elucidator8918 commited on
Commit
22c2c7c
1 Parent(s): 5873f67

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +49 -0
README.md ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+
3
+ # Transfer Learning Vision Transformer (ViT) - Google 224 ViT Base Patch
4
+
5
+ ## Description
6
+
7
+ This model is a Transfer Learning Vision Transformer (ViT) based on Google's 224 ViT Base Patch architecture. It has been fine-tuned on a dataset consisting of fungal images from Russia, with a specific focus on various fungi and lichen species.
8
+
9
+ ## Model Information
10
+
11
+ - Model Name: Transfer Learning ViT - Google 224 ViT Base Patch
12
+ - Model Architecture: Vision Transformer (ViT)
13
+ - Base Architecture: Google's 224 ViT Base Patch
14
+ - Pre-trained on General ImageNet dataset
15
+ - Fine-tuned on: Fungal image dataset from Russia
16
+
17
+ ## Performance
18
+
19
+ - Accuracy: 90.31%
20
+ - F1 Score: 86.33%
21
+
22
+ ## Training Details
23
+
24
+ - Training Loss:
25
+ - Initial: 1.043200
26
+ - Final: 0.116200
27
+ - Validation Loss:
28
+ - Initial: 0.822428
29
+ - Final: 0.335994
30
+ - Training Epochs: 10
31
+ - Training Runtime: 18575.04 seconds
32
+ - Training Samples per Second: 33.327
33
+ - Training Steps per Second: 1.042
34
+ - Total FLOPs: 4.801 x 10^19
35
+
36
+ ## Recommended Use Cases
37
+
38
+ - Species classification of various fungi and lichen in Russia.
39
+ - Fungal biodiversity studies.
40
+ - Image recognition tasks related to fungi and lichen species.
41
+
42
+ ## Limitations
43
+
44
+ - The model's performance is optimized for fungal species and may not generalize well to other domains.
45
+ - The model may not perform well on images of fungi and lichen species from regions other than Russia.
46
+
47
+ ## Model Author
48
+
49
+ Siddhant Dutta