Hugging Face's logo

Augusto777
/
swiftformer-xs-DMAE

TensorBoard
Safetensors
swiftformer
Generated from Trainer
Eval Results
Model card Files Metrics Community
Augusto777 commited on
Commit
a45f29f
·
verified ·
1 Parent(s): 285f387

End of training

Browse files
Files changed (12) hide show
  1. README.md +42 -42
  2. all_results.json +9 -9
  3. eval_results.json +5 -5
  4. model.safetensors +1 -1
  5. runs/Dec02_10-01-43_DESKTOP-SKBE9FB/events.out.tfevents.1733155304.DESKTOP-SKBE9FB.18252.0 +3 -0
  6. runs/Dec02_10-02-56_DESKTOP-SKBE9FB/events.out.tfevents.1733155376.DESKTOP-SKBE9FB.19304.0 +3 -0
  7. runs/Dec02_10-04-12_DESKTOP-SKBE9FB/events.out.tfevents.1733155453.DESKTOP-SKBE9FB.11208.0 +3 -0
  8. runs/Dec02_10-05-34_DESKTOP-SKBE9FB/events.out.tfevents.1733155534.DESKTOP-SKBE9FB.4368.0 +3 -0
  9. runs/Dec02_10-05-34_DESKTOP-SKBE9FB/events.out.tfevents.1733155669.DESKTOP-SKBE9FB.4368.1 +3 -0
  10. train_results.json +4 -4
  11. trainer_state.json +203 -203
  12. training_args.bin +1 -1
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
- value: 0.6956521739130435
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model is a fine-tuned version of [MBZUAI/swiftformer-xs](https://huggingface.co/MBZUAI/swiftformer-xs) on the imagefolder dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 250784738141274112.0000
35
- - Accuracy: 0.6957
36
 
37
  ## Model description
38
 
@@ -51,7 +51,7 @@ More information needed
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
54
- - learning_rate: 0.00055
55
  - train_batch_size: 16
56
  - eval_batch_size: 16
57
  - seed: 42
@@ -59,48 +59,48 @@ The following hyperparameters were used during training:
59
  - total_train_batch_size: 64
60
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
61
  - lr_scheduler_type: linear
62
- - lr_scheduler_warmup_ratio: 0.1
63
  - num_epochs: 40
64
 
65
  ### Training results
66
 
67
- | Training Loss | Epoch | Step | Validation Loss | Accuracy |
68
- |:-----------------------:|:-----:|:----:|:-----------------------:|:--------:|
69
- | No log | 0.86 | 3 | 250784738141274112.0000 | 0.4783 |
70
- | No log | 2.0 | 7 | 250784738141274112.0000 | 0.5435 |
71
- | 245863059525363296.0000 | 2.86 | 10 | 250784738141274112.0000 | 0.6087 |
72
- | 245863059525363296.0000 | 4.0 | 14 | 250784738141274112.0000 | 0.6304 |
73
- | 245863059525363296.0000 | 4.86 | 17 | 250784738141274112.0000 | 0.6087 |
74
- | 257399163011779776.0000 | 6.0 | 21 | 250784738141274112.0000 | 0.6087 |
75
- | 257399163011779776.0000 | 6.86 | 24 | 250784738141274112.0000 | 0.6304 |
76
- | 257399163011779776.0000 | 8.0 | 28 | 250784738141274112.0000 | 0.6957 |
77
- | 289123454471372800.0000 | 8.86 | 31 | 250784738141274112.0000 | 0.6739 |
78
- | 289123454471372800.0000 | 10.0 | 35 | 250784738141274112.0000 | 0.6739 |
79
- | 289123454471372800.0000 | 10.86 | 38 | 250784738141274112.0000 | 0.6522 |
80
- | 213417804547542208.0000 | 12.0 | 42 | 250784738141274112.0000 | 0.5435 |
81
- | 213417804547542208.0000 | 12.86 | 45 | 250784738141274112.0000 | 0.5652 |
82
- | 213417804547542208.0000 | 14.0 | 49 | 250784738141274112.0000 | 0.6739 |
83
- | 278308328247079744.0000 | 14.86 | 52 | 250784738141274112.0000 | 0.6087 |
84
- | 278308328247079744.0000 | 16.0 | 56 | 250784738141274112.0000 | 0.5870 |
85
- | 278308328247079744.0000 | 16.86 | 59 | 250784738141274112.0000 | 0.5870 |
86
- | 262446278724550656.0000 | 18.0 | 63 | 250784738141274112.0000 | 0.5652 |
87
- | 262446278724550656.0000 | 18.86 | 66 | 250784738141274112.0000 | 0.5435 |
88
- | 246584064275277408.0000 | 20.0 | 70 | 250784738141274112.0000 | 0.5652 |
89
- | 246584064275277408.0000 | 20.86 | 73 | 250784738141274112.0000 | 0.6087 |
90
- | 246584064275277408.0000 | 22.0 | 77 | 250784738141274112.0000 | 0.6304 |
91
- | 237932062251889472.0000 | 22.86 | 80 | 250784738141274112.0000 | 0.6304 |
92
- | 237932062251889472.0000 | 24.0 | 84 | 250784738141274112.0000 | 0.6522 |
93
- | 237932062251889472.0000 | 24.86 | 87 | 250784738141274112.0000 | 0.6957 |
94
- | 237932007276308064.0000 | 26.0 | 91 | 250784738141274112.0000 | 0.6739 |
95
- | 237932007276308064.0000 | 26.86 | 94 | 250784738141274112.0000 | 0.6739 |
96
- | 237932007276308064.0000 | 28.0 | 98 | 250784738141274112.0000 | 0.6739 |
97
- | 269656271248110400.0000 | 28.86 | 101 | 250784738141274112.0000 | 0.6739 |
98
- | 269656271248110400.0000 | 30.0 | 105 | 250784738141274112.0000 | 0.6739 |
99
- | 269656271248110400.0000 | 30.86 | 108 | 250784738141274112.0000 | 0.6739 |
100
- | 242258008288002048.0000 | 32.0 | 112 | 250784738141274112.0000 | 0.6739 |
101
- | 242258008288002048.0000 | 32.86 | 115 | 250784738141274112.0000 | 0.6739 |
102
- | 242258008288002048.0000 | 34.0 | 119 | 250784738141274112.0000 | 0.6739 |
103
- | 245863059525363296.0000 | 34.29 | 120 | 250784738141274112.0000 | 0.6739 |
104
 
105
 
106
  ### Framework versions
 
21
  metrics:
22
  - name: Accuracy
23
  type: accuracy
24
+ value: 0.7391304347826086
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
31
 
32
  This model is a fine-tuned version of [MBZUAI/swiftformer-xs](https://huggingface.co/MBZUAI/swiftformer-xs) on the imagefolder dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 0.8766
35
+ - Accuracy: 0.7391
36
 
37
  ## Model description
38
 
 
51
  ### Training hyperparameters
52
 
53
  The following hyperparameters were used during training:
54
+ - learning_rate: 0.0003
55
  - train_batch_size: 16
56
  - eval_batch_size: 16
57
  - seed: 42
 
59
  - total_train_batch_size: 64
60
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
61
  - lr_scheduler_type: linear
62
+ - lr_scheduler_warmup_ratio: 0.15
63
  - num_epochs: 40
64
 
65
  ### Training results
66
 
67
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
68
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
69
+ | No log | 0.86 | 3 | 1.3860 | 0.3261 |
70
+ | No log | 2.0 | 7 | 1.3830 | 0.5 |
71
+ | 1.3842 | 2.86 | 10 | 1.3756 | 0.5870 |
72
+ | 1.3842 | 4.0 | 14 | 1.3494 | 0.6087 |
73
+ | 1.3842 | 4.86 | 17 | 1.3039 | 0.6522 |
74
+ | 1.3427 | 6.0 | 21 | 1.1962 | 0.6522 |
75
+ | 1.3427 | 6.86 | 24 | 1.1310 | 0.6739 |
76
+ | 1.3427 | 8.0 | 28 | 1.0778 | 0.6522 |
77
+ | 1.1837 | 8.86 | 31 | 1.0276 | 0.6957 |
78
+ | 1.1837 | 10.0 | 35 | 0.9928 | 0.7174 |
79
+ | 1.1837 | 10.86 | 38 | 0.9710 | 0.7174 |
80
+ | 1.0531 | 12.0 | 42 | 0.9735 | 0.6739 |
81
+ | 1.0531 | 12.86 | 45 | 0.9385 | 0.6522 |
82
+ | 1.0531 | 14.0 | 49 | 0.9601 | 0.6739 |
83
+ | 0.9883 | 14.86 | 52 | 0.9369 | 0.6739 |
84
+ | 0.9883 | 16.0 | 56 | 0.8928 | 0.6739 |
85
+ | 0.9883 | 16.86 | 59 | 0.8825 | 0.6739 |
86
+ | 0.8632 | 18.0 | 63 | 0.8743 | 0.6739 |
87
+ | 0.8632 | 18.86 | 66 | 0.8479 | 0.6739 |
88
+ | 0.8732 | 20.0 | 70 | 0.8341 | 0.6739 |
89
+ | 0.8732 | 20.86 | 73 | 0.8134 | 0.7174 |
90
+ | 0.8732 | 22.0 | 77 | 0.8845 | 0.6739 |
91
+ | 0.7524 | 22.86 | 80 | 0.9037 | 0.6522 |
92
+ | 0.7524 | 24.0 | 84 | 0.8921 | 0.6304 |
93
+ | 0.7524 | 24.86 | 87 | 0.8766 | 0.7391 |
94
+ | 0.6843 | 26.0 | 91 | 0.8707 | 0.7391 |
95
+ | 0.6843 | 26.86 | 94 | 0.8922 | 0.7391 |
96
+ | 0.6843 | 28.0 | 98 | 0.8862 | 0.6957 |
97
+ | 0.7112 | 28.86 | 101 | 0.8954 | 0.7391 |
98
+ | 0.7112 | 30.0 | 105 | 0.9228 | 0.6957 |
99
+ | 0.7112 | 30.86 | 108 | 0.9173 | 0.6957 |
100
+ | 0.6885 | 32.0 | 112 | 0.9318 | 0.6957 |
101
+ | 0.6885 | 32.86 | 115 | 0.9049 | 0.7391 |
102
+ | 0.6885 | 34.0 | 119 | 0.9157 | 0.7174 |
103
+ | 0.6734 | 34.29 | 120 | 0.9051 | 0.7391 |
104
 
105
 
106
  ### Framework versions
all_results.json CHANGED
@@ -1,12 +1,12 @@
1
  {
2
  "epoch": 34.29,
3
- "eval_accuracy": 0.6956521739130435,
4
- "eval_loss": 2.507847381412741e+17,
5
- "eval_runtime": 0.5996,
6
- "eval_samples_per_second": 76.712,
7
- "eval_steps_per_second": 5.003,
8
- "train_loss": 2.522319634493866e+17,
9
- "train_runtime": 128.6883,
10
- "train_samples_per_second": 66.206,
11
- "train_steps_per_second": 0.932
12
  }
 
1
  {
2
  "epoch": 34.29,
3
+ "eval_accuracy": 0.7391304347826086,
4
+ "eval_loss": 0.8765698075294495,
5
+ "eval_runtime": 0.7857,
6
+ "eval_samples_per_second": 58.547,
7
+ "eval_steps_per_second": 3.818,
8
+ "train_loss": 0.9331819852193196,
9
+ "train_runtime": 134.062,
10
+ "train_samples_per_second": 63.553,
11
+ "train_steps_per_second": 0.895
12
  }
eval_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "epoch": 34.29,
3
- "eval_accuracy": 0.6956521739130435,
4
- "eval_loss": 2.507847381412741e+17,
5
- "eval_runtime": 0.5996,
6
- "eval_samples_per_second": 76.712,
7
- "eval_steps_per_second": 5.003
8
  }
 
1
  {
2
  "epoch": 34.29,
3
+ "eval_accuracy": 0.7391304347826086,
4
+ "eval_loss": 0.8765698075294495,
5
+ "eval_runtime": 0.7857,
6
+ "eval_samples_per_second": 58.547,
7
+ "eval_steps_per_second": 3.818
8
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ff605f9bb57030587e37bef680db6a91adb5369cfdb0a2ef6dc3d25a4db1f09f
3
  size 12203648
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e7622952d9d5948abcd9a1304b7d9404f4ee9485c6f5935c68e0310ec4fba9d
3
  size 12203648
runs/Dec02_10-01-43_DESKTOP-SKBE9FB/events.out.tfevents.1733155304.DESKTOP-SKBE9FB.18252.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e103f395fdf8d038c6af9200f4257192873d42adf715d86c8926a02ca6feae41
3
+ size 8175
runs/Dec02_10-02-56_DESKTOP-SKBE9FB/events.out.tfevents.1733155376.DESKTOP-SKBE9FB.19304.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:82f0985f287070c4698d16d679c5bf80b853f6ad422d458994f1e611bc723543
3
+ size 9280
runs/Dec02_10-04-12_DESKTOP-SKBE9FB/events.out.tfevents.1733155453.DESKTOP-SKBE9FB.11208.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b08f1a93fc11abdde08a4ff883043583a452bc4cd8dbaf4dc26fc7f108b5bbfe
3
+ size 8963
runs/Dec02_10-05-34_DESKTOP-SKBE9FB/events.out.tfevents.1733155534.DESKTOP-SKBE9FB.4368.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bab5687bef8bf468ac88e0e83926b6dbbd6ec294dd19c8c94533d518621da889
3
+ size 17835
runs/Dec02_10-05-34_DESKTOP-SKBE9FB/events.out.tfevents.1733155669.DESKTOP-SKBE9FB.4368.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6e7c88f25acbdd372c15f7363a18b1293a1391e3bc071208b5ffd34b1539b60
3
+ size 405
train_results.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "epoch": 34.29,
3
- "train_loss": 2.522319634493866e+17,
4
- "train_runtime": 128.6883,
5
- "train_samples_per_second": 66.206,
6
- "train_steps_per_second": 0.932
7
  }
 
1
  {
2
  "epoch": 34.29,
3
+ "train_loss": 0.9331819852193196,
4
+ "train_runtime": 134.062,
5
+ "train_samples_per_second": 63.553,
6
+ "train_steps_per_second": 0.895
7
  }
trainer_state.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
- "best_metric": 0.6956521739130435,
3
- "best_model_checkpoint": "swiftformer-xs-DMAE\\checkpoint-28",
4
  "epoch": 34.285714285714285,
5
  "eval_steps": 500,
6
  "global_step": 120,
@@ -10,399 +10,399 @@
10
  "log_history": [
11
  {
12
  "epoch": 0.86,
13
- "eval_accuracy": 0.4782608695652174,
14
- "eval_loss": 2.507847381412741e+17,
15
- "eval_runtime": 0.6662,
16
- "eval_samples_per_second": 69.052,
17
- "eval_steps_per_second": 4.503,
18
  "step": 3
19
  },
20
  {
21
  "epoch": 2.0,
22
- "eval_accuracy": 0.5434782608695652,
23
- "eval_loss": 2.507847381412741e+17,
24
- "eval_runtime": 0.6131,
25
- "eval_samples_per_second": 75.024,
26
- "eval_steps_per_second": 4.893,
27
  "step": 7
28
  },
29
  {
30
  "epoch": 2.86,
31
- "learning_rate": 0.0004583333333333334,
32
- "loss": 2.458630595253633e+17,
33
  "step": 10
34
  },
35
  {
36
  "epoch": 2.86,
37
- "eval_accuracy": 0.6086956521739131,
38
- "eval_loss": 2.507847381412741e+17,
39
- "eval_runtime": 0.6182,
40
- "eval_samples_per_second": 74.415,
41
- "eval_steps_per_second": 4.853,
42
  "step": 10
43
  },
44
  {
45
  "epoch": 4.0,
46
- "eval_accuracy": 0.6304347826086957,
47
- "eval_loss": 2.507847381412741e+17,
48
- "eval_runtime": 0.6201,
49
- "eval_samples_per_second": 74.177,
50
- "eval_steps_per_second": 4.838,
51
  "step": 14
52
  },
53
  {
54
  "epoch": 4.86,
55
- "eval_accuracy": 0.6086956521739131,
56
- "eval_loss": 2.507847381412741e+17,
57
- "eval_runtime": 0.6186,
58
- "eval_samples_per_second": 74.356,
59
- "eval_steps_per_second": 4.849,
60
  "step": 17
61
  },
62
  {
63
  "epoch": 5.71,
64
- "learning_rate": 0.0005092592592592593,
65
- "loss": 2.5739916301177978e+17,
66
  "step": 20
67
  },
68
  {
69
  "epoch": 6.0,
70
- "eval_accuracy": 0.6086956521739131,
71
- "eval_loss": 2.507847381412741e+17,
72
- "eval_runtime": 0.6116,
73
- "eval_samples_per_second": 75.207,
74
- "eval_steps_per_second": 4.905,
75
  "step": 21
76
  },
77
  {
78
  "epoch": 6.86,
79
- "eval_accuracy": 0.6304347826086957,
80
- "eval_loss": 2.507847381412741e+17,
81
- "eval_runtime": 0.6236,
82
- "eval_samples_per_second": 73.76,
83
- "eval_steps_per_second": 4.81,
84
  "step": 24
85
  },
86
  {
87
  "epoch": 8.0,
88
- "eval_accuracy": 0.6956521739130435,
89
- "eval_loss": 2.507847381412741e+17,
90
- "eval_runtime": 0.6072,
91
- "eval_samples_per_second": 75.764,
92
- "eval_steps_per_second": 4.941,
93
  "step": 28
94
  },
95
  {
96
  "epoch": 8.57,
97
- "learning_rate": 0.0004583333333333334,
98
- "loss": 2.891234544713728e+17,
99
  "step": 30
100
  },
101
  {
102
  "epoch": 8.86,
103
- "eval_accuracy": 0.6739130434782609,
104
- "eval_loss": 2.507847381412741e+17,
105
- "eval_runtime": 0.6201,
106
- "eval_samples_per_second": 74.177,
107
- "eval_steps_per_second": 4.838,
108
  "step": 31
109
  },
110
  {
111
  "epoch": 10.0,
112
- "eval_accuracy": 0.6739130434782609,
113
- "eval_loss": 2.507847381412741e+17,
114
- "eval_runtime": 0.6056,
115
- "eval_samples_per_second": 75.952,
116
- "eval_steps_per_second": 4.953,
117
  "step": 35
118
  },
119
  {
120
  "epoch": 10.86,
121
- "eval_accuracy": 0.6521739130434783,
122
- "eval_loss": 2.507847381412741e+17,
123
- "eval_runtime": 0.6161,
124
- "eval_samples_per_second": 74.658,
125
- "eval_steps_per_second": 4.869,
126
  "step": 38
127
  },
128
  {
129
  "epoch": 11.43,
130
- "learning_rate": 0.00040740740740740744,
131
- "loss": 2.134178045475422e+17,
132
  "step": 40
133
  },
134
  {
135
  "epoch": 12.0,
136
- "eval_accuracy": 0.5434782608695652,
137
- "eval_loss": 2.507847381412741e+17,
138
- "eval_runtime": 0.6192,
139
- "eval_samples_per_second": 74.294,
140
- "eval_steps_per_second": 4.845,
141
  "step": 42
142
  },
143
  {
144
  "epoch": 12.86,
145
- "eval_accuracy": 0.5652173913043478,
146
- "eval_loss": 2.507847381412741e+17,
147
- "eval_runtime": 0.6056,
148
- "eval_samples_per_second": 75.952,
149
- "eval_steps_per_second": 4.953,
150
  "step": 45
151
  },
152
  {
153
  "epoch": 14.0,
154
  "eval_accuracy": 0.6739130434782609,
155
- "eval_loss": 2.507847381412741e+17,
156
- "eval_runtime": 0.6272,
157
- "eval_samples_per_second": 73.346,
158
- "eval_steps_per_second": 4.783,
159
  "step": 49
160
  },
161
  {
162
  "epoch": 14.29,
163
- "learning_rate": 0.0003564814814814815,
164
- "loss": 2.7830832824707974e+17,
165
  "step": 50
166
  },
167
  {
168
  "epoch": 14.86,
169
- "eval_accuracy": 0.6086956521739131,
170
- "eval_loss": 2.507847381412741e+17,
171
- "eval_runtime": 0.6261,
172
- "eval_samples_per_second": 73.466,
173
- "eval_steps_per_second": 4.791,
174
  "step": 52
175
  },
176
  {
177
  "epoch": 16.0,
178
- "eval_accuracy": 0.5869565217391305,
179
- "eval_loss": 2.507847381412741e+17,
180
- "eval_runtime": 0.6111,
181
- "eval_samples_per_second": 75.27,
182
- "eval_steps_per_second": 4.909,
183
  "step": 56
184
  },
185
  {
186
  "epoch": 16.86,
187
- "eval_accuracy": 0.5869565217391305,
188
- "eval_loss": 2.507847381412741e+17,
189
- "eval_runtime": 0.6977,
190
- "eval_samples_per_second": 65.934,
191
- "eval_steps_per_second": 4.3,
192
  "step": 59
193
  },
194
  {
195
  "epoch": 17.14,
196
- "learning_rate": 0.0003055555555555556,
197
- "loss": 2.6244627872455066e+17,
198
  "step": 60
199
  },
200
  {
201
  "epoch": 18.0,
202
- "eval_accuracy": 0.5652173913043478,
203
- "eval_loss": 2.507847381412741e+17,
204
- "eval_runtime": 0.5916,
205
- "eval_samples_per_second": 77.75,
206
- "eval_steps_per_second": 5.071,
207
  "step": 63
208
  },
209
  {
210
  "epoch": 18.86,
211
- "eval_accuracy": 0.5434782608695652,
212
- "eval_loss": 2.507847381412741e+17,
213
- "eval_runtime": 0.6141,
214
- "eval_samples_per_second": 74.902,
215
- "eval_steps_per_second": 4.885,
216
  "step": 66
217
  },
218
  {
219
  "epoch": 20.0,
220
- "learning_rate": 0.00025462962962962966,
221
- "loss": 2.465840642752774e+17,
222
  "step": 70
223
  },
224
  {
225
  "epoch": 20.0,
226
- "eval_accuracy": 0.5652173913043478,
227
- "eval_loss": 2.507847381412741e+17,
228
- "eval_runtime": 0.5986,
229
- "eval_samples_per_second": 76.841,
230
- "eval_steps_per_second": 5.011,
231
  "step": 70
232
  },
233
  {
234
  "epoch": 20.86,
235
- "eval_accuracy": 0.6086956521739131,
236
- "eval_loss": 2.507847381412741e+17,
237
- "eval_runtime": 0.6411,
238
- "eval_samples_per_second": 71.747,
239
- "eval_steps_per_second": 4.679,
240
  "step": 73
241
  },
242
  {
243
  "epoch": 22.0,
244
- "eval_accuracy": 0.6304347826086957,
245
- "eval_loss": 2.507847381412741e+17,
246
- "eval_runtime": 0.6201,
247
- "eval_samples_per_second": 74.177,
248
- "eval_steps_per_second": 4.838,
249
  "step": 77
250
  },
251
  {
252
  "epoch": 22.86,
253
- "learning_rate": 0.00020370370370370372,
254
- "loss": 2.3793206225188947e+17,
255
  "step": 80
256
  },
257
  {
258
  "epoch": 22.86,
259
- "eval_accuracy": 0.6304347826086957,
260
- "eval_loss": 2.507847381412741e+17,
261
- "eval_runtime": 0.6112,
262
- "eval_samples_per_second": 75.268,
263
- "eval_steps_per_second": 4.909,
264
  "step": 80
265
  },
266
  {
267
  "epoch": 24.0,
268
- "eval_accuracy": 0.6521739130434783,
269
- "eval_loss": 2.507847381412741e+17,
270
- "eval_runtime": 0.6371,
271
- "eval_samples_per_second": 72.197,
272
- "eval_steps_per_second": 4.709,
273
  "step": 84
274
  },
275
  {
276
  "epoch": 24.86,
277
- "eval_accuracy": 0.6956521739130435,
278
- "eval_loss": 2.507847381412741e+17,
279
- "eval_runtime": 0.6086,
280
- "eval_samples_per_second": 75.578,
281
- "eval_steps_per_second": 4.929,
282
  "step": 87
283
  },
284
  {
285
  "epoch": 25.71,
286
- "learning_rate": 0.0001527777777777778,
287
- "loss": 2.3793200727630806e+17,
288
  "step": 90
289
  },
290
  {
291
  "epoch": 26.0,
292
- "eval_accuracy": 0.6739130434782609,
293
- "eval_loss": 2.507847381412741e+17,
294
- "eval_runtime": 0.6036,
295
- "eval_samples_per_second": 76.204,
296
- "eval_steps_per_second": 4.97,
297
  "step": 91
298
  },
299
  {
300
  "epoch": 26.86,
301
- "eval_accuracy": 0.6739130434782609,
302
- "eval_loss": 2.507847381412741e+17,
303
- "eval_runtime": 0.6431,
304
- "eval_samples_per_second": 71.524,
305
- "eval_steps_per_second": 4.665,
306
  "step": 94
307
  },
308
  {
309
  "epoch": 28.0,
310
- "eval_accuracy": 0.6739130434782609,
311
- "eval_loss": 2.507847381412741e+17,
312
- "eval_runtime": 0.6002,
313
- "eval_samples_per_second": 76.647,
314
- "eval_steps_per_second": 4.999,
315
  "step": 98
316
  },
317
  {
318
  "epoch": 28.57,
319
- "learning_rate": 0.00010185185185185186,
320
- "loss": 2.696562712481104e+17,
321
  "step": 100
322
  },
323
  {
324
  "epoch": 28.86,
325
- "eval_accuracy": 0.6739130434782609,
326
- "eval_loss": 2.507847381412741e+17,
327
- "eval_runtime": 0.5951,
328
- "eval_samples_per_second": 77.294,
329
- "eval_steps_per_second": 5.041,
330
  "step": 101
331
  },
332
  {
333
  "epoch": 30.0,
334
- "eval_accuracy": 0.6739130434782609,
335
- "eval_loss": 2.507847381412741e+17,
336
- "eval_runtime": 0.6152,
337
- "eval_samples_per_second": 74.778,
338
- "eval_steps_per_second": 4.877,
339
  "step": 105
340
  },
341
  {
342
  "epoch": 30.86,
343
- "eval_accuracy": 0.6739130434782609,
344
- "eval_loss": 2.507847381412741e+17,
345
- "eval_runtime": 0.6361,
346
- "eval_samples_per_second": 72.311,
347
- "eval_steps_per_second": 4.716,
348
  "step": 108
349
  },
350
  {
351
  "epoch": 31.43,
352
- "learning_rate": 5.092592592592593e-05,
353
- "loss": 2.4225800828800205e+17,
354
  "step": 110
355
  },
356
  {
357
  "epoch": 32.0,
358
- "eval_accuracy": 0.6739130434782609,
359
- "eval_loss": 2.507847381412741e+17,
360
- "eval_runtime": 0.6477,
361
- "eval_samples_per_second": 71.026,
362
- "eval_steps_per_second": 4.632,
363
  "step": 112
364
  },
365
  {
366
  "epoch": 32.86,
367
- "eval_accuracy": 0.6739130434782609,
368
- "eval_loss": 2.507847381412741e+17,
369
- "eval_runtime": 0.6032,
370
- "eval_samples_per_second": 76.266,
371
- "eval_steps_per_second": 4.974,
372
  "step": 115
373
  },
374
  {
375
  "epoch": 34.0,
376
- "eval_accuracy": 0.6739130434782609,
377
- "eval_loss": 2.507847381412741e+17,
378
- "eval_runtime": 0.6062,
379
- "eval_samples_per_second": 75.888,
380
- "eval_steps_per_second": 4.949,
381
  "step": 119
382
  },
383
  {
384
  "epoch": 34.29,
385
  "learning_rate": 0.0,
386
- "loss": 2.458630595253633e+17,
387
  "step": 120
388
  },
389
  {
390
  "epoch": 34.29,
391
- "eval_accuracy": 0.6739130434782609,
392
- "eval_loss": 2.507847381412741e+17,
393
- "eval_runtime": 0.6181,
394
- "eval_samples_per_second": 74.417,
395
- "eval_steps_per_second": 4.853,
396
  "step": 120
397
  },
398
  {
399
  "epoch": 34.29,
400
  "step": 120,
401
  "total_flos": 2.0027429927092224e+16,
402
- "train_loss": 2.522319634493866e+17,
403
- "train_runtime": 128.6883,
404
- "train_samples_per_second": 66.206,
405
- "train_steps_per_second": 0.932
406
  }
407
  ],
408
  "logging_steps": 10,
 
1
  {
2
+ "best_metric": 0.7391304347826086,
3
+ "best_model_checkpoint": "swiftformer-xs-DMAE\\checkpoint-87",
4
  "epoch": 34.285714285714285,
5
  "eval_steps": 500,
6
  "global_step": 120,
 
10
  "log_history": [
11
  {
12
  "epoch": 0.86,
13
+ "eval_accuracy": 0.32608695652173914,
14
+ "eval_loss": 1.3860292434692383,
15
+ "eval_runtime": 0.6541,
16
+ "eval_samples_per_second": 70.321,
17
+ "eval_steps_per_second": 4.586,
18
  "step": 3
19
  },
20
  {
21
  "epoch": 2.0,
22
+ "eval_accuracy": 0.5,
23
+ "eval_loss": 1.383039951324463,
24
+ "eval_runtime": 0.6141,
25
+ "eval_samples_per_second": 74.902,
26
+ "eval_steps_per_second": 4.885,
27
  "step": 7
28
  },
29
  {
30
  "epoch": 2.86,
31
+ "learning_rate": 0.00016666666666666666,
32
+ "loss": 1.3842,
33
  "step": 10
34
  },
35
  {
36
  "epoch": 2.86,
37
+ "eval_accuracy": 0.5869565217391305,
38
+ "eval_loss": 1.3756428956985474,
39
+ "eval_runtime": 0.6407,
40
+ "eval_samples_per_second": 71.802,
41
+ "eval_steps_per_second": 4.683,
42
  "step": 10
43
  },
44
  {
45
  "epoch": 4.0,
46
+ "eval_accuracy": 0.6086956521739131,
47
+ "eval_loss": 1.3493651151657104,
48
+ "eval_runtime": 0.5981,
49
+ "eval_samples_per_second": 76.904,
50
+ "eval_steps_per_second": 5.015,
51
  "step": 14
52
  },
53
  {
54
  "epoch": 4.86,
55
+ "eval_accuracy": 0.6521739130434783,
56
+ "eval_loss": 1.303892731666565,
57
+ "eval_runtime": 0.6221,
58
+ "eval_samples_per_second": 73.938,
59
+ "eval_steps_per_second": 4.822,
60
  "step": 17
61
  },
62
  {
63
  "epoch": 5.71,
64
+ "learning_rate": 0.0002941176470588235,
65
+ "loss": 1.3427,
66
  "step": 20
67
  },
68
  {
69
  "epoch": 6.0,
70
+ "eval_accuracy": 0.6521739130434783,
71
+ "eval_loss": 1.1961779594421387,
72
+ "eval_runtime": 0.6102,
73
+ "eval_samples_per_second": 75.391,
74
+ "eval_steps_per_second": 4.917,
75
  "step": 21
76
  },
77
  {
78
  "epoch": 6.86,
79
+ "eval_accuracy": 0.6739130434782609,
80
+ "eval_loss": 1.1310076713562012,
81
+ "eval_runtime": 0.6321,
82
+ "eval_samples_per_second": 72.769,
83
+ "eval_steps_per_second": 4.746,
84
  "step": 24
85
  },
86
  {
87
  "epoch": 8.0,
88
+ "eval_accuracy": 0.6521739130434783,
89
+ "eval_loss": 1.07782781124115,
90
+ "eval_runtime": 0.5902,
91
+ "eval_samples_per_second": 77.946,
92
+ "eval_steps_per_second": 5.083,
93
  "step": 28
94
  },
95
  {
96
  "epoch": 8.57,
97
+ "learning_rate": 0.00026470588235294115,
98
+ "loss": 1.1837,
99
  "step": 30
100
  },
101
  {
102
  "epoch": 8.86,
103
+ "eval_accuracy": 0.6956521739130435,
104
+ "eval_loss": 1.027555227279663,
105
+ "eval_runtime": 0.6977,
106
+ "eval_samples_per_second": 65.934,
107
+ "eval_steps_per_second": 4.3,
108
  "step": 31
109
  },
110
  {
111
  "epoch": 10.0,
112
+ "eval_accuracy": 0.717391304347826,
113
+ "eval_loss": 0.9927520751953125,
114
+ "eval_runtime": 0.6266,
115
+ "eval_samples_per_second": 73.407,
116
+ "eval_steps_per_second": 4.787,
117
  "step": 35
118
  },
119
  {
120
  "epoch": 10.86,
121
+ "eval_accuracy": 0.717391304347826,
122
+ "eval_loss": 0.9709724187850952,
123
+ "eval_runtime": 0.6331,
124
+ "eval_samples_per_second": 72.653,
125
+ "eval_steps_per_second": 4.738,
126
  "step": 38
127
  },
128
  {
129
  "epoch": 11.43,
130
+ "learning_rate": 0.0002352941176470588,
131
+ "loss": 1.0531,
132
  "step": 40
133
  },
134
  {
135
  "epoch": 12.0,
136
+ "eval_accuracy": 0.6739130434782609,
137
+ "eval_loss": 0.9734796285629272,
138
+ "eval_runtime": 0.6324,
139
+ "eval_samples_per_second": 72.735,
140
+ "eval_steps_per_second": 4.744,
141
  "step": 42
142
  },
143
  {
144
  "epoch": 12.86,
145
+ "eval_accuracy": 0.6521739130434783,
146
+ "eval_loss": 0.9384742379188538,
147
+ "eval_runtime": 0.6201,
148
+ "eval_samples_per_second": 74.177,
149
+ "eval_steps_per_second": 4.838,
150
  "step": 45
151
  },
152
  {
153
  "epoch": 14.0,
154
  "eval_accuracy": 0.6739130434782609,
155
+ "eval_loss": 0.9601091146469116,
156
+ "eval_runtime": 0.6196,
157
+ "eval_samples_per_second": 74.236,
158
+ "eval_steps_per_second": 4.841,
159
  "step": 49
160
  },
161
  {
162
  "epoch": 14.29,
163
+ "learning_rate": 0.00020588235294117645,
164
+ "loss": 0.9883,
165
  "step": 50
166
  },
167
  {
168
  "epoch": 14.86,
169
+ "eval_accuracy": 0.6739130434782609,
170
+ "eval_loss": 0.9368714094161987,
171
+ "eval_runtime": 0.5906,
172
+ "eval_samples_per_second": 77.882,
173
+ "eval_steps_per_second": 5.079,
174
  "step": 52
175
  },
176
  {
177
  "epoch": 16.0,
178
+ "eval_accuracy": 0.6739130434782609,
179
+ "eval_loss": 0.8928157687187195,
180
+ "eval_runtime": 0.6201,
181
+ "eval_samples_per_second": 74.177,
182
+ "eval_steps_per_second": 4.838,
183
  "step": 56
184
  },
185
  {
186
  "epoch": 16.86,
187
+ "eval_accuracy": 0.6739130434782609,
188
+ "eval_loss": 0.8825035691261292,
189
+ "eval_runtime": 0.6252,
190
+ "eval_samples_per_second": 73.581,
191
+ "eval_steps_per_second": 4.799,
192
  "step": 59
193
  },
194
  {
195
  "epoch": 17.14,
196
+ "learning_rate": 0.0001764705882352941,
197
+ "loss": 0.8632,
198
  "step": 60
199
  },
200
  {
201
  "epoch": 18.0,
202
+ "eval_accuracy": 0.6739130434782609,
203
+ "eval_loss": 0.8742826581001282,
204
+ "eval_runtime": 0.6216,
205
+ "eval_samples_per_second": 73.997,
206
+ "eval_steps_per_second": 4.826,
207
  "step": 63
208
  },
209
  {
210
  "epoch": 18.86,
211
+ "eval_accuracy": 0.6739130434782609,
212
+ "eval_loss": 0.8478592038154602,
213
+ "eval_runtime": 0.6091,
214
+ "eval_samples_per_second": 75.517,
215
+ "eval_steps_per_second": 4.925,
216
  "step": 66
217
  },
218
  {
219
  "epoch": 20.0,
220
+ "learning_rate": 0.00014705882352941175,
221
+ "loss": 0.8732,
222
  "step": 70
223
  },
224
  {
225
  "epoch": 20.0,
226
+ "eval_accuracy": 0.6739130434782609,
227
+ "eval_loss": 0.834061324596405,
228
+ "eval_runtime": 0.6332,
229
+ "eval_samples_per_second": 72.651,
230
+ "eval_steps_per_second": 4.738,
231
  "step": 70
232
  },
233
  {
234
  "epoch": 20.86,
235
+ "eval_accuracy": 0.717391304347826,
236
+ "eval_loss": 0.8133650422096252,
237
+ "eval_runtime": 0.7277,
238
+ "eval_samples_per_second": 63.215,
239
+ "eval_steps_per_second": 4.123,
240
  "step": 73
241
  },
242
  {
243
  "epoch": 22.0,
244
+ "eval_accuracy": 0.6739130434782609,
245
+ "eval_loss": 0.8844852447509766,
246
+ "eval_runtime": 0.6501,
247
+ "eval_samples_per_second": 70.753,
248
+ "eval_steps_per_second": 4.614,
249
  "step": 77
250
  },
251
  {
252
  "epoch": 22.86,
253
+ "learning_rate": 0.0001176470588235294,
254
+ "loss": 0.7524,
255
  "step": 80
256
  },
257
  {
258
  "epoch": 22.86,
259
+ "eval_accuracy": 0.6521739130434783,
260
+ "eval_loss": 0.9037047624588013,
261
+ "eval_runtime": 0.6262,
262
+ "eval_samples_per_second": 73.464,
263
+ "eval_steps_per_second": 4.791,
264
  "step": 80
265
  },
266
  {
267
  "epoch": 24.0,
268
+ "eval_accuracy": 0.6304347826086957,
269
+ "eval_loss": 0.8921377062797546,
270
+ "eval_runtime": 0.6157,
271
+ "eval_samples_per_second": 74.718,
272
+ "eval_steps_per_second": 4.873,
273
  "step": 84
274
  },
275
  {
276
  "epoch": 24.86,
277
+ "eval_accuracy": 0.7391304347826086,
278
+ "eval_loss": 0.8765698075294495,
279
+ "eval_runtime": 0.6211,
280
+ "eval_samples_per_second": 74.058,
281
+ "eval_steps_per_second": 4.83,
282
  "step": 87
283
  },
284
  {
285
  "epoch": 25.71,
286
+ "learning_rate": 8.823529411764705e-05,
287
+ "loss": 0.6843,
288
  "step": 90
289
  },
290
  {
291
  "epoch": 26.0,
292
+ "eval_accuracy": 0.7391304347826086,
293
+ "eval_loss": 0.8706851005554199,
294
+ "eval_runtime": 0.6322,
295
+ "eval_samples_per_second": 72.766,
296
+ "eval_steps_per_second": 4.746,
297
  "step": 91
298
  },
299
  {
300
  "epoch": 26.86,
301
+ "eval_accuracy": 0.7391304347826086,
302
+ "eval_loss": 0.8922207951545715,
303
+ "eval_runtime": 0.6461,
304
+ "eval_samples_per_second": 71.192,
305
+ "eval_steps_per_second": 4.643,
306
  "step": 94
307
  },
308
  {
309
  "epoch": 28.0,
310
+ "eval_accuracy": 0.6956521739130435,
311
+ "eval_loss": 0.8861702084541321,
312
+ "eval_runtime": 0.6096,
313
+ "eval_samples_per_second": 75.454,
314
+ "eval_steps_per_second": 4.921,
315
  "step": 98
316
  },
317
  {
318
  "epoch": 28.57,
319
+ "learning_rate": 5.88235294117647e-05,
320
+ "loss": 0.7112,
321
  "step": 100
322
  },
323
  {
324
  "epoch": 28.86,
325
+ "eval_accuracy": 0.7391304347826086,
326
+ "eval_loss": 0.8954079747200012,
327
+ "eval_runtime": 0.6282,
328
+ "eval_samples_per_second": 73.23,
329
+ "eval_steps_per_second": 4.776,
330
  "step": 101
331
  },
332
  {
333
  "epoch": 30.0,
334
+ "eval_accuracy": 0.6956521739130435,
335
+ "eval_loss": 0.9228313565254211,
336
+ "eval_runtime": 0.6091,
337
+ "eval_samples_per_second": 75.517,
338
+ "eval_steps_per_second": 4.925,
339
  "step": 105
340
  },
341
  {
342
  "epoch": 30.86,
343
+ "eval_accuracy": 0.6956521739130435,
344
+ "eval_loss": 0.9172708988189697,
345
+ "eval_runtime": 0.6367,
346
+ "eval_samples_per_second": 72.253,
347
+ "eval_steps_per_second": 4.712,
348
  "step": 108
349
  },
350
  {
351
  "epoch": 31.43,
352
+ "learning_rate": 2.941176470588235e-05,
353
+ "loss": 0.6885,
354
  "step": 110
355
  },
356
  {
357
  "epoch": 32.0,
358
+ "eval_accuracy": 0.6956521739130435,
359
+ "eval_loss": 0.9317991137504578,
360
+ "eval_runtime": 0.6256,
361
+ "eval_samples_per_second": 73.524,
362
+ "eval_steps_per_second": 4.795,
363
  "step": 112
364
  },
365
  {
366
  "epoch": 32.86,
367
+ "eval_accuracy": 0.7391304347826086,
368
+ "eval_loss": 0.9049164056777954,
369
+ "eval_runtime": 0.6101,
370
+ "eval_samples_per_second": 75.393,
371
+ "eval_steps_per_second": 4.917,
372
  "step": 115
373
  },
374
  {
375
  "epoch": 34.0,
376
+ "eval_accuracy": 0.717391304347826,
377
+ "eval_loss": 0.915714681148529,
378
+ "eval_runtime": 0.7362,
379
+ "eval_samples_per_second": 62.485,
380
+ "eval_steps_per_second": 4.075,
381
  "step": 119
382
  },
383
  {
384
  "epoch": 34.29,
385
  "learning_rate": 0.0,
386
+ "loss": 0.6734,
387
  "step": 120
388
  },
389
  {
390
  "epoch": 34.29,
391
+ "eval_accuracy": 0.7391304347826086,
392
+ "eval_loss": 0.9050928354263306,
393
+ "eval_runtime": 0.7057,
394
+ "eval_samples_per_second": 65.186,
395
+ "eval_steps_per_second": 4.251,
396
  "step": 120
397
  },
398
  {
399
  "epoch": 34.29,
400
  "step": 120,
401
  "total_flos": 2.0027429927092224e+16,
402
+ "train_loss": 0.9331819852193196,
403
+ "train_runtime": 134.062,
404
+ "train_samples_per_second": 63.553,
405
+ "train_steps_per_second": 0.895
406
  }
407
  ],
408
  "logging_steps": 10,
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c16711f63ef853b11d9c837e60ce87d83693556c4b754b9d71934c0fff892132
3
  size 4728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7673fec7b007a50d6f8a3168a28dc0e69f6038efdf738b3c08f114c5ccf36e4
3
  size 4728