File size: 2,999 Bytes
cbaa8e9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
license: cc-by-nc-4.0
base_model: MCG-NJU/videomae-base
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
model-index:
- name: videomae-base-finetuned-ucf-crimevbinaryv4
  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. -->

# videomae-base-finetuned-ucf-crimevbinaryv4

This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8055
- Accuracy: 0.8571
- Precision: 0.8571
- Recall: 0.8571
- Auc: 0.9328

## 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: 3e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | Auc    |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
| 0.4783        | 1.0   | 175  | 0.5235          | 0.78     | 0.7825    | 0.78   | 0.8269 |
| 0.6121        | 2.0   | 350  | 0.5076          | 0.816    | 0.8179    | 0.816  | 0.8518 |
| 0.3305        | 3.0   | 525  | 0.3910          | 0.856    | 0.8660    | 0.856  | 0.9222 |
| 0.2431        | 4.0   | 700  | 0.8684          | 0.728    | 0.8190    | 0.728  | 0.8933 |
| 0.3337        | 5.0   | 875  | 0.5653          | 0.808    | 0.8266    | 0.808  | 0.9149 |
| 0.1768        | 6.0   | 1050 | 0.4709          | 0.88     | 0.8803    | 0.88   | 0.9311 |
| 0.111         | 7.0   | 1225 | 0.6650          | 0.848    | 0.8600    | 0.848  | 0.9263 |
| 0.1507        | 8.0   | 1400 | 0.6151          | 0.852    | 0.8522    | 0.852  | 0.9232 |
| 0.1771        | 9.0   | 1575 | 0.6924          | 0.872    | 0.8729    | 0.872  | 0.9280 |
| 0.13          | 10.0  | 1750 | 0.6670          | 0.884    | 0.8846    | 0.884  | 0.9304 |
| 0.0062        | 11.0  | 1925 | 0.8110          | 0.86     | 0.8621    | 0.86   | 0.9198 |
| 0.0008        | 12.0  | 2100 | 0.8139          | 0.876    | 0.8768    | 0.876  | 0.9235 |
| 0.0004        | 13.0  | 2275 | 0.8727          | 0.868    | 0.8711    | 0.868  | 0.9258 |
| 0.0005        | 14.0  | 2450 | 1.0055          | 0.86     | 0.8658    | 0.86   | 0.9250 |
| 0.0002        | 15.0  | 2625 | 0.9728          | 0.864    | 0.8677    | 0.864  | 0.9269 |


### Framework versions

- Transformers 4.47.0.dev0
- Pytorch 2.5.1+cu124
- Datasets 3.1.0
- Tokenizers 0.20.3