File size: 4,136 Bytes
a89faf2
 
 
 
 
 
 
 
 
0dea54f
a89faf2
 
 
 
 
 
 
 
0dea54f
 
 
 
 
 
 
 
a89faf2
 
 
 
0dea54f
 
 
 
 
 
 
a89faf2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
26fe470
a89faf2
 
 
 
 
0dea54f
a89faf2
 
 
0dea54f
 
 
 
 
 
 
 
 
 
a89faf2
 
 
 
26fe470
 
 
 
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
---
license: mit
base_model: Amna100/PreTraining-MLM
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: fold_4
  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. -->

[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/rw6sjeap)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/g4pyaj7k)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/t4il24wd)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/qf2ywrxq)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/9xmjfnoc)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/vp363qmp)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/10xzvwgi)
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/wbresearch/FineTuning-ADE-DropOUT/runs/i1j07et7)
# fold_4

This model is a fine-tuned version of [Amna100/PreTraining-MLM](https://huggingface.co/Amna100/PreTraining-MLM) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0096
- Precision: 0.7641
- Recall: 0.6825
- F1: 0.7210
- Accuracy: 0.9995
- Roc Auc: 0.9961
- Pr Auc: 0.9999

## 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: 5
- eval_batch_size: 5
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy | Roc Auc | Pr Auc |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|:-------:|:------:|
| 0.0278        | 1.0   | 632  | 0.0107          | 0.5932    | 0.7270 | 0.6533 | 0.9993   | 0.9948  | 0.9999 |
| 0.01          | 2.0   | 1264 | 0.0096          | 0.7641    | 0.6825 | 0.7210 | 0.9995   | 0.9961  | 0.9999 |
| 0.0056        | 3.0   | 1896 | 0.0131          | 0.7207    | 0.6202 | 0.6667 | 0.9994   | 0.9827  | 0.9997 |
| 0.0025        | 4.0   | 2528 | 0.0114          | 0.7937    | 0.6736 | 0.7287 | 0.9995   | 0.9860  | 0.9997 |
| 0.0012        | 5.0   | 3160 | 0.0141          | 0.7727    | 0.6558 | 0.7095 | 0.9995   | 0.9860  | 0.9998 |
| 0.0008        | 6.0   | 3792 | 0.0126          | 0.7659    | 0.6795 | 0.7201 | 0.9995   | 0.9894  | 0.9998 |
| 0.0007        | 7.0   | 4424 | 0.0146          | 0.7845    | 0.6588 | 0.7161 | 0.9995   | 0.9917  | 0.9998 |
| 0.0003        | 8.0   | 5056 | 0.0161          | 0.7789    | 0.6588 | 0.7138 | 0.9995   | 0.9908  | 0.9998 |


### Framework versions

- Transformers 4.41.0.dev0
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1