luckyrabbits commited on
Commit
7de4c8f
·
verified ·
1 Parent(s): ec73c42

Upload trainer_state.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. trainer_state.json +3543 -0
trainer_state.json ADDED
@@ -0,0 +1,3543 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.11865211200759374,
6
+ "eval_steps": 500,
7
+ "global_step": 500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.00023730422401518748,
14
+ "grad_norm": 0.19874855875968933,
15
+ "learning_rate": 9.980000000000001e-06,
16
+ "loss": 2.3055,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.00047460844803037496,
21
+ "grad_norm": 0.3101324141025543,
22
+ "learning_rate": 9.960000000000001e-06,
23
+ "loss": 0.8548,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.0007119126720455624,
28
+ "grad_norm": 0.18851338326931,
29
+ "learning_rate": 9.940000000000001e-06,
30
+ "loss": 1.3783,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.0009492168960607499,
35
+ "grad_norm": 0.5474942326545715,
36
+ "learning_rate": 9.920000000000002e-06,
37
+ "loss": 2.5833,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.0011865211200759373,
42
+ "grad_norm": 0.28883224725723267,
43
+ "learning_rate": 9.9e-06,
44
+ "loss": 2.4045,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.0014238253440911248,
49
+ "grad_norm": 0.4798765778541565,
50
+ "learning_rate": 9.88e-06,
51
+ "loss": 1.2589,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.0016611295681063123,
56
+ "grad_norm": 0.48112648725509644,
57
+ "learning_rate": 9.86e-06,
58
+ "loss": 2.9112,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.0018984337921214998,
63
+ "grad_norm": 0.8760956525802612,
64
+ "learning_rate": 9.84e-06,
65
+ "loss": 2.3175,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.0021357380161366873,
70
+ "grad_norm": 0.44891074299812317,
71
+ "learning_rate": 9.820000000000001e-06,
72
+ "loss": 1.3196,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.0023730422401518746,
77
+ "grad_norm": 0.3395187556743622,
78
+ "learning_rate": 9.800000000000001e-06,
79
+ "loss": 1.588,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.0026103464641670624,
84
+ "grad_norm": 0.587505578994751,
85
+ "learning_rate": 9.780000000000001e-06,
86
+ "loss": 1.2509,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.0028476506881822496,
91
+ "grad_norm": 0.4817255437374115,
92
+ "learning_rate": 9.760000000000001e-06,
93
+ "loss": 1.3541,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.003084954912197437,
98
+ "grad_norm": 0.312285840511322,
99
+ "learning_rate": 9.74e-06,
100
+ "loss": 1.8525,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.0033222591362126247,
105
+ "grad_norm": 0.7621486783027649,
106
+ "learning_rate": 9.72e-06,
107
+ "loss": 1.3197,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.003559563360227812,
112
+ "grad_norm": 0.5916957259178162,
113
+ "learning_rate": 9.7e-06,
114
+ "loss": 1.0805,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 0.0037968675842429997,
119
+ "grad_norm": 0.7447299361228943,
120
+ "learning_rate": 9.68e-06,
121
+ "loss": 2.5366,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 0.004034171808258187,
126
+ "grad_norm": 0.38069915771484375,
127
+ "learning_rate": 9.66e-06,
128
+ "loss": 1.6038,
129
+ "step": 17
130
+ },
131
+ {
132
+ "epoch": 0.004271476032273375,
133
+ "grad_norm": 0.4375569224357605,
134
+ "learning_rate": 9.640000000000001e-06,
135
+ "loss": 2.8857,
136
+ "step": 18
137
+ },
138
+ {
139
+ "epoch": 0.004508780256288562,
140
+ "grad_norm": 0.2266787886619568,
141
+ "learning_rate": 9.620000000000001e-06,
142
+ "loss": 2.045,
143
+ "step": 19
144
+ },
145
+ {
146
+ "epoch": 0.004746084480303749,
147
+ "grad_norm": 0.3714630901813507,
148
+ "learning_rate": 9.600000000000001e-06,
149
+ "loss": 1.6152,
150
+ "step": 20
151
+ },
152
+ {
153
+ "epoch": 0.0049833887043189366,
154
+ "grad_norm": 0.45660167932510376,
155
+ "learning_rate": 9.58e-06,
156
+ "loss": 2.1712,
157
+ "step": 21
158
+ },
159
+ {
160
+ "epoch": 0.005220692928334125,
161
+ "grad_norm": 0.3804182708263397,
162
+ "learning_rate": 9.56e-06,
163
+ "loss": 1.7326,
164
+ "step": 22
165
+ },
166
+ {
167
+ "epoch": 0.005457997152349312,
168
+ "grad_norm": 0.5945218205451965,
169
+ "learning_rate": 9.54e-06,
170
+ "loss": 2.8499,
171
+ "step": 23
172
+ },
173
+ {
174
+ "epoch": 0.005695301376364499,
175
+ "grad_norm": 0.49986812472343445,
176
+ "learning_rate": 9.52e-06,
177
+ "loss": 1.887,
178
+ "step": 24
179
+ },
180
+ {
181
+ "epoch": 0.005932605600379687,
182
+ "grad_norm": 0.41545894742012024,
183
+ "learning_rate": 9.5e-06,
184
+ "loss": 1.0656,
185
+ "step": 25
186
+ },
187
+ {
188
+ "epoch": 0.006169909824394874,
189
+ "grad_norm": 0.6399343609809875,
190
+ "learning_rate": 9.48e-06,
191
+ "loss": 1.8672,
192
+ "step": 26
193
+ },
194
+ {
195
+ "epoch": 0.006407214048410062,
196
+ "grad_norm": 0.3765011727809906,
197
+ "learning_rate": 9.460000000000001e-06,
198
+ "loss": 0.9629,
199
+ "step": 27
200
+ },
201
+ {
202
+ "epoch": 0.006644518272425249,
203
+ "grad_norm": 0.48257043957710266,
204
+ "learning_rate": 9.440000000000001e-06,
205
+ "loss": 1.9923,
206
+ "step": 28
207
+ },
208
+ {
209
+ "epoch": 0.006881822496440437,
210
+ "grad_norm": 0.5844452977180481,
211
+ "learning_rate": 9.42e-06,
212
+ "loss": 2.0896,
213
+ "step": 29
214
+ },
215
+ {
216
+ "epoch": 0.007119126720455624,
217
+ "grad_norm": 0.6092790961265564,
218
+ "learning_rate": 9.4e-06,
219
+ "loss": 2.4446,
220
+ "step": 30
221
+ },
222
+ {
223
+ "epoch": 0.007356430944470811,
224
+ "grad_norm": 0.6862596869468689,
225
+ "learning_rate": 9.38e-06,
226
+ "loss": 1.9678,
227
+ "step": 31
228
+ },
229
+ {
230
+ "epoch": 0.007593735168485999,
231
+ "grad_norm": 0.5207828879356384,
232
+ "learning_rate": 9.360000000000002e-06,
233
+ "loss": 1.611,
234
+ "step": 32
235
+ },
236
+ {
237
+ "epoch": 0.007831039392501186,
238
+ "grad_norm": 0.5294057130813599,
239
+ "learning_rate": 9.340000000000002e-06,
240
+ "loss": 1.8447,
241
+ "step": 33
242
+ },
243
+ {
244
+ "epoch": 0.008068343616516375,
245
+ "grad_norm": 0.4702499210834503,
246
+ "learning_rate": 9.32e-06,
247
+ "loss": 2.035,
248
+ "step": 34
249
+ },
250
+ {
251
+ "epoch": 0.008305647840531562,
252
+ "grad_norm": 0.5244491696357727,
253
+ "learning_rate": 9.3e-06,
254
+ "loss": 1.4395,
255
+ "step": 35
256
+ },
257
+ {
258
+ "epoch": 0.00854295206454675,
259
+ "grad_norm": 0.644615650177002,
260
+ "learning_rate": 9.280000000000001e-06,
261
+ "loss": 1.7666,
262
+ "step": 36
263
+ },
264
+ {
265
+ "epoch": 0.008780256288561937,
266
+ "grad_norm": 0.5552751421928406,
267
+ "learning_rate": 9.260000000000001e-06,
268
+ "loss": 1.5753,
269
+ "step": 37
270
+ },
271
+ {
272
+ "epoch": 0.009017560512577124,
273
+ "grad_norm": 0.2882249057292938,
274
+ "learning_rate": 9.240000000000001e-06,
275
+ "loss": 1.5176,
276
+ "step": 38
277
+ },
278
+ {
279
+ "epoch": 0.009254864736592311,
280
+ "grad_norm": 0.38219153881073,
281
+ "learning_rate": 9.220000000000002e-06,
282
+ "loss": 1.7808,
283
+ "step": 39
284
+ },
285
+ {
286
+ "epoch": 0.009492168960607499,
287
+ "grad_norm": 0.6532744765281677,
288
+ "learning_rate": 9.200000000000002e-06,
289
+ "loss": 1.8763,
290
+ "step": 40
291
+ },
292
+ {
293
+ "epoch": 0.009729473184622686,
294
+ "grad_norm": 0.7350605726242065,
295
+ "learning_rate": 9.180000000000002e-06,
296
+ "loss": 2.5947,
297
+ "step": 41
298
+ },
299
+ {
300
+ "epoch": 0.009966777408637873,
301
+ "grad_norm": 0.44435158371925354,
302
+ "learning_rate": 9.16e-06,
303
+ "loss": 2.4829,
304
+ "step": 42
305
+ },
306
+ {
307
+ "epoch": 0.01020408163265306,
308
+ "grad_norm": 0.3937893509864807,
309
+ "learning_rate": 9.14e-06,
310
+ "loss": 2.106,
311
+ "step": 43
312
+ },
313
+ {
314
+ "epoch": 0.01044138585666825,
315
+ "grad_norm": 0.8209621906280518,
316
+ "learning_rate": 9.12e-06,
317
+ "loss": 1.8711,
318
+ "step": 44
319
+ },
320
+ {
321
+ "epoch": 0.010678690080683437,
322
+ "grad_norm": 1.220982313156128,
323
+ "learning_rate": 9.100000000000001e-06,
324
+ "loss": 2.4524,
325
+ "step": 45
326
+ },
327
+ {
328
+ "epoch": 0.010915994304698624,
329
+ "grad_norm": 0.4203420877456665,
330
+ "learning_rate": 9.080000000000001e-06,
331
+ "loss": 1.775,
332
+ "step": 46
333
+ },
334
+ {
335
+ "epoch": 0.011153298528713811,
336
+ "grad_norm": 0.28424403071403503,
337
+ "learning_rate": 9.060000000000001e-06,
338
+ "loss": 1.7594,
339
+ "step": 47
340
+ },
341
+ {
342
+ "epoch": 0.011390602752728999,
343
+ "grad_norm": 0.7268348336219788,
344
+ "learning_rate": 9.040000000000002e-06,
345
+ "loss": 0.9244,
346
+ "step": 48
347
+ },
348
+ {
349
+ "epoch": 0.011627906976744186,
350
+ "grad_norm": 0.4843103885650635,
351
+ "learning_rate": 9.020000000000002e-06,
352
+ "loss": 1.9955,
353
+ "step": 49
354
+ },
355
+ {
356
+ "epoch": 0.011865211200759373,
357
+ "grad_norm": 0.3405970335006714,
358
+ "learning_rate": 9e-06,
359
+ "loss": 1.6384,
360
+ "step": 50
361
+ },
362
+ {
363
+ "epoch": 0.01210251542477456,
364
+ "grad_norm": 1.1265524625778198,
365
+ "learning_rate": 8.98e-06,
366
+ "loss": 1.7042,
367
+ "step": 51
368
+ },
369
+ {
370
+ "epoch": 0.012339819648789748,
371
+ "grad_norm": 0.13442721962928772,
372
+ "learning_rate": 8.96e-06,
373
+ "loss": 1.3843,
374
+ "step": 52
375
+ },
376
+ {
377
+ "epoch": 0.012577123872804937,
378
+ "grad_norm": 0.4972473382949829,
379
+ "learning_rate": 8.94e-06,
380
+ "loss": 0.8437,
381
+ "step": 53
382
+ },
383
+ {
384
+ "epoch": 0.012814428096820124,
385
+ "grad_norm": 0.28667712211608887,
386
+ "learning_rate": 8.920000000000001e-06,
387
+ "loss": 0.9238,
388
+ "step": 54
389
+ },
390
+ {
391
+ "epoch": 0.013051732320835311,
392
+ "grad_norm": 0.2998306155204773,
393
+ "learning_rate": 8.900000000000001e-06,
394
+ "loss": 2.2332,
395
+ "step": 55
396
+ },
397
+ {
398
+ "epoch": 0.013289036544850499,
399
+ "grad_norm": 0.6133009195327759,
400
+ "learning_rate": 8.880000000000001e-06,
401
+ "loss": 2.2289,
402
+ "step": 56
403
+ },
404
+ {
405
+ "epoch": 0.013526340768865686,
406
+ "grad_norm": 0.48921748995780945,
407
+ "learning_rate": 8.860000000000002e-06,
408
+ "loss": 0.6597,
409
+ "step": 57
410
+ },
411
+ {
412
+ "epoch": 0.013763644992880873,
413
+ "grad_norm": 0.8161422610282898,
414
+ "learning_rate": 8.84e-06,
415
+ "loss": 1.3381,
416
+ "step": 58
417
+ },
418
+ {
419
+ "epoch": 0.01400094921689606,
420
+ "grad_norm": 0.4998335540294647,
421
+ "learning_rate": 8.82e-06,
422
+ "loss": 1.5505,
423
+ "step": 59
424
+ },
425
+ {
426
+ "epoch": 0.014238253440911248,
427
+ "grad_norm": 0.8967633843421936,
428
+ "learning_rate": 8.8e-06,
429
+ "loss": 1.2807,
430
+ "step": 60
431
+ },
432
+ {
433
+ "epoch": 0.014475557664926435,
434
+ "grad_norm": 0.8106015920639038,
435
+ "learning_rate": 8.78e-06,
436
+ "loss": 1.2792,
437
+ "step": 61
438
+ },
439
+ {
440
+ "epoch": 0.014712861888941622,
441
+ "grad_norm": 0.6022857427597046,
442
+ "learning_rate": 8.76e-06,
443
+ "loss": 2.023,
444
+ "step": 62
445
+ },
446
+ {
447
+ "epoch": 0.014950166112956811,
448
+ "grad_norm": 0.6330555081367493,
449
+ "learning_rate": 8.740000000000001e-06,
450
+ "loss": 0.931,
451
+ "step": 63
452
+ },
453
+ {
454
+ "epoch": 0.015187470336971999,
455
+ "grad_norm": 0.6975427269935608,
456
+ "learning_rate": 8.720000000000001e-06,
457
+ "loss": 1.5096,
458
+ "step": 64
459
+ },
460
+ {
461
+ "epoch": 0.015424774560987186,
462
+ "grad_norm": 0.9109779596328735,
463
+ "learning_rate": 8.700000000000001e-06,
464
+ "loss": 1.4295,
465
+ "step": 65
466
+ },
467
+ {
468
+ "epoch": 0.01566207878500237,
469
+ "grad_norm": 0.5679107904434204,
470
+ "learning_rate": 8.68e-06,
471
+ "loss": 1.6368,
472
+ "step": 66
473
+ },
474
+ {
475
+ "epoch": 0.01589938300901756,
476
+ "grad_norm": 0.4659746587276459,
477
+ "learning_rate": 8.66e-06,
478
+ "loss": 1.292,
479
+ "step": 67
480
+ },
481
+ {
482
+ "epoch": 0.01613668723303275,
483
+ "grad_norm": 0.6395153403282166,
484
+ "learning_rate": 8.64e-06,
485
+ "loss": 2.0611,
486
+ "step": 68
487
+ },
488
+ {
489
+ "epoch": 0.016373991457047935,
490
+ "grad_norm": 0.37920185923576355,
491
+ "learning_rate": 8.62e-06,
492
+ "loss": 1.1706,
493
+ "step": 69
494
+ },
495
+ {
496
+ "epoch": 0.016611295681063124,
497
+ "grad_norm": 0.2950354218482971,
498
+ "learning_rate": 8.6e-06,
499
+ "loss": 1.4619,
500
+ "step": 70
501
+ },
502
+ {
503
+ "epoch": 0.01684859990507831,
504
+ "grad_norm": 0.664068877696991,
505
+ "learning_rate": 8.580000000000001e-06,
506
+ "loss": 2.3014,
507
+ "step": 71
508
+ },
509
+ {
510
+ "epoch": 0.0170859041290935,
511
+ "grad_norm": 0.4015841782093048,
512
+ "learning_rate": 8.560000000000001e-06,
513
+ "loss": 2.6305,
514
+ "step": 72
515
+ },
516
+ {
517
+ "epoch": 0.017323208353108684,
518
+ "grad_norm": 0.8365151882171631,
519
+ "learning_rate": 8.540000000000001e-06,
520
+ "loss": 2.4402,
521
+ "step": 73
522
+ },
523
+ {
524
+ "epoch": 0.017560512577123873,
525
+ "grad_norm": 0.4587593376636505,
526
+ "learning_rate": 8.52e-06,
527
+ "loss": 1.1948,
528
+ "step": 74
529
+ },
530
+ {
531
+ "epoch": 0.01779781680113906,
532
+ "grad_norm": 0.5275629162788391,
533
+ "learning_rate": 8.5e-06,
534
+ "loss": 1.2889,
535
+ "step": 75
536
+ },
537
+ {
538
+ "epoch": 0.018035121025154248,
539
+ "grad_norm": 0.5322698354721069,
540
+ "learning_rate": 8.48e-06,
541
+ "loss": 2.5775,
542
+ "step": 76
543
+ },
544
+ {
545
+ "epoch": 0.018272425249169437,
546
+ "grad_norm": 0.7846812009811401,
547
+ "learning_rate": 8.46e-06,
548
+ "loss": 1.2905,
549
+ "step": 77
550
+ },
551
+ {
552
+ "epoch": 0.018509729473184623,
553
+ "grad_norm": 0.4759507179260254,
554
+ "learning_rate": 8.44e-06,
555
+ "loss": 1.1162,
556
+ "step": 78
557
+ },
558
+ {
559
+ "epoch": 0.01874703369719981,
560
+ "grad_norm": 0.596358597278595,
561
+ "learning_rate": 8.42e-06,
562
+ "loss": 1.6618,
563
+ "step": 79
564
+ },
565
+ {
566
+ "epoch": 0.018984337921214997,
567
+ "grad_norm": 0.5133060812950134,
568
+ "learning_rate": 8.400000000000001e-06,
569
+ "loss": 1.9102,
570
+ "step": 80
571
+ },
572
+ {
573
+ "epoch": 0.019221642145230186,
574
+ "grad_norm": 0.8989421129226685,
575
+ "learning_rate": 8.380000000000001e-06,
576
+ "loss": 2.1852,
577
+ "step": 81
578
+ },
579
+ {
580
+ "epoch": 0.01945894636924537,
581
+ "grad_norm": 0.4702143669128418,
582
+ "learning_rate": 8.36e-06,
583
+ "loss": 1.2668,
584
+ "step": 82
585
+ },
586
+ {
587
+ "epoch": 0.01969625059326056,
588
+ "grad_norm": 1.2159205675125122,
589
+ "learning_rate": 8.34e-06,
590
+ "loss": 1.1101,
591
+ "step": 83
592
+ },
593
+ {
594
+ "epoch": 0.019933554817275746,
595
+ "grad_norm": 0.5116935968399048,
596
+ "learning_rate": 8.32e-06,
597
+ "loss": 1.222,
598
+ "step": 84
599
+ },
600
+ {
601
+ "epoch": 0.020170859041290935,
602
+ "grad_norm": 1.3710129261016846,
603
+ "learning_rate": 8.3e-06,
604
+ "loss": 3.1337,
605
+ "step": 85
606
+ },
607
+ {
608
+ "epoch": 0.02040816326530612,
609
+ "grad_norm": 0.7467148303985596,
610
+ "learning_rate": 8.28e-06,
611
+ "loss": 2.1887,
612
+ "step": 86
613
+ },
614
+ {
615
+ "epoch": 0.02064546748932131,
616
+ "grad_norm": 0.6403272151947021,
617
+ "learning_rate": 8.26e-06,
618
+ "loss": 1.2013,
619
+ "step": 87
620
+ },
621
+ {
622
+ "epoch": 0.0208827717133365,
623
+ "grad_norm": 0.4310401678085327,
624
+ "learning_rate": 8.24e-06,
625
+ "loss": 1.4044,
626
+ "step": 88
627
+ },
628
+ {
629
+ "epoch": 0.021120075937351684,
630
+ "grad_norm": 0.6710259318351746,
631
+ "learning_rate": 8.220000000000001e-06,
632
+ "loss": 2.5627,
633
+ "step": 89
634
+ },
635
+ {
636
+ "epoch": 0.021357380161366873,
637
+ "grad_norm": 0.5828210115432739,
638
+ "learning_rate": 8.2e-06,
639
+ "loss": 1.7674,
640
+ "step": 90
641
+ },
642
+ {
643
+ "epoch": 0.02159468438538206,
644
+ "grad_norm": 0.5417571663856506,
645
+ "learning_rate": 8.18e-06,
646
+ "loss": 2.2125,
647
+ "step": 91
648
+ },
649
+ {
650
+ "epoch": 0.021831988609397248,
651
+ "grad_norm": 0.5095130205154419,
652
+ "learning_rate": 8.16e-06,
653
+ "loss": 3.1429,
654
+ "step": 92
655
+ },
656
+ {
657
+ "epoch": 0.022069292833412434,
658
+ "grad_norm": 0.2329273670911789,
659
+ "learning_rate": 8.14e-06,
660
+ "loss": 1.84,
661
+ "step": 93
662
+ },
663
+ {
664
+ "epoch": 0.022306597057427623,
665
+ "grad_norm": 0.4823262691497803,
666
+ "learning_rate": 8.120000000000002e-06,
667
+ "loss": 1.7104,
668
+ "step": 94
669
+ },
670
+ {
671
+ "epoch": 0.022543901281442808,
672
+ "grad_norm": 0.6276722550392151,
673
+ "learning_rate": 8.1e-06,
674
+ "loss": 0.7729,
675
+ "step": 95
676
+ },
677
+ {
678
+ "epoch": 0.022781205505457997,
679
+ "grad_norm": 0.642092764377594,
680
+ "learning_rate": 8.08e-06,
681
+ "loss": 2.1965,
682
+ "step": 96
683
+ },
684
+ {
685
+ "epoch": 0.023018509729473186,
686
+ "grad_norm": 0.34217798709869385,
687
+ "learning_rate": 8.06e-06,
688
+ "loss": 0.9981,
689
+ "step": 97
690
+ },
691
+ {
692
+ "epoch": 0.023255813953488372,
693
+ "grad_norm": 0.47839802503585815,
694
+ "learning_rate": 8.040000000000001e-06,
695
+ "loss": 1.594,
696
+ "step": 98
697
+ },
698
+ {
699
+ "epoch": 0.02349311817750356,
700
+ "grad_norm": 1.0722686052322388,
701
+ "learning_rate": 8.020000000000001e-06,
702
+ "loss": 2.0473,
703
+ "step": 99
704
+ },
705
+ {
706
+ "epoch": 0.023730422401518746,
707
+ "grad_norm": 0.7368118166923523,
708
+ "learning_rate": 8.000000000000001e-06,
709
+ "loss": 2.4467,
710
+ "step": 100
711
+ },
712
+ {
713
+ "epoch": 0.023967726625533935,
714
+ "grad_norm": 1.064618706703186,
715
+ "learning_rate": 7.980000000000002e-06,
716
+ "loss": 1.9095,
717
+ "step": 101
718
+ },
719
+ {
720
+ "epoch": 0.02420503084954912,
721
+ "grad_norm": 0.43928399682044983,
722
+ "learning_rate": 7.960000000000002e-06,
723
+ "loss": 1.7245,
724
+ "step": 102
725
+ },
726
+ {
727
+ "epoch": 0.02444233507356431,
728
+ "grad_norm": 0.6588628888130188,
729
+ "learning_rate": 7.94e-06,
730
+ "loss": 0.8569,
731
+ "step": 103
732
+ },
733
+ {
734
+ "epoch": 0.024679639297579496,
735
+ "grad_norm": 0.5403575301170349,
736
+ "learning_rate": 7.92e-06,
737
+ "loss": 1.2521,
738
+ "step": 104
739
+ },
740
+ {
741
+ "epoch": 0.024916943521594685,
742
+ "grad_norm": 0.6686379313468933,
743
+ "learning_rate": 7.9e-06,
744
+ "loss": 1.6869,
745
+ "step": 105
746
+ },
747
+ {
748
+ "epoch": 0.025154247745609874,
749
+ "grad_norm": 0.3803173005580902,
750
+ "learning_rate": 7.88e-06,
751
+ "loss": 2.1051,
752
+ "step": 106
753
+ },
754
+ {
755
+ "epoch": 0.02539155196962506,
756
+ "grad_norm": 0.12461218237876892,
757
+ "learning_rate": 7.860000000000001e-06,
758
+ "loss": 1.5027,
759
+ "step": 107
760
+ },
761
+ {
762
+ "epoch": 0.025628856193640248,
763
+ "grad_norm": 0.47309207916259766,
764
+ "learning_rate": 7.840000000000001e-06,
765
+ "loss": 1.156,
766
+ "step": 108
767
+ },
768
+ {
769
+ "epoch": 0.025866160417655434,
770
+ "grad_norm": 0.5996202826499939,
771
+ "learning_rate": 7.820000000000001e-06,
772
+ "loss": 1.8978,
773
+ "step": 109
774
+ },
775
+ {
776
+ "epoch": 0.026103464641670623,
777
+ "grad_norm": 1.3080424070358276,
778
+ "learning_rate": 7.800000000000002e-06,
779
+ "loss": 1.2111,
780
+ "step": 110
781
+ },
782
+ {
783
+ "epoch": 0.02634076886568581,
784
+ "grad_norm": 0.5753285884857178,
785
+ "learning_rate": 7.78e-06,
786
+ "loss": 1.2555,
787
+ "step": 111
788
+ },
789
+ {
790
+ "epoch": 0.026578073089700997,
791
+ "grad_norm": 0.4644084870815277,
792
+ "learning_rate": 7.76e-06,
793
+ "loss": 0.5773,
794
+ "step": 112
795
+ },
796
+ {
797
+ "epoch": 0.026815377313716183,
798
+ "grad_norm": 0.7260199189186096,
799
+ "learning_rate": 7.74e-06,
800
+ "loss": 1.3806,
801
+ "step": 113
802
+ },
803
+ {
804
+ "epoch": 0.027052681537731372,
805
+ "grad_norm": 3.247457265853882,
806
+ "learning_rate": 7.72e-06,
807
+ "loss": 3.6028,
808
+ "step": 114
809
+ },
810
+ {
811
+ "epoch": 0.027289985761746557,
812
+ "grad_norm": 0.6819869875907898,
813
+ "learning_rate": 7.7e-06,
814
+ "loss": 1.3892,
815
+ "step": 115
816
+ },
817
+ {
818
+ "epoch": 0.027527289985761746,
819
+ "grad_norm": 0.5939836502075195,
820
+ "learning_rate": 7.680000000000001e-06,
821
+ "loss": 0.8986,
822
+ "step": 116
823
+ },
824
+ {
825
+ "epoch": 0.027764594209776935,
826
+ "grad_norm": 0.9432902336120605,
827
+ "learning_rate": 7.660000000000001e-06,
828
+ "loss": 1.9631,
829
+ "step": 117
830
+ },
831
+ {
832
+ "epoch": 0.02800189843379212,
833
+ "grad_norm": 0.7726979851722717,
834
+ "learning_rate": 7.640000000000001e-06,
835
+ "loss": 1.8061,
836
+ "step": 118
837
+ },
838
+ {
839
+ "epoch": 0.02823920265780731,
840
+ "grad_norm": 1.1715900897979736,
841
+ "learning_rate": 7.620000000000001e-06,
842
+ "loss": 1.6768,
843
+ "step": 119
844
+ },
845
+ {
846
+ "epoch": 0.028476506881822496,
847
+ "grad_norm": 0.422097772359848,
848
+ "learning_rate": 7.600000000000001e-06,
849
+ "loss": 1.2938,
850
+ "step": 120
851
+ },
852
+ {
853
+ "epoch": 0.028713811105837685,
854
+ "grad_norm": 0.4177633225917816,
855
+ "learning_rate": 7.58e-06,
856
+ "loss": 2.4758,
857
+ "step": 121
858
+ },
859
+ {
860
+ "epoch": 0.02895111532985287,
861
+ "grad_norm": 0.8148231506347656,
862
+ "learning_rate": 7.5600000000000005e-06,
863
+ "loss": 1.6307,
864
+ "step": 122
865
+ },
866
+ {
867
+ "epoch": 0.02918841955386806,
868
+ "grad_norm": 0.36970993876457214,
869
+ "learning_rate": 7.540000000000001e-06,
870
+ "loss": 1.923,
871
+ "step": 123
872
+ },
873
+ {
874
+ "epoch": 0.029425723777883245,
875
+ "grad_norm": 0.9572102427482605,
876
+ "learning_rate": 7.520000000000001e-06,
877
+ "loss": 2.4889,
878
+ "step": 124
879
+ },
880
+ {
881
+ "epoch": 0.029663028001898434,
882
+ "grad_norm": 1.023463487625122,
883
+ "learning_rate": 7.500000000000001e-06,
884
+ "loss": 0.753,
885
+ "step": 125
886
+ },
887
+ {
888
+ "epoch": 0.029900332225913623,
889
+ "grad_norm": 0.7689042091369629,
890
+ "learning_rate": 7.48e-06,
891
+ "loss": 1.8955,
892
+ "step": 126
893
+ },
894
+ {
895
+ "epoch": 0.03013763644992881,
896
+ "grad_norm": 0.8892776370048523,
897
+ "learning_rate": 7.4600000000000006e-06,
898
+ "loss": 1.907,
899
+ "step": 127
900
+ },
901
+ {
902
+ "epoch": 0.030374940673943997,
903
+ "grad_norm": 0.6348279118537903,
904
+ "learning_rate": 7.440000000000001e-06,
905
+ "loss": 0.9907,
906
+ "step": 128
907
+ },
908
+ {
909
+ "epoch": 0.030612244897959183,
910
+ "grad_norm": 0.8271303772926331,
911
+ "learning_rate": 7.420000000000001e-06,
912
+ "loss": 0.8349,
913
+ "step": 129
914
+ },
915
+ {
916
+ "epoch": 0.030849549121974372,
917
+ "grad_norm": 0.567034900188446,
918
+ "learning_rate": 7.4e-06,
919
+ "loss": 1.5324,
920
+ "step": 130
921
+ },
922
+ {
923
+ "epoch": 0.031086853345989558,
924
+ "grad_norm": 0.7354723215103149,
925
+ "learning_rate": 7.3800000000000005e-06,
926
+ "loss": 1.8572,
927
+ "step": 131
928
+ },
929
+ {
930
+ "epoch": 0.03132415757000474,
931
+ "grad_norm": 0.7156671285629272,
932
+ "learning_rate": 7.360000000000001e-06,
933
+ "loss": 1.6312,
934
+ "step": 132
935
+ },
936
+ {
937
+ "epoch": 0.03156146179401993,
938
+ "grad_norm": 0.36890867352485657,
939
+ "learning_rate": 7.340000000000001e-06,
940
+ "loss": 0.8384,
941
+ "step": 133
942
+ },
943
+ {
944
+ "epoch": 0.03179876601803512,
945
+ "grad_norm": 0.6410567760467529,
946
+ "learning_rate": 7.32e-06,
947
+ "loss": 1.1728,
948
+ "step": 134
949
+ },
950
+ {
951
+ "epoch": 0.03203607024205031,
952
+ "grad_norm": 0.5395240187644958,
953
+ "learning_rate": 7.3e-06,
954
+ "loss": 1.3754,
955
+ "step": 135
956
+ },
957
+ {
958
+ "epoch": 0.0322733744660655,
959
+ "grad_norm": 0.9049538373947144,
960
+ "learning_rate": 7.280000000000001e-06,
961
+ "loss": 1.1323,
962
+ "step": 136
963
+ },
964
+ {
965
+ "epoch": 0.03251067869008068,
966
+ "grad_norm": 0.7486042380332947,
967
+ "learning_rate": 7.260000000000001e-06,
968
+ "loss": 2.6436,
969
+ "step": 137
970
+ },
971
+ {
972
+ "epoch": 0.03274798291409587,
973
+ "grad_norm": 0.2955069839954376,
974
+ "learning_rate": 7.24e-06,
975
+ "loss": 1.6258,
976
+ "step": 138
977
+ },
978
+ {
979
+ "epoch": 0.03298528713811106,
980
+ "grad_norm": 0.5585281848907471,
981
+ "learning_rate": 7.22e-06,
982
+ "loss": 1.0526,
983
+ "step": 139
984
+ },
985
+ {
986
+ "epoch": 0.03322259136212625,
987
+ "grad_norm": 0.6172239780426025,
988
+ "learning_rate": 7.2000000000000005e-06,
989
+ "loss": 2.0741,
990
+ "step": 140
991
+ },
992
+ {
993
+ "epoch": 0.03345989558614143,
994
+ "grad_norm": 1.0919370651245117,
995
+ "learning_rate": 7.180000000000001e-06,
996
+ "loss": 2.6546,
997
+ "step": 141
998
+ },
999
+ {
1000
+ "epoch": 0.03369719981015662,
1001
+ "grad_norm": 0.6277972459793091,
1002
+ "learning_rate": 7.16e-06,
1003
+ "loss": 1.4516,
1004
+ "step": 142
1005
+ },
1006
+ {
1007
+ "epoch": 0.03393450403417181,
1008
+ "grad_norm": 1.262568712234497,
1009
+ "learning_rate": 7.14e-06,
1010
+ "loss": 1.3782,
1011
+ "step": 143
1012
+ },
1013
+ {
1014
+ "epoch": 0.034171808258187,
1015
+ "grad_norm": 0.9131320714950562,
1016
+ "learning_rate": 7.1200000000000004e-06,
1017
+ "loss": 1.3814,
1018
+ "step": 144
1019
+ },
1020
+ {
1021
+ "epoch": 0.03440911248220219,
1022
+ "grad_norm": 0.4619258642196655,
1023
+ "learning_rate": 7.100000000000001e-06,
1024
+ "loss": 2.8978,
1025
+ "step": 145
1026
+ },
1027
+ {
1028
+ "epoch": 0.03464641670621737,
1029
+ "grad_norm": 0.17557378113269806,
1030
+ "learning_rate": 7.08e-06,
1031
+ "loss": 1.5882,
1032
+ "step": 146
1033
+ },
1034
+ {
1035
+ "epoch": 0.03488372093023256,
1036
+ "grad_norm": 0.6798960566520691,
1037
+ "learning_rate": 7.06e-06,
1038
+ "loss": 2.5057,
1039
+ "step": 147
1040
+ },
1041
+ {
1042
+ "epoch": 0.03512102515424775,
1043
+ "grad_norm": 0.7229902744293213,
1044
+ "learning_rate": 7.04e-06,
1045
+ "loss": 1.195,
1046
+ "step": 148
1047
+ },
1048
+ {
1049
+ "epoch": 0.035358329378262936,
1050
+ "grad_norm": 0.5356243848800659,
1051
+ "learning_rate": 7.0200000000000006e-06,
1052
+ "loss": 2.3255,
1053
+ "step": 149
1054
+ },
1055
+ {
1056
+ "epoch": 0.03559563360227812,
1057
+ "grad_norm": 0.735577404499054,
1058
+ "learning_rate": 7e-06,
1059
+ "loss": 0.5763,
1060
+ "step": 150
1061
+ },
1062
+ {
1063
+ "epoch": 0.03583293782629331,
1064
+ "grad_norm": 0.7191525101661682,
1065
+ "learning_rate": 6.98e-06,
1066
+ "loss": 1.5225,
1067
+ "step": 151
1068
+ },
1069
+ {
1070
+ "epoch": 0.036070242050308496,
1071
+ "grad_norm": 0.5553634166717529,
1072
+ "learning_rate": 6.96e-06,
1073
+ "loss": 2.0026,
1074
+ "step": 152
1075
+ },
1076
+ {
1077
+ "epoch": 0.036307546274323685,
1078
+ "grad_norm": 0.9029828310012817,
1079
+ "learning_rate": 6.9400000000000005e-06,
1080
+ "loss": 0.8248,
1081
+ "step": 153
1082
+ },
1083
+ {
1084
+ "epoch": 0.036544850498338874,
1085
+ "grad_norm": 0.8566996455192566,
1086
+ "learning_rate": 6.92e-06,
1087
+ "loss": 1.9291,
1088
+ "step": 154
1089
+ },
1090
+ {
1091
+ "epoch": 0.036782154722354056,
1092
+ "grad_norm": 0.677769124507904,
1093
+ "learning_rate": 6.9e-06,
1094
+ "loss": 1.5933,
1095
+ "step": 155
1096
+ },
1097
+ {
1098
+ "epoch": 0.037019458946369245,
1099
+ "grad_norm": 0.5915787220001221,
1100
+ "learning_rate": 6.88e-06,
1101
+ "loss": 0.9965,
1102
+ "step": 156
1103
+ },
1104
+ {
1105
+ "epoch": 0.037256763170384434,
1106
+ "grad_norm": 0.511048436164856,
1107
+ "learning_rate": 6.860000000000001e-06,
1108
+ "loss": 1.8256,
1109
+ "step": 157
1110
+ },
1111
+ {
1112
+ "epoch": 0.03749406739439962,
1113
+ "grad_norm": 0.49128812551498413,
1114
+ "learning_rate": 6.8400000000000014e-06,
1115
+ "loss": 1.6187,
1116
+ "step": 158
1117
+ },
1118
+ {
1119
+ "epoch": 0.037731371618414805,
1120
+ "grad_norm": 0.554414689540863,
1121
+ "learning_rate": 6.820000000000001e-06,
1122
+ "loss": 1.1991,
1123
+ "step": 159
1124
+ },
1125
+ {
1126
+ "epoch": 0.037968675842429994,
1127
+ "grad_norm": 0.9458298683166504,
1128
+ "learning_rate": 6.800000000000001e-06,
1129
+ "loss": 2.2793,
1130
+ "step": 160
1131
+ },
1132
+ {
1133
+ "epoch": 0.03820598006644518,
1134
+ "grad_norm": 2.0072669982910156,
1135
+ "learning_rate": 6.780000000000001e-06,
1136
+ "loss": 1.89,
1137
+ "step": 161
1138
+ },
1139
+ {
1140
+ "epoch": 0.03844328429046037,
1141
+ "grad_norm": 1.146154761314392,
1142
+ "learning_rate": 6.760000000000001e-06,
1143
+ "loss": 2.607,
1144
+ "step": 162
1145
+ },
1146
+ {
1147
+ "epoch": 0.038680588514475554,
1148
+ "grad_norm": 0.6168065667152405,
1149
+ "learning_rate": 6.740000000000001e-06,
1150
+ "loss": 1.459,
1151
+ "step": 163
1152
+ },
1153
+ {
1154
+ "epoch": 0.03891789273849074,
1155
+ "grad_norm": 0.4497089684009552,
1156
+ "learning_rate": 6.720000000000001e-06,
1157
+ "loss": 1.2466,
1158
+ "step": 164
1159
+ },
1160
+ {
1161
+ "epoch": 0.03915519696250593,
1162
+ "grad_norm": 0.7007705569267273,
1163
+ "learning_rate": 6.700000000000001e-06,
1164
+ "loss": 2.4729,
1165
+ "step": 165
1166
+ },
1167
+ {
1168
+ "epoch": 0.03939250118652112,
1169
+ "grad_norm": 0.8613377809524536,
1170
+ "learning_rate": 6.680000000000001e-06,
1171
+ "loss": 2.2219,
1172
+ "step": 166
1173
+ },
1174
+ {
1175
+ "epoch": 0.03962980541053631,
1176
+ "grad_norm": 0.539036750793457,
1177
+ "learning_rate": 6.660000000000001e-06,
1178
+ "loss": 0.5526,
1179
+ "step": 167
1180
+ },
1181
+ {
1182
+ "epoch": 0.03986710963455149,
1183
+ "grad_norm": 0.6085006594657898,
1184
+ "learning_rate": 6.640000000000001e-06,
1185
+ "loss": 2.1238,
1186
+ "step": 168
1187
+ },
1188
+ {
1189
+ "epoch": 0.04010441385856668,
1190
+ "grad_norm": 1.8861020803451538,
1191
+ "learning_rate": 6.620000000000001e-06,
1192
+ "loss": 2.6934,
1193
+ "step": 169
1194
+ },
1195
+ {
1196
+ "epoch": 0.04034171808258187,
1197
+ "grad_norm": 0.5803074240684509,
1198
+ "learning_rate": 6.600000000000001e-06,
1199
+ "loss": 1.2254,
1200
+ "step": 170
1201
+ },
1202
+ {
1203
+ "epoch": 0.04057902230659706,
1204
+ "grad_norm": 1.0310598611831665,
1205
+ "learning_rate": 6.5800000000000005e-06,
1206
+ "loss": 1.1757,
1207
+ "step": 171
1208
+ },
1209
+ {
1210
+ "epoch": 0.04081632653061224,
1211
+ "grad_norm": 0.9412042498588562,
1212
+ "learning_rate": 6.560000000000001e-06,
1213
+ "loss": 1.3748,
1214
+ "step": 172
1215
+ },
1216
+ {
1217
+ "epoch": 0.04105363075462743,
1218
+ "grad_norm": 0.6556461453437805,
1219
+ "learning_rate": 6.540000000000001e-06,
1220
+ "loss": 1.0809,
1221
+ "step": 173
1222
+ },
1223
+ {
1224
+ "epoch": 0.04129093497864262,
1225
+ "grad_norm": 0.4990858733654022,
1226
+ "learning_rate": 6.520000000000001e-06,
1227
+ "loss": 1.2934,
1228
+ "step": 174
1229
+ },
1230
+ {
1231
+ "epoch": 0.04152823920265781,
1232
+ "grad_norm": 0.6699053645133972,
1233
+ "learning_rate": 6.5000000000000004e-06,
1234
+ "loss": 2.4814,
1235
+ "step": 175
1236
+ },
1237
+ {
1238
+ "epoch": 0.041765543426673,
1239
+ "grad_norm": 0.8843134641647339,
1240
+ "learning_rate": 6.480000000000001e-06,
1241
+ "loss": 1.9615,
1242
+ "step": 176
1243
+ },
1244
+ {
1245
+ "epoch": 0.04200284765068818,
1246
+ "grad_norm": 0.6757798790931702,
1247
+ "learning_rate": 6.460000000000001e-06,
1248
+ "loss": 1.7584,
1249
+ "step": 177
1250
+ },
1251
+ {
1252
+ "epoch": 0.04224015187470337,
1253
+ "grad_norm": 0.46641185879707336,
1254
+ "learning_rate": 6.440000000000001e-06,
1255
+ "loss": 1.3054,
1256
+ "step": 178
1257
+ },
1258
+ {
1259
+ "epoch": 0.04247745609871856,
1260
+ "grad_norm": 0.7104260325431824,
1261
+ "learning_rate": 6.42e-06,
1262
+ "loss": 1.1761,
1263
+ "step": 179
1264
+ },
1265
+ {
1266
+ "epoch": 0.04271476032273375,
1267
+ "grad_norm": 0.3888971209526062,
1268
+ "learning_rate": 6.4000000000000006e-06,
1269
+ "loss": 2.2055,
1270
+ "step": 180
1271
+ },
1272
+ {
1273
+ "epoch": 0.04295206454674893,
1274
+ "grad_norm": 1.1661548614501953,
1275
+ "learning_rate": 6.380000000000001e-06,
1276
+ "loss": 1.1329,
1277
+ "step": 181
1278
+ },
1279
+ {
1280
+ "epoch": 0.04318936877076412,
1281
+ "grad_norm": 0.40143099427223206,
1282
+ "learning_rate": 6.360000000000001e-06,
1283
+ "loss": 1.0135,
1284
+ "step": 182
1285
+ },
1286
+ {
1287
+ "epoch": 0.04342667299477931,
1288
+ "grad_norm": 0.692574679851532,
1289
+ "learning_rate": 6.34e-06,
1290
+ "loss": 1.3553,
1291
+ "step": 183
1292
+ },
1293
+ {
1294
+ "epoch": 0.043663977218794496,
1295
+ "grad_norm": 0.6210284233093262,
1296
+ "learning_rate": 6.3200000000000005e-06,
1297
+ "loss": 1.4226,
1298
+ "step": 184
1299
+ },
1300
+ {
1301
+ "epoch": 0.043901281442809685,
1302
+ "grad_norm": 0.6218913197517395,
1303
+ "learning_rate": 6.300000000000001e-06,
1304
+ "loss": 1.8475,
1305
+ "step": 185
1306
+ },
1307
+ {
1308
+ "epoch": 0.04413858566682487,
1309
+ "grad_norm": 0.6004898548126221,
1310
+ "learning_rate": 6.280000000000001e-06,
1311
+ "loss": 1.8829,
1312
+ "step": 186
1313
+ },
1314
+ {
1315
+ "epoch": 0.044375889890840056,
1316
+ "grad_norm": 0.8992873430252075,
1317
+ "learning_rate": 6.26e-06,
1318
+ "loss": 1.0599,
1319
+ "step": 187
1320
+ },
1321
+ {
1322
+ "epoch": 0.044613194114855245,
1323
+ "grad_norm": 0.5328947305679321,
1324
+ "learning_rate": 6.24e-06,
1325
+ "loss": 0.7189,
1326
+ "step": 188
1327
+ },
1328
+ {
1329
+ "epoch": 0.044850498338870434,
1330
+ "grad_norm": 0.5015738010406494,
1331
+ "learning_rate": 6.220000000000001e-06,
1332
+ "loss": 2.0798,
1333
+ "step": 189
1334
+ },
1335
+ {
1336
+ "epoch": 0.045087802562885616,
1337
+ "grad_norm": 0.47242438793182373,
1338
+ "learning_rate": 6.200000000000001e-06,
1339
+ "loss": 1.3423,
1340
+ "step": 190
1341
+ },
1342
+ {
1343
+ "epoch": 0.045325106786900805,
1344
+ "grad_norm": 0.9236852526664734,
1345
+ "learning_rate": 6.18e-06,
1346
+ "loss": 1.2769,
1347
+ "step": 191
1348
+ },
1349
+ {
1350
+ "epoch": 0.045562411010915994,
1351
+ "grad_norm": 0.47883141040802,
1352
+ "learning_rate": 6.16e-06,
1353
+ "loss": 1.5209,
1354
+ "step": 192
1355
+ },
1356
+ {
1357
+ "epoch": 0.04579971523493118,
1358
+ "grad_norm": 0.760746955871582,
1359
+ "learning_rate": 6.1400000000000005e-06,
1360
+ "loss": 1.6353,
1361
+ "step": 193
1362
+ },
1363
+ {
1364
+ "epoch": 0.04603701945894637,
1365
+ "grad_norm": 0.5924590826034546,
1366
+ "learning_rate": 6.120000000000001e-06,
1367
+ "loss": 1.3738,
1368
+ "step": 194
1369
+ },
1370
+ {
1371
+ "epoch": 0.046274323682961555,
1372
+ "grad_norm": 0.7453778386116028,
1373
+ "learning_rate": 6.1e-06,
1374
+ "loss": 0.8792,
1375
+ "step": 195
1376
+ },
1377
+ {
1378
+ "epoch": 0.046511627906976744,
1379
+ "grad_norm": 0.8486724495887756,
1380
+ "learning_rate": 6.08e-06,
1381
+ "loss": 1.7324,
1382
+ "step": 196
1383
+ },
1384
+ {
1385
+ "epoch": 0.04674893213099193,
1386
+ "grad_norm": 0.5349167585372925,
1387
+ "learning_rate": 6.0600000000000004e-06,
1388
+ "loss": 1.1938,
1389
+ "step": 197
1390
+ },
1391
+ {
1392
+ "epoch": 0.04698623635500712,
1393
+ "grad_norm": 0.4522935748100281,
1394
+ "learning_rate": 6.040000000000001e-06,
1395
+ "loss": 1.282,
1396
+ "step": 198
1397
+ },
1398
+ {
1399
+ "epoch": 0.047223540579022304,
1400
+ "grad_norm": 0.6473549008369446,
1401
+ "learning_rate": 6.02e-06,
1402
+ "loss": 0.9753,
1403
+ "step": 199
1404
+ },
1405
+ {
1406
+ "epoch": 0.04746084480303749,
1407
+ "grad_norm": 0.8137974739074707,
1408
+ "learning_rate": 6e-06,
1409
+ "loss": 1.6364,
1410
+ "step": 200
1411
+ },
1412
+ {
1413
+ "epoch": 0.04769814902705268,
1414
+ "grad_norm": 0.48920387029647827,
1415
+ "learning_rate": 5.98e-06,
1416
+ "loss": 1.8822,
1417
+ "step": 201
1418
+ },
1419
+ {
1420
+ "epoch": 0.04793545325106787,
1421
+ "grad_norm": 0.7745688557624817,
1422
+ "learning_rate": 5.9600000000000005e-06,
1423
+ "loss": 1.1183,
1424
+ "step": 202
1425
+ },
1426
+ {
1427
+ "epoch": 0.04817275747508306,
1428
+ "grad_norm": 1.2582346200942993,
1429
+ "learning_rate": 5.94e-06,
1430
+ "loss": 2.3109,
1431
+ "step": 203
1432
+ },
1433
+ {
1434
+ "epoch": 0.04841006169909824,
1435
+ "grad_norm": 0.6862124800682068,
1436
+ "learning_rate": 5.92e-06,
1437
+ "loss": 1.972,
1438
+ "step": 204
1439
+ },
1440
+ {
1441
+ "epoch": 0.04864736592311343,
1442
+ "grad_norm": 0.2611483931541443,
1443
+ "learning_rate": 5.9e-06,
1444
+ "loss": 1.657,
1445
+ "step": 205
1446
+ },
1447
+ {
1448
+ "epoch": 0.04888467014712862,
1449
+ "grad_norm": 0.6788746118545532,
1450
+ "learning_rate": 5.8800000000000005e-06,
1451
+ "loss": 1.4215,
1452
+ "step": 206
1453
+ },
1454
+ {
1455
+ "epoch": 0.04912197437114381,
1456
+ "grad_norm": 0.7344043254852295,
1457
+ "learning_rate": 5.86e-06,
1458
+ "loss": 1.8161,
1459
+ "step": 207
1460
+ },
1461
+ {
1462
+ "epoch": 0.04935927859515899,
1463
+ "grad_norm": 0.9730067849159241,
1464
+ "learning_rate": 5.84e-06,
1465
+ "loss": 2.009,
1466
+ "step": 208
1467
+ },
1468
+ {
1469
+ "epoch": 0.04959658281917418,
1470
+ "grad_norm": 0.5687656998634338,
1471
+ "learning_rate": 5.82e-06,
1472
+ "loss": 1.371,
1473
+ "step": 209
1474
+ },
1475
+ {
1476
+ "epoch": 0.04983388704318937,
1477
+ "grad_norm": 0.9074623584747314,
1478
+ "learning_rate": 5.8e-06,
1479
+ "loss": 0.9096,
1480
+ "step": 210
1481
+ },
1482
+ {
1483
+ "epoch": 0.05007119126720456,
1484
+ "grad_norm": 0.5359849333763123,
1485
+ "learning_rate": 5.78e-06,
1486
+ "loss": 1.7314,
1487
+ "step": 211
1488
+ },
1489
+ {
1490
+ "epoch": 0.05030849549121975,
1491
+ "grad_norm": 0.5226656198501587,
1492
+ "learning_rate": 5.76e-06,
1493
+ "loss": 0.7389,
1494
+ "step": 212
1495
+ },
1496
+ {
1497
+ "epoch": 0.05054579971523493,
1498
+ "grad_norm": 0.8956894874572754,
1499
+ "learning_rate": 5.74e-06,
1500
+ "loss": 1.1937,
1501
+ "step": 213
1502
+ },
1503
+ {
1504
+ "epoch": 0.05078310393925012,
1505
+ "grad_norm": 1.2422311305999756,
1506
+ "learning_rate": 5.72e-06,
1507
+ "loss": 1.8545,
1508
+ "step": 214
1509
+ },
1510
+ {
1511
+ "epoch": 0.05102040816326531,
1512
+ "grad_norm": 0.5634399056434631,
1513
+ "learning_rate": 5.7e-06,
1514
+ "loss": 1.446,
1515
+ "step": 215
1516
+ },
1517
+ {
1518
+ "epoch": 0.051257712387280496,
1519
+ "grad_norm": 0.7447414398193359,
1520
+ "learning_rate": 5.68e-06,
1521
+ "loss": 1.7231,
1522
+ "step": 216
1523
+ },
1524
+ {
1525
+ "epoch": 0.05149501661129568,
1526
+ "grad_norm": 1.131425380706787,
1527
+ "learning_rate": 5.66e-06,
1528
+ "loss": 1.4652,
1529
+ "step": 217
1530
+ },
1531
+ {
1532
+ "epoch": 0.05173232083531087,
1533
+ "grad_norm": 0.8668593764305115,
1534
+ "learning_rate": 5.64e-06,
1535
+ "loss": 2.0443,
1536
+ "step": 218
1537
+ },
1538
+ {
1539
+ "epoch": 0.051969625059326056,
1540
+ "grad_norm": 1.2437993288040161,
1541
+ "learning_rate": 5.620000000000001e-06,
1542
+ "loss": 2.1323,
1543
+ "step": 219
1544
+ },
1545
+ {
1546
+ "epoch": 0.052206929283341245,
1547
+ "grad_norm": 0.914955198764801,
1548
+ "learning_rate": 5.600000000000001e-06,
1549
+ "loss": 2.2205,
1550
+ "step": 220
1551
+ },
1552
+ {
1553
+ "epoch": 0.05244423350735643,
1554
+ "grad_norm": 1.0641348361968994,
1555
+ "learning_rate": 5.580000000000001e-06,
1556
+ "loss": 2.4159,
1557
+ "step": 221
1558
+ },
1559
+ {
1560
+ "epoch": 0.05268153773137162,
1561
+ "grad_norm": 0.17957435548305511,
1562
+ "learning_rate": 5.560000000000001e-06,
1563
+ "loss": 1.1683,
1564
+ "step": 222
1565
+ },
1566
+ {
1567
+ "epoch": 0.052918841955386806,
1568
+ "grad_norm": 1.058635950088501,
1569
+ "learning_rate": 5.540000000000001e-06,
1570
+ "loss": 1.8851,
1571
+ "step": 223
1572
+ },
1573
+ {
1574
+ "epoch": 0.053156146179401995,
1575
+ "grad_norm": 0.6216885447502136,
1576
+ "learning_rate": 5.5200000000000005e-06,
1577
+ "loss": 2.0638,
1578
+ "step": 224
1579
+ },
1580
+ {
1581
+ "epoch": 0.053393450403417184,
1582
+ "grad_norm": 0.32197320461273193,
1583
+ "learning_rate": 5.500000000000001e-06,
1584
+ "loss": 4.2387,
1585
+ "step": 225
1586
+ },
1587
+ {
1588
+ "epoch": 0.053630754627432366,
1589
+ "grad_norm": 0.5334311127662659,
1590
+ "learning_rate": 5.480000000000001e-06,
1591
+ "loss": 1.199,
1592
+ "step": 226
1593
+ },
1594
+ {
1595
+ "epoch": 0.053868058851447555,
1596
+ "grad_norm": 0.8656753301620483,
1597
+ "learning_rate": 5.460000000000001e-06,
1598
+ "loss": 1.0053,
1599
+ "step": 227
1600
+ },
1601
+ {
1602
+ "epoch": 0.054105363075462744,
1603
+ "grad_norm": 0.9633522629737854,
1604
+ "learning_rate": 5.4400000000000004e-06,
1605
+ "loss": 2.6078,
1606
+ "step": 228
1607
+ },
1608
+ {
1609
+ "epoch": 0.05434266729947793,
1610
+ "grad_norm": 0.8262597322463989,
1611
+ "learning_rate": 5.420000000000001e-06,
1612
+ "loss": 2.4638,
1613
+ "step": 229
1614
+ },
1615
+ {
1616
+ "epoch": 0.054579971523493115,
1617
+ "grad_norm": 0.9816875457763672,
1618
+ "learning_rate": 5.400000000000001e-06,
1619
+ "loss": 2.3545,
1620
+ "step": 230
1621
+ },
1622
+ {
1623
+ "epoch": 0.054817275747508304,
1624
+ "grad_norm": 0.7079796195030212,
1625
+ "learning_rate": 5.380000000000001e-06,
1626
+ "loss": 1.7035,
1627
+ "step": 231
1628
+ },
1629
+ {
1630
+ "epoch": 0.05505457997152349,
1631
+ "grad_norm": 0.2551076412200928,
1632
+ "learning_rate": 5.36e-06,
1633
+ "loss": 2.3941,
1634
+ "step": 232
1635
+ },
1636
+ {
1637
+ "epoch": 0.05529188419553868,
1638
+ "grad_norm": 0.8667798042297363,
1639
+ "learning_rate": 5.3400000000000005e-06,
1640
+ "loss": 2.5876,
1641
+ "step": 233
1642
+ },
1643
+ {
1644
+ "epoch": 0.05552918841955387,
1645
+ "grad_norm": 0.48972687125205994,
1646
+ "learning_rate": 5.320000000000001e-06,
1647
+ "loss": 1.9343,
1648
+ "step": 234
1649
+ },
1650
+ {
1651
+ "epoch": 0.05576649264356905,
1652
+ "grad_norm": 1.166282057762146,
1653
+ "learning_rate": 5.300000000000001e-06,
1654
+ "loss": 1.5903,
1655
+ "step": 235
1656
+ },
1657
+ {
1658
+ "epoch": 0.05600379686758424,
1659
+ "grad_norm": 1.0136897563934326,
1660
+ "learning_rate": 5.28e-06,
1661
+ "loss": 1.9003,
1662
+ "step": 236
1663
+ },
1664
+ {
1665
+ "epoch": 0.05624110109159943,
1666
+ "grad_norm": 0.9301249384880066,
1667
+ "learning_rate": 5.2600000000000005e-06,
1668
+ "loss": 2.2514,
1669
+ "step": 237
1670
+ },
1671
+ {
1672
+ "epoch": 0.05647840531561462,
1673
+ "grad_norm": 0.6378384232521057,
1674
+ "learning_rate": 5.240000000000001e-06,
1675
+ "loss": 1.0407,
1676
+ "step": 238
1677
+ },
1678
+ {
1679
+ "epoch": 0.0567157095396298,
1680
+ "grad_norm": 0.7191042900085449,
1681
+ "learning_rate": 5.220000000000001e-06,
1682
+ "loss": 1.8487,
1683
+ "step": 239
1684
+ },
1685
+ {
1686
+ "epoch": 0.05695301376364499,
1687
+ "grad_norm": 0.8724852204322815,
1688
+ "learning_rate": 5.2e-06,
1689
+ "loss": 1.9766,
1690
+ "step": 240
1691
+ },
1692
+ {
1693
+ "epoch": 0.05719031798766018,
1694
+ "grad_norm": 1.2465623617172241,
1695
+ "learning_rate": 5.18e-06,
1696
+ "loss": 2.0123,
1697
+ "step": 241
1698
+ },
1699
+ {
1700
+ "epoch": 0.05742762221167537,
1701
+ "grad_norm": 0.6831521987915039,
1702
+ "learning_rate": 5.1600000000000006e-06,
1703
+ "loss": 2.5164,
1704
+ "step": 242
1705
+ },
1706
+ {
1707
+ "epoch": 0.05766492643569056,
1708
+ "grad_norm": 0.9678359627723694,
1709
+ "learning_rate": 5.140000000000001e-06,
1710
+ "loss": 1.1767,
1711
+ "step": 243
1712
+ },
1713
+ {
1714
+ "epoch": 0.05790223065970574,
1715
+ "grad_norm": 0.7171100378036499,
1716
+ "learning_rate": 5.12e-06,
1717
+ "loss": 1.5144,
1718
+ "step": 244
1719
+ },
1720
+ {
1721
+ "epoch": 0.05813953488372093,
1722
+ "grad_norm": 1.8189458847045898,
1723
+ "learning_rate": 5.1e-06,
1724
+ "loss": 0.8019,
1725
+ "step": 245
1726
+ },
1727
+ {
1728
+ "epoch": 0.05837683910773612,
1729
+ "grad_norm": 0.8519846796989441,
1730
+ "learning_rate": 5.0800000000000005e-06,
1731
+ "loss": 0.5605,
1732
+ "step": 246
1733
+ },
1734
+ {
1735
+ "epoch": 0.05861414333175131,
1736
+ "grad_norm": 0.8352647423744202,
1737
+ "learning_rate": 5.060000000000001e-06,
1738
+ "loss": 1.83,
1739
+ "step": 247
1740
+ },
1741
+ {
1742
+ "epoch": 0.05885144755576649,
1743
+ "grad_norm": 0.8553798198699951,
1744
+ "learning_rate": 5.04e-06,
1745
+ "loss": 0.8078,
1746
+ "step": 248
1747
+ },
1748
+ {
1749
+ "epoch": 0.05908875177978168,
1750
+ "grad_norm": 0.5836890935897827,
1751
+ "learning_rate": 5.02e-06,
1752
+ "loss": 1.805,
1753
+ "step": 249
1754
+ },
1755
+ {
1756
+ "epoch": 0.05932605600379687,
1757
+ "grad_norm": 1.2686045169830322,
1758
+ "learning_rate": 5e-06,
1759
+ "loss": 2.9273,
1760
+ "step": 250
1761
+ },
1762
+ {
1763
+ "epoch": 0.05956336022781206,
1764
+ "grad_norm": 1.0224002599716187,
1765
+ "learning_rate": 4.980000000000001e-06,
1766
+ "loss": 1.5689,
1767
+ "step": 251
1768
+ },
1769
+ {
1770
+ "epoch": 0.059800664451827246,
1771
+ "grad_norm": 0.6084955334663391,
1772
+ "learning_rate": 4.960000000000001e-06,
1773
+ "loss": 1.4633,
1774
+ "step": 252
1775
+ },
1776
+ {
1777
+ "epoch": 0.06003796867584243,
1778
+ "grad_norm": 1.0595405101776123,
1779
+ "learning_rate": 4.94e-06,
1780
+ "loss": 1.799,
1781
+ "step": 253
1782
+ },
1783
+ {
1784
+ "epoch": 0.06027527289985762,
1785
+ "grad_norm": 0.6184794306755066,
1786
+ "learning_rate": 4.92e-06,
1787
+ "loss": 0.936,
1788
+ "step": 254
1789
+ },
1790
+ {
1791
+ "epoch": 0.060512577123872806,
1792
+ "grad_norm": 0.8224856853485107,
1793
+ "learning_rate": 4.9000000000000005e-06,
1794
+ "loss": 1.2481,
1795
+ "step": 255
1796
+ },
1797
+ {
1798
+ "epoch": 0.060749881347887995,
1799
+ "grad_norm": 0.4918515086174011,
1800
+ "learning_rate": 4.880000000000001e-06,
1801
+ "loss": 1.6237,
1802
+ "step": 256
1803
+ },
1804
+ {
1805
+ "epoch": 0.06098718557190318,
1806
+ "grad_norm": 0.7177821397781372,
1807
+ "learning_rate": 4.86e-06,
1808
+ "loss": 1.6721,
1809
+ "step": 257
1810
+ },
1811
+ {
1812
+ "epoch": 0.061224489795918366,
1813
+ "grad_norm": 0.654577374458313,
1814
+ "learning_rate": 4.84e-06,
1815
+ "loss": 1.6336,
1816
+ "step": 258
1817
+ },
1818
+ {
1819
+ "epoch": 0.061461794019933555,
1820
+ "grad_norm": 0.6433861255645752,
1821
+ "learning_rate": 4.8200000000000004e-06,
1822
+ "loss": 0.7603,
1823
+ "step": 259
1824
+ },
1825
+ {
1826
+ "epoch": 0.061699098243948744,
1827
+ "grad_norm": 0.8999320864677429,
1828
+ "learning_rate": 4.800000000000001e-06,
1829
+ "loss": 1.5193,
1830
+ "step": 260
1831
+ },
1832
+ {
1833
+ "epoch": 0.06193640246796393,
1834
+ "grad_norm": 1.071006417274475,
1835
+ "learning_rate": 4.78e-06,
1836
+ "loss": 2.1792,
1837
+ "step": 261
1838
+ },
1839
+ {
1840
+ "epoch": 0.062173706691979115,
1841
+ "grad_norm": 0.950939416885376,
1842
+ "learning_rate": 4.76e-06,
1843
+ "loss": 1.7475,
1844
+ "step": 262
1845
+ },
1846
+ {
1847
+ "epoch": 0.062411010915994304,
1848
+ "grad_norm": 0.6791463494300842,
1849
+ "learning_rate": 4.74e-06,
1850
+ "loss": 1.5734,
1851
+ "step": 263
1852
+ },
1853
+ {
1854
+ "epoch": 0.06264831514000949,
1855
+ "grad_norm": 0.8757117986679077,
1856
+ "learning_rate": 4.7200000000000005e-06,
1857
+ "loss": 0.9807,
1858
+ "step": 264
1859
+ },
1860
+ {
1861
+ "epoch": 0.06288561936402468,
1862
+ "grad_norm": 1.6328972578048706,
1863
+ "learning_rate": 4.7e-06,
1864
+ "loss": 1.3517,
1865
+ "step": 265
1866
+ },
1867
+ {
1868
+ "epoch": 0.06312292358803986,
1869
+ "grad_norm": 1.2624143362045288,
1870
+ "learning_rate": 4.680000000000001e-06,
1871
+ "loss": 1.935,
1872
+ "step": 266
1873
+ },
1874
+ {
1875
+ "epoch": 0.06336022781205505,
1876
+ "grad_norm": 0.8143572211265564,
1877
+ "learning_rate": 4.66e-06,
1878
+ "loss": 1.2027,
1879
+ "step": 267
1880
+ },
1881
+ {
1882
+ "epoch": 0.06359753203607024,
1883
+ "grad_norm": 1.05904221534729,
1884
+ "learning_rate": 4.6400000000000005e-06,
1885
+ "loss": 2.0852,
1886
+ "step": 268
1887
+ },
1888
+ {
1889
+ "epoch": 0.06383483626008543,
1890
+ "grad_norm": 0.5743423104286194,
1891
+ "learning_rate": 4.620000000000001e-06,
1892
+ "loss": 1.649,
1893
+ "step": 269
1894
+ },
1895
+ {
1896
+ "epoch": 0.06407214048410062,
1897
+ "grad_norm": 1.2721295356750488,
1898
+ "learning_rate": 4.600000000000001e-06,
1899
+ "loss": 1.7225,
1900
+ "step": 270
1901
+ },
1902
+ {
1903
+ "epoch": 0.06430944470811581,
1904
+ "grad_norm": 0.8935225009918213,
1905
+ "learning_rate": 4.58e-06,
1906
+ "loss": 1.7969,
1907
+ "step": 271
1908
+ },
1909
+ {
1910
+ "epoch": 0.064546748932131,
1911
+ "grad_norm": 0.4470404088497162,
1912
+ "learning_rate": 4.56e-06,
1913
+ "loss": 1.4834,
1914
+ "step": 272
1915
+ },
1916
+ {
1917
+ "epoch": 0.06478405315614617,
1918
+ "grad_norm": 0.41209957003593445,
1919
+ "learning_rate": 4.540000000000001e-06,
1920
+ "loss": 1.5577,
1921
+ "step": 273
1922
+ },
1923
+ {
1924
+ "epoch": 0.06502135738016136,
1925
+ "grad_norm": 1.3055024147033691,
1926
+ "learning_rate": 4.520000000000001e-06,
1927
+ "loss": 2.0445,
1928
+ "step": 274
1929
+ },
1930
+ {
1931
+ "epoch": 0.06525866160417655,
1932
+ "grad_norm": 1.182127594947815,
1933
+ "learning_rate": 4.5e-06,
1934
+ "loss": 2.5339,
1935
+ "step": 275
1936
+ },
1937
+ {
1938
+ "epoch": 0.06549596582819174,
1939
+ "grad_norm": 1.030988335609436,
1940
+ "learning_rate": 4.48e-06,
1941
+ "loss": 2.8977,
1942
+ "step": 276
1943
+ },
1944
+ {
1945
+ "epoch": 0.06573327005220693,
1946
+ "grad_norm": 1.023729920387268,
1947
+ "learning_rate": 4.4600000000000005e-06,
1948
+ "loss": 3.7561,
1949
+ "step": 277
1950
+ },
1951
+ {
1952
+ "epoch": 0.06597057427622212,
1953
+ "grad_norm": 0.8441396951675415,
1954
+ "learning_rate": 4.440000000000001e-06,
1955
+ "loss": 0.9105,
1956
+ "step": 278
1957
+ },
1958
+ {
1959
+ "epoch": 0.06620787850023731,
1960
+ "grad_norm": 0.781200647354126,
1961
+ "learning_rate": 4.42e-06,
1962
+ "loss": 1.1585,
1963
+ "step": 279
1964
+ },
1965
+ {
1966
+ "epoch": 0.0664451827242525,
1967
+ "grad_norm": 1.0872159004211426,
1968
+ "learning_rate": 4.4e-06,
1969
+ "loss": 1.8595,
1970
+ "step": 280
1971
+ },
1972
+ {
1973
+ "epoch": 0.06668248694826769,
1974
+ "grad_norm": 0.7548374533653259,
1975
+ "learning_rate": 4.38e-06,
1976
+ "loss": 1.5494,
1977
+ "step": 281
1978
+ },
1979
+ {
1980
+ "epoch": 0.06691979117228286,
1981
+ "grad_norm": 0.7101930975914001,
1982
+ "learning_rate": 4.360000000000001e-06,
1983
+ "loss": 0.9779,
1984
+ "step": 282
1985
+ },
1986
+ {
1987
+ "epoch": 0.06715709539629805,
1988
+ "grad_norm": 2.0798416137695312,
1989
+ "learning_rate": 4.34e-06,
1990
+ "loss": 2.2541,
1991
+ "step": 283
1992
+ },
1993
+ {
1994
+ "epoch": 0.06739439962031324,
1995
+ "grad_norm": 0.45076173543930054,
1996
+ "learning_rate": 4.32e-06,
1997
+ "loss": 2.0375,
1998
+ "step": 284
1999
+ },
2000
+ {
2001
+ "epoch": 0.06763170384432843,
2002
+ "grad_norm": 1.132407784461975,
2003
+ "learning_rate": 4.3e-06,
2004
+ "loss": 1.8917,
2005
+ "step": 285
2006
+ },
2007
+ {
2008
+ "epoch": 0.06786900806834362,
2009
+ "grad_norm": 0.4013515114784241,
2010
+ "learning_rate": 4.2800000000000005e-06,
2011
+ "loss": 0.3564,
2012
+ "step": 286
2013
+ },
2014
+ {
2015
+ "epoch": 0.0681063122923588,
2016
+ "grad_norm": 0.7896368503570557,
2017
+ "learning_rate": 4.26e-06,
2018
+ "loss": 1.5522,
2019
+ "step": 287
2020
+ },
2021
+ {
2022
+ "epoch": 0.068343616516374,
2023
+ "grad_norm": 0.5332828760147095,
2024
+ "learning_rate": 4.24e-06,
2025
+ "loss": 0.56,
2026
+ "step": 288
2027
+ },
2028
+ {
2029
+ "epoch": 0.06858092074038918,
2030
+ "grad_norm": 0.6611258387565613,
2031
+ "learning_rate": 4.22e-06,
2032
+ "loss": 1.5975,
2033
+ "step": 289
2034
+ },
2035
+ {
2036
+ "epoch": 0.06881822496440437,
2037
+ "grad_norm": 0.8199064135551453,
2038
+ "learning_rate": 4.2000000000000004e-06,
2039
+ "loss": 2.1785,
2040
+ "step": 290
2041
+ },
2042
+ {
2043
+ "epoch": 0.06905552918841955,
2044
+ "grad_norm": 0.8175731301307678,
2045
+ "learning_rate": 4.18e-06,
2046
+ "loss": 2.1028,
2047
+ "step": 291
2048
+ },
2049
+ {
2050
+ "epoch": 0.06929283341243474,
2051
+ "grad_norm": 0.7817385196685791,
2052
+ "learning_rate": 4.16e-06,
2053
+ "loss": 1.6518,
2054
+ "step": 292
2055
+ },
2056
+ {
2057
+ "epoch": 0.06953013763644993,
2058
+ "grad_norm": 0.7397461533546448,
2059
+ "learning_rate": 4.14e-06,
2060
+ "loss": 1.4586,
2061
+ "step": 293
2062
+ },
2063
+ {
2064
+ "epoch": 0.06976744186046512,
2065
+ "grad_norm": 0.39525383710861206,
2066
+ "learning_rate": 4.12e-06,
2067
+ "loss": 2.1693,
2068
+ "step": 294
2069
+ },
2070
+ {
2071
+ "epoch": 0.0700047460844803,
2072
+ "grad_norm": 0.16452622413635254,
2073
+ "learning_rate": 4.1e-06,
2074
+ "loss": 2.0744,
2075
+ "step": 295
2076
+ },
2077
+ {
2078
+ "epoch": 0.0702420503084955,
2079
+ "grad_norm": 0.49008700251579285,
2080
+ "learning_rate": 4.08e-06,
2081
+ "loss": 1.5107,
2082
+ "step": 296
2083
+ },
2084
+ {
2085
+ "epoch": 0.07047935453251068,
2086
+ "grad_norm": 0.6975173354148865,
2087
+ "learning_rate": 4.060000000000001e-06,
2088
+ "loss": 1.4174,
2089
+ "step": 297
2090
+ },
2091
+ {
2092
+ "epoch": 0.07071665875652587,
2093
+ "grad_norm": 0.6213851571083069,
2094
+ "learning_rate": 4.04e-06,
2095
+ "loss": 1.4038,
2096
+ "step": 298
2097
+ },
2098
+ {
2099
+ "epoch": 0.07095396298054106,
2100
+ "grad_norm": 0.5320644378662109,
2101
+ "learning_rate": 4.0200000000000005e-06,
2102
+ "loss": 0.8553,
2103
+ "step": 299
2104
+ },
2105
+ {
2106
+ "epoch": 0.07119126720455624,
2107
+ "grad_norm": 1.0871286392211914,
2108
+ "learning_rate": 4.000000000000001e-06,
2109
+ "loss": 2.0419,
2110
+ "step": 300
2111
+ },
2112
+ {
2113
+ "epoch": 0.07142857142857142,
2114
+ "grad_norm": 0.3801209330558777,
2115
+ "learning_rate": 3.980000000000001e-06,
2116
+ "loss": 0.5386,
2117
+ "step": 301
2118
+ },
2119
+ {
2120
+ "epoch": 0.07166587565258661,
2121
+ "grad_norm": 0.39513254165649414,
2122
+ "learning_rate": 3.96e-06,
2123
+ "loss": 2.0508,
2124
+ "step": 302
2125
+ },
2126
+ {
2127
+ "epoch": 0.0719031798766018,
2128
+ "grad_norm": 0.45403411984443665,
2129
+ "learning_rate": 3.94e-06,
2130
+ "loss": 1.3929,
2131
+ "step": 303
2132
+ },
2133
+ {
2134
+ "epoch": 0.07214048410061699,
2135
+ "grad_norm": 0.9575373530387878,
2136
+ "learning_rate": 3.920000000000001e-06,
2137
+ "loss": 2.3445,
2138
+ "step": 304
2139
+ },
2140
+ {
2141
+ "epoch": 0.07237778832463218,
2142
+ "grad_norm": 0.7246173620223999,
2143
+ "learning_rate": 3.900000000000001e-06,
2144
+ "loss": 1.1485,
2145
+ "step": 305
2146
+ },
2147
+ {
2148
+ "epoch": 0.07261509254864737,
2149
+ "grad_norm": 0.4268713891506195,
2150
+ "learning_rate": 3.88e-06,
2151
+ "loss": 1.7111,
2152
+ "step": 306
2153
+ },
2154
+ {
2155
+ "epoch": 0.07285239677266256,
2156
+ "grad_norm": 0.3135124742984772,
2157
+ "learning_rate": 3.86e-06,
2158
+ "loss": 1.4731,
2159
+ "step": 307
2160
+ },
2161
+ {
2162
+ "epoch": 0.07308970099667775,
2163
+ "grad_norm": 0.647866427898407,
2164
+ "learning_rate": 3.8400000000000005e-06,
2165
+ "loss": 0.9345,
2166
+ "step": 308
2167
+ },
2168
+ {
2169
+ "epoch": 0.07332700522069292,
2170
+ "grad_norm": 0.6480103135108948,
2171
+ "learning_rate": 3.820000000000001e-06,
2172
+ "loss": 0.5553,
2173
+ "step": 309
2174
+ },
2175
+ {
2176
+ "epoch": 0.07356430944470811,
2177
+ "grad_norm": 0.7226047515869141,
2178
+ "learning_rate": 3.8000000000000005e-06,
2179
+ "loss": 1.1579,
2180
+ "step": 310
2181
+ },
2182
+ {
2183
+ "epoch": 0.0738016136687233,
2184
+ "grad_norm": 0.6003654599189758,
2185
+ "learning_rate": 3.7800000000000002e-06,
2186
+ "loss": 1.2483,
2187
+ "step": 311
2188
+ },
2189
+ {
2190
+ "epoch": 0.07403891789273849,
2191
+ "grad_norm": 0.6066475510597229,
2192
+ "learning_rate": 3.7600000000000004e-06,
2193
+ "loss": 0.7838,
2194
+ "step": 312
2195
+ },
2196
+ {
2197
+ "epoch": 0.07427622211675368,
2198
+ "grad_norm": 0.5357272624969482,
2199
+ "learning_rate": 3.74e-06,
2200
+ "loss": 1.8871,
2201
+ "step": 313
2202
+ },
2203
+ {
2204
+ "epoch": 0.07451352634076887,
2205
+ "grad_norm": 0.8201131820678711,
2206
+ "learning_rate": 3.7200000000000004e-06,
2207
+ "loss": 1.6833,
2208
+ "step": 314
2209
+ },
2210
+ {
2211
+ "epoch": 0.07475083056478406,
2212
+ "grad_norm": 0.6600767970085144,
2213
+ "learning_rate": 3.7e-06,
2214
+ "loss": 1.5297,
2215
+ "step": 315
2216
+ },
2217
+ {
2218
+ "epoch": 0.07498813478879925,
2219
+ "grad_norm": 0.6373748779296875,
2220
+ "learning_rate": 3.6800000000000003e-06,
2221
+ "loss": 1.5093,
2222
+ "step": 316
2223
+ },
2224
+ {
2225
+ "epoch": 0.07522543901281442,
2226
+ "grad_norm": 0.7496886849403381,
2227
+ "learning_rate": 3.66e-06,
2228
+ "loss": 2.4157,
2229
+ "step": 317
2230
+ },
2231
+ {
2232
+ "epoch": 0.07546274323682961,
2233
+ "grad_norm": 0.9333056211471558,
2234
+ "learning_rate": 3.6400000000000003e-06,
2235
+ "loss": 1.5617,
2236
+ "step": 318
2237
+ },
2238
+ {
2239
+ "epoch": 0.0757000474608448,
2240
+ "grad_norm": 1.0693997144699097,
2241
+ "learning_rate": 3.62e-06,
2242
+ "loss": 1.5375,
2243
+ "step": 319
2244
+ },
2245
+ {
2246
+ "epoch": 0.07593735168485999,
2247
+ "grad_norm": 0.5746883749961853,
2248
+ "learning_rate": 3.6000000000000003e-06,
2249
+ "loss": 1.4276,
2250
+ "step": 320
2251
+ },
2252
+ {
2253
+ "epoch": 0.07617465590887518,
2254
+ "grad_norm": 0.9793761968612671,
2255
+ "learning_rate": 3.58e-06,
2256
+ "loss": 0.9466,
2257
+ "step": 321
2258
+ },
2259
+ {
2260
+ "epoch": 0.07641196013289037,
2261
+ "grad_norm": 0.6006602048873901,
2262
+ "learning_rate": 3.5600000000000002e-06,
2263
+ "loss": 1.028,
2264
+ "step": 322
2265
+ },
2266
+ {
2267
+ "epoch": 0.07664926435690556,
2268
+ "grad_norm": 0.7533923983573914,
2269
+ "learning_rate": 3.54e-06,
2270
+ "loss": 1.9837,
2271
+ "step": 323
2272
+ },
2273
+ {
2274
+ "epoch": 0.07688656858092074,
2275
+ "grad_norm": 1.0659205913543701,
2276
+ "learning_rate": 3.52e-06,
2277
+ "loss": 1.4845,
2278
+ "step": 324
2279
+ },
2280
+ {
2281
+ "epoch": 0.07712387280493593,
2282
+ "grad_norm": 0.9382888078689575,
2283
+ "learning_rate": 3.5e-06,
2284
+ "loss": 1.8868,
2285
+ "step": 325
2286
+ },
2287
+ {
2288
+ "epoch": 0.07736117702895111,
2289
+ "grad_norm": 0.5263766050338745,
2290
+ "learning_rate": 3.48e-06,
2291
+ "loss": 1.6957,
2292
+ "step": 326
2293
+ },
2294
+ {
2295
+ "epoch": 0.0775984812529663,
2296
+ "grad_norm": 1.1845793724060059,
2297
+ "learning_rate": 3.46e-06,
2298
+ "loss": 1.7971,
2299
+ "step": 327
2300
+ },
2301
+ {
2302
+ "epoch": 0.07783578547698149,
2303
+ "grad_norm": 0.7983663082122803,
2304
+ "learning_rate": 3.44e-06,
2305
+ "loss": 1.7826,
2306
+ "step": 328
2307
+ },
2308
+ {
2309
+ "epoch": 0.07807308970099668,
2310
+ "grad_norm": 0.999782145023346,
2311
+ "learning_rate": 3.4200000000000007e-06,
2312
+ "loss": 1.4238,
2313
+ "step": 329
2314
+ },
2315
+ {
2316
+ "epoch": 0.07831039392501186,
2317
+ "grad_norm": 0.44168537855148315,
2318
+ "learning_rate": 3.4000000000000005e-06,
2319
+ "loss": 1.3762,
2320
+ "step": 330
2321
+ },
2322
+ {
2323
+ "epoch": 0.07854769814902705,
2324
+ "grad_norm": 3.451951026916504,
2325
+ "learning_rate": 3.3800000000000007e-06,
2326
+ "loss": 1.8961,
2327
+ "step": 331
2328
+ },
2329
+ {
2330
+ "epoch": 0.07878500237304224,
2331
+ "grad_norm": 1.2203079462051392,
2332
+ "learning_rate": 3.3600000000000004e-06,
2333
+ "loss": 2.4091,
2334
+ "step": 332
2335
+ },
2336
+ {
2337
+ "epoch": 0.07902230659705743,
2338
+ "grad_norm": 0.7909596562385559,
2339
+ "learning_rate": 3.3400000000000006e-06,
2340
+ "loss": 1.288,
2341
+ "step": 333
2342
+ },
2343
+ {
2344
+ "epoch": 0.07925961082107262,
2345
+ "grad_norm": 1.0289673805236816,
2346
+ "learning_rate": 3.3200000000000004e-06,
2347
+ "loss": 2.7401,
2348
+ "step": 334
2349
+ },
2350
+ {
2351
+ "epoch": 0.0794969150450878,
2352
+ "grad_norm": 1.550726056098938,
2353
+ "learning_rate": 3.3000000000000006e-06,
2354
+ "loss": 2.0937,
2355
+ "step": 335
2356
+ },
2357
+ {
2358
+ "epoch": 0.07973421926910298,
2359
+ "grad_norm": 0.9550947546958923,
2360
+ "learning_rate": 3.2800000000000004e-06,
2361
+ "loss": 1.3554,
2362
+ "step": 336
2363
+ },
2364
+ {
2365
+ "epoch": 0.07997152349311817,
2366
+ "grad_norm": 0.5482783913612366,
2367
+ "learning_rate": 3.2600000000000006e-06,
2368
+ "loss": 2.0287,
2369
+ "step": 337
2370
+ },
2371
+ {
2372
+ "epoch": 0.08020882771713336,
2373
+ "grad_norm": 1.071254014968872,
2374
+ "learning_rate": 3.2400000000000003e-06,
2375
+ "loss": 1.6622,
2376
+ "step": 338
2377
+ },
2378
+ {
2379
+ "epoch": 0.08044613194114855,
2380
+ "grad_norm": 1.0661407709121704,
2381
+ "learning_rate": 3.2200000000000005e-06,
2382
+ "loss": 0.9957,
2383
+ "step": 339
2384
+ },
2385
+ {
2386
+ "epoch": 0.08068343616516374,
2387
+ "grad_norm": 0.91053307056427,
2388
+ "learning_rate": 3.2000000000000003e-06,
2389
+ "loss": 0.8996,
2390
+ "step": 340
2391
+ },
2392
+ {
2393
+ "epoch": 0.08092074038917893,
2394
+ "grad_norm": 1.1073331832885742,
2395
+ "learning_rate": 3.1800000000000005e-06,
2396
+ "loss": 2.591,
2397
+ "step": 341
2398
+ },
2399
+ {
2400
+ "epoch": 0.08115804461319412,
2401
+ "grad_norm": 1.430469274520874,
2402
+ "learning_rate": 3.1600000000000002e-06,
2403
+ "loss": 2.6224,
2404
+ "step": 342
2405
+ },
2406
+ {
2407
+ "epoch": 0.08139534883720931,
2408
+ "grad_norm": 0.8856847882270813,
2409
+ "learning_rate": 3.1400000000000004e-06,
2410
+ "loss": 1.8773,
2411
+ "step": 343
2412
+ },
2413
+ {
2414
+ "epoch": 0.08163265306122448,
2415
+ "grad_norm": 0.7149975895881653,
2416
+ "learning_rate": 3.12e-06,
2417
+ "loss": 1.9744,
2418
+ "step": 344
2419
+ },
2420
+ {
2421
+ "epoch": 0.08186995728523967,
2422
+ "grad_norm": 1.156784176826477,
2423
+ "learning_rate": 3.1000000000000004e-06,
2424
+ "loss": 2.718,
2425
+ "step": 345
2426
+ },
2427
+ {
2428
+ "epoch": 0.08210726150925486,
2429
+ "grad_norm": 1.512473464012146,
2430
+ "learning_rate": 3.08e-06,
2431
+ "loss": 1.0725,
2432
+ "step": 346
2433
+ },
2434
+ {
2435
+ "epoch": 0.08234456573327005,
2436
+ "grad_norm": 0.5760719776153564,
2437
+ "learning_rate": 3.0600000000000003e-06,
2438
+ "loss": 1.3986,
2439
+ "step": 347
2440
+ },
2441
+ {
2442
+ "epoch": 0.08258186995728524,
2443
+ "grad_norm": 1.1456931829452515,
2444
+ "learning_rate": 3.04e-06,
2445
+ "loss": 2.6081,
2446
+ "step": 348
2447
+ },
2448
+ {
2449
+ "epoch": 0.08281917418130043,
2450
+ "grad_norm": 0.6848717927932739,
2451
+ "learning_rate": 3.0200000000000003e-06,
2452
+ "loss": 2.8127,
2453
+ "step": 349
2454
+ },
2455
+ {
2456
+ "epoch": 0.08305647840531562,
2457
+ "grad_norm": 1.2799925804138184,
2458
+ "learning_rate": 3e-06,
2459
+ "loss": 1.1195,
2460
+ "step": 350
2461
+ },
2462
+ {
2463
+ "epoch": 0.0832937826293308,
2464
+ "grad_norm": 1.1718430519104004,
2465
+ "learning_rate": 2.9800000000000003e-06,
2466
+ "loss": 1.6587,
2467
+ "step": 351
2468
+ },
2469
+ {
2470
+ "epoch": 0.083531086853346,
2471
+ "grad_norm": 0.5741757750511169,
2472
+ "learning_rate": 2.96e-06,
2473
+ "loss": 1.6265,
2474
+ "step": 352
2475
+ },
2476
+ {
2477
+ "epoch": 0.08376839107736117,
2478
+ "grad_norm": 0.8194566965103149,
2479
+ "learning_rate": 2.9400000000000002e-06,
2480
+ "loss": 1.2456,
2481
+ "step": 353
2482
+ },
2483
+ {
2484
+ "epoch": 0.08400569530137636,
2485
+ "grad_norm": 0.49410197138786316,
2486
+ "learning_rate": 2.92e-06,
2487
+ "loss": 1.5949,
2488
+ "step": 354
2489
+ },
2490
+ {
2491
+ "epoch": 0.08424299952539155,
2492
+ "grad_norm": 0.9407163858413696,
2493
+ "learning_rate": 2.9e-06,
2494
+ "loss": 1.1365,
2495
+ "step": 355
2496
+ },
2497
+ {
2498
+ "epoch": 0.08448030374940674,
2499
+ "grad_norm": 0.764671266078949,
2500
+ "learning_rate": 2.88e-06,
2501
+ "loss": 0.9231,
2502
+ "step": 356
2503
+ },
2504
+ {
2505
+ "epoch": 0.08471760797342193,
2506
+ "grad_norm": 0.6322979927062988,
2507
+ "learning_rate": 2.86e-06,
2508
+ "loss": 0.477,
2509
+ "step": 357
2510
+ },
2511
+ {
2512
+ "epoch": 0.08495491219743712,
2513
+ "grad_norm": 0.7397903800010681,
2514
+ "learning_rate": 2.84e-06,
2515
+ "loss": 1.2955,
2516
+ "step": 358
2517
+ },
2518
+ {
2519
+ "epoch": 0.0851922164214523,
2520
+ "grad_norm": 1.3564809560775757,
2521
+ "learning_rate": 2.82e-06,
2522
+ "loss": 1.0156,
2523
+ "step": 359
2524
+ },
2525
+ {
2526
+ "epoch": 0.0854295206454675,
2527
+ "grad_norm": 0.9099704027175903,
2528
+ "learning_rate": 2.8000000000000003e-06,
2529
+ "loss": 2.5199,
2530
+ "step": 360
2531
+ },
2532
+ {
2533
+ "epoch": 0.08566682486948268,
2534
+ "grad_norm": 0.596339225769043,
2535
+ "learning_rate": 2.7800000000000005e-06,
2536
+ "loss": 1.9148,
2537
+ "step": 361
2538
+ },
2539
+ {
2540
+ "epoch": 0.08590412909349786,
2541
+ "grad_norm": 0.8687242269515991,
2542
+ "learning_rate": 2.7600000000000003e-06,
2543
+ "loss": 1.6878,
2544
+ "step": 362
2545
+ },
2546
+ {
2547
+ "epoch": 0.08614143331751305,
2548
+ "grad_norm": 1.1088467836380005,
2549
+ "learning_rate": 2.7400000000000004e-06,
2550
+ "loss": 1.3883,
2551
+ "step": 363
2552
+ },
2553
+ {
2554
+ "epoch": 0.08637873754152824,
2555
+ "grad_norm": 0.26222410798072815,
2556
+ "learning_rate": 2.7200000000000002e-06,
2557
+ "loss": 1.6821,
2558
+ "step": 364
2559
+ },
2560
+ {
2561
+ "epoch": 0.08661604176554343,
2562
+ "grad_norm": 0.9348676800727844,
2563
+ "learning_rate": 2.7000000000000004e-06,
2564
+ "loss": 1.3761,
2565
+ "step": 365
2566
+ },
2567
+ {
2568
+ "epoch": 0.08685334598955861,
2569
+ "grad_norm": 0.6078274250030518,
2570
+ "learning_rate": 2.68e-06,
2571
+ "loss": 0.8237,
2572
+ "step": 366
2573
+ },
2574
+ {
2575
+ "epoch": 0.0870906502135738,
2576
+ "grad_norm": 2.2705135345458984,
2577
+ "learning_rate": 2.6600000000000004e-06,
2578
+ "loss": 2.1154,
2579
+ "step": 367
2580
+ },
2581
+ {
2582
+ "epoch": 0.08732795443758899,
2583
+ "grad_norm": 0.5769693851470947,
2584
+ "learning_rate": 2.64e-06,
2585
+ "loss": 0.7956,
2586
+ "step": 368
2587
+ },
2588
+ {
2589
+ "epoch": 0.08756525866160418,
2590
+ "grad_norm": 0.9362378120422363,
2591
+ "learning_rate": 2.6200000000000003e-06,
2592
+ "loss": 0.8264,
2593
+ "step": 369
2594
+ },
2595
+ {
2596
+ "epoch": 0.08780256288561937,
2597
+ "grad_norm": 1.3028643131256104,
2598
+ "learning_rate": 2.6e-06,
2599
+ "loss": 2.0418,
2600
+ "step": 370
2601
+ },
2602
+ {
2603
+ "epoch": 0.08803986710963455,
2604
+ "grad_norm": 1.020676612854004,
2605
+ "learning_rate": 2.5800000000000003e-06,
2606
+ "loss": 1.6203,
2607
+ "step": 371
2608
+ },
2609
+ {
2610
+ "epoch": 0.08827717133364973,
2611
+ "grad_norm": 0.5925426483154297,
2612
+ "learning_rate": 2.56e-06,
2613
+ "loss": 2.0088,
2614
+ "step": 372
2615
+ },
2616
+ {
2617
+ "epoch": 0.08851447555766492,
2618
+ "grad_norm": 0.9247467517852783,
2619
+ "learning_rate": 2.5400000000000002e-06,
2620
+ "loss": 1.4702,
2621
+ "step": 373
2622
+ },
2623
+ {
2624
+ "epoch": 0.08875177978168011,
2625
+ "grad_norm": 0.8355708718299866,
2626
+ "learning_rate": 2.52e-06,
2627
+ "loss": 1.511,
2628
+ "step": 374
2629
+ },
2630
+ {
2631
+ "epoch": 0.0889890840056953,
2632
+ "grad_norm": 0.8599538207054138,
2633
+ "learning_rate": 2.5e-06,
2634
+ "loss": 1.1829,
2635
+ "step": 375
2636
+ },
2637
+ {
2638
+ "epoch": 0.08922638822971049,
2639
+ "grad_norm": 0.7480601668357849,
2640
+ "learning_rate": 2.4800000000000004e-06,
2641
+ "loss": 0.5197,
2642
+ "step": 376
2643
+ },
2644
+ {
2645
+ "epoch": 0.08946369245372568,
2646
+ "grad_norm": 0.6589898467063904,
2647
+ "learning_rate": 2.46e-06,
2648
+ "loss": 1.8403,
2649
+ "step": 377
2650
+ },
2651
+ {
2652
+ "epoch": 0.08970099667774087,
2653
+ "grad_norm": 1.1234686374664307,
2654
+ "learning_rate": 2.4400000000000004e-06,
2655
+ "loss": 1.1402,
2656
+ "step": 378
2657
+ },
2658
+ {
2659
+ "epoch": 0.08993830090175606,
2660
+ "grad_norm": 0.4725819528102875,
2661
+ "learning_rate": 2.42e-06,
2662
+ "loss": 1.3948,
2663
+ "step": 379
2664
+ },
2665
+ {
2666
+ "epoch": 0.09017560512577123,
2667
+ "grad_norm": 0.579430878162384,
2668
+ "learning_rate": 2.4000000000000003e-06,
2669
+ "loss": 1.2087,
2670
+ "step": 380
2671
+ },
2672
+ {
2673
+ "epoch": 0.09041290934978642,
2674
+ "grad_norm": 0.9052660465240479,
2675
+ "learning_rate": 2.38e-06,
2676
+ "loss": 0.9077,
2677
+ "step": 381
2678
+ },
2679
+ {
2680
+ "epoch": 0.09065021357380161,
2681
+ "grad_norm": 0.5743672847747803,
2682
+ "learning_rate": 2.3600000000000003e-06,
2683
+ "loss": 2.4246,
2684
+ "step": 382
2685
+ },
2686
+ {
2687
+ "epoch": 0.0908875177978168,
2688
+ "grad_norm": 0.5894416570663452,
2689
+ "learning_rate": 2.3400000000000005e-06,
2690
+ "loss": 0.9541,
2691
+ "step": 383
2692
+ },
2693
+ {
2694
+ "epoch": 0.09112482202183199,
2695
+ "grad_norm": 0.5887079238891602,
2696
+ "learning_rate": 2.3200000000000002e-06,
2697
+ "loss": 1.5845,
2698
+ "step": 384
2699
+ },
2700
+ {
2701
+ "epoch": 0.09136212624584718,
2702
+ "grad_norm": 0.6027985215187073,
2703
+ "learning_rate": 2.3000000000000004e-06,
2704
+ "loss": 1.8314,
2705
+ "step": 385
2706
+ },
2707
+ {
2708
+ "epoch": 0.09159943046986237,
2709
+ "grad_norm": 1.0281093120574951,
2710
+ "learning_rate": 2.28e-06,
2711
+ "loss": 1.0648,
2712
+ "step": 386
2713
+ },
2714
+ {
2715
+ "epoch": 0.09183673469387756,
2716
+ "grad_norm": 1.0283626317977905,
2717
+ "learning_rate": 2.2600000000000004e-06,
2718
+ "loss": 2.3067,
2719
+ "step": 387
2720
+ },
2721
+ {
2722
+ "epoch": 0.09207403891789274,
2723
+ "grad_norm": 1.1482547521591187,
2724
+ "learning_rate": 2.24e-06,
2725
+ "loss": 1.8681,
2726
+ "step": 388
2727
+ },
2728
+ {
2729
+ "epoch": 0.09231134314190792,
2730
+ "grad_norm": 0.8530061841011047,
2731
+ "learning_rate": 2.2200000000000003e-06,
2732
+ "loss": 0.8857,
2733
+ "step": 389
2734
+ },
2735
+ {
2736
+ "epoch": 0.09254864736592311,
2737
+ "grad_norm": 0.836930513381958,
2738
+ "learning_rate": 2.2e-06,
2739
+ "loss": 1.8586,
2740
+ "step": 390
2741
+ },
2742
+ {
2743
+ "epoch": 0.0927859515899383,
2744
+ "grad_norm": 0.8930790424346924,
2745
+ "learning_rate": 2.1800000000000003e-06,
2746
+ "loss": 1.6977,
2747
+ "step": 391
2748
+ },
2749
+ {
2750
+ "epoch": 0.09302325581395349,
2751
+ "grad_norm": 0.4565262496471405,
2752
+ "learning_rate": 2.16e-06,
2753
+ "loss": 2.1539,
2754
+ "step": 392
2755
+ },
2756
+ {
2757
+ "epoch": 0.09326056003796868,
2758
+ "grad_norm": 1.635286569595337,
2759
+ "learning_rate": 2.1400000000000003e-06,
2760
+ "loss": 2.4099,
2761
+ "step": 393
2762
+ },
2763
+ {
2764
+ "epoch": 0.09349786426198387,
2765
+ "grad_norm": 1.1868668794631958,
2766
+ "learning_rate": 2.12e-06,
2767
+ "loss": 1.6283,
2768
+ "step": 394
2769
+ },
2770
+ {
2771
+ "epoch": 0.09373516848599905,
2772
+ "grad_norm": 1.4009878635406494,
2773
+ "learning_rate": 2.1000000000000002e-06,
2774
+ "loss": 1.4061,
2775
+ "step": 395
2776
+ },
2777
+ {
2778
+ "epoch": 0.09397247271001424,
2779
+ "grad_norm": 0.46099644899368286,
2780
+ "learning_rate": 2.08e-06,
2781
+ "loss": 0.4871,
2782
+ "step": 396
2783
+ },
2784
+ {
2785
+ "epoch": 0.09420977693402943,
2786
+ "grad_norm": 0.7012650370597839,
2787
+ "learning_rate": 2.06e-06,
2788
+ "loss": 1.5595,
2789
+ "step": 397
2790
+ },
2791
+ {
2792
+ "epoch": 0.09444708115804461,
2793
+ "grad_norm": 0.7766276001930237,
2794
+ "learning_rate": 2.04e-06,
2795
+ "loss": 1.0104,
2796
+ "step": 398
2797
+ },
2798
+ {
2799
+ "epoch": 0.0946843853820598,
2800
+ "grad_norm": 1.029155969619751,
2801
+ "learning_rate": 2.02e-06,
2802
+ "loss": 2.6892,
2803
+ "step": 399
2804
+ },
2805
+ {
2806
+ "epoch": 0.09492168960607499,
2807
+ "grad_norm": 1.6015249490737915,
2808
+ "learning_rate": 2.0000000000000003e-06,
2809
+ "loss": 1.7859,
2810
+ "step": 400
2811
+ },
2812
+ {
2813
+ "epoch": 0.09515899383009017,
2814
+ "grad_norm": 0.3838267922401428,
2815
+ "learning_rate": 1.98e-06,
2816
+ "loss": 1.5704,
2817
+ "step": 401
2818
+ },
2819
+ {
2820
+ "epoch": 0.09539629805410536,
2821
+ "grad_norm": 0.6478832364082336,
2822
+ "learning_rate": 1.9600000000000003e-06,
2823
+ "loss": 1.2718,
2824
+ "step": 402
2825
+ },
2826
+ {
2827
+ "epoch": 0.09563360227812055,
2828
+ "grad_norm": 0.9483569860458374,
2829
+ "learning_rate": 1.94e-06,
2830
+ "loss": 1.9627,
2831
+ "step": 403
2832
+ },
2833
+ {
2834
+ "epoch": 0.09587090650213574,
2835
+ "grad_norm": 0.5800157785415649,
2836
+ "learning_rate": 1.9200000000000003e-06,
2837
+ "loss": 0.9114,
2838
+ "step": 404
2839
+ },
2840
+ {
2841
+ "epoch": 0.09610821072615093,
2842
+ "grad_norm": 0.5951704978942871,
2843
+ "learning_rate": 1.9000000000000002e-06,
2844
+ "loss": 1.2684,
2845
+ "step": 405
2846
+ },
2847
+ {
2848
+ "epoch": 0.09634551495016612,
2849
+ "grad_norm": 0.7555820345878601,
2850
+ "learning_rate": 1.8800000000000002e-06,
2851
+ "loss": 1.7506,
2852
+ "step": 406
2853
+ },
2854
+ {
2855
+ "epoch": 0.0965828191741813,
2856
+ "grad_norm": 1.0840675830841064,
2857
+ "learning_rate": 1.8600000000000002e-06,
2858
+ "loss": 2.3729,
2859
+ "step": 407
2860
+ },
2861
+ {
2862
+ "epoch": 0.09682012339819648,
2863
+ "grad_norm": 1.0009734630584717,
2864
+ "learning_rate": 1.8400000000000002e-06,
2865
+ "loss": 1.0095,
2866
+ "step": 408
2867
+ },
2868
+ {
2869
+ "epoch": 0.09705742762221167,
2870
+ "grad_norm": 0.8436226844787598,
2871
+ "learning_rate": 1.8200000000000002e-06,
2872
+ "loss": 1.6425,
2873
+ "step": 409
2874
+ },
2875
+ {
2876
+ "epoch": 0.09729473184622686,
2877
+ "grad_norm": 0.6967753767967224,
2878
+ "learning_rate": 1.8000000000000001e-06,
2879
+ "loss": 2.2186,
2880
+ "step": 410
2881
+ },
2882
+ {
2883
+ "epoch": 0.09753203607024205,
2884
+ "grad_norm": 0.9439252614974976,
2885
+ "learning_rate": 1.7800000000000001e-06,
2886
+ "loss": 1.0762,
2887
+ "step": 411
2888
+ },
2889
+ {
2890
+ "epoch": 0.09776934029425724,
2891
+ "grad_norm": 1.6090588569641113,
2892
+ "learning_rate": 1.76e-06,
2893
+ "loss": 1.6685,
2894
+ "step": 412
2895
+ },
2896
+ {
2897
+ "epoch": 0.09800664451827243,
2898
+ "grad_norm": 0.6204804182052612,
2899
+ "learning_rate": 1.74e-06,
2900
+ "loss": 0.7148,
2901
+ "step": 413
2902
+ },
2903
+ {
2904
+ "epoch": 0.09824394874228762,
2905
+ "grad_norm": 0.9542770385742188,
2906
+ "learning_rate": 1.72e-06,
2907
+ "loss": 0.7676,
2908
+ "step": 414
2909
+ },
2910
+ {
2911
+ "epoch": 0.0984812529663028,
2912
+ "grad_norm": 1.0385842323303223,
2913
+ "learning_rate": 1.7000000000000002e-06,
2914
+ "loss": 1.6284,
2915
+ "step": 415
2916
+ },
2917
+ {
2918
+ "epoch": 0.09871855719031798,
2919
+ "grad_norm": 0.643661379814148,
2920
+ "learning_rate": 1.6800000000000002e-06,
2921
+ "loss": 0.9284,
2922
+ "step": 416
2923
+ },
2924
+ {
2925
+ "epoch": 0.09895586141433317,
2926
+ "grad_norm": 0.7566413283348083,
2927
+ "learning_rate": 1.6600000000000002e-06,
2928
+ "loss": 1.8893,
2929
+ "step": 417
2930
+ },
2931
+ {
2932
+ "epoch": 0.09919316563834836,
2933
+ "grad_norm": 0.805566132068634,
2934
+ "learning_rate": 1.6400000000000002e-06,
2935
+ "loss": 1.8842,
2936
+ "step": 418
2937
+ },
2938
+ {
2939
+ "epoch": 0.09943046986236355,
2940
+ "grad_norm": 0.503933310508728,
2941
+ "learning_rate": 1.6200000000000002e-06,
2942
+ "loss": 1.4513,
2943
+ "step": 419
2944
+ },
2945
+ {
2946
+ "epoch": 0.09966777408637874,
2947
+ "grad_norm": 2.3548262119293213,
2948
+ "learning_rate": 1.6000000000000001e-06,
2949
+ "loss": 1.381,
2950
+ "step": 420
2951
+ },
2952
+ {
2953
+ "epoch": 0.09990507831039393,
2954
+ "grad_norm": 1.1577256917953491,
2955
+ "learning_rate": 1.5800000000000001e-06,
2956
+ "loss": 2.0068,
2957
+ "step": 421
2958
+ },
2959
+ {
2960
+ "epoch": 0.10014238253440912,
2961
+ "grad_norm": 0.8693385124206543,
2962
+ "learning_rate": 1.56e-06,
2963
+ "loss": 1.38,
2964
+ "step": 422
2965
+ },
2966
+ {
2967
+ "epoch": 0.1003796867584243,
2968
+ "grad_norm": 0.27834704518318176,
2969
+ "learning_rate": 1.54e-06,
2970
+ "loss": 0.3649,
2971
+ "step": 423
2972
+ },
2973
+ {
2974
+ "epoch": 0.1006169909824395,
2975
+ "grad_norm": 0.6906237006187439,
2976
+ "learning_rate": 1.52e-06,
2977
+ "loss": 2.2498,
2978
+ "step": 424
2979
+ },
2980
+ {
2981
+ "epoch": 0.10085429520645467,
2982
+ "grad_norm": 1.5801548957824707,
2983
+ "learning_rate": 1.5e-06,
2984
+ "loss": 1.6734,
2985
+ "step": 425
2986
+ },
2987
+ {
2988
+ "epoch": 0.10109159943046986,
2989
+ "grad_norm": 0.6525102853775024,
2990
+ "learning_rate": 1.48e-06,
2991
+ "loss": 1.7765,
2992
+ "step": 426
2993
+ },
2994
+ {
2995
+ "epoch": 0.10132890365448505,
2996
+ "grad_norm": 0.6370388865470886,
2997
+ "learning_rate": 1.46e-06,
2998
+ "loss": 1.497,
2999
+ "step": 427
3000
+ },
3001
+ {
3002
+ "epoch": 0.10156620787850024,
3003
+ "grad_norm": 0.9169662594795227,
3004
+ "learning_rate": 1.44e-06,
3005
+ "loss": 1.7245,
3006
+ "step": 428
3007
+ },
3008
+ {
3009
+ "epoch": 0.10180351210251543,
3010
+ "grad_norm": 0.8349008560180664,
3011
+ "learning_rate": 1.42e-06,
3012
+ "loss": 1.2525,
3013
+ "step": 429
3014
+ },
3015
+ {
3016
+ "epoch": 0.10204081632653061,
3017
+ "grad_norm": 0.5627273917198181,
3018
+ "learning_rate": 1.4000000000000001e-06,
3019
+ "loss": 1.5604,
3020
+ "step": 430
3021
+ },
3022
+ {
3023
+ "epoch": 0.1022781205505458,
3024
+ "grad_norm": 0.6167902946472168,
3025
+ "learning_rate": 1.3800000000000001e-06,
3026
+ "loss": 0.4832,
3027
+ "step": 431
3028
+ },
3029
+ {
3030
+ "epoch": 0.10251542477456099,
3031
+ "grad_norm": 0.6913139224052429,
3032
+ "learning_rate": 1.3600000000000001e-06,
3033
+ "loss": 0.4277,
3034
+ "step": 432
3035
+ },
3036
+ {
3037
+ "epoch": 0.10275272899857618,
3038
+ "grad_norm": 0.9053479433059692,
3039
+ "learning_rate": 1.34e-06,
3040
+ "loss": 1.1996,
3041
+ "step": 433
3042
+ },
3043
+ {
3044
+ "epoch": 0.10299003322259136,
3045
+ "grad_norm": 0.5853822827339172,
3046
+ "learning_rate": 1.32e-06,
3047
+ "loss": 1.5833,
3048
+ "step": 434
3049
+ },
3050
+ {
3051
+ "epoch": 0.10322733744660655,
3052
+ "grad_norm": 0.6833449602127075,
3053
+ "learning_rate": 1.3e-06,
3054
+ "loss": 1.5597,
3055
+ "step": 435
3056
+ },
3057
+ {
3058
+ "epoch": 0.10346464167062173,
3059
+ "grad_norm": 0.48403409123420715,
3060
+ "learning_rate": 1.28e-06,
3061
+ "loss": 1.7934,
3062
+ "step": 436
3063
+ },
3064
+ {
3065
+ "epoch": 0.10370194589463692,
3066
+ "grad_norm": 0.8773030638694763,
3067
+ "learning_rate": 1.26e-06,
3068
+ "loss": 0.5695,
3069
+ "step": 437
3070
+ },
3071
+ {
3072
+ "epoch": 0.10393925011865211,
3073
+ "grad_norm": 0.6054906249046326,
3074
+ "learning_rate": 1.2400000000000002e-06,
3075
+ "loss": 2.3307,
3076
+ "step": 438
3077
+ },
3078
+ {
3079
+ "epoch": 0.1041765543426673,
3080
+ "grad_norm": 0.6956624388694763,
3081
+ "learning_rate": 1.2200000000000002e-06,
3082
+ "loss": 1.0714,
3083
+ "step": 439
3084
+ },
3085
+ {
3086
+ "epoch": 0.10441385856668249,
3087
+ "grad_norm": 1.519985556602478,
3088
+ "learning_rate": 1.2000000000000002e-06,
3089
+ "loss": 1.695,
3090
+ "step": 440
3091
+ },
3092
+ {
3093
+ "epoch": 0.10465116279069768,
3094
+ "grad_norm": 1.0551714897155762,
3095
+ "learning_rate": 1.1800000000000001e-06,
3096
+ "loss": 2.457,
3097
+ "step": 441
3098
+ },
3099
+ {
3100
+ "epoch": 0.10488846701471286,
3101
+ "grad_norm": 0.7493646740913391,
3102
+ "learning_rate": 1.1600000000000001e-06,
3103
+ "loss": 1.965,
3104
+ "step": 442
3105
+ },
3106
+ {
3107
+ "epoch": 0.10512577123872804,
3108
+ "grad_norm": 0.7913329601287842,
3109
+ "learning_rate": 1.14e-06,
3110
+ "loss": 1.5939,
3111
+ "step": 443
3112
+ },
3113
+ {
3114
+ "epoch": 0.10536307546274323,
3115
+ "grad_norm": 1.10833740234375,
3116
+ "learning_rate": 1.12e-06,
3117
+ "loss": 2.1043,
3118
+ "step": 444
3119
+ },
3120
+ {
3121
+ "epoch": 0.10560037968675842,
3122
+ "grad_norm": 0.8675681948661804,
3123
+ "learning_rate": 1.1e-06,
3124
+ "loss": 1.6333,
3125
+ "step": 445
3126
+ },
3127
+ {
3128
+ "epoch": 0.10583768391077361,
3129
+ "grad_norm": 0.8735470771789551,
3130
+ "learning_rate": 1.08e-06,
3131
+ "loss": 2.1604,
3132
+ "step": 446
3133
+ },
3134
+ {
3135
+ "epoch": 0.1060749881347888,
3136
+ "grad_norm": 0.9015608429908752,
3137
+ "learning_rate": 1.06e-06,
3138
+ "loss": 0.9548,
3139
+ "step": 447
3140
+ },
3141
+ {
3142
+ "epoch": 0.10631229235880399,
3143
+ "grad_norm": 0.7339662313461304,
3144
+ "learning_rate": 1.04e-06,
3145
+ "loss": 1.3065,
3146
+ "step": 448
3147
+ },
3148
+ {
3149
+ "epoch": 0.10654959658281918,
3150
+ "grad_norm": 0.5532211661338806,
3151
+ "learning_rate": 1.02e-06,
3152
+ "loss": 1.8098,
3153
+ "step": 449
3154
+ },
3155
+ {
3156
+ "epoch": 0.10678690080683437,
3157
+ "grad_norm": 0.8225467801094055,
3158
+ "learning_rate": 1.0000000000000002e-06,
3159
+ "loss": 2.2966,
3160
+ "step": 450
3161
+ },
3162
+ {
3163
+ "epoch": 0.10702420503084954,
3164
+ "grad_norm": 0.5000866651535034,
3165
+ "learning_rate": 9.800000000000001e-07,
3166
+ "loss": 0.2921,
3167
+ "step": 451
3168
+ },
3169
+ {
3170
+ "epoch": 0.10726150925486473,
3171
+ "grad_norm": 1.0391067266464233,
3172
+ "learning_rate": 9.600000000000001e-07,
3173
+ "loss": 2.2974,
3174
+ "step": 452
3175
+ },
3176
+ {
3177
+ "epoch": 0.10749881347887992,
3178
+ "grad_norm": 0.685451865196228,
3179
+ "learning_rate": 9.400000000000001e-07,
3180
+ "loss": 2.0784,
3181
+ "step": 453
3182
+ },
3183
+ {
3184
+ "epoch": 0.10773611770289511,
3185
+ "grad_norm": 0.5864785313606262,
3186
+ "learning_rate": 9.200000000000001e-07,
3187
+ "loss": 2.2062,
3188
+ "step": 454
3189
+ },
3190
+ {
3191
+ "epoch": 0.1079734219269103,
3192
+ "grad_norm": 0.6014403104782104,
3193
+ "learning_rate": 9.000000000000001e-07,
3194
+ "loss": 0.8348,
3195
+ "step": 455
3196
+ },
3197
+ {
3198
+ "epoch": 0.10821072615092549,
3199
+ "grad_norm": 1.6633024215698242,
3200
+ "learning_rate": 8.8e-07,
3201
+ "loss": 1.2606,
3202
+ "step": 456
3203
+ },
3204
+ {
3205
+ "epoch": 0.10844803037494068,
3206
+ "grad_norm": 0.7936999797821045,
3207
+ "learning_rate": 8.6e-07,
3208
+ "loss": 2.0804,
3209
+ "step": 457
3210
+ },
3211
+ {
3212
+ "epoch": 0.10868533459895587,
3213
+ "grad_norm": 0.22339628636837006,
3214
+ "learning_rate": 8.400000000000001e-07,
3215
+ "loss": 2.1541,
3216
+ "step": 458
3217
+ },
3218
+ {
3219
+ "epoch": 0.10892263882297105,
3220
+ "grad_norm": 0.6188535690307617,
3221
+ "learning_rate": 8.200000000000001e-07,
3222
+ "loss": 1.4218,
3223
+ "step": 459
3224
+ },
3225
+ {
3226
+ "epoch": 0.10915994304698623,
3227
+ "grad_norm": 0.8845232129096985,
3228
+ "learning_rate": 8.000000000000001e-07,
3229
+ "loss": 1.4453,
3230
+ "step": 460
3231
+ },
3232
+ {
3233
+ "epoch": 0.10939724727100142,
3234
+ "grad_norm": 0.5522229671478271,
3235
+ "learning_rate": 7.8e-07,
3236
+ "loss": 1.4562,
3237
+ "step": 461
3238
+ },
3239
+ {
3240
+ "epoch": 0.10963455149501661,
3241
+ "grad_norm": 0.49053800106048584,
3242
+ "learning_rate": 7.6e-07,
3243
+ "loss": 0.6361,
3244
+ "step": 462
3245
+ },
3246
+ {
3247
+ "epoch": 0.1098718557190318,
3248
+ "grad_norm": 0.5572338104248047,
3249
+ "learning_rate": 7.4e-07,
3250
+ "loss": 0.8394,
3251
+ "step": 463
3252
+ },
3253
+ {
3254
+ "epoch": 0.11010915994304699,
3255
+ "grad_norm": 0.9283513426780701,
3256
+ "learning_rate": 7.2e-07,
3257
+ "loss": 1.6316,
3258
+ "step": 464
3259
+ },
3260
+ {
3261
+ "epoch": 0.11034646416706217,
3262
+ "grad_norm": 0.8812064528465271,
3263
+ "learning_rate": 7.000000000000001e-07,
3264
+ "loss": 1.2302,
3265
+ "step": 465
3266
+ },
3267
+ {
3268
+ "epoch": 0.11058376839107736,
3269
+ "grad_norm": 0.7392125129699707,
3270
+ "learning_rate": 6.800000000000001e-07,
3271
+ "loss": 0.8159,
3272
+ "step": 466
3273
+ },
3274
+ {
3275
+ "epoch": 0.11082107261509255,
3276
+ "grad_norm": 0.7616608738899231,
3277
+ "learning_rate": 6.6e-07,
3278
+ "loss": 1.0404,
3279
+ "step": 467
3280
+ },
3281
+ {
3282
+ "epoch": 0.11105837683910774,
3283
+ "grad_norm": 0.7175336480140686,
3284
+ "learning_rate": 6.4e-07,
3285
+ "loss": 1.7509,
3286
+ "step": 468
3287
+ },
3288
+ {
3289
+ "epoch": 0.11129568106312292,
3290
+ "grad_norm": 0.7180752754211426,
3291
+ "learning_rate": 6.200000000000001e-07,
3292
+ "loss": 0.9795,
3293
+ "step": 469
3294
+ },
3295
+ {
3296
+ "epoch": 0.1115329852871381,
3297
+ "grad_norm": 0.875347912311554,
3298
+ "learning_rate": 6.000000000000001e-07,
3299
+ "loss": 1.8271,
3300
+ "step": 470
3301
+ },
3302
+ {
3303
+ "epoch": 0.1117702895111533,
3304
+ "grad_norm": 0.6006546020507812,
3305
+ "learning_rate": 5.800000000000001e-07,
3306
+ "loss": 2.0545,
3307
+ "step": 471
3308
+ },
3309
+ {
3310
+ "epoch": 0.11200759373516848,
3311
+ "grad_norm": 1.1124011278152466,
3312
+ "learning_rate": 5.6e-07,
3313
+ "loss": 2.2172,
3314
+ "step": 472
3315
+ },
3316
+ {
3317
+ "epoch": 0.11224489795918367,
3318
+ "grad_norm": 1.5857324600219727,
3319
+ "learning_rate": 5.4e-07,
3320
+ "loss": 1.2116,
3321
+ "step": 473
3322
+ },
3323
+ {
3324
+ "epoch": 0.11248220218319886,
3325
+ "grad_norm": 0.6902075409889221,
3326
+ "learning_rate": 5.2e-07,
3327
+ "loss": 2.0822,
3328
+ "step": 474
3329
+ },
3330
+ {
3331
+ "epoch": 0.11271950640721405,
3332
+ "grad_norm": 0.9990330934524536,
3333
+ "learning_rate": 5.000000000000001e-07,
3334
+ "loss": 1.9775,
3335
+ "step": 475
3336
+ },
3337
+ {
3338
+ "epoch": 0.11295681063122924,
3339
+ "grad_norm": 1.281016230583191,
3340
+ "learning_rate": 4.800000000000001e-07,
3341
+ "loss": 1.9021,
3342
+ "step": 476
3343
+ },
3344
+ {
3345
+ "epoch": 0.11319411485524443,
3346
+ "grad_norm": 0.3223126232624054,
3347
+ "learning_rate": 4.6000000000000004e-07,
3348
+ "loss": 1.8383,
3349
+ "step": 477
3350
+ },
3351
+ {
3352
+ "epoch": 0.1134314190792596,
3353
+ "grad_norm": 1.1316232681274414,
3354
+ "learning_rate": 4.4e-07,
3355
+ "loss": 1.6243,
3356
+ "step": 478
3357
+ },
3358
+ {
3359
+ "epoch": 0.1136687233032748,
3360
+ "grad_norm": 0.6887989640235901,
3361
+ "learning_rate": 4.2000000000000006e-07,
3362
+ "loss": 1.5909,
3363
+ "step": 479
3364
+ },
3365
+ {
3366
+ "epoch": 0.11390602752728998,
3367
+ "grad_norm": 0.8150919675827026,
3368
+ "learning_rate": 4.0000000000000003e-07,
3369
+ "loss": 2.1949,
3370
+ "step": 480
3371
+ },
3372
+ {
3373
+ "epoch": 0.11414333175130517,
3374
+ "grad_norm": 0.6823549866676331,
3375
+ "learning_rate": 3.8e-07,
3376
+ "loss": 2.095,
3377
+ "step": 481
3378
+ },
3379
+ {
3380
+ "epoch": 0.11438063597532036,
3381
+ "grad_norm": 1.052901268005371,
3382
+ "learning_rate": 3.6e-07,
3383
+ "loss": 1.7599,
3384
+ "step": 482
3385
+ },
3386
+ {
3387
+ "epoch": 0.11461794019933555,
3388
+ "grad_norm": 1.048052430152893,
3389
+ "learning_rate": 3.4000000000000003e-07,
3390
+ "loss": 2.1143,
3391
+ "step": 483
3392
+ },
3393
+ {
3394
+ "epoch": 0.11485524442335074,
3395
+ "grad_norm": 1.2748647928237915,
3396
+ "learning_rate": 3.2e-07,
3397
+ "loss": 1.8183,
3398
+ "step": 484
3399
+ },
3400
+ {
3401
+ "epoch": 0.11509254864736593,
3402
+ "grad_norm": 1.2471035718917847,
3403
+ "learning_rate": 3.0000000000000004e-07,
3404
+ "loss": 1.8442,
3405
+ "step": 485
3406
+ },
3407
+ {
3408
+ "epoch": 0.11532985287138112,
3409
+ "grad_norm": 0.46195486187934875,
3410
+ "learning_rate": 2.8e-07,
3411
+ "loss": 1.1617,
3412
+ "step": 486
3413
+ },
3414
+ {
3415
+ "epoch": 0.11556715709539629,
3416
+ "grad_norm": 0.6743305325508118,
3417
+ "learning_rate": 2.6e-07,
3418
+ "loss": 1.4748,
3419
+ "step": 487
3420
+ },
3421
+ {
3422
+ "epoch": 0.11580446131941148,
3423
+ "grad_norm": 1.0564024448394775,
3424
+ "learning_rate": 2.4000000000000003e-07,
3425
+ "loss": 1.5775,
3426
+ "step": 488
3427
+ },
3428
+ {
3429
+ "epoch": 0.11604176554342667,
3430
+ "grad_norm": 0.6965152025222778,
3431
+ "learning_rate": 2.2e-07,
3432
+ "loss": 2.5251,
3433
+ "step": 489
3434
+ },
3435
+ {
3436
+ "epoch": 0.11627906976744186,
3437
+ "grad_norm": 0.8700504899024963,
3438
+ "learning_rate": 2.0000000000000002e-07,
3439
+ "loss": 1.6123,
3440
+ "step": 490
3441
+ },
3442
+ {
3443
+ "epoch": 0.11651637399145705,
3444
+ "grad_norm": 0.7157378196716309,
3445
+ "learning_rate": 1.8e-07,
3446
+ "loss": 1.7864,
3447
+ "step": 491
3448
+ },
3449
+ {
3450
+ "epoch": 0.11675367821547224,
3451
+ "grad_norm": 0.9464967250823975,
3452
+ "learning_rate": 1.6e-07,
3453
+ "loss": 2.0064,
3454
+ "step": 492
3455
+ },
3456
+ {
3457
+ "epoch": 0.11699098243948743,
3458
+ "grad_norm": 0.734511137008667,
3459
+ "learning_rate": 1.4e-07,
3460
+ "loss": 0.8449,
3461
+ "step": 493
3462
+ },
3463
+ {
3464
+ "epoch": 0.11722828666350261,
3465
+ "grad_norm": 1.0827577114105225,
3466
+ "learning_rate": 1.2000000000000002e-07,
3467
+ "loss": 1.8062,
3468
+ "step": 494
3469
+ },
3470
+ {
3471
+ "epoch": 0.1174655908875178,
3472
+ "grad_norm": 1.1060395240783691,
3473
+ "learning_rate": 1.0000000000000001e-07,
3474
+ "loss": 1.9935,
3475
+ "step": 495
3476
+ },
3477
+ {
3478
+ "epoch": 0.11770289511153298,
3479
+ "grad_norm": 0.9821236729621887,
3480
+ "learning_rate": 8e-08,
3481
+ "loss": 1.1501,
3482
+ "step": 496
3483
+ },
3484
+ {
3485
+ "epoch": 0.11794019933554817,
3486
+ "grad_norm": 0.46705755591392517,
3487
+ "learning_rate": 6.000000000000001e-08,
3488
+ "loss": 2.2727,
3489
+ "step": 497
3490
+ },
3491
+ {
3492
+ "epoch": 0.11817750355956336,
3493
+ "grad_norm": 0.7528263926506042,
3494
+ "learning_rate": 4e-08,
3495
+ "loss": 2.2196,
3496
+ "step": 498
3497
+ },
3498
+ {
3499
+ "epoch": 0.11841480778357855,
3500
+ "grad_norm": 1.029765009880066,
3501
+ "learning_rate": 2e-08,
3502
+ "loss": 1.3983,
3503
+ "step": 499
3504
+ },
3505
+ {
3506
+ "epoch": 0.11865211200759374,
3507
+ "grad_norm": 0.9231687188148499,
3508
+ "learning_rate": 9.000000000000001e-11,
3509
+ "loss": 0.7083,
3510
+ "step": 500
3511
+ },
3512
+ {
3513
+ "epoch": 0.11865211200759374,
3514
+ "step": 500,
3515
+ "total_flos": 8.201567207424e+16,
3516
+ "train_loss": 1.6423830435276032,
3517
+ "train_runtime": 704.5977,
3518
+ "train_samples_per_second": 1.419,
3519
+ "train_steps_per_second": 0.71
3520
+ }
3521
+ ],
3522
+ "logging_steps": 1.0,
3523
+ "max_steps": 500,
3524
+ "num_input_tokens_seen": 0,
3525
+ "num_train_epochs": 1,
3526
+ "save_steps": 100,
3527
+ "stateful_callbacks": {
3528
+ "TrainerControl": {
3529
+ "args": {
3530
+ "should_epoch_stop": false,
3531
+ "should_evaluate": false,
3532
+ "should_log": false,
3533
+ "should_save": true,
3534
+ "should_training_stop": true
3535
+ },
3536
+ "attributes": {}
3537
+ }
3538
+ },
3539
+ "total_flos": 8.201567207424e+16,
3540
+ "train_batch_size": 1,
3541
+ "trial_name": null,
3542
+ "trial_params": null
3543
+ }