DrishtiSharma
commited on
Commit
•
3ac8f96
1
Parent(s):
5911eb7
update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
metrics:
|
6 |
+
- accuracy
|
7 |
+
model-index:
|
8 |
+
- name: finetuned-ViT-human-action-recognition-v1
|
9 |
+
results: []
|
10 |
+
---
|
11 |
+
|
12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
13 |
+
should probably proofread and complete it, then remove this comment. -->
|
14 |
+
|
15 |
+
# finetuned-ViT-human-action-recognition-v1
|
16 |
+
|
17 |
+
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.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 5.7529
|
20 |
+
- Accuracy: 0.0680
|
21 |
+
|
22 |
+
## Model description
|
23 |
+
|
24 |
+
More information needed
|
25 |
+
|
26 |
+
## Intended uses & limitations
|
27 |
+
|
28 |
+
More information needed
|
29 |
+
|
30 |
+
## Training and evaluation data
|
31 |
+
|
32 |
+
More information needed
|
33 |
+
|
34 |
+
## Training procedure
|
35 |
+
|
36 |
+
### Training hyperparameters
|
37 |
+
|
38 |
+
The following hyperparameters were used during training:
|
39 |
+
- learning_rate: 0.0002
|
40 |
+
- train_batch_size: 16
|
41 |
+
- eval_batch_size: 8
|
42 |
+
- seed: 42
|
43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
44 |
+
- lr_scheduler_type: linear
|
45 |
+
- num_epochs: 4
|
46 |
+
- mixed_precision_training: Native AMP
|
47 |
+
|
48 |
+
### Training results
|
49 |
+
|
50 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
51 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
52 |
+
| 1.4986 | 0.13 | 100 | 3.1427 | 0.0791 |
|
53 |
+
| 1.1929 | 0.25 | 200 | 3.4083 | 0.0726 |
|
54 |
+
| 1.2673 | 0.38 | 300 | 3.4615 | 0.0769 |
|
55 |
+
| 0.9805 | 0.51 | 400 | 3.9192 | 0.0824 |
|
56 |
+
| 1.158 | 0.63 | 500 | 4.2648 | 0.0698 |
|
57 |
+
| 1.2544 | 0.76 | 600 | 4.5536 | 0.0574 |
|
58 |
+
| 1.0073 | 0.89 | 700 | 4.0310 | 0.0819 |
|
59 |
+
| 0.9315 | 1.02 | 800 | 4.5154 | 0.0702 |
|
60 |
+
| 0.9063 | 1.14 | 900 | 4.7162 | 0.0633 |
|
61 |
+
| 0.6756 | 1.27 | 1000 | 4.6482 | 0.0626 |
|
62 |
+
| 1.0239 | 1.4 | 1100 | 4.6437 | 0.0635 |
|
63 |
+
| 0.7634 | 1.52 | 1200 | 4.5625 | 0.0752 |
|
64 |
+
| 0.8365 | 1.65 | 1300 | 4.9912 | 0.0561 |
|
65 |
+
| 0.8979 | 1.78 | 1400 | 5.1739 | 0.0356 |
|
66 |
+
| 0.9448 | 1.9 | 1500 | 4.8946 | 0.0541 |
|
67 |
+
| 0.697 | 2.03 | 1600 | 4.9516 | 0.0741 |
|
68 |
+
| 0.7861 | 2.16 | 1700 | 5.0090 | 0.0776 |
|
69 |
+
| 0.6404 | 2.28 | 1800 | 5.3905 | 0.0643 |
|
70 |
+
| 0.7939 | 2.41 | 1900 | 4.9159 | 0.1015 |
|
71 |
+
| 0.6331 | 2.54 | 2000 | 5.3083 | 0.0589 |
|
72 |
+
| 0.6082 | 2.66 | 2100 | 4.8538 | 0.0857 |
|
73 |
+
| 0.6229 | 2.79 | 2200 | 5.3086 | 0.0689 |
|
74 |
+
| 0.6964 | 2.92 | 2300 | 5.3745 | 0.0713 |
|
75 |
+
| 0.5246 | 3.05 | 2400 | 5.0369 | 0.0796 |
|
76 |
+
| 0.6097 | 3.17 | 2500 | 5.2935 | 0.0743 |
|
77 |
+
| 0.5778 | 3.3 | 2600 | 5.5431 | 0.0709 |
|
78 |
+
| 0.4196 | 3.43 | 2700 | 5.5508 | 0.0759 |
|
79 |
+
| 0.5495 | 3.55 | 2800 | 5.5728 | 0.0813 |
|
80 |
+
| 0.5932 | 3.68 | 2900 | 5.7992 | 0.0663 |
|
81 |
+
| 0.4382 | 3.81 | 3000 | 5.8010 | 0.0643 |
|
82 |
+
| 0.4827 | 3.93 | 3100 | 5.7529 | 0.0680 |
|
83 |
+
|
84 |
+
|
85 |
+
### Framework versions
|
86 |
+
|
87 |
+
- Transformers 4.21.2
|
88 |
+
- Pytorch 1.12.1+cu113
|
89 |
+
- Datasets 2.4.0
|
90 |
+
- Tokenizers 0.12.1
|