Hugging Face's logo

dennisjooo
/
emotion_classification

Image Classification
Transformers
PyTorch
Safetensors
vit
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
9
dennisjooo commited on
Commit
ed770c1
•
1 Parent(s): 9e21d9b

End of training

Browse files
Files changed (4) hide show
  1. README.md +80 -83
  2. config.json +1 -1
  3. pytorch_model.bin +1 -1
  4. training_args.bin +1 -1
README.md CHANGED
@@ -4,7 +4,7 @@ base_model: google/vit-base-patch16-224-in21k
4
  tags:
5
  - generated_from_trainer
6
  datasets:
7
- - FastJobs/Visual_Emotional_Analysis
8
  metrics:
9
  - accuracy
10
  - precision
@@ -16,128 +16,125 @@ model-index:
16
  name: Image Classification
17
  type: image-classification
18
  dataset:
19
- name: FastJobs/Visual_Emotional_Analysis
20
- type: FastJobs/Visual_Emotional_Analysis
21
- config: default
22
  split: train
23
- args: default
24
  metrics:
25
  - name: Accuracy
26
  type: accuracy
27
- value: 0.63125
28
  - name: Precision
29
  type: precision
30
- value: 0.6430986797647803
31
  - name: F1
32
  type: f1
33
- value: 0.6224944698106615
34
  ---
35
 
36
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
37
  should probably proofread and complete it, then remove this comment. -->
38
 
39
- # Emotion Classification
40
-
41
- This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k)
42
- on the [FastJobs/Visual_Emotional_Analysis](https://huggingface.co/datasets/FastJobs/Visual_Emotional_Analysis) dataset.
43
-
44
- In theory, the accuracy for a random guess on this dataset is 0.1429.
45
 
 
46
  It achieves the following results on the evaluation set:
47
- - Loss: 1.1031
48
- - Accuracy: 0.6312
49
- - Precision: 0.6431
50
- - F1: 0.6225
51
 
52
  ## Model description
53
 
54
- The Vision Transformer base version trained on ImageNet-21K released by Google.
55
- Further details can be found on their [repo](https://huggingface.co/google/vit-base-patch16-224-in21k).
56
-
57
- ## Training and evaluation data
58
-
59
- ### Data Split
60
 
61
- Used a 4:1 ratio for training and development sets and a random seed of 42.
62
- Also used a seed of 42 for batching the data, completely unrelated lol.
63
 
64
- ### Pre-processing Augmentation
65
 
66
- The main pre-processing phase for both training and evaluation includes:
67
- - Bilinear interpolation to resize the image to (224, 224, 3) because it uses ImageNet images to train the original model
68
- - Normalizing images using a mean and standard deviation of [0.5, 0.5, 0.5] just like the original model
69
 
70
- Other than the aforementioned pre-processing, the training set was augmented using:
71
- - Random horizontal & vertical flip
72
- - Color jitter
73
- - Random resized crop
74
 
75
  ## Training procedure
76
 
77
  ### Training hyperparameters
78
 
79
  The following hyperparameters were used during training:
80
- - learning_rate: 0.0002
81
  - train_batch_size: 64
82
  - eval_batch_size: 64
83
  - seed: 42
84
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
85
  - lr_scheduler_type: cosine_with_restarts
86
- - lr_scheduler_warmup_steps: 20
87
- - num_epochs: 100
88
 
89
  ### Training results
90
 
91
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | F1 |
92
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|
93
- | 2.0742 | 1.0 | 10 | 2.0533 | 0.1938 | 0.1942 | 0.1858 |
94
- | 2.0081 | 2.0 | 20 | 1.8908 | 0.3438 | 0.3701 | 0.3368 |
95
- | 1.7211 | 3.0 | 30 | 1.5199 | 0.5312 | 0.4821 | 0.4844 |
96
- | 1.5641 | 4.0 | 40 | 1.4248 | 0.4875 | 0.5314 | 0.4532 |
97
- | 1.3979 | 5.0 | 50 | 1.2973 | 0.5375 | 0.5162 | 0.5023 |
98
- | 1.2997 | 6.0 | 60 | 1.2016 | 0.525 | 0.4828 | 0.4826 |
99
- | 1.2348 | 7.0 | 70 | 1.1670 | 0.5875 | 0.6375 | 0.5941 |
100
- | 1.1481 | 8.0 | 80 | 1.1292 | 0.6 | 0.6111 | 0.5961 |
101
- | 1.079 | 9.0 | 90 | 1.1782 | 0.5188 | 0.5265 | 0.5005 |
102
- | 0.9909 | 10.0 | 100 | 1.1115 | 0.5813 | 0.5892 | 0.5668 |
103
- | 0.9662 | 11.0 | 110 | 1.1047 | 0.5938 | 0.6336 | 0.5723 |
104
- | 0.8149 | 12.0 | 120 | 1.0944 | 0.5563 | 0.5648 | 0.5499 |
105
- | 0.7661 | 13.0 | 130 | 1.0932 | 0.5625 | 0.5738 | 0.5499 |
106
- | 0.7067 | 14.0 | 140 | 1.0787 | 0.6062 | 0.6318 | 0.6045 |
107
- | 0.6708 | 15.0 | 150 | 1.1140 | 0.6188 | 0.6463 | 0.6134 |
108
- | 0.6268 | 16.0 | 160 | 1.0875 | 0.5813 | 0.6016 | 0.5815 |
109
- | 0.5473 | 17.0 | 170 | 1.1483 | 0.5938 | 0.6027 | 0.5844 |
110
- | 0.5228 | 18.0 | 180 | 1.1031 | 0.6312 | 0.6431 | 0.6225 |
111
- | 0.4805 | 19.0 | 190 | 1.1747 | 0.5813 | 0.6057 | 0.5848 |
112
- | 0.4995 | 20.0 | 200 | 1.1865 | 0.6062 | 0.6062 | 0.5980 |
113
- | 0.456 | 21.0 | 210 | 1.2619 | 0.6 | 0.6020 | 0.5843 |
114
- | 0.4697 | 22.0 | 220 | 1.2476 | 0.5625 | 0.5804 | 0.5647 |
115
- | 0.3656 | 23.0 | 230 | 1.3106 | 0.6125 | 0.6645 | 0.6130 |
116
- | 0.394 | 24.0 | 240 | 1.3398 | 0.5437 | 0.5627 | 0.5460 |
117
- | 0.35 | 25.0 | 250 | 1.3391 | 0.5938 | 0.5940 | 0.5860 |
118
- | 0.3508 | 26.0 | 260 | 1.2846 | 0.575 | 0.6070 | 0.5821 |
119
- | 0.3106 | 27.0 | 270 | 1.3495 | 0.575 | 0.6258 | 0.5663 |
120
- | 0.3265 | 28.0 | 280 | 1.4450 | 0.5375 | 0.6512 | 0.5248 |
121
- | 0.2806 | 29.0 | 290 | 1.5145 | 0.5188 | 0.5840 | 0.5151 |
122
- | 0.3276 | 30.0 | 300 | 1.5207 | 0.5188 | 0.5741 | 0.5164 |
123
- | 0.2932 | 31.0 | 310 | 1.3179 | 0.6312 | 0.6421 | 0.6298 |
124
- | 0.3542 | 32.0 | 320 | 1.3720 | 0.5875 | 0.6157 | 0.5780 |
125
- | 0.3321 | 33.0 | 330 | 1.4787 | 0.5625 | 0.6088 | 0.5714 |
126
- | 0.2641 | 34.0 | 340 | 1.5468 | 0.5375 | 0.5817 | 0.5385 |
127
- | 0.2432 | 35.0 | 350 | 1.4893 | 0.5687 | 0.6012 | 0.5538 |
128
- | 0.275 | 36.0 | 360 | 1.4775 | 0.575 | 0.5827 | 0.5710 |
129
- | 0.239 | 37.0 | 370 | 1.4812 | 0.575 | 0.6100 | 0.5739 |
130
- | 0.2658 | 38.0 | 380 | 1.7335 | 0.5563 | 0.6547 | 0.5436 |
131
- | 0.3026 | 39.0 | 390 | 1.5692 | 0.5875 | 0.6401 | 0.5854 |
132
- | 0.1867 | 40.0 | 400 | 1.4908 | 0.5687 | 0.5921 | 0.5741 |
133
- | 0.1931 | 41.0 | 410 | 1.6608 | 0.5375 | 0.5834 | 0.5396 |
134
- | 0.2416 | 42.0 | 420 | 1.5172 | 0.5938 | 0.6259 | 0.5935 |
135
- | 0.1943 | 43.0 | 430 | 1.5260 | 0.5437 | 0.5775 | 0.5498 |
 
 
 
 
 
 
 
 
 
 
 
 
136
 
137
 
138
  ### Framework versions
139
 
140
- - Transformers 4.33.1
141
- - Pytorch 2.0.1+cu118
142
- - Datasets 2.14.5
143
  - Tokenizers 0.13.3
 
4
  tags:
5
  - generated_from_trainer
6
  datasets:
7
+ - image_folder
8
  metrics:
9
  - accuracy
10
  - precision
 
16
  name: Image Classification
17
  type: image-classification
18
  dataset:
19
+ name: image_folder
20
+ type: image_folder
21
+ config: FastJobs--Visual_Emotional_Analysis
22
  split: train
23
+ args: FastJobs--Visual_Emotional_Analysis
24
  metrics:
25
  - name: Accuracy
26
  type: accuracy
27
+ value: 0.64375
28
  - name: Precision
29
  type: precision
30
+ value: 0.6639732142857142
31
  - name: F1
32
  type: f1
33
+ value: 0.640682001352849
34
  ---
35
 
36
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
37
  should probably proofread and complete it, then remove this comment. -->
38
 
39
+ # emotion_classification
 
 
 
 
 
40
 
41
+ 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 image_folder dataset.
42
  It achieves the following results on the evaluation set:
43
+ - Loss: 1.0750
44
+ - Accuracy: 0.6438
45
+ - Precision: 0.6640
46
+ - F1: 0.6407
47
 
48
  ## Model description
49
 
50
+ More information needed
 
 
 
 
 
51
 
52
+ ## Intended uses & limitations
 
53
 
54
+ More information needed
55
 
56
+ ## Training and evaluation data
 
 
57
 
58
+ More information needed
 
 
 
59
 
60
  ## Training procedure
61
 
62
  ### Training hyperparameters
63
 
64
  The following hyperparameters were used during training:
65
+ - learning_rate: 5e-05
66
  - train_batch_size: 64
67
  - eval_batch_size: 64
68
  - seed: 42
69
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
70
  - lr_scheduler_type: cosine_with_restarts
71
+ - lr_scheduler_warmup_steps: 50
72
+ - num_epochs: 200
73
 
74
  ### Training results
75
 
76
  | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | F1 |
77
  |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|
78
+ | 2.0755 | 1.0 | 10 | 2.0787 | 0.1437 | 0.1529 | 0.1414 |
79
+ | 2.0711 | 2.0 | 20 | 2.0698 | 0.1875 | 0.1926 | 0.1832 |
80
+ | 2.0533 | 3.0 | 30 | 2.0520 | 0.2 | 0.2127 | 0.1961 |
81
+ | 2.0225 | 4.0 | 40 | 2.0173 | 0.225 | 0.2228 | 0.2054 |
82
+ | 1.9569 | 5.0 | 50 | 1.9289 | 0.2812 | 0.3345 | 0.2544 |
83
+ | 1.8501 | 6.0 | 60 | 1.7792 | 0.3688 | 0.4904 | 0.3225 |
84
+ | 1.7072 | 7.0 | 70 | 1.6236 | 0.4313 | 0.4131 | 0.3883 |
85
+ | 1.6065 | 8.0 | 80 | 1.5276 | 0.45 | 0.4533 | 0.3920 |
86
+ | 1.539 | 9.0 | 90 | 1.4747 | 0.4938 | 0.4748 | 0.4563 |
87
+ | 1.5086 | 10.0 | 100 | 1.4393 | 0.4938 | 0.4557 | 0.4466 |
88
+ | 1.4479 | 11.0 | 110 | 1.3893 | 0.5188 | 0.4563 | 0.4696 |
89
+ | 1.3683 | 12.0 | 120 | 1.3534 | 0.5437 | 0.5081 | 0.5149 |
90
+ | 1.3288 | 13.0 | 130 | 1.3392 | 0.5563 | 0.5569 | 0.5323 |
91
+ | 1.2514 | 14.0 | 140 | 1.2723 | 0.5625 | 0.5467 | 0.5246 |
92
+ | 1.2116 | 15.0 | 150 | 1.2526 | 0.5875 | 0.5554 | 0.5601 |
93
+ | 1.1824 | 16.0 | 160 | 1.2047 | 0.5938 | 0.6100 | 0.5697 |
94
+ | 1.1323 | 17.0 | 170 | 1.1950 | 0.5813 | 0.5331 | 0.5472 |
95
+ | 1.0782 | 18.0 | 180 | 1.1802 | 0.5875 | 0.5911 | 0.5807 |
96
+ | 1.0304 | 19.0 | 190 | 1.1534 | 0.6125 | 0.6133 | 0.6012 |
97
+ | 0.982 | 20.0 | 200 | 1.1302 | 0.6 | 0.5923 | 0.5806 |
98
+ | 0.9309 | 21.0 | 210 | 1.1849 | 0.5938 | 0.6157 | 0.5723 |
99
+ | 0.9205 | 22.0 | 220 | 1.1483 | 0.6 | 0.6137 | 0.5882 |
100
+ | 0.8275 | 23.0 | 230 | 1.1332 | 0.5938 | 0.6192 | 0.5894 |
101
+ | 0.8472 | 24.0 | 240 | 1.1195 | 0.625 | 0.6444 | 0.6242 |
102
+ | 0.7974 | 25.0 | 250 | 1.1444 | 0.6062 | 0.6277 | 0.6035 |
103
+ | 0.7532 | 26.0 | 260 | 1.1312 | 0.5875 | 0.6036 | 0.5832 |
104
+ | 0.7596 | 27.0 | 270 | 1.1217 | 0.6062 | 0.6412 | 0.6098 |
105
+ | 0.6818 | 28.0 | 280 | 1.1736 | 0.5625 | 0.6180 | 0.5473 |
106
+ | 0.6484 | 29.0 | 290 | 1.1630 | 0.5563 | 0.5887 | 0.5367 |
107
+ | 0.6578 | 30.0 | 300 | 1.0750 | 0.6438 | 0.6640 | 0.6407 |
108
+ | 0.6235 | 31.0 | 310 | 1.0676 | 0.6438 | 0.6556 | 0.6422 |
109
+ | 0.5966 | 32.0 | 320 | 1.0531 | 0.6438 | 0.6421 | 0.6385 |
110
+ | 0.5819 | 33.0 | 330 | 1.1244 | 0.6188 | 0.6315 | 0.6176 |
111
+ | 0.5585 | 34.0 | 340 | 1.1466 | 0.5813 | 0.6136 | 0.5790 |
112
+ | 0.5696 | 35.0 | 350 | 1.0703 | 0.6438 | 0.6614 | 0.6481 |
113
+ | 0.5476 | 36.0 | 360 | 1.1136 | 0.6438 | 0.6764 | 0.6466 |
114
+ | 0.475 | 37.0 | 370 | 1.1122 | 0.6375 | 0.6612 | 0.6340 |
115
+ | 0.5381 | 38.0 | 380 | 1.1547 | 0.6188 | 0.6570 | 0.6122 |
116
+ | 0.5161 | 39.0 | 390 | 1.2268 | 0.5875 | 0.6161 | 0.5704 |
117
+ | 0.4528 | 40.0 | 400 | 1.1065 | 0.6188 | 0.6314 | 0.6122 |
118
+ | 0.401 | 41.0 | 410 | 1.1209 | 0.6438 | 0.6550 | 0.6440 |
119
+ | 0.4067 | 42.0 | 420 | 1.1440 | 0.6312 | 0.6345 | 0.6251 |
120
+ | 0.3831 | 43.0 | 430 | 1.1972 | 0.6188 | 0.6480 | 0.6075 |
121
+ | 0.4073 | 44.0 | 440 | 1.2422 | 0.6062 | 0.6644 | 0.6028 |
122
+ | 0.371 | 45.0 | 450 | 1.2152 | 0.5875 | 0.6087 | 0.5848 |
123
+ | 0.396 | 46.0 | 460 | 1.1972 | 0.6125 | 0.6306 | 0.6106 |
124
+ | 0.3322 | 47.0 | 470 | 1.2979 | 0.5813 | 0.6158 | 0.5811 |
125
+ | 0.3691 | 48.0 | 480 | 1.1657 | 0.625 | 0.6371 | 0.6162 |
126
+ | 0.3219 | 49.0 | 490 | 1.1786 | 0.6 | 0.6417 | 0.5997 |
127
+ | 0.3371 | 50.0 | 500 | 1.2126 | 0.6188 | 0.6396 | 0.6149 |
128
+ | 0.3781 | 51.0 | 510 | 1.2246 | 0.6 | 0.6244 | 0.5972 |
129
+ | 0.3629 | 52.0 | 520 | 1.1820 | 0.6188 | 0.6437 | 0.6122 |
130
+ | 0.3025 | 53.0 | 530 | 1.1795 | 0.6062 | 0.6326 | 0.6063 |
131
+ | 0.309 | 54.0 | 540 | 1.1647 | 0.625 | 0.6510 | 0.6252 |
132
+ | 0.2999 | 55.0 | 550 | 1.2023 | 0.6375 | 0.6449 | 0.6373 |
133
 
134
 
135
  ### Framework versions
136
 
137
+ - Transformers 4.33.0
138
+ - Pytorch 2.0.0
139
+ - Datasets 2.1.0
140
  - Tokenizers 0.13.3
config.json CHANGED
@@ -40,5 +40,5 @@
40
  "problem_type": "single_label_classification",
41
  "qkv_bias": true,
42
  "torch_dtype": "float32",
43
- "transformers_version": "4.33.1"
44
  }
 
40
  "problem_type": "single_label_classification",
41
  "qkv_bias": true,
42
  "torch_dtype": "float32",
43
+ "transformers_version": "4.33.0"
44
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:748f0dd2bd7134254a547fcf47842053881f8de66aa2463ddb1a3233e0d0afb4
3
  size 343287149
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24d162410d963bd2b0a0e8866c5bfa7f9a106b1fb29271c8ab38843aeb47b55c
3
  size 343287149
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c2b5806c6f8c8517c7ec802a410b81f44d1d97b52dded21076fb20f3779beb54
3
  size 4027
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbd66263591fa13de290a574867e058eb60f972cf9760917a9c1aad8c1b99bfc
3
  size 4027