tanlq commited on
Commit
b180bca
1 Parent(s): 9e15ded

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +13 -0
  2. eval_results.json +8 -0
  3. train_results.json +8 -0
  4. trainer_state.json +2860 -0
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.9875,
4
+ "eval_loss": 0.0503208190202713,
5
+ "eval_runtime": 377.3903,
6
+ "eval_samples_per_second": 26.498,
7
+ "eval_steps_per_second": 3.312,
8
+ "total_flos": 1.1623391924884734e+19,
9
+ "train_loss": 0.39782794105462216,
10
+ "train_runtime": 35181.3105,
11
+ "train_samples_per_second": 4.264,
12
+ "train_steps_per_second": 0.133
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.9875,
4
+ "eval_loss": 0.0503208190202713,
5
+ "eval_runtime": 377.3903,
6
+ "eval_samples_per_second": 26.498,
7
+ "eval_steps_per_second": 3.312
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "total_flos": 1.1623391924884734e+19,
4
+ "train_loss": 0.39782794105462216,
5
+ "train_runtime": 35181.3105,
6
+ "train_samples_per_second": 4.264,
7
+ "train_steps_per_second": 0.133
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2860 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9875,
3
+ "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-cifar10/checkpoint-4686",
4
+ "epoch": 2.99968,
5
+ "global_step": 4686,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 1.066098081023454e-06,
13
+ "loss": 2.3047,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 2.132196162046908e-06,
19
+ "loss": 2.3074,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 3.198294243070363e-06,
25
+ "loss": 2.2927,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.03,
30
+ "learning_rate": 4.264392324093816e-06,
31
+ "loss": 2.2861,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 5.330490405117271e-06,
37
+ "loss": 2.2798,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.04,
42
+ "learning_rate": 6.396588486140726e-06,
43
+ "loss": 2.2716,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "learning_rate": 7.4626865671641785e-06,
49
+ "loss": 2.2492,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 8.528784648187633e-06,
55
+ "loss": 2.2232,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.06,
60
+ "learning_rate": 9.594882729211089e-06,
61
+ "loss": 2.2032,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.06,
66
+ "learning_rate": 1.0660980810234541e-05,
67
+ "loss": 2.1674,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.07,
72
+ "learning_rate": 1.1727078891257996e-05,
73
+ "loss": 2.1286,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.08,
78
+ "learning_rate": 1.2793176972281452e-05,
79
+ "loss": 2.0904,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.08,
84
+ "learning_rate": 1.3859275053304904e-05,
85
+ "loss": 2.0274,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.09,
90
+ "learning_rate": 1.4925373134328357e-05,
91
+ "loss": 1.9892,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.1,
96
+ "learning_rate": 1.5991471215351813e-05,
97
+ "loss": 1.9114,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.1,
102
+ "learning_rate": 1.7057569296375266e-05,
103
+ "loss": 1.8492,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.11,
108
+ "learning_rate": 1.812366737739872e-05,
109
+ "loss": 1.7891,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.12,
114
+ "learning_rate": 1.9189765458422178e-05,
115
+ "loss": 1.6913,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.12,
120
+ "learning_rate": 2.025586353944563e-05,
121
+ "loss": 1.5952,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.13,
126
+ "learning_rate": 2.1321961620469083e-05,
127
+ "loss": 1.5431,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.13,
132
+ "learning_rate": 2.238805970149254e-05,
133
+ "loss": 1.4303,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.14,
138
+ "learning_rate": 2.345415778251599e-05,
139
+ "loss": 1.3897,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.15,
144
+ "learning_rate": 2.4520255863539444e-05,
145
+ "loss": 1.2901,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.15,
150
+ "learning_rate": 2.5586353944562904e-05,
151
+ "loss": 1.2243,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.16,
156
+ "learning_rate": 2.6652452025586356e-05,
157
+ "loss": 1.1776,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.17,
162
+ "learning_rate": 2.771855010660981e-05,
163
+ "loss": 1.137,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.17,
168
+ "learning_rate": 2.878464818763326e-05,
169
+ "loss": 1.1444,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.18,
174
+ "learning_rate": 2.9850746268656714e-05,
175
+ "loss": 1.0676,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.19,
180
+ "learning_rate": 3.0916844349680173e-05,
181
+ "loss": 1.0014,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.19,
186
+ "learning_rate": 3.1982942430703626e-05,
187
+ "loss": 0.9586,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.2,
192
+ "learning_rate": 3.304904051172708e-05,
193
+ "loss": 0.9382,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.2,
198
+ "learning_rate": 3.411513859275053e-05,
199
+ "loss": 0.9358,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.21,
204
+ "learning_rate": 3.518123667377399e-05,
205
+ "loss": 0.9262,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.22,
210
+ "learning_rate": 3.624733475479744e-05,
211
+ "loss": 0.844,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.22,
216
+ "learning_rate": 3.73134328358209e-05,
217
+ "loss": 0.8245,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.23,
222
+ "learning_rate": 3.8379530916844355e-05,
223
+ "loss": 0.8027,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.24,
228
+ "learning_rate": 3.944562899786781e-05,
229
+ "loss": 0.7783,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.24,
234
+ "learning_rate": 4.051172707889126e-05,
235
+ "loss": 0.7367,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.25,
240
+ "learning_rate": 4.157782515991471e-05,
241
+ "loss": 0.7316,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.26,
246
+ "learning_rate": 4.2643923240938166e-05,
247
+ "loss": 0.7364,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.26,
252
+ "learning_rate": 4.3710021321961625e-05,
253
+ "loss": 0.7266,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.27,
258
+ "learning_rate": 4.477611940298508e-05,
259
+ "loss": 0.7036,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.28,
264
+ "learning_rate": 4.584221748400853e-05,
265
+ "loss": 0.7039,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.28,
270
+ "learning_rate": 4.690831556503198e-05,
271
+ "loss": 0.6398,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.29,
276
+ "learning_rate": 4.7974413646055436e-05,
277
+ "loss": 0.6432,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.29,
282
+ "learning_rate": 4.904051172707889e-05,
283
+ "loss": 0.5952,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.3,
288
+ "learning_rate": 4.9988143229784203e-05,
289
+ "loss": 0.586,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.31,
294
+ "learning_rate": 4.9869575527626275e-05,
295
+ "loss": 0.5641,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.31,
300
+ "learning_rate": 4.975100782546834e-05,
301
+ "loss": 0.6042,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.32,
306
+ "learning_rate": 4.963244012331041e-05,
307
+ "loss": 0.5384,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.33,
312
+ "learning_rate": 4.9513872421152476e-05,
313
+ "loss": 0.5275,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.33,
318
+ "learning_rate": 4.939530471899455e-05,
319
+ "loss": 0.5853,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.34,
324
+ "learning_rate": 4.927673701683661e-05,
325
+ "loss": 0.6605,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.35,
330
+ "learning_rate": 4.9158169314678684e-05,
331
+ "loss": 0.6153,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.35,
336
+ "learning_rate": 4.903960161252075e-05,
337
+ "loss": 0.5895,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.36,
342
+ "learning_rate": 4.892103391036282e-05,
343
+ "loss": 0.54,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.36,
348
+ "learning_rate": 4.8802466208204886e-05,
349
+ "loss": 0.5053,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.37,
354
+ "learning_rate": 4.868389850604696e-05,
355
+ "loss": 0.4973,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.38,
360
+ "learning_rate": 4.856533080388903e-05,
361
+ "loss": 0.5406,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.38,
366
+ "learning_rate": 4.8446763101731094e-05,
367
+ "loss": 0.4995,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.39,
372
+ "learning_rate": 4.832819539957316e-05,
373
+ "loss": 0.445,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.4,
378
+ "learning_rate": 4.820962769741523e-05,
379
+ "loss": 0.503,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.4,
384
+ "learning_rate": 4.8091059995257295e-05,
385
+ "loss": 0.4842,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.41,
390
+ "learning_rate": 4.7972492293099366e-05,
391
+ "loss": 0.4961,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.42,
396
+ "learning_rate": 4.785392459094143e-05,
397
+ "loss": 0.5153,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.42,
402
+ "learning_rate": 4.77353568887835e-05,
403
+ "loss": 0.4224,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.43,
408
+ "learning_rate": 4.761678918662557e-05,
409
+ "loss": 0.5129,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.44,
414
+ "learning_rate": 4.749822148446763e-05,
415
+ "loss": 0.4767,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.44,
420
+ "learning_rate": 4.7379653782309704e-05,
421
+ "loss": 0.4884,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.45,
426
+ "learning_rate": 4.726108608015177e-05,
427
+ "loss": 0.5008,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.45,
432
+ "learning_rate": 4.714251837799384e-05,
433
+ "loss": 0.4454,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.46,
438
+ "learning_rate": 4.7023950675835905e-05,
439
+ "loss": 0.4604,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.47,
444
+ "learning_rate": 4.690538297367798e-05,
445
+ "loss": 0.5052,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.47,
450
+ "learning_rate": 4.678681527152004e-05,
451
+ "loss": 0.5529,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.48,
456
+ "learning_rate": 4.6668247569362107e-05,
457
+ "loss": 0.4555,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.49,
462
+ "learning_rate": 4.654967986720418e-05,
463
+ "loss": 0.4728,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.49,
468
+ "learning_rate": 4.643111216504624e-05,
469
+ "loss": 0.4299,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.5,
474
+ "learning_rate": 4.6312544462888315e-05,
475
+ "loss": 0.4924,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.51,
480
+ "learning_rate": 4.619397676073038e-05,
481
+ "loss": 0.4357,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.51,
486
+ "learning_rate": 4.607540905857245e-05,
487
+ "loss": 0.3772,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.52,
492
+ "learning_rate": 4.5956841356414516e-05,
493
+ "loss": 0.3952,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.52,
498
+ "learning_rate": 4.583827365425658e-05,
499
+ "loss": 0.5014,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.53,
504
+ "learning_rate": 4.571970595209865e-05,
505
+ "loss": 0.4596,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.54,
510
+ "learning_rate": 4.560113824994072e-05,
511
+ "loss": 0.3872,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.54,
516
+ "learning_rate": 4.548257054778279e-05,
517
+ "loss": 0.3455,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.55,
522
+ "learning_rate": 4.5364002845624853e-05,
523
+ "loss": 0.3995,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.56,
528
+ "learning_rate": 4.5245435143466925e-05,
529
+ "loss": 0.419,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.56,
534
+ "learning_rate": 4.512686744130899e-05,
535
+ "loss": 0.3945,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.57,
540
+ "learning_rate": 4.5008299739151055e-05,
541
+ "loss": 0.4771,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.58,
546
+ "learning_rate": 4.4889732036993126e-05,
547
+ "loss": 0.4154,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.58,
552
+ "learning_rate": 4.477116433483519e-05,
553
+ "loss": 0.4645,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.59,
558
+ "learning_rate": 4.465259663267726e-05,
559
+ "loss": 0.4333,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.6,
564
+ "learning_rate": 4.453402893051933e-05,
565
+ "loss": 0.4068,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.6,
570
+ "learning_rate": 4.44154612283614e-05,
571
+ "loss": 0.4505,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.61,
576
+ "learning_rate": 4.4296893526203464e-05,
577
+ "loss": 0.4101,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.61,
582
+ "learning_rate": 4.417832582404553e-05,
583
+ "loss": 0.3332,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.62,
588
+ "learning_rate": 4.40597581218876e-05,
589
+ "loss": 0.362,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.63,
594
+ "learning_rate": 4.3941190419729665e-05,
595
+ "loss": 0.4307,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.63,
600
+ "learning_rate": 4.382262271757174e-05,
601
+ "loss": 0.4096,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.64,
606
+ "learning_rate": 4.37040550154138e-05,
607
+ "loss": 0.4387,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.65,
612
+ "learning_rate": 4.358548731325587e-05,
613
+ "loss": 0.4004,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.65,
618
+ "learning_rate": 4.346691961109794e-05,
619
+ "loss": 0.3721,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.66,
624
+ "learning_rate": 4.334835190894e-05,
625
+ "loss": 0.4278,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.67,
630
+ "learning_rate": 4.3229784206782074e-05,
631
+ "loss": 0.3304,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.67,
636
+ "learning_rate": 4.311121650462414e-05,
637
+ "loss": 0.3647,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.68,
642
+ "learning_rate": 4.299264880246621e-05,
643
+ "loss": 0.4029,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.68,
648
+ "learning_rate": 4.2874081100308276e-05,
649
+ "loss": 0.3775,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.69,
654
+ "learning_rate": 4.275551339815035e-05,
655
+ "loss": 0.4088,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.7,
660
+ "learning_rate": 4.263694569599241e-05,
661
+ "loss": 0.3043,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.7,
666
+ "learning_rate": 4.251837799383448e-05,
667
+ "loss": 0.4541,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.71,
672
+ "learning_rate": 4.239981029167655e-05,
673
+ "loss": 0.3975,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.72,
678
+ "learning_rate": 4.228124258951861e-05,
679
+ "loss": 0.4155,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.72,
684
+ "learning_rate": 4.2162674887360685e-05,
685
+ "loss": 0.3831,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.73,
690
+ "learning_rate": 4.204410718520275e-05,
691
+ "loss": 0.4689,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.74,
696
+ "learning_rate": 4.192553948304482e-05,
697
+ "loss": 0.3641,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.74,
702
+ "learning_rate": 4.1806971780886886e-05,
703
+ "loss": 0.3748,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.75,
708
+ "learning_rate": 4.168840407872895e-05,
709
+ "loss": 0.4045,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.76,
714
+ "learning_rate": 4.156983637657102e-05,
715
+ "loss": 0.3324,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.76,
720
+ "learning_rate": 4.145126867441309e-05,
721
+ "loss": 0.4005,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.77,
726
+ "learning_rate": 4.133270097225516e-05,
727
+ "loss": 0.3305,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.77,
732
+ "learning_rate": 4.121413327009723e-05,
733
+ "loss": 0.2942,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.78,
738
+ "learning_rate": 4.1095565567939295e-05,
739
+ "loss": 0.397,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.79,
744
+ "learning_rate": 4.097699786578137e-05,
745
+ "loss": 0.3656,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.79,
750
+ "learning_rate": 4.085843016362343e-05,
751
+ "loss": 0.3399,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.8,
756
+ "learning_rate": 4.07398624614655e-05,
757
+ "loss": 0.4376,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.81,
762
+ "learning_rate": 4.062129475930757e-05,
763
+ "loss": 0.3958,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.81,
768
+ "learning_rate": 4.050272705714964e-05,
769
+ "loss": 0.3063,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.82,
774
+ "learning_rate": 4.0384159354991705e-05,
775
+ "loss": 0.3652,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.83,
780
+ "learning_rate": 4.026559165283377e-05,
781
+ "loss": 0.3659,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.83,
786
+ "learning_rate": 4.014702395067584e-05,
787
+ "loss": 0.3644,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.84,
792
+ "learning_rate": 4.0028456248517906e-05,
793
+ "loss": 0.3597,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.84,
798
+ "learning_rate": 3.990988854635998e-05,
799
+ "loss": 0.3046,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.85,
804
+ "learning_rate": 3.979132084420204e-05,
805
+ "loss": 0.2584,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.86,
810
+ "learning_rate": 3.9672753142044114e-05,
811
+ "loss": 0.4337,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.86,
816
+ "learning_rate": 3.955418543988618e-05,
817
+ "loss": 0.3598,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.87,
822
+ "learning_rate": 3.9435617737728243e-05,
823
+ "loss": 0.3817,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.88,
828
+ "learning_rate": 3.9317050035570315e-05,
829
+ "loss": 0.3243,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.88,
834
+ "learning_rate": 3.919848233341238e-05,
835
+ "loss": 0.4476,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.89,
840
+ "learning_rate": 3.907991463125445e-05,
841
+ "loss": 0.3555,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.9,
846
+ "learning_rate": 3.8961346929096516e-05,
847
+ "loss": 0.2918,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.9,
852
+ "learning_rate": 3.884277922693859e-05,
853
+ "loss": 0.3575,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.91,
858
+ "learning_rate": 3.872421152478065e-05,
859
+ "loss": 0.327,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.92,
864
+ "learning_rate": 3.860564382262272e-05,
865
+ "loss": 0.3016,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.92,
870
+ "learning_rate": 3.848707612046479e-05,
871
+ "loss": 0.2876,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.93,
876
+ "learning_rate": 3.8368508418306854e-05,
877
+ "loss": 0.312,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.93,
882
+ "learning_rate": 3.8249940716148926e-05,
883
+ "loss": 0.356,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.94,
888
+ "learning_rate": 3.813137301399099e-05,
889
+ "loss": 0.3757,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.95,
894
+ "learning_rate": 3.801280531183306e-05,
895
+ "loss": 0.3686,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.95,
900
+ "learning_rate": 3.789423760967513e-05,
901
+ "loss": 0.3574,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.96,
906
+ "learning_rate": 3.777566990751719e-05,
907
+ "loss": 0.3462,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.97,
912
+ "learning_rate": 3.765710220535926e-05,
913
+ "loss": 0.3756,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.97,
918
+ "learning_rate": 3.753853450320133e-05,
919
+ "loss": 0.3462,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.98,
924
+ "learning_rate": 3.74199668010434e-05,
925
+ "loss": 0.301,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.99,
930
+ "learning_rate": 3.7301399098885464e-05,
931
+ "loss": 0.2796,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.99,
936
+ "learning_rate": 3.7182831396727536e-05,
937
+ "loss": 0.396,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 1.0,
942
+ "learning_rate": 3.70642636945696e-05,
943
+ "loss": 0.3118,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 1.0,
948
+ "eval_accuracy": 0.9778,
949
+ "eval_loss": 0.11352841556072235,
950
+ "eval_runtime": 795.9024,
951
+ "eval_samples_per_second": 12.564,
952
+ "eval_steps_per_second": 1.571,
953
+ "step": 1562
954
+ },
955
+ {
956
+ "epoch": 1.01,
957
+ "learning_rate": 3.6945695992411666e-05,
958
+ "loss": 0.3787,
959
+ "step": 1570
960
+ },
961
+ {
962
+ "epoch": 1.01,
963
+ "learning_rate": 3.682712829025374e-05,
964
+ "loss": 0.3559,
965
+ "step": 1580
966
+ },
967
+ {
968
+ "epoch": 1.02,
969
+ "learning_rate": 3.67085605880958e-05,
970
+ "loss": 0.354,
971
+ "step": 1590
972
+ },
973
+ {
974
+ "epoch": 1.02,
975
+ "learning_rate": 3.6589992885937874e-05,
976
+ "loss": 0.3028,
977
+ "step": 1600
978
+ },
979
+ {
980
+ "epoch": 1.03,
981
+ "learning_rate": 3.647142518377994e-05,
982
+ "loss": 0.3173,
983
+ "step": 1610
984
+ },
985
+ {
986
+ "epoch": 1.04,
987
+ "learning_rate": 3.635285748162201e-05,
988
+ "loss": 0.3481,
989
+ "step": 1620
990
+ },
991
+ {
992
+ "epoch": 1.04,
993
+ "learning_rate": 3.6234289779464075e-05,
994
+ "loss": 0.294,
995
+ "step": 1630
996
+ },
997
+ {
998
+ "epoch": 1.05,
999
+ "learning_rate": 3.611572207730614e-05,
1000
+ "loss": 0.2834,
1001
+ "step": 1640
1002
+ },
1003
+ {
1004
+ "epoch": 1.06,
1005
+ "learning_rate": 3.599715437514821e-05,
1006
+ "loss": 0.3062,
1007
+ "step": 1650
1008
+ },
1009
+ {
1010
+ "epoch": 1.06,
1011
+ "learning_rate": 3.5878586672990276e-05,
1012
+ "loss": 0.279,
1013
+ "step": 1660
1014
+ },
1015
+ {
1016
+ "epoch": 1.07,
1017
+ "learning_rate": 3.576001897083235e-05,
1018
+ "loss": 0.2972,
1019
+ "step": 1670
1020
+ },
1021
+ {
1022
+ "epoch": 1.08,
1023
+ "learning_rate": 3.564145126867441e-05,
1024
+ "loss": 0.2582,
1025
+ "step": 1680
1026
+ },
1027
+ {
1028
+ "epoch": 1.08,
1029
+ "learning_rate": 3.5522883566516484e-05,
1030
+ "loss": 0.2836,
1031
+ "step": 1690
1032
+ },
1033
+ {
1034
+ "epoch": 1.09,
1035
+ "learning_rate": 3.540431586435855e-05,
1036
+ "loss": 0.3459,
1037
+ "step": 1700
1038
+ },
1039
+ {
1040
+ "epoch": 1.09,
1041
+ "learning_rate": 3.5285748162200614e-05,
1042
+ "loss": 0.3492,
1043
+ "step": 1710
1044
+ },
1045
+ {
1046
+ "epoch": 1.1,
1047
+ "learning_rate": 3.5167180460042685e-05,
1048
+ "loss": 0.3747,
1049
+ "step": 1720
1050
+ },
1051
+ {
1052
+ "epoch": 1.11,
1053
+ "learning_rate": 3.504861275788475e-05,
1054
+ "loss": 0.3482,
1055
+ "step": 1730
1056
+ },
1057
+ {
1058
+ "epoch": 1.11,
1059
+ "learning_rate": 3.493004505572682e-05,
1060
+ "loss": 0.3396,
1061
+ "step": 1740
1062
+ },
1063
+ {
1064
+ "epoch": 1.12,
1065
+ "learning_rate": 3.481147735356889e-05,
1066
+ "loss": 0.2672,
1067
+ "step": 1750
1068
+ },
1069
+ {
1070
+ "epoch": 1.13,
1071
+ "learning_rate": 3.469290965141096e-05,
1072
+ "loss": 0.2904,
1073
+ "step": 1760
1074
+ },
1075
+ {
1076
+ "epoch": 1.13,
1077
+ "learning_rate": 3.457434194925302e-05,
1078
+ "loss": 0.4062,
1079
+ "step": 1770
1080
+ },
1081
+ {
1082
+ "epoch": 1.14,
1083
+ "learning_rate": 3.445577424709509e-05,
1084
+ "loss": 0.3078,
1085
+ "step": 1780
1086
+ },
1087
+ {
1088
+ "epoch": 1.15,
1089
+ "learning_rate": 3.433720654493716e-05,
1090
+ "loss": 0.3316,
1091
+ "step": 1790
1092
+ },
1093
+ {
1094
+ "epoch": 1.15,
1095
+ "learning_rate": 3.4218638842779224e-05,
1096
+ "loss": 0.3346,
1097
+ "step": 1800
1098
+ },
1099
+ {
1100
+ "epoch": 1.16,
1101
+ "learning_rate": 3.4100071140621296e-05,
1102
+ "loss": 0.2539,
1103
+ "step": 1810
1104
+ },
1105
+ {
1106
+ "epoch": 1.17,
1107
+ "learning_rate": 3.398150343846336e-05,
1108
+ "loss": 0.3525,
1109
+ "step": 1820
1110
+ },
1111
+ {
1112
+ "epoch": 1.17,
1113
+ "learning_rate": 3.386293573630543e-05,
1114
+ "loss": 0.3179,
1115
+ "step": 1830
1116
+ },
1117
+ {
1118
+ "epoch": 1.18,
1119
+ "learning_rate": 3.37443680341475e-05,
1120
+ "loss": 0.2437,
1121
+ "step": 1840
1122
+ },
1123
+ {
1124
+ "epoch": 1.18,
1125
+ "learning_rate": 3.362580033198957e-05,
1126
+ "loss": 0.2792,
1127
+ "step": 1850
1128
+ },
1129
+ {
1130
+ "epoch": 1.19,
1131
+ "learning_rate": 3.3507232629831634e-05,
1132
+ "loss": 0.2653,
1133
+ "step": 1860
1134
+ },
1135
+ {
1136
+ "epoch": 1.2,
1137
+ "learning_rate": 3.3388664927673705e-05,
1138
+ "loss": 0.3883,
1139
+ "step": 1870
1140
+ },
1141
+ {
1142
+ "epoch": 1.2,
1143
+ "learning_rate": 3.327009722551577e-05,
1144
+ "loss": 0.3113,
1145
+ "step": 1880
1146
+ },
1147
+ {
1148
+ "epoch": 1.21,
1149
+ "learning_rate": 3.315152952335784e-05,
1150
+ "loss": 0.2777,
1151
+ "step": 1890
1152
+ },
1153
+ {
1154
+ "epoch": 1.22,
1155
+ "learning_rate": 3.3032961821199906e-05,
1156
+ "loss": 0.2625,
1157
+ "step": 1900
1158
+ },
1159
+ {
1160
+ "epoch": 1.22,
1161
+ "learning_rate": 3.291439411904198e-05,
1162
+ "loss": 0.3033,
1163
+ "step": 1910
1164
+ },
1165
+ {
1166
+ "epoch": 1.23,
1167
+ "learning_rate": 3.279582641688404e-05,
1168
+ "loss": 0.3028,
1169
+ "step": 1920
1170
+ },
1171
+ {
1172
+ "epoch": 1.24,
1173
+ "learning_rate": 3.2677258714726114e-05,
1174
+ "loss": 0.3201,
1175
+ "step": 1930
1176
+ },
1177
+ {
1178
+ "epoch": 1.24,
1179
+ "learning_rate": 3.255869101256818e-05,
1180
+ "loss": 0.2813,
1181
+ "step": 1940
1182
+ },
1183
+ {
1184
+ "epoch": 1.25,
1185
+ "learning_rate": 3.244012331041025e-05,
1186
+ "loss": 0.1829,
1187
+ "step": 1950
1188
+ },
1189
+ {
1190
+ "epoch": 1.25,
1191
+ "learning_rate": 3.2321555608252316e-05,
1192
+ "loss": 0.3424,
1193
+ "step": 1960
1194
+ },
1195
+ {
1196
+ "epoch": 1.26,
1197
+ "learning_rate": 3.220298790609439e-05,
1198
+ "loss": 0.2576,
1199
+ "step": 1970
1200
+ },
1201
+ {
1202
+ "epoch": 1.27,
1203
+ "learning_rate": 3.208442020393645e-05,
1204
+ "loss": 0.2928,
1205
+ "step": 1980
1206
+ },
1207
+ {
1208
+ "epoch": 1.27,
1209
+ "learning_rate": 3.196585250177852e-05,
1210
+ "loss": 0.312,
1211
+ "step": 1990
1212
+ },
1213
+ {
1214
+ "epoch": 1.28,
1215
+ "learning_rate": 3.184728479962059e-05,
1216
+ "loss": 0.3058,
1217
+ "step": 2000
1218
+ },
1219
+ {
1220
+ "epoch": 1.29,
1221
+ "learning_rate": 3.172871709746265e-05,
1222
+ "loss": 0.2928,
1223
+ "step": 2010
1224
+ },
1225
+ {
1226
+ "epoch": 1.29,
1227
+ "learning_rate": 3.1610149395304725e-05,
1228
+ "loss": 0.2554,
1229
+ "step": 2020
1230
+ },
1231
+ {
1232
+ "epoch": 1.3,
1233
+ "learning_rate": 3.149158169314679e-05,
1234
+ "loss": 0.2874,
1235
+ "step": 2030
1236
+ },
1237
+ {
1238
+ "epoch": 1.31,
1239
+ "learning_rate": 3.137301399098886e-05,
1240
+ "loss": 0.2791,
1241
+ "step": 2040
1242
+ },
1243
+ {
1244
+ "epoch": 1.31,
1245
+ "learning_rate": 3.1254446288830926e-05,
1246
+ "loss": 0.3548,
1247
+ "step": 2050
1248
+ },
1249
+ {
1250
+ "epoch": 1.32,
1251
+ "learning_rate": 3.113587858667299e-05,
1252
+ "loss": 0.2915,
1253
+ "step": 2060
1254
+ },
1255
+ {
1256
+ "epoch": 1.33,
1257
+ "learning_rate": 3.101731088451506e-05,
1258
+ "loss": 0.221,
1259
+ "step": 2070
1260
+ },
1261
+ {
1262
+ "epoch": 1.33,
1263
+ "learning_rate": 3.089874318235713e-05,
1264
+ "loss": 0.2891,
1265
+ "step": 2080
1266
+ },
1267
+ {
1268
+ "epoch": 1.34,
1269
+ "learning_rate": 3.07801754801992e-05,
1270
+ "loss": 0.2505,
1271
+ "step": 2090
1272
+ },
1273
+ {
1274
+ "epoch": 1.34,
1275
+ "learning_rate": 3.0661607778041264e-05,
1276
+ "loss": 0.2632,
1277
+ "step": 2100
1278
+ },
1279
+ {
1280
+ "epoch": 1.35,
1281
+ "learning_rate": 3.0543040075883335e-05,
1282
+ "loss": 0.2336,
1283
+ "step": 2110
1284
+ },
1285
+ {
1286
+ "epoch": 1.36,
1287
+ "learning_rate": 3.04244723737254e-05,
1288
+ "loss": 0.265,
1289
+ "step": 2120
1290
+ },
1291
+ {
1292
+ "epoch": 1.36,
1293
+ "learning_rate": 3.0305904671567465e-05,
1294
+ "loss": 0.3128,
1295
+ "step": 2130
1296
+ },
1297
+ {
1298
+ "epoch": 1.37,
1299
+ "learning_rate": 3.0187336969409537e-05,
1300
+ "loss": 0.3218,
1301
+ "step": 2140
1302
+ },
1303
+ {
1304
+ "epoch": 1.38,
1305
+ "learning_rate": 3.00687692672516e-05,
1306
+ "loss": 0.2648,
1307
+ "step": 2150
1308
+ },
1309
+ {
1310
+ "epoch": 1.38,
1311
+ "learning_rate": 2.9950201565093673e-05,
1312
+ "loss": 0.3387,
1313
+ "step": 2160
1314
+ },
1315
+ {
1316
+ "epoch": 1.39,
1317
+ "learning_rate": 2.9831633862935738e-05,
1318
+ "loss": 0.2822,
1319
+ "step": 2170
1320
+ },
1321
+ {
1322
+ "epoch": 1.4,
1323
+ "learning_rate": 2.971306616077781e-05,
1324
+ "loss": 0.2359,
1325
+ "step": 2180
1326
+ },
1327
+ {
1328
+ "epoch": 1.4,
1329
+ "learning_rate": 2.9594498458619874e-05,
1330
+ "loss": 0.2544,
1331
+ "step": 2190
1332
+ },
1333
+ {
1334
+ "epoch": 1.41,
1335
+ "learning_rate": 2.947593075646194e-05,
1336
+ "loss": 0.3222,
1337
+ "step": 2200
1338
+ },
1339
+ {
1340
+ "epoch": 1.41,
1341
+ "learning_rate": 2.935736305430401e-05,
1342
+ "loss": 0.3221,
1343
+ "step": 2210
1344
+ },
1345
+ {
1346
+ "epoch": 1.42,
1347
+ "learning_rate": 2.9238795352146075e-05,
1348
+ "loss": 0.2123,
1349
+ "step": 2220
1350
+ },
1351
+ {
1352
+ "epoch": 1.43,
1353
+ "learning_rate": 2.9120227649988147e-05,
1354
+ "loss": 0.3861,
1355
+ "step": 2230
1356
+ },
1357
+ {
1358
+ "epoch": 1.43,
1359
+ "learning_rate": 2.9001659947830212e-05,
1360
+ "loss": 0.2087,
1361
+ "step": 2240
1362
+ },
1363
+ {
1364
+ "epoch": 1.44,
1365
+ "learning_rate": 2.8883092245672283e-05,
1366
+ "loss": 0.3096,
1367
+ "step": 2250
1368
+ },
1369
+ {
1370
+ "epoch": 1.45,
1371
+ "learning_rate": 2.8764524543514348e-05,
1372
+ "loss": 0.3355,
1373
+ "step": 2260
1374
+ },
1375
+ {
1376
+ "epoch": 1.45,
1377
+ "learning_rate": 2.8645956841356413e-05,
1378
+ "loss": 0.2005,
1379
+ "step": 2270
1380
+ },
1381
+ {
1382
+ "epoch": 1.46,
1383
+ "learning_rate": 2.8527389139198485e-05,
1384
+ "loss": 0.3197,
1385
+ "step": 2280
1386
+ },
1387
+ {
1388
+ "epoch": 1.47,
1389
+ "learning_rate": 2.840882143704055e-05,
1390
+ "loss": 0.2659,
1391
+ "step": 2290
1392
+ },
1393
+ {
1394
+ "epoch": 1.47,
1395
+ "learning_rate": 2.829025373488262e-05,
1396
+ "loss": 0.2898,
1397
+ "step": 2300
1398
+ },
1399
+ {
1400
+ "epoch": 1.48,
1401
+ "learning_rate": 2.8171686032724686e-05,
1402
+ "loss": 0.3266,
1403
+ "step": 2310
1404
+ },
1405
+ {
1406
+ "epoch": 1.49,
1407
+ "learning_rate": 2.805311833056675e-05,
1408
+ "loss": 0.2504,
1409
+ "step": 2320
1410
+ },
1411
+ {
1412
+ "epoch": 1.49,
1413
+ "learning_rate": 2.7934550628408822e-05,
1414
+ "loss": 0.1921,
1415
+ "step": 2330
1416
+ },
1417
+ {
1418
+ "epoch": 1.5,
1419
+ "learning_rate": 2.7815982926250887e-05,
1420
+ "loss": 0.2172,
1421
+ "step": 2340
1422
+ },
1423
+ {
1424
+ "epoch": 1.5,
1425
+ "learning_rate": 2.769741522409296e-05,
1426
+ "loss": 0.2125,
1427
+ "step": 2350
1428
+ },
1429
+ {
1430
+ "epoch": 1.51,
1431
+ "learning_rate": 2.7578847521935024e-05,
1432
+ "loss": 0.2438,
1433
+ "step": 2360
1434
+ },
1435
+ {
1436
+ "epoch": 1.52,
1437
+ "learning_rate": 2.7460279819777095e-05,
1438
+ "loss": 0.2219,
1439
+ "step": 2370
1440
+ },
1441
+ {
1442
+ "epoch": 1.52,
1443
+ "learning_rate": 2.734171211761916e-05,
1444
+ "loss": 0.3002,
1445
+ "step": 2380
1446
+ },
1447
+ {
1448
+ "epoch": 1.53,
1449
+ "learning_rate": 2.7223144415461228e-05,
1450
+ "loss": 0.2161,
1451
+ "step": 2390
1452
+ },
1453
+ {
1454
+ "epoch": 1.54,
1455
+ "learning_rate": 2.7104576713303296e-05,
1456
+ "loss": 0.3084,
1457
+ "step": 2400
1458
+ },
1459
+ {
1460
+ "epoch": 1.54,
1461
+ "learning_rate": 2.6986009011145365e-05,
1462
+ "loss": 0.3025,
1463
+ "step": 2410
1464
+ },
1465
+ {
1466
+ "epoch": 1.55,
1467
+ "learning_rate": 2.6867441308987433e-05,
1468
+ "loss": 0.3288,
1469
+ "step": 2420
1470
+ },
1471
+ {
1472
+ "epoch": 1.56,
1473
+ "learning_rate": 2.67488736068295e-05,
1474
+ "loss": 0.2618,
1475
+ "step": 2430
1476
+ },
1477
+ {
1478
+ "epoch": 1.56,
1479
+ "learning_rate": 2.663030590467157e-05,
1480
+ "loss": 0.2627,
1481
+ "step": 2440
1482
+ },
1483
+ {
1484
+ "epoch": 1.57,
1485
+ "learning_rate": 2.6511738202513637e-05,
1486
+ "loss": 0.1764,
1487
+ "step": 2450
1488
+ },
1489
+ {
1490
+ "epoch": 1.57,
1491
+ "learning_rate": 2.6393170500355702e-05,
1492
+ "loss": 0.3657,
1493
+ "step": 2460
1494
+ },
1495
+ {
1496
+ "epoch": 1.58,
1497
+ "learning_rate": 2.6274602798197774e-05,
1498
+ "loss": 0.2533,
1499
+ "step": 2470
1500
+ },
1501
+ {
1502
+ "epoch": 1.59,
1503
+ "learning_rate": 2.615603509603984e-05,
1504
+ "loss": 0.2534,
1505
+ "step": 2480
1506
+ },
1507
+ {
1508
+ "epoch": 1.59,
1509
+ "learning_rate": 2.603746739388191e-05,
1510
+ "loss": 0.2104,
1511
+ "step": 2490
1512
+ },
1513
+ {
1514
+ "epoch": 1.6,
1515
+ "learning_rate": 2.5918899691723975e-05,
1516
+ "loss": 0.2663,
1517
+ "step": 2500
1518
+ },
1519
+ {
1520
+ "epoch": 1.61,
1521
+ "learning_rate": 2.5800331989566047e-05,
1522
+ "loss": 0.237,
1523
+ "step": 2510
1524
+ },
1525
+ {
1526
+ "epoch": 1.61,
1527
+ "learning_rate": 2.568176428740811e-05,
1528
+ "loss": 0.2769,
1529
+ "step": 2520
1530
+ },
1531
+ {
1532
+ "epoch": 1.62,
1533
+ "learning_rate": 2.5563196585250176e-05,
1534
+ "loss": 0.2478,
1535
+ "step": 2530
1536
+ },
1537
+ {
1538
+ "epoch": 1.63,
1539
+ "learning_rate": 2.5444628883092248e-05,
1540
+ "loss": 0.2914,
1541
+ "step": 2540
1542
+ },
1543
+ {
1544
+ "epoch": 1.63,
1545
+ "learning_rate": 2.5326061180934313e-05,
1546
+ "loss": 0.2729,
1547
+ "step": 2550
1548
+ },
1549
+ {
1550
+ "epoch": 1.64,
1551
+ "learning_rate": 2.5207493478776384e-05,
1552
+ "loss": 0.2714,
1553
+ "step": 2560
1554
+ },
1555
+ {
1556
+ "epoch": 1.65,
1557
+ "learning_rate": 2.508892577661845e-05,
1558
+ "loss": 0.2168,
1559
+ "step": 2570
1560
+ },
1561
+ {
1562
+ "epoch": 1.65,
1563
+ "learning_rate": 2.4970358074460517e-05,
1564
+ "loss": 0.2612,
1565
+ "step": 2580
1566
+ },
1567
+ {
1568
+ "epoch": 1.66,
1569
+ "learning_rate": 2.4851790372302586e-05,
1570
+ "loss": 0.2308,
1571
+ "step": 2590
1572
+ },
1573
+ {
1574
+ "epoch": 1.66,
1575
+ "learning_rate": 2.4733222670144654e-05,
1576
+ "loss": 0.2405,
1577
+ "step": 2600
1578
+ },
1579
+ {
1580
+ "epoch": 1.67,
1581
+ "learning_rate": 2.4614654967986722e-05,
1582
+ "loss": 0.335,
1583
+ "step": 2610
1584
+ },
1585
+ {
1586
+ "epoch": 1.68,
1587
+ "learning_rate": 2.449608726582879e-05,
1588
+ "loss": 0.2793,
1589
+ "step": 2620
1590
+ },
1591
+ {
1592
+ "epoch": 1.68,
1593
+ "learning_rate": 2.4377519563670855e-05,
1594
+ "loss": 0.3362,
1595
+ "step": 2630
1596
+ },
1597
+ {
1598
+ "epoch": 1.69,
1599
+ "learning_rate": 2.4258951861512923e-05,
1600
+ "loss": 0.2275,
1601
+ "step": 2640
1602
+ },
1603
+ {
1604
+ "epoch": 1.7,
1605
+ "learning_rate": 2.414038415935499e-05,
1606
+ "loss": 0.3178,
1607
+ "step": 2650
1608
+ },
1609
+ {
1610
+ "epoch": 1.7,
1611
+ "learning_rate": 2.402181645719706e-05,
1612
+ "loss": 0.2352,
1613
+ "step": 2660
1614
+ },
1615
+ {
1616
+ "epoch": 1.71,
1617
+ "learning_rate": 2.3903248755039128e-05,
1618
+ "loss": 0.3398,
1619
+ "step": 2670
1620
+ },
1621
+ {
1622
+ "epoch": 1.72,
1623
+ "learning_rate": 2.3784681052881196e-05,
1624
+ "loss": 0.2568,
1625
+ "step": 2680
1626
+ },
1627
+ {
1628
+ "epoch": 1.72,
1629
+ "learning_rate": 2.3666113350723264e-05,
1630
+ "loss": 0.2928,
1631
+ "step": 2690
1632
+ },
1633
+ {
1634
+ "epoch": 1.73,
1635
+ "learning_rate": 2.354754564856533e-05,
1636
+ "loss": 0.211,
1637
+ "step": 2700
1638
+ },
1639
+ {
1640
+ "epoch": 1.73,
1641
+ "learning_rate": 2.3428977946407397e-05,
1642
+ "loss": 0.2666,
1643
+ "step": 2710
1644
+ },
1645
+ {
1646
+ "epoch": 1.74,
1647
+ "learning_rate": 2.3310410244249466e-05,
1648
+ "loss": 0.2174,
1649
+ "step": 2720
1650
+ },
1651
+ {
1652
+ "epoch": 1.75,
1653
+ "learning_rate": 2.3191842542091534e-05,
1654
+ "loss": 0.2172,
1655
+ "step": 2730
1656
+ },
1657
+ {
1658
+ "epoch": 1.75,
1659
+ "learning_rate": 2.3073274839933602e-05,
1660
+ "loss": 0.2173,
1661
+ "step": 2740
1662
+ },
1663
+ {
1664
+ "epoch": 1.76,
1665
+ "learning_rate": 2.295470713777567e-05,
1666
+ "loss": 0.2972,
1667
+ "step": 2750
1668
+ },
1669
+ {
1670
+ "epoch": 1.77,
1671
+ "learning_rate": 2.283613943561774e-05,
1672
+ "loss": 0.2733,
1673
+ "step": 2760
1674
+ },
1675
+ {
1676
+ "epoch": 1.77,
1677
+ "learning_rate": 2.2717571733459807e-05,
1678
+ "loss": 0.2385,
1679
+ "step": 2770
1680
+ },
1681
+ {
1682
+ "epoch": 1.78,
1683
+ "learning_rate": 2.2599004031301875e-05,
1684
+ "loss": 0.2272,
1685
+ "step": 2780
1686
+ },
1687
+ {
1688
+ "epoch": 1.79,
1689
+ "learning_rate": 2.2480436329143943e-05,
1690
+ "loss": 0.2721,
1691
+ "step": 2790
1692
+ },
1693
+ {
1694
+ "epoch": 1.79,
1695
+ "learning_rate": 2.236186862698601e-05,
1696
+ "loss": 0.3015,
1697
+ "step": 2800
1698
+ },
1699
+ {
1700
+ "epoch": 1.8,
1701
+ "learning_rate": 2.224330092482808e-05,
1702
+ "loss": 0.2605,
1703
+ "step": 2810
1704
+ },
1705
+ {
1706
+ "epoch": 1.81,
1707
+ "learning_rate": 2.2124733222670148e-05,
1708
+ "loss": 0.2269,
1709
+ "step": 2820
1710
+ },
1711
+ {
1712
+ "epoch": 1.81,
1713
+ "learning_rate": 2.2006165520512216e-05,
1714
+ "loss": 0.2565,
1715
+ "step": 2830
1716
+ },
1717
+ {
1718
+ "epoch": 1.82,
1719
+ "learning_rate": 2.188759781835428e-05,
1720
+ "loss": 0.1987,
1721
+ "step": 2840
1722
+ },
1723
+ {
1724
+ "epoch": 1.82,
1725
+ "learning_rate": 2.176903011619635e-05,
1726
+ "loss": 0.2089,
1727
+ "step": 2850
1728
+ },
1729
+ {
1730
+ "epoch": 1.83,
1731
+ "learning_rate": 2.1650462414038417e-05,
1732
+ "loss": 0.2291,
1733
+ "step": 2860
1734
+ },
1735
+ {
1736
+ "epoch": 1.84,
1737
+ "learning_rate": 2.1531894711880485e-05,
1738
+ "loss": 0.202,
1739
+ "step": 2870
1740
+ },
1741
+ {
1742
+ "epoch": 1.84,
1743
+ "learning_rate": 2.1413327009722553e-05,
1744
+ "loss": 0.1743,
1745
+ "step": 2880
1746
+ },
1747
+ {
1748
+ "epoch": 1.85,
1749
+ "learning_rate": 2.129475930756462e-05,
1750
+ "loss": 0.2867,
1751
+ "step": 2890
1752
+ },
1753
+ {
1754
+ "epoch": 1.86,
1755
+ "learning_rate": 2.117619160540669e-05,
1756
+ "loss": 0.3186,
1757
+ "step": 2900
1758
+ },
1759
+ {
1760
+ "epoch": 1.86,
1761
+ "learning_rate": 2.1057623903248755e-05,
1762
+ "loss": 0.1718,
1763
+ "step": 2910
1764
+ },
1765
+ {
1766
+ "epoch": 1.87,
1767
+ "learning_rate": 2.0939056201090823e-05,
1768
+ "loss": 0.2629,
1769
+ "step": 2920
1770
+ },
1771
+ {
1772
+ "epoch": 1.88,
1773
+ "learning_rate": 2.082048849893289e-05,
1774
+ "loss": 0.2223,
1775
+ "step": 2930
1776
+ },
1777
+ {
1778
+ "epoch": 1.88,
1779
+ "learning_rate": 2.070192079677496e-05,
1780
+ "loss": 0.1843,
1781
+ "step": 2940
1782
+ },
1783
+ {
1784
+ "epoch": 1.89,
1785
+ "learning_rate": 2.0583353094617028e-05,
1786
+ "loss": 0.2813,
1787
+ "step": 2950
1788
+ },
1789
+ {
1790
+ "epoch": 1.89,
1791
+ "learning_rate": 2.0464785392459096e-05,
1792
+ "loss": 0.28,
1793
+ "step": 2960
1794
+ },
1795
+ {
1796
+ "epoch": 1.9,
1797
+ "learning_rate": 2.0346217690301164e-05,
1798
+ "loss": 0.2057,
1799
+ "step": 2970
1800
+ },
1801
+ {
1802
+ "epoch": 1.91,
1803
+ "learning_rate": 2.022764998814323e-05,
1804
+ "loss": 0.3016,
1805
+ "step": 2980
1806
+ },
1807
+ {
1808
+ "epoch": 1.91,
1809
+ "learning_rate": 2.0109082285985297e-05,
1810
+ "loss": 0.2579,
1811
+ "step": 2990
1812
+ },
1813
+ {
1814
+ "epoch": 1.92,
1815
+ "learning_rate": 1.9990514583827365e-05,
1816
+ "loss": 0.2419,
1817
+ "step": 3000
1818
+ },
1819
+ {
1820
+ "epoch": 1.93,
1821
+ "learning_rate": 1.9871946881669433e-05,
1822
+ "loss": 0.2841,
1823
+ "step": 3010
1824
+ },
1825
+ {
1826
+ "epoch": 1.93,
1827
+ "learning_rate": 1.97533791795115e-05,
1828
+ "loss": 0.227,
1829
+ "step": 3020
1830
+ },
1831
+ {
1832
+ "epoch": 1.94,
1833
+ "learning_rate": 1.963481147735357e-05,
1834
+ "loss": 0.2019,
1835
+ "step": 3030
1836
+ },
1837
+ {
1838
+ "epoch": 1.95,
1839
+ "learning_rate": 1.9516243775195638e-05,
1840
+ "loss": 0.296,
1841
+ "step": 3040
1842
+ },
1843
+ {
1844
+ "epoch": 1.95,
1845
+ "learning_rate": 1.9397676073037703e-05,
1846
+ "loss": 0.2055,
1847
+ "step": 3050
1848
+ },
1849
+ {
1850
+ "epoch": 1.96,
1851
+ "learning_rate": 1.927910837087977e-05,
1852
+ "loss": 0.2495,
1853
+ "step": 3060
1854
+ },
1855
+ {
1856
+ "epoch": 1.97,
1857
+ "learning_rate": 1.916054066872184e-05,
1858
+ "loss": 0.1625,
1859
+ "step": 3070
1860
+ },
1861
+ {
1862
+ "epoch": 1.97,
1863
+ "learning_rate": 1.9041972966563907e-05,
1864
+ "loss": 0.2483,
1865
+ "step": 3080
1866
+ },
1867
+ {
1868
+ "epoch": 1.98,
1869
+ "learning_rate": 1.8923405264405976e-05,
1870
+ "loss": 0.2782,
1871
+ "step": 3090
1872
+ },
1873
+ {
1874
+ "epoch": 1.98,
1875
+ "learning_rate": 1.8804837562248044e-05,
1876
+ "loss": 0.2128,
1877
+ "step": 3100
1878
+ },
1879
+ {
1880
+ "epoch": 1.99,
1881
+ "learning_rate": 1.8686269860090112e-05,
1882
+ "loss": 0.1631,
1883
+ "step": 3110
1884
+ },
1885
+ {
1886
+ "epoch": 2.0,
1887
+ "learning_rate": 1.856770215793218e-05,
1888
+ "loss": 0.2717,
1889
+ "step": 3120
1890
+ },
1891
+ {
1892
+ "epoch": 2.0,
1893
+ "eval_accuracy": 0.9867,
1894
+ "eval_loss": 0.06189415976405144,
1895
+ "eval_runtime": 794.0573,
1896
+ "eval_samples_per_second": 12.594,
1897
+ "eval_steps_per_second": 1.574,
1898
+ "step": 3124
1899
+ },
1900
+ {
1901
+ "epoch": 2.0,
1902
+ "learning_rate": 1.844913445577425e-05,
1903
+ "loss": 0.2339,
1904
+ "step": 3130
1905
+ },
1906
+ {
1907
+ "epoch": 2.01,
1908
+ "learning_rate": 1.8330566753616317e-05,
1909
+ "loss": 0.2318,
1910
+ "step": 3140
1911
+ },
1912
+ {
1913
+ "epoch": 2.02,
1914
+ "learning_rate": 1.8211999051458385e-05,
1915
+ "loss": 0.2117,
1916
+ "step": 3150
1917
+ },
1918
+ {
1919
+ "epoch": 2.02,
1920
+ "learning_rate": 1.8093431349300453e-05,
1921
+ "loss": 0.2235,
1922
+ "step": 3160
1923
+ },
1924
+ {
1925
+ "epoch": 2.03,
1926
+ "learning_rate": 1.797486364714252e-05,
1927
+ "loss": 0.2447,
1928
+ "step": 3170
1929
+ },
1930
+ {
1931
+ "epoch": 2.04,
1932
+ "learning_rate": 1.7856295944984586e-05,
1933
+ "loss": 0.2343,
1934
+ "step": 3180
1935
+ },
1936
+ {
1937
+ "epoch": 2.04,
1938
+ "learning_rate": 1.7737728242826654e-05,
1939
+ "loss": 0.2205,
1940
+ "step": 3190
1941
+ },
1942
+ {
1943
+ "epoch": 2.05,
1944
+ "learning_rate": 1.7619160540668723e-05,
1945
+ "loss": 0.245,
1946
+ "step": 3200
1947
+ },
1948
+ {
1949
+ "epoch": 2.06,
1950
+ "learning_rate": 1.750059283851079e-05,
1951
+ "loss": 0.1824,
1952
+ "step": 3210
1953
+ },
1954
+ {
1955
+ "epoch": 2.06,
1956
+ "learning_rate": 1.738202513635286e-05,
1957
+ "loss": 0.1872,
1958
+ "step": 3220
1959
+ },
1960
+ {
1961
+ "epoch": 2.07,
1962
+ "learning_rate": 1.7263457434194927e-05,
1963
+ "loss": 0.201,
1964
+ "step": 3230
1965
+ },
1966
+ {
1967
+ "epoch": 2.07,
1968
+ "learning_rate": 1.7144889732036995e-05,
1969
+ "loss": 0.1929,
1970
+ "step": 3240
1971
+ },
1972
+ {
1973
+ "epoch": 2.08,
1974
+ "learning_rate": 1.702632202987906e-05,
1975
+ "loss": 0.1707,
1976
+ "step": 3250
1977
+ },
1978
+ {
1979
+ "epoch": 2.09,
1980
+ "learning_rate": 1.690775432772113e-05,
1981
+ "loss": 0.26,
1982
+ "step": 3260
1983
+ },
1984
+ {
1985
+ "epoch": 2.09,
1986
+ "learning_rate": 1.6789186625563197e-05,
1987
+ "loss": 0.2135,
1988
+ "step": 3270
1989
+ },
1990
+ {
1991
+ "epoch": 2.1,
1992
+ "learning_rate": 1.6670618923405265e-05,
1993
+ "loss": 0.2477,
1994
+ "step": 3280
1995
+ },
1996
+ {
1997
+ "epoch": 2.11,
1998
+ "learning_rate": 1.6552051221247333e-05,
1999
+ "loss": 0.3029,
2000
+ "step": 3290
2001
+ },
2002
+ {
2003
+ "epoch": 2.11,
2004
+ "learning_rate": 1.64334835190894e-05,
2005
+ "loss": 0.2445,
2006
+ "step": 3300
2007
+ },
2008
+ {
2009
+ "epoch": 2.12,
2010
+ "learning_rate": 1.631491581693147e-05,
2011
+ "loss": 0.2268,
2012
+ "step": 3310
2013
+ },
2014
+ {
2015
+ "epoch": 2.13,
2016
+ "learning_rate": 1.6196348114773534e-05,
2017
+ "loss": 0.3133,
2018
+ "step": 3320
2019
+ },
2020
+ {
2021
+ "epoch": 2.13,
2022
+ "learning_rate": 1.6077780412615603e-05,
2023
+ "loss": 0.2255,
2024
+ "step": 3330
2025
+ },
2026
+ {
2027
+ "epoch": 2.14,
2028
+ "learning_rate": 1.595921271045767e-05,
2029
+ "loss": 0.2406,
2030
+ "step": 3340
2031
+ },
2032
+ {
2033
+ "epoch": 2.14,
2034
+ "learning_rate": 1.584064500829974e-05,
2035
+ "loss": 0.1899,
2036
+ "step": 3350
2037
+ },
2038
+ {
2039
+ "epoch": 2.15,
2040
+ "learning_rate": 1.5722077306141807e-05,
2041
+ "loss": 0.2539,
2042
+ "step": 3360
2043
+ },
2044
+ {
2045
+ "epoch": 2.16,
2046
+ "learning_rate": 1.5603509603983875e-05,
2047
+ "loss": 0.2009,
2048
+ "step": 3370
2049
+ },
2050
+ {
2051
+ "epoch": 2.16,
2052
+ "learning_rate": 1.5484941901825944e-05,
2053
+ "loss": 0.2638,
2054
+ "step": 3380
2055
+ },
2056
+ {
2057
+ "epoch": 2.17,
2058
+ "learning_rate": 1.536637419966801e-05,
2059
+ "loss": 0.1588,
2060
+ "step": 3390
2061
+ },
2062
+ {
2063
+ "epoch": 2.18,
2064
+ "learning_rate": 1.5247806497510078e-05,
2065
+ "loss": 0.2059,
2066
+ "step": 3400
2067
+ },
2068
+ {
2069
+ "epoch": 2.18,
2070
+ "learning_rate": 1.5129238795352146e-05,
2071
+ "loss": 0.1815,
2072
+ "step": 3410
2073
+ },
2074
+ {
2075
+ "epoch": 2.19,
2076
+ "learning_rate": 1.5010671093194215e-05,
2077
+ "loss": 0.2102,
2078
+ "step": 3420
2079
+ },
2080
+ {
2081
+ "epoch": 2.2,
2082
+ "learning_rate": 1.4892103391036283e-05,
2083
+ "loss": 0.3203,
2084
+ "step": 3430
2085
+ },
2086
+ {
2087
+ "epoch": 2.2,
2088
+ "learning_rate": 1.4773535688878351e-05,
2089
+ "loss": 0.17,
2090
+ "step": 3440
2091
+ },
2092
+ {
2093
+ "epoch": 2.21,
2094
+ "learning_rate": 1.465496798672042e-05,
2095
+ "loss": 0.2176,
2096
+ "step": 3450
2097
+ },
2098
+ {
2099
+ "epoch": 2.22,
2100
+ "learning_rate": 1.4536400284562484e-05,
2101
+ "loss": 0.1528,
2102
+ "step": 3460
2103
+ },
2104
+ {
2105
+ "epoch": 2.22,
2106
+ "learning_rate": 1.4417832582404552e-05,
2107
+ "loss": 0.1842,
2108
+ "step": 3470
2109
+ },
2110
+ {
2111
+ "epoch": 2.23,
2112
+ "learning_rate": 1.429926488024662e-05,
2113
+ "loss": 0.1791,
2114
+ "step": 3480
2115
+ },
2116
+ {
2117
+ "epoch": 2.23,
2118
+ "learning_rate": 1.4180697178088689e-05,
2119
+ "loss": 0.2646,
2120
+ "step": 3490
2121
+ },
2122
+ {
2123
+ "epoch": 2.24,
2124
+ "learning_rate": 1.4062129475930757e-05,
2125
+ "loss": 0.3129,
2126
+ "step": 3500
2127
+ },
2128
+ {
2129
+ "epoch": 2.25,
2130
+ "learning_rate": 1.3943561773772825e-05,
2131
+ "loss": 0.1282,
2132
+ "step": 3510
2133
+ },
2134
+ {
2135
+ "epoch": 2.25,
2136
+ "learning_rate": 1.3824994071614893e-05,
2137
+ "loss": 0.1876,
2138
+ "step": 3520
2139
+ },
2140
+ {
2141
+ "epoch": 2.26,
2142
+ "learning_rate": 1.370642636945696e-05,
2143
+ "loss": 0.2883,
2144
+ "step": 3530
2145
+ },
2146
+ {
2147
+ "epoch": 2.27,
2148
+ "learning_rate": 1.3587858667299028e-05,
2149
+ "loss": 0.2572,
2150
+ "step": 3540
2151
+ },
2152
+ {
2153
+ "epoch": 2.27,
2154
+ "learning_rate": 1.3469290965141096e-05,
2155
+ "loss": 0.2848,
2156
+ "step": 3550
2157
+ },
2158
+ {
2159
+ "epoch": 2.28,
2160
+ "learning_rate": 1.3350723262983165e-05,
2161
+ "loss": 0.1814,
2162
+ "step": 3560
2163
+ },
2164
+ {
2165
+ "epoch": 2.29,
2166
+ "learning_rate": 1.3232155560825233e-05,
2167
+ "loss": 0.1811,
2168
+ "step": 3570
2169
+ },
2170
+ {
2171
+ "epoch": 2.29,
2172
+ "learning_rate": 1.3113587858667301e-05,
2173
+ "loss": 0.1992,
2174
+ "step": 3580
2175
+ },
2176
+ {
2177
+ "epoch": 2.3,
2178
+ "learning_rate": 1.2995020156509369e-05,
2179
+ "loss": 0.2808,
2180
+ "step": 3590
2181
+ },
2182
+ {
2183
+ "epoch": 2.3,
2184
+ "learning_rate": 1.2876452454351434e-05,
2185
+ "loss": 0.2771,
2186
+ "step": 3600
2187
+ },
2188
+ {
2189
+ "epoch": 2.31,
2190
+ "learning_rate": 1.2757884752193502e-05,
2191
+ "loss": 0.1892,
2192
+ "step": 3610
2193
+ },
2194
+ {
2195
+ "epoch": 2.32,
2196
+ "learning_rate": 1.263931705003557e-05,
2197
+ "loss": 0.1927,
2198
+ "step": 3620
2199
+ },
2200
+ {
2201
+ "epoch": 2.32,
2202
+ "learning_rate": 1.2520749347877639e-05,
2203
+ "loss": 0.2016,
2204
+ "step": 3630
2205
+ },
2206
+ {
2207
+ "epoch": 2.33,
2208
+ "learning_rate": 1.2402181645719707e-05,
2209
+ "loss": 0.2299,
2210
+ "step": 3640
2211
+ },
2212
+ {
2213
+ "epoch": 2.34,
2214
+ "learning_rate": 1.2283613943561773e-05,
2215
+ "loss": 0.1871,
2216
+ "step": 3650
2217
+ },
2218
+ {
2219
+ "epoch": 2.34,
2220
+ "learning_rate": 1.2165046241403842e-05,
2221
+ "loss": 0.1911,
2222
+ "step": 3660
2223
+ },
2224
+ {
2225
+ "epoch": 2.35,
2226
+ "learning_rate": 1.204647853924591e-05,
2227
+ "loss": 0.2504,
2228
+ "step": 3670
2229
+ },
2230
+ {
2231
+ "epoch": 2.36,
2232
+ "learning_rate": 1.1927910837087978e-05,
2233
+ "loss": 0.1751,
2234
+ "step": 3680
2235
+ },
2236
+ {
2237
+ "epoch": 2.36,
2238
+ "learning_rate": 1.1809343134930046e-05,
2239
+ "loss": 0.1574,
2240
+ "step": 3690
2241
+ },
2242
+ {
2243
+ "epoch": 2.37,
2244
+ "learning_rate": 1.1690775432772114e-05,
2245
+ "loss": 0.1423,
2246
+ "step": 3700
2247
+ },
2248
+ {
2249
+ "epoch": 2.38,
2250
+ "learning_rate": 1.1572207730614183e-05,
2251
+ "loss": 0.1892,
2252
+ "step": 3710
2253
+ },
2254
+ {
2255
+ "epoch": 2.38,
2256
+ "learning_rate": 1.1453640028456249e-05,
2257
+ "loss": 0.2249,
2258
+ "step": 3720
2259
+ },
2260
+ {
2261
+ "epoch": 2.39,
2262
+ "learning_rate": 1.1335072326298317e-05,
2263
+ "loss": 0.2122,
2264
+ "step": 3730
2265
+ },
2266
+ {
2267
+ "epoch": 2.39,
2268
+ "learning_rate": 1.1216504624140385e-05,
2269
+ "loss": 0.2073,
2270
+ "step": 3740
2271
+ },
2272
+ {
2273
+ "epoch": 2.4,
2274
+ "learning_rate": 1.1097936921982452e-05,
2275
+ "loss": 0.2123,
2276
+ "step": 3750
2277
+ },
2278
+ {
2279
+ "epoch": 2.41,
2280
+ "learning_rate": 1.097936921982452e-05,
2281
+ "loss": 0.1781,
2282
+ "step": 3760
2283
+ },
2284
+ {
2285
+ "epoch": 2.41,
2286
+ "learning_rate": 1.0860801517666588e-05,
2287
+ "loss": 0.224,
2288
+ "step": 3770
2289
+ },
2290
+ {
2291
+ "epoch": 2.42,
2292
+ "learning_rate": 1.0742233815508657e-05,
2293
+ "loss": 0.188,
2294
+ "step": 3780
2295
+ },
2296
+ {
2297
+ "epoch": 2.43,
2298
+ "learning_rate": 1.0623666113350723e-05,
2299
+ "loss": 0.2385,
2300
+ "step": 3790
2301
+ },
2302
+ {
2303
+ "epoch": 2.43,
2304
+ "learning_rate": 1.0505098411192791e-05,
2305
+ "loss": 0.2363,
2306
+ "step": 3800
2307
+ },
2308
+ {
2309
+ "epoch": 2.44,
2310
+ "learning_rate": 1.038653070903486e-05,
2311
+ "loss": 0.1909,
2312
+ "step": 3810
2313
+ },
2314
+ {
2315
+ "epoch": 2.45,
2316
+ "learning_rate": 1.0267963006876926e-05,
2317
+ "loss": 0.2612,
2318
+ "step": 3820
2319
+ },
2320
+ {
2321
+ "epoch": 2.45,
2322
+ "learning_rate": 1.0149395304718994e-05,
2323
+ "loss": 0.2269,
2324
+ "step": 3830
2325
+ },
2326
+ {
2327
+ "epoch": 2.46,
2328
+ "learning_rate": 1.0030827602561062e-05,
2329
+ "loss": 0.2408,
2330
+ "step": 3840
2331
+ },
2332
+ {
2333
+ "epoch": 2.46,
2334
+ "learning_rate": 9.91225990040313e-06,
2335
+ "loss": 0.229,
2336
+ "step": 3850
2337
+ },
2338
+ {
2339
+ "epoch": 2.47,
2340
+ "learning_rate": 9.793692198245199e-06,
2341
+ "loss": 0.2482,
2342
+ "step": 3860
2343
+ },
2344
+ {
2345
+ "epoch": 2.48,
2346
+ "learning_rate": 9.675124496087267e-06,
2347
+ "loss": 0.1652,
2348
+ "step": 3870
2349
+ },
2350
+ {
2351
+ "epoch": 2.48,
2352
+ "learning_rate": 9.556556793929335e-06,
2353
+ "loss": 0.1671,
2354
+ "step": 3880
2355
+ },
2356
+ {
2357
+ "epoch": 2.49,
2358
+ "learning_rate": 9.437989091771402e-06,
2359
+ "loss": 0.1846,
2360
+ "step": 3890
2361
+ },
2362
+ {
2363
+ "epoch": 2.5,
2364
+ "learning_rate": 9.31942138961347e-06,
2365
+ "loss": 0.1745,
2366
+ "step": 3900
2367
+ },
2368
+ {
2369
+ "epoch": 2.5,
2370
+ "learning_rate": 9.200853687455538e-06,
2371
+ "loss": 0.2112,
2372
+ "step": 3910
2373
+ },
2374
+ {
2375
+ "epoch": 2.51,
2376
+ "learning_rate": 9.082285985297605e-06,
2377
+ "loss": 0.2274,
2378
+ "step": 3920
2379
+ },
2380
+ {
2381
+ "epoch": 2.52,
2382
+ "learning_rate": 8.963718283139673e-06,
2383
+ "loss": 0.2371,
2384
+ "step": 3930
2385
+ },
2386
+ {
2387
+ "epoch": 2.52,
2388
+ "learning_rate": 8.845150580981741e-06,
2389
+ "loss": 0.2615,
2390
+ "step": 3940
2391
+ },
2392
+ {
2393
+ "epoch": 2.53,
2394
+ "learning_rate": 8.72658287882381e-06,
2395
+ "loss": 0.2454,
2396
+ "step": 3950
2397
+ },
2398
+ {
2399
+ "epoch": 2.54,
2400
+ "learning_rate": 8.608015176665876e-06,
2401
+ "loss": 0.1898,
2402
+ "step": 3960
2403
+ },
2404
+ {
2405
+ "epoch": 2.54,
2406
+ "learning_rate": 8.489447474507944e-06,
2407
+ "loss": 0.207,
2408
+ "step": 3970
2409
+ },
2410
+ {
2411
+ "epoch": 2.55,
2412
+ "learning_rate": 8.370879772350012e-06,
2413
+ "loss": 0.2036,
2414
+ "step": 3980
2415
+ },
2416
+ {
2417
+ "epoch": 2.55,
2418
+ "learning_rate": 8.252312070192079e-06,
2419
+ "loss": 0.2158,
2420
+ "step": 3990
2421
+ },
2422
+ {
2423
+ "epoch": 2.56,
2424
+ "learning_rate": 8.133744368034147e-06,
2425
+ "loss": 0.1839,
2426
+ "step": 4000
2427
+ },
2428
+ {
2429
+ "epoch": 2.57,
2430
+ "learning_rate": 8.015176665876215e-06,
2431
+ "loss": 0.2519,
2432
+ "step": 4010
2433
+ },
2434
+ {
2435
+ "epoch": 2.57,
2436
+ "learning_rate": 7.896608963718283e-06,
2437
+ "loss": 0.1437,
2438
+ "step": 4020
2439
+ },
2440
+ {
2441
+ "epoch": 2.58,
2442
+ "learning_rate": 7.778041261560352e-06,
2443
+ "loss": 0.2306,
2444
+ "step": 4030
2445
+ },
2446
+ {
2447
+ "epoch": 2.59,
2448
+ "learning_rate": 7.65947355940242e-06,
2449
+ "loss": 0.2353,
2450
+ "step": 4040
2451
+ },
2452
+ {
2453
+ "epoch": 2.59,
2454
+ "learning_rate": 7.540905857244487e-06,
2455
+ "loss": 0.2149,
2456
+ "step": 4050
2457
+ },
2458
+ {
2459
+ "epoch": 2.6,
2460
+ "learning_rate": 7.422338155086555e-06,
2461
+ "loss": 0.2845,
2462
+ "step": 4060
2463
+ },
2464
+ {
2465
+ "epoch": 2.61,
2466
+ "learning_rate": 7.303770452928623e-06,
2467
+ "loss": 0.1821,
2468
+ "step": 4070
2469
+ },
2470
+ {
2471
+ "epoch": 2.61,
2472
+ "learning_rate": 7.185202750770691e-06,
2473
+ "loss": 0.2109,
2474
+ "step": 4080
2475
+ },
2476
+ {
2477
+ "epoch": 2.62,
2478
+ "learning_rate": 7.066635048612759e-06,
2479
+ "loss": 0.2036,
2480
+ "step": 4090
2481
+ },
2482
+ {
2483
+ "epoch": 2.62,
2484
+ "learning_rate": 6.948067346454826e-06,
2485
+ "loss": 0.2048,
2486
+ "step": 4100
2487
+ },
2488
+ {
2489
+ "epoch": 2.63,
2490
+ "learning_rate": 6.829499644296894e-06,
2491
+ "loss": 0.2418,
2492
+ "step": 4110
2493
+ },
2494
+ {
2495
+ "epoch": 2.64,
2496
+ "learning_rate": 6.710931942138962e-06,
2497
+ "loss": 0.1733,
2498
+ "step": 4120
2499
+ },
2500
+ {
2501
+ "epoch": 2.64,
2502
+ "learning_rate": 6.592364239981029e-06,
2503
+ "loss": 0.1938,
2504
+ "step": 4130
2505
+ },
2506
+ {
2507
+ "epoch": 2.65,
2508
+ "learning_rate": 6.473796537823097e-06,
2509
+ "loss": 0.1931,
2510
+ "step": 4140
2511
+ },
2512
+ {
2513
+ "epoch": 2.66,
2514
+ "learning_rate": 6.355228835665165e-06,
2515
+ "loss": 0.1563,
2516
+ "step": 4150
2517
+ },
2518
+ {
2519
+ "epoch": 2.66,
2520
+ "learning_rate": 6.236661133507233e-06,
2521
+ "loss": 0.2386,
2522
+ "step": 4160
2523
+ },
2524
+ {
2525
+ "epoch": 2.67,
2526
+ "learning_rate": 6.118093431349301e-06,
2527
+ "loss": 0.1682,
2528
+ "step": 4170
2529
+ },
2530
+ {
2531
+ "epoch": 2.68,
2532
+ "learning_rate": 5.999525729191369e-06,
2533
+ "loss": 0.1493,
2534
+ "step": 4180
2535
+ },
2536
+ {
2537
+ "epoch": 2.68,
2538
+ "learning_rate": 5.880958027033436e-06,
2539
+ "loss": 0.1525,
2540
+ "step": 4190
2541
+ },
2542
+ {
2543
+ "epoch": 2.69,
2544
+ "learning_rate": 5.762390324875504e-06,
2545
+ "loss": 0.1471,
2546
+ "step": 4200
2547
+ },
2548
+ {
2549
+ "epoch": 2.7,
2550
+ "learning_rate": 5.643822622717572e-06,
2551
+ "loss": 0.1232,
2552
+ "step": 4210
2553
+ },
2554
+ {
2555
+ "epoch": 2.7,
2556
+ "learning_rate": 5.525254920559639e-06,
2557
+ "loss": 0.1545,
2558
+ "step": 4220
2559
+ },
2560
+ {
2561
+ "epoch": 2.71,
2562
+ "learning_rate": 5.406687218401707e-06,
2563
+ "loss": 0.1835,
2564
+ "step": 4230
2565
+ },
2566
+ {
2567
+ "epoch": 2.71,
2568
+ "learning_rate": 5.2881195162437755e-06,
2569
+ "loss": 0.2369,
2570
+ "step": 4240
2571
+ },
2572
+ {
2573
+ "epoch": 2.72,
2574
+ "learning_rate": 5.169551814085844e-06,
2575
+ "loss": 0.1412,
2576
+ "step": 4250
2577
+ },
2578
+ {
2579
+ "epoch": 2.73,
2580
+ "learning_rate": 5.050984111927911e-06,
2581
+ "loss": 0.1897,
2582
+ "step": 4260
2583
+ },
2584
+ {
2585
+ "epoch": 2.73,
2586
+ "learning_rate": 4.932416409769979e-06,
2587
+ "loss": 0.1865,
2588
+ "step": 4270
2589
+ },
2590
+ {
2591
+ "epoch": 2.74,
2592
+ "learning_rate": 4.813848707612047e-06,
2593
+ "loss": 0.1815,
2594
+ "step": 4280
2595
+ },
2596
+ {
2597
+ "epoch": 2.75,
2598
+ "learning_rate": 4.695281005454114e-06,
2599
+ "loss": 0.2102,
2600
+ "step": 4290
2601
+ },
2602
+ {
2603
+ "epoch": 2.75,
2604
+ "learning_rate": 4.576713303296182e-06,
2605
+ "loss": 0.1086,
2606
+ "step": 4300
2607
+ },
2608
+ {
2609
+ "epoch": 2.76,
2610
+ "learning_rate": 4.45814560113825e-06,
2611
+ "loss": 0.117,
2612
+ "step": 4310
2613
+ },
2614
+ {
2615
+ "epoch": 2.77,
2616
+ "learning_rate": 4.339577898980318e-06,
2617
+ "loss": 0.2686,
2618
+ "step": 4320
2619
+ },
2620
+ {
2621
+ "epoch": 2.77,
2622
+ "learning_rate": 4.221010196822386e-06,
2623
+ "loss": 0.2048,
2624
+ "step": 4330
2625
+ },
2626
+ {
2627
+ "epoch": 2.78,
2628
+ "learning_rate": 4.102442494664454e-06,
2629
+ "loss": 0.2178,
2630
+ "step": 4340
2631
+ },
2632
+ {
2633
+ "epoch": 2.78,
2634
+ "learning_rate": 3.983874792506522e-06,
2635
+ "loss": 0.2273,
2636
+ "step": 4350
2637
+ },
2638
+ {
2639
+ "epoch": 2.79,
2640
+ "learning_rate": 3.865307090348589e-06,
2641
+ "loss": 0.1873,
2642
+ "step": 4360
2643
+ },
2644
+ {
2645
+ "epoch": 2.8,
2646
+ "learning_rate": 3.746739388190657e-06,
2647
+ "loss": 0.2187,
2648
+ "step": 4370
2649
+ },
2650
+ {
2651
+ "epoch": 2.8,
2652
+ "learning_rate": 3.6281716860327245e-06,
2653
+ "loss": 0.2283,
2654
+ "step": 4380
2655
+ },
2656
+ {
2657
+ "epoch": 2.81,
2658
+ "learning_rate": 3.5096039838747927e-06,
2659
+ "loss": 0.2117,
2660
+ "step": 4390
2661
+ },
2662
+ {
2663
+ "epoch": 2.82,
2664
+ "learning_rate": 3.3910362817168605e-06,
2665
+ "loss": 0.1951,
2666
+ "step": 4400
2667
+ },
2668
+ {
2669
+ "epoch": 2.82,
2670
+ "learning_rate": 3.272468579558928e-06,
2671
+ "loss": 0.1994,
2672
+ "step": 4410
2673
+ },
2674
+ {
2675
+ "epoch": 2.83,
2676
+ "learning_rate": 3.153900877400996e-06,
2677
+ "loss": 0.1735,
2678
+ "step": 4420
2679
+ },
2680
+ {
2681
+ "epoch": 2.84,
2682
+ "learning_rate": 3.0353331752430643e-06,
2683
+ "loss": 0.2796,
2684
+ "step": 4430
2685
+ },
2686
+ {
2687
+ "epoch": 2.84,
2688
+ "learning_rate": 2.9167654730851317e-06,
2689
+ "loss": 0.1308,
2690
+ "step": 4440
2691
+ },
2692
+ {
2693
+ "epoch": 2.85,
2694
+ "learning_rate": 2.7981977709271994e-06,
2695
+ "loss": 0.2232,
2696
+ "step": 4450
2697
+ },
2698
+ {
2699
+ "epoch": 2.86,
2700
+ "learning_rate": 2.6796300687692672e-06,
2701
+ "loss": 0.2338,
2702
+ "step": 4460
2703
+ },
2704
+ {
2705
+ "epoch": 2.86,
2706
+ "learning_rate": 2.5610623666113354e-06,
2707
+ "loss": 0.2609,
2708
+ "step": 4470
2709
+ },
2710
+ {
2711
+ "epoch": 2.87,
2712
+ "learning_rate": 2.4424946644534032e-06,
2713
+ "loss": 0.1971,
2714
+ "step": 4480
2715
+ },
2716
+ {
2717
+ "epoch": 2.87,
2718
+ "learning_rate": 2.323926962295471e-06,
2719
+ "loss": 0.1864,
2720
+ "step": 4490
2721
+ },
2722
+ {
2723
+ "epoch": 2.88,
2724
+ "learning_rate": 2.2053592601375384e-06,
2725
+ "loss": 0.1813,
2726
+ "step": 4500
2727
+ },
2728
+ {
2729
+ "epoch": 2.89,
2730
+ "learning_rate": 2.0867915579796066e-06,
2731
+ "loss": 0.2013,
2732
+ "step": 4510
2733
+ },
2734
+ {
2735
+ "epoch": 2.89,
2736
+ "learning_rate": 1.9682238558216744e-06,
2737
+ "loss": 0.1666,
2738
+ "step": 4520
2739
+ },
2740
+ {
2741
+ "epoch": 2.9,
2742
+ "learning_rate": 1.8496561536637421e-06,
2743
+ "loss": 0.2143,
2744
+ "step": 4530
2745
+ },
2746
+ {
2747
+ "epoch": 2.91,
2748
+ "learning_rate": 1.73108845150581e-06,
2749
+ "loss": 0.1793,
2750
+ "step": 4540
2751
+ },
2752
+ {
2753
+ "epoch": 2.91,
2754
+ "learning_rate": 1.612520749347878e-06,
2755
+ "loss": 0.1433,
2756
+ "step": 4550
2757
+ },
2758
+ {
2759
+ "epoch": 2.92,
2760
+ "learning_rate": 1.4939530471899455e-06,
2761
+ "loss": 0.1603,
2762
+ "step": 4560
2763
+ },
2764
+ {
2765
+ "epoch": 2.93,
2766
+ "learning_rate": 1.3753853450320135e-06,
2767
+ "loss": 0.136,
2768
+ "step": 4570
2769
+ },
2770
+ {
2771
+ "epoch": 2.93,
2772
+ "learning_rate": 1.256817642874081e-06,
2773
+ "loss": 0.1805,
2774
+ "step": 4580
2775
+ },
2776
+ {
2777
+ "epoch": 2.94,
2778
+ "learning_rate": 1.138249940716149e-06,
2779
+ "loss": 0.2584,
2780
+ "step": 4590
2781
+ },
2782
+ {
2783
+ "epoch": 2.94,
2784
+ "learning_rate": 1.0196822385582168e-06,
2785
+ "loss": 0.1895,
2786
+ "step": 4600
2787
+ },
2788
+ {
2789
+ "epoch": 2.95,
2790
+ "learning_rate": 9.011145364002845e-07,
2791
+ "loss": 0.2353,
2792
+ "step": 4610
2793
+ },
2794
+ {
2795
+ "epoch": 2.96,
2796
+ "learning_rate": 7.825468342423524e-07,
2797
+ "loss": 0.1917,
2798
+ "step": 4620
2799
+ },
2800
+ {
2801
+ "epoch": 2.96,
2802
+ "learning_rate": 6.639791320844202e-07,
2803
+ "loss": 0.1414,
2804
+ "step": 4630
2805
+ },
2806
+ {
2807
+ "epoch": 2.97,
2808
+ "learning_rate": 5.454114299264881e-07,
2809
+ "loss": 0.2053,
2810
+ "step": 4640
2811
+ },
2812
+ {
2813
+ "epoch": 2.98,
2814
+ "learning_rate": 4.2684372776855587e-07,
2815
+ "loss": 0.1612,
2816
+ "step": 4650
2817
+ },
2818
+ {
2819
+ "epoch": 2.98,
2820
+ "learning_rate": 3.082760256106237e-07,
2821
+ "loss": 0.2274,
2822
+ "step": 4660
2823
+ },
2824
+ {
2825
+ "epoch": 2.99,
2826
+ "learning_rate": 1.8970832345269152e-07,
2827
+ "loss": 0.2354,
2828
+ "step": 4670
2829
+ },
2830
+ {
2831
+ "epoch": 3.0,
2832
+ "learning_rate": 7.114062129475932e-08,
2833
+ "loss": 0.1964,
2834
+ "step": 4680
2835
+ },
2836
+ {
2837
+ "epoch": 3.0,
2838
+ "eval_accuracy": 0.9875,
2839
+ "eval_loss": 0.0503208190202713,
2840
+ "eval_runtime": 861.4693,
2841
+ "eval_samples_per_second": 11.608,
2842
+ "eval_steps_per_second": 1.451,
2843
+ "step": 4686
2844
+ },
2845
+ {
2846
+ "epoch": 3.0,
2847
+ "step": 4686,
2848
+ "total_flos": 1.1623391924884734e+19,
2849
+ "train_loss": 0.39782794105462216,
2850
+ "train_runtime": 35181.3105,
2851
+ "train_samples_per_second": 4.264,
2852
+ "train_steps_per_second": 0.133
2853
+ }
2854
+ ],
2855
+ "max_steps": 4686,
2856
+ "num_train_epochs": 3,
2857
+ "total_flos": 1.1623391924884734e+19,
2858
+ "trial_name": null,
2859
+ "trial_params": null
2860
+ }