igorktech commited on
Commit
89d1638
·
1 Parent(s): 0480361

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +15 -0
  2. eval_results.json +10 -0
  3. train_results.json +8 -0
  4. trainer_state.json +3043 -0
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.1,
3
+ "eval_accuracy": 0.5611792082891779,
4
+ "eval_loss": 2.292884111404419,
5
+ "eval_runtime": 6152.2021,
6
+ "eval_samples": 217361,
7
+ "eval_samples_per_second": 35.331,
8
+ "eval_steps_per_second": 2.208,
9
+ "perplexity": 9.903459211422307,
10
+ "train_loss": 3.1285950134277343,
11
+ "train_runtime": 151231.41,
12
+ "train_samples": 4129855,
13
+ "train_samples_per_second": 84.639,
14
+ "train_steps_per_second": 0.331
15
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.1,
3
+ "eval_accuracy": 0.5611792082891779,
4
+ "eval_loss": 2.292884111404419,
5
+ "eval_runtime": 6152.2021,
6
+ "eval_samples": 217361,
7
+ "eval_samples_per_second": 35.331,
8
+ "eval_steps_per_second": 2.208,
9
+ "perplexity": 9.903459211422307
10
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.1,
3
+ "train_loss": 3.1285950134277343,
4
+ "train_runtime": 151231.41,
5
+ "train_samples": 4129855,
6
+ "train_samples_per_second": 84.639,
7
+ "train_steps_per_second": 0.331
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0993816733561657,
5
+ "global_step": 50000,
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": 4e-06,
13
+ "loss": 10.285,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 8e-06,
19
+ "loss": 8.5629,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 1.2e-05,
25
+ "loss": 7.7941,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 1.6e-05,
31
+ "loss": 7.2214,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 2e-05,
37
+ "loss": 6.7445,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.04,
42
+ "learning_rate": 2.4e-05,
43
+ "loss": 6.5017,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "learning_rate": 2.8e-05,
49
+ "loss": 6.3973,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 3.2e-05,
55
+ "loss": 6.3519,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.06,
60
+ "learning_rate": 3.6e-05,
61
+ "loss": 6.341,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.06,
66
+ "learning_rate": 4e-05,
67
+ "loss": 6.3197,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.07,
72
+ "learning_rate": 4.4e-05,
73
+ "loss": 6.3186,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.07,
78
+ "learning_rate": 4.8e-05,
79
+ "loss": 6.3114,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 0.08,
84
+ "learning_rate": 5.2e-05,
85
+ "loss": 6.3023,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 0.09,
90
+ "learning_rate": 5.6e-05,
91
+ "loss": 6.2749,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 0.09,
96
+ "learning_rate": 6e-05,
97
+ "loss": 6.2705,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 0.1,
102
+ "learning_rate": 6.4e-05,
103
+ "loss": 6.2538,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 0.11,
108
+ "learning_rate": 6.800000000000001e-05,
109
+ "loss": 6.2464,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 0.11,
114
+ "learning_rate": 7.2e-05,
115
+ "loss": 6.2073,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 0.12,
120
+ "learning_rate": 7.6e-05,
121
+ "loss": 6.101,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 0.12,
126
+ "learning_rate": 8e-05,
127
+ "loss": 5.9768,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 0.13,
132
+ "learning_rate": 8.400000000000001e-05,
133
+ "loss": 5.7986,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 0.14,
138
+ "learning_rate": 8.8e-05,
139
+ "loss": 5.6088,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 0.14,
144
+ "learning_rate": 9.2e-05,
145
+ "loss": 5.4903,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 0.15,
150
+ "learning_rate": 9.6e-05,
151
+ "loss": 5.3685,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 0.15,
156
+ "learning_rate": 0.0001,
157
+ "loss": 5.282,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 0.16,
162
+ "learning_rate": 0.000104,
163
+ "loss": 5.1831,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 0.17,
168
+ "learning_rate": 0.000108,
169
+ "loss": 5.1007,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 0.17,
174
+ "learning_rate": 0.000112,
175
+ "loss": 5.0023,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 0.18,
180
+ "learning_rate": 0.00011600000000000001,
181
+ "loss": 4.9239,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 0.19,
186
+ "learning_rate": 0.00012,
187
+ "loss": 4.8625,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 0.19,
192
+ "learning_rate": 0.000124,
193
+ "loss": 4.7771,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 0.2,
198
+ "learning_rate": 0.000128,
199
+ "loss": 4.7051,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 0.2,
204
+ "learning_rate": 0.000132,
205
+ "loss": 4.6168,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 0.21,
210
+ "learning_rate": 0.00013600000000000003,
211
+ "loss": 4.5591,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 0.22,
216
+ "learning_rate": 0.00014000000000000001,
217
+ "loss": 4.5026,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 0.22,
222
+ "learning_rate": 0.000144,
223
+ "loss": 4.4251,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 0.23,
228
+ "learning_rate": 0.000148,
229
+ "loss": 4.3889,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 0.24,
234
+ "learning_rate": 0.000152,
235
+ "loss": 4.3326,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 0.24,
240
+ "learning_rate": 0.000156,
241
+ "loss": 4.288,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 0.25,
246
+ "learning_rate": 0.00016,
247
+ "loss": 4.2549,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 0.25,
252
+ "learning_rate": 0.000164,
253
+ "loss": 4.19,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 0.26,
258
+ "learning_rate": 0.00016800000000000002,
259
+ "loss": 4.1428,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 0.27,
264
+ "learning_rate": 0.00017199999999999998,
265
+ "loss": 4.1085,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 0.27,
270
+ "learning_rate": 0.000176,
271
+ "loss": 4.078,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 0.28,
276
+ "learning_rate": 0.00017999999999999998,
277
+ "loss": 4.0206,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 0.29,
282
+ "learning_rate": 0.000184,
283
+ "loss": 4.0012,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 0.29,
288
+ "learning_rate": 0.00018800000000000002,
289
+ "loss": 3.98,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 0.3,
294
+ "learning_rate": 0.000192,
295
+ "loss": 3.9526,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 0.3,
300
+ "learning_rate": 0.00019600000000000002,
301
+ "loss": 3.8986,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 0.31,
306
+ "learning_rate": 0.0002,
307
+ "loss": 3.8794,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 0.32,
312
+ "learning_rate": 0.000204,
313
+ "loss": 3.84,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 0.32,
318
+ "learning_rate": 0.000208,
319
+ "loss": 3.8412,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 0.33,
324
+ "learning_rate": 0.000212,
325
+ "loss": 3.8021,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 0.33,
330
+ "learning_rate": 0.000216,
331
+ "loss": 3.7633,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 0.34,
336
+ "learning_rate": 0.00022,
337
+ "loss": 3.7475,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 0.35,
342
+ "learning_rate": 0.000224,
343
+ "loss": 3.7366,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 0.35,
348
+ "learning_rate": 0.000228,
349
+ "loss": 3.7149,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 0.36,
354
+ "learning_rate": 0.00023200000000000003,
355
+ "loss": 3.6964,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 0.37,
360
+ "learning_rate": 0.000236,
361
+ "loss": 3.6868,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 0.37,
366
+ "learning_rate": 0.00024,
367
+ "loss": 3.6684,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 0.38,
372
+ "learning_rate": 0.000244,
373
+ "loss": 3.6452,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 0.38,
378
+ "learning_rate": 0.000248,
379
+ "loss": 3.6353,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 0.39,
384
+ "learning_rate": 0.000252,
385
+ "loss": 3.6115,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 0.4,
390
+ "learning_rate": 0.000256,
391
+ "loss": 3.6012,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 0.4,
396
+ "learning_rate": 0.00026000000000000003,
397
+ "loss": 3.5799,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 0.41,
402
+ "learning_rate": 0.000264,
403
+ "loss": 3.5517,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 0.42,
408
+ "learning_rate": 0.000268,
409
+ "loss": 3.5509,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 0.42,
414
+ "learning_rate": 0.00027200000000000005,
415
+ "loss": 3.539,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 0.43,
420
+ "learning_rate": 0.00027600000000000004,
421
+ "loss": 3.5227,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 0.43,
426
+ "learning_rate": 0.00028000000000000003,
427
+ "loss": 3.5066,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 0.44,
432
+ "learning_rate": 0.00028399999999999996,
433
+ "loss": 3.5071,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 0.45,
438
+ "learning_rate": 0.000288,
439
+ "loss": 3.5012,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 0.45,
444
+ "learning_rate": 0.000292,
445
+ "loss": 3.4881,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 0.46,
450
+ "learning_rate": 0.000296,
451
+ "loss": 3.483,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 0.46,
456
+ "learning_rate": 0.0003,
457
+ "loss": 3.4781,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 0.47,
462
+ "learning_rate": 0.000304,
463
+ "loss": 3.4433,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 0.48,
468
+ "learning_rate": 0.000308,
469
+ "loss": 3.4457,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 0.48,
474
+ "learning_rate": 0.000312,
475
+ "loss": 3.4491,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 0.49,
480
+ "learning_rate": 0.000316,
481
+ "loss": 3.4211,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 0.5,
486
+ "learning_rate": 0.00032,
487
+ "loss": 3.416,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 0.5,
492
+ "learning_rate": 0.000324,
493
+ "loss": 3.4149,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 0.51,
498
+ "learning_rate": 0.000328,
499
+ "loss": 3.3943,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 0.51,
504
+ "learning_rate": 0.00033200000000000005,
505
+ "loss": 3.3854,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 0.52,
510
+ "learning_rate": 0.00033600000000000004,
511
+ "loss": 3.389,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 0.53,
516
+ "learning_rate": 0.00034,
517
+ "loss": 3.3672,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 0.53,
522
+ "learning_rate": 0.00034399999999999996,
523
+ "loss": 3.3656,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 0.54,
528
+ "learning_rate": 0.000348,
529
+ "loss": 3.3347,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 0.55,
534
+ "learning_rate": 0.000352,
535
+ "loss": 3.3448,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 0.55,
540
+ "learning_rate": 0.000356,
541
+ "loss": 3.3228,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 0.56,
546
+ "learning_rate": 0.00035999999999999997,
547
+ "loss": 3.3337,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 0.56,
552
+ "learning_rate": 0.000364,
553
+ "loss": 3.3183,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 0.57,
558
+ "learning_rate": 0.000368,
559
+ "loss": 3.33,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 0.58,
564
+ "learning_rate": 0.000372,
565
+ "loss": 3.3125,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 0.58,
570
+ "learning_rate": 0.00037600000000000003,
571
+ "loss": 3.2962,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 0.59,
576
+ "learning_rate": 0.00038,
577
+ "loss": 3.2982,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 0.6,
582
+ "learning_rate": 0.000384,
583
+ "loss": 3.295,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 0.6,
588
+ "learning_rate": 0.000388,
589
+ "loss": 3.2839,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 0.61,
594
+ "learning_rate": 0.00039200000000000004,
595
+ "loss": 3.2859,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 0.61,
600
+ "learning_rate": 0.00039600000000000003,
601
+ "loss": 3.2746,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 0.62,
606
+ "learning_rate": 0.0004,
607
+ "loss": 3.2595,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 0.63,
612
+ "learning_rate": 0.000404,
613
+ "loss": 3.2802,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 0.63,
618
+ "learning_rate": 0.000408,
619
+ "loss": 3.2474,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 0.64,
624
+ "learning_rate": 0.000412,
625
+ "loss": 3.2535,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 0.64,
630
+ "learning_rate": 0.000416,
631
+ "loss": 3.2276,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 0.65,
636
+ "learning_rate": 0.00042,
637
+ "loss": 3.2365,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 0.66,
642
+ "learning_rate": 0.000424,
643
+ "loss": 3.2413,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 0.66,
648
+ "learning_rate": 0.000428,
649
+ "loss": 3.2175,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 0.67,
654
+ "learning_rate": 0.000432,
655
+ "loss": 3.2231,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 0.68,
660
+ "learning_rate": 0.000436,
661
+ "loss": 3.2403,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 0.68,
666
+ "learning_rate": 0.00044,
667
+ "loss": 3.2022,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 0.69,
672
+ "learning_rate": 0.000444,
673
+ "loss": 3.2017,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 0.69,
678
+ "learning_rate": 0.000448,
679
+ "loss": 3.2083,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 0.7,
684
+ "learning_rate": 0.00045200000000000004,
685
+ "loss": 3.1954,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 0.71,
690
+ "learning_rate": 0.000456,
691
+ "loss": 3.1911,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 0.71,
696
+ "learning_rate": 0.00046,
697
+ "loss": 3.1823,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 0.72,
702
+ "learning_rate": 0.00046400000000000006,
703
+ "loss": 3.1863,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 0.73,
708
+ "learning_rate": 0.00046800000000000005,
709
+ "loss": 3.1878,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 0.73,
714
+ "learning_rate": 0.000472,
715
+ "loss": 3.1734,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 0.74,
720
+ "learning_rate": 0.00047599999999999997,
721
+ "loss": 3.1782,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 0.74,
726
+ "learning_rate": 0.00048,
727
+ "loss": 3.1711,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 0.75,
732
+ "learning_rate": 0.000484,
733
+ "loss": 3.1747,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 0.76,
738
+ "learning_rate": 0.000488,
739
+ "loss": 3.1527,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 0.76,
744
+ "learning_rate": 0.000492,
745
+ "loss": 3.15,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 0.77,
750
+ "learning_rate": 0.000496,
751
+ "loss": 3.1416,
752
+ "step": 12400
753
+ },
754
+ {
755
+ "epoch": 0.77,
756
+ "learning_rate": 0.0005,
757
+ "loss": 3.1455,
758
+ "step": 12500
759
+ },
760
+ {
761
+ "epoch": 0.78,
762
+ "learning_rate": 0.000504,
763
+ "loss": 3.1441,
764
+ "step": 12600
765
+ },
766
+ {
767
+ "epoch": 0.79,
768
+ "learning_rate": 0.000508,
769
+ "loss": 3.1462,
770
+ "step": 12700
771
+ },
772
+ {
773
+ "epoch": 0.79,
774
+ "learning_rate": 0.000512,
775
+ "loss": 3.1531,
776
+ "step": 12800
777
+ },
778
+ {
779
+ "epoch": 0.8,
780
+ "learning_rate": 0.0005160000000000001,
781
+ "loss": 3.1394,
782
+ "step": 12900
783
+ },
784
+ {
785
+ "epoch": 0.81,
786
+ "learning_rate": 0.0005200000000000001,
787
+ "loss": 3.1304,
788
+ "step": 13000
789
+ },
790
+ {
791
+ "epoch": 0.81,
792
+ "learning_rate": 0.000524,
793
+ "loss": 3.1292,
794
+ "step": 13100
795
+ },
796
+ {
797
+ "epoch": 0.82,
798
+ "learning_rate": 0.000528,
799
+ "loss": 3.1334,
800
+ "step": 13200
801
+ },
802
+ {
803
+ "epoch": 0.82,
804
+ "learning_rate": 0.000532,
805
+ "loss": 3.1129,
806
+ "step": 13300
807
+ },
808
+ {
809
+ "epoch": 0.83,
810
+ "learning_rate": 0.000536,
811
+ "loss": 3.1171,
812
+ "step": 13400
813
+ },
814
+ {
815
+ "epoch": 0.84,
816
+ "learning_rate": 0.00054,
817
+ "loss": 3.1147,
818
+ "step": 13500
819
+ },
820
+ {
821
+ "epoch": 0.84,
822
+ "learning_rate": 0.0005440000000000001,
823
+ "loss": 3.1122,
824
+ "step": 13600
825
+ },
826
+ {
827
+ "epoch": 0.85,
828
+ "learning_rate": 0.0005480000000000001,
829
+ "loss": 3.101,
830
+ "step": 13700
831
+ },
832
+ {
833
+ "epoch": 0.86,
834
+ "learning_rate": 0.0005520000000000001,
835
+ "loss": 3.1152,
836
+ "step": 13800
837
+ },
838
+ {
839
+ "epoch": 0.86,
840
+ "learning_rate": 0.0005560000000000001,
841
+ "loss": 3.0982,
842
+ "step": 13900
843
+ },
844
+ {
845
+ "epoch": 0.87,
846
+ "learning_rate": 0.0005600000000000001,
847
+ "loss": 3.1003,
848
+ "step": 14000
849
+ },
850
+ {
851
+ "epoch": 0.87,
852
+ "learning_rate": 0.0005639999999999999,
853
+ "loss": 3.1038,
854
+ "step": 14100
855
+ },
856
+ {
857
+ "epoch": 0.88,
858
+ "learning_rate": 0.0005679999999999999,
859
+ "loss": 3.1052,
860
+ "step": 14200
861
+ },
862
+ {
863
+ "epoch": 0.89,
864
+ "learning_rate": 0.0005719999999999999,
865
+ "loss": 3.0987,
866
+ "step": 14300
867
+ },
868
+ {
869
+ "epoch": 0.89,
870
+ "learning_rate": 0.000576,
871
+ "loss": 3.0798,
872
+ "step": 14400
873
+ },
874
+ {
875
+ "epoch": 0.9,
876
+ "learning_rate": 0.00058,
877
+ "loss": 3.0898,
878
+ "step": 14500
879
+ },
880
+ {
881
+ "epoch": 0.91,
882
+ "learning_rate": 0.000584,
883
+ "loss": 3.0645,
884
+ "step": 14600
885
+ },
886
+ {
887
+ "epoch": 0.91,
888
+ "learning_rate": 0.000588,
889
+ "loss": 3.085,
890
+ "step": 14700
891
+ },
892
+ {
893
+ "epoch": 0.92,
894
+ "learning_rate": 0.000592,
895
+ "loss": 3.0773,
896
+ "step": 14800
897
+ },
898
+ {
899
+ "epoch": 0.92,
900
+ "learning_rate": 0.000596,
901
+ "loss": 3.081,
902
+ "step": 14900
903
+ },
904
+ {
905
+ "epoch": 0.93,
906
+ "learning_rate": 0.0006,
907
+ "loss": 3.0728,
908
+ "step": 15000
909
+ },
910
+ {
911
+ "epoch": 0.94,
912
+ "learning_rate": 0.000604,
913
+ "loss": 3.0774,
914
+ "step": 15100
915
+ },
916
+ {
917
+ "epoch": 0.94,
918
+ "learning_rate": 0.000608,
919
+ "loss": 3.0598,
920
+ "step": 15200
921
+ },
922
+ {
923
+ "epoch": 0.95,
924
+ "learning_rate": 0.000612,
925
+ "loss": 3.064,
926
+ "step": 15300
927
+ },
928
+ {
929
+ "epoch": 0.95,
930
+ "learning_rate": 0.000616,
931
+ "loss": 3.0875,
932
+ "step": 15400
933
+ },
934
+ {
935
+ "epoch": 0.96,
936
+ "learning_rate": 0.00062,
937
+ "loss": 3.0779,
938
+ "step": 15500
939
+ },
940
+ {
941
+ "epoch": 0.97,
942
+ "learning_rate": 0.000624,
943
+ "loss": 3.0626,
944
+ "step": 15600
945
+ },
946
+ {
947
+ "epoch": 0.97,
948
+ "learning_rate": 0.000628,
949
+ "loss": 3.0632,
950
+ "step": 15700
951
+ },
952
+ {
953
+ "epoch": 0.98,
954
+ "learning_rate": 0.000632,
955
+ "loss": 3.0534,
956
+ "step": 15800
957
+ },
958
+ {
959
+ "epoch": 0.99,
960
+ "learning_rate": 0.0006360000000000001,
961
+ "loss": 3.0666,
962
+ "step": 15900
963
+ },
964
+ {
965
+ "epoch": 0.99,
966
+ "learning_rate": 0.00064,
967
+ "loss": 3.0665,
968
+ "step": 16000
969
+ },
970
+ {
971
+ "epoch": 1.0,
972
+ "learning_rate": 0.000644,
973
+ "loss": 3.0614,
974
+ "step": 16100
975
+ },
976
+ {
977
+ "epoch": 1.0,
978
+ "learning_rate": 0.000648,
979
+ "loss": 3.0527,
980
+ "step": 16200
981
+ },
982
+ {
983
+ "epoch": 1.01,
984
+ "learning_rate": 0.000652,
985
+ "loss": 3.0384,
986
+ "step": 16300
987
+ },
988
+ {
989
+ "epoch": 1.02,
990
+ "learning_rate": 0.000656,
991
+ "loss": 3.0633,
992
+ "step": 16400
993
+ },
994
+ {
995
+ "epoch": 1.02,
996
+ "learning_rate": 0.00066,
997
+ "loss": 3.0308,
998
+ "step": 16500
999
+ },
1000
+ {
1001
+ "epoch": 1.03,
1002
+ "learning_rate": 0.0006640000000000001,
1003
+ "loss": 3.0476,
1004
+ "step": 16600
1005
+ },
1006
+ {
1007
+ "epoch": 1.04,
1008
+ "learning_rate": 0.0006680000000000001,
1009
+ "loss": 3.0514,
1010
+ "step": 16700
1011
+ },
1012
+ {
1013
+ "epoch": 1.04,
1014
+ "learning_rate": 0.0006720000000000001,
1015
+ "loss": 3.057,
1016
+ "step": 16800
1017
+ },
1018
+ {
1019
+ "epoch": 1.05,
1020
+ "learning_rate": 0.0006760000000000001,
1021
+ "loss": 3.0202,
1022
+ "step": 16900
1023
+ },
1024
+ {
1025
+ "epoch": 1.05,
1026
+ "learning_rate": 0.00068,
1027
+ "loss": 3.0328,
1028
+ "step": 17000
1029
+ },
1030
+ {
1031
+ "epoch": 1.06,
1032
+ "learning_rate": 0.000684,
1033
+ "loss": 3.0422,
1034
+ "step": 17100
1035
+ },
1036
+ {
1037
+ "epoch": 1.07,
1038
+ "learning_rate": 0.0006879999999999999,
1039
+ "loss": 3.0236,
1040
+ "step": 17200
1041
+ },
1042
+ {
1043
+ "epoch": 1.07,
1044
+ "learning_rate": 0.000692,
1045
+ "loss": 3.0257,
1046
+ "step": 17300
1047
+ },
1048
+ {
1049
+ "epoch": 1.08,
1050
+ "learning_rate": 0.000696,
1051
+ "loss": 3.0346,
1052
+ "step": 17400
1053
+ },
1054
+ {
1055
+ "epoch": 1.08,
1056
+ "learning_rate": 0.0007,
1057
+ "loss": 3.0428,
1058
+ "step": 17500
1059
+ },
1060
+ {
1061
+ "epoch": 1.09,
1062
+ "learning_rate": 0.000704,
1063
+ "loss": 3.0297,
1064
+ "step": 17600
1065
+ },
1066
+ {
1067
+ "epoch": 1.1,
1068
+ "learning_rate": 0.000708,
1069
+ "loss": 3.035,
1070
+ "step": 17700
1071
+ },
1072
+ {
1073
+ "epoch": 1.1,
1074
+ "learning_rate": 0.000712,
1075
+ "loss": 3.0265,
1076
+ "step": 17800
1077
+ },
1078
+ {
1079
+ "epoch": 1.11,
1080
+ "learning_rate": 0.000716,
1081
+ "loss": 3.0069,
1082
+ "step": 17900
1083
+ },
1084
+ {
1085
+ "epoch": 1.12,
1086
+ "learning_rate": 0.0007199999999999999,
1087
+ "loss": 3.044,
1088
+ "step": 18000
1089
+ },
1090
+ {
1091
+ "epoch": 1.12,
1092
+ "learning_rate": 0.000724,
1093
+ "loss": 3.0174,
1094
+ "step": 18100
1095
+ },
1096
+ {
1097
+ "epoch": 1.13,
1098
+ "learning_rate": 0.000728,
1099
+ "loss": 3.0153,
1100
+ "step": 18200
1101
+ },
1102
+ {
1103
+ "epoch": 1.13,
1104
+ "learning_rate": 0.000732,
1105
+ "loss": 3.0149,
1106
+ "step": 18300
1107
+ },
1108
+ {
1109
+ "epoch": 1.14,
1110
+ "learning_rate": 0.000736,
1111
+ "loss": 3.0094,
1112
+ "step": 18400
1113
+ },
1114
+ {
1115
+ "epoch": 1.15,
1116
+ "learning_rate": 0.00074,
1117
+ "loss": 3.0139,
1118
+ "step": 18500
1119
+ },
1120
+ {
1121
+ "epoch": 1.15,
1122
+ "learning_rate": 0.000744,
1123
+ "loss": 3.0115,
1124
+ "step": 18600
1125
+ },
1126
+ {
1127
+ "epoch": 1.16,
1128
+ "learning_rate": 0.000748,
1129
+ "loss": 3.0049,
1130
+ "step": 18700
1131
+ },
1132
+ {
1133
+ "epoch": 1.17,
1134
+ "learning_rate": 0.0007520000000000001,
1135
+ "loss": 3.0011,
1136
+ "step": 18800
1137
+ },
1138
+ {
1139
+ "epoch": 1.17,
1140
+ "learning_rate": 0.000756,
1141
+ "loss": 3.0098,
1142
+ "step": 18900
1143
+ },
1144
+ {
1145
+ "epoch": 1.18,
1146
+ "learning_rate": 0.00076,
1147
+ "loss": 2.9991,
1148
+ "step": 19000
1149
+ },
1150
+ {
1151
+ "epoch": 1.18,
1152
+ "learning_rate": 0.000764,
1153
+ "loss": 3.0096,
1154
+ "step": 19100
1155
+ },
1156
+ {
1157
+ "epoch": 1.19,
1158
+ "learning_rate": 0.000768,
1159
+ "loss": 3.0174,
1160
+ "step": 19200
1161
+ },
1162
+ {
1163
+ "epoch": 1.2,
1164
+ "learning_rate": 0.000772,
1165
+ "loss": 3.0131,
1166
+ "step": 19300
1167
+ },
1168
+ {
1169
+ "epoch": 1.2,
1170
+ "learning_rate": 0.000776,
1171
+ "loss": 3.003,
1172
+ "step": 19400
1173
+ },
1174
+ {
1175
+ "epoch": 1.21,
1176
+ "learning_rate": 0.0007800000000000001,
1177
+ "loss": 3.0081,
1178
+ "step": 19500
1179
+ },
1180
+ {
1181
+ "epoch": 1.21,
1182
+ "learning_rate": 0.0007840000000000001,
1183
+ "loss": 2.9973,
1184
+ "step": 19600
1185
+ },
1186
+ {
1187
+ "epoch": 1.22,
1188
+ "learning_rate": 0.0007880000000000001,
1189
+ "loss": 3.0083,
1190
+ "step": 19700
1191
+ },
1192
+ {
1193
+ "epoch": 1.23,
1194
+ "learning_rate": 0.0007920000000000001,
1195
+ "loss": 3.0031,
1196
+ "step": 19800
1197
+ },
1198
+ {
1199
+ "epoch": 1.23,
1200
+ "learning_rate": 0.000796,
1201
+ "loss": 2.9853,
1202
+ "step": 19900
1203
+ },
1204
+ {
1205
+ "epoch": 1.24,
1206
+ "learning_rate": 0.0008,
1207
+ "loss": 2.9972,
1208
+ "step": 20000
1209
+ },
1210
+ {
1211
+ "epoch": 1.25,
1212
+ "learning_rate": 0.000804,
1213
+ "loss": 2.9945,
1214
+ "step": 20100
1215
+ },
1216
+ {
1217
+ "epoch": 1.25,
1218
+ "learning_rate": 0.000808,
1219
+ "loss": 2.9802,
1220
+ "step": 20200
1221
+ },
1222
+ {
1223
+ "epoch": 1.26,
1224
+ "learning_rate": 0.0008120000000000001,
1225
+ "loss": 2.9833,
1226
+ "step": 20300
1227
+ },
1228
+ {
1229
+ "epoch": 1.26,
1230
+ "learning_rate": 0.000816,
1231
+ "loss": 2.9901,
1232
+ "step": 20400
1233
+ },
1234
+ {
1235
+ "epoch": 1.27,
1236
+ "learning_rate": 0.00082,
1237
+ "loss": 2.9983,
1238
+ "step": 20500
1239
+ },
1240
+ {
1241
+ "epoch": 1.28,
1242
+ "learning_rate": 0.000824,
1243
+ "loss": 2.9861,
1244
+ "step": 20600
1245
+ },
1246
+ {
1247
+ "epoch": 1.28,
1248
+ "learning_rate": 0.000828,
1249
+ "loss": 2.9787,
1250
+ "step": 20700
1251
+ },
1252
+ {
1253
+ "epoch": 1.29,
1254
+ "learning_rate": 0.000832,
1255
+ "loss": 2.9893,
1256
+ "step": 20800
1257
+ },
1258
+ {
1259
+ "epoch": 1.3,
1260
+ "learning_rate": 0.0008359999999999999,
1261
+ "loss": 2.9919,
1262
+ "step": 20900
1263
+ },
1264
+ {
1265
+ "epoch": 1.3,
1266
+ "learning_rate": 0.00084,
1267
+ "loss": 2.9547,
1268
+ "step": 21000
1269
+ },
1270
+ {
1271
+ "epoch": 1.31,
1272
+ "learning_rate": 0.000844,
1273
+ "loss": 2.9825,
1274
+ "step": 21100
1275
+ },
1276
+ {
1277
+ "epoch": 1.31,
1278
+ "learning_rate": 0.000848,
1279
+ "loss": 2.9905,
1280
+ "step": 21200
1281
+ },
1282
+ {
1283
+ "epoch": 1.32,
1284
+ "learning_rate": 0.000852,
1285
+ "loss": 2.982,
1286
+ "step": 21300
1287
+ },
1288
+ {
1289
+ "epoch": 1.33,
1290
+ "learning_rate": 0.000856,
1291
+ "loss": 2.9726,
1292
+ "step": 21400
1293
+ },
1294
+ {
1295
+ "epoch": 1.33,
1296
+ "learning_rate": 0.00086,
1297
+ "loss": 2.9559,
1298
+ "step": 21500
1299
+ },
1300
+ {
1301
+ "epoch": 1.34,
1302
+ "learning_rate": 0.000864,
1303
+ "loss": 2.9818,
1304
+ "step": 21600
1305
+ },
1306
+ {
1307
+ "epoch": 1.35,
1308
+ "learning_rate": 0.0008680000000000001,
1309
+ "loss": 2.978,
1310
+ "step": 21700
1311
+ },
1312
+ {
1313
+ "epoch": 1.35,
1314
+ "learning_rate": 0.000872,
1315
+ "loss": 2.9851,
1316
+ "step": 21800
1317
+ },
1318
+ {
1319
+ "epoch": 1.36,
1320
+ "learning_rate": 0.000876,
1321
+ "loss": 2.9761,
1322
+ "step": 21900
1323
+ },
1324
+ {
1325
+ "epoch": 1.36,
1326
+ "learning_rate": 0.00088,
1327
+ "loss": 2.9584,
1328
+ "step": 22000
1329
+ },
1330
+ {
1331
+ "epoch": 1.37,
1332
+ "learning_rate": 0.000884,
1333
+ "loss": 2.9792,
1334
+ "step": 22100
1335
+ },
1336
+ {
1337
+ "epoch": 1.38,
1338
+ "learning_rate": 0.000888,
1339
+ "loss": 2.9692,
1340
+ "step": 22200
1341
+ },
1342
+ {
1343
+ "epoch": 1.38,
1344
+ "learning_rate": 0.000892,
1345
+ "loss": 2.9636,
1346
+ "step": 22300
1347
+ },
1348
+ {
1349
+ "epoch": 1.39,
1350
+ "learning_rate": 0.000896,
1351
+ "loss": 2.9668,
1352
+ "step": 22400
1353
+ },
1354
+ {
1355
+ "epoch": 1.39,
1356
+ "learning_rate": 0.0009000000000000001,
1357
+ "loss": 2.9632,
1358
+ "step": 22500
1359
+ },
1360
+ {
1361
+ "epoch": 1.4,
1362
+ "learning_rate": 0.0009040000000000001,
1363
+ "loss": 2.9857,
1364
+ "step": 22600
1365
+ },
1366
+ {
1367
+ "epoch": 1.41,
1368
+ "learning_rate": 0.0009080000000000001,
1369
+ "loss": 2.9574,
1370
+ "step": 22700
1371
+ },
1372
+ {
1373
+ "epoch": 1.41,
1374
+ "learning_rate": 0.000912,
1375
+ "loss": 2.9548,
1376
+ "step": 22800
1377
+ },
1378
+ {
1379
+ "epoch": 1.42,
1380
+ "learning_rate": 0.000916,
1381
+ "loss": 2.9772,
1382
+ "step": 22900
1383
+ },
1384
+ {
1385
+ "epoch": 1.43,
1386
+ "learning_rate": 0.00092,
1387
+ "loss": 2.9661,
1388
+ "step": 23000
1389
+ },
1390
+ {
1391
+ "epoch": 1.43,
1392
+ "learning_rate": 0.000924,
1393
+ "loss": 2.9418,
1394
+ "step": 23100
1395
+ },
1396
+ {
1397
+ "epoch": 1.44,
1398
+ "learning_rate": 0.0009280000000000001,
1399
+ "loss": 2.9544,
1400
+ "step": 23200
1401
+ },
1402
+ {
1403
+ "epoch": 1.44,
1404
+ "learning_rate": 0.0009320000000000001,
1405
+ "loss": 2.978,
1406
+ "step": 23300
1407
+ },
1408
+ {
1409
+ "epoch": 1.45,
1410
+ "learning_rate": 0.0009360000000000001,
1411
+ "loss": 2.9759,
1412
+ "step": 23400
1413
+ },
1414
+ {
1415
+ "epoch": 1.46,
1416
+ "learning_rate": 0.00094,
1417
+ "loss": 2.9494,
1418
+ "step": 23500
1419
+ },
1420
+ {
1421
+ "epoch": 1.46,
1422
+ "learning_rate": 0.000944,
1423
+ "loss": 2.9474,
1424
+ "step": 23600
1425
+ },
1426
+ {
1427
+ "epoch": 1.47,
1428
+ "learning_rate": 0.000948,
1429
+ "loss": 2.9466,
1430
+ "step": 23700
1431
+ },
1432
+ {
1433
+ "epoch": 1.48,
1434
+ "learning_rate": 0.0009519999999999999,
1435
+ "loss": 2.9378,
1436
+ "step": 23800
1437
+ },
1438
+ {
1439
+ "epoch": 1.48,
1440
+ "learning_rate": 0.0009559999999999999,
1441
+ "loss": 3.0109,
1442
+ "step": 23900
1443
+ },
1444
+ {
1445
+ "epoch": 1.49,
1446
+ "learning_rate": 0.00096,
1447
+ "loss": 2.969,
1448
+ "step": 24000
1449
+ },
1450
+ {
1451
+ "epoch": 1.49,
1452
+ "learning_rate": 0.000964,
1453
+ "loss": 2.9366,
1454
+ "step": 24100
1455
+ },
1456
+ {
1457
+ "epoch": 1.5,
1458
+ "learning_rate": 0.000968,
1459
+ "loss": 2.9381,
1460
+ "step": 24200
1461
+ },
1462
+ {
1463
+ "epoch": 1.51,
1464
+ "learning_rate": 0.000972,
1465
+ "loss": 2.9382,
1466
+ "step": 24300
1467
+ },
1468
+ {
1469
+ "epoch": 1.51,
1470
+ "learning_rate": 0.000976,
1471
+ "loss": 2.9519,
1472
+ "step": 24400
1473
+ },
1474
+ {
1475
+ "epoch": 1.52,
1476
+ "learning_rate": 0.00098,
1477
+ "loss": 2.9452,
1478
+ "step": 24500
1479
+ },
1480
+ {
1481
+ "epoch": 1.52,
1482
+ "learning_rate": 0.000984,
1483
+ "loss": 2.942,
1484
+ "step": 24600
1485
+ },
1486
+ {
1487
+ "epoch": 1.53,
1488
+ "learning_rate": 0.000988,
1489
+ "loss": 2.9315,
1490
+ "step": 24700
1491
+ },
1492
+ {
1493
+ "epoch": 1.54,
1494
+ "learning_rate": 0.000992,
1495
+ "loss": 2.938,
1496
+ "step": 24800
1497
+ },
1498
+ {
1499
+ "epoch": 1.54,
1500
+ "learning_rate": 0.000996,
1501
+ "loss": 2.9331,
1502
+ "step": 24900
1503
+ },
1504
+ {
1505
+ "epoch": 1.55,
1506
+ "learning_rate": 0.001,
1507
+ "loss": 2.9488,
1508
+ "step": 25000
1509
+ },
1510
+ {
1511
+ "epoch": 1.55,
1512
+ "eval_accuracy": 0.49347509891775837,
1513
+ "eval_loss": 2.7666990756988525,
1514
+ "eval_runtime": 6163.6406,
1515
+ "eval_samples_per_second": 35.265,
1516
+ "eval_steps_per_second": 2.204,
1517
+ "step": 25000
1518
+ },
1519
+ {
1520
+ "epoch": 1.56,
1521
+ "learning_rate": 0.000996,
1522
+ "loss": 2.9298,
1523
+ "step": 25100
1524
+ },
1525
+ {
1526
+ "epoch": 1.56,
1527
+ "learning_rate": 0.000992,
1528
+ "loss": 2.93,
1529
+ "step": 25200
1530
+ },
1531
+ {
1532
+ "epoch": 1.57,
1533
+ "learning_rate": 0.000988,
1534
+ "loss": 2.9365,
1535
+ "step": 25300
1536
+ },
1537
+ {
1538
+ "epoch": 1.57,
1539
+ "learning_rate": 0.000984,
1540
+ "loss": 2.9522,
1541
+ "step": 25400
1542
+ },
1543
+ {
1544
+ "epoch": 1.58,
1545
+ "learning_rate": 0.00098,
1546
+ "loss": 2.9344,
1547
+ "step": 25500
1548
+ },
1549
+ {
1550
+ "epoch": 1.59,
1551
+ "learning_rate": 0.000976,
1552
+ "loss": 2.9349,
1553
+ "step": 25600
1554
+ },
1555
+ {
1556
+ "epoch": 1.59,
1557
+ "learning_rate": 0.000972,
1558
+ "loss": 2.922,
1559
+ "step": 25700
1560
+ },
1561
+ {
1562
+ "epoch": 1.6,
1563
+ "learning_rate": 0.000968,
1564
+ "loss": 2.9094,
1565
+ "step": 25800
1566
+ },
1567
+ {
1568
+ "epoch": 1.61,
1569
+ "learning_rate": 0.000964,
1570
+ "loss": 2.9079,
1571
+ "step": 25900
1572
+ },
1573
+ {
1574
+ "epoch": 1.61,
1575
+ "learning_rate": 0.00096,
1576
+ "loss": 2.9033,
1577
+ "step": 26000
1578
+ },
1579
+ {
1580
+ "epoch": 1.62,
1581
+ "learning_rate": 0.0009559999999999999,
1582
+ "loss": 2.8957,
1583
+ "step": 26100
1584
+ },
1585
+ {
1586
+ "epoch": 1.62,
1587
+ "learning_rate": 0.0009519999999999999,
1588
+ "loss": 2.9106,
1589
+ "step": 26200
1590
+ },
1591
+ {
1592
+ "epoch": 1.63,
1593
+ "learning_rate": 0.000948,
1594
+ "loss": 2.9186,
1595
+ "step": 26300
1596
+ },
1597
+ {
1598
+ "epoch": 1.64,
1599
+ "learning_rate": 0.000944,
1600
+ "loss": 2.8864,
1601
+ "step": 26400
1602
+ },
1603
+ {
1604
+ "epoch": 1.64,
1605
+ "learning_rate": 0.00094,
1606
+ "loss": 2.9045,
1607
+ "step": 26500
1608
+ },
1609
+ {
1610
+ "epoch": 1.65,
1611
+ "learning_rate": 0.0009360000000000001,
1612
+ "loss": 2.8999,
1613
+ "step": 26600
1614
+ },
1615
+ {
1616
+ "epoch": 1.66,
1617
+ "learning_rate": 0.0009320000000000001,
1618
+ "loss": 2.8869,
1619
+ "step": 26700
1620
+ },
1621
+ {
1622
+ "epoch": 1.66,
1623
+ "learning_rate": 0.0009280000000000001,
1624
+ "loss": 2.8616,
1625
+ "step": 26800
1626
+ },
1627
+ {
1628
+ "epoch": 1.67,
1629
+ "learning_rate": 0.000924,
1630
+ "loss": 2.8826,
1631
+ "step": 26900
1632
+ },
1633
+ {
1634
+ "epoch": 1.67,
1635
+ "learning_rate": 0.00092,
1636
+ "loss": 2.8858,
1637
+ "step": 27000
1638
+ },
1639
+ {
1640
+ "epoch": 1.68,
1641
+ "learning_rate": 0.000916,
1642
+ "loss": 2.8868,
1643
+ "step": 27100
1644
+ },
1645
+ {
1646
+ "epoch": 1.69,
1647
+ "learning_rate": 0.000912,
1648
+ "loss": 2.8586,
1649
+ "step": 27200
1650
+ },
1651
+ {
1652
+ "epoch": 1.69,
1653
+ "learning_rate": 0.0009080000000000001,
1654
+ "loss": 2.872,
1655
+ "step": 27300
1656
+ },
1657
+ {
1658
+ "epoch": 1.7,
1659
+ "learning_rate": 0.0009040000000000001,
1660
+ "loss": 2.8575,
1661
+ "step": 27400
1662
+ },
1663
+ {
1664
+ "epoch": 1.7,
1665
+ "learning_rate": 0.0009000000000000001,
1666
+ "loss": 2.8648,
1667
+ "step": 27500
1668
+ },
1669
+ {
1670
+ "epoch": 1.71,
1671
+ "learning_rate": 0.000896,
1672
+ "loss": 2.8639,
1673
+ "step": 27600
1674
+ },
1675
+ {
1676
+ "epoch": 1.72,
1677
+ "learning_rate": 0.000892,
1678
+ "loss": 2.8573,
1679
+ "step": 27700
1680
+ },
1681
+ {
1682
+ "epoch": 1.72,
1683
+ "learning_rate": 0.000888,
1684
+ "loss": 2.849,
1685
+ "step": 27800
1686
+ },
1687
+ {
1688
+ "epoch": 1.73,
1689
+ "learning_rate": 0.000884,
1690
+ "loss": 2.8566,
1691
+ "step": 27900
1692
+ },
1693
+ {
1694
+ "epoch": 1.74,
1695
+ "learning_rate": 0.00088,
1696
+ "loss": 2.8574,
1697
+ "step": 28000
1698
+ },
1699
+ {
1700
+ "epoch": 1.74,
1701
+ "learning_rate": 0.000876,
1702
+ "loss": 2.8519,
1703
+ "step": 28100
1704
+ },
1705
+ {
1706
+ "epoch": 1.75,
1707
+ "learning_rate": 0.000872,
1708
+ "loss": 2.8346,
1709
+ "step": 28200
1710
+ },
1711
+ {
1712
+ "epoch": 1.75,
1713
+ "learning_rate": 0.0008680000000000001,
1714
+ "loss": 2.8502,
1715
+ "step": 28300
1716
+ },
1717
+ {
1718
+ "epoch": 1.76,
1719
+ "learning_rate": 0.000864,
1720
+ "loss": 2.8269,
1721
+ "step": 28400
1722
+ },
1723
+ {
1724
+ "epoch": 1.77,
1725
+ "learning_rate": 0.00086,
1726
+ "loss": 2.842,
1727
+ "step": 28500
1728
+ },
1729
+ {
1730
+ "epoch": 1.77,
1731
+ "learning_rate": 0.000856,
1732
+ "loss": 2.8386,
1733
+ "step": 28600
1734
+ },
1735
+ {
1736
+ "epoch": 1.78,
1737
+ "learning_rate": 0.000852,
1738
+ "loss": 2.8188,
1739
+ "step": 28700
1740
+ },
1741
+ {
1742
+ "epoch": 1.79,
1743
+ "learning_rate": 0.000848,
1744
+ "loss": 2.8147,
1745
+ "step": 28800
1746
+ },
1747
+ {
1748
+ "epoch": 1.79,
1749
+ "learning_rate": 0.000844,
1750
+ "loss": 2.82,
1751
+ "step": 28900
1752
+ },
1753
+ {
1754
+ "epoch": 1.8,
1755
+ "learning_rate": 0.00084,
1756
+ "loss": 2.8251,
1757
+ "step": 29000
1758
+ },
1759
+ {
1760
+ "epoch": 1.8,
1761
+ "learning_rate": 0.0008359999999999999,
1762
+ "loss": 2.8202,
1763
+ "step": 29100
1764
+ },
1765
+ {
1766
+ "epoch": 1.81,
1767
+ "learning_rate": 0.000832,
1768
+ "loss": 2.808,
1769
+ "step": 29200
1770
+ },
1771
+ {
1772
+ "epoch": 1.82,
1773
+ "learning_rate": 0.000828,
1774
+ "loss": 2.819,
1775
+ "step": 29300
1776
+ },
1777
+ {
1778
+ "epoch": 1.82,
1779
+ "learning_rate": 0.000824,
1780
+ "loss": 2.8127,
1781
+ "step": 29400
1782
+ },
1783
+ {
1784
+ "epoch": 1.83,
1785
+ "learning_rate": 0.00082,
1786
+ "loss": 2.7945,
1787
+ "step": 29500
1788
+ },
1789
+ {
1790
+ "epoch": 1.83,
1791
+ "learning_rate": 0.000816,
1792
+ "loss": 2.8057,
1793
+ "step": 29600
1794
+ },
1795
+ {
1796
+ "epoch": 1.84,
1797
+ "learning_rate": 0.0008120000000000001,
1798
+ "loss": 2.8077,
1799
+ "step": 29700
1800
+ },
1801
+ {
1802
+ "epoch": 1.85,
1803
+ "learning_rate": 0.000808,
1804
+ "loss": 2.809,
1805
+ "step": 29800
1806
+ },
1807
+ {
1808
+ "epoch": 1.85,
1809
+ "learning_rate": 0.000804,
1810
+ "loss": 2.8029,
1811
+ "step": 29900
1812
+ },
1813
+ {
1814
+ "epoch": 1.86,
1815
+ "learning_rate": 0.0008,
1816
+ "loss": 2.7945,
1817
+ "step": 30000
1818
+ },
1819
+ {
1820
+ "epoch": 1.87,
1821
+ "learning_rate": 0.000796,
1822
+ "loss": 2.7991,
1823
+ "step": 30100
1824
+ },
1825
+ {
1826
+ "epoch": 1.87,
1827
+ "learning_rate": 0.0007920000000000001,
1828
+ "loss": 2.7869,
1829
+ "step": 30200
1830
+ },
1831
+ {
1832
+ "epoch": 1.88,
1833
+ "learning_rate": 0.0007880000000000001,
1834
+ "loss": 2.7931,
1835
+ "step": 30300
1836
+ },
1837
+ {
1838
+ "epoch": 1.88,
1839
+ "learning_rate": 0.0007840000000000001,
1840
+ "loss": 2.7927,
1841
+ "step": 30400
1842
+ },
1843
+ {
1844
+ "epoch": 1.89,
1845
+ "learning_rate": 0.0007800000000000001,
1846
+ "loss": 2.787,
1847
+ "step": 30500
1848
+ },
1849
+ {
1850
+ "epoch": 1.9,
1851
+ "learning_rate": 0.000776,
1852
+ "loss": 2.7747,
1853
+ "step": 30600
1854
+ },
1855
+ {
1856
+ "epoch": 1.9,
1857
+ "learning_rate": 0.000772,
1858
+ "loss": 2.7916,
1859
+ "step": 30700
1860
+ },
1861
+ {
1862
+ "epoch": 1.91,
1863
+ "learning_rate": 0.000768,
1864
+ "loss": 2.7759,
1865
+ "step": 30800
1866
+ },
1867
+ {
1868
+ "epoch": 1.92,
1869
+ "learning_rate": 0.000764,
1870
+ "loss": 2.7599,
1871
+ "step": 30900
1872
+ },
1873
+ {
1874
+ "epoch": 1.92,
1875
+ "learning_rate": 0.00076,
1876
+ "loss": 2.7578,
1877
+ "step": 31000
1878
+ },
1879
+ {
1880
+ "epoch": 1.93,
1881
+ "learning_rate": 0.000756,
1882
+ "loss": 2.7688,
1883
+ "step": 31100
1884
+ },
1885
+ {
1886
+ "epoch": 1.93,
1887
+ "learning_rate": 0.0007520000000000001,
1888
+ "loss": 2.7586,
1889
+ "step": 31200
1890
+ },
1891
+ {
1892
+ "epoch": 1.94,
1893
+ "learning_rate": 0.000748,
1894
+ "loss": 2.7641,
1895
+ "step": 31300
1896
+ },
1897
+ {
1898
+ "epoch": 1.95,
1899
+ "learning_rate": 0.000744,
1900
+ "loss": 2.7688,
1901
+ "step": 31400
1902
+ },
1903
+ {
1904
+ "epoch": 1.95,
1905
+ "learning_rate": 0.00074,
1906
+ "loss": 2.761,
1907
+ "step": 31500
1908
+ },
1909
+ {
1910
+ "epoch": 1.96,
1911
+ "learning_rate": 0.000736,
1912
+ "loss": 2.7543,
1913
+ "step": 31600
1914
+ },
1915
+ {
1916
+ "epoch": 1.97,
1917
+ "learning_rate": 0.000732,
1918
+ "loss": 2.7434,
1919
+ "step": 31700
1920
+ },
1921
+ {
1922
+ "epoch": 1.97,
1923
+ "learning_rate": 0.000728,
1924
+ "loss": 2.7458,
1925
+ "step": 31800
1926
+ },
1927
+ {
1928
+ "epoch": 1.98,
1929
+ "learning_rate": 0.000724,
1930
+ "loss": 2.7583,
1931
+ "step": 31900
1932
+ },
1933
+ {
1934
+ "epoch": 1.98,
1935
+ "learning_rate": 0.0007199999999999999,
1936
+ "loss": 2.7356,
1937
+ "step": 32000
1938
+ },
1939
+ {
1940
+ "epoch": 1.99,
1941
+ "learning_rate": 0.000716,
1942
+ "loss": 2.7437,
1943
+ "step": 32100
1944
+ },
1945
+ {
1946
+ "epoch": 2.0,
1947
+ "learning_rate": 0.000712,
1948
+ "loss": 2.7392,
1949
+ "step": 32200
1950
+ },
1951
+ {
1952
+ "epoch": 2.0,
1953
+ "learning_rate": 0.000708,
1954
+ "loss": 2.7443,
1955
+ "step": 32300
1956
+ },
1957
+ {
1958
+ "epoch": 2.01,
1959
+ "learning_rate": 0.000704,
1960
+ "loss": 2.7234,
1961
+ "step": 32400
1962
+ },
1963
+ {
1964
+ "epoch": 2.01,
1965
+ "learning_rate": 0.0007,
1966
+ "loss": 2.7302,
1967
+ "step": 32500
1968
+ },
1969
+ {
1970
+ "epoch": 2.02,
1971
+ "learning_rate": 0.000696,
1972
+ "loss": 2.72,
1973
+ "step": 32600
1974
+ },
1975
+ {
1976
+ "epoch": 2.03,
1977
+ "learning_rate": 0.000692,
1978
+ "loss": 2.7183,
1979
+ "step": 32700
1980
+ },
1981
+ {
1982
+ "epoch": 2.03,
1983
+ "learning_rate": 0.0006879999999999999,
1984
+ "loss": 2.7129,
1985
+ "step": 32800
1986
+ },
1987
+ {
1988
+ "epoch": 2.04,
1989
+ "learning_rate": 0.000684,
1990
+ "loss": 2.7253,
1991
+ "step": 32900
1992
+ },
1993
+ {
1994
+ "epoch": 2.05,
1995
+ "learning_rate": 0.00068,
1996
+ "loss": 2.7299,
1997
+ "step": 33000
1998
+ },
1999
+ {
2000
+ "epoch": 2.05,
2001
+ "learning_rate": 0.0006760000000000001,
2002
+ "loss": 2.7195,
2003
+ "step": 33100
2004
+ },
2005
+ {
2006
+ "epoch": 2.06,
2007
+ "learning_rate": 0.0006720000000000001,
2008
+ "loss": 2.7026,
2009
+ "step": 33200
2010
+ },
2011
+ {
2012
+ "epoch": 2.06,
2013
+ "learning_rate": 0.0006680000000000001,
2014
+ "loss": 2.7002,
2015
+ "step": 33300
2016
+ },
2017
+ {
2018
+ "epoch": 2.07,
2019
+ "learning_rate": 0.0006640000000000001,
2020
+ "loss": 2.7065,
2021
+ "step": 33400
2022
+ },
2023
+ {
2024
+ "epoch": 2.08,
2025
+ "learning_rate": 0.00066,
2026
+ "loss": 2.7092,
2027
+ "step": 33500
2028
+ },
2029
+ {
2030
+ "epoch": 2.08,
2031
+ "learning_rate": 0.000656,
2032
+ "loss": 2.7014,
2033
+ "step": 33600
2034
+ },
2035
+ {
2036
+ "epoch": 2.09,
2037
+ "learning_rate": 0.000652,
2038
+ "loss": 2.7083,
2039
+ "step": 33700
2040
+ },
2041
+ {
2042
+ "epoch": 2.1,
2043
+ "learning_rate": 0.000648,
2044
+ "loss": 2.6784,
2045
+ "step": 33800
2046
+ },
2047
+ {
2048
+ "epoch": 2.1,
2049
+ "learning_rate": 0.000644,
2050
+ "loss": 2.6933,
2051
+ "step": 33900
2052
+ },
2053
+ {
2054
+ "epoch": 2.11,
2055
+ "learning_rate": 0.00064,
2056
+ "loss": 2.7003,
2057
+ "step": 34000
2058
+ },
2059
+ {
2060
+ "epoch": 2.11,
2061
+ "learning_rate": 0.0006360000000000001,
2062
+ "loss": 2.6866,
2063
+ "step": 34100
2064
+ },
2065
+ {
2066
+ "epoch": 2.12,
2067
+ "learning_rate": 0.000632,
2068
+ "loss": 2.6984,
2069
+ "step": 34200
2070
+ },
2071
+ {
2072
+ "epoch": 2.13,
2073
+ "learning_rate": 0.000628,
2074
+ "loss": 2.678,
2075
+ "step": 34300
2076
+ },
2077
+ {
2078
+ "epoch": 2.13,
2079
+ "learning_rate": 0.000624,
2080
+ "loss": 2.698,
2081
+ "step": 34400
2082
+ },
2083
+ {
2084
+ "epoch": 2.14,
2085
+ "learning_rate": 0.00062,
2086
+ "loss": 2.6927,
2087
+ "step": 34500
2088
+ },
2089
+ {
2090
+ "epoch": 2.14,
2091
+ "learning_rate": 0.000616,
2092
+ "loss": 2.6787,
2093
+ "step": 34600
2094
+ },
2095
+ {
2096
+ "epoch": 2.15,
2097
+ "learning_rate": 0.000612,
2098
+ "loss": 2.6916,
2099
+ "step": 34700
2100
+ },
2101
+ {
2102
+ "epoch": 2.16,
2103
+ "learning_rate": 0.000608,
2104
+ "loss": 2.6794,
2105
+ "step": 34800
2106
+ },
2107
+ {
2108
+ "epoch": 2.16,
2109
+ "learning_rate": 0.000604,
2110
+ "loss": 2.6723,
2111
+ "step": 34900
2112
+ },
2113
+ {
2114
+ "epoch": 2.17,
2115
+ "learning_rate": 0.0006,
2116
+ "loss": 2.6582,
2117
+ "step": 35000
2118
+ },
2119
+ {
2120
+ "epoch": 2.18,
2121
+ "learning_rate": 0.000596,
2122
+ "loss": 2.6683,
2123
+ "step": 35100
2124
+ },
2125
+ {
2126
+ "epoch": 2.18,
2127
+ "learning_rate": 0.000592,
2128
+ "loss": 2.6748,
2129
+ "step": 35200
2130
+ },
2131
+ {
2132
+ "epoch": 2.19,
2133
+ "learning_rate": 0.000588,
2134
+ "loss": 2.6711,
2135
+ "step": 35300
2136
+ },
2137
+ {
2138
+ "epoch": 2.19,
2139
+ "learning_rate": 0.000584,
2140
+ "loss": 2.6566,
2141
+ "step": 35400
2142
+ },
2143
+ {
2144
+ "epoch": 2.2,
2145
+ "learning_rate": 0.00058,
2146
+ "loss": 2.6589,
2147
+ "step": 35500
2148
+ },
2149
+ {
2150
+ "epoch": 2.21,
2151
+ "learning_rate": 0.000576,
2152
+ "loss": 2.6656,
2153
+ "step": 35600
2154
+ },
2155
+ {
2156
+ "epoch": 2.21,
2157
+ "learning_rate": 0.0005719999999999999,
2158
+ "loss": 2.6423,
2159
+ "step": 35700
2160
+ },
2161
+ {
2162
+ "epoch": 2.22,
2163
+ "learning_rate": 0.0005679999999999999,
2164
+ "loss": 2.6561,
2165
+ "step": 35800
2166
+ },
2167
+ {
2168
+ "epoch": 2.23,
2169
+ "learning_rate": 0.0005639999999999999,
2170
+ "loss": 2.6579,
2171
+ "step": 35900
2172
+ },
2173
+ {
2174
+ "epoch": 2.23,
2175
+ "learning_rate": 0.0005600000000000001,
2176
+ "loss": 2.6596,
2177
+ "step": 36000
2178
+ },
2179
+ {
2180
+ "epoch": 2.24,
2181
+ "learning_rate": 0.0005560000000000001,
2182
+ "loss": 2.6419,
2183
+ "step": 36100
2184
+ },
2185
+ {
2186
+ "epoch": 2.24,
2187
+ "learning_rate": 0.0005520000000000001,
2188
+ "loss": 2.6477,
2189
+ "step": 36200
2190
+ },
2191
+ {
2192
+ "epoch": 2.25,
2193
+ "learning_rate": 0.0005480000000000001,
2194
+ "loss": 2.6391,
2195
+ "step": 36300
2196
+ },
2197
+ {
2198
+ "epoch": 2.26,
2199
+ "learning_rate": 0.0005440000000000001,
2200
+ "loss": 2.6337,
2201
+ "step": 36400
2202
+ },
2203
+ {
2204
+ "epoch": 2.26,
2205
+ "learning_rate": 0.00054,
2206
+ "loss": 2.6417,
2207
+ "step": 36500
2208
+ },
2209
+ {
2210
+ "epoch": 2.27,
2211
+ "learning_rate": 0.000536,
2212
+ "loss": 2.6318,
2213
+ "step": 36600
2214
+ },
2215
+ {
2216
+ "epoch": 2.27,
2217
+ "learning_rate": 0.000532,
2218
+ "loss": 2.6409,
2219
+ "step": 36700
2220
+ },
2221
+ {
2222
+ "epoch": 2.28,
2223
+ "learning_rate": 0.000528,
2224
+ "loss": 2.6469,
2225
+ "step": 36800
2226
+ },
2227
+ {
2228
+ "epoch": 2.29,
2229
+ "learning_rate": 0.000524,
2230
+ "loss": 2.6394,
2231
+ "step": 36900
2232
+ },
2233
+ {
2234
+ "epoch": 2.29,
2235
+ "learning_rate": 0.0005200000000000001,
2236
+ "loss": 2.6342,
2237
+ "step": 37000
2238
+ },
2239
+ {
2240
+ "epoch": 2.3,
2241
+ "learning_rate": 0.0005160000000000001,
2242
+ "loss": 2.6335,
2243
+ "step": 37100
2244
+ },
2245
+ {
2246
+ "epoch": 2.31,
2247
+ "learning_rate": 0.000512,
2248
+ "loss": 2.6274,
2249
+ "step": 37200
2250
+ },
2251
+ {
2252
+ "epoch": 2.31,
2253
+ "learning_rate": 0.000508,
2254
+ "loss": 2.6209,
2255
+ "step": 37300
2256
+ },
2257
+ {
2258
+ "epoch": 2.32,
2259
+ "learning_rate": 0.000504,
2260
+ "loss": 2.6213,
2261
+ "step": 37400
2262
+ },
2263
+ {
2264
+ "epoch": 2.32,
2265
+ "learning_rate": 0.0005,
2266
+ "loss": 2.623,
2267
+ "step": 37500
2268
+ },
2269
+ {
2270
+ "epoch": 2.33,
2271
+ "learning_rate": 0.000496,
2272
+ "loss": 2.617,
2273
+ "step": 37600
2274
+ },
2275
+ {
2276
+ "epoch": 2.34,
2277
+ "learning_rate": 0.000492,
2278
+ "loss": 2.6276,
2279
+ "step": 37700
2280
+ },
2281
+ {
2282
+ "epoch": 2.34,
2283
+ "learning_rate": 0.000488,
2284
+ "loss": 2.627,
2285
+ "step": 37800
2286
+ },
2287
+ {
2288
+ "epoch": 2.35,
2289
+ "learning_rate": 0.000484,
2290
+ "loss": 2.6128,
2291
+ "step": 37900
2292
+ },
2293
+ {
2294
+ "epoch": 2.36,
2295
+ "learning_rate": 0.00048,
2296
+ "loss": 2.5874,
2297
+ "step": 38000
2298
+ },
2299
+ {
2300
+ "epoch": 2.36,
2301
+ "learning_rate": 0.00047599999999999997,
2302
+ "loss": 2.6047,
2303
+ "step": 38100
2304
+ },
2305
+ {
2306
+ "epoch": 2.37,
2307
+ "learning_rate": 0.000472,
2308
+ "loss": 2.6107,
2309
+ "step": 38200
2310
+ },
2311
+ {
2312
+ "epoch": 2.37,
2313
+ "learning_rate": 0.00046800000000000005,
2314
+ "loss": 2.6277,
2315
+ "step": 38300
2316
+ },
2317
+ {
2318
+ "epoch": 2.38,
2319
+ "learning_rate": 0.00046400000000000006,
2320
+ "loss": 2.6046,
2321
+ "step": 38400
2322
+ },
2323
+ {
2324
+ "epoch": 2.39,
2325
+ "learning_rate": 0.00046,
2326
+ "loss": 2.6021,
2327
+ "step": 38500
2328
+ },
2329
+ {
2330
+ "epoch": 2.39,
2331
+ "learning_rate": 0.000456,
2332
+ "loss": 2.5964,
2333
+ "step": 38600
2334
+ },
2335
+ {
2336
+ "epoch": 2.4,
2337
+ "learning_rate": 0.00045200000000000004,
2338
+ "loss": 2.5984,
2339
+ "step": 38700
2340
+ },
2341
+ {
2342
+ "epoch": 2.41,
2343
+ "learning_rate": 0.000448,
2344
+ "loss": 2.6017,
2345
+ "step": 38800
2346
+ },
2347
+ {
2348
+ "epoch": 2.41,
2349
+ "learning_rate": 0.000444,
2350
+ "loss": 2.5829,
2351
+ "step": 38900
2352
+ },
2353
+ {
2354
+ "epoch": 2.42,
2355
+ "learning_rate": 0.00044,
2356
+ "loss": 2.6043,
2357
+ "step": 39000
2358
+ },
2359
+ {
2360
+ "epoch": 2.42,
2361
+ "learning_rate": 0.000436,
2362
+ "loss": 2.5952,
2363
+ "step": 39100
2364
+ },
2365
+ {
2366
+ "epoch": 2.43,
2367
+ "learning_rate": 0.000432,
2368
+ "loss": 2.5914,
2369
+ "step": 39200
2370
+ },
2371
+ {
2372
+ "epoch": 2.44,
2373
+ "learning_rate": 0.000428,
2374
+ "loss": 2.5807,
2375
+ "step": 39300
2376
+ },
2377
+ {
2378
+ "epoch": 2.44,
2379
+ "learning_rate": 0.000424,
2380
+ "loss": 2.5741,
2381
+ "step": 39400
2382
+ },
2383
+ {
2384
+ "epoch": 2.45,
2385
+ "learning_rate": 0.00042,
2386
+ "loss": 2.5896,
2387
+ "step": 39500
2388
+ },
2389
+ {
2390
+ "epoch": 2.45,
2391
+ "learning_rate": 0.000416,
2392
+ "loss": 2.5787,
2393
+ "step": 39600
2394
+ },
2395
+ {
2396
+ "epoch": 2.46,
2397
+ "learning_rate": 0.000412,
2398
+ "loss": 2.5855,
2399
+ "step": 39700
2400
+ },
2401
+ {
2402
+ "epoch": 2.47,
2403
+ "learning_rate": 0.000408,
2404
+ "loss": 2.5841,
2405
+ "step": 39800
2406
+ },
2407
+ {
2408
+ "epoch": 2.47,
2409
+ "learning_rate": 0.000404,
2410
+ "loss": 2.5796,
2411
+ "step": 39900
2412
+ },
2413
+ {
2414
+ "epoch": 2.48,
2415
+ "learning_rate": 0.0004,
2416
+ "loss": 2.5741,
2417
+ "step": 40000
2418
+ },
2419
+ {
2420
+ "epoch": 2.49,
2421
+ "learning_rate": 0.00039600000000000003,
2422
+ "loss": 2.5621,
2423
+ "step": 40100
2424
+ },
2425
+ {
2426
+ "epoch": 2.49,
2427
+ "learning_rate": 0.00039200000000000004,
2428
+ "loss": 2.5595,
2429
+ "step": 40200
2430
+ },
2431
+ {
2432
+ "epoch": 2.5,
2433
+ "learning_rate": 0.000388,
2434
+ "loss": 2.5762,
2435
+ "step": 40300
2436
+ },
2437
+ {
2438
+ "epoch": 2.5,
2439
+ "learning_rate": 0.000384,
2440
+ "loss": 2.557,
2441
+ "step": 40400
2442
+ },
2443
+ {
2444
+ "epoch": 2.51,
2445
+ "learning_rate": 0.00038,
2446
+ "loss": 2.572,
2447
+ "step": 40500
2448
+ },
2449
+ {
2450
+ "epoch": 2.52,
2451
+ "learning_rate": 0.00037600000000000003,
2452
+ "loss": 2.5602,
2453
+ "step": 40600
2454
+ },
2455
+ {
2456
+ "epoch": 2.52,
2457
+ "learning_rate": 0.000372,
2458
+ "loss": 2.5448,
2459
+ "step": 40700
2460
+ },
2461
+ {
2462
+ "epoch": 2.53,
2463
+ "learning_rate": 0.000368,
2464
+ "loss": 2.5611,
2465
+ "step": 40800
2466
+ },
2467
+ {
2468
+ "epoch": 2.54,
2469
+ "learning_rate": 0.000364,
2470
+ "loss": 2.5426,
2471
+ "step": 40900
2472
+ },
2473
+ {
2474
+ "epoch": 2.54,
2475
+ "learning_rate": 0.00035999999999999997,
2476
+ "loss": 2.5544,
2477
+ "step": 41000
2478
+ },
2479
+ {
2480
+ "epoch": 2.55,
2481
+ "learning_rate": 0.000356,
2482
+ "loss": 2.5621,
2483
+ "step": 41100
2484
+ },
2485
+ {
2486
+ "epoch": 2.55,
2487
+ "learning_rate": 0.000352,
2488
+ "loss": 2.5451,
2489
+ "step": 41200
2490
+ },
2491
+ {
2492
+ "epoch": 2.56,
2493
+ "learning_rate": 0.000348,
2494
+ "loss": 2.5358,
2495
+ "step": 41300
2496
+ },
2497
+ {
2498
+ "epoch": 2.57,
2499
+ "learning_rate": 0.00034399999999999996,
2500
+ "loss": 2.5481,
2501
+ "step": 41400
2502
+ },
2503
+ {
2504
+ "epoch": 2.57,
2505
+ "learning_rate": 0.00034,
2506
+ "loss": 2.5407,
2507
+ "step": 41500
2508
+ },
2509
+ {
2510
+ "epoch": 2.58,
2511
+ "learning_rate": 0.00033600000000000004,
2512
+ "loss": 2.5451,
2513
+ "step": 41600
2514
+ },
2515
+ {
2516
+ "epoch": 2.58,
2517
+ "learning_rate": 0.00033200000000000005,
2518
+ "loss": 2.5317,
2519
+ "step": 41700
2520
+ },
2521
+ {
2522
+ "epoch": 2.59,
2523
+ "learning_rate": 0.000328,
2524
+ "loss": 2.5421,
2525
+ "step": 41800
2526
+ },
2527
+ {
2528
+ "epoch": 2.6,
2529
+ "learning_rate": 0.000324,
2530
+ "loss": 2.5286,
2531
+ "step": 41900
2532
+ },
2533
+ {
2534
+ "epoch": 2.6,
2535
+ "learning_rate": 0.00032,
2536
+ "loss": 2.5366,
2537
+ "step": 42000
2538
+ },
2539
+ {
2540
+ "epoch": 2.61,
2541
+ "learning_rate": 0.000316,
2542
+ "loss": 2.5288,
2543
+ "step": 42100
2544
+ },
2545
+ {
2546
+ "epoch": 2.62,
2547
+ "learning_rate": 0.000312,
2548
+ "loss": 2.5221,
2549
+ "step": 42200
2550
+ },
2551
+ {
2552
+ "epoch": 2.62,
2553
+ "learning_rate": 0.000308,
2554
+ "loss": 2.5278,
2555
+ "step": 42300
2556
+ },
2557
+ {
2558
+ "epoch": 2.63,
2559
+ "learning_rate": 0.000304,
2560
+ "loss": 2.545,
2561
+ "step": 42400
2562
+ },
2563
+ {
2564
+ "epoch": 2.63,
2565
+ "learning_rate": 0.0003,
2566
+ "loss": 2.5246,
2567
+ "step": 42500
2568
+ },
2569
+ {
2570
+ "epoch": 2.64,
2571
+ "learning_rate": 0.000296,
2572
+ "loss": 2.5333,
2573
+ "step": 42600
2574
+ },
2575
+ {
2576
+ "epoch": 2.65,
2577
+ "learning_rate": 0.000292,
2578
+ "loss": 2.5258,
2579
+ "step": 42700
2580
+ },
2581
+ {
2582
+ "epoch": 2.65,
2583
+ "learning_rate": 0.000288,
2584
+ "loss": 2.5334,
2585
+ "step": 42800
2586
+ },
2587
+ {
2588
+ "epoch": 2.66,
2589
+ "learning_rate": 0.00028399999999999996,
2590
+ "loss": 2.5101,
2591
+ "step": 42900
2592
+ },
2593
+ {
2594
+ "epoch": 2.67,
2595
+ "learning_rate": 0.00028000000000000003,
2596
+ "loss": 2.5168,
2597
+ "step": 43000
2598
+ },
2599
+ {
2600
+ "epoch": 2.67,
2601
+ "learning_rate": 0.00027600000000000004,
2602
+ "loss": 2.5184,
2603
+ "step": 43100
2604
+ },
2605
+ {
2606
+ "epoch": 2.68,
2607
+ "learning_rate": 0.00027200000000000005,
2608
+ "loss": 2.5236,
2609
+ "step": 43200
2610
+ },
2611
+ {
2612
+ "epoch": 2.68,
2613
+ "learning_rate": 0.000268,
2614
+ "loss": 2.5015,
2615
+ "step": 43300
2616
+ },
2617
+ {
2618
+ "epoch": 2.69,
2619
+ "learning_rate": 0.000264,
2620
+ "loss": 2.5106,
2621
+ "step": 43400
2622
+ },
2623
+ {
2624
+ "epoch": 2.7,
2625
+ "learning_rate": 0.00026000000000000003,
2626
+ "loss": 2.5095,
2627
+ "step": 43500
2628
+ },
2629
+ {
2630
+ "epoch": 2.7,
2631
+ "learning_rate": 0.000256,
2632
+ "loss": 2.5089,
2633
+ "step": 43600
2634
+ },
2635
+ {
2636
+ "epoch": 2.71,
2637
+ "learning_rate": 0.000252,
2638
+ "loss": 2.5086,
2639
+ "step": 43700
2640
+ },
2641
+ {
2642
+ "epoch": 2.72,
2643
+ "learning_rate": 0.000248,
2644
+ "loss": 2.4996,
2645
+ "step": 43800
2646
+ },
2647
+ {
2648
+ "epoch": 2.72,
2649
+ "learning_rate": 0.000244,
2650
+ "loss": 2.5202,
2651
+ "step": 43900
2652
+ },
2653
+ {
2654
+ "epoch": 2.73,
2655
+ "learning_rate": 0.00024,
2656
+ "loss": 2.5012,
2657
+ "step": 44000
2658
+ },
2659
+ {
2660
+ "epoch": 2.73,
2661
+ "learning_rate": 0.000236,
2662
+ "loss": 2.51,
2663
+ "step": 44100
2664
+ },
2665
+ {
2666
+ "epoch": 2.74,
2667
+ "learning_rate": 0.00023200000000000003,
2668
+ "loss": 2.5011,
2669
+ "step": 44200
2670
+ },
2671
+ {
2672
+ "epoch": 2.75,
2673
+ "learning_rate": 0.000228,
2674
+ "loss": 2.4794,
2675
+ "step": 44300
2676
+ },
2677
+ {
2678
+ "epoch": 2.75,
2679
+ "learning_rate": 0.000224,
2680
+ "loss": 2.4736,
2681
+ "step": 44400
2682
+ },
2683
+ {
2684
+ "epoch": 2.76,
2685
+ "learning_rate": 0.00022,
2686
+ "loss": 2.5016,
2687
+ "step": 44500
2688
+ },
2689
+ {
2690
+ "epoch": 2.76,
2691
+ "learning_rate": 0.000216,
2692
+ "loss": 2.4874,
2693
+ "step": 44600
2694
+ },
2695
+ {
2696
+ "epoch": 2.77,
2697
+ "learning_rate": 0.000212,
2698
+ "loss": 2.4932,
2699
+ "step": 44700
2700
+ },
2701
+ {
2702
+ "epoch": 2.78,
2703
+ "learning_rate": 0.000208,
2704
+ "loss": 2.5032,
2705
+ "step": 44800
2706
+ },
2707
+ {
2708
+ "epoch": 2.78,
2709
+ "learning_rate": 0.000204,
2710
+ "loss": 2.4956,
2711
+ "step": 44900
2712
+ },
2713
+ {
2714
+ "epoch": 2.79,
2715
+ "learning_rate": 0.0002,
2716
+ "loss": 2.5014,
2717
+ "step": 45000
2718
+ },
2719
+ {
2720
+ "epoch": 2.8,
2721
+ "learning_rate": 0.00019600000000000002,
2722
+ "loss": 2.4986,
2723
+ "step": 45100
2724
+ },
2725
+ {
2726
+ "epoch": 2.8,
2727
+ "learning_rate": 0.000192,
2728
+ "loss": 2.4685,
2729
+ "step": 45200
2730
+ },
2731
+ {
2732
+ "epoch": 2.81,
2733
+ "learning_rate": 0.00018800000000000002,
2734
+ "loss": 2.4901,
2735
+ "step": 45300
2736
+ },
2737
+ {
2738
+ "epoch": 2.81,
2739
+ "learning_rate": 0.000184,
2740
+ "loss": 2.4783,
2741
+ "step": 45400
2742
+ },
2743
+ {
2744
+ "epoch": 2.82,
2745
+ "learning_rate": 0.00017999999999999998,
2746
+ "loss": 2.4947,
2747
+ "step": 45500
2748
+ },
2749
+ {
2750
+ "epoch": 2.83,
2751
+ "learning_rate": 0.000176,
2752
+ "loss": 2.4696,
2753
+ "step": 45600
2754
+ },
2755
+ {
2756
+ "epoch": 2.83,
2757
+ "learning_rate": 0.00017199999999999998,
2758
+ "loss": 2.4728,
2759
+ "step": 45700
2760
+ },
2761
+ {
2762
+ "epoch": 2.84,
2763
+ "learning_rate": 0.00016800000000000002,
2764
+ "loss": 2.4791,
2765
+ "step": 45800
2766
+ },
2767
+ {
2768
+ "epoch": 2.85,
2769
+ "learning_rate": 0.000164,
2770
+ "loss": 2.4764,
2771
+ "step": 45900
2772
+ },
2773
+ {
2774
+ "epoch": 2.85,
2775
+ "learning_rate": 0.00016,
2776
+ "loss": 2.4861,
2777
+ "step": 46000
2778
+ },
2779
+ {
2780
+ "epoch": 2.86,
2781
+ "learning_rate": 0.000156,
2782
+ "loss": 2.4619,
2783
+ "step": 46100
2784
+ },
2785
+ {
2786
+ "epoch": 2.86,
2787
+ "learning_rate": 0.000152,
2788
+ "loss": 2.4698,
2789
+ "step": 46200
2790
+ },
2791
+ {
2792
+ "epoch": 2.87,
2793
+ "learning_rate": 0.000148,
2794
+ "loss": 2.489,
2795
+ "step": 46300
2796
+ },
2797
+ {
2798
+ "epoch": 2.88,
2799
+ "learning_rate": 0.000144,
2800
+ "loss": 2.4607,
2801
+ "step": 46400
2802
+ },
2803
+ {
2804
+ "epoch": 2.88,
2805
+ "learning_rate": 0.00014000000000000001,
2806
+ "loss": 2.4727,
2807
+ "step": 46500
2808
+ },
2809
+ {
2810
+ "epoch": 2.89,
2811
+ "learning_rate": 0.00013600000000000003,
2812
+ "loss": 2.4548,
2813
+ "step": 46600
2814
+ },
2815
+ {
2816
+ "epoch": 2.89,
2817
+ "learning_rate": 0.000132,
2818
+ "loss": 2.4553,
2819
+ "step": 46700
2820
+ },
2821
+ {
2822
+ "epoch": 2.9,
2823
+ "learning_rate": 0.000128,
2824
+ "loss": 2.465,
2825
+ "step": 46800
2826
+ },
2827
+ {
2828
+ "epoch": 2.91,
2829
+ "learning_rate": 0.000124,
2830
+ "loss": 2.4601,
2831
+ "step": 46900
2832
+ },
2833
+ {
2834
+ "epoch": 2.91,
2835
+ "learning_rate": 0.00012,
2836
+ "loss": 2.4554,
2837
+ "step": 47000
2838
+ },
2839
+ {
2840
+ "epoch": 2.92,
2841
+ "learning_rate": 0.00011600000000000001,
2842
+ "loss": 2.4575,
2843
+ "step": 47100
2844
+ },
2845
+ {
2846
+ "epoch": 2.93,
2847
+ "learning_rate": 0.000112,
2848
+ "loss": 2.4515,
2849
+ "step": 47200
2850
+ },
2851
+ {
2852
+ "epoch": 2.93,
2853
+ "learning_rate": 0.000108,
2854
+ "loss": 2.4613,
2855
+ "step": 47300
2856
+ },
2857
+ {
2858
+ "epoch": 2.94,
2859
+ "learning_rate": 0.000104,
2860
+ "loss": 2.4583,
2861
+ "step": 47400
2862
+ },
2863
+ {
2864
+ "epoch": 2.94,
2865
+ "learning_rate": 0.0001,
2866
+ "loss": 2.4455,
2867
+ "step": 47500
2868
+ },
2869
+ {
2870
+ "epoch": 2.95,
2871
+ "learning_rate": 9.6e-05,
2872
+ "loss": 2.4446,
2873
+ "step": 47600
2874
+ },
2875
+ {
2876
+ "epoch": 2.96,
2877
+ "learning_rate": 9.2e-05,
2878
+ "loss": 2.4488,
2879
+ "step": 47700
2880
+ },
2881
+ {
2882
+ "epoch": 2.96,
2883
+ "learning_rate": 8.8e-05,
2884
+ "loss": 2.4454,
2885
+ "step": 47800
2886
+ },
2887
+ {
2888
+ "epoch": 2.97,
2889
+ "learning_rate": 8.400000000000001e-05,
2890
+ "loss": 2.4445,
2891
+ "step": 47900
2892
+ },
2893
+ {
2894
+ "epoch": 2.98,
2895
+ "learning_rate": 8e-05,
2896
+ "loss": 2.4388,
2897
+ "step": 48000
2898
+ },
2899
+ {
2900
+ "epoch": 2.98,
2901
+ "learning_rate": 7.6e-05,
2902
+ "loss": 2.4506,
2903
+ "step": 48100
2904
+ },
2905
+ {
2906
+ "epoch": 2.99,
2907
+ "learning_rate": 7.2e-05,
2908
+ "loss": 2.4392,
2909
+ "step": 48200
2910
+ },
2911
+ {
2912
+ "epoch": 2.99,
2913
+ "learning_rate": 6.800000000000001e-05,
2914
+ "loss": 2.4408,
2915
+ "step": 48300
2916
+ },
2917
+ {
2918
+ "epoch": 3.0,
2919
+ "learning_rate": 6.4e-05,
2920
+ "loss": 2.4391,
2921
+ "step": 48400
2922
+ },
2923
+ {
2924
+ "epoch": 3.01,
2925
+ "learning_rate": 6e-05,
2926
+ "loss": 2.4281,
2927
+ "step": 48500
2928
+ },
2929
+ {
2930
+ "epoch": 3.01,
2931
+ "learning_rate": 5.6e-05,
2932
+ "loss": 2.4451,
2933
+ "step": 48600
2934
+ },
2935
+ {
2936
+ "epoch": 3.02,
2937
+ "learning_rate": 5.2e-05,
2938
+ "loss": 2.4277,
2939
+ "step": 48700
2940
+ },
2941
+ {
2942
+ "epoch": 3.02,
2943
+ "learning_rate": 4.8e-05,
2944
+ "loss": 2.4388,
2945
+ "step": 48800
2946
+ },
2947
+ {
2948
+ "epoch": 3.03,
2949
+ "learning_rate": 4.4e-05,
2950
+ "loss": 2.4368,
2951
+ "step": 48900
2952
+ },
2953
+ {
2954
+ "epoch": 3.04,
2955
+ "learning_rate": 4e-05,
2956
+ "loss": 2.423,
2957
+ "step": 49000
2958
+ },
2959
+ {
2960
+ "epoch": 3.04,
2961
+ "learning_rate": 3.6e-05,
2962
+ "loss": 2.4458,
2963
+ "step": 49100
2964
+ },
2965
+ {
2966
+ "epoch": 3.05,
2967
+ "learning_rate": 3.2e-05,
2968
+ "loss": 2.4218,
2969
+ "step": 49200
2970
+ },
2971
+ {
2972
+ "epoch": 3.06,
2973
+ "learning_rate": 2.8e-05,
2974
+ "loss": 2.4226,
2975
+ "step": 49300
2976
+ },
2977
+ {
2978
+ "epoch": 3.06,
2979
+ "learning_rate": 2.4e-05,
2980
+ "loss": 2.4287,
2981
+ "step": 49400
2982
+ },
2983
+ {
2984
+ "epoch": 3.07,
2985
+ "learning_rate": 2e-05,
2986
+ "loss": 2.4237,
2987
+ "step": 49500
2988
+ },
2989
+ {
2990
+ "epoch": 3.07,
2991
+ "learning_rate": 1.6e-05,
2992
+ "loss": 2.421,
2993
+ "step": 49600
2994
+ },
2995
+ {
2996
+ "epoch": 3.08,
2997
+ "learning_rate": 1.2e-05,
2998
+ "loss": 2.4214,
2999
+ "step": 49700
3000
+ },
3001
+ {
3002
+ "epoch": 3.09,
3003
+ "learning_rate": 8e-06,
3004
+ "loss": 2.413,
3005
+ "step": 49800
3006
+ },
3007
+ {
3008
+ "epoch": 3.09,
3009
+ "learning_rate": 4e-06,
3010
+ "loss": 2.4309,
3011
+ "step": 49900
3012
+ },
3013
+ {
3014
+ "epoch": 3.1,
3015
+ "learning_rate": 0.0,
3016
+ "loss": 2.4233,
3017
+ "step": 50000
3018
+ },
3019
+ {
3020
+ "epoch": 3.1,
3021
+ "eval_accuracy": 0.5611543501568519,
3022
+ "eval_loss": 2.2921805381774902,
3023
+ "eval_runtime": 6159.6175,
3024
+ "eval_samples_per_second": 35.288,
3025
+ "eval_steps_per_second": 2.206,
3026
+ "step": 50000
3027
+ },
3028
+ {
3029
+ "epoch": 3.1,
3030
+ "step": 50000,
3031
+ "total_flos": 7.554075721728e+17,
3032
+ "train_loss": 3.1285950134277343,
3033
+ "train_runtime": 151231.41,
3034
+ "train_samples_per_second": 84.639,
3035
+ "train_steps_per_second": 0.331
3036
+ }
3037
+ ],
3038
+ "max_steps": 50000,
3039
+ "num_train_epochs": 4,
3040
+ "total_flos": 7.554075721728e+17,
3041
+ "trial_name": null,
3042
+ "trial_params": null
3043
+ }