sidushdid commited on
Commit
e20387a
·
verified ·
1 Parent(s): 0d4d397

Upload 9 files

Browse files
all_results.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 143.48,
3
+ "eval_accuracy": 0.9545454545454546,
4
+ "eval_loss": 0.2613329589366913,
5
+ "eval_runtime": 2.0486,
6
+ "eval_samples_per_second": 75.175,
7
+ "eval_steps_per_second": 4.881,
8
+ "test_accuracy": 0.9483870967741935,
9
+ "test_loss": 0.3847464323043823,
10
+ "test_runtime": 1.8406,
11
+ "test_samples_per_second": 84.211,
12
+ "test_steps_per_second": 5.433,
13
+ "total_flos": 8.017005638819359e+18,
14
+ "train_loss": 0.028458750352940775,
15
+ "train_runtime": 2286.6106,
16
+ "train_samples_per_second": 47.297,
17
+ "train_steps_per_second": 0.722
18
+ }
config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/vit-base-patch16-224",
3
+ "architectures": [
4
+ "ViTForImageClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "encoder_stride": 16,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.0,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "benign",
13
+ "1": "malignant",
14
+ "2": "normal"
15
+ },
16
+ "image_size": 224,
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 3072,
19
+ "label2id": {
20
+ "benign": 0,
21
+ "malignant": 1,
22
+ "normal": 2
23
+ },
24
+ "layer_norm_eps": 1e-12,
25
+ "model_type": "vit",
26
+ "num_attention_heads": 12,
27
+ "num_channels": 3,
28
+ "num_hidden_layers": 12,
29
+ "patch_size": 16,
30
+ "problem_type": "single_label_classification",
31
+ "qkv_bias": true,
32
+ "torch_dtype": "float32",
33
+ "transformers_version": "4.35.2"
34
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 143.48,
3
+ "eval_accuracy": 0.9545454545454546,
4
+ "eval_loss": 0.2613329589366913,
5
+ "eval_runtime": 2.0486,
6
+ "eval_samples_per_second": 75.175,
7
+ "eval_steps_per_second": 4.881
8
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24935da771bbd606f46dcd50c7b33ee49771f900acfe2817fbae9e382e968b26
3
+ size 343227052
pred_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "test_accuracy": 0.9483870967741935,
3
+ "test_loss": 0.3847464323043823,
4
+ "test_runtime": 1.8406,
5
+ "test_samples_per_second": 84.211,
6
+ "test_steps_per_second": 5.433
7
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "do_rescale": true,
4
+ "do_resize": true,
5
+ "image_mean": [
6
+ 0.5,
7
+ 0.5,
8
+ 0.5
9
+ ],
10
+ "image_processor_type": "ViTImageProcessor",
11
+ "image_std": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "resample": 2,
17
+ "rescale_factor": 0.00392156862745098,
18
+ "size": {
19
+ "height": 224,
20
+ "width": 224
21
+ }
22
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 143.48,
3
+ "total_flos": 8.017005638819359e+18,
4
+ "train_loss": 0.028458750352940775,
5
+ "train_runtime": 2286.6106,
6
+ "train_samples_per_second": 47.297,
7
+ "train_steps_per_second": 0.722
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1816 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9545454545454546,
3
+ "best_model_checkpoint": "ViT-base-16-224-7.5-1.5-1.5-split-lion-4\\checkpoint-943",
4
+ "epoch": 143.47826086956522,
5
+ "eval_steps": 500,
6
+ "global_step": 1650,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.96,
13
+ "eval_accuracy": 0.7077922077922078,
14
+ "eval_loss": 0.7381948232650757,
15
+ "eval_runtime": 1.6508,
16
+ "eval_samples_per_second": 93.289,
17
+ "eval_steps_per_second": 6.058,
18
+ "step": 11
19
+ },
20
+ {
21
+ "epoch": 1.74,
22
+ "learning_rate": 1.5780272002695102e-05,
23
+ "loss": 0.8102,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 2.0,
28
+ "eval_accuracy": 0.8181818181818182,
29
+ "eval_loss": 0.5202796459197998,
30
+ "eval_runtime": 1.6447,
31
+ "eval_samples_per_second": 93.636,
32
+ "eval_steps_per_second": 6.08,
33
+ "step": 23
34
+ },
35
+ {
36
+ "epoch": 2.96,
37
+ "eval_accuracy": 0.8701298701298701,
38
+ "eval_loss": 0.4076531231403351,
39
+ "eval_runtime": 1.6584,
40
+ "eval_samples_per_second": 92.862,
41
+ "eval_steps_per_second": 6.03,
42
+ "step": 34
43
+ },
44
+ {
45
+ "epoch": 3.48,
46
+ "learning_rate": 1.3309654930856552e-05,
47
+ "loss": 0.4016,
48
+ "step": 40
49
+ },
50
+ {
51
+ "epoch": 4.0,
52
+ "eval_accuracy": 0.8636363636363636,
53
+ "eval_loss": 0.3777826428413391,
54
+ "eval_runtime": 1.6734,
55
+ "eval_samples_per_second": 92.03,
56
+ "eval_steps_per_second": 5.976,
57
+ "step": 46
58
+ },
59
+ {
60
+ "epoch": 4.96,
61
+ "eval_accuracy": 0.8766233766233766,
62
+ "eval_loss": 0.35903552174568176,
63
+ "eval_runtime": 1.6866,
64
+ "eval_samples_per_second": 91.308,
65
+ "eval_steps_per_second": 5.929,
66
+ "step": 57
67
+ },
68
+ {
69
+ "epoch": 5.22,
70
+ "learning_rate": 9.780401480557754e-06,
71
+ "loss": 0.2052,
72
+ "step": 60
73
+ },
74
+ {
75
+ "epoch": 6.0,
76
+ "eval_accuracy": 0.9025974025974026,
77
+ "eval_loss": 0.29355403780937195,
78
+ "eval_runtime": 1.7256,
79
+ "eval_samples_per_second": 89.245,
80
+ "eval_steps_per_second": 5.795,
81
+ "step": 69
82
+ },
83
+ {
84
+ "epoch": 6.96,
85
+ "learning_rate": 5.943306394074249e-06,
86
+ "loss": 0.0838,
87
+ "step": 80
88
+ },
89
+ {
90
+ "epoch": 6.96,
91
+ "eval_accuracy": 0.8961038961038961,
92
+ "eval_loss": 0.2710248529911041,
93
+ "eval_runtime": 1.6959,
94
+ "eval_samples_per_second": 90.805,
95
+ "eval_steps_per_second": 5.896,
96
+ "step": 80
97
+ },
98
+ {
99
+ "epoch": 8.0,
100
+ "eval_accuracy": 0.8961038961038961,
101
+ "eval_loss": 0.2775874435901642,
102
+ "eval_runtime": 1.6736,
103
+ "eval_samples_per_second": 92.018,
104
+ "eval_steps_per_second": 5.975,
105
+ "step": 92
106
+ },
107
+ {
108
+ "epoch": 8.7,
109
+ "learning_rate": 2.6146530177605546e-06,
110
+ "loss": 0.0407,
111
+ "step": 100
112
+ },
113
+ {
114
+ "epoch": 8.96,
115
+ "eval_accuracy": 0.9415584415584416,
116
+ "eval_loss": 0.22880351543426514,
117
+ "eval_runtime": 1.6916,
118
+ "eval_samples_per_second": 91.039,
119
+ "eval_steps_per_second": 5.912,
120
+ "step": 103
121
+ },
122
+ {
123
+ "epoch": 10.0,
124
+ "eval_accuracy": 0.9415584415584416,
125
+ "eval_loss": 0.2207733541727066,
126
+ "eval_runtime": 1.6896,
127
+ "eval_samples_per_second": 91.147,
128
+ "eval_steps_per_second": 5.919,
129
+ "step": 115
130
+ },
131
+ {
132
+ "epoch": 10.43,
133
+ "learning_rate": 5.025614934507641e-07,
134
+ "loss": 0.039,
135
+ "step": 120
136
+ },
137
+ {
138
+ "epoch": 10.96,
139
+ "eval_accuracy": 0.9415584415584416,
140
+ "eval_loss": 0.22476842999458313,
141
+ "eval_runtime": 1.7237,
142
+ "eval_samples_per_second": 89.342,
143
+ "eval_steps_per_second": 5.801,
144
+ "step": 126
145
+ },
146
+ {
147
+ "epoch": 12.0,
148
+ "eval_accuracy": 0.9025974025974026,
149
+ "eval_loss": 0.3085295557975769,
150
+ "eval_runtime": 1.727,
151
+ "eval_samples_per_second": 89.173,
152
+ "eval_steps_per_second": 5.79,
153
+ "step": 138
154
+ },
155
+ {
156
+ "epoch": 12.17,
157
+ "learning_rate": 1.6610319647849526e-05,
158
+ "loss": 0.0324,
159
+ "step": 140
160
+ },
161
+ {
162
+ "epoch": 12.96,
163
+ "eval_accuracy": 0.9285714285714286,
164
+ "eval_loss": 0.26784980297088623,
165
+ "eval_runtime": 1.7027,
166
+ "eval_samples_per_second": 90.446,
167
+ "eval_steps_per_second": 5.873,
168
+ "step": 149
169
+ },
170
+ {
171
+ "epoch": 13.91,
172
+ "learning_rate": 1.529573176177447e-05,
173
+ "loss": 0.022,
174
+ "step": 160
175
+ },
176
+ {
177
+ "epoch": 14.0,
178
+ "eval_accuracy": 0.948051948051948,
179
+ "eval_loss": 0.2529321610927582,
180
+ "eval_runtime": 1.7126,
181
+ "eval_samples_per_second": 89.923,
182
+ "eval_steps_per_second": 5.839,
183
+ "step": 161
184
+ },
185
+ {
186
+ "epoch": 14.96,
187
+ "eval_accuracy": 0.922077922077922,
188
+ "eval_loss": 0.24028430879116058,
189
+ "eval_runtime": 1.7176,
190
+ "eval_samples_per_second": 89.659,
191
+ "eval_steps_per_second": 5.822,
192
+ "step": 172
193
+ },
194
+ {
195
+ "epoch": 15.65,
196
+ "learning_rate": 1.25e-05,
197
+ "loss": 0.012,
198
+ "step": 180
199
+ },
200
+ {
201
+ "epoch": 16.0,
202
+ "eval_accuracy": 0.9090909090909091,
203
+ "eval_loss": 0.3513343632221222,
204
+ "eval_runtime": 1.7476,
205
+ "eval_samples_per_second": 88.12,
206
+ "eval_steps_per_second": 5.722,
207
+ "step": 184
208
+ },
209
+ {
210
+ "epoch": 16.96,
211
+ "eval_accuracy": 0.935064935064935,
212
+ "eval_loss": 0.3014402389526367,
213
+ "eval_runtime": 1.7276,
214
+ "eval_samples_per_second": 89.139,
215
+ "eval_steps_per_second": 5.788,
216
+ "step": 195
217
+ },
218
+ {
219
+ "epoch": 17.39,
220
+ "learning_rate": 8.817873574253966e-06,
221
+ "loss": 0.0097,
222
+ "step": 200
223
+ },
224
+ {
225
+ "epoch": 18.0,
226
+ "eval_accuracy": 0.935064935064935,
227
+ "eval_loss": 0.31746622920036316,
228
+ "eval_runtime": 1.9143,
229
+ "eval_samples_per_second": 80.446,
230
+ "eval_steps_per_second": 5.224,
231
+ "step": 207
232
+ },
233
+ {
234
+ "epoch": 18.96,
235
+ "eval_accuracy": 0.935064935064935,
236
+ "eval_loss": 0.2747339904308319,
237
+ "eval_runtime": 1.7322,
238
+ "eval_samples_per_second": 88.904,
239
+ "eval_steps_per_second": 5.773,
240
+ "step": 218
241
+ },
242
+ {
243
+ "epoch": 19.13,
244
+ "learning_rate": 5.03266861634036e-06,
245
+ "loss": 0.0052,
246
+ "step": 220
247
+ },
248
+ {
249
+ "epoch": 20.0,
250
+ "eval_accuracy": 0.935064935064935,
251
+ "eval_loss": 0.2932997941970825,
252
+ "eval_runtime": 1.7167,
253
+ "eval_samples_per_second": 89.706,
254
+ "eval_steps_per_second": 5.825,
255
+ "step": 230
256
+ },
257
+ {
258
+ "epoch": 20.87,
259
+ "learning_rate": 1.9496296406751813e-06,
260
+ "loss": 0.009,
261
+ "step": 240
262
+ },
263
+ {
264
+ "epoch": 20.96,
265
+ "eval_accuracy": 0.9415584415584416,
266
+ "eval_loss": 0.28077924251556396,
267
+ "eval_runtime": 1.7005,
268
+ "eval_samples_per_second": 90.563,
269
+ "eval_steps_per_second": 5.881,
270
+ "step": 241
271
+ },
272
+ {
273
+ "epoch": 22.0,
274
+ "eval_accuracy": 0.935064935064935,
275
+ "eval_loss": 0.29576078057289124,
276
+ "eval_runtime": 1.7368,
277
+ "eval_samples_per_second": 88.67,
278
+ "eval_steps_per_second": 5.758,
279
+ "step": 253
280
+ },
281
+ {
282
+ "epoch": 22.61,
283
+ "learning_rate": 2.246260785014683e-07,
284
+ "loss": 0.0115,
285
+ "step": 260
286
+ },
287
+ {
288
+ "epoch": 22.96,
289
+ "eval_accuracy": 0.935064935064935,
290
+ "eval_loss": 0.2983975112438202,
291
+ "eval_runtime": 1.7116,
292
+ "eval_samples_per_second": 89.976,
293
+ "eval_steps_per_second": 5.843,
294
+ "step": 264
295
+ },
296
+ {
297
+ "epoch": 24.0,
298
+ "eval_accuracy": 0.9285714285714286,
299
+ "eval_loss": 0.3521440327167511,
300
+ "eval_runtime": 1.7503,
301
+ "eval_samples_per_second": 87.986,
302
+ "eval_steps_per_second": 5.713,
303
+ "step": 276
304
+ },
305
+ {
306
+ "epoch": 24.35,
307
+ "learning_rate": 1.64420405881652e-05,
308
+ "loss": 0.0104,
309
+ "step": 280
310
+ },
311
+ {
312
+ "epoch": 24.96,
313
+ "eval_accuracy": 0.9025974025974026,
314
+ "eval_loss": 0.4289417862892151,
315
+ "eval_runtime": 1.7466,
316
+ "eval_samples_per_second": 88.173,
317
+ "eval_steps_per_second": 5.726,
318
+ "step": 287
319
+ },
320
+ {
321
+ "epoch": 26.0,
322
+ "eval_accuracy": 0.9025974025974026,
323
+ "eval_loss": 0.6231942176818848,
324
+ "eval_runtime": 1.7359,
325
+ "eval_samples_per_second": 88.715,
326
+ "eval_steps_per_second": 5.761,
327
+ "step": 299
328
+ },
329
+ {
330
+ "epoch": 26.09,
331
+ "learning_rate": 1.4717037025991483e-05,
332
+ "loss": 0.0086,
333
+ "step": 300
334
+ },
335
+ {
336
+ "epoch": 26.96,
337
+ "eval_accuracy": 0.9090909090909091,
338
+ "eval_loss": 0.5161650776863098,
339
+ "eval_runtime": 1.7289,
340
+ "eval_samples_per_second": 89.075,
341
+ "eval_steps_per_second": 5.784,
342
+ "step": 310
343
+ },
344
+ {
345
+ "epoch": 27.83,
346
+ "learning_rate": 1.1633998050326307e-05,
347
+ "loss": 0.0205,
348
+ "step": 320
349
+ },
350
+ {
351
+ "epoch": 28.0,
352
+ "eval_accuracy": 0.8896103896103896,
353
+ "eval_loss": 0.6762561798095703,
354
+ "eval_runtime": 1.7546,
355
+ "eval_samples_per_second": 87.77,
356
+ "eval_steps_per_second": 5.699,
357
+ "step": 322
358
+ },
359
+ {
360
+ "epoch": 28.96,
361
+ "eval_accuracy": 0.9155844155844156,
362
+ "eval_loss": 0.4664335250854492,
363
+ "eval_runtime": 1.7617,
364
+ "eval_samples_per_second": 87.414,
365
+ "eval_steps_per_second": 5.676,
366
+ "step": 333
367
+ },
368
+ {
369
+ "epoch": 29.57,
370
+ "learning_rate": 7.848793092412702e-06,
371
+ "loss": 0.012,
372
+ "step": 340
373
+ },
374
+ {
375
+ "epoch": 30.0,
376
+ "eval_accuracy": 0.9285714285714286,
377
+ "eval_loss": 0.3841441869735718,
378
+ "eval_runtime": 1.7019,
379
+ "eval_samples_per_second": 90.487,
380
+ "eval_steps_per_second": 5.876,
381
+ "step": 345
382
+ },
383
+ {
384
+ "epoch": 30.96,
385
+ "eval_accuracy": 0.9155844155844156,
386
+ "eval_loss": 0.42462781071662903,
387
+ "eval_runtime": 1.741,
388
+ "eval_samples_per_second": 88.455,
389
+ "eval_steps_per_second": 5.744,
390
+ "step": 356
391
+ },
392
+ {
393
+ "epoch": 31.3,
394
+ "learning_rate": 4.1666666666666686e-06,
395
+ "loss": 0.0061,
396
+ "step": 360
397
+ },
398
+ {
399
+ "epoch": 32.0,
400
+ "eval_accuracy": 0.9155844155844156,
401
+ "eval_loss": 0.43087005615234375,
402
+ "eval_runtime": 1.7385,
403
+ "eval_samples_per_second": 88.582,
404
+ "eval_steps_per_second": 5.752,
405
+ "step": 368
406
+ },
407
+ {
408
+ "epoch": 32.96,
409
+ "eval_accuracy": 0.9155844155844156,
410
+ "eval_loss": 0.45713886618614197,
411
+ "eval_runtime": 1.7098,
412
+ "eval_samples_per_second": 90.069,
413
+ "eval_steps_per_second": 5.849,
414
+ "step": 379
415
+ },
416
+ {
417
+ "epoch": 33.04,
418
+ "learning_rate": 1.3709349048921951e-06,
419
+ "loss": 0.0093,
420
+ "step": 380
421
+ },
422
+ {
423
+ "epoch": 34.0,
424
+ "eval_accuracy": 0.9090909090909091,
425
+ "eval_loss": 0.48606640100479126,
426
+ "eval_runtime": 1.7668,
427
+ "eval_samples_per_second": 87.163,
428
+ "eval_steps_per_second": 5.66,
429
+ "step": 391
430
+ },
431
+ {
432
+ "epoch": 34.78,
433
+ "learning_rate": 5.634701881714148e-08,
434
+ "loss": 0.0101,
435
+ "step": 400
436
+ },
437
+ {
438
+ "epoch": 34.96,
439
+ "eval_accuracy": 0.9090909090909091,
440
+ "eval_loss": 0.49100440740585327,
441
+ "eval_runtime": 1.7267,
442
+ "eval_samples_per_second": 89.186,
443
+ "eval_steps_per_second": 5.791,
444
+ "step": 402
445
+ },
446
+ {
447
+ "epoch": 36.0,
448
+ "eval_accuracy": 0.8961038961038961,
449
+ "eval_loss": 0.5978976488113403,
450
+ "eval_runtime": 1.7357,
451
+ "eval_samples_per_second": 88.725,
452
+ "eval_steps_per_second": 5.761,
453
+ "step": 414
454
+ },
455
+ {
456
+ "epoch": 36.52,
457
+ "learning_rate": 1.6164105173215904e-05,
458
+ "loss": 0.011,
459
+ "step": 420
460
+ },
461
+ {
462
+ "epoch": 36.96,
463
+ "eval_accuracy": 0.9090909090909091,
464
+ "eval_loss": 0.3977736830711365,
465
+ "eval_runtime": 1.7968,
466
+ "eval_samples_per_second": 85.708,
467
+ "eval_steps_per_second": 5.565,
468
+ "step": 425
469
+ },
470
+ {
471
+ "epoch": 38.0,
472
+ "eval_accuracy": 0.935064935064935,
473
+ "eval_loss": 0.40591639280319214,
474
+ "eval_runtime": 1.7136,
475
+ "eval_samples_per_second": 89.87,
476
+ "eval_steps_per_second": 5.836,
477
+ "step": 437
478
+ },
479
+ {
480
+ "epoch": 38.26,
481
+ "learning_rate": 1.4052013648906114e-05,
482
+ "loss": 0.0226,
483
+ "step": 440
484
+ },
485
+ {
486
+ "epoch": 38.96,
487
+ "eval_accuracy": 0.9090909090909091,
488
+ "eval_loss": 0.4942101240158081,
489
+ "eval_runtime": 1.7437,
490
+ "eval_samples_per_second": 88.318,
491
+ "eval_steps_per_second": 5.735,
492
+ "step": 448
493
+ },
494
+ {
495
+ "epoch": 40.0,
496
+ "learning_rate": 1.0723360272592418e-05,
497
+ "loss": 0.0118,
498
+ "step": 460
499
+ },
500
+ {
501
+ "epoch": 40.0,
502
+ "eval_accuracy": 0.8896103896103896,
503
+ "eval_loss": 0.7233626842498779,
504
+ "eval_runtime": 1.7317,
505
+ "eval_samples_per_second": 88.928,
506
+ "eval_steps_per_second": 5.775,
507
+ "step": 460
508
+ },
509
+ {
510
+ "epoch": 40.96,
511
+ "eval_accuracy": 0.9155844155844156,
512
+ "eval_loss": 0.5826935172080994,
513
+ "eval_runtime": 1.7439,
514
+ "eval_samples_per_second": 88.31,
515
+ "eval_steps_per_second": 5.734,
516
+ "step": 471
517
+ },
518
+ {
519
+ "epoch": 41.74,
520
+ "learning_rate": 6.886265186108914e-06,
521
+ "loss": 0.011,
522
+ "step": 480
523
+ },
524
+ {
525
+ "epoch": 42.0,
526
+ "eval_accuracy": 0.8961038961038961,
527
+ "eval_loss": 0.662550687789917,
528
+ "eval_runtime": 1.7627,
529
+ "eval_samples_per_second": 87.366,
530
+ "eval_steps_per_second": 5.673,
531
+ "step": 483
532
+ },
533
+ {
534
+ "epoch": 42.96,
535
+ "eval_accuracy": 0.9090909090909091,
536
+ "eval_loss": 0.5871102213859558,
537
+ "eval_runtime": 1.7486,
538
+ "eval_samples_per_second": 88.072,
539
+ "eval_steps_per_second": 5.719,
540
+ "step": 494
541
+ },
542
+ {
543
+ "epoch": 43.48,
544
+ "learning_rate": 3.3570117358101172e-06,
545
+ "loss": 0.0003,
546
+ "step": 500
547
+ },
548
+ {
549
+ "epoch": 44.0,
550
+ "eval_accuracy": 0.9090909090909091,
551
+ "eval_loss": 0.5640321969985962,
552
+ "eval_runtime": 1.6975,
553
+ "eval_samples_per_second": 90.721,
554
+ "eval_steps_per_second": 5.891,
555
+ "step": 506
556
+ },
557
+ {
558
+ "epoch": 44.96,
559
+ "eval_accuracy": 0.9090909090909091,
560
+ "eval_loss": 0.5655084848403931,
561
+ "eval_runtime": 1.7184,
562
+ "eval_samples_per_second": 89.619,
563
+ "eval_steps_per_second": 5.819,
564
+ "step": 517
565
+ },
566
+ {
567
+ "epoch": 45.22,
568
+ "learning_rate": 8.863946639715635e-07,
569
+ "loss": 0.0005,
570
+ "step": 520
571
+ },
572
+ {
573
+ "epoch": 46.0,
574
+ "eval_accuracy": 0.9090909090909091,
575
+ "eval_loss": 0.5844298601150513,
576
+ "eval_runtime": 1.7344,
577
+ "eval_samples_per_second": 88.793,
578
+ "eval_steps_per_second": 5.766,
579
+ "step": 529
580
+ },
581
+ {
582
+ "epoch": 46.96,
583
+ "learning_rate": 1.6666666666666667e-05,
584
+ "loss": 0.0064,
585
+ "step": 540
586
+ },
587
+ {
588
+ "epoch": 46.96,
589
+ "eval_accuracy": 0.9090909090909091,
590
+ "eval_loss": 0.5887525677680969,
591
+ "eval_runtime": 1.7094,
592
+ "eval_samples_per_second": 90.091,
593
+ "eval_steps_per_second": 5.85,
594
+ "step": 540
595
+ },
596
+ {
597
+ "epoch": 48.0,
598
+ "eval_accuracy": 0.8896103896103896,
599
+ "eval_loss": 0.6656709313392639,
600
+ "eval_runtime": 1.7414,
601
+ "eval_samples_per_second": 88.435,
602
+ "eval_steps_per_second": 5.743,
603
+ "step": 552
604
+ },
605
+ {
606
+ "epoch": 48.7,
607
+ "learning_rate": 1.5780272002695102e-05,
608
+ "loss": 0.0084,
609
+ "step": 560
610
+ },
611
+ {
612
+ "epoch": 48.96,
613
+ "eval_accuracy": 0.8961038961038961,
614
+ "eval_loss": 0.7156269550323486,
615
+ "eval_runtime": 1.7014,
616
+ "eval_samples_per_second": 90.514,
617
+ "eval_steps_per_second": 5.878,
618
+ "step": 563
619
+ },
620
+ {
621
+ "epoch": 50.0,
622
+ "eval_accuracy": 0.8766233766233766,
623
+ "eval_loss": 0.9346238970756531,
624
+ "eval_runtime": 1.7164,
625
+ "eval_samples_per_second": 89.724,
626
+ "eval_steps_per_second": 5.826,
627
+ "step": 575
628
+ },
629
+ {
630
+ "epoch": 50.43,
631
+ "learning_rate": 1.3309654930856552e-05,
632
+ "loss": 0.0318,
633
+ "step": 580
634
+ },
635
+ {
636
+ "epoch": 50.96,
637
+ "eval_accuracy": 0.8961038961038961,
638
+ "eval_loss": 0.8030693531036377,
639
+ "eval_runtime": 1.6994,
640
+ "eval_samples_per_second": 90.621,
641
+ "eval_steps_per_second": 5.884,
642
+ "step": 586
643
+ },
644
+ {
645
+ "epoch": 52.0,
646
+ "eval_accuracy": 0.8961038961038961,
647
+ "eval_loss": 0.5700052976608276,
648
+ "eval_runtime": 1.7294,
649
+ "eval_samples_per_second": 89.049,
650
+ "eval_steps_per_second": 5.782,
651
+ "step": 598
652
+ },
653
+ {
654
+ "epoch": 52.17,
655
+ "learning_rate": 9.780401480557754e-06,
656
+ "loss": 0.0338,
657
+ "step": 600
658
+ },
659
+ {
660
+ "epoch": 52.96,
661
+ "eval_accuracy": 0.9155844155844156,
662
+ "eval_loss": 0.40834710001945496,
663
+ "eval_runtime": 1.6964,
664
+ "eval_samples_per_second": 90.781,
665
+ "eval_steps_per_second": 5.895,
666
+ "step": 609
667
+ },
668
+ {
669
+ "epoch": 53.91,
670
+ "learning_rate": 5.943306394074249e-06,
671
+ "loss": 0.0147,
672
+ "step": 620
673
+ },
674
+ {
675
+ "epoch": 54.0,
676
+ "eval_accuracy": 0.9155844155844156,
677
+ "eval_loss": 0.41295498609542847,
678
+ "eval_runtime": 1.7144,
679
+ "eval_samples_per_second": 89.828,
680
+ "eval_steps_per_second": 5.833,
681
+ "step": 621
682
+ },
683
+ {
684
+ "epoch": 54.96,
685
+ "eval_accuracy": 0.9285714285714286,
686
+ "eval_loss": 0.40372058749198914,
687
+ "eval_runtime": 1.7344,
688
+ "eval_samples_per_second": 88.792,
689
+ "eval_steps_per_second": 5.766,
690
+ "step": 632
691
+ },
692
+ {
693
+ "epoch": 55.65,
694
+ "learning_rate": 2.6146530177605546e-06,
695
+ "loss": 0.0011,
696
+ "step": 640
697
+ },
698
+ {
699
+ "epoch": 56.0,
700
+ "eval_accuracy": 0.9090909090909091,
701
+ "eval_loss": 0.36978378891944885,
702
+ "eval_runtime": 1.7234,
703
+ "eval_samples_per_second": 89.359,
704
+ "eval_steps_per_second": 5.803,
705
+ "step": 644
706
+ },
707
+ {
708
+ "epoch": 56.96,
709
+ "eval_accuracy": 0.9155844155844156,
710
+ "eval_loss": 0.38696253299713135,
711
+ "eval_runtime": 1.7014,
712
+ "eval_samples_per_second": 90.515,
713
+ "eval_steps_per_second": 5.878,
714
+ "step": 655
715
+ },
716
+ {
717
+ "epoch": 57.39,
718
+ "learning_rate": 5.025614934507641e-07,
719
+ "loss": 0.0021,
720
+ "step": 660
721
+ },
722
+ {
723
+ "epoch": 58.0,
724
+ "eval_accuracy": 0.9090909090909091,
725
+ "eval_loss": 0.39924710988998413,
726
+ "eval_runtime": 1.6854,
727
+ "eval_samples_per_second": 91.374,
728
+ "eval_steps_per_second": 5.933,
729
+ "step": 667
730
+ },
731
+ {
732
+ "epoch": 58.96,
733
+ "eval_accuracy": 0.9155844155844156,
734
+ "eval_loss": 0.44311344623565674,
735
+ "eval_runtime": 1.6854,
736
+ "eval_samples_per_second": 91.374,
737
+ "eval_steps_per_second": 5.933,
738
+ "step": 678
739
+ },
740
+ {
741
+ "epoch": 59.13,
742
+ "learning_rate": 1.6610319647849526e-05,
743
+ "loss": 0.0002,
744
+ "step": 680
745
+ },
746
+ {
747
+ "epoch": 60.0,
748
+ "eval_accuracy": 0.9090909090909091,
749
+ "eval_loss": 0.44143199920654297,
750
+ "eval_runtime": 1.7504,
751
+ "eval_samples_per_second": 87.98,
752
+ "eval_steps_per_second": 5.713,
753
+ "step": 690
754
+ },
755
+ {
756
+ "epoch": 60.87,
757
+ "learning_rate": 1.529573176177447e-05,
758
+ "loss": 0.0088,
759
+ "step": 700
760
+ },
761
+ {
762
+ "epoch": 60.96,
763
+ "eval_accuracy": 0.9415584415584416,
764
+ "eval_loss": 0.4255146086215973,
765
+ "eval_runtime": 1.7144,
766
+ "eval_samples_per_second": 89.828,
767
+ "eval_steps_per_second": 5.833,
768
+ "step": 701
769
+ },
770
+ {
771
+ "epoch": 62.0,
772
+ "eval_accuracy": 0.922077922077922,
773
+ "eval_loss": 0.4168331027030945,
774
+ "eval_runtime": 1.7434,
775
+ "eval_samples_per_second": 88.334,
776
+ "eval_steps_per_second": 5.736,
777
+ "step": 713
778
+ },
779
+ {
780
+ "epoch": 62.61,
781
+ "learning_rate": 1.25e-05,
782
+ "loss": 0.0061,
783
+ "step": 720
784
+ },
785
+ {
786
+ "epoch": 62.96,
787
+ "eval_accuracy": 0.922077922077922,
788
+ "eval_loss": 0.49312305450439453,
789
+ "eval_runtime": 1.7334,
790
+ "eval_samples_per_second": 88.843,
791
+ "eval_steps_per_second": 5.769,
792
+ "step": 724
793
+ },
794
+ {
795
+ "epoch": 64.0,
796
+ "eval_accuracy": 0.948051948051948,
797
+ "eval_loss": 0.28522124886512756,
798
+ "eval_runtime": 1.7184,
799
+ "eval_samples_per_second": 89.619,
800
+ "eval_steps_per_second": 5.819,
801
+ "step": 736
802
+ },
803
+ {
804
+ "epoch": 64.35,
805
+ "learning_rate": 8.817873574253966e-06,
806
+ "loss": 0.0179,
807
+ "step": 740
808
+ },
809
+ {
810
+ "epoch": 64.96,
811
+ "eval_accuracy": 0.9285714285714286,
812
+ "eval_loss": 0.44120827317237854,
813
+ "eval_runtime": 1.7364,
814
+ "eval_samples_per_second": 88.69,
815
+ "eval_steps_per_second": 5.759,
816
+ "step": 747
817
+ },
818
+ {
819
+ "epoch": 66.0,
820
+ "eval_accuracy": 0.935064935064935,
821
+ "eval_loss": 0.331409215927124,
822
+ "eval_runtime": 1.7124,
823
+ "eval_samples_per_second": 89.933,
824
+ "eval_steps_per_second": 5.84,
825
+ "step": 759
826
+ },
827
+ {
828
+ "epoch": 66.09,
829
+ "learning_rate": 5.03266861634036e-06,
830
+ "loss": 0.0014,
831
+ "step": 760
832
+ },
833
+ {
834
+ "epoch": 66.96,
835
+ "eval_accuracy": 0.9415584415584416,
836
+ "eval_loss": 0.2971489727497101,
837
+ "eval_runtime": 1.7494,
838
+ "eval_samples_per_second": 88.031,
839
+ "eval_steps_per_second": 5.716,
840
+ "step": 770
841
+ },
842
+ {
843
+ "epoch": 67.83,
844
+ "learning_rate": 1.9496296406751813e-06,
845
+ "loss": 0.0199,
846
+ "step": 780
847
+ },
848
+ {
849
+ "epoch": 68.0,
850
+ "eval_accuracy": 0.9285714285714286,
851
+ "eval_loss": 0.3261590301990509,
852
+ "eval_runtime": 1.7504,
853
+ "eval_samples_per_second": 87.98,
854
+ "eval_steps_per_second": 5.713,
855
+ "step": 782
856
+ },
857
+ {
858
+ "epoch": 68.96,
859
+ "eval_accuracy": 0.935064935064935,
860
+ "eval_loss": 0.3835467994213104,
861
+ "eval_runtime": 1.7824,
862
+ "eval_samples_per_second": 86.401,
863
+ "eval_steps_per_second": 5.61,
864
+ "step": 793
865
+ },
866
+ {
867
+ "epoch": 69.57,
868
+ "learning_rate": 2.246260785014683e-07,
869
+ "loss": 0.0091,
870
+ "step": 800
871
+ },
872
+ {
873
+ "epoch": 70.0,
874
+ "eval_accuracy": 0.9285714285714286,
875
+ "eval_loss": 0.3903743326663971,
876
+ "eval_runtime": 1.8425,
877
+ "eval_samples_per_second": 83.582,
878
+ "eval_steps_per_second": 5.427,
879
+ "step": 805
880
+ },
881
+ {
882
+ "epoch": 70.96,
883
+ "eval_accuracy": 0.9285714285714286,
884
+ "eval_loss": 0.47730717062950134,
885
+ "eval_runtime": 1.8005,
886
+ "eval_samples_per_second": 85.531,
887
+ "eval_steps_per_second": 5.554,
888
+ "step": 816
889
+ },
890
+ {
891
+ "epoch": 71.3,
892
+ "learning_rate": 1.64420405881652e-05,
893
+ "loss": 0.0029,
894
+ "step": 820
895
+ },
896
+ {
897
+ "epoch": 72.0,
898
+ "eval_accuracy": 0.922077922077922,
899
+ "eval_loss": 0.4937627613544464,
900
+ "eval_runtime": 1.7875,
901
+ "eval_samples_per_second": 86.152,
902
+ "eval_steps_per_second": 5.594,
903
+ "step": 828
904
+ },
905
+ {
906
+ "epoch": 72.96,
907
+ "eval_accuracy": 0.948051948051948,
908
+ "eval_loss": 0.456775426864624,
909
+ "eval_runtime": 1.7718,
910
+ "eval_samples_per_second": 86.918,
911
+ "eval_steps_per_second": 5.644,
912
+ "step": 839
913
+ },
914
+ {
915
+ "epoch": 73.04,
916
+ "learning_rate": 1.4717037025991483e-05,
917
+ "loss": 0.0224,
918
+ "step": 840
919
+ },
920
+ {
921
+ "epoch": 74.0,
922
+ "eval_accuracy": 0.9285714285714286,
923
+ "eval_loss": 0.4465040862560272,
924
+ "eval_runtime": 1.7516,
925
+ "eval_samples_per_second": 87.92,
926
+ "eval_steps_per_second": 5.709,
927
+ "step": 851
928
+ },
929
+ {
930
+ "epoch": 74.78,
931
+ "learning_rate": 1.1633998050326307e-05,
932
+ "loss": 0.0045,
933
+ "step": 860
934
+ },
935
+ {
936
+ "epoch": 74.96,
937
+ "eval_accuracy": 0.9155844155844156,
938
+ "eval_loss": 0.6093705296516418,
939
+ "eval_runtime": 1.8028,
940
+ "eval_samples_per_second": 85.421,
941
+ "eval_steps_per_second": 5.547,
942
+ "step": 862
943
+ },
944
+ {
945
+ "epoch": 76.0,
946
+ "eval_accuracy": 0.9155844155844156,
947
+ "eval_loss": 0.6924065947532654,
948
+ "eval_runtime": 1.7779,
949
+ "eval_samples_per_second": 86.617,
950
+ "eval_steps_per_second": 5.624,
951
+ "step": 874
952
+ },
953
+ {
954
+ "epoch": 76.52,
955
+ "learning_rate": 7.848793092412702e-06,
956
+ "loss": 0.0088,
957
+ "step": 880
958
+ },
959
+ {
960
+ "epoch": 76.96,
961
+ "eval_accuracy": 0.9155844155844156,
962
+ "eval_loss": 0.48120445013046265,
963
+ "eval_runtime": 1.7635,
964
+ "eval_samples_per_second": 87.327,
965
+ "eval_steps_per_second": 5.671,
966
+ "step": 885
967
+ },
968
+ {
969
+ "epoch": 78.0,
970
+ "eval_accuracy": 0.9090909090909091,
971
+ "eval_loss": 0.6367415189743042,
972
+ "eval_runtime": 1.7799,
973
+ "eval_samples_per_second": 86.522,
974
+ "eval_steps_per_second": 5.618,
975
+ "step": 897
976
+ },
977
+ {
978
+ "epoch": 78.26,
979
+ "learning_rate": 4.1666666666666686e-06,
980
+ "loss": 0.0033,
981
+ "step": 900
982
+ },
983
+ {
984
+ "epoch": 78.96,
985
+ "eval_accuracy": 0.922077922077922,
986
+ "eval_loss": 0.49070408940315247,
987
+ "eval_runtime": 1.7676,
988
+ "eval_samples_per_second": 87.125,
989
+ "eval_steps_per_second": 5.657,
990
+ "step": 908
991
+ },
992
+ {
993
+ "epoch": 80.0,
994
+ "learning_rate": 1.3709349048921951e-06,
995
+ "loss": 0.0076,
996
+ "step": 920
997
+ },
998
+ {
999
+ "epoch": 80.0,
1000
+ "eval_accuracy": 0.9415584415584416,
1001
+ "eval_loss": 0.31148040294647217,
1002
+ "eval_runtime": 1.7496,
1003
+ "eval_samples_per_second": 88.021,
1004
+ "eval_steps_per_second": 5.716,
1005
+ "step": 920
1006
+ },
1007
+ {
1008
+ "epoch": 80.96,
1009
+ "eval_accuracy": 0.948051948051948,
1010
+ "eval_loss": 0.2700955271720886,
1011
+ "eval_runtime": 1.7827,
1012
+ "eval_samples_per_second": 86.387,
1013
+ "eval_steps_per_second": 5.61,
1014
+ "step": 931
1015
+ },
1016
+ {
1017
+ "epoch": 81.74,
1018
+ "learning_rate": 5.634701881714148e-08,
1019
+ "loss": 0.0002,
1020
+ "step": 940
1021
+ },
1022
+ {
1023
+ "epoch": 82.0,
1024
+ "eval_accuracy": 0.9545454545454546,
1025
+ "eval_loss": 0.2613329589366913,
1026
+ "eval_runtime": 1.7597,
1027
+ "eval_samples_per_second": 87.515,
1028
+ "eval_steps_per_second": 5.683,
1029
+ "step": 943
1030
+ },
1031
+ {
1032
+ "epoch": 82.96,
1033
+ "eval_accuracy": 0.9155844155844156,
1034
+ "eval_loss": 0.40443289279937744,
1035
+ "eval_runtime": 1.7637,
1036
+ "eval_samples_per_second": 87.314,
1037
+ "eval_steps_per_second": 5.67,
1038
+ "step": 954
1039
+ },
1040
+ {
1041
+ "epoch": 83.48,
1042
+ "learning_rate": 1.6164105173215904e-05,
1043
+ "loss": 0.0193,
1044
+ "step": 960
1045
+ },
1046
+ {
1047
+ "epoch": 84.0,
1048
+ "eval_accuracy": 0.9025974025974026,
1049
+ "eval_loss": 0.9613493084907532,
1050
+ "eval_runtime": 1.7854,
1051
+ "eval_samples_per_second": 86.255,
1052
+ "eval_steps_per_second": 5.601,
1053
+ "step": 966
1054
+ },
1055
+ {
1056
+ "epoch": 84.96,
1057
+ "eval_accuracy": 0.9025974025974026,
1058
+ "eval_loss": 0.6934040784835815,
1059
+ "eval_runtime": 1.7617,
1060
+ "eval_samples_per_second": 87.414,
1061
+ "eval_steps_per_second": 5.676,
1062
+ "step": 977
1063
+ },
1064
+ {
1065
+ "epoch": 85.22,
1066
+ "learning_rate": 1.4052013648906114e-05,
1067
+ "loss": 0.0238,
1068
+ "step": 980
1069
+ },
1070
+ {
1071
+ "epoch": 86.0,
1072
+ "eval_accuracy": 0.8896103896103896,
1073
+ "eval_loss": 0.9348794221878052,
1074
+ "eval_runtime": 1.7915,
1075
+ "eval_samples_per_second": 85.962,
1076
+ "eval_steps_per_second": 5.582,
1077
+ "step": 989
1078
+ },
1079
+ {
1080
+ "epoch": 86.96,
1081
+ "learning_rate": 1.0723360272592418e-05,
1082
+ "loss": 0.011,
1083
+ "step": 1000
1084
+ },
1085
+ {
1086
+ "epoch": 86.96,
1087
+ "eval_accuracy": 0.9025974025974026,
1088
+ "eval_loss": 0.8836289644241333,
1089
+ "eval_runtime": 1.8098,
1090
+ "eval_samples_per_second": 85.095,
1091
+ "eval_steps_per_second": 5.526,
1092
+ "step": 1000
1093
+ },
1094
+ {
1095
+ "epoch": 88.0,
1096
+ "eval_accuracy": 0.9025974025974026,
1097
+ "eval_loss": 0.7403988838195801,
1098
+ "eval_runtime": 1.7608,
1099
+ "eval_samples_per_second": 87.459,
1100
+ "eval_steps_per_second": 5.679,
1101
+ "step": 1012
1102
+ },
1103
+ {
1104
+ "epoch": 88.7,
1105
+ "learning_rate": 6.886265186108914e-06,
1106
+ "loss": 0.018,
1107
+ "step": 1020
1108
+ },
1109
+ {
1110
+ "epoch": 88.96,
1111
+ "eval_accuracy": 0.9155844155844156,
1112
+ "eval_loss": 0.5259799957275391,
1113
+ "eval_runtime": 1.7749,
1114
+ "eval_samples_per_second": 86.764,
1115
+ "eval_steps_per_second": 5.634,
1116
+ "step": 1023
1117
+ },
1118
+ {
1119
+ "epoch": 90.0,
1120
+ "eval_accuracy": 0.9155844155844156,
1121
+ "eval_loss": 0.5202356576919556,
1122
+ "eval_runtime": 1.7868,
1123
+ "eval_samples_per_second": 86.187,
1124
+ "eval_steps_per_second": 5.597,
1125
+ "step": 1035
1126
+ },
1127
+ {
1128
+ "epoch": 90.43,
1129
+ "learning_rate": 3.3570117358101172e-06,
1130
+ "loss": 0.0041,
1131
+ "step": 1040
1132
+ },
1133
+ {
1134
+ "epoch": 90.96,
1135
+ "eval_accuracy": 0.922077922077922,
1136
+ "eval_loss": 0.48472902178764343,
1137
+ "eval_runtime": 1.7845,
1138
+ "eval_samples_per_second": 86.297,
1139
+ "eval_steps_per_second": 5.604,
1140
+ "step": 1046
1141
+ },
1142
+ {
1143
+ "epoch": 92.0,
1144
+ "eval_accuracy": 0.9155844155844156,
1145
+ "eval_loss": 0.48678579926490784,
1146
+ "eval_runtime": 1.7317,
1147
+ "eval_samples_per_second": 88.93,
1148
+ "eval_steps_per_second": 5.775,
1149
+ "step": 1058
1150
+ },
1151
+ {
1152
+ "epoch": 92.17,
1153
+ "learning_rate": 8.863946639715635e-07,
1154
+ "loss": 0.001,
1155
+ "step": 1060
1156
+ },
1157
+ {
1158
+ "epoch": 92.96,
1159
+ "eval_accuracy": 0.9090909090909091,
1160
+ "eval_loss": 0.4975196421146393,
1161
+ "eval_runtime": 1.7866,
1162
+ "eval_samples_per_second": 86.198,
1163
+ "eval_steps_per_second": 5.597,
1164
+ "step": 1069
1165
+ },
1166
+ {
1167
+ "epoch": 93.91,
1168
+ "learning_rate": 1.6666666666666667e-05,
1169
+ "loss": 0.0014,
1170
+ "step": 1080
1171
+ },
1172
+ {
1173
+ "epoch": 94.0,
1174
+ "eval_accuracy": 0.922077922077922,
1175
+ "eval_loss": 0.6255179643630981,
1176
+ "eval_runtime": 1.7777,
1177
+ "eval_samples_per_second": 86.631,
1178
+ "eval_steps_per_second": 5.625,
1179
+ "step": 1081
1180
+ },
1181
+ {
1182
+ "epoch": 94.96,
1183
+ "eval_accuracy": 0.8766233766233766,
1184
+ "eval_loss": 0.9968315362930298,
1185
+ "eval_runtime": 1.7619,
1186
+ "eval_samples_per_second": 87.408,
1187
+ "eval_steps_per_second": 5.676,
1188
+ "step": 1092
1189
+ },
1190
+ {
1191
+ "epoch": 95.65,
1192
+ "learning_rate": 1.5780272002695102e-05,
1193
+ "loss": 0.0165,
1194
+ "step": 1100
1195
+ },
1196
+ {
1197
+ "epoch": 96.0,
1198
+ "eval_accuracy": 0.9155844155844156,
1199
+ "eval_loss": 0.6173205971717834,
1200
+ "eval_runtime": 1.8105,
1201
+ "eval_samples_per_second": 85.06,
1202
+ "eval_steps_per_second": 5.523,
1203
+ "step": 1104
1204
+ },
1205
+ {
1206
+ "epoch": 96.96,
1207
+ "eval_accuracy": 0.9155844155844156,
1208
+ "eval_loss": 0.7537987232208252,
1209
+ "eval_runtime": 1.7727,
1210
+ "eval_samples_per_second": 86.873,
1211
+ "eval_steps_per_second": 5.641,
1212
+ "step": 1115
1213
+ },
1214
+ {
1215
+ "epoch": 97.39,
1216
+ "learning_rate": 1.3309654930856552e-05,
1217
+ "loss": 0.013,
1218
+ "step": 1120
1219
+ },
1220
+ {
1221
+ "epoch": 98.0,
1222
+ "eval_accuracy": 0.8766233766233766,
1223
+ "eval_loss": 0.7560279965400696,
1224
+ "eval_runtime": 1.7715,
1225
+ "eval_samples_per_second": 86.931,
1226
+ "eval_steps_per_second": 5.645,
1227
+ "step": 1127
1228
+ },
1229
+ {
1230
+ "epoch": 98.96,
1231
+ "eval_accuracy": 0.9155844155844156,
1232
+ "eval_loss": 0.5807818174362183,
1233
+ "eval_runtime": 1.79,
1234
+ "eval_samples_per_second": 86.032,
1235
+ "eval_steps_per_second": 5.586,
1236
+ "step": 1138
1237
+ },
1238
+ {
1239
+ "epoch": 99.13,
1240
+ "learning_rate": 9.780401480557754e-06,
1241
+ "loss": 0.0237,
1242
+ "step": 1140
1243
+ },
1244
+ {
1245
+ "epoch": 100.0,
1246
+ "eval_accuracy": 0.9090909090909091,
1247
+ "eval_loss": 0.5147651433944702,
1248
+ "eval_runtime": 1.7527,
1249
+ "eval_samples_per_second": 87.864,
1250
+ "eval_steps_per_second": 5.705,
1251
+ "step": 1150
1252
+ },
1253
+ {
1254
+ "epoch": 100.87,
1255
+ "learning_rate": 5.943306394074249e-06,
1256
+ "loss": 0.0061,
1257
+ "step": 1160
1258
+ },
1259
+ {
1260
+ "epoch": 100.96,
1261
+ "eval_accuracy": 0.9090909090909091,
1262
+ "eval_loss": 0.5672047734260559,
1263
+ "eval_runtime": 1.7885,
1264
+ "eval_samples_per_second": 86.106,
1265
+ "eval_steps_per_second": 5.591,
1266
+ "step": 1161
1267
+ },
1268
+ {
1269
+ "epoch": 102.0,
1270
+ "eval_accuracy": 0.9155844155844156,
1271
+ "eval_loss": 0.4343276619911194,
1272
+ "eval_runtime": 1.7759,
1273
+ "eval_samples_per_second": 86.714,
1274
+ "eval_steps_per_second": 5.631,
1275
+ "step": 1173
1276
+ },
1277
+ {
1278
+ "epoch": 102.61,
1279
+ "learning_rate": 2.6146530177605546e-06,
1280
+ "loss": 0.002,
1281
+ "step": 1180
1282
+ },
1283
+ {
1284
+ "epoch": 102.96,
1285
+ "eval_accuracy": 0.9155844155844156,
1286
+ "eval_loss": 0.32392024993896484,
1287
+ "eval_runtime": 1.8027,
1288
+ "eval_samples_per_second": 85.425,
1289
+ "eval_steps_per_second": 5.547,
1290
+ "step": 1184
1291
+ },
1292
+ {
1293
+ "epoch": 104.0,
1294
+ "eval_accuracy": 0.9285714285714286,
1295
+ "eval_loss": 0.2951604127883911,
1296
+ "eval_runtime": 1.8006,
1297
+ "eval_samples_per_second": 85.525,
1298
+ "eval_steps_per_second": 5.554,
1299
+ "step": 1196
1300
+ },
1301
+ {
1302
+ "epoch": 104.35,
1303
+ "learning_rate": 5.025614934507641e-07,
1304
+ "loss": 0.0005,
1305
+ "step": 1200
1306
+ },
1307
+ {
1308
+ "epoch": 104.96,
1309
+ "eval_accuracy": 0.9285714285714286,
1310
+ "eval_loss": 0.2926579713821411,
1311
+ "eval_runtime": 1.7666,
1312
+ "eval_samples_per_second": 87.174,
1313
+ "eval_steps_per_second": 5.661,
1314
+ "step": 1207
1315
+ },
1316
+ {
1317
+ "epoch": 106.0,
1318
+ "eval_accuracy": 0.9285714285714286,
1319
+ "eval_loss": 0.3511568605899811,
1320
+ "eval_runtime": 1.7794,
1321
+ "eval_samples_per_second": 86.546,
1322
+ "eval_steps_per_second": 5.62,
1323
+ "step": 1219
1324
+ },
1325
+ {
1326
+ "epoch": 106.09,
1327
+ "learning_rate": 1.6610319647849526e-05,
1328
+ "loss": 0.0003,
1329
+ "step": 1220
1330
+ },
1331
+ {
1332
+ "epoch": 106.96,
1333
+ "eval_accuracy": 0.922077922077922,
1334
+ "eval_loss": 0.4030219316482544,
1335
+ "eval_runtime": 1.7918,
1336
+ "eval_samples_per_second": 85.945,
1337
+ "eval_steps_per_second": 5.581,
1338
+ "step": 1230
1339
+ },
1340
+ {
1341
+ "epoch": 107.83,
1342
+ "learning_rate": 1.529573176177447e-05,
1343
+ "loss": 0.0023,
1344
+ "step": 1240
1345
+ },
1346
+ {
1347
+ "epoch": 108.0,
1348
+ "eval_accuracy": 0.8896103896103896,
1349
+ "eval_loss": 0.3911021053791046,
1350
+ "eval_runtime": 1.7635,
1351
+ "eval_samples_per_second": 87.328,
1352
+ "eval_steps_per_second": 5.671,
1353
+ "step": 1242
1354
+ },
1355
+ {
1356
+ "epoch": 108.96,
1357
+ "eval_accuracy": 0.9090909090909091,
1358
+ "eval_loss": 0.5156851410865784,
1359
+ "eval_runtime": 1.7917,
1360
+ "eval_samples_per_second": 85.953,
1361
+ "eval_steps_per_second": 5.581,
1362
+ "step": 1253
1363
+ },
1364
+ {
1365
+ "epoch": 109.57,
1366
+ "learning_rate": 1.25e-05,
1367
+ "loss": 0.0114,
1368
+ "step": 1260
1369
+ },
1370
+ {
1371
+ "epoch": 110.0,
1372
+ "eval_accuracy": 0.9155844155844156,
1373
+ "eval_loss": 0.5531629323959351,
1374
+ "eval_runtime": 1.7244,
1375
+ "eval_samples_per_second": 89.307,
1376
+ "eval_steps_per_second": 5.799,
1377
+ "step": 1265
1378
+ },
1379
+ {
1380
+ "epoch": 110.96,
1381
+ "eval_accuracy": 0.9090909090909091,
1382
+ "eval_loss": 0.6447522640228271,
1383
+ "eval_runtime": 1.7304,
1384
+ "eval_samples_per_second": 88.997,
1385
+ "eval_steps_per_second": 5.779,
1386
+ "step": 1276
1387
+ },
1388
+ {
1389
+ "epoch": 111.3,
1390
+ "learning_rate": 8.817873574253966e-06,
1391
+ "loss": 0.0003,
1392
+ "step": 1280
1393
+ },
1394
+ {
1395
+ "epoch": 112.0,
1396
+ "eval_accuracy": 0.9090909090909091,
1397
+ "eval_loss": 0.6866676807403564,
1398
+ "eval_runtime": 1.7494,
1399
+ "eval_samples_per_second": 88.031,
1400
+ "eval_steps_per_second": 5.716,
1401
+ "step": 1288
1402
+ },
1403
+ {
1404
+ "epoch": 112.96,
1405
+ "eval_accuracy": 0.9025974025974026,
1406
+ "eval_loss": 0.7411206960678101,
1407
+ "eval_runtime": 1.7444,
1408
+ "eval_samples_per_second": 88.283,
1409
+ "eval_steps_per_second": 5.733,
1410
+ "step": 1299
1411
+ },
1412
+ {
1413
+ "epoch": 113.04,
1414
+ "learning_rate": 5.03266861634036e-06,
1415
+ "loss": 0.0153,
1416
+ "step": 1300
1417
+ },
1418
+ {
1419
+ "epoch": 114.0,
1420
+ "eval_accuracy": 0.8896103896103896,
1421
+ "eval_loss": 0.7488161325454712,
1422
+ "eval_runtime": 1.7644,
1423
+ "eval_samples_per_second": 87.282,
1424
+ "eval_steps_per_second": 5.668,
1425
+ "step": 1311
1426
+ },
1427
+ {
1428
+ "epoch": 114.78,
1429
+ "learning_rate": 1.9496296406751813e-06,
1430
+ "loss": 0.0039,
1431
+ "step": 1320
1432
+ },
1433
+ {
1434
+ "epoch": 114.96,
1435
+ "eval_accuracy": 0.9025974025974026,
1436
+ "eval_loss": 0.6945769786834717,
1437
+ "eval_runtime": 1.7344,
1438
+ "eval_samples_per_second": 88.792,
1439
+ "eval_steps_per_second": 5.766,
1440
+ "step": 1322
1441
+ },
1442
+ {
1443
+ "epoch": 116.0,
1444
+ "eval_accuracy": 0.9025974025974026,
1445
+ "eval_loss": 0.7218338847160339,
1446
+ "eval_runtime": 1.7644,
1447
+ "eval_samples_per_second": 87.281,
1448
+ "eval_steps_per_second": 5.668,
1449
+ "step": 1334
1450
+ },
1451
+ {
1452
+ "epoch": 116.52,
1453
+ "learning_rate": 2.246260785014683e-07,
1454
+ "loss": 0.0002,
1455
+ "step": 1340
1456
+ },
1457
+ {
1458
+ "epoch": 116.96,
1459
+ "eval_accuracy": 0.9025974025974026,
1460
+ "eval_loss": 0.7305352091789246,
1461
+ "eval_runtime": 1.7634,
1462
+ "eval_samples_per_second": 87.332,
1463
+ "eval_steps_per_second": 5.671,
1464
+ "step": 1345
1465
+ },
1466
+ {
1467
+ "epoch": 118.0,
1468
+ "eval_accuracy": 0.8701298701298701,
1469
+ "eval_loss": 1.0061231851577759,
1470
+ "eval_runtime": 1.7854,
1471
+ "eval_samples_per_second": 86.255,
1472
+ "eval_steps_per_second": 5.601,
1473
+ "step": 1357
1474
+ },
1475
+ {
1476
+ "epoch": 118.26,
1477
+ "learning_rate": 1.64420405881652e-05,
1478
+ "loss": 0.0066,
1479
+ "step": 1360
1480
+ },
1481
+ {
1482
+ "epoch": 118.96,
1483
+ "eval_accuracy": 0.9025974025974026,
1484
+ "eval_loss": 0.5966177582740784,
1485
+ "eval_runtime": 1.7834,
1486
+ "eval_samples_per_second": 86.352,
1487
+ "eval_steps_per_second": 5.607,
1488
+ "step": 1368
1489
+ },
1490
+ {
1491
+ "epoch": 120.0,
1492
+ "learning_rate": 1.4717037025991483e-05,
1493
+ "loss": 0.0083,
1494
+ "step": 1380
1495
+ },
1496
+ {
1497
+ "epoch": 120.0,
1498
+ "eval_accuracy": 0.8636363636363636,
1499
+ "eval_loss": 1.1087840795516968,
1500
+ "eval_runtime": 1.7834,
1501
+ "eval_samples_per_second": 86.352,
1502
+ "eval_steps_per_second": 5.607,
1503
+ "step": 1380
1504
+ },
1505
+ {
1506
+ "epoch": 120.96,
1507
+ "eval_accuracy": 0.8831168831168831,
1508
+ "eval_loss": 0.821342945098877,
1509
+ "eval_runtime": 1.7534,
1510
+ "eval_samples_per_second": 87.83,
1511
+ "eval_steps_per_second": 5.703,
1512
+ "step": 1391
1513
+ },
1514
+ {
1515
+ "epoch": 121.74,
1516
+ "learning_rate": 1.1633998050326307e-05,
1517
+ "loss": 0.0202,
1518
+ "step": 1400
1519
+ },
1520
+ {
1521
+ "epoch": 122.0,
1522
+ "eval_accuracy": 0.9155844155844156,
1523
+ "eval_loss": 0.5226480960845947,
1524
+ "eval_runtime": 1.7954,
1525
+ "eval_samples_per_second": 85.775,
1526
+ "eval_steps_per_second": 5.57,
1527
+ "step": 1403
1528
+ },
1529
+ {
1530
+ "epoch": 122.96,
1531
+ "eval_accuracy": 0.922077922077922,
1532
+ "eval_loss": 0.44853323698043823,
1533
+ "eval_runtime": 1.7774,
1534
+ "eval_samples_per_second": 86.643,
1535
+ "eval_steps_per_second": 5.626,
1536
+ "step": 1414
1537
+ },
1538
+ {
1539
+ "epoch": 123.48,
1540
+ "learning_rate": 7.848793092412702e-06,
1541
+ "loss": 0.0033,
1542
+ "step": 1420
1543
+ },
1544
+ {
1545
+ "epoch": 124.0,
1546
+ "eval_accuracy": 0.9090909090909091,
1547
+ "eval_loss": 0.7281427979469299,
1548
+ "eval_runtime": 1.8004,
1549
+ "eval_samples_per_second": 85.537,
1550
+ "eval_steps_per_second": 5.554,
1551
+ "step": 1426
1552
+ },
1553
+ {
1554
+ "epoch": 124.96,
1555
+ "eval_accuracy": 0.9090909090909091,
1556
+ "eval_loss": 0.6655176281929016,
1557
+ "eval_runtime": 1.7664,
1558
+ "eval_samples_per_second": 87.183,
1559
+ "eval_steps_per_second": 5.661,
1560
+ "step": 1437
1561
+ },
1562
+ {
1563
+ "epoch": 125.22,
1564
+ "learning_rate": 4.1666666666666686e-06,
1565
+ "loss": 0.0185,
1566
+ "step": 1440
1567
+ },
1568
+ {
1569
+ "epoch": 126.0,
1570
+ "eval_accuracy": 0.9155844155844156,
1571
+ "eval_loss": 0.5161400437355042,
1572
+ "eval_runtime": 1.7704,
1573
+ "eval_samples_per_second": 86.986,
1574
+ "eval_steps_per_second": 5.648,
1575
+ "step": 1449
1576
+ },
1577
+ {
1578
+ "epoch": 126.96,
1579
+ "learning_rate": 1.3709349048921951e-06,
1580
+ "loss": 0.0001,
1581
+ "step": 1460
1582
+ },
1583
+ {
1584
+ "epoch": 126.96,
1585
+ "eval_accuracy": 0.9025974025974026,
1586
+ "eval_loss": 0.4859886169433594,
1587
+ "eval_runtime": 1.7874,
1588
+ "eval_samples_per_second": 86.159,
1589
+ "eval_steps_per_second": 5.595,
1590
+ "step": 1460
1591
+ },
1592
+ {
1593
+ "epoch": 128.0,
1594
+ "eval_accuracy": 0.9155844155844156,
1595
+ "eval_loss": 0.4834165573120117,
1596
+ "eval_runtime": 1.7944,
1597
+ "eval_samples_per_second": 85.822,
1598
+ "eval_steps_per_second": 5.573,
1599
+ "step": 1472
1600
+ },
1601
+ {
1602
+ "epoch": 128.7,
1603
+ "learning_rate": 5.634701881714148e-08,
1604
+ "loss": 0.0047,
1605
+ "step": 1480
1606
+ },
1607
+ {
1608
+ "epoch": 128.96,
1609
+ "eval_accuracy": 0.9155844155844156,
1610
+ "eval_loss": 0.48360273241996765,
1611
+ "eval_runtime": 1.7574,
1612
+ "eval_samples_per_second": 87.63,
1613
+ "eval_steps_per_second": 5.69,
1614
+ "step": 1483
1615
+ },
1616
+ {
1617
+ "epoch": 130.0,
1618
+ "eval_accuracy": 0.9155844155844156,
1619
+ "eval_loss": 0.6164301037788391,
1620
+ "eval_runtime": 1.7394,
1621
+ "eval_samples_per_second": 88.537,
1622
+ "eval_steps_per_second": 5.749,
1623
+ "step": 1495
1624
+ },
1625
+ {
1626
+ "epoch": 130.43,
1627
+ "learning_rate": 1.6164105173215904e-05,
1628
+ "loss": 0.011,
1629
+ "step": 1500
1630
+ },
1631
+ {
1632
+ "epoch": 130.96,
1633
+ "eval_accuracy": 0.8961038961038961,
1634
+ "eval_loss": 0.7818012237548828,
1635
+ "eval_runtime": 1.7684,
1636
+ "eval_samples_per_second": 87.085,
1637
+ "eval_steps_per_second": 5.655,
1638
+ "step": 1506
1639
+ },
1640
+ {
1641
+ "epoch": 132.0,
1642
+ "eval_accuracy": 0.8636363636363636,
1643
+ "eval_loss": 0.8022345900535583,
1644
+ "eval_runtime": 1.7544,
1645
+ "eval_samples_per_second": 87.779,
1646
+ "eval_steps_per_second": 5.7,
1647
+ "step": 1518
1648
+ },
1649
+ {
1650
+ "epoch": 132.17,
1651
+ "learning_rate": 1.4052013648906114e-05,
1652
+ "loss": 0.0023,
1653
+ "step": 1520
1654
+ },
1655
+ {
1656
+ "epoch": 132.96,
1657
+ "eval_accuracy": 0.8636363636363636,
1658
+ "eval_loss": 0.8653693199157715,
1659
+ "eval_runtime": 1.7604,
1660
+ "eval_samples_per_second": 87.48,
1661
+ "eval_steps_per_second": 5.681,
1662
+ "step": 1529
1663
+ },
1664
+ {
1665
+ "epoch": 133.91,
1666
+ "learning_rate": 1.0723360272592418e-05,
1667
+ "loss": 0.0222,
1668
+ "step": 1540
1669
+ },
1670
+ {
1671
+ "epoch": 134.0,
1672
+ "eval_accuracy": 0.8896103896103896,
1673
+ "eval_loss": 0.6707515716552734,
1674
+ "eval_runtime": 1.7434,
1675
+ "eval_samples_per_second": 88.334,
1676
+ "eval_steps_per_second": 5.736,
1677
+ "step": 1541
1678
+ },
1679
+ {
1680
+ "epoch": 134.96,
1681
+ "eval_accuracy": 0.935064935064935,
1682
+ "eval_loss": 0.4996984004974365,
1683
+ "eval_runtime": 1.7534,
1684
+ "eval_samples_per_second": 87.83,
1685
+ "eval_steps_per_second": 5.703,
1686
+ "step": 1552
1687
+ },
1688
+ {
1689
+ "epoch": 135.65,
1690
+ "learning_rate": 6.886265186108914e-06,
1691
+ "loss": 0.0126,
1692
+ "step": 1560
1693
+ },
1694
+ {
1695
+ "epoch": 136.0,
1696
+ "eval_accuracy": 0.922077922077922,
1697
+ "eval_loss": 0.5560286641120911,
1698
+ "eval_runtime": 1.7314,
1699
+ "eval_samples_per_second": 88.946,
1700
+ "eval_steps_per_second": 5.776,
1701
+ "step": 1564
1702
+ },
1703
+ {
1704
+ "epoch": 136.96,
1705
+ "eval_accuracy": 0.8961038961038961,
1706
+ "eval_loss": 0.6162758469581604,
1707
+ "eval_runtime": 1.7204,
1708
+ "eval_samples_per_second": 89.515,
1709
+ "eval_steps_per_second": 5.813,
1710
+ "step": 1575
1711
+ },
1712
+ {
1713
+ "epoch": 137.39,
1714
+ "learning_rate": 3.3570117358101172e-06,
1715
+ "loss": 0.014,
1716
+ "step": 1580
1717
+ },
1718
+ {
1719
+ "epoch": 138.0,
1720
+ "eval_accuracy": 0.9025974025974026,
1721
+ "eval_loss": 0.6488694548606873,
1722
+ "eval_runtime": 1.7194,
1723
+ "eval_samples_per_second": 89.567,
1724
+ "eval_steps_per_second": 5.816,
1725
+ "step": 1587
1726
+ },
1727
+ {
1728
+ "epoch": 138.96,
1729
+ "eval_accuracy": 0.8701298701298701,
1730
+ "eval_loss": 0.684516191482544,
1731
+ "eval_runtime": 1.7304,
1732
+ "eval_samples_per_second": 88.997,
1733
+ "eval_steps_per_second": 5.779,
1734
+ "step": 1598
1735
+ },
1736
+ {
1737
+ "epoch": 139.13,
1738
+ "learning_rate": 8.863946639715635e-07,
1739
+ "loss": 0.0088,
1740
+ "step": 1600
1741
+ },
1742
+ {
1743
+ "epoch": 140.0,
1744
+ "eval_accuracy": 0.8766233766233766,
1745
+ "eval_loss": 0.7022619247436523,
1746
+ "eval_runtime": 1.7374,
1747
+ "eval_samples_per_second": 88.639,
1748
+ "eval_steps_per_second": 5.756,
1749
+ "step": 1610
1750
+ },
1751
+ {
1752
+ "epoch": 140.87,
1753
+ "learning_rate": 1.6666666666666667e-05,
1754
+ "loss": 0.0022,
1755
+ "step": 1620
1756
+ },
1757
+ {
1758
+ "epoch": 140.96,
1759
+ "eval_accuracy": 0.8701298701298701,
1760
+ "eval_loss": 0.762221097946167,
1761
+ "eval_runtime": 1.6984,
1762
+ "eval_samples_per_second": 90.675,
1763
+ "eval_steps_per_second": 5.888,
1764
+ "step": 1621
1765
+ },
1766
+ {
1767
+ "epoch": 142.0,
1768
+ "eval_accuracy": 0.8961038961038961,
1769
+ "eval_loss": 0.6736029982566833,
1770
+ "eval_runtime": 1.7594,
1771
+ "eval_samples_per_second": 87.53,
1772
+ "eval_steps_per_second": 5.684,
1773
+ "step": 1633
1774
+ },
1775
+ {
1776
+ "epoch": 142.61,
1777
+ "learning_rate": 1.5780272002695102e-05,
1778
+ "loss": 0.0017,
1779
+ "step": 1640
1780
+ },
1781
+ {
1782
+ "epoch": 142.96,
1783
+ "eval_accuracy": 0.9090909090909091,
1784
+ "eval_loss": 0.5298991799354553,
1785
+ "eval_runtime": 1.6954,
1786
+ "eval_samples_per_second": 90.835,
1787
+ "eval_steps_per_second": 5.898,
1788
+ "step": 1644
1789
+ },
1790
+ {
1791
+ "epoch": 143.48,
1792
+ "eval_accuracy": 0.9025974025974026,
1793
+ "eval_loss": 0.5584802031517029,
1794
+ "eval_runtime": 1.6964,
1795
+ "eval_samples_per_second": 90.782,
1796
+ "eval_steps_per_second": 5.895,
1797
+ "step": 1650
1798
+ },
1799
+ {
1800
+ "epoch": 143.48,
1801
+ "step": 1650,
1802
+ "total_flos": 8.017005638819359e+18,
1803
+ "train_loss": 0.028458750352940775,
1804
+ "train_runtime": 2286.6106,
1805
+ "train_samples_per_second": 47.297,
1806
+ "train_steps_per_second": 0.722
1807
+ }
1808
+ ],
1809
+ "logging_steps": 20,
1810
+ "max_steps": 1650,
1811
+ "num_train_epochs": 150,
1812
+ "save_steps": 500,
1813
+ "total_flos": 8.017005638819359e+18,
1814
+ "trial_name": null,
1815
+ "trial_params": null
1816
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee48722e8d443342cee852ce36ed0c4d288098a9d3d580fb816532456bc52a90
3
+ size 4664