Model save
Browse files- README.md +84 -0
- model.safetensors +1 -1
README.md
ADDED
@@ -0,0 +1,84 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
base_model: google/vit-base-patch16-224
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- accuracy
|
8 |
+
- precision
|
9 |
+
- recall
|
10 |
+
- f1
|
11 |
+
model-index:
|
12 |
+
- name: vit-lr-linear
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
# vit-lr-linear
|
20 |
+
|
21 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on an unknown dataset.
|
22 |
+
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 0.8046
|
24 |
+
- Accuracy: 0.8499
|
25 |
+
- Precision: 0.8547
|
26 |
+
- Recall: 0.8499
|
27 |
+
- F1: 0.8506
|
28 |
+
|
29 |
+
## Model description
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Intended uses & limitations
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training and evaluation data
|
38 |
+
|
39 |
+
More information needed
|
40 |
+
|
41 |
+
## Training procedure
|
42 |
+
|
43 |
+
### Training hyperparameters
|
44 |
+
|
45 |
+
The following hyperparameters were used during training:
|
46 |
+
- learning_rate: 0.0001
|
47 |
+
- train_batch_size: 16
|
48 |
+
- eval_batch_size: 8
|
49 |
+
- seed: 42
|
50 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
51 |
+
- lr_scheduler_type: linear
|
52 |
+
- num_epochs: 100
|
53 |
+
- mixed_precision_training: Native AMP
|
54 |
+
|
55 |
+
### Training results
|
56 |
+
|
57 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|
58 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
|
59 |
+
| 0.6029 | 0.31 | 100 | 0.6126 | 0.7805 | 0.7602 | 0.7805 | 0.7529 |
|
60 |
+
| 0.5726 | 0.62 | 200 | 0.6950 | 0.7649 | 0.7613 | 0.7649 | 0.7177 |
|
61 |
+
| 0.6521 | 0.93 | 300 | 0.5102 | 0.8124 | 0.8149 | 0.8124 | 0.8060 |
|
62 |
+
| 0.3803 | 1.25 | 400 | 0.6125 | 0.7843 | 0.8128 | 0.7843 | 0.7934 |
|
63 |
+
| 0.4048 | 1.56 | 500 | 0.5059 | 0.8214 | 0.8156 | 0.8214 | 0.8078 |
|
64 |
+
| 0.2939 | 1.87 | 600 | 0.6723 | 0.7680 | 0.8366 | 0.7680 | 0.7818 |
|
65 |
+
| 0.2138 | 2.18 | 700 | 0.6351 | 0.8128 | 0.8480 | 0.8128 | 0.8170 |
|
66 |
+
| 0.2615 | 2.49 | 800 | 0.4920 | 0.8322 | 0.8400 | 0.8322 | 0.8323 |
|
67 |
+
| 0.2125 | 2.8 | 900 | 0.5596 | 0.8492 | 0.8509 | 0.8492 | 0.8432 |
|
68 |
+
| 0.0768 | 3.12 | 1000 | 0.8239 | 0.8291 | 0.8500 | 0.8291 | 0.8235 |
|
69 |
+
| 0.0649 | 3.43 | 1100 | 0.6827 | 0.8367 | 0.8481 | 0.8367 | 0.8360 |
|
70 |
+
| 0.1382 | 3.74 | 1200 | 0.6838 | 0.8450 | 0.8467 | 0.8450 | 0.8399 |
|
71 |
+
| 0.0486 | 4.05 | 1300 | 0.6367 | 0.8578 | 0.8548 | 0.8578 | 0.8494 |
|
72 |
+
| 0.1122 | 4.36 | 1400 | 0.7330 | 0.8398 | 0.8368 | 0.8398 | 0.8330 |
|
73 |
+
| 0.0302 | 4.67 | 1500 | 0.7137 | 0.8450 | 0.8470 | 0.8450 | 0.8442 |
|
74 |
+
| 0.0462 | 4.98 | 1600 | 0.8198 | 0.8516 | 0.8519 | 0.8516 | 0.8456 |
|
75 |
+
| 0.0109 | 5.3 | 1700 | 0.8482 | 0.8478 | 0.8384 | 0.8478 | 0.8378 |
|
76 |
+
| 0.0545 | 5.61 | 1800 | 0.8046 | 0.8499 | 0.8547 | 0.8499 | 0.8506 |
|
77 |
+
|
78 |
+
|
79 |
+
### Framework versions
|
80 |
+
|
81 |
+
- Transformers 4.38.2
|
82 |
+
- Pytorch 2.2.1+cu121
|
83 |
+
- Datasets 2.18.0
|
84 |
+
- Tokenizers 0.15.2
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 343239356
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2586d8dae96c44631679ea93321fcf0f8c3c08bb567ca2f01e9d6a21a4d8dba1
|
3 |
size 343239356
|