Hugging Face's logo

Augusto777
/
beit-base-patch16-224-OT-2

TensorBoard
Safetensors
beit
Generated from Trainer
Eval Results
Model card Files Metrics Community
Augusto777 commited on
Commit
d82dc21
•
1 Parent(s): ab2b520

End of training

Browse files
Files changed (11) hide show
  1. README.md +114 -0
  2. all_results.json +12 -0
  3. config.json +77 -0
  4. eval_results.json +8 -0
  5. model.safetensors +3 -0
  6. preprocessor_config.json +28 -0
  7. runs/Oct11_09-00-01_DESKTOP-SKBE9FB/events.out.tfevents.1728658803.DESKTOP-SKBE9FB.1940.0 +3 -0
  8. runs/Oct11_09-00-01_DESKTOP-SKBE9FB/events.out.tfevents.1728659525.DESKTOP-SKBE9FB.1940.1 +3 -0
  9. train_results.json +7 -0
  10. trainer_state.json +483 -0
  11. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: microsoft/beit-base-patch16-224
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - imagefolder
8
+ metrics:
9
+ - accuracy
10
+ model-index:
11
+ - name: beit-base-patch16-224-OT-2
12
+ results:
13
+ - task:
14
+ name: Image Classification
15
+ type: image-classification
16
+ dataset:
17
+ name: imagefolder
18
+ type: imagefolder
19
+ config: default
20
+ split: train
21
+ args: default
22
+ metrics:
23
+ - name: Accuracy
24
+ type: accuracy
25
+ value: 0.8387096774193549
26
+ ---
27
+
28
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
+ should probably proofread and complete it, then remove this comment. -->
30
+
31
+ # beit-base-patch16-224-OT-2
32
+
33
+ This model is a fine-tuned version of [microsoft/beit-base-patch16-224](https://huggingface.co/microsoft/beit-base-patch16-224) on the imagefolder dataset.
34
+ It achieves the following results on the evaluation set:
35
+ - Loss: 0.5047
36
+ - Accuracy: 0.8387
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 3.5e-05
56
+ - train_batch_size: 16
57
+ - eval_batch_size: 16
58
+ - seed: 42
59
+ - gradient_accumulation_steps: 4
60
+ - total_train_batch_size: 64
61
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
+ - lr_scheduler_type: linear
63
+ - lr_scheduler_warmup_ratio: 0.1
64
+ - num_epochs: 40
65
+
66
+ ### Training results
67
+
68
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
69
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
70
+ | No log | 0.91 | 5 | 1.8532 | 0.0806 |
71
+ | 1.7494 | 2.0 | 11 | 1.7818 | 0.0806 |
72
+ | 1.7494 | 2.91 | 16 | 1.6613 | 0.0806 |
73
+ | 1.6235 | 4.0 | 22 | 1.4651 | 0.0806 |
74
+ | 1.6235 | 4.91 | 27 | 1.3293 | 0.0806 |
75
+ | 1.3836 | 6.0 | 33 | 1.2034 | 0.5161 |
76
+ | 1.3836 | 6.91 | 38 | 1.1748 | 0.3710 |
77
+ | 1.2192 | 8.0 | 44 | 1.0815 | 0.4677 |
78
+ | 1.2192 | 8.91 | 49 | 1.0238 | 0.5 |
79
+ | 1.093 | 10.0 | 55 | 1.0225 | 0.4516 |
80
+ | 0.9938 | 10.91 | 60 | 0.9650 | 0.6452 |
81
+ | 0.9938 | 12.0 | 66 | 0.9314 | 0.6935 |
82
+ | 0.9235 | 12.91 | 71 | 0.9490 | 0.6452 |
83
+ | 0.9235 | 14.0 | 77 | 0.8234 | 0.7258 |
84
+ | 0.8258 | 14.91 | 82 | 0.8159 | 0.7258 |
85
+ | 0.8258 | 16.0 | 88 | 0.7514 | 0.7419 |
86
+ | 0.716 | 16.91 | 93 | 0.7469 | 0.7419 |
87
+ | 0.716 | 18.0 | 99 | 0.6734 | 0.7903 |
88
+ | 0.6026 | 18.91 | 104 | 0.6926 | 0.7581 |
89
+ | 0.5725 | 20.0 | 110 | 0.7952 | 0.7258 |
90
+ | 0.5725 | 20.91 | 115 | 0.6284 | 0.7742 |
91
+ | 0.554 | 22.0 | 121 | 0.6317 | 0.7742 |
92
+ | 0.554 | 22.91 | 126 | 0.6361 | 0.7419 |
93
+ | 0.5162 | 24.0 | 132 | 0.5501 | 0.8226 |
94
+ | 0.5162 | 24.91 | 137 | 0.6278 | 0.7581 |
95
+ | 0.4768 | 26.0 | 143 | 0.5868 | 0.7903 |
96
+ | 0.4768 | 26.91 | 148 | 0.5047 | 0.8387 |
97
+ | 0.4488 | 28.0 | 154 | 0.5264 | 0.7903 |
98
+ | 0.4488 | 28.91 | 159 | 0.4942 | 0.8387 |
99
+ | 0.4281 | 30.0 | 165 | 0.5127 | 0.8387 |
100
+ | 0.4126 | 30.91 | 170 | 0.5027 | 0.8387 |
101
+ | 0.4126 | 32.0 | 176 | 0.5387 | 0.7742 |
102
+ | 0.4326 | 32.91 | 181 | 0.5251 | 0.7903 |
103
+ | 0.4326 | 34.0 | 187 | 0.5091 | 0.8065 |
104
+ | 0.3765 | 34.91 | 192 | 0.5142 | 0.8065 |
105
+ | 0.3765 | 36.0 | 198 | 0.5142 | 0.7903 |
106
+ | 0.3913 | 36.36 | 200 | 0.5144 | 0.7903 |
107
+
108
+
109
+ ### Framework versions
110
+
111
+ - Transformers 4.36.2
112
+ - Pytorch 2.1.2+cu118
113
+ - Datasets 2.16.1
114
+ - Tokenizers 0.15.0
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 36.36,
3
+ "eval_accuracy": 0.8387096774193549,
4
+ "eval_loss": 0.5047271251678467,
5
+ "eval_runtime": 2.5816,
6
+ "eval_samples_per_second": 24.016,
7
+ "eval_steps_per_second": 1.549,
8
+ "train_loss": 0.786993111371994,
9
+ "train_runtime": 718.789,
10
+ "train_samples_per_second": 19.477,
11
+ "train_steps_per_second": 0.278
12
+ }
config.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/beit-base-patch16-224",
3
+ "add_fpn": false,
4
+ "architectures": [
5
+ "BeitForImageClassification"
6
+ ],
7
+ "attention_probs_dropout_prob": 0.0,
8
+ "auxiliary_channels": 256,
9
+ "auxiliary_concat_input": false,
10
+ "auxiliary_loss_weight": 0.4,
11
+ "auxiliary_num_convs": 1,
12
+ "drop_path_rate": 0.1,
13
+ "hidden_act": "gelu",
14
+ "hidden_dropout_prob": 0.0,
15
+ "hidden_size": 768,
16
+ "id2label": {
17
+ "0": "active",
18
+ "1": "active-inactive",
19
+ "2": "healthy",
20
+ "3": "inactive"
21
+ },
22
+ "image_size": 224,
23
+ "initializer_range": 0.02,
24
+ "intermediate_size": 3072,
25
+ "label2id": {
26
+ "active": 0,
27
+ "active-inactive": 1,
28
+ "healthy": 2,
29
+ "inactive": 3
30
+ },
31
+ "layer_norm_eps": 1e-12,
32
+ "layer_scale_init_value": 0.1,
33
+ "model_type": "beit",
34
+ "num_attention_heads": 12,
35
+ "num_channels": 3,
36
+ "num_hidden_layers": 12,
37
+ "out_features": [
38
+ "stage12"
39
+ ],
40
+ "out_indices": [
41
+ 12
42
+ ],
43
+ "patch_size": 16,
44
+ "pool_scales": [
45
+ 1,
46
+ 2,
47
+ 3,
48
+ 6
49
+ ],
50
+ "problem_type": "single_label_classification",
51
+ "reshape_hidden_states": true,
52
+ "semantic_loss_ignore_index": 255,
53
+ "stage_names": [
54
+ "stem",
55
+ "stage1",
56
+ "stage2",
57
+ "stage3",
58
+ "stage4",
59
+ "stage5",
60
+ "stage6",
61
+ "stage7",
62
+ "stage8",
63
+ "stage9",
64
+ "stage10",
65
+ "stage11",
66
+ "stage12"
67
+ ],
68
+ "torch_dtype": "float32",
69
+ "transformers_version": "4.36.2",
70
+ "use_absolute_position_embeddings": false,
71
+ "use_auxiliary_head": true,
72
+ "use_mask_token": false,
73
+ "use_mean_pooling": true,
74
+ "use_relative_position_bias": true,
75
+ "use_shared_relative_position_bias": false,
76
+ "vocab_size": 8192
77
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 36.36,
3
+ "eval_accuracy": 0.8387096774193549,
4
+ "eval_loss": 0.5047271251678467,
5
+ "eval_runtime": 2.5816,
6
+ "eval_samples_per_second": 24.016,
7
+ "eval_steps_per_second": 1.549
8
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93dde8d21ab8ea2857b23b374bc6d37e993a6935173a9fdb1d0fd88d358c1479
3
+ size 343086480
preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 224,
4
+ "width": 224
5
+ },
6
+ "do_center_crop": false,
7
+ "do_normalize": true,
8
+ "do_reduce_labels": false,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "image_processor_type": "BeitImageProcessor",
17
+ "image_std": [
18
+ 0.5,
19
+ 0.5,
20
+ 0.5
21
+ ],
22
+ "resample": 2,
23
+ "rescale_factor": 0.00392156862745098,
24
+ "size": {
25
+ "height": 224,
26
+ "width": 224
27
+ }
28
+ }
runs/Oct11_09-00-01_DESKTOP-SKBE9FB/events.out.tfevents.1728658803.DESKTOP-SKBE9FB.1940.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a588f4a1e22228d8787c4c1cc1e531561b3cc30771d1f98c59d780049418a4cd
3
+ size 20600
runs/Oct11_09-00-01_DESKTOP-SKBE9FB/events.out.tfevents.1728659525.DESKTOP-SKBE9FB.1940.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69a2bff249dba7a18d0930e95e4a83309621d3a82e52743fe3399615fcf573f3
3
+ size 411
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 36.36,
3
+ "train_loss": 0.786993111371994,
4
+ "train_runtime": 718.789,
5
+ "train_samples_per_second": 19.477,
6
+ "train_steps_per_second": 0.278
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,483 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8387096774193549,
3
+ "best_model_checkpoint": "beit-base-patch16-224-OT-2\\checkpoint-148",
4
+ "epoch": 36.36363636363637,
5
+ "eval_steps": 500,
6
+ "global_step": 200,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.91,
13
+ "eval_accuracy": 0.08064516129032258,
14
+ "eval_loss": 1.853175163269043,
15
+ "eval_runtime": 2.2425,
16
+ "eval_samples_per_second": 27.647,
17
+ "eval_steps_per_second": 1.784,
18
+ "step": 5
19
+ },
20
+ {
21
+ "epoch": 1.82,
22
+ "learning_rate": 1.75e-05,
23
+ "loss": 1.7494,
24
+ "step": 10
25
+ },
26
+ {
27
+ "epoch": 2.0,
28
+ "eval_accuracy": 0.08064516129032258,
29
+ "eval_loss": 1.7818024158477783,
30
+ "eval_runtime": 2.2481,
31
+ "eval_samples_per_second": 27.579,
32
+ "eval_steps_per_second": 1.779,
33
+ "step": 11
34
+ },
35
+ {
36
+ "epoch": 2.91,
37
+ "eval_accuracy": 0.08064516129032258,
38
+ "eval_loss": 1.6613073348999023,
39
+ "eval_runtime": 2.2572,
40
+ "eval_samples_per_second": 27.468,
41
+ "eval_steps_per_second": 1.772,
42
+ "step": 16
43
+ },
44
+ {
45
+ "epoch": 3.64,
46
+ "learning_rate": 3.5e-05,
47
+ "loss": 1.6235,
48
+ "step": 20
49
+ },
50
+ {
51
+ "epoch": 4.0,
52
+ "eval_accuracy": 0.08064516129032258,
53
+ "eval_loss": 1.4651364088058472,
54
+ "eval_runtime": 2.2621,
55
+ "eval_samples_per_second": 27.409,
56
+ "eval_steps_per_second": 1.768,
57
+ "step": 22
58
+ },
59
+ {
60
+ "epoch": 4.91,
61
+ "eval_accuracy": 0.08064516129032258,
62
+ "eval_loss": 1.3293113708496094,
63
+ "eval_runtime": 2.221,
64
+ "eval_samples_per_second": 27.915,
65
+ "eval_steps_per_second": 1.801,
66
+ "step": 27
67
+ },
68
+ {
69
+ "epoch": 5.45,
70
+ "learning_rate": 3.3055555555555553e-05,
71
+ "loss": 1.3836,
72
+ "step": 30
73
+ },
74
+ {
75
+ "epoch": 6.0,
76
+ "eval_accuracy": 0.5161290322580645,
77
+ "eval_loss": 1.2034481763839722,
78
+ "eval_runtime": 2.214,
79
+ "eval_samples_per_second": 28.003,
80
+ "eval_steps_per_second": 1.807,
81
+ "step": 33
82
+ },
83
+ {
84
+ "epoch": 6.91,
85
+ "eval_accuracy": 0.3709677419354839,
86
+ "eval_loss": 1.1748452186584473,
87
+ "eval_runtime": 2.2906,
88
+ "eval_samples_per_second": 27.068,
89
+ "eval_steps_per_second": 1.746,
90
+ "step": 38
91
+ },
92
+ {
93
+ "epoch": 7.27,
94
+ "learning_rate": 3.111111111111111e-05,
95
+ "loss": 1.2192,
96
+ "step": 40
97
+ },
98
+ {
99
+ "epoch": 8.0,
100
+ "eval_accuracy": 0.46774193548387094,
101
+ "eval_loss": 1.0814906358718872,
102
+ "eval_runtime": 2.3561,
103
+ "eval_samples_per_second": 26.315,
104
+ "eval_steps_per_second": 1.698,
105
+ "step": 44
106
+ },
107
+ {
108
+ "epoch": 8.91,
109
+ "eval_accuracy": 0.5,
110
+ "eval_loss": 1.0237908363342285,
111
+ "eval_runtime": 2.2371,
112
+ "eval_samples_per_second": 27.715,
113
+ "eval_steps_per_second": 1.788,
114
+ "step": 49
115
+ },
116
+ {
117
+ "epoch": 9.09,
118
+ "learning_rate": 2.9166666666666666e-05,
119
+ "loss": 1.093,
120
+ "step": 50
121
+ },
122
+ {
123
+ "epoch": 10.0,
124
+ "eval_accuracy": 0.45161290322580644,
125
+ "eval_loss": 1.0225210189819336,
126
+ "eval_runtime": 2.1705,
127
+ "eval_samples_per_second": 28.565,
128
+ "eval_steps_per_second": 1.843,
129
+ "step": 55
130
+ },
131
+ {
132
+ "epoch": 10.91,
133
+ "learning_rate": 2.722222222222222e-05,
134
+ "loss": 0.9938,
135
+ "step": 60
136
+ },
137
+ {
138
+ "epoch": 10.91,
139
+ "eval_accuracy": 0.6451612903225806,
140
+ "eval_loss": 0.9649767279624939,
141
+ "eval_runtime": 2.178,
142
+ "eval_samples_per_second": 28.466,
143
+ "eval_steps_per_second": 1.837,
144
+ "step": 60
145
+ },
146
+ {
147
+ "epoch": 12.0,
148
+ "eval_accuracy": 0.6935483870967742,
149
+ "eval_loss": 0.9313719868659973,
150
+ "eval_runtime": 2.2561,
151
+ "eval_samples_per_second": 27.482,
152
+ "eval_steps_per_second": 1.773,
153
+ "step": 66
154
+ },
155
+ {
156
+ "epoch": 12.73,
157
+ "learning_rate": 2.5277777777777776e-05,
158
+ "loss": 0.9235,
159
+ "step": 70
160
+ },
161
+ {
162
+ "epoch": 12.91,
163
+ "eval_accuracy": 0.6451612903225806,
164
+ "eval_loss": 0.9490375518798828,
165
+ "eval_runtime": 2.2295,
166
+ "eval_samples_per_second": 27.809,
167
+ "eval_steps_per_second": 1.794,
168
+ "step": 71
169
+ },
170
+ {
171
+ "epoch": 14.0,
172
+ "eval_accuracy": 0.7258064516129032,
173
+ "eval_loss": 0.8233911395072937,
174
+ "eval_runtime": 2.1996,
175
+ "eval_samples_per_second": 28.187,
176
+ "eval_steps_per_second": 1.819,
177
+ "step": 77
178
+ },
179
+ {
180
+ "epoch": 14.55,
181
+ "learning_rate": 2.333333333333333e-05,
182
+ "loss": 0.8258,
183
+ "step": 80
184
+ },
185
+ {
186
+ "epoch": 14.91,
187
+ "eval_accuracy": 0.7258064516129032,
188
+ "eval_loss": 0.8159334659576416,
189
+ "eval_runtime": 2.2264,
190
+ "eval_samples_per_second": 27.848,
191
+ "eval_steps_per_second": 1.797,
192
+ "step": 82
193
+ },
194
+ {
195
+ "epoch": 16.0,
196
+ "eval_accuracy": 0.7419354838709677,
197
+ "eval_loss": 0.7513789534568787,
198
+ "eval_runtime": 2.242,
199
+ "eval_samples_per_second": 27.653,
200
+ "eval_steps_per_second": 1.784,
201
+ "step": 88
202
+ },
203
+ {
204
+ "epoch": 16.36,
205
+ "learning_rate": 2.138888888888889e-05,
206
+ "loss": 0.716,
207
+ "step": 90
208
+ },
209
+ {
210
+ "epoch": 16.91,
211
+ "eval_accuracy": 0.7419354838709677,
212
+ "eval_loss": 0.7468788027763367,
213
+ "eval_runtime": 2.2824,
214
+ "eval_samples_per_second": 27.165,
215
+ "eval_steps_per_second": 1.753,
216
+ "step": 93
217
+ },
218
+ {
219
+ "epoch": 18.0,
220
+ "eval_accuracy": 0.7903225806451613,
221
+ "eval_loss": 0.6733937859535217,
222
+ "eval_runtime": 2.2143,
223
+ "eval_samples_per_second": 28.0,
224
+ "eval_steps_per_second": 1.806,
225
+ "step": 99
226
+ },
227
+ {
228
+ "epoch": 18.18,
229
+ "learning_rate": 1.9444444444444442e-05,
230
+ "loss": 0.6026,
231
+ "step": 100
232
+ },
233
+ {
234
+ "epoch": 18.91,
235
+ "eval_accuracy": 0.7580645161290323,
236
+ "eval_loss": 0.6926217079162598,
237
+ "eval_runtime": 2.2096,
238
+ "eval_samples_per_second": 28.06,
239
+ "eval_steps_per_second": 1.81,
240
+ "step": 104
241
+ },
242
+ {
243
+ "epoch": 20.0,
244
+ "learning_rate": 1.75e-05,
245
+ "loss": 0.5725,
246
+ "step": 110
247
+ },
248
+ {
249
+ "epoch": 20.0,
250
+ "eval_accuracy": 0.7258064516129032,
251
+ "eval_loss": 0.7951985001564026,
252
+ "eval_runtime": 2.1841,
253
+ "eval_samples_per_second": 28.388,
254
+ "eval_steps_per_second": 1.831,
255
+ "step": 110
256
+ },
257
+ {
258
+ "epoch": 20.91,
259
+ "eval_accuracy": 0.7741935483870968,
260
+ "eval_loss": 0.6283877491950989,
261
+ "eval_runtime": 2.2365,
262
+ "eval_samples_per_second": 27.721,
263
+ "eval_steps_per_second": 1.788,
264
+ "step": 115
265
+ },
266
+ {
267
+ "epoch": 21.82,
268
+ "learning_rate": 1.5555555555555555e-05,
269
+ "loss": 0.554,
270
+ "step": 120
271
+ },
272
+ {
273
+ "epoch": 22.0,
274
+ "eval_accuracy": 0.7741935483870968,
275
+ "eval_loss": 0.6316818594932556,
276
+ "eval_runtime": 2.1575,
277
+ "eval_samples_per_second": 28.737,
278
+ "eval_steps_per_second": 1.854,
279
+ "step": 121
280
+ },
281
+ {
282
+ "epoch": 22.91,
283
+ "eval_accuracy": 0.7419354838709677,
284
+ "eval_loss": 0.6361408233642578,
285
+ "eval_runtime": 2.2285,
286
+ "eval_samples_per_second": 27.821,
287
+ "eval_steps_per_second": 1.795,
288
+ "step": 126
289
+ },
290
+ {
291
+ "epoch": 23.64,
292
+ "learning_rate": 1.361111111111111e-05,
293
+ "loss": 0.5162,
294
+ "step": 130
295
+ },
296
+ {
297
+ "epoch": 24.0,
298
+ "eval_accuracy": 0.8225806451612904,
299
+ "eval_loss": 0.5500665903091431,
300
+ "eval_runtime": 2.2304,
301
+ "eval_samples_per_second": 27.798,
302
+ "eval_steps_per_second": 1.793,
303
+ "step": 132
304
+ },
305
+ {
306
+ "epoch": 24.91,
307
+ "eval_accuracy": 0.7580645161290323,
308
+ "eval_loss": 0.6277778744697571,
309
+ "eval_runtime": 2.2482,
310
+ "eval_samples_per_second": 27.577,
311
+ "eval_steps_per_second": 1.779,
312
+ "step": 137
313
+ },
314
+ {
315
+ "epoch": 25.45,
316
+ "learning_rate": 1.1666666666666665e-05,
317
+ "loss": 0.4768,
318
+ "step": 140
319
+ },
320
+ {
321
+ "epoch": 26.0,
322
+ "eval_accuracy": 0.7903225806451613,
323
+ "eval_loss": 0.5867522358894348,
324
+ "eval_runtime": 2.211,
325
+ "eval_samples_per_second": 28.041,
326
+ "eval_steps_per_second": 1.809,
327
+ "step": 143
328
+ },
329
+ {
330
+ "epoch": 26.91,
331
+ "eval_accuracy": 0.8387096774193549,
332
+ "eval_loss": 0.5047271251678467,
333
+ "eval_runtime": 2.199,
334
+ "eval_samples_per_second": 28.194,
335
+ "eval_steps_per_second": 1.819,
336
+ "step": 148
337
+ },
338
+ {
339
+ "epoch": 27.27,
340
+ "learning_rate": 9.722222222222221e-06,
341
+ "loss": 0.4488,
342
+ "step": 150
343
+ },
344
+ {
345
+ "epoch": 28.0,
346
+ "eval_accuracy": 0.7903225806451613,
347
+ "eval_loss": 0.5263858437538147,
348
+ "eval_runtime": 2.226,
349
+ "eval_samples_per_second": 27.852,
350
+ "eval_steps_per_second": 1.797,
351
+ "step": 154
352
+ },
353
+ {
354
+ "epoch": 28.91,
355
+ "eval_accuracy": 0.8387096774193549,
356
+ "eval_loss": 0.49419379234313965,
357
+ "eval_runtime": 2.2036,
358
+ "eval_samples_per_second": 28.136,
359
+ "eval_steps_per_second": 1.815,
360
+ "step": 159
361
+ },
362
+ {
363
+ "epoch": 29.09,
364
+ "learning_rate": 7.777777777777777e-06,
365
+ "loss": 0.4281,
366
+ "step": 160
367
+ },
368
+ {
369
+ "epoch": 30.0,
370
+ "eval_accuracy": 0.8387096774193549,
371
+ "eval_loss": 0.5126697421073914,
372
+ "eval_runtime": 2.1663,
373
+ "eval_samples_per_second": 28.62,
374
+ "eval_steps_per_second": 1.846,
375
+ "step": 165
376
+ },
377
+ {
378
+ "epoch": 30.91,
379
+ "learning_rate": 5.833333333333332e-06,
380
+ "loss": 0.4126,
381
+ "step": 170
382
+ },
383
+ {
384
+ "epoch": 30.91,
385
+ "eval_accuracy": 0.8387096774193549,
386
+ "eval_loss": 0.5026701092720032,
387
+ "eval_runtime": 2.1795,
388
+ "eval_samples_per_second": 28.447,
389
+ "eval_steps_per_second": 1.835,
390
+ "step": 170
391
+ },
392
+ {
393
+ "epoch": 32.0,
394
+ "eval_accuracy": 0.7741935483870968,
395
+ "eval_loss": 0.5387425422668457,
396
+ "eval_runtime": 2.5256,
397
+ "eval_samples_per_second": 24.549,
398
+ "eval_steps_per_second": 1.584,
399
+ "step": 176
400
+ },
401
+ {
402
+ "epoch": 32.73,
403
+ "learning_rate": 3.888888888888889e-06,
404
+ "loss": 0.4326,
405
+ "step": 180
406
+ },
407
+ {
408
+ "epoch": 32.91,
409
+ "eval_accuracy": 0.7903225806451613,
410
+ "eval_loss": 0.5250905156135559,
411
+ "eval_runtime": 2.2035,
412
+ "eval_samples_per_second": 28.137,
413
+ "eval_steps_per_second": 1.815,
414
+ "step": 181
415
+ },
416
+ {
417
+ "epoch": 34.0,
418
+ "eval_accuracy": 0.8064516129032258,
419
+ "eval_loss": 0.5091389417648315,
420
+ "eval_runtime": 2.3211,
421
+ "eval_samples_per_second": 26.712,
422
+ "eval_steps_per_second": 1.723,
423
+ "step": 187
424
+ },
425
+ {
426
+ "epoch": 34.55,
427
+ "learning_rate": 1.9444444444444444e-06,
428
+ "loss": 0.3765,
429
+ "step": 190
430
+ },
431
+ {
432
+ "epoch": 34.91,
433
+ "eval_accuracy": 0.8064516129032258,
434
+ "eval_loss": 0.5141739845275879,
435
+ "eval_runtime": 2.2025,
436
+ "eval_samples_per_second": 28.15,
437
+ "eval_steps_per_second": 1.816,
438
+ "step": 192
439
+ },
440
+ {
441
+ "epoch": 36.0,
442
+ "eval_accuracy": 0.7903225806451613,
443
+ "eval_loss": 0.514202356338501,
444
+ "eval_runtime": 2.1798,
445
+ "eval_samples_per_second": 28.443,
446
+ "eval_steps_per_second": 1.835,
447
+ "step": 198
448
+ },
449
+ {
450
+ "epoch": 36.36,
451
+ "learning_rate": 0.0,
452
+ "loss": 0.3913,
453
+ "step": 200
454
+ },
455
+ {
456
+ "epoch": 36.36,
457
+ "eval_accuracy": 0.7903225806451613,
458
+ "eval_loss": 0.5144444108009338,
459
+ "eval_runtime": 2.196,
460
+ "eval_samples_per_second": 28.233,
461
+ "eval_steps_per_second": 1.821,
462
+ "step": 200
463
+ },
464
+ {
465
+ "epoch": 36.36,
466
+ "step": 200,
467
+ "total_flos": 9.859141601338982e+17,
468
+ "train_loss": 0.786993111371994,
469
+ "train_runtime": 718.789,
470
+ "train_samples_per_second": 19.477,
471
+ "train_steps_per_second": 0.278
472
+ }
473
+ ],
474
+ "logging_steps": 10,
475
+ "max_steps": 200,
476
+ "num_input_tokens_seen": 0,
477
+ "num_train_epochs": 40,
478
+ "save_steps": 500,
479
+ "total_flos": 9.859141601338982e+17,
480
+ "train_batch_size": 16,
481
+ "trial_name": null,
482
+ "trial_params": null
483
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36e96da6446cba44d6ae08cb924e141f94e6c3bff23eb872eca89a73e9a155de
3
+ size 4728