9272 | 1.0       | 0.8643 |
| 0.1579        | 4.96  | 88   | 0.2696          | 0.9331   | 0.9294 | 0.9690    | 0.8929 |
| 0.1626        | 5.97  | 106  | 0.2726          | 0.9225   | 0.9179 | 0.9609    | 0.8786 |


### Framework versions

- Transformers 4.39.3
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2