Hugging Face's logo

sudo-s
/
exper7_mesum5

Image Classification
Transformers
PyTorch
TensorBoard
vit
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
sudo-s commited on
Commit
a8a8b69
•
1 Parent(s): f2f7fdc

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +102 -0
README.md ADDED
@@ -0,0 +1,102 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - accuracy
7
+ model-index:
8
+ - name: exper7_mesum5
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
+ # exper7_mesum5
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: 0.5954
20
+ - Accuracy: 0.8538
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: 0.0001
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: 10
46
+ - mixed_precision_training: Native AMP
47
+
48
+ ### Training results
49
+
50
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
51
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
52
+ | 4.2072 | 0.23 | 100 | 4.1532 | 0.1923 |
53
+ | 3.5433 | 0.47 | 200 | 3.5680 | 0.2888 |
54
+ | 3.1388 | 0.7 | 300 | 3.1202 | 0.3911 |
55
+ | 2.7924 | 0.93 | 400 | 2.7434 | 0.4787 |
56
+ | 2.1269 | 1.16 | 500 | 2.3262 | 0.5781 |
57
+ | 1.8589 | 1.4 | 600 | 1.9754 | 0.6272 |
58
+ | 1.7155 | 1.63 | 700 | 1.7627 | 0.6840 |
59
+ | 1.4689 | 1.86 | 800 | 1.5937 | 0.6994 |
60
+ | 1.0149 | 2.09 | 900 | 1.3168 | 0.7497 |
61
+ | 0.8148 | 2.33 | 1000 | 1.1630 | 0.7615 |
62
+ | 0.7159 | 2.56 | 1100 | 1.0869 | 0.7675 |
63
+ | 0.7257 | 2.79 | 1200 | 0.9607 | 0.7893 |
64
+ | 0.4171 | 3.02 | 1300 | 0.8835 | 0.7935 |
65
+ | 0.2969 | 3.26 | 1400 | 0.8259 | 0.8130 |
66
+ | 0.2405 | 3.49 | 1500 | 0.7711 | 0.8142 |
67
+ | 0.2948 | 3.72 | 1600 | 0.7629 | 0.8112 |
68
+ | 0.1765 | 3.95 | 1700 | 0.7117 | 0.8124 |
69
+ | 0.1603 | 4.19 | 1800 | 0.6946 | 0.8237 |
70
+ | 0.0955 | 4.42 | 1900 | 0.6597 | 0.8349 |
71
+ | 0.0769 | 4.65 | 2000 | 0.6531 | 0.8266 |
72
+ | 0.0816 | 4.88 | 2100 | 0.6335 | 0.8337 |
73
+ | 0.0315 | 5.12 | 2200 | 0.6087 | 0.8402 |
74
+ | 0.0368 | 5.35 | 2300 | 0.6026 | 0.8444 |
75
+ | 0.0377 | 5.58 | 2400 | 0.6450 | 0.8278 |
76
+ | 0.0603 | 5.81 | 2500 | 0.6564 | 0.8343 |
77
+ | 0.0205 | 6.05 | 2600 | 0.6119 | 0.8467 |
78
+ | 0.019 | 6.28 | 2700 | 0.6070 | 0.8479 |
79
+ | 0.0249 | 6.51 | 2800 | 0.6002 | 0.8538 |
80
+ | 0.0145 | 6.74 | 2900 | 0.6012 | 0.8497 |
81
+ | 0.0134 | 6.98 | 3000 | 0.5991 | 0.8521 |
82
+ | 0.0271 | 7.21 | 3100 | 0.5972 | 0.8503 |
83
+ | 0.0128 | 7.44 | 3200 | 0.5911 | 0.8521 |
84
+ | 0.0123 | 7.67 | 3300 | 0.5889 | 0.8538 |
85
+ | 0.0278 | 7.91 | 3400 | 0.6135 | 0.8491 |
86
+ | 0.0106 | 8.14 | 3500 | 0.5934 | 0.8533 |
87
+ | 0.0109 | 8.37 | 3600 | 0.5929 | 0.8533 |
88
+ | 0.0095 | 8.6 | 3700 | 0.5953 | 0.8550 |
89
+ | 0.009 | 8.84 | 3800 | 0.5933 | 0.8574 |
90
+ | 0.009 | 9.07 | 3900 | 0.5948 | 0.8550 |
91
+ | 0.0089 | 9.3 | 4000 | 0.5953 | 0.8556 |
92
+ | 0.0086 | 9.53 | 4100 | 0.5956 | 0.8544 |
93
+ | 0.0085 | 9.77 | 4200 | 0.5955 | 0.8556 |
94
+ | 0.0087 | 10.0 | 4300 | 0.5954 | 0.8538 |
95
+
96
+
97
+ ### Framework versions
98
+
99
+ - Transformers 4.20.1
100
+ - Pytorch 1.12.0+cu113
101
+ - Datasets 2.3.2
102
+ - Tokenizers 0.12.1