File size: 1,784 Bytes
2cccaa6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: Fsoft-AIC/videberta-base
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: ComOM-VIDeBERTa-2
  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. -->

# ComOM-VIDeBERTa-2

This model is a fine-tuned version of [Fsoft-AIC/videberta-base](https://huggingface.co/Fsoft-AIC/videberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3046
- Accuracy: 0.5357

## 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: 2e-05
- 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: 8

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log        | 1.0   | 77   | 1.4249          | 0.4708   |
| No log        | 2.0   | 154  | 1.4096          | 0.4708   |
| No log        | 3.0   | 231  | 1.3871          | 0.4708   |
| No log        | 4.0   | 308  | 1.3809          | 0.5032   |
| No log        | 5.0   | 385  | 1.3529          | 0.5195   |
| No log        | 6.0   | 462  | 1.3257          | 0.5260   |
| 1.4302        | 7.0   | 539  | 1.3101          | 0.5325   |
| 1.4302        | 8.0   | 616  | 1.3046          | 0.5357   |


### Framework versions

- Transformers 4.35.0
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.14.1