File size: 2,285 Bytes
f0464a0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ff74d22
f0464a0
6eed9bf
 
 
f0464a0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ff74d22
f0464a0
 
 
 
 
 
 
 
 
 
 
 
6eed9bf
 
 
 
 
 
 
 
 
 
 
f0464a0
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
- generated_from_trainer
model-index:
- name: ryan_model314_3
  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. -->

# ryan_model314_3

This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2783
- Na Accuracy: 0.9389
- Ordinal Mae: 0.8154

## 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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Na Accuracy | Ordinal Mae |
|:-------------:|:-----:|:----:|:---------------:|:-----------:|:-----------:|
| 0.3676        | 0.05  | 100  | 0.3423          | 0.9274      | 1.1293      |
| 0.3329        | 0.09  | 200  | 0.3136          | 0.9314      | 1.0706      |
| 0.3134        | 0.14  | 300  | 0.3302          | 0.9166      | 1.1220      |
| 0.314         | 0.19  | 400  | 0.2992          | 0.9256      | 0.8202      |
| 0.2965        | 0.23  | 500  | 0.3198          | 0.9249      | 1.2210      |
| 0.3068        | 0.28  | 600  | 0.2673          | 0.9372      | 1.1036      |
| 0.2824        | 0.32  | 700  | 0.2922          | 0.9372      | 1.4977      |
| 0.2914        | 0.37  | 800  | 0.2798          | 0.9384      | 0.7789      |
| 0.2968        | 0.42  | 900  | 0.2710          | 0.9369      | 0.9694      |
| 0.2433        | 0.46  | 1000 | 0.2652          | 0.9372      | 1.0212      |
| 0.2438        | 0.51  | 1100 | 0.2783          | 0.9389      | 0.8154      |


### Framework versions

- Transformers 4.39.1
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2