diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,87278 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12464,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.7782012296174035,
+      "learning_rate": 2.6737967914438503e-08,
+      "loss": 1.3154,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.401347519306018,
+      "learning_rate": 5.3475935828877005e-08,
+      "loss": 1.372,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.5985544146511925,
+      "learning_rate": 8.021390374331552e-08,
+      "loss": 1.3674,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.416751268109562,
+      "learning_rate": 1.0695187165775401e-07,
+      "loss": 1.3505,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.499206370703229,
+      "learning_rate": 1.3368983957219251e-07,
+      "loss": 1.4123,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.857202154933578,
+      "learning_rate": 1.6042780748663104e-07,
+      "loss": 1.4046,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.066284844533282,
+      "learning_rate": 1.8716577540106952e-07,
+      "loss": 1.4246,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.130085775570846,
+      "learning_rate": 2.1390374331550802e-07,
+      "loss": 1.2694,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.260778613597487,
+      "learning_rate": 2.4064171122994655e-07,
+      "loss": 1.3858,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.2677734603592175,
+      "learning_rate": 2.6737967914438503e-07,
+      "loss": 1.3965,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.198162166213362,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 1.4326,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.548656471974227,
+      "learning_rate": 3.208556149732621e-07,
+      "loss": 1.4168,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.9047660966760445,
+      "learning_rate": 3.4759358288770056e-07,
+      "loss": 1.3972,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.068196959142517,
+      "learning_rate": 3.7433155080213904e-07,
+      "loss": 1.4404,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.100133406123299,
+      "learning_rate": 4.0106951871657757e-07,
+      "loss": 1.37,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.38010898431197,
+      "learning_rate": 4.2780748663101604e-07,
+      "loss": 1.4224,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.970892738698906,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 1.2825,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.659361259195055,
+      "learning_rate": 4.812834224598931e-07,
+      "loss": 1.3848,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.351085437989884,
+      "learning_rate": 5.080213903743316e-07,
+      "loss": 1.5052,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.48655672103779,
+      "learning_rate": 5.347593582887701e-07,
+      "loss": 1.3905,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.434512621861015,
+      "learning_rate": 5.614973262032086e-07,
+      "loss": 1.3759,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.9997292246294105,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 1.3716,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.23579727651579,
+      "learning_rate": 6.149732620320856e-07,
+      "loss": 1.3089,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.287208076755039,
+      "learning_rate": 6.417112299465242e-07,
+      "loss": 1.4492,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.8678880297581495,
+      "learning_rate": 6.684491978609627e-07,
+      "loss": 1.3663,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.117495029815131,
+      "learning_rate": 6.951871657754011e-07,
+      "loss": 1.3751,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.902439861327881,
+      "learning_rate": 7.219251336898397e-07,
+      "loss": 1.3404,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.394737690984814,
+      "learning_rate": 7.486631016042781e-07,
+      "loss": 1.4086,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.40858594970431,
+      "learning_rate": 7.754010695187167e-07,
+      "loss": 1.385,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.849514966612606,
+      "learning_rate": 8.021390374331551e-07,
+      "loss": 1.3826,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.4908123251624765,
+      "learning_rate": 8.288770053475937e-07,
+      "loss": 1.3148,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.81061860321785,
+      "learning_rate": 8.556149732620321e-07,
+      "loss": 1.3841,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.012893559459193,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 1.3631,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.76203032889139,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 1.3023,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.576812547841188,
+      "learning_rate": 9.358288770053477e-07,
+      "loss": 1.3222,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.660928583706245,
+      "learning_rate": 9.625668449197862e-07,
+      "loss": 1.3277,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.714892183621894,
+      "learning_rate": 9.893048128342248e-07,
+      "loss": 1.3456,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.303008621046351,
+      "learning_rate": 1.0160427807486633e-06,
+      "loss": 1.2667,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.665456931966554,
+      "learning_rate": 1.0427807486631017e-06,
+      "loss": 1.3254,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.145129369536635,
+      "learning_rate": 1.0695187165775401e-06,
+      "loss": 1.3165,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.324919286220631,
+      "learning_rate": 1.0962566844919787e-06,
+      "loss": 1.3135,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8932822587645965,
+      "learning_rate": 1.1229946524064172e-06,
+      "loss": 1.3801,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7533254164181646,
+      "learning_rate": 1.1497326203208558e-06,
+      "loss": 1.3273,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6310745974008025,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 1.2343,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.316246853177768,
+      "learning_rate": 1.2032085561497326e-06,
+      "loss": 1.2036,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4306426605255624,
+      "learning_rate": 1.2299465240641713e-06,
+      "loss": 1.2371,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2239862034255538,
+      "learning_rate": 1.2566844919786097e-06,
+      "loss": 1.2778,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0936291849127606,
+      "learning_rate": 1.2834224598930483e-06,
+      "loss": 1.2062,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2666271086267358,
+      "learning_rate": 1.3101604278074868e-06,
+      "loss": 1.1664,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.800576975918958,
+      "learning_rate": 1.3368983957219254e-06,
+      "loss": 1.2244,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4502554235688105,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 1.1918,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3007937184620317,
+      "learning_rate": 1.3903743315508022e-06,
+      "loss": 1.1664,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1931360501804686,
+      "learning_rate": 1.4171122994652409e-06,
+      "loss": 1.2175,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.111515784653167,
+      "learning_rate": 1.4438502673796793e-06,
+      "loss": 1.2425,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.144436947205643,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 1.1785,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1765083947098374,
+      "learning_rate": 1.4973262032085562e-06,
+      "loss": 1.2607,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.07671033670113,
+      "learning_rate": 1.5240641711229948e-06,
+      "loss": 1.2282,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5725341201883896,
+      "learning_rate": 1.5508021390374334e-06,
+      "loss": 1.3187,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9335282059481156,
+      "learning_rate": 1.5775401069518716e-06,
+      "loss": 1.195,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.009007179067198,
+      "learning_rate": 1.6042780748663103e-06,
+      "loss": 1.1317,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9694706636693953,
+      "learning_rate": 1.631016042780749e-06,
+      "loss": 1.1673,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9359575209049924,
+      "learning_rate": 1.6577540106951873e-06,
+      "loss": 1.1938,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9420895367049347,
+      "learning_rate": 1.684491978609626e-06,
+      "loss": 1.2261,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8773321072783626,
+      "learning_rate": 1.7112299465240642e-06,
+      "loss": 1.1594,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9401452492937292,
+      "learning_rate": 1.7379679144385028e-06,
+      "loss": 1.1125,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7415120315561607,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 1.1536,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.77922139292567,
+      "learning_rate": 1.7914438502673799e-06,
+      "loss": 1.1667,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.943977056834105,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.0904,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6636556551761967,
+      "learning_rate": 1.8449197860962567e-06,
+      "loss": 1.1664,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.863746578896042,
+      "learning_rate": 1.8716577540106954e-06,
+      "loss": 1.1581,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.406990767360614,
+      "learning_rate": 1.898395721925134e-06,
+      "loss": 1.3855,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7471377534749015,
+      "learning_rate": 1.9251336898395724e-06,
+      "loss": 1.1278,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7772450562842448,
+      "learning_rate": 1.951871657754011e-06,
+      "loss": 1.0773,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6416503558736697,
+      "learning_rate": 1.9786096256684497e-06,
+      "loss": 1.1633,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7553053622221313,
+      "learning_rate": 2.0053475935828877e-06,
+      "loss": 1.0293,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9311381021274994,
+      "learning_rate": 2.0320855614973265e-06,
+      "loss": 1.1742,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7984719062272627,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 1.1453,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9571998762030964,
+      "learning_rate": 2.0855614973262034e-06,
+      "loss": 1.1183,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7348247070925025,
+      "learning_rate": 2.112299465240642e-06,
+      "loss": 1.0005,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5404069057449405,
+      "learning_rate": 2.1390374331550802e-06,
+      "loss": 1.1682,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6872028718730576,
+      "learning_rate": 2.165775401069519e-06,
+      "loss": 1.0626,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.708502165239471,
+      "learning_rate": 2.1925133689839575e-06,
+      "loss": 1.1357,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6395311626671796,
+      "learning_rate": 2.219251336898396e-06,
+      "loss": 1.1261,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7113249341359087,
+      "learning_rate": 2.2459893048128343e-06,
+      "loss": 1.128,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.614926585183808,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.1306,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8131438441645662,
+      "learning_rate": 2.2994652406417116e-06,
+      "loss": 1.1958,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.680183820477123,
+      "learning_rate": 2.32620320855615e-06,
+      "loss": 1.0452,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.750740867921424,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 1.1231,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6238086067073154,
+      "learning_rate": 2.379679144385027e-06,
+      "loss": 1.0713,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6032620037399257,
+      "learning_rate": 2.4064171122994653e-06,
+      "loss": 1.1075,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6454970916236602,
+      "learning_rate": 2.433155080213904e-06,
+      "loss": 1.1106,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6997486806165423,
+      "learning_rate": 2.4598930481283426e-06,
+      "loss": 1.1078,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5688177252105697,
+      "learning_rate": 2.486631016042781e-06,
+      "loss": 1.0592,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6740780022823822,
+      "learning_rate": 2.5133689839572194e-06,
+      "loss": 1.0854,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.36462046075457,
+      "learning_rate": 2.5401069518716583e-06,
+      "loss": 1.3855,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6674656216077521,
+      "learning_rate": 2.5668449197860967e-06,
+      "loss": 1.0836,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1659189286585874,
+      "learning_rate": 2.5935828877005347e-06,
+      "loss": 1.3713,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7331399996139798,
+      "learning_rate": 2.6203208556149735e-06,
+      "loss": 1.0612,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6904850517612737,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 1.1184,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9867508790537807,
+      "learning_rate": 2.673796791443851e-06,
+      "loss": 1.3398,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6174079261988488,
+      "learning_rate": 2.7005347593582892e-06,
+      "loss": 1.063,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6733551851126403,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.0697,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6394187648579581,
+      "learning_rate": 2.754010695187166e-06,
+      "loss": 1.3216,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3311306269101206,
+      "learning_rate": 2.7807486631016045e-06,
+      "loss": 1.0237,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6841728277715124,
+      "learning_rate": 2.807486631016043e-06,
+      "loss": 1.0376,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.728148138182015,
+      "learning_rate": 2.8342245989304818e-06,
+      "loss": 1.0421,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6563947480680483,
+      "learning_rate": 2.8609625668449198e-06,
+      "loss": 1.0944,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7968837108862217,
+      "learning_rate": 2.8877005347593586e-06,
+      "loss": 1.0684,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8330604511521116,
+      "learning_rate": 2.914438502673797e-06,
+      "loss": 1.0887,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5449051036528985,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 1.0419,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.325087987969179,
+      "learning_rate": 2.9679144385026743e-06,
+      "loss": 1.3336,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.26116633828592,
+      "learning_rate": 2.9946524064171123e-06,
+      "loss": 1.3216,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2182907540723362,
+      "learning_rate": 3.0213903743315507e-06,
+      "loss": 1.2995,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7007166762191699,
+      "learning_rate": 3.0481283422459896e-06,
+      "loss": 1.0124,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7079128672425548,
+      "learning_rate": 3.074866310160428e-06,
+      "loss": 1.1157,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8116776644026351,
+      "learning_rate": 3.101604278074867e-06,
+      "loss": 1.0315,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9250369310126723,
+      "learning_rate": 3.128342245989305e-06,
+      "loss": 1.0841,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6158353263107859,
+      "learning_rate": 3.1550802139037433e-06,
+      "loss": 1.0489,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6815143584372974,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.9644,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.660051493858394,
+      "learning_rate": 3.2085561497326205e-06,
+      "loss": 1.1633,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6947857591941244,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 1.0905,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7580202446509152,
+      "learning_rate": 3.262032085561498e-06,
+      "loss": 1.0099,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8053001612938018,
+      "learning_rate": 3.288770053475936e-06,
+      "loss": 1.0476,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.831633578791614,
+      "learning_rate": 3.3155080213903747e-06,
+      "loss": 1.1014,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0423276707915021,
+      "learning_rate": 3.342245989304813e-06,
+      "loss": 1.2979,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7816211885515854,
+      "learning_rate": 3.368983957219252e-06,
+      "loss": 1.0508,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8065126762895813,
+      "learning_rate": 3.3957219251336904e-06,
+      "loss": 0.9765,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5990954510363118,
+      "learning_rate": 3.4224598930481284e-06,
+      "loss": 1.0768,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.585782750176935,
+      "learning_rate": 3.449197860962567e-06,
+      "loss": 1.0432,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.732520450159707,
+      "learning_rate": 3.4759358288770056e-06,
+      "loss": 1.0168,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1258475246457988,
+      "learning_rate": 3.5026737967914445e-06,
+      "loss": 1.3368,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7058790970058442,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 1.0296,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6418877269220324,
+      "learning_rate": 3.556149732620321e-06,
+      "loss": 1.0455,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5492695592309698,
+      "learning_rate": 3.5828877005347597e-06,
+      "loss": 0.9839,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8799833208278507,
+      "learning_rate": 3.609625668449198e-06,
+      "loss": 1.1338,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.052951148669076,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.3087,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.682742643586665,
+      "learning_rate": 3.6631016042780754e-06,
+      "loss": 1.0169,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0155504941743332,
+      "learning_rate": 3.6898395721925134e-06,
+      "loss": 1.3072,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6240264674493865,
+      "learning_rate": 3.716577540106952e-06,
+      "loss": 1.0004,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7338249148680371,
+      "learning_rate": 3.7433155080213907e-06,
+      "loss": 1.1341,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6660281755339907,
+      "learning_rate": 3.770053475935829e-06,
+      "loss": 1.0311,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5784557321392956,
+      "learning_rate": 3.796791443850268e-06,
+      "loss": 1.0118,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6200212404624486,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 1.0138,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7518680081010531,
+      "learning_rate": 3.850267379679145e-06,
+      "loss": 1.0325,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8853775007445426,
+      "learning_rate": 3.877005347593583e-06,
+      "loss": 1.3215,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8546331448502054,
+      "learning_rate": 3.903743315508022e-06,
+      "loss": 1.0077,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6513207288177008,
+      "learning_rate": 3.93048128342246e-06,
+      "loss": 1.0415,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6789714170086183,
+      "learning_rate": 3.957219251336899e-06,
+      "loss": 1.0329,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7581412197477366,
+      "learning_rate": 3.983957219251337e-06,
+      "loss": 0.9085,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7460174171385607,
+      "learning_rate": 4.010695187165775e-06,
+      "loss": 1.02,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6677576903886222,
+      "learning_rate": 4.037433155080215e-06,
+      "loss": 0.94,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6149469889320314,
+      "learning_rate": 4.064171122994653e-06,
+      "loss": 1.027,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9416237721141347,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 1.3041,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6761266057035356,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.985,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6845151349092773,
+      "learning_rate": 4.144385026737968e-06,
+      "loss": 1.0277,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6456095057810234,
+      "learning_rate": 4.171122994652407e-06,
+      "loss": 0.9715,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6427589795129955,
+      "learning_rate": 4.197860962566845e-06,
+      "loss": 0.9857,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.942636164299908,
+      "learning_rate": 4.224598930481284e-06,
+      "loss": 1.2801,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6818495173237056,
+      "learning_rate": 4.251336898395722e-06,
+      "loss": 1.0736,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.673317004087276,
+      "learning_rate": 4.2780748663101604e-06,
+      "loss": 1.0476,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6579551062032738,
+      "learning_rate": 4.304812834224599e-06,
+      "loss": 1.1425,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6402153038560567,
+      "learning_rate": 4.331550802139038e-06,
+      "loss": 1.0614,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7889679705277794,
+      "learning_rate": 4.3582887700534766e-06,
+      "loss": 1.1135,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6118486078317327,
+      "learning_rate": 4.385026737967915e-06,
+      "loss": 1.0414,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.89673461751575,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 1.0867,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7143734112255669,
+      "learning_rate": 4.438502673796792e-06,
+      "loss": 1.0716,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5999482780004672,
+      "learning_rate": 4.46524064171123e-06,
+      "loss": 1.0208,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7281296508062456,
+      "learning_rate": 4.491978609625669e-06,
+      "loss": 0.989,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9433879335264962,
+      "learning_rate": 4.518716577540107e-06,
+      "loss": 1.3175,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6337703097025655,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.1339,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.804087677783003,
+      "learning_rate": 4.572192513368984e-06,
+      "loss": 1.0788,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6721395533867491,
+      "learning_rate": 4.598930481283423e-06,
+      "loss": 1.0327,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6175828742311005,
+      "learning_rate": 4.625668449197862e-06,
+      "loss": 1.0366,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6472291697591595,
+      "learning_rate": 4.6524064171123e-06,
+      "loss": 0.9914,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8964001300033594,
+      "learning_rate": 4.6791443850267385e-06,
+      "loss": 1.2853,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5686764709503878,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.9891,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7595835927415242,
+      "learning_rate": 4.732620320855615e-06,
+      "loss": 1.0183,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.00075721370661,
+      "learning_rate": 4.759358288770054e-06,
+      "loss": 0.9857,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8945538515945914,
+      "learning_rate": 4.786096256684493e-06,
+      "loss": 1.274,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.581446601392783,
+      "learning_rate": 4.812834224598931e-06,
+      "loss": 1.1132,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7677277932551543,
+      "learning_rate": 4.839572192513369e-06,
+      "loss": 1.0201,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.601127301311296,
+      "learning_rate": 4.866310160427808e-06,
+      "loss": 0.9692,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7120255063263343,
+      "learning_rate": 4.893048128342247e-06,
+      "loss": 1.1433,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7243519838011778,
+      "learning_rate": 4.919786096256685e-06,
+      "loss": 1.0281,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9685886849863812,
+      "learning_rate": 4.9465240641711236e-06,
+      "loss": 1.0211,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.567250784740927,
+      "learning_rate": 4.973262032085562e-06,
+      "loss": 1.0142,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6965801728703,
+      "learning_rate": 5e-06,
+      "loss": 0.9994,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6182534867163023,
+      "learning_rate": 5.026737967914439e-06,
+      "loss": 1.0344,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.839686587793268,
+      "learning_rate": 5.053475935828877e-06,
+      "loss": 1.0805,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8911490269228441,
+      "learning_rate": 5.0802139037433165e-06,
+      "loss": 1.3445,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9136077670033413,
+      "learning_rate": 5.106951871657755e-06,
+      "loss": 1.288,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7067331670819195,
+      "learning_rate": 5.133689839572193e-06,
+      "loss": 0.9915,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6831864211212457,
+      "learning_rate": 5.160427807486631e-06,
+      "loss": 1.0045,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6391612895852137,
+      "learning_rate": 5.187165775401069e-06,
+      "loss": 1.0169,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8254036582210817,
+      "learning_rate": 5.213903743315508e-06,
+      "loss": 1.017,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8898312374429613,
+      "learning_rate": 5.240641711229947e-06,
+      "loss": 0.9066,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7309979971666078,
+      "learning_rate": 5.2673796791443855e-06,
+      "loss": 0.9044,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.725961084471421,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 1.0472,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.693516064454112,
+      "learning_rate": 5.320855614973262e-06,
+      "loss": 0.9931,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6025343266244938,
+      "learning_rate": 5.347593582887702e-06,
+      "loss": 0.9461,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7509743685523245,
+      "learning_rate": 5.37433155080214e-06,
+      "loss": 0.959,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8315515418222552,
+      "learning_rate": 5.4010695187165785e-06,
+      "loss": 1.016,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6106657161327298,
+      "learning_rate": 5.427807486631016e-06,
+      "loss": 1.0601,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6566123491722393,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.0186,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6314345288305356,
+      "learning_rate": 5.481283422459893e-06,
+      "loss": 1.0039,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.681194950959268,
+      "learning_rate": 5.508021390374332e-06,
+      "loss": 1.0076,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.628406486640975,
+      "learning_rate": 5.5347593582887706e-06,
+      "loss": 1.0995,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.740064519856547,
+      "learning_rate": 5.561497326203209e-06,
+      "loss": 0.9441,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.700201957673078,
+      "learning_rate": 5.588235294117647e-06,
+      "loss": 1.005,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.720457954350711,
+      "learning_rate": 5.614973262032086e-06,
+      "loss": 0.9218,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6059634958287854,
+      "learning_rate": 5.641711229946525e-06,
+      "loss": 0.9719,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9730707625064416,
+      "learning_rate": 5.6684491978609635e-06,
+      "loss": 1.3031,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9472987656004846,
+      "learning_rate": 5.695187165775401e-06,
+      "loss": 1.3051,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.719301753728789,
+      "learning_rate": 5.7219251336898395e-06,
+      "loss": 1.0337,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.862038263117422,
+      "learning_rate": 5.748663101604278e-06,
+      "loss": 1.2741,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8169719951778056,
+      "learning_rate": 5.775401069518717e-06,
+      "loss": 1.0638,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8240506806345749,
+      "learning_rate": 5.802139037433156e-06,
+      "loss": 0.9513,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.711573112497114,
+      "learning_rate": 5.828877005347594e-06,
+      "loss": 0.8484,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6418927389674376,
+      "learning_rate": 5.8556149732620325e-06,
+      "loss": 0.9835,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6443398047697138,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 1.0394,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6140767743927953,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 1.0312,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.599461408294564,
+      "learning_rate": 5.935828877005349e-06,
+      "loss": 0.9694,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9712802482387606,
+      "learning_rate": 5.962566844919787e-06,
+      "loss": 1.2997,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6267235857564968,
+      "learning_rate": 5.989304812834225e-06,
+      "loss": 0.9027,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5585354363542179,
+      "learning_rate": 6.016042780748663e-06,
+      "loss": 0.935,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7750118993434965,
+      "learning_rate": 6.0427807486631015e-06,
+      "loss": 0.985,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5802599756579798,
+      "learning_rate": 6.069518716577541e-06,
+      "loss": 1.0869,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8547978937904448,
+      "learning_rate": 6.096256684491979e-06,
+      "loss": 1.3026,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5899711735595479,
+      "learning_rate": 6.122994652406418e-06,
+      "loss": 0.911,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.827550234078434,
+      "learning_rate": 6.149732620320856e-06,
+      "loss": 1.0285,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8757858970945072,
+      "learning_rate": 6.176470588235295e-06,
+      "loss": 1.284,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8692741927138358,
+      "learning_rate": 6.203208556149734e-06,
+      "loss": 1.2999,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6783208799409806,
+      "learning_rate": 6.229946524064172e-06,
+      "loss": 0.9687,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6521359340339714,
+      "learning_rate": 6.25668449197861e-06,
+      "loss": 0.9094,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0490886608174566,
+      "learning_rate": 6.283422459893048e-06,
+      "loss": 1.0102,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.817985850694022,
+      "learning_rate": 6.3101604278074865e-06,
+      "loss": 1.0262,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.699216894618949,
+      "learning_rate": 6.336898395721926e-06,
+      "loss": 0.995,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6359773881893027,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.9094,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7047799148498088,
+      "learning_rate": 6.390374331550803e-06,
+      "loss": 1.0249,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5496327081352343,
+      "learning_rate": 6.417112299465241e-06,
+      "loss": 0.9852,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.546985729807792,
+      "learning_rate": 6.4438502673796795e-06,
+      "loss": 0.988,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6235319580404075,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 1.0401,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7858326314118382,
+      "learning_rate": 6.497326203208557e-06,
+      "loss": 0.977,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9215192342065,
+      "learning_rate": 6.524064171122996e-06,
+      "loss": 0.9439,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6234861968143297,
+      "learning_rate": 6.550802139037433e-06,
+      "loss": 0.9796,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5428677185434194,
+      "learning_rate": 6.577540106951872e-06,
+      "loss": 0.9237,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.646665709681037,
+      "learning_rate": 6.60427807486631e-06,
+      "loss": 0.8975,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5202862255953613,
+      "learning_rate": 6.631016042780749e-06,
+      "loss": 1.0537,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7035913347433047,
+      "learning_rate": 6.657754010695188e-06,
+      "loss": 0.9191,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0833285085287374,
+      "learning_rate": 6.684491978609626e-06,
+      "loss": 1.2878,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6996945195613555,
+      "learning_rate": 6.711229946524065e-06,
+      "loss": 1.0344,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7205883772778936,
+      "learning_rate": 6.737967914438504e-06,
+      "loss": 1.0155,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8879522750767714,
+      "learning_rate": 6.764705882352942e-06,
+      "loss": 0.8822,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6924076626852802,
+      "learning_rate": 6.791443850267381e-06,
+      "loss": 1.0087,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.707080849706548,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.9837,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5250609261388683,
+      "learning_rate": 6.844919786096257e-06,
+      "loss": 0.9946,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8860791458013021,
+      "learning_rate": 6.871657754010695e-06,
+      "loss": 1.2398,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6893195031147463,
+      "learning_rate": 6.898395721925134e-06,
+      "loss": 0.9228,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7756464652648931,
+      "learning_rate": 6.925133689839573e-06,
+      "loss": 1.048,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.57256545872689,
+      "learning_rate": 6.951871657754011e-06,
+      "loss": 0.8823,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6828440093329764,
+      "learning_rate": 6.97860962566845e-06,
+      "loss": 1.0102,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5037501274704808,
+      "learning_rate": 7.005347593582889e-06,
+      "loss": 0.9849,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.747158669018813,
+      "learning_rate": 7.032085561497327e-06,
+      "loss": 0.9902,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8311578597982793,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 1.2745,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5616266099103957,
+      "learning_rate": 7.085561497326203e-06,
+      "loss": 0.9442,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8932123161178773,
+      "learning_rate": 7.112299465240642e-06,
+      "loss": 0.9597,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7976550405472533,
+      "learning_rate": 7.13903743315508e-06,
+      "loss": 0.9656,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7266421741458091,
+      "learning_rate": 7.1657754010695195e-06,
+      "loss": 1.0149,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5830867914599367,
+      "learning_rate": 7.192513368983958e-06,
+      "loss": 0.9254,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6728946333943526,
+      "learning_rate": 7.219251336898396e-06,
+      "loss": 1.0008,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6960637229779523,
+      "learning_rate": 7.245989304812835e-06,
+      "loss": 0.9458,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5910388033396852,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.9572,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7273540990510239,
+      "learning_rate": 7.2994652406417124e-06,
+      "loss": 0.9379,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6699737001101487,
+      "learning_rate": 7.326203208556151e-06,
+      "loss": 0.9862,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0125117471817617,
+      "learning_rate": 7.352941176470589e-06,
+      "loss": 1.2909,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8888979041409342,
+      "learning_rate": 7.379679144385027e-06,
+      "loss": 0.8539,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6843254542057433,
+      "learning_rate": 7.406417112299465e-06,
+      "loss": 0.9129,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7150599174851449,
+      "learning_rate": 7.433155080213904e-06,
+      "loss": 0.9316,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9344576564988232,
+      "learning_rate": 7.459893048128343e-06,
+      "loss": 0.9501,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.722406243916013,
+      "learning_rate": 7.486631016042781e-06,
+      "loss": 0.9627,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6541527004561716,
+      "learning_rate": 7.51336898395722e-06,
+      "loss": 0.9587,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9612983022631244,
+      "learning_rate": 7.540106951871658e-06,
+      "loss": 1.2829,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6260207978616372,
+      "learning_rate": 7.5668449197860975e-06,
+      "loss": 0.999,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6983558234316496,
+      "learning_rate": 7.593582887700536e-06,
+      "loss": 0.9444,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.728134534297393,
+      "learning_rate": 7.620320855614974e-06,
+      "loss": 1.0072,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.608036154103953,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.9667,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5954525816110636,
+      "learning_rate": 7.67379679144385e-06,
+      "loss": 0.9614,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.949390713502756,
+      "learning_rate": 7.70053475935829e-06,
+      "loss": 1.2866,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7064434731267104,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 1.0574,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7750140327384567,
+      "learning_rate": 7.754010695187166e-06,
+      "loss": 0.8476,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8158884685606924,
+      "learning_rate": 7.780748663101606e-06,
+      "loss": 1.0228,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5938517616645385,
+      "learning_rate": 7.807486631016043e-06,
+      "loss": 0.9129,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8891577501766359,
+      "learning_rate": 7.834224598930483e-06,
+      "loss": 1.3145,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5262988165709592,
+      "learning_rate": 7.86096256684492e-06,
+      "loss": 0.8343,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5270838162567688,
+      "learning_rate": 7.88770053475936e-06,
+      "loss": 0.9989,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.548058369483167,
+      "learning_rate": 7.914438502673799e-06,
+      "loss": 0.9297,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8684740476400858,
+      "learning_rate": 7.941176470588236e-06,
+      "loss": 0.9603,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.649763370830464,
+      "learning_rate": 7.967914438502674e-06,
+      "loss": 1.0185,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7479959350069487,
+      "learning_rate": 7.994652406417113e-06,
+      "loss": 1.0523,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8553081083531295,
+      "learning_rate": 8.02139037433155e-06,
+      "loss": 1.261,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5506693313123037,
+      "learning_rate": 8.04812834224599e-06,
+      "loss": 0.9031,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7866068718090098,
+      "learning_rate": 8.07486631016043e-06,
+      "loss": 0.957,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8095175631029055,
+      "learning_rate": 8.101604278074867e-06,
+      "loss": 1.282,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6361471081758008,
+      "learning_rate": 8.128342245989306e-06,
+      "loss": 1.0175,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7497482368844108,
+      "learning_rate": 8.155080213903744e-06,
+      "loss": 0.8959,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8688365878907724,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 1.254,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6560811189952516,
+      "learning_rate": 8.20855614973262e-06,
+      "loss": 0.9925,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6814079201571874,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.9509,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8496793258592301,
+      "learning_rate": 8.262032085561497e-06,
+      "loss": 1.3053,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.605953265274145,
+      "learning_rate": 8.288770053475937e-06,
+      "loss": 0.8953,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6042779172543353,
+      "learning_rate": 8.315508021390374e-06,
+      "loss": 1.0336,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8654892703516974,
+      "learning_rate": 8.342245989304813e-06,
+      "loss": 1.3111,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5472422647163915,
+      "learning_rate": 8.368983957219253e-06,
+      "loss": 0.8905,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7329507872119396,
+      "learning_rate": 8.39572192513369e-06,
+      "loss": 0.9063,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7122893310256027,
+      "learning_rate": 8.42245989304813e-06,
+      "loss": 1.0017,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6884654330228583,
+      "learning_rate": 8.449197860962567e-06,
+      "loss": 1.009,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.916277685593287,
+      "learning_rate": 8.475935828877005e-06,
+      "loss": 0.9124,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8763280648151808,
+      "learning_rate": 8.502673796791444e-06,
+      "loss": 1.273,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.667673774920465,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 0.9552,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8402045760269671,
+      "learning_rate": 8.556149732620321e-06,
+      "loss": 1.2751,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6896959178398792,
+      "learning_rate": 8.58288770053476e-06,
+      "loss": 0.9334,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8202310020228063,
+      "learning_rate": 8.609625668449198e-06,
+      "loss": 1.064,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.657559430622702,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.9057,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7138438137155545,
+      "learning_rate": 8.663101604278076e-06,
+      "loss": 0.9105,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6971731752446384,
+      "learning_rate": 8.689839572192514e-06,
+      "loss": 0.9514,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9432398940495196,
+      "learning_rate": 8.716577540106953e-06,
+      "loss": 1.2672,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8971779119961532,
+      "learning_rate": 8.743315508021392e-06,
+      "loss": 1.3042,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7117447718424714,
+      "learning_rate": 8.77005347593583e-06,
+      "loss": 0.958,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8102997436760981,
+      "learning_rate": 8.796791443850268e-06,
+      "loss": 1.2761,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.813828877687282,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 1.2846,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5449758393383635,
+      "learning_rate": 8.850267379679144e-06,
+      "loss": 1.0099,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7655860369688334,
+      "learning_rate": 8.877005347593584e-06,
+      "loss": 0.8669,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5785143616236195,
+      "learning_rate": 8.903743315508023e-06,
+      "loss": 0.8427,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6358973467391211,
+      "learning_rate": 8.93048128342246e-06,
+      "loss": 0.9039,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6286817818750994,
+      "learning_rate": 8.9572192513369e-06,
+      "loss": 0.9735,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6155763610544203,
+      "learning_rate": 8.983957219251337e-06,
+      "loss": 0.971,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7557332778993522,
+      "learning_rate": 9.010695187165777e-06,
+      "loss": 1.0301,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6047531042873184,
+      "learning_rate": 9.037433155080214e-06,
+      "loss": 0.9479,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8418487792719318,
+      "learning_rate": 9.064171122994653e-06,
+      "loss": 0.9315,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7893492458699531,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.9656,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6452995896562552,
+      "learning_rate": 9.11764705882353e-06,
+      "loss": 0.9057,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6859366686041781,
+      "learning_rate": 9.144385026737968e-06,
+      "loss": 0.9259,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0465615979633054,
+      "learning_rate": 9.171122994652407e-06,
+      "loss": 1.2266,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9021971524278636,
+      "learning_rate": 9.197860962566846e-06,
+      "loss": 1.2314,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8788333460307477,
+      "learning_rate": 9.224598930481284e-06,
+      "loss": 1.287,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7473589081901069,
+      "learning_rate": 9.251336898395723e-06,
+      "loss": 0.9958,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8215791072083818,
+      "learning_rate": 9.278074866310161e-06,
+      "loss": 0.8234,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6288486950596233,
+      "learning_rate": 9.3048128342246e-06,
+      "loss": 0.9347,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6628947162214622,
+      "learning_rate": 9.331550802139038e-06,
+      "loss": 0.8916,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7398928821579398,
+      "learning_rate": 9.358288770053477e-06,
+      "loss": 0.9581,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7300763952703848,
+      "learning_rate": 9.385026737967915e-06,
+      "loss": 0.9696,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7642186601242813,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.9615,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.628333920910577,
+      "learning_rate": 9.438502673796791e-06,
+      "loss": 0.909,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5731596416651115,
+      "learning_rate": 9.46524064171123e-06,
+      "loss": 1.0685,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3955115534245028,
+      "learning_rate": 9.49197860962567e-06,
+      "loss": 1.3058,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.60648085798287,
+      "learning_rate": 9.518716577540108e-06,
+      "loss": 0.9221,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.666065988871153,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.9936,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6294006007537274,
+      "learning_rate": 9.572192513368986e-06,
+      "loss": 0.9437,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7483773104421196,
+      "learning_rate": 9.598930481283422e-06,
+      "loss": 0.892,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7247476441286762,
+      "learning_rate": 9.625668449197861e-06,
+      "loss": 0.971,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5821126524961222,
+      "learning_rate": 9.6524064171123e-06,
+      "loss": 0.8899,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5999572505257,
+      "learning_rate": 9.679144385026738e-06,
+      "loss": 0.9065,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9151428325252255,
+      "learning_rate": 9.705882352941177e-06,
+      "loss": 0.9727,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.811509411319549,
+      "learning_rate": 9.732620320855617e-06,
+      "loss": 0.9686,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7867004700225126,
+      "learning_rate": 9.759358288770054e-06,
+      "loss": 0.8992,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6037002049017484,
+      "learning_rate": 9.786096256684493e-06,
+      "loss": 0.9359,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.64679571661576,
+      "learning_rate": 9.812834224598931e-06,
+      "loss": 0.89,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6178854784673113,
+      "learning_rate": 9.83957219251337e-06,
+      "loss": 0.9553,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.579457724331343,
+      "learning_rate": 9.866310160427808e-06,
+      "loss": 0.8985,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7374424163478515,
+      "learning_rate": 9.893048128342247e-06,
+      "loss": 0.9167,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.684017113550306,
+      "learning_rate": 9.919786096256685e-06,
+      "loss": 0.8826,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.701861404588261,
+      "learning_rate": 9.946524064171124e-06,
+      "loss": 0.91,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6090868520163608,
+      "learning_rate": 9.973262032085562e-06,
+      "loss": 0.8827,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6520436655859032,
+      "learning_rate": 1e-05,
+      "loss": 0.9056,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6243794348049438,
+      "learning_rate": 9.999999831194285e-06,
+      "loss": 0.9581,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.551574413749861,
+      "learning_rate": 9.999999324777145e-06,
+      "loss": 0.9347,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.390114445135984,
+      "learning_rate": 9.99999848074862e-06,
+      "loss": 1.2743,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.787028272704262,
+      "learning_rate": 9.999997299108763e-06,
+      "loss": 0.9842,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1903372318951186,
+      "learning_rate": 9.999995779857656e-06,
+      "loss": 0.9636,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8018731055283046,
+      "learning_rate": 9.9999939229954e-06,
+      "loss": 0.9551,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9344489078516135,
+      "learning_rate": 9.999991728522121e-06,
+      "loss": 1.291,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6394082574765416,
+      "learning_rate": 9.99998919643797e-06,
+      "loss": 0.9651,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9172563410887771,
+      "learning_rate": 9.999986326743111e-06,
+      "loss": 1.2319,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6432180415049298,
+      "learning_rate": 9.999983119437745e-06,
+      "loss": 1.0143,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6638150027747303,
+      "learning_rate": 9.999979574522085e-06,
+      "loss": 0.9233,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.545287995515525,
+      "learning_rate": 9.99997569199637e-06,
+      "loss": 0.953,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0220271690567286,
+      "learning_rate": 9.999971471860864e-06,
+      "loss": 1.2658,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6881111738570367,
+      "learning_rate": 9.99996691411585e-06,
+      "loss": 0.9825,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6124409104293327,
+      "learning_rate": 9.99996201876164e-06,
+      "loss": 0.9518,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.534545140458619,
+      "learning_rate": 9.99995678579856e-06,
+      "loss": 0.9232,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5251052016625228,
+      "learning_rate": 9.999951215226962e-06,
+      "loss": 0.8441,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5963204254120489,
+      "learning_rate": 9.999945307047228e-06,
+      "loss": 0.9252,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6322385152938577,
+      "learning_rate": 9.999939061259751e-06,
+      "loss": 0.9571,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5995817861085302,
+      "learning_rate": 9.999932477864958e-06,
+      "loss": 0.9701,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6239612447743899,
+      "learning_rate": 9.99992555686329e-06,
+      "loss": 0.9427,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7108357788213928,
+      "learning_rate": 9.999918298255215e-06,
+      "loss": 1.0103,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7440732981296752,
+      "learning_rate": 9.999910702041225e-06,
+      "loss": 0.84,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6949964614460773,
+      "learning_rate": 9.99990276822183e-06,
+      "loss": 0.9718,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.787382784168786,
+      "learning_rate": 9.999894496797569e-06,
+      "loss": 1.0048,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.659609798714727,
+      "learning_rate": 9.999885887768996e-06,
+      "loss": 0.8776,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9228044140085423,
+      "learning_rate": 9.999876941136697e-06,
+      "loss": 0.9901,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7529652531809627,
+      "learning_rate": 9.999867656901273e-06,
+      "loss": 0.88,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5250613889327829,
+      "learning_rate": 9.999858035063353e-06,
+      "loss": 0.9798,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7619461159888476,
+      "learning_rate": 9.999848075623584e-06,
+      "loss": 0.9949,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0561991856729636,
+      "learning_rate": 9.999837778582641e-06,
+      "loss": 1.2836,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.92681210621099,
+      "learning_rate": 9.999827143941217e-06,
+      "loss": 1.2826,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.681315556539873,
+      "learning_rate": 9.999816171700034e-06,
+      "loss": 0.8997,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7495770520936509,
+      "learning_rate": 9.999804861859828e-06,
+      "loss": 0.8305,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7647805360266469,
+      "learning_rate": 9.999793214421366e-06,
+      "loss": 0.9522,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6329353867917615,
+      "learning_rate": 9.999781229385433e-06,
+      "loss": 0.9091,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7174912605692367,
+      "learning_rate": 9.99976890675284e-06,
+      "loss": 0.8806,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6010419763471635,
+      "learning_rate": 9.999756246524416e-06,
+      "loss": 0.9584,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.747255689396597,
+      "learning_rate": 9.99974324870102e-06,
+      "loss": 0.8597,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9348825283079543,
+      "learning_rate": 9.999729913283525e-06,
+      "loss": 1.2965,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5696759401062859,
+      "learning_rate": 9.999716240272834e-06,
+      "loss": 0.8643,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.520368120172788,
+      "learning_rate": 9.99970222966987e-06,
+      "loss": 0.9118,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1941450079809515,
+      "learning_rate": 9.99968788147558e-06,
+      "loss": 1.264,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6572101313148722,
+      "learning_rate": 9.999673195690931e-06,
+      "loss": 0.9812,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.82528756433914,
+      "learning_rate": 9.999658172316915e-06,
+      "loss": 0.849,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.806457508443531,
+      "learning_rate": 9.999642811354545e-06,
+      "loss": 0.9743,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7046973363366897,
+      "learning_rate": 9.999627112804863e-06,
+      "loss": 0.8863,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.746524894819276,
+      "learning_rate": 9.999611076668926e-06,
+      "loss": 1.2741,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7425078545663912,
+      "learning_rate": 9.999594702947817e-06,
+      "loss": 0.9435,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.767315942542136,
+      "learning_rate": 9.999577991642639e-06,
+      "loss": 0.9825,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6702346837860618,
+      "learning_rate": 9.999560942754525e-06,
+      "loss": 0.8508,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5734237744581654,
+      "learning_rate": 9.999543556284623e-06,
+      "loss": 0.9614,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9809453845491745,
+      "learning_rate": 9.999525832234107e-06,
+      "loss": 0.8673,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6435487971039735,
+      "learning_rate": 9.999507770604177e-06,
+      "loss": 0.9654,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5653056216237808,
+      "learning_rate": 9.999489371396049e-06,
+      "loss": 0.8478,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9815977599537147,
+      "learning_rate": 9.999470634610966e-06,
+      "loss": 0.9588,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7757758904493655,
+      "learning_rate": 9.999451560250196e-06,
+      "loss": 0.9365,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.735111945701849,
+      "learning_rate": 9.999432148315022e-06,
+      "loss": 0.9867,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9591930291021465,
+      "learning_rate": 9.999412398806758e-06,
+      "loss": 1.2311,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5799794587699147,
+      "learning_rate": 9.999392311726738e-06,
+      "loss": 0.9355,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8333285339838076,
+      "learning_rate": 9.999371887076317e-06,
+      "loss": 1.2618,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.799588790289206,
+      "learning_rate": 9.999351124856873e-06,
+      "loss": 1.2627,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.700043263455576,
+      "learning_rate": 9.999330025069812e-06,
+      "loss": 1.0191,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5805374389348381,
+      "learning_rate": 9.999308587716554e-06,
+      "loss": 0.9201,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6951041251944257,
+      "learning_rate": 9.99928681279855e-06,
+      "loss": 0.885,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8731187741216433,
+      "learning_rate": 9.999264700317268e-06,
+      "loss": 0.897,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8296784188937354,
+      "learning_rate": 9.999242250274201e-06,
+      "loss": 0.9396,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1243571314785823,
+      "learning_rate": 9.999219462670867e-06,
+      "loss": 1.2621,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6923038597516902,
+      "learning_rate": 9.999196337508804e-06,
+      "loss": 0.9493,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.607177321689406,
+      "learning_rate": 9.999172874789572e-06,
+      "loss": 0.862,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5891323735076752,
+      "learning_rate": 9.999149074514757e-06,
+      "loss": 0.9944,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8838148719690812,
+      "learning_rate": 9.999124936685965e-06,
+      "loss": 1.2631,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.563782081104667,
+      "learning_rate": 9.999100461304825e-06,
+      "loss": 0.9901,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5336731531971337,
+      "learning_rate": 9.999075648372991e-06,
+      "loss": 0.8467,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5474563845348597,
+      "learning_rate": 9.99905049789214e-06,
+      "loss": 0.8726,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9850284613117485,
+      "learning_rate": 9.999025009863967e-06,
+      "loss": 1.2528,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5313709653756034,
+      "learning_rate": 9.998999184290194e-06,
+      "loss": 0.8875,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5741386583618306,
+      "learning_rate": 9.998973021172564e-06,
+      "loss": 0.8189,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5789443175873787,
+      "learning_rate": 9.998946520512847e-06,
+      "loss": 0.8695,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5684631550614898,
+      "learning_rate": 9.99891968231283e-06,
+      "loss": 0.9885,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9007688030773668,
+      "learning_rate": 9.998892506574325e-06,
+      "loss": 1.2603,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5583066302768747,
+      "learning_rate": 9.998864993299167e-06,
+      "loss": 0.8531,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6115227654509239,
+      "learning_rate": 9.998837142489213e-06,
+      "loss": 0.8908,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5672610972532492,
+      "learning_rate": 9.998808954146347e-06,
+      "loss": 0.8991,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3914220138293192,
+      "learning_rate": 9.998780428272467e-06,
+      "loss": 0.8893,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8793608060421403,
+      "learning_rate": 9.998751564869504e-06,
+      "loss": 0.8796,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5752520182530787,
+      "learning_rate": 9.998722363939407e-06,
+      "loss": 0.8969,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6165856956405618,
+      "learning_rate": 9.998692825484142e-06,
+      "loss": 0.9758,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6493681139196776,
+      "learning_rate": 9.998662949505708e-06,
+      "loss": 0.9003,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8764917987604242,
+      "learning_rate": 9.998632736006124e-06,
+      "loss": 1.2573,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.573399940478985,
+      "learning_rate": 9.998602184987425e-06,
+      "loss": 0.8808,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6468968805043236,
+      "learning_rate": 9.998571296451677e-06,
+      "loss": 0.949,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.986551763307262,
+      "learning_rate": 9.998540070400966e-06,
+      "loss": 0.8755,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.757450531122584,
+      "learning_rate": 9.998508506837398e-06,
+      "loss": 0.8433,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9491703744019895,
+      "learning_rate": 9.998476605763107e-06,
+      "loss": 0.9387,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.899977480061322,
+      "learning_rate": 9.998444367180247e-06,
+      "loss": 1.2475,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8549146049737377,
+      "learning_rate": 9.998411791090992e-06,
+      "loss": 0.9733,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8567296404077995,
+      "learning_rate": 9.998378877497543e-06,
+      "loss": 0.8655,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5299359127547578,
+      "learning_rate": 9.998345626402124e-06,
+      "loss": 0.8116,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6874725348895478,
+      "learning_rate": 9.998312037806978e-06,
+      "loss": 0.8258,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4494965684918637,
+      "learning_rate": 9.998278111714374e-06,
+      "loss": 0.9289,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6015726873260077,
+      "learning_rate": 9.998243848126604e-06,
+      "loss": 0.8999,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7656568248434399,
+      "learning_rate": 9.998209247045978e-06,
+      "loss": 0.8936,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6085268692275922,
+      "learning_rate": 9.998174308474836e-06,
+      "loss": 0.8998,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6942513697776738,
+      "learning_rate": 9.998139032415534e-06,
+      "loss": 0.8594,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9884116921183193,
+      "learning_rate": 9.998103418870459e-06,
+      "loss": 1.2684,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6595450239198377,
+      "learning_rate": 9.998067467842009e-06,
+      "loss": 0.9525,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8279241638857247,
+      "learning_rate": 9.998031179332618e-06,
+      "loss": 1.2366,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8613005337877088,
+      "learning_rate": 9.99799455334473e-06,
+      "loss": 1.235,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9029121265924763,
+      "learning_rate": 9.997957589880823e-06,
+      "loss": 1.2391,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8115637439600158,
+      "learning_rate": 9.997920288943388e-06,
+      "loss": 1.2395,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.749345629544039,
+      "learning_rate": 9.99788265053495e-06,
+      "loss": 0.946,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7000515965620613,
+      "learning_rate": 9.997844674658046e-06,
+      "loss": 0.7775,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7343184224722692,
+      "learning_rate": 9.99780636131524e-06,
+      "loss": 0.9099,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.589715732675177,
+      "learning_rate": 9.997767710509123e-06,
+      "loss": 0.8861,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5313660285620998,
+      "learning_rate": 9.9977287222423e-06,
+      "loss": 0.8574,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.597332022101124,
+      "learning_rate": 9.997689396517408e-06,
+      "loss": 0.832,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6224510464653679,
+      "learning_rate": 9.997649733337097e-06,
+      "loss": 0.9924,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4748902636403407,
+      "learning_rate": 9.99760973270405e-06,
+      "loss": 1.2782,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.162276791513521,
+      "learning_rate": 9.997569394620965e-06,
+      "loss": 1.2356,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8347024982060705,
+      "learning_rate": 9.997528719090567e-06,
+      "loss": 0.946,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.64819431145681,
+      "learning_rate": 9.997487706115604e-06,
+      "loss": 0.9524,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8763102605397277,
+      "learning_rate": 9.997446355698843e-06,
+      "loss": 0.9231,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6694220706832703,
+      "learning_rate": 9.997404667843076e-06,
+      "loss": 0.8081,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8323873433359208,
+      "learning_rate": 9.997362642551118e-06,
+      "loss": 0.8759,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6738505520574647,
+      "learning_rate": 9.99732027982581e-06,
+      "loss": 0.9464,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.653498491201669,
+      "learning_rate": 9.997277579670007e-06,
+      "loss": 0.8697,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.537275229888343,
+      "learning_rate": 9.997234542086595e-06,
+      "loss": 0.8666,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5801384157809617,
+      "learning_rate": 9.997191167078479e-06,
+      "loss": 0.9432,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8937147518405208,
+      "learning_rate": 9.99714745464859e-06,
+      "loss": 0.8627,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7628763307939042,
+      "learning_rate": 9.997103404799879e-06,
+      "loss": 0.9767,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0991500160472127,
+      "learning_rate": 9.99705901753532e-06,
+      "loss": 1.2741,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8986050017446972,
+      "learning_rate": 9.997014292857907e-06,
+      "loss": 0.9643,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.638679397836925,
+      "learning_rate": 9.996969230770665e-06,
+      "loss": 0.8487,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6360459056178525,
+      "learning_rate": 9.996923831276632e-06,
+      "loss": 0.952,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.711916397036682,
+      "learning_rate": 9.996878094378878e-06,
+      "loss": 0.8952,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5641866021629862,
+      "learning_rate": 9.996832020080488e-06,
+      "loss": 0.8311,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9739476368531711,
+      "learning_rate": 9.996785608384573e-06,
+      "loss": 1.2281,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.534359993741927,
+      "learning_rate": 9.99673885929427e-06,
+      "loss": 0.9411,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.527861611324215,
+      "learning_rate": 9.996691772812733e-06,
+      "loss": 0.8473,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7207582918012991,
+      "learning_rate": 9.996644348943141e-06,
+      "loss": 0.9302,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6378467215926737,
+      "learning_rate": 9.996596587688697e-06,
+      "loss": 0.8768,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.673046656354974,
+      "learning_rate": 9.996548489052627e-06,
+      "loss": 0.9053,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7510180395657364,
+      "learning_rate": 9.996500053038176e-06,
+      "loss": 0.8892,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3683029432047211,
+      "learning_rate": 9.996451279648618e-06,
+      "loss": 1.2197,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5062711123405232,
+      "learning_rate": 9.996402168887243e-06,
+      "loss": 0.9609,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5808593627357073,
+      "learning_rate": 9.99635272075737e-06,
+      "loss": 0.8981,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7561868963204466,
+      "learning_rate": 9.996302935262337e-06,
+      "loss": 0.9481,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6423204785829326,
+      "learning_rate": 9.996252812405503e-06,
+      "loss": 0.871,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.749309031025904,
+      "learning_rate": 9.996202352190256e-06,
+      "loss": 0.8546,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6434650989279864,
+      "learning_rate": 9.996151554620001e-06,
+      "loss": 0.8435,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8571766691478683,
+      "learning_rate": 9.996100419698168e-06,
+      "loss": 1.2331,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.608118832698862,
+      "learning_rate": 9.996048947428212e-06,
+      "loss": 0.8707,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6229247756000789,
+      "learning_rate": 9.995997137813606e-06,
+      "loss": 0.8998,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6887980655650596,
+      "learning_rate": 9.995944990857848e-06,
+      "loss": 0.8708,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8482670601679823,
+      "learning_rate": 9.995892506564461e-06,
+      "loss": 0.9478,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6494954313277905,
+      "learning_rate": 9.99583968493699e-06,
+      "loss": 0.8775,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9430354874142154,
+      "learning_rate": 9.995786525978998e-06,
+      "loss": 1.2639,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6602049074855336,
+      "learning_rate": 9.995733029694077e-06,
+      "loss": 0.8766,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.810467583152964,
+      "learning_rate": 9.99567919608584e-06,
+      "loss": 1.1945,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5644503989725047,
+      "learning_rate": 9.995625025157918e-06,
+      "loss": 0.9197,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.659735984300469,
+      "learning_rate": 9.995570516913971e-06,
+      "loss": 0.9214,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.49522612022363,
+      "learning_rate": 9.995515671357681e-06,
+      "loss": 0.8636,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6669462794406056,
+      "learning_rate": 9.995460488492749e-06,
+      "loss": 0.9339,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5454763962002922,
+      "learning_rate": 9.995404968322902e-06,
+      "loss": 0.9158,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5025009251840864,
+      "learning_rate": 9.99534911085189e-06,
+      "loss": 0.9333,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5092647622177995,
+      "learning_rate": 9.995292916083482e-06,
+      "loss": 0.7782,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5360754095581055,
+      "learning_rate": 9.995236384021474e-06,
+      "loss": 0.8158,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0911125085420645,
+      "learning_rate": 9.995179514669683e-06,
+      "loss": 1.2155,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9846391808132821,
+      "learning_rate": 9.995122308031951e-06,
+      "loss": 1.2169,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.119711133683519,
+      "learning_rate": 9.995064764112135e-06,
+      "loss": 0.9179,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8869390509546895,
+      "learning_rate": 9.995006882914127e-06,
+      "loss": 1.2032,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7609955593494857,
+      "learning_rate": 9.994948664441832e-06,
+      "loss": 0.9029,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6610543371469302,
+      "learning_rate": 9.994890108699182e-06,
+      "loss": 0.9351,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6123857058586843,
+      "learning_rate": 9.99483121569013e-06,
+      "loss": 0.865,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6083944492443754,
+      "learning_rate": 9.994771985418653e-06,
+      "loss": 0.8647,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5959016478107604,
+      "learning_rate": 9.99471241788875e-06,
+      "loss": 0.9241,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5919004567567854,
+      "learning_rate": 9.994652513104443e-06,
+      "loss": 0.855,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.710211387541457,
+      "learning_rate": 9.994592271069778e-06,
+      "loss": 0.9691,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5042139723841543,
+      "learning_rate": 9.994531691788822e-06,
+      "loss": 0.9477,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5367310849988596,
+      "learning_rate": 9.994470775265665e-06,
+      "loss": 1.2088,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4831720639619212,
+      "learning_rate": 9.99440952150442e-06,
+      "loss": 0.8747,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6151909973235963,
+      "learning_rate": 9.994347930509225e-06,
+      "loss": 0.8923,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7065905303547673,
+      "learning_rate": 9.994286002284238e-06,
+      "loss": 0.8577,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6677749827378603,
+      "learning_rate": 9.994223736833638e-06,
+      "loss": 0.8453,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.58826101692309,
+      "learning_rate": 9.994161134161635e-06,
+      "loss": 0.8675,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5648526352982406,
+      "learning_rate": 9.994098194272449e-06,
+      "loss": 0.8729,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6039288089554729,
+      "learning_rate": 9.994034917170334e-06,
+      "loss": 0.9103,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5969341875232201,
+      "learning_rate": 9.993971302859561e-06,
+      "loss": 0.8989,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6001751682564351,
+      "learning_rate": 9.993907351344427e-06,
+      "loss": 0.9074,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5152697728804674,
+      "learning_rate": 9.993843062629252e-06,
+      "loss": 0.8708,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2762873156748002,
+      "learning_rate": 9.99377843671837e-06,
+      "loss": 1.2133,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0488258187183748,
+      "learning_rate": 9.993713473616151e-06,
+      "loss": 1.2448,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8302823334761562,
+      "learning_rate": 9.99364817332698e-06,
+      "loss": 1.2179,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7478522933950016,
+      "learning_rate": 9.993582535855265e-06,
+      "loss": 0.8843,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8010008166730094,
+      "learning_rate": 9.993516561205439e-06,
+      "loss": 0.8445,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6580887441813017,
+      "learning_rate": 9.993450249381955e-06,
+      "loss": 0.8709,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6130303398140315,
+      "learning_rate": 9.993383600389294e-06,
+      "loss": 0.9551,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.541957169675673,
+      "learning_rate": 9.993316614231954e-06,
+      "loss": 0.8371,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.630107462392517,
+      "learning_rate": 9.993249290914457e-06,
+      "loss": 0.8282,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6110711383662422,
+      "learning_rate": 9.993181630441352e-06,
+      "loss": 0.8558,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6765435883047133,
+      "learning_rate": 9.993113632817203e-06,
+      "loss": 0.8896,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6571740147127099,
+      "learning_rate": 9.993045298046605e-06,
+      "loss": 0.9455,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5629925174600878,
+      "learning_rate": 9.992976626134171e-06,
+      "loss": 0.8485,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4406178321139556,
+      "learning_rate": 9.99290761708454e-06,
+      "loss": 1.2458,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6788244620966644,
+      "learning_rate": 9.992838270902367e-06,
+      "loss": 0.9033,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4964194969985083,
+      "learning_rate": 9.99276858759234e-06,
+      "loss": 0.8512,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6938564136054508,
+      "learning_rate": 9.99269856715916e-06,
+      "loss": 0.8532,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7802682045493063,
+      "learning_rate": 9.992628209607556e-06,
+      "loss": 0.9366,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.486744420231548,
+      "learning_rate": 9.992557514942278e-06,
+      "loss": 0.8586,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.580408205444079,
+      "learning_rate": 9.992486483168103e-06,
+      "loss": 0.9601,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5561523297814972,
+      "learning_rate": 9.992415114289822e-06,
+      "loss": 0.9062,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8403719743874054,
+      "learning_rate": 9.992343408312258e-06,
+      "loss": 0.7997,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6522894687190806,
+      "learning_rate": 9.992271365240251e-06,
+      "loss": 0.9172,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.613680937844445,
+      "learning_rate": 9.992198985078667e-06,
+      "loss": 0.7513,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7951403310189302,
+      "learning_rate": 9.992126267832392e-06,
+      "loss": 0.907,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6666747091665577,
+      "learning_rate": 9.992053213506333e-06,
+      "loss": 0.946,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5174424251431535,
+      "learning_rate": 9.99197982210543e-06,
+      "loss": 0.7802,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.57905431067547,
+      "learning_rate": 9.991906093634633e-06,
+      "loss": 0.8486,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.927276188212811,
+      "learning_rate": 9.991832028098923e-06,
+      "loss": 1.2057,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5455666145700724,
+      "learning_rate": 9.991757625503298e-06,
+      "loss": 0.8989,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4946866303217154,
+      "learning_rate": 9.991682885852784e-06,
+      "loss": 0.9335,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4733671090024694,
+      "learning_rate": 9.991607809152428e-06,
+      "loss": 0.763,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7381609630529256,
+      "learning_rate": 9.991532395407299e-06,
+      "loss": 0.8846,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.768201055694979,
+      "learning_rate": 9.991456644622489e-06,
+      "loss": 0.9364,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.621717442507501,
+      "learning_rate": 9.991380556803113e-06,
+      "loss": 0.7931,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6440221177127945,
+      "learning_rate": 9.991304131954307e-06,
+      "loss": 0.8169,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5362828411071268,
+      "learning_rate": 9.991227370081233e-06,
+      "loss": 0.9469,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0037931545702514,
+      "learning_rate": 9.991150271189074e-06,
+      "loss": 1.1896,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8795664631760622,
+      "learning_rate": 9.991072835283035e-06,
+      "loss": 1.2511,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5754226940928284,
+      "learning_rate": 9.990995062368346e-06,
+      "loss": 0.8803,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5880314045473087,
+      "learning_rate": 9.99091695245026e-06,
+      "loss": 0.9296,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5410767803454686,
+      "learning_rate": 9.990838505534047e-06,
+      "loss": 0.8992,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.704934495640686,
+      "learning_rate": 9.990759721625005e-06,
+      "loss": 0.8759,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0957230453121363,
+      "learning_rate": 9.990680600728456e-06,
+      "loss": 1.2116,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.538624042535066,
+      "learning_rate": 9.99060114284974e-06,
+      "loss": 0.9854,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.711206233378193,
+      "learning_rate": 9.990521347994224e-06,
+      "loss": 0.9945,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5498646427037646,
+      "learning_rate": 9.990441216167295e-06,
+      "loss": 0.8909,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5204349330275777,
+      "learning_rate": 9.990360747374363e-06,
+      "loss": 0.9561,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.557888200095083,
+      "learning_rate": 9.990279941620861e-06,
+      "loss": 0.86,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5229902516586238,
+      "learning_rate": 9.990198798912249e-06,
+      "loss": 0.766,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8672973522704668,
+      "learning_rate": 9.990117319254002e-06,
+      "loss": 1.2268,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7336154199377132,
+      "learning_rate": 9.990035502651624e-06,
+      "loss": 0.8863,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5702541511971044,
+      "learning_rate": 9.989953349110637e-06,
+      "loss": 0.9203,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5675104621469005,
+      "learning_rate": 9.98987085863659e-06,
+      "loss": 0.8137,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7478097807567814,
+      "learning_rate": 9.989788031235054e-06,
+      "loss": 0.9069,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5816514734533278,
+      "learning_rate": 9.989704866911617e-06,
+      "loss": 0.9474,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5307604203283434,
+      "learning_rate": 9.989621365671902e-06,
+      "loss": 0.9027,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7099158693623115,
+      "learning_rate": 9.98953752752154e-06,
+      "loss": 0.8647,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6031767482718504,
+      "learning_rate": 9.989453352466196e-06,
+      "loss": 0.8469,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.603681108229473,
+      "learning_rate": 9.989368840511553e-06,
+      "loss": 0.9328,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6542118265919625,
+      "learning_rate": 9.989283991663316e-06,
+      "loss": 0.9612,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7026090617826815,
+      "learning_rate": 9.989198805927216e-06,
+      "loss": 0.9237,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6802956360336565,
+      "learning_rate": 9.989113283309003e-06,
+      "loss": 0.8404,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7098201471561392,
+      "learning_rate": 9.989027423814454e-06,
+      "loss": 0.8346,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4249275244651514,
+      "learning_rate": 9.988941227449365e-06,
+      "loss": 0.8522,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9768677434350856,
+      "learning_rate": 9.988854694219556e-06,
+      "loss": 1.2524,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6397667580516584,
+      "learning_rate": 9.98876782413087e-06,
+      "loss": 0.9268,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8404328183385964,
+      "learning_rate": 9.988680617189173e-06,
+      "loss": 1.2356,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.509760364395218,
+      "learning_rate": 9.988593073400354e-06,
+      "loss": 0.9461,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8403025964964629,
+      "learning_rate": 9.988505192770324e-06,
+      "loss": 1.2676,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5747824208245265,
+      "learning_rate": 9.988416975305016e-06,
+      "loss": 1.2729,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0050449092714855,
+      "learning_rate": 9.988328421010387e-06,
+      "loss": 0.8385,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5456888845915189,
+      "learning_rate": 9.988239529892416e-06,
+      "loss": 0.859,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5725474684614429,
+      "learning_rate": 9.988150301957107e-06,
+      "loss": 0.9187,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9406963468307236,
+      "learning_rate": 9.988060737210483e-06,
+      "loss": 1.2383,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5091036443477883,
+      "learning_rate": 9.987970835658592e-06,
+      "loss": 0.9659,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5248098770384597,
+      "learning_rate": 9.987880597307504e-06,
+      "loss": 0.8901,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4934607510261524,
+      "learning_rate": 9.987790022163312e-06,
+      "loss": 0.9549,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8669290362240009,
+      "learning_rate": 9.987699110232134e-06,
+      "loss": 1.2622,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.555145461021801,
+      "learning_rate": 9.987607861520107e-06,
+      "loss": 0.8903,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8167948456544997,
+      "learning_rate": 9.987516276033392e-06,
+      "loss": 1.2434,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.573000541361675,
+      "learning_rate": 9.987424353778172e-06,
+      "loss": 0.883,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6452898824730504,
+      "learning_rate": 9.987332094760657e-06,
+      "loss": 0.84,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5505766672269359,
+      "learning_rate": 9.987239498987074e-06,
+      "loss": 0.9677,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5597647220195894,
+      "learning_rate": 9.987146566463677e-06,
+      "loss": 0.8992,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7278327778950622,
+      "learning_rate": 9.987053297196739e-06,
+      "loss": 0.8776,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6095740584926126,
+      "learning_rate": 9.986959691192558e-06,
+      "loss": 0.8749,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.658747598103537,
+      "learning_rate": 9.986865748457457e-06,
+      "loss": 0.8319,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.374034712631668,
+      "learning_rate": 9.986771468997775e-06,
+      "loss": 0.9244,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.03444190148565,
+      "learning_rate": 9.986676852819883e-06,
+      "loss": 1.2029,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9148144746449407,
+      "learning_rate": 9.986581899930167e-06,
+      "loss": 1.2383,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5606565181319547,
+      "learning_rate": 9.986486610335038e-06,
+      "loss": 0.9174,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8710684720213665,
+      "learning_rate": 9.98639098404093e-06,
+      "loss": 1.2552,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5312592018774447,
+      "learning_rate": 9.986295021054302e-06,
+      "loss": 0.9297,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7093262945412542,
+      "learning_rate": 9.98619872138163e-06,
+      "loss": 0.8883,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4895729487095428,
+      "learning_rate": 9.986102085029422e-06,
+      "loss": 0.8958,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0806053871780958,
+      "learning_rate": 9.986005112004198e-06,
+      "loss": 1.2015,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5223136808938662,
+      "learning_rate": 9.985907802312509e-06,
+      "loss": 0.9453,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.559165079945979,
+      "learning_rate": 9.985810155960921e-06,
+      "loss": 0.9099,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6459652934260187,
+      "learning_rate": 9.985712172956035e-06,
+      "loss": 0.8524,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8286421654293442,
+      "learning_rate": 9.985613853304459e-06,
+      "loss": 1.2233,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9420887536025467,
+      "learning_rate": 9.985515197012835e-06,
+      "loss": 1.1933,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6595594912318345,
+      "learning_rate": 9.985416204087828e-06,
+      "loss": 0.915,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5888052057616477,
+      "learning_rate": 9.985316874536117e-06,
+      "loss": 0.9199,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.839121537061138,
+      "learning_rate": 9.985217208364413e-06,
+      "loss": 0.888,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5907781771823426,
+      "learning_rate": 9.985117205579442e-06,
+      "loss": 0.9079,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9829742380070137,
+      "learning_rate": 9.985016866187958e-06,
+      "loss": 1.2284,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7103142340605464,
+      "learning_rate": 9.984916190196736e-06,
+      "loss": 0.9585,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5896713394004327,
+      "learning_rate": 9.984815177612574e-06,
+      "loss": 0.8648,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.633415888881344,
+      "learning_rate": 9.984713828442294e-06,
+      "loss": 0.9065,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6350011716002886,
+      "learning_rate": 9.984612142692738e-06,
+      "loss": 0.9171,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.667908928147892,
+      "learning_rate": 9.984510120370771e-06,
+      "loss": 0.8634,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6537885089182556,
+      "learning_rate": 9.984407761483283e-06,
+      "loss": 0.8037,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5561847863622056,
+      "learning_rate": 9.984305066037186e-06,
+      "loss": 0.9414,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5812995329230877,
+      "learning_rate": 9.984202034039414e-06,
+      "loss": 0.8331,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5989725026429866,
+      "learning_rate": 9.984098665496923e-06,
+      "loss": 0.8515,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5944374234149283,
+      "learning_rate": 9.983994960416694e-06,
+      "loss": 0.9365,
+      "step": 682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6772548456317202,
+      "learning_rate": 9.983890918805727e-06,
+      "loss": 0.8367,
+      "step": 683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5131124911678855,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.9453,
+      "step": 684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9755789630286947,
+      "learning_rate": 9.98368182601971e-06,
+      "loss": 1.2229,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.555723489823332,
+      "learning_rate": 9.983576774858776e-06,
+      "loss": 0.8555,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6190411536338511,
+      "learning_rate": 9.983471387195344e-06,
+      "loss": 0.8755,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.484149186152568,
+      "learning_rate": 9.983365663036528e-06,
+      "loss": 0.8814,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6786912403174066,
+      "learning_rate": 9.983259602389469e-06,
+      "loss": 0.826,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6068327199795442,
+      "learning_rate": 9.983153205261324e-06,
+      "loss": 0.8497,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5445774210221108,
+      "learning_rate": 9.98304647165928e-06,
+      "loss": 0.9258,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5315113041813724,
+      "learning_rate": 9.982939401590545e-06,
+      "loss": 0.9213,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5899543418513098,
+      "learning_rate": 9.982831995062346e-06,
+      "loss": 0.926,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6751246762494365,
+      "learning_rate": 9.982724252081939e-06,
+      "loss": 0.9669,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5516510769708267,
+      "learning_rate": 9.982616172656594e-06,
+      "loss": 0.8255,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.706260872923349,
+      "learning_rate": 9.982507756793613e-06,
+      "loss": 0.8482,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.598076169056707,
+      "learning_rate": 9.982399004500317e-06,
+      "loss": 0.9474,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2136121244158218,
+      "learning_rate": 9.982289915784044e-06,
+      "loss": 1.1819,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.720504965562063,
+      "learning_rate": 9.982180490652165e-06,
+      "loss": 0.8769,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5383289037586392,
+      "learning_rate": 9.982070729112068e-06,
+      "loss": 0.9425,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5963427159188732,
+      "learning_rate": 9.981960631171162e-06,
+      "loss": 0.9602,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.627023960602584,
+      "learning_rate": 9.98185019683688e-06,
+      "loss": 0.8973,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5683098037613818,
+      "learning_rate": 9.981739426116683e-06,
+      "loss": 0.9381,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5990424212653616,
+      "learning_rate": 9.98162831901805e-06,
+      "loss": 0.8936,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.588155402693836,
+      "learning_rate": 9.98151687554848e-06,
+      "loss": 0.8985,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.636514694246623,
+      "learning_rate": 9.9814050957155e-06,
+      "loss": 0.8852,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.55467141455963,
+      "learning_rate": 9.981292979526656e-06,
+      "loss": 0.857,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6410517232357096,
+      "learning_rate": 9.981180526989521e-06,
+      "loss": 0.835,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.580974143948201,
+      "learning_rate": 9.981067738111688e-06,
+      "loss": 0.9093,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5174794488558365,
+      "learning_rate": 9.980954612900768e-06,
+      "loss": 0.8953,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7920119434215045,
+      "learning_rate": 9.980841151364405e-06,
+      "loss": 0.898,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4628579953318983,
+      "learning_rate": 9.980727353510257e-06,
+      "loss": 1.215,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5887559776008755,
+      "learning_rate": 9.980613219346012e-06,
+      "loss": 0.8952,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5657538135950384,
+      "learning_rate": 9.98049874887937e-06,
+      "loss": 0.8115,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6039110285504001,
+      "learning_rate": 9.980383942118066e-06,
+      "loss": 0.8548,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5438521834723717,
+      "learning_rate": 9.980268799069848e-06,
+      "loss": 0.8248,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7472637221841802,
+      "learning_rate": 9.980153319742494e-06,
+      "loss": 0.9288,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6303226675408926,
+      "learning_rate": 9.9800375041438e-06,
+      "loss": 0.8262,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8730337927007267,
+      "learning_rate": 9.979921352281585e-06,
+      "loss": 1.2351,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.666848112413282,
+      "learning_rate": 9.979804864163695e-06,
+      "loss": 0.8641,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6167557433585886,
+      "learning_rate": 9.979688039797993e-06,
+      "loss": 0.8687,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6495922239507568,
+      "learning_rate": 9.979570879192365e-06,
+      "loss": 0.8167,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6603340424215878,
+      "learning_rate": 9.97945338235473e-06,
+      "loss": 0.7958,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5799663922895817,
+      "learning_rate": 9.979335549293013e-06,
+      "loss": 0.9047,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5124978712176635,
+      "learning_rate": 9.979217380015173e-06,
+      "loss": 0.8814,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5965181659074317,
+      "learning_rate": 9.979098874529192e-06,
+      "loss": 0.7939,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5077987423420147,
+      "learning_rate": 9.978980032843068e-06,
+      "loss": 0.8388,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.553440561187466,
+      "learning_rate": 9.978860854964827e-06,
+      "loss": 0.9127,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8264754858935193,
+      "learning_rate": 9.978741340902518e-06,
+      "loss": 0.8823,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5545555089631016,
+      "learning_rate": 9.978621490664208e-06,
+      "loss": 0.8589,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6266049848090909,
+      "learning_rate": 9.978501304257991e-06,
+      "loss": 0.9402,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5614390447484343,
+      "learning_rate": 9.978380781691982e-06,
+      "loss": 0.8536,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5554993156724606,
+      "learning_rate": 9.978259922974318e-06,
+      "loss": 0.8689,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5046641498486533,
+      "learning_rate": 9.97813872811316e-06,
+      "loss": 0.858,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.606611499411753,
+      "learning_rate": 9.978017197116694e-06,
+      "loss": 0.8776,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5210110312767826,
+      "learning_rate": 9.97789532999312e-06,
+      "loss": 0.8236,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7465464581178427,
+      "learning_rate": 9.977773126750677e-06,
+      "loss": 0.9767,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6111250587942993,
+      "learning_rate": 9.977650587397606e-06,
+      "loss": 0.8865,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.480647135578578,
+      "learning_rate": 9.977527711942186e-06,
+      "loss": 0.8715,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0666711139883507,
+      "learning_rate": 9.977404500392711e-06,
+      "loss": 1.2361,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5396949612043953,
+      "learning_rate": 9.977280952757505e-06,
+      "loss": 0.8523,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5749007841252676,
+      "learning_rate": 9.977157069044907e-06,
+      "loss": 0.869,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8209669877614383,
+      "learning_rate": 9.977032849263284e-06,
+      "loss": 1.2254,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5371445876897547,
+      "learning_rate": 9.976908293421022e-06,
+      "loss": 0.809,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6175345896676243,
+      "learning_rate": 9.97678340152653e-06,
+      "loss": 0.9048,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6654187789642825,
+      "learning_rate": 9.976658173588244e-06,
+      "loss": 0.9158,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5590356266672274,
+      "learning_rate": 9.976532609614617e-06,
+      "loss": 0.8143,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5892721295071321,
+      "learning_rate": 9.97640670961413e-06,
+      "loss": 1.0021,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5206359572073573,
+      "learning_rate": 9.976280473595284e-06,
+      "loss": 0.822,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4996376307617352,
+      "learning_rate": 9.976153901566598e-06,
+      "loss": 0.7713,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5643904527388357,
+      "learning_rate": 9.976026993536625e-06,
+      "loss": 0.9362,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5978878376660144,
+      "learning_rate": 9.975899749513928e-06,
+      "loss": 0.8737,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.941335736585203,
+      "learning_rate": 9.975772169507106e-06,
+      "loss": 0.8129,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5498726969354564,
+      "learning_rate": 9.975644253524766e-06,
+      "loss": 0.8792,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2373839718074042,
+      "learning_rate": 9.975516001575549e-06,
+      "loss": 1.2587,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6190563426852662,
+      "learning_rate": 9.975387413668115e-06,
+      "loss": 0.9161,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9786834931376954,
+      "learning_rate": 9.975258489811146e-06,
+      "loss": 1.2285,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5904891919408752,
+      "learning_rate": 9.975129230013347e-06,
+      "loss": 0.9175,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6491689405453176,
+      "learning_rate": 9.974999634283447e-06,
+      "loss": 0.9621,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5425456155206803,
+      "learning_rate": 9.974869702630193e-06,
+      "loss": 0.8477,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6323690984462185,
+      "learning_rate": 9.974739435062364e-06,
+      "loss": 0.9159,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.597473102522369,
+      "learning_rate": 9.97460883158875e-06,
+      "loss": 0.8274,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7591030012424824,
+      "learning_rate": 9.974477892218175e-06,
+      "loss": 0.9676,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4515123197981594,
+      "learning_rate": 9.974346616959476e-06,
+      "loss": 1.2239,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.623752285977115,
+      "learning_rate": 9.97421500582152e-06,
+      "loss": 0.9116,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5259449040384303,
+      "learning_rate": 9.974083058813192e-06,
+      "loss": 0.8631,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0480510879441411,
+      "learning_rate": 9.973950775943403e-06,
+      "loss": 1.2349,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8704545487560228,
+      "learning_rate": 9.973818157221084e-06,
+      "loss": 1.1815,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5269801868692314,
+      "learning_rate": 9.973685202655187e-06,
+      "loss": 0.8548,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6544800795913106,
+      "learning_rate": 9.973551912254696e-06,
+      "loss": 0.7791,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5514811024722737,
+      "learning_rate": 9.973418286028604e-06,
+      "loss": 0.8721,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.621469922705513,
+      "learning_rate": 9.97328432398594e-06,
+      "loss": 0.9063,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4964916349472792,
+      "learning_rate": 9.973150026135743e-06,
+      "loss": 1.2128,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.341534462295757,
+      "learning_rate": 9.973015392487087e-06,
+      "loss": 1.2319,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5823867981938509,
+      "learning_rate": 9.972880423049058e-06,
+      "loss": 0.9171,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.844438164131746,
+      "learning_rate": 9.972745117830774e-06,
+      "loss": 0.819,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.612375097820019,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.8955,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5785070904723566,
+      "learning_rate": 9.972473500089998e-06,
+      "loss": 0.8625,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.538494166835394,
+      "learning_rate": 9.972337187585848e-06,
+      "loss": 0.8907,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5283398002076578,
+      "learning_rate": 9.972200539338122e-06,
+      "loss": 1.2961,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5859564775424169,
+      "learning_rate": 9.972063555356047e-06,
+      "loss": 0.8669,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5090774919281724,
+      "learning_rate": 9.971926235648868e-06,
+      "loss": 0.835,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9445687642109515,
+      "learning_rate": 9.971788580225864e-06,
+      "loss": 0.8807,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6083399799055387,
+      "learning_rate": 9.971650589096324e-06,
+      "loss": 0.9281,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9813029896629548,
+      "learning_rate": 9.971512262269568e-06,
+      "loss": 1.2021,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6046692968507614,
+      "learning_rate": 9.971373599754936e-06,
+      "loss": 0.8743,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5651722015270195,
+      "learning_rate": 9.971234601561793e-06,
+      "loss": 0.8819,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.537118391205696,
+      "learning_rate": 9.97109526769952e-06,
+      "loss": 0.8208,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5672592648201562,
+      "learning_rate": 9.970955598177527e-06,
+      "loss": 0.8875,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5678209152764904,
+      "learning_rate": 9.970815593005248e-06,
+      "loss": 0.8423,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.531325471259386,
+      "learning_rate": 9.970675252192133e-06,
+      "loss": 0.8477,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.656475536341903,
+      "learning_rate": 9.970534575747658e-06,
+      "loss": 0.8787,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6635111196453198,
+      "learning_rate": 9.97039356368132e-06,
+      "loss": 0.8332,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6422252576877494,
+      "learning_rate": 9.970252216002647e-06,
+      "loss": 0.9709,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6090937231293232,
+      "learning_rate": 9.970110532721178e-06,
+      "loss": 0.8753,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5462977286701667,
+      "learning_rate": 9.96996851384648e-06,
+      "loss": 0.8564,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3282387274437408,
+      "learning_rate": 9.969826159388145e-06,
+      "loss": 1.2532,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4674759499030112,
+      "learning_rate": 9.969683469355781e-06,
+      "loss": 0.8873,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5443975266399053,
+      "learning_rate": 9.969540443759027e-06,
+      "loss": 0.7424,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.785355030482525,
+      "learning_rate": 9.96939708260754e-06,
+      "loss": 0.838,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4410309972787625,
+      "learning_rate": 9.969253385910997e-06,
+      "loss": 0.8712,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6673873896844216,
+      "learning_rate": 9.969109353679104e-06,
+      "loss": 0.8428,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.595269550655003,
+      "learning_rate": 9.968964985921584e-06,
+      "loss": 0.8763,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5080419817774886,
+      "learning_rate": 9.968820282648186e-06,
+      "loss": 0.8848,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9376162111387027,
+      "learning_rate": 9.96867524386868e-06,
+      "loss": 1.215,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6630040066781884,
+      "learning_rate": 9.96852986959286e-06,
+      "loss": 0.8812,
+      "step": 806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8556451627071315,
+      "learning_rate": 9.968384159830542e-06,
+      "loss": 1.213,
+      "step": 807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7165800515064054,
+      "learning_rate": 9.968238114591567e-06,
+      "loss": 0.8601,
+      "step": 808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5021242657808525,
+      "learning_rate": 9.96809173388579e-06,
+      "loss": 0.838,
+      "step": 809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8262725675577803,
+      "learning_rate": 9.967945017723102e-06,
+      "loss": 1.2221,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.823374721998848,
+      "learning_rate": 9.967797966113404e-06,
+      "loss": 1.1861,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5704877437669507,
+      "learning_rate": 9.96765057906663e-06,
+      "loss": 0.8668,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.609655047484496,
+      "learning_rate": 9.967502856592728e-06,
+      "loss": 0.8939,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5023129617247735,
+      "learning_rate": 9.967354798701676e-06,
+      "loss": 0.8581,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5513696158423875,
+      "learning_rate": 9.967206405403468e-06,
+      "loss": 0.8089,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.480325824586333,
+      "learning_rate": 9.967057676708126e-06,
+      "loss": 0.8899,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5791232959809558,
+      "learning_rate": 9.966908612625693e-06,
+      "loss": 0.898,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5435787004076655,
+      "learning_rate": 9.966759213166231e-06,
+      "loss": 0.8969,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.491859491751482,
+      "learning_rate": 9.96660947833983e-06,
+      "loss": 0.9832,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6134884929501343,
+      "learning_rate": 9.966459408156601e-06,
+      "loss": 0.8548,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6129626840203988,
+      "learning_rate": 9.966309002626676e-06,
+      "loss": 0.8981,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9876694438889108,
+      "learning_rate": 9.966158261760211e-06,
+      "loss": 1.2057,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4481304968213926,
+      "learning_rate": 9.966007185567383e-06,
+      "loss": 0.8799,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8124541295541208,
+      "learning_rate": 9.965855774058395e-06,
+      "loss": 1.2072,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7162508680321158,
+      "learning_rate": 9.96570402724347e-06,
+      "loss": 0.8587,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5821215459936224,
+      "learning_rate": 9.965551945132857e-06,
+      "loss": 0.887,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.56419392195959,
+      "learning_rate": 9.965399527736819e-06,
+      "loss": 0.8165,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6293281869621739,
+      "learning_rate": 9.965246775065652e-06,
+      "loss": 0.8051,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9726554641881016,
+      "learning_rate": 9.965093687129669e-06,
+      "loss": 1.225,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4965604333569407,
+      "learning_rate": 9.964940263939206e-06,
+      "loss": 0.8082,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.696273361150681,
+      "learning_rate": 9.964786505504624e-06,
+      "loss": 0.8512,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5109942544994217,
+      "learning_rate": 9.964632411836306e-06,
+      "loss": 0.8501,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8587970481718141,
+      "learning_rate": 9.964477982944654e-06,
+      "loss": 1.1877,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5279342918380323,
+      "learning_rate": 9.964323218840095e-06,
+      "loss": 0.8103,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5595903501863457,
+      "learning_rate": 9.964168119533084e-06,
+      "loss": 0.8384,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5979927257460829,
+      "learning_rate": 9.964012685034087e-06,
+      "loss": 0.8243,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.690872532926303,
+      "learning_rate": 9.963856915353604e-06,
+      "loss": 0.9339,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8790289567986185,
+      "learning_rate": 9.963700810502154e-06,
+      "loss": 1.2626,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.465410231999522,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.9177,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4658161959685483,
+      "learning_rate": 9.963387595328524e-06,
+      "loss": 0.8501,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5831185515196096,
+      "learning_rate": 9.963230485027498e-06,
+      "loss": 0.8988,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.468366848240649,
+      "learning_rate": 9.963073039597798e-06,
+      "loss": 0.8387,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5807375552592093,
+      "learning_rate": 9.962915259050058e-06,
+      "loss": 0.867,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6579131494058128,
+      "learning_rate": 9.962757143394934e-06,
+      "loss": 0.9019,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5243215409962914,
+      "learning_rate": 9.962598692643098e-06,
+      "loss": 0.91,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6755319096627121,
+      "learning_rate": 9.96243990680525e-06,
+      "loss": 0.9577,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5894058638274384,
+      "learning_rate": 9.962280785892113e-06,
+      "loss": 0.8963,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9345554922338601,
+      "learning_rate": 9.962121329914432e-06,
+      "loss": 1.2323,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.65930238891711,
+      "learning_rate": 9.96196153888297e-06,
+      "loss": 0.8531,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5225564704758765,
+      "learning_rate": 9.96180141280852e-06,
+      "loss": 0.9075,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8542727834613469,
+      "learning_rate": 9.961640951701892e-06,
+      "loss": 1.23,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8160117491424631,
+      "learning_rate": 9.961480155573921e-06,
+      "loss": 1.1977,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5911479201817575,
+      "learning_rate": 9.961319024435465e-06,
+      "loss": 0.9465,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6384427925034404,
+      "learning_rate": 9.961157558297404e-06,
+      "loss": 0.8725,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4620914600178492,
+      "learning_rate": 9.960995757170639e-06,
+      "loss": 0.8515,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5610457928687695,
+      "learning_rate": 9.9608336210661e-06,
+      "loss": 0.9082,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5816539275334973,
+      "learning_rate": 9.960671149994727e-06,
+      "loss": 0.9179,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4298912907931096,
+      "learning_rate": 9.960508343967497e-06,
+      "loss": 0.88,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4559016872395851,
+      "learning_rate": 9.960345202995401e-06,
+      "loss": 0.8663,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5895417029091696,
+      "learning_rate": 9.960181727089455e-06,
+      "loss": 0.8654,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8876913838201386,
+      "learning_rate": 9.960017916260695e-06,
+      "loss": 0.8921,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6488919618165974,
+      "learning_rate": 9.959853770520184e-06,
+      "loss": 0.9062,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6468957590819309,
+      "learning_rate": 9.959689289879003e-06,
+      "loss": 0.7902,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.636536067445133,
+      "learning_rate": 9.959524474348263e-06,
+      "loss": 1.0091,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.505287671297433,
+      "learning_rate": 9.95935932393909e-06,
+      "loss": 0.8594,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5136664485681097,
+      "learning_rate": 9.959193838662634e-06,
+      "loss": 0.8617,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4405864271679185,
+      "learning_rate": 9.95902801853007e-06,
+      "loss": 0.8202,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6017390454120775,
+      "learning_rate": 9.958861863552596e-06,
+      "loss": 0.8785,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6065819357687168,
+      "learning_rate": 9.958695373741428e-06,
+      "loss": 0.7879,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.132440165936126,
+      "learning_rate": 9.958528549107812e-06,
+      "loss": 1.2008,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5197884971071811,
+      "learning_rate": 9.958361389663007e-06,
+      "loss": 0.8491,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.576604646765031,
+      "learning_rate": 9.958193895418305e-06,
+      "loss": 0.9584,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6245211033233342,
+      "learning_rate": 9.958026066385014e-06,
+      "loss": 0.8651,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.876308172032014,
+      "learning_rate": 9.957857902574464e-06,
+      "loss": 1.2343,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5730547260320016,
+      "learning_rate": 9.957689403998012e-06,
+      "loss": 0.8485,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6183600606595228,
+      "learning_rate": 9.957520570667036e-06,
+      "loss": 1.0281,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5215872860332218,
+      "learning_rate": 9.957351402592933e-06,
+      "loss": 0.8868,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6198581805666281,
+      "learning_rate": 9.95718189978713e-06,
+      "loss": 0.8681,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.499073540604396,
+      "learning_rate": 9.95701206226107e-06,
+      "loss": 0.8472,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0290245541061973,
+      "learning_rate": 9.956841890026218e-06,
+      "loss": 1.2433,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.597372438886083,
+      "learning_rate": 9.95667138309407e-06,
+      "loss": 0.8655,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8670633449494914,
+      "learning_rate": 9.956500541476135e-06,
+      "loss": 1.2232,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8023267121904394,
+      "learning_rate": 9.956329365183948e-06,
+      "loss": 0.8918,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8403670938989496,
+      "learning_rate": 9.956157854229072e-06,
+      "loss": 1.2084,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5220605541357521,
+      "learning_rate": 9.955986008623083e-06,
+      "loss": 0.816,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5766209460689375,
+      "learning_rate": 9.955813828377585e-06,
+      "loss": 0.8964,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7968329731912558,
+      "learning_rate": 9.955641313504208e-06,
+      "loss": 0.8917,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.31269975837571,
+      "learning_rate": 9.955468464014595e-06,
+      "loss": 0.9481,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.565988950782471,
+      "learning_rate": 9.955295279920422e-06,
+      "loss": 0.9358,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5498484177096614,
+      "learning_rate": 9.95512176123338e-06,
+      "loss": 0.8871,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5764026661093893,
+      "learning_rate": 9.954947907965186e-06,
+      "loss": 0.9162,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4343010486300414,
+      "learning_rate": 9.954773720127579e-06,
+      "loss": 0.7625,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.483432701379875,
+      "learning_rate": 9.95459919773232e-06,
+      "loss": 0.7859,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5006845700511382,
+      "learning_rate": 9.954424340791195e-06,
+      "loss": 0.865,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5432522428558055,
+      "learning_rate": 9.95424914931601e-06,
+      "loss": 0.8889,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.508478572845355,
+      "learning_rate": 9.954073623318593e-06,
+      "loss": 0.9132,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.685373554909562,
+      "learning_rate": 9.953897762810796e-06,
+      "loss": 0.926,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0509692178331513,
+      "learning_rate": 9.953721567804496e-06,
+      "loss": 1.1793,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6457077056726912,
+      "learning_rate": 9.953545038311587e-06,
+      "loss": 0.8336,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5051217144543587,
+      "learning_rate": 9.95336817434399e-06,
+      "loss": 0.8139,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.509006638078777,
+      "learning_rate": 9.953190975913648e-06,
+      "loss": 0.813,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5407290794333866,
+      "learning_rate": 9.953013443032524e-06,
+      "loss": 0.8367,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.148708886736417,
+      "learning_rate": 9.952835575712607e-06,
+      "loss": 0.883,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4351348165143027,
+      "learning_rate": 9.952657373965908e-06,
+      "loss": 0.8516,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9163739034550332,
+      "learning_rate": 9.952478837804459e-06,
+      "loss": 1.2375,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.596995845901004,
+      "learning_rate": 9.95229996724031e-06,
+      "loss": 0.8367,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6241139398126248,
+      "learning_rate": 9.952120762285546e-06,
+      "loss": 0.9035,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7389202527739365,
+      "learning_rate": 9.951941222952264e-06,
+      "loss": 0.9981,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8881784023247613,
+      "learning_rate": 9.951761349252587e-06,
+      "loss": 1.164,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.745527381650594,
+      "learning_rate": 9.951581141198662e-06,
+      "loss": 0.8893,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4603734816595164,
+      "learning_rate": 9.951400598802655e-06,
+      "loss": 0.8596,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6758564415412476,
+      "learning_rate": 9.951219722076759e-06,
+      "loss": 0.8865,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.554165973561126,
+      "learning_rate": 9.951038511033182e-06,
+      "loss": 0.8569,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8294022262654263,
+      "learning_rate": 9.950856965684167e-06,
+      "loss": 1.2292,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5887727342285896,
+      "learning_rate": 9.950675086041967e-06,
+      "loss": 0.8965,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5873872244248364,
+      "learning_rate": 9.950492872118867e-06,
+      "loss": 0.8617,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4561416679150612,
+      "learning_rate": 9.950310323927165e-06,
+      "loss": 0.7397,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.569213865174641,
+      "learning_rate": 9.950127441479193e-06,
+      "loss": 0.9019,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7220188100836231,
+      "learning_rate": 9.949944224787296e-06,
+      "loss": 0.907,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6776562914672004,
+      "learning_rate": 9.949760673863846e-06,
+      "loss": 0.9423,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4972164724285582,
+      "learning_rate": 9.949576788721237e-06,
+      "loss": 0.877,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.443475702180261,
+      "learning_rate": 9.949392569371886e-06,
+      "loss": 0.7778,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.860554014325369,
+      "learning_rate": 9.949208015828232e-06,
+      "loss": 1.1936,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6219505745666818,
+      "learning_rate": 9.949023128102734e-06,
+      "loss": 0.9293,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5469613469048205,
+      "learning_rate": 9.948837906207878e-06,
+      "loss": 0.819,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5759415908270265,
+      "learning_rate": 9.948652350156172e-06,
+      "loss": 0.8659,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1399759358035464,
+      "learning_rate": 9.948466459960142e-06,
+      "loss": 0.7979,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.534666105384267,
+      "learning_rate": 9.948280235632341e-06,
+      "loss": 0.9091,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4610704585460546,
+      "learning_rate": 9.948093677185345e-06,
+      "loss": 0.8987,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.519563061823506,
+      "learning_rate": 9.947906784631749e-06,
+      "loss": 0.8853,
+      "step": 930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6430968184663322,
+      "learning_rate": 9.947719557984172e-06,
+      "loss": 0.8796,
+      "step": 931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5331526850000736,
+      "learning_rate": 9.947531997255256e-06,
+      "loss": 0.9565,
+      "step": 932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4856893161716964,
+      "learning_rate": 9.947344102457669e-06,
+      "loss": 0.8796,
+      "step": 933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5657934606103434,
+      "learning_rate": 9.947155873604093e-06,
+      "loss": 0.951,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5821595327846476,
+      "learning_rate": 9.946967310707241e-06,
+      "loss": 0.7768,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6180640920062845,
+      "learning_rate": 9.946778413779844e-06,
+      "loss": 0.8243,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7297272143683036,
+      "learning_rate": 9.946589182834657e-06,
+      "loss": 0.9456,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9824388706503082,
+      "learning_rate": 9.946399617884457e-06,
+      "loss": 0.9136,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6628193105182445,
+      "learning_rate": 9.946209718942046e-06,
+      "loss": 0.8556,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6179862584998124,
+      "learning_rate": 9.946019486020242e-06,
+      "loss": 0.8255,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5413442363317122,
+      "learning_rate": 9.945828919131894e-06,
+      "loss": 0.9084,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.631281259492192,
+      "learning_rate": 9.94563801828987e-06,
+      "loss": 0.7806,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5136807338796188,
+      "learning_rate": 9.945446783507056e-06,
+      "loss": 0.8918,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.605564196752661,
+      "learning_rate": 9.945255214796366e-06,
+      "loss": 0.8638,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5496755674473295,
+      "learning_rate": 9.945063312170736e-06,
+      "loss": 0.9507,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5981523373121949,
+      "learning_rate": 9.944871075643125e-06,
+      "loss": 0.9116,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4576302343313638,
+      "learning_rate": 9.944678505226511e-06,
+      "loss": 0.8427,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6232678569968226,
+      "learning_rate": 9.944485600933898e-06,
+      "loss": 0.8038,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8700567164123805,
+      "learning_rate": 9.94429236277831e-06,
+      "loss": 0.8559,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6592827706886064,
+      "learning_rate": 9.944098790772797e-06,
+      "loss": 0.8687,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6709037132298505,
+      "learning_rate": 9.943904884930427e-06,
+      "loss": 0.922,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5381899241454413,
+      "learning_rate": 9.943710645264295e-06,
+      "loss": 0.8483,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5614291075175473,
+      "learning_rate": 9.943516071787517e-06,
+      "loss": 0.9734,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5961762762811158,
+      "learning_rate": 9.943321164513229e-06,
+      "loss": 0.8162,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0210905831306314,
+      "learning_rate": 9.94312592345459e-06,
+      "loss": 1.2453,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6169314694183436,
+      "learning_rate": 9.942930348624788e-06,
+      "loss": 0.9288,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4918095732996235,
+      "learning_rate": 9.942734440037028e-06,
+      "loss": 0.8517,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.64156294740685,
+      "learning_rate": 9.942538197704533e-06,
+      "loss": 0.881,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8148778287909048,
+      "learning_rate": 9.942341621640558e-06,
+      "loss": 1.1763,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5743707770346411,
+      "learning_rate": 9.942144711858374e-06,
+      "loss": 0.9441,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.586839715239287,
+      "learning_rate": 9.94194746837128e-06,
+      "loss": 0.8847,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.604869335109098,
+      "learning_rate": 9.941749891192594e-06,
+      "loss": 0.8805,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.637202468195142,
+      "learning_rate": 9.941551980335653e-06,
+      "loss": 0.7827,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.374547597815223,
+      "learning_rate": 9.941353735813824e-06,
+      "loss": 0.7999,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.597958564554555,
+      "learning_rate": 9.94115515764049e-06,
+      "loss": 0.8718,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9021014230172983,
+      "learning_rate": 9.940956245829061e-06,
+      "loss": 1.1964,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5312997885390878,
+      "learning_rate": 9.940757000392971e-06,
+      "loss": 0.8741,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5440300544113965,
+      "learning_rate": 9.940557421345667e-06,
+      "loss": 0.9808,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5445796018965658,
+      "learning_rate": 9.94035750870063e-06,
+      "loss": 0.8889,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6666604033169776,
+      "learning_rate": 9.940157262471359e-06,
+      "loss": 0.7208,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8842428952972322,
+      "learning_rate": 9.939956682671372e-06,
+      "loss": 1.2271,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7581795636181663,
+      "learning_rate": 9.939755769314215e-06,
+      "loss": 0.8037,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5391243599005493,
+      "learning_rate": 9.93955452241345e-06,
+      "loss": 0.8918,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5359442733189441,
+      "learning_rate": 9.939352941982671e-06,
+      "loss": 0.8655,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8062949802875369,
+      "learning_rate": 9.939151028035487e-06,
+      "loss": 1.2372,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5216981529084044,
+      "learning_rate": 9.93894878058553e-06,
+      "loss": 0.8752,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8032581072848206,
+      "learning_rate": 9.938746199646458e-06,
+      "loss": 1.1947,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5425433427122017,
+      "learning_rate": 9.938543285231952e-06,
+      "loss": 0.8646,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8766476602216222,
+      "learning_rate": 9.938340037355709e-06,
+      "loss": 0.8965,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5530331778329138,
+      "learning_rate": 9.938136456031454e-06,
+      "loss": 0.8668,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7312236697410663,
+      "learning_rate": 9.937932541272935e-06,
+      "loss": 0.9714,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.53502276308394,
+      "learning_rate": 9.937728293093918e-06,
+      "loss": 0.8559,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.873787567066599,
+      "learning_rate": 9.937523711508196e-06,
+      "loss": 1.1908,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.583836588337417,
+      "learning_rate": 9.937318796529583e-06,
+      "loss": 0.93,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.595658369403494,
+      "learning_rate": 9.937113548171914e-06,
+      "loss": 0.9399,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8016914893362079,
+      "learning_rate": 9.93690796644905e-06,
+      "loss": 1.1929,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6791446924191054,
+      "learning_rate": 9.93670205137487e-06,
+      "loss": 0.9495,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.515781764995385,
+      "learning_rate": 9.936495802963282e-06,
+      "loss": 0.9064,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.60690294009617,
+      "learning_rate": 9.936289221228207e-06,
+      "loss": 0.7737,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5146465229404158,
+      "learning_rate": 9.936082306183598e-06,
+      "loss": 0.8108,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8504574430651369,
+      "learning_rate": 9.935875057843423e-06,
+      "loss": 1.1846,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7182963666124904,
+      "learning_rate": 9.935667476221678e-06,
+      "loss": 1.002,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5884988118240264,
+      "learning_rate": 9.93545956133238e-06,
+      "loss": 0.7536,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7254944733755144,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.9031,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5504381889644234,
+      "learning_rate": 9.935042731807297e-06,
+      "loss": 0.8186,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5130669231709604,
+      "learning_rate": 9.93483381719966e-06,
+      "loss": 0.8472,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5185782026311627,
+      "learning_rate": 9.934624569380757e-06,
+      "loss": 0.8774,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5647256269570495,
+      "learning_rate": 9.934414988364722e-06,
+      "loss": 0.9294,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8972949851881172,
+      "learning_rate": 9.934205074165703e-06,
+      "loss": 1.1737,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.665495518116886,
+      "learning_rate": 9.933994826797874e-06,
+      "loss": 0.9056,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5531067781664374,
+      "learning_rate": 9.933784246275432e-06,
+      "loss": 0.877,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.53748757501442,
+      "learning_rate": 9.933573332612597e-06,
+      "loss": 0.8487,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5843208437155938,
+      "learning_rate": 9.933362085823607e-06,
+      "loss": 0.8863,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5492327815742086,
+      "learning_rate": 9.93315050592273e-06,
+      "loss": 0.9022,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6243674850768384,
+      "learning_rate": 9.93293859292425e-06,
+      "loss": 0.8727,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5174011916970498,
+      "learning_rate": 9.932726346842479e-06,
+      "loss": 0.863,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6042732116648475,
+      "learning_rate": 9.932513767691743e-06,
+      "loss": 0.8812,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5394724745453752,
+      "learning_rate": 9.932300855486397e-06,
+      "loss": 0.8525,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5453099659151175,
+      "learning_rate": 9.932087610240822e-06,
+      "loss": 0.8639,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7275330068236703,
+      "learning_rate": 9.931874031969411e-06,
+      "loss": 0.842,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7201351796931614,
+      "learning_rate": 9.93166012068659e-06,
+      "loss": 0.8436,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.582786772520307,
+      "learning_rate": 9.9314458764068e-06,
+      "loss": 0.7336,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9531460852511578,
+      "learning_rate": 9.931231299144509e-06,
+      "loss": 1.203,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4798361473330093,
+      "learning_rate": 9.931016388914203e-06,
+      "loss": 0.9,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.600552365583872,
+      "learning_rate": 9.930801145730395e-06,
+      "loss": 0.8708,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6394589132287902,
+      "learning_rate": 9.93058556960762e-06,
+      "loss": 0.8994,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5195910122642113,
+      "learning_rate": 9.930369660560432e-06,
+      "loss": 0.7773,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.632863694125496,
+      "learning_rate": 9.930153418603411e-06,
+      "loss": 0.845,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0662260827843473,
+      "learning_rate": 9.929936843751158e-06,
+      "loss": 1.2221,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5664288412922165,
+      "learning_rate": 9.929719936018296e-06,
+      "loss": 0.941,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5698506426366516,
+      "learning_rate": 9.92950269541947e-06,
+      "loss": 0.9052,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8310566856172155,
+      "learning_rate": 9.929285121969352e-06,
+      "loss": 1.1871,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8732153692891029,
+      "learning_rate": 9.929067215682631e-06,
+      "loss": 1.2318,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.046394045987021,
+      "learning_rate": 9.92884897657402e-06,
+      "loss": 0.9203,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.567175079880734,
+      "learning_rate": 9.928630404658255e-06,
+      "loss": 0.8439,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9090350751602232,
+      "learning_rate": 9.928411499950096e-06,
+      "loss": 1.2207,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5002832370544328,
+      "learning_rate": 9.928192262464322e-06,
+      "loss": 0.8756,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5352448946562403,
+      "learning_rate": 9.927972692215739e-06,
+      "loss": 0.8771,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6336150628605406,
+      "learning_rate": 9.927752789219171e-06,
+      "loss": 0.7678,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6818809068422826,
+      "learning_rate": 9.927532553489465e-06,
+      "loss": 0.8266,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5708989484033575,
+      "learning_rate": 9.927311985041495e-06,
+      "loss": 0.943,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5027899322917295,
+      "learning_rate": 9.927091083890152e-06,
+      "loss": 0.8621,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4992683125154507,
+      "learning_rate": 9.926869850050353e-06,
+      "loss": 0.7401,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4824835062092483,
+      "learning_rate": 9.926648283537037e-06,
+      "loss": 0.8743,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5354069234637797,
+      "learning_rate": 9.926426384365162e-06,
+      "loss": 0.8563,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0987647945752776,
+      "learning_rate": 9.926204152549711e-06,
+      "loss": 1.1839,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5525296555975288,
+      "learning_rate": 9.925981588105695e-06,
+      "loss": 0.8212,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8568660608055731,
+      "learning_rate": 9.925758691048134e-06,
+      "loss": 1.2212,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5258758737977265,
+      "learning_rate": 9.925535461392086e-06,
+      "loss": 0.8393,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4817903993309793,
+      "learning_rate": 9.92531189915262e-06,
+      "loss": 0.845,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6972602229571543,
+      "learning_rate": 9.925088004344832e-06,
+      "loss": 0.9327,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.909799237114476,
+      "learning_rate": 9.92486377698384e-06,
+      "loss": 1.2105,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7024101955125066,
+      "learning_rate": 9.924639217084783e-06,
+      "loss": 0.9141,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5110794413414614,
+      "learning_rate": 9.924414324662827e-06,
+      "loss": 0.9045,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5066057699860382,
+      "learning_rate": 9.924189099733155e-06,
+      "loss": 0.7925,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8702900988726628,
+      "learning_rate": 9.923963542310975e-06,
+      "loss": 1.1937,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.736162303163339,
+      "learning_rate": 9.923737652411518e-06,
+      "loss": 0.8332,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5198858096184873,
+      "learning_rate": 9.923511430050034e-06,
+      "loss": 0.8473,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4527317209393917,
+      "learning_rate": 9.923284875241802e-06,
+      "loss": 0.7719,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4891813388016397,
+      "learning_rate": 9.923057988002117e-06,
+      "loss": 0.9117,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8729613917353225,
+      "learning_rate": 9.922830768346298e-06,
+      "loss": 1.2137,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4835843078850388,
+      "learning_rate": 9.92260321628969e-06,
+      "loss": 0.7782,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.786138353739257,
+      "learning_rate": 9.922375331847658e-06,
+      "loss": 1.1841,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5722831660808285,
+      "learning_rate": 9.922147115035585e-06,
+      "loss": 0.8429,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5294361807362953,
+      "learning_rate": 9.921918565868887e-06,
+      "loss": 0.9203,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.472894075752094,
+      "learning_rate": 9.921689684362989e-06,
+      "loss": 0.8603,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.464298819187418,
+      "learning_rate": 9.921460470533352e-06,
+      "loss": 0.8238,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.91131972026508,
+      "learning_rate": 9.921230924395449e-06,
+      "loss": 1.1832,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8713129941443456,
+      "learning_rate": 9.921001045964781e-06,
+      "loss": 1.1831,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6823759060121664,
+      "learning_rate": 9.920770835256871e-06,
+      "loss": 0.9078,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6424460203350244,
+      "learning_rate": 9.920540292287262e-06,
+      "loss": 0.9276,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5198904477845627,
+      "learning_rate": 9.92030941707152e-06,
+      "loss": 0.946,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5076187627365758,
+      "learning_rate": 9.920078209625235e-06,
+      "loss": 0.8346,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5020431667943692,
+      "learning_rate": 9.91984666996402e-06,
+      "loss": 0.9164,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5395527010476282,
+      "learning_rate": 9.919614798103507e-06,
+      "loss": 0.8482,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5533403247446196,
+      "learning_rate": 9.919382594059355e-06,
+      "loss": 0.903,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5167212409535804,
+      "learning_rate": 9.91915005784724e-06,
+      "loss": 0.7482,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2914236185044998,
+      "learning_rate": 9.918917189482865e-06,
+      "loss": 1.251,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4555784770399571,
+      "learning_rate": 9.918683988981955e-06,
+      "loss": 0.7842,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5160918002045114,
+      "learning_rate": 9.918450456360252e-06,
+      "loss": 0.7749,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.609977507953848,
+      "learning_rate": 9.91821659163353e-06,
+      "loss": 0.8249,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.579525264431761,
+      "learning_rate": 9.917982394817576e-06,
+      "loss": 0.9153,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5418925623560698,
+      "learning_rate": 9.917747865928206e-06,
+      "loss": 0.7731,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5288245582176816,
+      "learning_rate": 9.917513004981253e-06,
+      "loss": 0.9565,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4619844993559827,
+      "learning_rate": 9.917277811992581e-06,
+      "loss": 0.8472,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5730241206034477,
+      "learning_rate": 9.917042286978064e-06,
+      "loss": 0.8843,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7141542324201973,
+      "learning_rate": 9.91680642995361e-06,
+      "loss": 0.8374,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2169815829014885,
+      "learning_rate": 9.916570240935141e-06,
+      "loss": 1.1773,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6097672967008727,
+      "learning_rate": 9.916333719938608e-06,
+      "loss": 0.9599,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.344097721529167,
+      "learning_rate": 9.91609686697998e-06,
+      "loss": 0.8377,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6316441228779275,
+      "learning_rate": 9.915859682075252e-06,
+      "loss": 0.8202,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5486254248979179,
+      "learning_rate": 9.915622165240435e-06,
+      "loss": 0.8237,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6247933979351792,
+      "learning_rate": 9.915384316491572e-06,
+      "loss": 0.8637,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.459428823926388,
+      "learning_rate": 9.915146135844718e-06,
+      "loss": 0.82,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5344320703812198,
+      "learning_rate": 9.914907623315958e-06,
+      "loss": 0.8219,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4836243220108631,
+      "learning_rate": 9.914668778921398e-06,
+      "loss": 0.8555,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5856890990603516,
+      "learning_rate": 9.914429602677163e-06,
+      "loss": 0.8011,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5548503293413205,
+      "learning_rate": 9.914190094599403e-06,
+      "loss": 0.8195,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6908458661639212,
+      "learning_rate": 9.913950254704291e-06,
+      "loss": 0.8285,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.630572128957167,
+      "learning_rate": 9.913710083008021e-06,
+      "loss": 0.8455,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.598683503942084,
+      "learning_rate": 9.913469579526811e-06,
+      "loss": 0.8437,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5601603919710103,
+      "learning_rate": 9.9132287442769e-06,
+      "loss": 0.7611,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6090610272391923,
+      "learning_rate": 9.91298757727455e-06,
+      "loss": 0.8685,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5139364606929304,
+      "learning_rate": 9.912746078536044e-06,
+      "loss": 0.8942,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5366755768643061,
+      "learning_rate": 9.912504248077688e-06,
+      "loss": 0.7907,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4509300835573726,
+      "learning_rate": 9.912262085915813e-06,
+      "loss": 0.8253,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5753821866681386,
+      "learning_rate": 9.91201959206677e-06,
+      "loss": 0.9057,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6530624321200773,
+      "learning_rate": 9.911776766546931e-06,
+      "loss": 0.8657,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5248727947254477,
+      "learning_rate": 9.911533609372694e-06,
+      "loss": 0.8792,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6866733858683758,
+      "learning_rate": 9.911290120560477e-06,
+      "loss": 0.8792,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.171399721776323,
+      "learning_rate": 9.911046300126719e-06,
+      "loss": 1.2299,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6891617334926026,
+      "learning_rate": 9.910802148087887e-06,
+      "loss": 0.8796,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.535237975178472,
+      "learning_rate": 9.910557664460464e-06,
+      "loss": 0.8837,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5129704526911545,
+      "learning_rate": 9.910312849260959e-06,
+      "loss": 0.958,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9220043339547658,
+      "learning_rate": 9.9100677025059e-06,
+      "loss": 1.2046,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.784846926630246,
+      "learning_rate": 9.909822224211845e-06,
+      "loss": 0.9383,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.514460963986215,
+      "learning_rate": 9.909576414395367e-06,
+      "loss": 0.8208,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.521264286141833,
+      "learning_rate": 9.909330273073062e-06,
+      "loss": 0.9664,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5200117206841075,
+      "learning_rate": 9.90908380026155e-06,
+      "loss": 0.9416,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8792569407413832,
+      "learning_rate": 9.908836995977475e-06,
+      "loss": 1.2216,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.570973813922727,
+      "learning_rate": 9.908589860237503e-06,
+      "loss": 0.9358,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.550819864566989,
+      "learning_rate": 9.908342393058317e-06,
+      "loss": 0.85,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.118708779521882,
+      "learning_rate": 9.908094594456631e-06,
+      "loss": 0.8399,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6512662053032676,
+      "learning_rate": 9.907846464449174e-06,
+      "loss": 0.8493,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.62276289173124,
+      "learning_rate": 9.907598003052701e-06,
+      "loss": 0.9437,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5674265861098502,
+      "learning_rate": 9.90734921028399e-06,
+      "loss": 0.8414,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5451079032907697,
+      "learning_rate": 9.907100086159838e-06,
+      "loss": 0.7977,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4785922671367702,
+      "learning_rate": 9.906850630697068e-06,
+      "loss": 0.8055,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.587094961471265,
+      "learning_rate": 9.906600843912523e-06,
+      "loss": 0.8646,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4526678098687626,
+      "learning_rate": 9.90635072582307e-06,
+      "loss": 0.8927,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5463857958282818,
+      "learning_rate": 9.906100276445596e-06,
+      "loss": 0.7829,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1807653472659774,
+      "learning_rate": 9.905849495797014e-06,
+      "loss": 1.2406,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.63948515815072,
+      "learning_rate": 9.905598383894254e-06,
+      "loss": 0.8754,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5223887128561004,
+      "learning_rate": 9.905346940754274e-06,
+      "loss": 0.7713,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4300524512584327,
+      "learning_rate": 9.905095166394055e-06,
+      "loss": 0.8,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.624434536546282,
+      "learning_rate": 9.90484306083059e-06,
+      "loss": 0.8801,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4194621891840205,
+      "learning_rate": 9.90459062408091e-06,
+      "loss": 0.8817,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0209018765780313,
+      "learning_rate": 9.904337856162054e-06,
+      "loss": 1.2278,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5574826714228565,
+      "learning_rate": 9.904084757091091e-06,
+      "loss": 0.8016,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4955509123193762,
+      "learning_rate": 9.903831326885112e-06,
+      "loss": 0.8899,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6916403146279122,
+      "learning_rate": 9.903577565561229e-06,
+      "loss": 0.9612,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.478638540478408,
+      "learning_rate": 9.903323473136575e-06,
+      "loss": 0.8891,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4542566163402708,
+      "learning_rate": 9.90306904962831e-06,
+      "loss": 0.8421,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5171907465491326,
+      "learning_rate": 9.902814295053608e-06,
+      "loss": 0.9647,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4377286820578683,
+      "learning_rate": 9.902559209429676e-06,
+      "loss": 0.8935,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.506615842362728,
+      "learning_rate": 9.902303792773736e-06,
+      "loss": 0.8391,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6930207740617287,
+      "learning_rate": 9.902048045103031e-06,
+      "loss": 0.9168,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6149285982630328,
+      "learning_rate": 9.901791966434838e-06,
+      "loss": 0.8755,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9078728509504543,
+      "learning_rate": 9.90153555678644e-06,
+      "loss": 1.1742,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.639294434120774,
+      "learning_rate": 9.901278816175152e-06,
+      "loss": 0.9032,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5834602250884953,
+      "learning_rate": 9.90102174461831e-06,
+      "loss": 0.8432,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.728071903217369,
+      "learning_rate": 9.900764342133277e-06,
+      "loss": 0.9073,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5617181078667028,
+      "learning_rate": 9.900506608737427e-06,
+      "loss": 0.8592,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8670616800731159,
+      "learning_rate": 9.900248544448166e-06,
+      "loss": 1.2156,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5264199845585185,
+      "learning_rate": 9.899990149282917e-06,
+      "loss": 0.8329,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5763591753268233,
+      "learning_rate": 9.899731423259129e-06,
+      "loss": 0.7854,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5581618396775654,
+      "learning_rate": 9.899472366394272e-06,
+      "loss": 0.8635,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.524802436119131,
+      "learning_rate": 9.899212978705836e-06,
+      "loss": 0.8511,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9461813365221302,
+      "learning_rate": 9.89895326021134e-06,
+      "loss": 1.2242,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5934375961221818,
+      "learning_rate": 9.898693210928315e-06,
+      "loss": 0.8575,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8177545336906779,
+      "learning_rate": 9.898432830874324e-06,
+      "loss": 1.1945,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6507124546737442,
+      "learning_rate": 9.898172120066947e-06,
+      "loss": 0.9113,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4963074703378922,
+      "learning_rate": 9.897911078523787e-06,
+      "loss": 0.8705,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.621155849917538,
+      "learning_rate": 9.897649706262474e-06,
+      "loss": 0.8986,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5692567111905145,
+      "learning_rate": 9.897388003300652e-06,
+      "loss": 0.895,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.645268669549263,
+      "learning_rate": 9.897125969655993e-06,
+      "loss": 0.9499,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.481465500600821,
+      "learning_rate": 9.896863605346191e-06,
+      "loss": 0.8011,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.062999489797315,
+      "learning_rate": 9.89660091038896e-06,
+      "loss": 1.1873,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6463121471874338,
+      "learning_rate": 9.896337884802041e-06,
+      "loss": 0.9552,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5696978312823517,
+      "learning_rate": 9.89607452860319e-06,
+      "loss": 0.838,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.801389252910241,
+      "learning_rate": 9.895810841810191e-06,
+      "loss": 1.1902,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5817728664880302,
+      "learning_rate": 9.895546824440852e-06,
+      "loss": 0.8703,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5451104085043792,
+      "learning_rate": 9.895282476512995e-06,
+      "loss": 0.8254,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.718846582803934,
+      "learning_rate": 9.89501779804447e-06,
+      "loss": 0.8769,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6278556824204597,
+      "learning_rate": 9.894752789053153e-06,
+      "loss": 0.8371,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4303309211167965,
+      "learning_rate": 9.894487449556934e-06,
+      "loss": 0.7878,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5587649644443693,
+      "learning_rate": 9.894221779573729e-06,
+      "loss": 0.8388,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4743462116990227,
+      "learning_rate": 9.89395577912148e-06,
+      "loss": 0.9034,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6285272101420099,
+      "learning_rate": 9.893689448218146e-06,
+      "loss": 0.9346,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8609240438837328,
+      "learning_rate": 9.89342278688171e-06,
+      "loss": 1.1954,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8379287176936888,
+      "learning_rate": 9.893155795130177e-06,
+      "loss": 1.1985,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5995032803149256,
+      "learning_rate": 9.892888472981577e-06,
+      "loss": 0.8818,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.59624011749449,
+      "learning_rate": 9.892620820453959e-06,
+      "loss": 0.8093,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8662669558907233,
+      "learning_rate": 9.892352837565395e-06,
+      "loss": 1.1694,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8646642927054429,
+      "learning_rate": 9.89208452433398e-06,
+      "loss": 1.195,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4936008096193696,
+      "learning_rate": 9.891815880777834e-06,
+      "loss": 0.8531,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5847014248547993,
+      "learning_rate": 9.891546906915091e-06,
+      "loss": 0.9319,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8423058678231164,
+      "learning_rate": 9.891277602763916e-06,
+      "loss": 1.2195,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4170008341534437,
+      "learning_rate": 9.891007968342493e-06,
+      "loss": 0.809,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5350499010970065,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.7639,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.470641389458826,
+      "learning_rate": 9.89046770876175e-06,
+      "loss": 0.7469,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5180644769971037,
+      "learning_rate": 9.89019708363891e-06,
+      "loss": 0.8169,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.896302808603409,
+      "learning_rate": 9.88992612831878e-06,
+      "loss": 1.2118,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4363488405314828,
+      "learning_rate": 9.889654842819658e-06,
+      "loss": 0.8802,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4963303480368686,
+      "learning_rate": 9.88938322715986e-06,
+      "loss": 0.8113,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5079940885764052,
+      "learning_rate": 9.889111281357725e-06,
+      "loss": 0.8417,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4420228775634831,
+      "learning_rate": 9.888839005431615e-06,
+      "loss": 0.7966,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4217592483158636,
+      "learning_rate": 9.888566399399918e-06,
+      "loss": 0.8243,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9596237029887802,
+      "learning_rate": 9.88829346328104e-06,
+      "loss": 1.1973,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5807244492235648,
+      "learning_rate": 9.888020197093409e-06,
+      "loss": 0.7818,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6094023180268835,
+      "learning_rate": 9.887746600855477e-06,
+      "loss": 0.9041,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4315510844110082,
+      "learning_rate": 9.887472674585718e-06,
+      "loss": 0.8847,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6143849420957126,
+      "learning_rate": 9.887198418302629e-06,
+      "loss": 0.8722,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8736413049553033,
+      "learning_rate": 9.886923832024726e-06,
+      "loss": 1.1919,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4876563566682297,
+      "learning_rate": 9.886648915770553e-06,
+      "loss": 0.8599,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4922819788171222,
+      "learning_rate": 9.886373669558669e-06,
+      "loss": 0.8139,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5551088110903455,
+      "learning_rate": 9.886098093407664e-06,
+      "loss": 0.7953,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5044351452307718,
+      "learning_rate": 9.885822187336142e-06,
+      "loss": 0.872,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5884547010397598,
+      "learning_rate": 9.885545951362733e-06,
+      "loss": 0.8428,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.505602855132224,
+      "learning_rate": 9.88526938550609e-06,
+      "loss": 0.7956,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5422232321602634,
+      "learning_rate": 9.884992489784888e-06,
+      "loss": 0.7777,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5526181719588474,
+      "learning_rate": 9.884715264217823e-06,
+      "loss": 0.9298,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4935116108163888,
+      "learning_rate": 9.884437708823614e-06,
+      "loss": 0.7999,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5376788293044548,
+      "learning_rate": 9.884159823621e-06,
+      "loss": 0.7795,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9331977794353743,
+      "learning_rate": 9.883881608628748e-06,
+      "loss": 1.2014,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8756825225718817,
+      "learning_rate": 9.883603063865642e-06,
+      "loss": 1.1707,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5765964141280344,
+      "learning_rate": 9.88332418935049e-06,
+      "loss": 0.8703,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.853175936019261,
+      "learning_rate": 9.883044985102122e-06,
+      "loss": 1.1873,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6361721109061373,
+      "learning_rate": 9.882765451139391e-06,
+      "loss": 0.8048,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8840592149869639,
+      "learning_rate": 9.882485587481172e-06,
+      "loss": 1.2146,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5680679306183298,
+      "learning_rate": 9.882205394146362e-06,
+      "loss": 0.8792,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6423825607086926,
+      "learning_rate": 9.881924871153879e-06,
+      "loss": 0.9045,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6278985205730483,
+      "learning_rate": 9.881644018522665e-06,
+      "loss": 0.8729,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5592678817310412,
+      "learning_rate": 9.881362836271686e-06,
+      "loss": 0.7889,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9323916723592967,
+      "learning_rate": 9.881081324419925e-06,
+      "loss": 1.148,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.498353559239422,
+      "learning_rate": 9.880799482986392e-06,
+      "loss": 0.865,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.741714375667812,
+      "learning_rate": 9.880517311990118e-06,
+      "loss": 0.8332,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.531632018796856,
+      "learning_rate": 9.880234811450154e-06,
+      "loss": 0.8458,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.547892073304275,
+      "learning_rate": 9.879951981385577e-06,
+      "loss": 0.7555,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4623411094987229,
+      "learning_rate": 9.879668821815484e-06,
+      "loss": 0.8779,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9117430969888959,
+      "learning_rate": 9.879385332758994e-06,
+      "loss": 1.2091,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5910541555968742,
+      "learning_rate": 9.879101514235248e-06,
+      "loss": 0.8156,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.530674721681562,
+      "learning_rate": 9.878817366263412e-06,
+      "loss": 0.8341,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.537143374975027,
+      "learning_rate": 9.878532888862671e-06,
+      "loss": 0.8712,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.438194046880818,
+      "learning_rate": 9.878248082052233e-06,
+      "loss": 0.8724,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.469020163955216,
+      "learning_rate": 9.87796294585133e-06,
+      "loss": 0.8774,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9045402846986408,
+      "learning_rate": 9.877677480279215e-06,
+      "loss": 1.1877,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8485465763884734,
+      "learning_rate": 9.877391685355164e-06,
+      "loss": 1.1939,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5349571510053912,
+      "learning_rate": 9.877105561098473e-06,
+      "loss": 0.8907,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5231731842172158,
+      "learning_rate": 9.876819107528461e-06,
+      "loss": 0.835,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6079268720640238,
+      "learning_rate": 9.876532324664473e-06,
+      "loss": 0.8379,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4404650649228092,
+      "learning_rate": 9.87624521252587e-06,
+      "loss": 0.7746,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.573042300152001,
+      "learning_rate": 9.875957771132042e-06,
+      "loss": 0.8227,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4782603594155683,
+      "learning_rate": 9.875670000502394e-06,
+      "loss": 0.8051,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5490029843574642,
+      "learning_rate": 9.87538190065636e-06,
+      "loss": 0.8982,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.520006558494078,
+      "learning_rate": 9.875093471613392e-06,
+      "loss": 0.7148,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5517389488432893,
+      "learning_rate": 9.874804713392964e-06,
+      "loss": 0.8479,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2903052314724248,
+      "learning_rate": 9.874515626014576e-06,
+      "loss": 1.1894,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1237688510319628,
+      "learning_rate": 9.874226209497745e-06,
+      "loss": 1.2088,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.588593261587381,
+      "learning_rate": 9.873936463862017e-06,
+      "loss": 0.8539,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5673792588622841,
+      "learning_rate": 9.873646389126954e-06,
+      "loss": 0.8058,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.956677178028655,
+      "learning_rate": 9.873355985312141e-06,
+      "loss": 0.8613,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5887984839563987,
+      "learning_rate": 9.873065252437189e-06,
+      "loss": 0.8421,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6146774662495709,
+      "learning_rate": 9.872774190521727e-06,
+      "loss": 0.8584,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9486224620946708,
+      "learning_rate": 9.87248279958541e-06,
+      "loss": 0.7851,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5886960173019136,
+      "learning_rate": 9.872191079647913e-06,
+      "loss": 0.9202,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5278281640509506,
+      "learning_rate": 9.871899030728932e-06,
+      "loss": 0.8494,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.595886011184496,
+      "learning_rate": 9.871606652848191e-06,
+      "loss": 0.849,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5471501250115154,
+      "learning_rate": 9.871313946025427e-06,
+      "loss": 0.7933,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5653286903242623,
+      "learning_rate": 9.871020910280408e-06,
+      "loss": 0.9047,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4825047931837663,
+      "learning_rate": 9.87072754563292e-06,
+      "loss": 1.2111,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.438869923833787,
+      "learning_rate": 9.870433852102766e-06,
+      "loss": 0.7566,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4688938720814233,
+      "learning_rate": 9.870139829709784e-06,
+      "loss": 0.8079,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5547898485997957,
+      "learning_rate": 9.869845478473826e-06,
+      "loss": 0.7872,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7793062260624724,
+      "learning_rate": 9.869550798414763e-06,
+      "loss": 0.8893,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6369441462934284,
+      "learning_rate": 9.869255789552496e-06,
+      "loss": 0.9565,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.026470008140896,
+      "learning_rate": 9.868960451906945e-06,
+      "loss": 0.8732,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5542806129200308,
+      "learning_rate": 9.868664785498049e-06,
+      "loss": 0.8932,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5747825171958425,
+      "learning_rate": 9.868368790345777e-06,
+      "loss": 0.8428,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5981271931048509,
+      "learning_rate": 9.868072466470109e-06,
+      "loss": 0.8302,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.936162562503447,
+      "learning_rate": 9.867775813891056e-06,
+      "loss": 1.1908,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9071525844636011,
+      "learning_rate": 9.867478832628652e-06,
+      "loss": 1.1804,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5562349094598218,
+      "learning_rate": 9.867181522702945e-06,
+      "loss": 0.8913,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4443955872843803,
+      "learning_rate": 9.866883884134012e-06,
+      "loss": 0.9088,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4809995842689774,
+      "learning_rate": 9.866585916941951e-06,
+      "loss": 0.9139,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.496507161727315,
+      "learning_rate": 9.866287621146882e-06,
+      "loss": 0.8198,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4955253721812956,
+      "learning_rate": 9.865988996768945e-06,
+      "loss": 0.8259,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6666385141954962,
+      "learning_rate": 9.865690043828302e-06,
+      "loss": 0.9153,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5171236315540129,
+      "learning_rate": 9.865390762345143e-06,
+      "loss": 0.8697,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6354602865317125,
+      "learning_rate": 9.865091152339674e-06,
+      "loss": 0.9007,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5689026535447894,
+      "learning_rate": 9.864791213832125e-06,
+      "loss": 0.8492,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5659722753495828,
+      "learning_rate": 9.86449094684275e-06,
+      "loss": 0.86,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4771556642637091,
+      "learning_rate": 9.864190351391822e-06,
+      "loss": 0.8549,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3284379296792292,
+      "learning_rate": 9.863889427499641e-06,
+      "loss": 1.1914,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.591508108745014,
+      "learning_rate": 9.863588175186522e-06,
+      "loss": 0.8972,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5414392588606192,
+      "learning_rate": 9.863286594472809e-06,
+      "loss": 0.872,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.667243685766783,
+      "learning_rate": 9.862984685378864e-06,
+      "loss": 0.738,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5744172694909626,
+      "learning_rate": 9.862682447925073e-06,
+      "loss": 0.8596,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6953900439190548,
+      "learning_rate": 9.862379882131844e-06,
+      "loss": 0.932,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5298377238644922,
+      "learning_rate": 9.862076988019609e-06,
+      "loss": 0.9113,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4729571933496919,
+      "learning_rate": 9.861773765608816e-06,
+      "loss": 0.8325,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6070729794782581,
+      "learning_rate": 9.86147021491994e-06,
+      "loss": 0.8781,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9221028350118031,
+      "learning_rate": 9.86116633597348e-06,
+      "loss": 1.1878,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5225635092739078,
+      "learning_rate": 9.860862128789954e-06,
+      "loss": 0.8356,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5351457652965883,
+      "learning_rate": 9.860557593389901e-06,
+      "loss": 0.8227,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6052197914722888,
+      "learning_rate": 9.860252729793885e-06,
+      "loss": 0.8858,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8158616010700757,
+      "learning_rate": 9.859947538022493e-06,
+      "loss": 1.1954,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4108127920660092,
+      "learning_rate": 9.859642018096326e-06,
+      "loss": 0.828,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5418038115503787,
+      "learning_rate": 9.859336170036022e-06,
+      "loss": 0.9367,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.546929474515776,
+      "learning_rate": 9.859029993862225e-06,
+      "loss": 0.9134,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7300549261260982,
+      "learning_rate": 9.858723489595613e-06,
+      "loss": 0.9416,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6836381900352337,
+      "learning_rate": 9.858416657256883e-06,
+      "loss": 0.901,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.289018750690642,
+      "learning_rate": 9.85810949686675e-06,
+      "loss": 0.82,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9242258865474157,
+      "learning_rate": 9.857802008445953e-06,
+      "loss": 1.2152,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4736484530680223,
+      "learning_rate": 9.857494192015258e-06,
+      "loss": 0.9157,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6566500046015487,
+      "learning_rate": 9.857186047595448e-06,
+      "loss": 0.8569,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7055677039529196,
+      "learning_rate": 9.85687757520733e-06,
+      "loss": 0.8578,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4070556213737988,
+      "learning_rate": 9.85656877487173e-06,
+      "loss": 0.7889,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4129318827285564,
+      "learning_rate": 9.856259646609502e-06,
+      "loss": 0.7902,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9754546732047004,
+      "learning_rate": 9.85595019044152e-06,
+      "loss": 1.1929,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6490818380054377,
+      "learning_rate": 9.855640406388673e-06,
+      "loss": 0.929,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6112090513989958,
+      "learning_rate": 9.855330294471886e-06,
+      "loss": 0.7994,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5926109815547944,
+      "learning_rate": 9.855019854712094e-06,
+      "loss": 0.8724,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9784351763031909,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 1.2002,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6299940449234447,
+      "learning_rate": 9.85439799174737e-06,
+      "loss": 0.8581,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5349689049540285,
+      "learning_rate": 9.854086568584425e-06,
+      "loss": 0.8728,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.687897348710694,
+      "learning_rate": 9.853774817662453e-06,
+      "loss": 0.7981,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4854390251710325,
+      "learning_rate": 9.85346273900251e-06,
+      "loss": 0.8525,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5643721308759893,
+      "learning_rate": 9.853150332625663e-06,
+      "loss": 0.8589,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.467465638269256,
+      "learning_rate": 9.85283759855301e-06,
+      "loss": 0.8462,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5261348890192472,
+      "learning_rate": 9.852524536805663e-06,
+      "loss": 0.8514,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5259279326559416,
+      "learning_rate": 9.852211147404765e-06,
+      "loss": 0.8678,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5371570267510344,
+      "learning_rate": 9.851897430371475e-06,
+      "loss": 0.8571,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.450758035026696,
+      "learning_rate": 9.851583385726975e-06,
+      "loss": 0.8335,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.641953556168877,
+      "learning_rate": 9.851269013492473e-06,
+      "loss": 0.9522,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.431164765337068,
+      "learning_rate": 9.850954313689193e-06,
+      "loss": 0.8876,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.49865279944874,
+      "learning_rate": 9.850639286338385e-06,
+      "loss": 0.8355,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4538855531246735,
+      "learning_rate": 9.850323931461321e-06,
+      "loss": 0.8226,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5798041629490462,
+      "learning_rate": 9.850008249079295e-06,
+      "loss": 0.8621,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.464627903690832,
+      "learning_rate": 9.84969223921362e-06,
+      "loss": 0.8131,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9260955724761041,
+      "learning_rate": 9.849375901885636e-06,
+      "loss": 1.1874,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8986381017899657,
+      "learning_rate": 9.849059237116702e-06,
+      "loss": 1.1666,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.533701088336803,
+      "learning_rate": 9.848742244928202e-06,
+      "loss": 0.8333,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4644545161832554,
+      "learning_rate": 9.848424925341537e-06,
+      "loss": 0.7996,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9247337474623049,
+      "learning_rate": 9.848107278378136e-06,
+      "loss": 1.2204,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6635449759521905,
+      "learning_rate": 9.847789304059445e-06,
+      "loss": 0.998,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.479228470814932,
+      "learning_rate": 9.847471002406935e-06,
+      "loss": 0.8962,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9279910086761578,
+      "learning_rate": 9.847152373442096e-06,
+      "loss": 1.1762,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4333696439023775,
+      "learning_rate": 9.846833417186448e-06,
+      "loss": 0.848,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5276918188363302,
+      "learning_rate": 9.846514133661524e-06,
+      "loss": 0.8787,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5809737046192498,
+      "learning_rate": 9.846194522888884e-06,
+      "loss": 0.8455,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4855007643805271,
+      "learning_rate": 9.845874584890106e-06,
+      "loss": 0.7949,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5016534184499017,
+      "learning_rate": 9.845554319686799e-06,
+      "loss": 0.8553,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0583734228798338,
+      "learning_rate": 9.84523372730058e-06,
+      "loss": 1.1944,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4948105985914302,
+      "learning_rate": 9.844912807753105e-06,
+      "loss": 0.8048,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.549981597895823,
+      "learning_rate": 9.844591561066035e-06,
+      "loss": 0.8597,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5397088050533996,
+      "learning_rate": 9.844269987261066e-06,
+      "loss": 0.8023,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4976514792560012,
+      "learning_rate": 9.843948086359912e-06,
+      "loss": 0.8148,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5970778161381713,
+      "learning_rate": 9.843625858384305e-06,
+      "loss": 0.9347,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4376779464012985,
+      "learning_rate": 9.843303303356005e-06,
+      "loss": 0.8763,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9373975731809215,
+      "learning_rate": 9.84298042129679e-06,
+      "loss": 1.1849,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6350707255511154,
+      "learning_rate": 9.842657212228464e-06,
+      "loss": 0.9309,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5039739292986363,
+      "learning_rate": 9.84233367617285e-06,
+      "loss": 0.8529,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5865616158002578,
+      "learning_rate": 9.842009813151793e-06,
+      "loss": 0.9316,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.508166700126012,
+      "learning_rate": 9.841685623187162e-06,
+      "loss": 0.8487,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.582670270272094,
+      "learning_rate": 9.841361106300846e-06,
+      "loss": 0.9394,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4243645307350397,
+      "learning_rate": 9.841036262514756e-06,
+      "loss": 0.8543,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4596442543767099,
+      "learning_rate": 9.84071109185083e-06,
+      "loss": 0.8383,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5103195211903802,
+      "learning_rate": 9.840385594331022e-06,
+      "loss": 0.8385,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0356195386223297,
+      "learning_rate": 9.84005976997731e-06,
+      "loss": 1.1976,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.610580585463436,
+      "learning_rate": 9.839733618811695e-06,
+      "loss": 0.7779,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4362731185806386,
+      "learning_rate": 9.839407140856199e-06,
+      "loss": 0.8681,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.684087985304583,
+      "learning_rate": 9.839080336132867e-06,
+      "loss": 0.9215,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6142422860259236,
+      "learning_rate": 9.838753204663766e-06,
+      "loss": 0.8348,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5927316913953349,
+      "learning_rate": 9.838425746470984e-06,
+      "loss": 0.8466,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6364811177703218,
+      "learning_rate": 9.838097961576632e-06,
+      "loss": 0.9265,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4531917493494853,
+      "learning_rate": 9.837769850002842e-06,
+      "loss": 0.8171,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5843963754970225,
+      "learning_rate": 9.837441411771771e-06,
+      "loss": 0.8449,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0211902152211878,
+      "learning_rate": 9.837112646905593e-06,
+      "loss": 1.2037,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4799796641075675,
+      "learning_rate": 9.836783555426508e-06,
+      "loss": 0.7816,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5104608898802458,
+      "learning_rate": 9.836454137356739e-06,
+      "loss": 0.8397,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5986710279649745,
+      "learning_rate": 9.836124392718526e-06,
+      "loss": 0.8931,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5327145870916294,
+      "learning_rate": 9.835794321534136e-06,
+      "loss": 0.7759,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6431471114587126,
+      "learning_rate": 9.835463923825854e-06,
+      "loss": 0.8484,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4390853701463031,
+      "learning_rate": 9.835133199615994e-06,
+      "loss": 0.8117,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4472150698218422,
+      "learning_rate": 9.834802148926883e-06,
+      "loss": 0.71,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0461701179326943,
+      "learning_rate": 9.834470771780875e-06,
+      "loss": 1.2117,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9118172566859502,
+      "learning_rate": 9.834139068200346e-06,
+      "loss": 1.1839,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5478388917734731,
+      "learning_rate": 9.833807038207693e-06,
+      "loss": 0.8947,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6115358619188718,
+      "learning_rate": 9.833474681825334e-06,
+      "loss": 0.8313,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.605303375468737,
+      "learning_rate": 9.833141999075715e-06,
+      "loss": 0.9124,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5676137176313567,
+      "learning_rate": 9.832808989981296e-06,
+      "loss": 0.8648,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2447719994848394,
+      "learning_rate": 9.832475654564562e-06,
+      "loss": 0.8052,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5951523608946647,
+      "learning_rate": 9.832141992848022e-06,
+      "loss": 0.8798,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4909927414318729,
+      "learning_rate": 9.831808004854207e-06,
+      "loss": 0.8175,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5350463398282166,
+      "learning_rate": 9.831473690605664e-06,
+      "loss": 0.8448,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.619771994945871,
+      "learning_rate": 9.831139050124972e-06,
+      "loss": 0.9307,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5795033045615345,
+      "learning_rate": 9.830804083434722e-06,
+      "loss": 0.9465,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7465772773765422,
+      "learning_rate": 9.830468790557536e-06,
+      "loss": 1.1725,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4366774512116443,
+      "learning_rate": 9.83013317151605e-06,
+      "loss": 1.1944,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.54944758466393,
+      "learning_rate": 9.829797226332928e-06,
+      "loss": 0.8501,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8108328247645301,
+      "learning_rate": 9.829460955030854e-06,
+      "loss": 1.168,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9471264346778719,
+      "learning_rate": 9.829124357632533e-06,
+      "loss": 1.167,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5824945124402159,
+      "learning_rate": 9.828787434160694e-06,
+      "loss": 0.7757,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6296089234138476,
+      "learning_rate": 9.828450184638082e-06,
+      "loss": 0.781,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5168811759476448,
+      "learning_rate": 9.828112609087477e-06,
+      "loss": 0.8173,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4641586970613358,
+      "learning_rate": 9.827774707531667e-06,
+      "loss": 0.8237,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5652592366624505,
+      "learning_rate": 9.827436479993468e-06,
+      "loss": 0.824,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5380373868838944,
+      "learning_rate": 9.827097926495722e-06,
+      "loss": 0.7754,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.498154961271221,
+      "learning_rate": 9.826759047061283e-06,
+      "loss": 0.8721,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.520452054808714,
+      "learning_rate": 9.826419841713038e-06,
+      "loss": 0.8127,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4999675509528054,
+      "learning_rate": 9.826080310473888e-06,
+      "loss": 0.8486,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5869869206789395,
+      "learning_rate": 9.825740453366761e-06,
+      "loss": 0.8818,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6242947500917713,
+      "learning_rate": 9.825400270414602e-06,
+      "loss": 0.8402,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.479651042019983,
+      "learning_rate": 9.825059761640386e-06,
+      "loss": 0.7545,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5197623788977601,
+      "learning_rate": 9.824718927067098e-06,
+      "loss": 0.7946,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4493370271122041,
+      "learning_rate": 9.824377766717758e-06,
+      "loss": 0.8174,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6071322849901584,
+      "learning_rate": 9.824036280615398e-06,
+      "loss": 0.8973,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5205270050291608,
+      "learning_rate": 9.82369446878308e-06,
+      "loss": 0.8092,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4170956171852072,
+      "learning_rate": 9.823352331243881e-06,
+      "loss": 0.8179,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.495142648736861,
+      "learning_rate": 9.823009868020901e-06,
+      "loss": 0.8784,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7685272216454342,
+      "learning_rate": 9.822667079137268e-06,
+      "loss": 0.9017,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5183626075829533,
+      "learning_rate": 9.822323964616125e-06,
+      "loss": 0.7997,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7470646161691972,
+      "learning_rate": 9.821980524480641e-06,
+      "loss": 1.1873,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.429186462052708,
+      "learning_rate": 9.821636758754007e-06,
+      "loss": 0.8153,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1084619603492618,
+      "learning_rate": 9.821292667459435e-06,
+      "loss": 1.1843,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8988174886017172,
+      "learning_rate": 9.820948250620155e-06,
+      "loss": 1.1764,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6693095102999058,
+      "learning_rate": 9.820603508259425e-06,
+      "loss": 0.8289,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6867170159085598,
+      "learning_rate": 9.820258440400525e-06,
+      "loss": 0.8981,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1837697770917035,
+      "learning_rate": 9.819913047066752e-06,
+      "loss": 1.1551,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.47722073073164,
+      "learning_rate": 9.81956732828143e-06,
+      "loss": 0.8417,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.296421444259182,
+      "learning_rate": 9.8192212840679e-06,
+      "loss": 1.1881,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.687541674674507,
+      "learning_rate": 9.818874914449528e-06,
+      "loss": 0.8185,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4710423676489746,
+      "learning_rate": 9.818528219449705e-06,
+      "loss": 0.8521,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.616671725725048,
+      "learning_rate": 9.818181199091838e-06,
+      "loss": 0.7511,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4261890079336377,
+      "learning_rate": 9.817833853399358e-06,
+      "loss": 0.7955,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6409360040228838,
+      "learning_rate": 9.81748618239572e-06,
+      "loss": 0.9141,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6216273692196497,
+      "learning_rate": 9.817138186104401e-06,
+      "loss": 0.76,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.063935879852533,
+      "learning_rate": 9.816789864548893e-06,
+      "loss": 1.2028,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5800870537974188,
+      "learning_rate": 9.816441217752721e-06,
+      "loss": 0.9143,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5537999519646133,
+      "learning_rate": 9.816092245739426e-06,
+      "loss": 0.8306,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4393604420606185,
+      "learning_rate": 9.815742948532568e-06,
+      "loss": 0.8153,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5679317888267674,
+      "learning_rate": 9.815393326155734e-06,
+      "loss": 0.8041,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5260223643283375,
+      "learning_rate": 9.81504337863253e-06,
+      "loss": 0.8428,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5373357530411829,
+      "learning_rate": 9.81469310598659e-06,
+      "loss": 0.8429,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5631309644465936,
+      "learning_rate": 9.814342508241561e-06,
+      "loss": 0.8683,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4509390970918752,
+      "learning_rate": 9.813991585421118e-06,
+      "loss": 0.8087,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1233512121323803,
+      "learning_rate": 9.813640337548955e-06,
+      "loss": 1.1819,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5336450112502695,
+      "learning_rate": 9.813288764648787e-06,
+      "loss": 0.829,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.545749266813217,
+      "learning_rate": 9.812936866744358e-06,
+      "loss": 0.7875,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4431320648369386,
+      "learning_rate": 9.812584643859426e-06,
+      "loss": 0.843,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5258267217906833,
+      "learning_rate": 9.812232096017773e-06,
+      "loss": 0.8933,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5303516472823941,
+      "learning_rate": 9.811879223243207e-06,
+      "loss": 0.8757,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4277562074627048,
+      "learning_rate": 9.81152602555955e-06,
+      "loss": 0.7984,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5397463552019395,
+      "learning_rate": 9.811172502990656e-06,
+      "loss": 0.7918,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4516082156336023,
+      "learning_rate": 9.810818655560393e-06,
+      "loss": 0.7881,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4768224070579294,
+      "learning_rate": 9.810464483292653e-06,
+      "loss": 0.8238,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5003525044063502,
+      "learning_rate": 9.81010998621135e-06,
+      "loss": 0.8339,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.489859004665234,
+      "learning_rate": 9.809755164340423e-06,
+      "loss": 0.8019,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0518296593142906,
+      "learning_rate": 9.80940001770383e-06,
+      "loss": 1.1595,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8766815480270564,
+      "learning_rate": 9.80904454632555e-06,
+      "loss": 1.1913,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5368395043072665,
+      "learning_rate": 9.808688750229584e-06,
+      "loss": 0.7743,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4537216444180296,
+      "learning_rate": 9.808332629439961e-06,
+      "loss": 0.7998,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7008063387753622,
+      "learning_rate": 9.80797618398072e-06,
+      "loss": 0.8099,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4677348412520916,
+      "learning_rate": 9.807619413875937e-06,
+      "loss": 0.7965,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5208447333326456,
+      "learning_rate": 9.807262319149695e-06,
+      "loss": 0.8557,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.42775197658619,
+      "learning_rate": 9.80690489982611e-06,
+      "loss": 0.7997,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4833874630850366,
+      "learning_rate": 9.806547155929315e-06,
+      "loss": 0.8494,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6021441839810697,
+      "learning_rate": 9.806189087483465e-06,
+      "loss": 0.8959,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4769055730644622,
+      "learning_rate": 9.805830694512736e-06,
+      "loss": 0.8331,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7147908340655862,
+      "learning_rate": 9.80547197704133e-06,
+      "loss": 1.1938,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5508334248365785,
+      "learning_rate": 9.805112935093469e-06,
+      "loss": 0.8128,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.652445688019367,
+      "learning_rate": 9.804753568693395e-06,
+      "loss": 0.8933,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5706119466869262,
+      "learning_rate": 9.804393877865373e-06,
+      "loss": 0.7907,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4910739003796312,
+      "learning_rate": 9.80403386263369e-06,
+      "loss": 0.8202,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4921426614704876,
+      "learning_rate": 9.803673523022655e-06,
+      "loss": 0.818,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6236117416339702,
+      "learning_rate": 9.8033128590566e-06,
+      "loss": 0.8422,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9816857878795604,
+      "learning_rate": 9.802951870759878e-06,
+      "loss": 1.1497,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4825176137391896,
+      "learning_rate": 9.802590558156863e-06,
+      "loss": 0.8794,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.64229382344146,
+      "learning_rate": 9.80222892127195e-06,
+      "loss": 0.9127,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4356716503948197,
+      "learning_rate": 9.801866960129561e-06,
+      "loss": 0.7799,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.572036977908545,
+      "learning_rate": 9.801504674754134e-06,
+      "loss": 0.8608,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5817357085829638,
+      "learning_rate": 9.801142065170132e-06,
+      "loss": 0.9045,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.495132733617185,
+      "learning_rate": 9.80077913140204e-06,
+      "loss": 0.8792,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4955302647396005,
+      "learning_rate": 9.800415873474363e-06,
+      "loss": 0.8211,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5802661811514742,
+      "learning_rate": 9.80005229141163e-06,
+      "loss": 0.8688,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8842380797754821,
+      "learning_rate": 9.799688385238388e-06,
+      "loss": 1.1488,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5840031616198111,
+      "learning_rate": 9.799324154979215e-06,
+      "loss": 0.7911,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7999383203826839,
+      "learning_rate": 9.798959600658697e-06,
+      "loss": 1.1829,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.512162564849784,
+      "learning_rate": 9.798594722301455e-06,
+      "loss": 0.8454,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6902123568426313,
+      "learning_rate": 9.798229519932125e-06,
+      "loss": 0.8018,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4787318483175864,
+      "learning_rate": 9.797863993575365e-06,
+      "loss": 0.744,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.506274356964622,
+      "learning_rate": 9.797498143255859e-06,
+      "loss": 0.6985,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6210371067957121,
+      "learning_rate": 9.797131968998305e-06,
+      "loss": 0.9201,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9276205140553885,
+      "learning_rate": 9.796765470827435e-06,
+      "loss": 1.2148,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5595260770203656,
+      "learning_rate": 9.796398648767989e-06,
+      "loss": 0.8049,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4807356023861347,
+      "learning_rate": 9.79603150284474e-06,
+      "loss": 0.7859,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4144421454948075,
+      "learning_rate": 9.795664033082476e-06,
+      "loss": 0.8067,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5984690238716062,
+      "learning_rate": 9.795296239506011e-06,
+      "loss": 0.8363,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7086357408694295,
+      "learning_rate": 9.794928122140179e-06,
+      "loss": 0.8154,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6046329493160347,
+      "learning_rate": 9.794559681009837e-06,
+      "loss": 0.7742,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4861206995305734,
+      "learning_rate": 9.794190916139861e-06,
+      "loss": 0.8119,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.646525136148379,
+      "learning_rate": 9.79382182755515e-06,
+      "loss": 0.8068,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5592221437211087,
+      "learning_rate": 9.79345241528063e-06,
+      "loss": 0.8823,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.385837525038439,
+      "learning_rate": 9.79308267934124e-06,
+      "loss": 0.7938,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5882318770697132,
+      "learning_rate": 9.79271261976195e-06,
+      "loss": 0.8718,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9924734439766923,
+      "learning_rate": 9.792342236567743e-06,
+      "loss": 1.1529,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5180753555449649,
+      "learning_rate": 9.79197152978363e-06,
+      "loss": 0.8271,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5575423371826715,
+      "learning_rate": 9.791600499434642e-06,
+      "loss": 0.8882,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5807953831444395,
+      "learning_rate": 9.791229145545832e-06,
+      "loss": 0.8461,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9030204631492412,
+      "learning_rate": 9.790857468142274e-06,
+      "loss": 1.19,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4742362590822566,
+      "learning_rate": 9.790485467249065e-06,
+      "loss": 0.8686,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6644699931148292,
+      "learning_rate": 9.790113142891323e-06,
+      "loss": 0.9355,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4882826852943436,
+      "learning_rate": 9.789740495094186e-06,
+      "loss": 0.7823,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8543869463904564,
+      "learning_rate": 9.789367523882822e-06,
+      "loss": 1.1814,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.460642725609597,
+      "learning_rate": 9.78899422928241e-06,
+      "loss": 0.8275,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4911956750238113,
+      "learning_rate": 9.788620611318157e-06,
+      "loss": 0.8702,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6122796984548178,
+      "learning_rate": 9.788246670015289e-06,
+      "loss": 0.9353,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8684723600441826,
+      "learning_rate": 9.787872405399059e-06,
+      "loss": 1.1727,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4535354234294837,
+      "learning_rate": 9.787497817494734e-06,
+      "loss": 0.8486,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8607205042600619,
+      "learning_rate": 9.78712290632761e-06,
+      "loss": 1.1782,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8005297584029993,
+      "learning_rate": 9.786747671923003e-06,
+      "loss": 1.1533,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4322755033143728,
+      "learning_rate": 9.786372114306244e-06,
+      "loss": 0.8055,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6949315201653319,
+      "learning_rate": 9.785996233502697e-06,
+      "loss": 0.8137,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5014190562268706,
+      "learning_rate": 9.785620029537741e-06,
+      "loss": 0.7736,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4009907262020476,
+      "learning_rate": 9.785243502436776e-06,
+      "loss": 0.8824,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5103592923137423,
+      "learning_rate": 9.78486665222523e-06,
+      "loss": 0.892,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7978399952160373,
+      "learning_rate": 9.784489478928545e-06,
+      "loss": 0.8985,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5970410404323987,
+      "learning_rate": 9.784111982572188e-06,
+      "loss": 0.8033,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1592982381505892,
+      "learning_rate": 9.783734163181653e-06,
+      "loss": 1.158,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4491139836061901,
+      "learning_rate": 9.783356020782448e-06,
+      "loss": 0.8536,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5309159256057085,
+      "learning_rate": 9.782977555400106e-06,
+      "loss": 0.9239,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4467469986087207,
+      "learning_rate": 9.782598767060186e-06,
+      "loss": 0.8158,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4916075589874864,
+      "learning_rate": 9.782219655788257e-06,
+      "loss": 0.8621,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5996329867560244,
+      "learning_rate": 9.781840221609922e-06,
+      "loss": 0.8734,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4481352930461173,
+      "learning_rate": 9.781460464550802e-06,
+      "loss": 0.8206,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5387136935042627,
+      "learning_rate": 9.781080384636539e-06,
+      "loss": 0.821,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4524478710416522,
+      "learning_rate": 9.780699981892793e-06,
+      "loss": 0.8188,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0502036669747203,
+      "learning_rate": 9.780319256345255e-06,
+      "loss": 1.1834,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4827645758652848,
+      "learning_rate": 9.77993820801963e-06,
+      "loss": 0.7462,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5676902291255275,
+      "learning_rate": 9.779556836941646e-06,
+      "loss": 0.7887,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.570302620741426,
+      "learning_rate": 9.779175143137055e-06,
+      "loss": 0.8601,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5130651986180583,
+      "learning_rate": 9.778793126631632e-06,
+      "loss": 0.7974,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5460217504723186,
+      "learning_rate": 9.778410787451168e-06,
+      "loss": 0.8215,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5482815896105826,
+      "learning_rate": 9.778028125621481e-06,
+      "loss": 0.8432,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.425089310779897,
+      "learning_rate": 9.777645141168411e-06,
+      "loss": 0.8456,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.447066841273666,
+      "learning_rate": 9.777261834117814e-06,
+      "loss": 0.802,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6027690940761827,
+      "learning_rate": 9.776878204495574e-06,
+      "loss": 0.8732,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5018584687698269,
+      "learning_rate": 9.776494252327597e-06,
+      "loss": 0.7605,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0469217417949142,
+      "learning_rate": 9.776109977639804e-06,
+      "loss": 1.148,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4886294184825286,
+      "learning_rate": 9.775725380458145e-06,
+      "loss": 0.7644,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.541150848069218,
+      "learning_rate": 9.775340460808589e-06,
+      "loss": 0.8111,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4055461771970452,
+      "learning_rate": 9.774955218717123e-06,
+      "loss": 0.7777,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4823678578145854,
+      "learning_rate": 9.774569654209764e-06,
+      "loss": 0.7948,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.827097810162567,
+      "learning_rate": 9.774183767312545e-06,
+      "loss": 1.1593,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8379624305681712,
+      "learning_rate": 9.77379755805152e-06,
+      "loss": 1.1903,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.429181293296838,
+      "learning_rate": 9.773411026452768e-06,
+      "loss": 0.8137,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3976240793968402,
+      "learning_rate": 9.773024172542389e-06,
+      "loss": 0.8866,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5318989016343512,
+      "learning_rate": 9.772636996346503e-06,
+      "loss": 0.8699,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6253614944609107,
+      "learning_rate": 9.772249497891254e-06,
+      "loss": 0.8171,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0469555977503313,
+      "learning_rate": 9.771861677202804e-06,
+      "loss": 1.1813,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6177812148047643,
+      "learning_rate": 9.771473534307345e-06,
+      "loss": 0.8568,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4767819250951477,
+      "learning_rate": 9.771085069231082e-06,
+      "loss": 0.9074,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5602706941923075,
+      "learning_rate": 9.770696282000245e-06,
+      "loss": 0.8657,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.511199648796229,
+      "learning_rate": 9.770307172641088e-06,
+      "loss": 0.8177,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5269963836368348,
+      "learning_rate": 9.76991774117988e-06,
+      "loss": 0.793,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.010890221253354,
+      "learning_rate": 9.76952798764292e-06,
+      "loss": 1.2296,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6302748602203143,
+      "learning_rate": 9.769137912056523e-06,
+      "loss": 0.9168,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8426255534483389,
+      "learning_rate": 9.76874751444703e-06,
+      "loss": 1.1568,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5922851810237115,
+      "learning_rate": 9.7683567948408e-06,
+      "loss": 0.7692,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.566652154364714,
+      "learning_rate": 9.767965753264218e-06,
+      "loss": 0.7822,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4300587126065634,
+      "learning_rate": 9.767574389743683e-06,
+      "loss": 0.7466,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.558900846418108,
+      "learning_rate": 9.767182704305625e-06,
+      "loss": 0.8102,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4941730998603502,
+      "learning_rate": 9.766790696976489e-06,
+      "loss": 0.8531,
+      "step": 1554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5629299857961287,
+      "learning_rate": 9.766398367782744e-06,
+      "loss": 0.8043,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.529365923869202,
+      "learning_rate": 9.766005716750884e-06,
+      "loss": 0.8151,
+      "step": 1556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.515383482519452,
+      "learning_rate": 9.76561274390742e-06,
+      "loss": 0.8664,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6261631614332202,
+      "learning_rate": 9.765219449278885e-06,
+      "loss": 0.8843,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.419538000836456,
+      "learning_rate": 9.764825832891837e-06,
+      "loss": 0.7278,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0606748374123418,
+      "learning_rate": 9.764431894772855e-06,
+      "loss": 1.1815,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.513236066606246,
+      "learning_rate": 9.764037634948536e-06,
+      "loss": 0.8176,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8597541750032723,
+      "learning_rate": 9.7636430534455e-06,
+      "loss": 1.155,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6011073045904582,
+      "learning_rate": 9.763248150290394e-06,
+      "loss": 0.844,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5256822446620584,
+      "learning_rate": 9.762852925509882e-06,
+      "loss": 0.911,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.618662847792621,
+      "learning_rate": 9.762457379130649e-06,
+      "loss": 0.8191,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.601523173857137,
+      "learning_rate": 9.762061511179404e-06,
+      "loss": 0.8384,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4883050780964193,
+      "learning_rate": 9.761665321682875e-06,
+      "loss": 0.7799,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.562189650546599,
+      "learning_rate": 9.761268810667817e-06,
+      "loss": 0.8477,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.193667133727517,
+      "learning_rate": 9.760871978161e-06,
+      "loss": 1.2182,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5791481548647144,
+      "learning_rate": 9.760474824189222e-06,
+      "loss": 0.7709,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5643770932118597,
+      "learning_rate": 9.760077348779298e-06,
+      "loss": 0.8382,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7273908665768951,
+      "learning_rate": 9.759679551958068e-06,
+      "loss": 0.8113,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8131383980359979,
+      "learning_rate": 9.759281433752389e-06,
+      "loss": 1.141,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6165712698132797,
+      "learning_rate": 9.758882994189145e-06,
+      "loss": 0.8206,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7753753442074367,
+      "learning_rate": 9.75848423329524e-06,
+      "loss": 0.8438,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5264704262010795,
+      "learning_rate": 9.7580851510976e-06,
+      "loss": 0.8779,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.631113433724481,
+      "learning_rate": 9.757685747623169e-06,
+      "loss": 0.7932,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5479028858885668,
+      "learning_rate": 9.757286022898918e-06,
+      "loss": 0.885,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4445042117203173,
+      "learning_rate": 9.756885976951835e-06,
+      "loss": 0.8372,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5081997649156735,
+      "learning_rate": 9.756485609808934e-06,
+      "loss": 0.8447,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4155712115763956,
+      "learning_rate": 9.75608492149725e-06,
+      "loss": 0.866,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6566785977571656,
+      "learning_rate": 9.755683912043836e-06,
+      "loss": 0.8761,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6474805475883003,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.854,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4853855333542811,
+      "learning_rate": 9.754880929820149e-06,
+      "loss": 0.7515,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5044393710760613,
+      "learning_rate": 9.754478957104094e-06,
+      "loss": 0.838,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6174230905492268,
+      "learning_rate": 9.75407666335475e-06,
+      "loss": 0.862,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1999052919767088,
+      "learning_rate": 9.75367404859928e-06,
+      "loss": 1.1896,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9775660498868867,
+      "learning_rate": 9.753271112864866e-06,
+      "loss": 1.1426,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5607453626554773,
+      "learning_rate": 9.752867856178719e-06,
+      "loss": 0.8262,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5355587884381978,
+      "learning_rate": 9.752464278568066e-06,
+      "loss": 0.833,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.552507424276035,
+      "learning_rate": 9.752060380060156e-06,
+      "loss": 0.8163,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5448047551595698,
+      "learning_rate": 9.751656160682265e-06,
+      "loss": 0.9331,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4905546644728287,
+      "learning_rate": 9.751251620461683e-06,
+      "loss": 0.8142,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4920693436656194,
+      "learning_rate": 9.75084675942573e-06,
+      "loss": 0.8839,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8640971351172513,
+      "learning_rate": 9.750441577601738e-06,
+      "loss": 1.1824,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6794332669502086,
+      "learning_rate": 9.750036075017068e-06,
+      "loss": 0.8747,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5443385615478114,
+      "learning_rate": 9.7496302516991e-06,
+      "loss": 0.8118,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6498921883344098,
+      "learning_rate": 9.749224107675239e-06,
+      "loss": 0.8177,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4392998257719518,
+      "learning_rate": 9.748817642972905e-06,
+      "loss": 0.8275,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5418585095034405,
+      "learning_rate": 9.748410857619547e-06,
+      "loss": 0.8588,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.468319085881354,
+      "learning_rate": 9.748003751642628e-06,
+      "loss": 0.8244,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5544739155997838,
+      "learning_rate": 9.747596325069638e-06,
+      "loss": 0.8278,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5857674438921594,
+      "learning_rate": 9.747188577928089e-06,
+      "loss": 0.9021,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6330246634932593,
+      "learning_rate": 9.746780510245512e-06,
+      "loss": 0.8882,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.020640255631947,
+      "learning_rate": 9.74637212204946e-06,
+      "loss": 0.9223,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4763635857374666,
+      "learning_rate": 9.745963413367511e-06,
+      "loss": 0.8214,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.522480455458243,
+      "learning_rate": 9.74555438422726e-06,
+      "loss": 0.8077,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.545304232439133,
+      "learning_rate": 9.745145034656325e-06,
+      "loss": 0.8768,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.462005431515541,
+      "learning_rate": 9.744735364682347e-06,
+      "loss": 0.7482,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5309857620212377,
+      "learning_rate": 9.744325374332986e-06,
+      "loss": 0.8582,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5750460300317697,
+      "learning_rate": 9.74391506363593e-06,
+      "loss": 0.843,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.506357040075656,
+      "learning_rate": 9.74350443261888e-06,
+      "loss": 0.8294,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6059452791451545,
+      "learning_rate": 9.743093481309563e-06,
+      "loss": 0.8354,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5212094461268337,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.8456,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4914197812146246,
+      "learning_rate": 9.742270617925148e-06,
+      "loss": 0.8993,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4445695564865404,
+      "learning_rate": 9.741858705905609e-06,
+      "loss": 0.8626,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5311151263751501,
+      "learning_rate": 9.74144647370493e-06,
+      "loss": 0.8097,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5540372202517865,
+      "learning_rate": 9.741033921350941e-06,
+      "loss": 0.8272,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.902374857213976,
+      "learning_rate": 9.740621048871501e-06,
+      "loss": 1.1535,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.516597058394549,
+      "learning_rate": 9.74020785629449e-06,
+      "loss": 0.8848,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4783985389734733,
+      "learning_rate": 9.739794343647802e-06,
+      "loss": 0.8706,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4985533830083768,
+      "learning_rate": 9.739380510959365e-06,
+      "loss": 0.8503,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5211375808335061,
+      "learning_rate": 9.738966358257116e-06,
+      "loss": 0.9254,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5229077503549107,
+      "learning_rate": 9.738551885569022e-06,
+      "loss": 0.8359,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.820993167648609,
+      "learning_rate": 9.738137092923072e-06,
+      "loss": 1.1883,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.562891885063231,
+      "learning_rate": 9.73772198034727e-06,
+      "loss": 0.8385,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8270678503712992,
+      "learning_rate": 9.737306547869645e-06,
+      "loss": 1.1241,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4828856995959205,
+      "learning_rate": 9.73689079551825e-06,
+      "loss": 0.8357,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5924636544676867,
+      "learning_rate": 9.736474723321159e-06,
+      "loss": 0.8112,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4639223366458662,
+      "learning_rate": 9.736058331306461e-06,
+      "loss": 0.8454,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.541009111588045,
+      "learning_rate": 9.735641619502277e-06,
+      "loss": 0.809,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.437058264494758,
+      "learning_rate": 9.735224587936743e-06,
+      "loss": 0.7504,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.490602647157641,
+      "learning_rate": 9.734807236638015e-06,
+      "loss": 0.8573,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4525107881689685,
+      "learning_rate": 9.734389565634277e-06,
+      "loss": 0.7874,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.575529978469036,
+      "learning_rate": 9.733971574953726e-06,
+      "loss": 0.7654,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.004436012358606,
+      "learning_rate": 9.733553264624593e-06,
+      "loss": 0.8332,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.978151023509133,
+      "learning_rate": 9.73313463467512e-06,
+      "loss": 1.173,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4578954697664561,
+      "learning_rate": 9.732715685133572e-06,
+      "loss": 0.7895,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5855147013660655,
+      "learning_rate": 9.732296416028239e-06,
+      "loss": 0.8316,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8309499242629935,
+      "learning_rate": 9.731876827387433e-06,
+      "loss": 1.1713,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5135406747756008,
+      "learning_rate": 9.73145691923948e-06,
+      "loss": 0.852,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6245707033690688,
+      "learning_rate": 9.73103669161274e-06,
+      "loss": 0.8243,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.450746219526346,
+      "learning_rate": 9.730616144535581e-06,
+      "loss": 0.839,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5363749401661804,
+      "learning_rate": 9.730195278036405e-06,
+      "loss": 0.8135,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5532327193979731,
+      "learning_rate": 9.729774092143627e-06,
+      "loss": 0.8142,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.477348693834012,
+      "learning_rate": 9.729352586885687e-06,
+      "loss": 0.8481,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.982004071184762,
+      "learning_rate": 9.728930762291046e-06,
+      "loss": 1.1643,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9184945549533521,
+      "learning_rate": 9.728508618388186e-06,
+      "loss": 1.2233,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8206029141756019,
+      "learning_rate": 9.728086155205614e-06,
+      "loss": 1.1858,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6320246874260242,
+      "learning_rate": 9.727663372771852e-06,
+      "loss": 0.9328,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5676439956669872,
+      "learning_rate": 9.727240271115448e-06,
+      "loss": 0.7884,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9871983160944695,
+      "learning_rate": 9.726816850264971e-06,
+      "loss": 0.7755,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4911241582228956,
+      "learning_rate": 9.726393110249012e-06,
+      "loss": 0.825,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4602639356893825,
+      "learning_rate": 9.725969051096185e-06,
+      "loss": 0.8068,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3886021944492761,
+      "learning_rate": 9.725544672835118e-06,
+      "loss": 1.1916,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5131917405124988,
+      "learning_rate": 9.72511997549447e-06,
+      "loss": 0.8899,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4705134975539087,
+      "learning_rate": 9.724694959102918e-06,
+      "loss": 0.8234,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.128559687342653,
+      "learning_rate": 9.724269623689158e-06,
+      "loss": 1.1929,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5025403322874118,
+      "learning_rate": 9.72384396928191e-06,
+      "loss": 0.81,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.555435411350934,
+      "learning_rate": 9.723417995909915e-06,
+      "loss": 0.8151,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.514204807186842,
+      "learning_rate": 9.722991703601936e-06,
+      "loss": 0.8844,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5702605979905422,
+      "learning_rate": 9.72256509238676e-06,
+      "loss": 0.7571,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5417940807373565,
+      "learning_rate": 9.722138162293187e-06,
+      "loss": 0.8097,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1852843789942191,
+      "learning_rate": 9.721710913350048e-06,
+      "loss": 1.1835,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5278967487269908,
+      "learning_rate": 9.721283345586191e-06,
+      "loss": 0.8881,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5557144690578057,
+      "learning_rate": 9.720855459030489e-06,
+      "loss": 0.9165,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.475850216630695,
+      "learning_rate": 9.720427253711831e-06,
+      "loss": 0.8597,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8964989655734535,
+      "learning_rate": 9.719998729659129e-06,
+      "loss": 1.1653,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.38238178685736,
+      "learning_rate": 9.71956988690132e-06,
+      "loss": 0.7778,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7232355346277546,
+      "learning_rate": 9.719140725467362e-06,
+      "loss": 0.8164,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5585898877866649,
+      "learning_rate": 9.718711245386232e-06,
+      "loss": 0.8672,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3973544877900228,
+      "learning_rate": 9.718281446686926e-06,
+      "loss": 0.8284,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4752334985864726,
+      "learning_rate": 9.717851329398469e-06,
+      "loss": 0.7674,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5296708480556587,
+      "learning_rate": 9.717420893549902e-06,
+      "loss": 0.8824,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.548199121009184,
+      "learning_rate": 9.71699013917029e-06,
+      "loss": 0.8609,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0407842058765289,
+      "learning_rate": 9.716559066288716e-06,
+      "loss": 1.1901,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4881735800219105,
+      "learning_rate": 9.716127674934291e-06,
+      "loss": 0.7915,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8098622496458227,
+      "learning_rate": 9.715695965136139e-06,
+      "loss": 1.1799,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.466229642879152,
+      "learning_rate": 9.715263936923413e-06,
+      "loss": 0.8702,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.589982846154434,
+      "learning_rate": 9.714831590325286e-06,
+      "loss": 0.8377,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4760628580750468,
+      "learning_rate": 9.714398925370946e-06,
+      "loss": 0.8133,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4681820268432904,
+      "learning_rate": 9.713965942089612e-06,
+      "loss": 0.8062,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5914508007456492,
+      "learning_rate": 9.71353264051052e-06,
+      "loss": 0.9552,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4149153146506217,
+      "learning_rate": 9.713099020662922e-06,
+      "loss": 0.8842,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3767279424585575,
+      "learning_rate": 9.712665082576104e-06,
+      "loss": 0.7917,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6455039654789445,
+      "learning_rate": 9.712230826279363e-06,
+      "loss": 0.793,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5443356530905545,
+      "learning_rate": 9.71179625180202e-06,
+      "loss": 0.9405,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5945932195345813,
+      "learning_rate": 9.711361359173422e-06,
+      "loss": 0.8455,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.213131833909528,
+      "learning_rate": 9.71092614842293e-06,
+      "loss": 1.1983,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5333978683809595,
+      "learning_rate": 9.710490619579933e-06,
+      "loss": 0.8661,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.534402845248015,
+      "learning_rate": 9.710054772673839e-06,
+      "loss": 0.8026,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4759418252359835,
+      "learning_rate": 9.709618607734075e-06,
+      "loss": 0.8948,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8173881197474613,
+      "learning_rate": 9.709182124790094e-06,
+      "loss": 1.1829,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.49791759692554,
+      "learning_rate": 9.708745323871369e-06,
+      "loss": 0.7922,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3852993818674915,
+      "learning_rate": 9.708308205007391e-06,
+      "loss": 0.7752,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6008010360542064,
+      "learning_rate": 9.707870768227677e-06,
+      "loss": 0.7751,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.881584166135359,
+      "learning_rate": 9.707433013561765e-06,
+      "loss": 1.1321,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4908590197680862,
+      "learning_rate": 9.706994941039209e-06,
+      "loss": 0.7502,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3632819123284918,
+      "learning_rate": 9.706556550689593e-06,
+      "loss": 0.8297,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.500361643935751,
+      "learning_rate": 9.706117842542517e-06,
+      "loss": 0.7485,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.412177840869574,
+      "learning_rate": 9.705678816627601e-06,
+      "loss": 0.8444,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4053669329528855,
+      "learning_rate": 9.705239472974495e-06,
+      "loss": 0.7415,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6369343494640611,
+      "learning_rate": 9.704799811612858e-06,
+      "loss": 0.8382,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4399529872397383,
+      "learning_rate": 9.704359832572378e-06,
+      "loss": 0.7943,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8819670372697286,
+      "learning_rate": 9.703919535882767e-06,
+      "loss": 1.1698,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8284847659341459,
+      "learning_rate": 9.703478921573753e-06,
+      "loss": 1.1445,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8097446502843204,
+      "learning_rate": 9.703037989675088e-06,
+      "loss": 1.1868,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5347071011631346,
+      "learning_rate": 9.702596740216541e-06,
+      "loss": 0.8518,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5870662861378397,
+      "learning_rate": 9.702155173227911e-06,
+      "loss": 0.8512,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5379043403593504,
+      "learning_rate": 9.70171328873901e-06,
+      "loss": 0.8222,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.845065133551567,
+      "learning_rate": 9.701271086779678e-06,
+      "loss": 1.156,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5488072125359502,
+      "learning_rate": 9.700828567379772e-06,
+      "loss": 0.9055,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6891927046268491,
+      "learning_rate": 9.700385730569171e-06,
+      "loss": 0.8476,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4939929915660417,
+      "learning_rate": 9.699942576377779e-06,
+      "loss": 0.7776,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5925000552209982,
+      "learning_rate": 9.699499104835514e-06,
+      "loss": 0.9079,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5722214073225893,
+      "learning_rate": 9.699055315972328e-06,
+      "loss": 0.8073,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5545814010449013,
+      "learning_rate": 9.698611209818178e-06,
+      "loss": 0.8066,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8806955295773359,
+      "learning_rate": 9.698166786403057e-06,
+      "loss": 1.1803,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4686405393031277,
+      "learning_rate": 9.697722045756973e-06,
+      "loss": 0.7562,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5699556054544272,
+      "learning_rate": 9.697276987909951e-06,
+      "loss": 0.8435,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5128329517989345,
+      "learning_rate": 9.696831612892048e-06,
+      "loss": 0.8269,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8431566538279116,
+      "learning_rate": 9.696385920733335e-06,
+      "loss": 1.1715,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7200554272170396,
+      "learning_rate": 9.695939911463904e-06,
+      "loss": 0.829,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6162504723424296,
+      "learning_rate": 9.695493585113873e-06,
+      "loss": 0.7914,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8270729569141857,
+      "learning_rate": 9.695046941713379e-06,
+      "loss": 1.1894,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4696685078097274,
+      "learning_rate": 9.694599981292578e-06,
+      "loss": 0.7616,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6114130894525227,
+      "learning_rate": 9.694152703881653e-06,
+      "loss": 0.7942,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5874158109834735,
+      "learning_rate": 9.693705109510803e-06,
+      "loss": 0.8807,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5242717956570906,
+      "learning_rate": 9.693257198210251e-06,
+      "loss": 0.8851,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4181308437409006,
+      "learning_rate": 9.69280897001024e-06,
+      "loss": 0.8883,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6331227105813944,
+      "learning_rate": 9.69236042494104e-06,
+      "loss": 0.8319,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4522987714088507,
+      "learning_rate": 9.691911563032932e-06,
+      "loss": 0.8675,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.440827238388004,
+      "learning_rate": 9.691462384316226e-06,
+      "loss": 0.8664,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6223928363189393,
+      "learning_rate": 9.691012888821254e-06,
+      "loss": 0.9025,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.464324892960736,
+      "learning_rate": 9.690563076578364e-06,
+      "loss": 0.8914,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5829757379714613,
+      "learning_rate": 9.690112947617929e-06,
+      "loss": 0.7316,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4012009351665144,
+      "learning_rate": 9.689662501970343e-06,
+      "loss": 0.7813,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.981365538394024,
+      "learning_rate": 9.689211739666023e-06,
+      "loss": 1.2022,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4941257252165787,
+      "learning_rate": 9.688760660735403e-06,
+      "loss": 0.8789,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.663006714337196,
+      "learning_rate": 9.688309265208941e-06,
+      "loss": 0.8431,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8469568396807724,
+      "learning_rate": 9.687857553117119e-06,
+      "loss": 1.1719,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.520666432172101,
+      "learning_rate": 9.687405524490433e-06,
+      "loss": 0.7752,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.464014520658535,
+      "learning_rate": 9.686953179359408e-06,
+      "loss": 0.8272,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5526223490259756,
+      "learning_rate": 9.686500517754589e-06,
+      "loss": 0.7765,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5800301252872575,
+      "learning_rate": 9.686047539706536e-06,
+      "loss": 0.903,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.561692851940971,
+      "learning_rate": 9.68559424524584e-06,
+      "loss": 0.7937,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8583825366753453,
+      "learning_rate": 9.685140634403106e-06,
+      "loss": 1.1621,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8801039464378387,
+      "learning_rate": 9.684686707208962e-06,
+      "loss": 1.1957,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4569117609816349,
+      "learning_rate": 9.68423246369406e-06,
+      "loss": 0.8619,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4406207939514377,
+      "learning_rate": 9.68377790388907e-06,
+      "loss": 0.7772,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5253511878443835,
+      "learning_rate": 9.683323027824687e-06,
+      "loss": 0.7999,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8365798442417645,
+      "learning_rate": 9.682867835531624e-06,
+      "loss": 1.1777,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5742170997246374,
+      "learning_rate": 9.682412327040617e-06,
+      "loss": 0.8897,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.668988513159721,
+      "learning_rate": 9.681956502382423e-06,
+      "loss": 0.9119,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5716584749091955,
+      "learning_rate": 9.681500361587818e-06,
+      "loss": 0.9084,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3578454896936722,
+      "learning_rate": 9.681043904687605e-06,
+      "loss": 0.7841,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6065361944319299,
+      "learning_rate": 9.680587131712605e-06,
+      "loss": 0.9119,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5205484398859332,
+      "learning_rate": 9.680130042693657e-06,
+      "loss": 0.835,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5306054478117836,
+      "learning_rate": 9.679672637661627e-06,
+      "loss": 0.8301,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3510576427864385,
+      "learning_rate": 9.6792149166474e-06,
+      "loss": 0.8059,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5781590120740918,
+      "learning_rate": 9.678756879681884e-06,
+      "loss": 0.7731,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6688549350329707,
+      "learning_rate": 9.678298526796002e-06,
+      "loss": 0.8226,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5862792942823696,
+      "learning_rate": 9.677839858020709e-06,
+      "loss": 0.8524,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4487138889108717,
+      "learning_rate": 9.677380873386968e-06,
+      "loss": 0.8052,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6050776691542659,
+      "learning_rate": 9.676921572925777e-06,
+      "loss": 0.8389,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4527464183103627,
+      "learning_rate": 9.676461956668148e-06,
+      "loss": 0.7924,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6100323349956902,
+      "learning_rate": 9.676002024645114e-06,
+      "loss": 0.8305,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4744206179733308,
+      "learning_rate": 9.675541776887731e-06,
+      "loss": 0.8078,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9540264302670897,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 1.1727,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5586735225640893,
+      "learning_rate": 9.674620334294246e-06,
+      "loss": 0.8913,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6136987113322039,
+      "learning_rate": 9.674159139520363e-06,
+      "loss": 0.8612,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6435939542342457,
+      "learning_rate": 9.673697629136566e-06,
+      "loss": 0.8432,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6254283917331394,
+      "learning_rate": 9.673235803174018e-06,
+      "loss": 0.8067,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9389270769073029,
+      "learning_rate": 9.672773661663903e-06,
+      "loss": 1.1524,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4837075164280675,
+      "learning_rate": 9.672311204637426e-06,
+      "loss": 0.8768,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5767991968162345,
+      "learning_rate": 9.671848432125812e-06,
+      "loss": 0.8347,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6535477450131855,
+      "learning_rate": 9.671385344160309e-06,
+      "loss": 0.8516,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.513176100383834,
+      "learning_rate": 9.670921940772186e-06,
+      "loss": 0.7918,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5335213234117717,
+      "learning_rate": 9.670458221992733e-06,
+      "loss": 0.9053,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4993628500123712,
+      "learning_rate": 9.66999418785326e-06,
+      "loss": 0.8964,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5679626871207248,
+      "learning_rate": 9.669529838385102e-06,
+      "loss": 0.8191,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4816658730262284,
+      "learning_rate": 9.669065173619612e-06,
+      "loss": 0.8692,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9700913678395203,
+      "learning_rate": 9.668600193588165e-06,
+      "loss": 1.1645,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.468160328747445,
+      "learning_rate": 9.668134898322157e-06,
+      "loss": 0.8437,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5107270157570762,
+      "learning_rate": 9.667669287853006e-06,
+      "loss": 0.875,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5435556474230339,
+      "learning_rate": 9.667203362212152e-06,
+      "loss": 0.9047,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5510476452922575,
+      "learning_rate": 9.666737121431055e-06,
+      "loss": 0.8127,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.553092568257882,
+      "learning_rate": 9.666270565541196e-06,
+      "loss": 0.822,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9546332086625076,
+      "learning_rate": 9.665803694574079e-06,
+      "loss": 1.1728,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5084168363121073,
+      "learning_rate": 9.665336508561225e-06,
+      "loss": 0.8489,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6698513623819695,
+      "learning_rate": 9.664869007534185e-06,
+      "loss": 0.9352,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6591394351454614,
+      "learning_rate": 9.664401191524522e-06,
+      "loss": 0.7893,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5129262070241574,
+      "learning_rate": 9.663933060563824e-06,
+      "loss": 0.8436,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5192580700707576,
+      "learning_rate": 9.663464614683702e-06,
+      "loss": 0.8233,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6460297487561517,
+      "learning_rate": 9.662995853915785e-06,
+      "loss": 0.8645,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7052106840703782,
+      "learning_rate": 9.662526778291725e-06,
+      "loss": 0.8707,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5208134976902332,
+      "learning_rate": 9.662057387843196e-06,
+      "loss": 0.7294,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5846465073453435,
+      "learning_rate": 9.66158768260189e-06,
+      "loss": 0.8496,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6143837844637892,
+      "learning_rate": 9.661117662599527e-06,
+      "loss": 0.9072,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5528409262348821,
+      "learning_rate": 9.66064732786784e-06,
+      "loss": 0.8341,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6034731646820828,
+      "learning_rate": 9.660176678438588e-06,
+      "loss": 0.7606,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1485660012523888,
+      "learning_rate": 9.659705714343551e-06,
+      "loss": 1.1838,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.942463094180893,
+      "learning_rate": 9.659234435614529e-06,
+      "loss": 1.1757,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5729860579640866,
+      "learning_rate": 9.658762842283343e-06,
+      "loss": 0.8553,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6418064827736678,
+      "learning_rate": 9.658290934381837e-06,
+      "loss": 0.8452,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4643190953413758,
+      "learning_rate": 9.657818711941877e-06,
+      "loss": 0.8631,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.465402551095835,
+      "learning_rate": 9.657346174995346e-06,
+      "loss": 0.7899,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5352074145945847,
+      "learning_rate": 9.656873323574152e-06,
+      "loss": 0.8929,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.449831483293168,
+      "learning_rate": 9.656400157710221e-06,
+      "loss": 0.8194,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5206170485420745,
+      "learning_rate": 9.655926677435506e-06,
+      "loss": 0.8191,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5922879442628086,
+      "learning_rate": 9.655452882781972e-06,
+      "loss": 0.8223,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9004144095111877,
+      "learning_rate": 9.654978773781617e-06,
+      "loss": 1.194,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5918330055100864,
+      "learning_rate": 9.65450435046645e-06,
+      "loss": 0.9016,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4864201664851162,
+      "learning_rate": 9.654029612868507e-06,
+      "loss": 0.8024,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0799982012085143,
+      "learning_rate": 9.653554561019843e-06,
+      "loss": 1.2023,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4693907542152291,
+      "learning_rate": 9.653079194952532e-06,
+      "loss": 0.8172,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6978351899966562,
+      "learning_rate": 9.652603514698674e-06,
+      "loss": 0.859,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5316583722538475,
+      "learning_rate": 9.652127520290388e-06,
+      "loss": 0.7953,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5944718165226932,
+      "learning_rate": 9.651651211759814e-06,
+      "loss": 0.8765,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.691153840271507,
+      "learning_rate": 9.651174589139115e-06,
+      "loss": 0.9134,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5928018952598135,
+      "learning_rate": 9.650697652460471e-06,
+      "loss": 1.1802,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4575409701539865,
+      "learning_rate": 9.650220401756088e-06,
+      "loss": 0.791,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.420906813906397,
+      "learning_rate": 9.649742837058189e-06,
+      "loss": 0.8431,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3935946370901644,
+      "learning_rate": 9.649264958399022e-06,
+      "loss": 0.836,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6397407284761742,
+      "learning_rate": 9.648786765810853e-06,
+      "loss": 0.9505,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.004196427702683,
+      "learning_rate": 9.648308259325973e-06,
+      "loss": 1.1863,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8371272119375583,
+      "learning_rate": 9.647829438976689e-06,
+      "loss": 1.1771,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5494113910947722,
+      "learning_rate": 9.647350304795333e-06,
+      "loss": 0.8044,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4829373228421028,
+      "learning_rate": 9.646870856814259e-06,
+      "loss": 0.7746,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4223975837845726,
+      "learning_rate": 9.646391095065838e-06,
+      "loss": 0.7936,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5277439315278891,
+      "learning_rate": 9.645911019582467e-06,
+      "loss": 0.885,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4960155050835333,
+      "learning_rate": 9.64543063039656e-06,
+      "loss": 0.8184,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6172671488436552,
+      "learning_rate": 9.644949927540553e-06,
+      "loss": 0.8108,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6599409473511135,
+      "learning_rate": 9.644468911046906e-06,
+      "loss": 1.1598,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6976516394521568,
+      "learning_rate": 9.6439875809481e-06,
+      "loss": 0.8224,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4330036058999358,
+      "learning_rate": 9.64350593727663e-06,
+      "loss": 0.8043,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4988459876749862,
+      "learning_rate": 9.643023980065025e-06,
+      "loss": 0.7638,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3990705807693076,
+      "learning_rate": 9.64254170934582e-06,
+      "loss": 0.7856,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.45729867400443,
+      "learning_rate": 9.642059125151586e-06,
+      "loss": 0.8615,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5661666821009343,
+      "learning_rate": 9.641576227514903e-06,
+      "loss": 0.8643,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.529652947545323,
+      "learning_rate": 9.641093016468381e-06,
+      "loss": 0.7785,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5282067353219049,
+      "learning_rate": 9.640609492044646e-06,
+      "loss": 0.793,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5961346165456891,
+      "learning_rate": 9.640125654276347e-06,
+      "loss": 0.8773,
+      "step": 1843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.608136513525393,
+      "learning_rate": 9.639641503196152e-06,
+      "loss": 0.8363,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6011860828587843,
+      "learning_rate": 9.639157038836755e-06,
+      "loss": 0.8059,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7847742494734866,
+      "learning_rate": 9.638672261230866e-06,
+      "loss": 0.8634,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6085087575125747,
+      "learning_rate": 9.638187170411218e-06,
+      "loss": 0.7925,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6105788528642824,
+      "learning_rate": 9.637701766410568e-06,
+      "loss": 0.8748,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4956439065009206,
+      "learning_rate": 9.63721604926169e-06,
+      "loss": 0.7961,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5730577589379264,
+      "learning_rate": 9.63673001899738e-06,
+      "loss": 0.8023,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6750585084901053,
+      "learning_rate": 9.636243675650456e-06,
+      "loss": 0.903,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5140501616418316,
+      "learning_rate": 9.635757019253758e-06,
+      "loss": 0.819,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5488672183551293,
+      "learning_rate": 9.635270049840146e-06,
+      "loss": 0.8091,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5512724062080712,
+      "learning_rate": 9.634782767442501e-06,
+      "loss": 0.8442,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5016854700880238,
+      "learning_rate": 9.634295172093727e-06,
+      "loss": 0.8592,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.453975023028802,
+      "learning_rate": 9.633807263826745e-06,
+      "loss": 0.8873,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4848079009089288,
+      "learning_rate": 9.633319042674497e-06,
+      "loss": 0.8333,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5440876797994385,
+      "learning_rate": 9.632830508669957e-06,
+      "loss": 0.8307,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7645069403788376,
+      "learning_rate": 9.632341661846107e-06,
+      "loss": 0.8262,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.566634535503002,
+      "learning_rate": 9.631852502235954e-06,
+      "loss": 0.7956,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5596251130748158,
+      "learning_rate": 9.631363029872529e-06,
+      "loss": 0.8929,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5179854677025637,
+      "learning_rate": 9.630873244788884e-06,
+      "loss": 0.805,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.090096791941229,
+      "learning_rate": 9.630383147018086e-06,
+      "loss": 1.1363,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9430071435023304,
+      "learning_rate": 9.629892736593231e-06,
+      "loss": 1.1606,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5453346312647547,
+      "learning_rate": 9.629402013547432e-06,
+      "loss": 0.8189,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6004283871551463,
+      "learning_rate": 9.628910977913821e-06,
+      "loss": 0.8366,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6779674885803768,
+      "learning_rate": 9.628419629725558e-06,
+      "loss": 0.855,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1112261931216882,
+      "learning_rate": 9.627927969015817e-06,
+      "loss": 1.1756,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.617339152264514,
+      "learning_rate": 9.627435995817799e-06,
+      "loss": 0.8174,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6176686794045037,
+      "learning_rate": 9.62694371016472e-06,
+      "loss": 0.8144,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4551864310463187,
+      "learning_rate": 9.62645111208982e-06,
+      "loss": 0.8049,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.500558989547229,
+      "learning_rate": 9.625958201626365e-06,
+      "loss": 0.8802,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4847802025673165,
+      "learning_rate": 9.625464978807633e-06,
+      "loss": 0.8953,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6388772642875407,
+      "learning_rate": 9.62497144366693e-06,
+      "loss": 0.8518,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.581868826155573,
+      "learning_rate": 9.624477596237577e-06,
+      "loss": 0.8313,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5557398423833004,
+      "learning_rate": 9.623983436552924e-06,
+      "loss": 0.762,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.606046016753044,
+      "learning_rate": 9.623488964646334e-06,
+      "loss": 0.8427,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.449851429012584,
+      "learning_rate": 9.6229941805512e-06,
+      "loss": 0.8233,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4510607379891773,
+      "learning_rate": 9.622499084300924e-06,
+      "loss": 0.899,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4704394143513242,
+      "learning_rate": 9.622003675928943e-06,
+      "loss": 0.7686,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.481190483383334,
+      "learning_rate": 9.621507955468704e-06,
+      "loss": 0.8059,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5877354816884341,
+      "learning_rate": 9.621011922953681e-06,
+      "loss": 0.7473,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6622911969097016,
+      "learning_rate": 9.620515578417364e-06,
+      "loss": 0.8036,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5823729385497614,
+      "learning_rate": 9.620018921893272e-06,
+      "loss": 0.8257,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5125782339126712,
+      "learning_rate": 9.619521953414936e-06,
+      "loss": 0.8121,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7126401164519731,
+      "learning_rate": 9.619024673015916e-06,
+      "loss": 0.863,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8957481621484896,
+      "learning_rate": 9.618527080729789e-06,
+      "loss": 1.1752,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4585206774743154,
+      "learning_rate": 9.618029176590152e-06,
+      "loss": 0.8013,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.469168193796979,
+      "learning_rate": 9.617530960630624e-06,
+      "loss": 0.7789,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5289400520077705,
+      "learning_rate": 9.617032432884847e-06,
+      "loss": 0.8235,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5185767951338127,
+      "learning_rate": 9.616533593386484e-06,
+      "loss": 0.7904,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.444557590913687,
+      "learning_rate": 9.616034442169214e-06,
+      "loss": 0.8816,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9029375239582471,
+      "learning_rate": 9.615534979266745e-06,
+      "loss": 1.1914,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8304838852731932,
+      "learning_rate": 9.6150352047128e-06,
+      "loss": 1.1768,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5806368107679758,
+      "learning_rate": 9.614535118541126e-06,
+      "loss": 0.9242,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4347680283690143,
+      "learning_rate": 9.614034720785488e-06,
+      "loss": 0.7653,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8296597903754166,
+      "learning_rate": 9.613534011479675e-06,
+      "loss": 0.7371,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4590699707067372,
+      "learning_rate": 9.613032990657495e-06,
+      "loss": 0.8686,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8609330094284481,
+      "learning_rate": 9.612531658352782e-06,
+      "loss": 1.137,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5709311651799216,
+      "learning_rate": 9.612030014599381e-06,
+      "loss": 0.817,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8627446563438778,
+      "learning_rate": 9.61152805943117e-06,
+      "loss": 1.154,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.808761666850305,
+      "learning_rate": 9.611025792882038e-06,
+      "loss": 1.1731,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6537067174113689,
+      "learning_rate": 9.6105232149859e-06,
+      "loss": 0.8358,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8006899805807882,
+      "learning_rate": 9.610020325776694e-06,
+      "loss": 1.1253,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5995874645563164,
+      "learning_rate": 9.609517125288373e-06,
+      "loss": 0.8371,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.64395169708802,
+      "learning_rate": 9.609013613554917e-06,
+      "loss": 0.7476,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3823231042324897,
+      "learning_rate": 9.608509790610322e-06,
+      "loss": 0.7927,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4188303104669986,
+      "learning_rate": 9.608005656488605e-06,
+      "loss": 0.7651,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8875494826376961,
+      "learning_rate": 9.607501211223812e-06,
+      "loss": 1.1497,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6343710275658958,
+      "learning_rate": 9.606996454850002e-06,
+      "loss": 0.8216,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8411611534455249,
+      "learning_rate": 9.606491387401256e-06,
+      "loss": 1.1532,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4592087109381038,
+      "learning_rate": 9.605986008911677e-06,
+      "loss": 0.8018,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4271140318066142,
+      "learning_rate": 9.605480319415391e-06,
+      "loss": 0.8076,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4031341931286325,
+      "learning_rate": 9.604974318946544e-06,
+      "loss": 0.7655,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5157193237988733,
+      "learning_rate": 9.6044680075393e-06,
+      "loss": 0.8691,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6025534141134328,
+      "learning_rate": 9.603961385227848e-06,
+      "loss": 0.7923,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4126408589981612,
+      "learning_rate": 9.603454452046395e-06,
+      "loss": 0.8057,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5140043897353115,
+      "learning_rate": 9.602947208029172e-06,
+      "loss": 0.7787,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4458827783841817,
+      "learning_rate": 9.602439653210426e-06,
+      "loss": 0.8325,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5065544985895871,
+      "learning_rate": 9.601931787624432e-06,
+      "loss": 0.9228,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5500625454414283,
+      "learning_rate": 9.601423611305481e-06,
+      "loss": 0.865,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4859931888949134,
+      "learning_rate": 9.600915124287886e-06,
+      "loss": 0.8372,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0559445035420005,
+      "learning_rate": 9.600406326605983e-06,
+      "loss": 1.2052,
+      "step": 1923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4571971603518485,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.9132,
+      "step": 1924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5753094972092905,
+      "learning_rate": 9.599387799386684e-06,
+      "loss": 0.8819,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5823569573061436,
+      "learning_rate": 9.598878069918064e-06,
+      "loss": 0.8316,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8271540781876586,
+      "learning_rate": 9.598368029922681e-06,
+      "loss": 1.134,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6059994765595487,
+      "learning_rate": 9.597857679434974e-06,
+      "loss": 0.9165,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5831001769555655,
+      "learning_rate": 9.597347018489403e-06,
+      "loss": 0.8513,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4450721335626313,
+      "learning_rate": 9.596836047120449e-06,
+      "loss": 0.808,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8980032014616626,
+      "learning_rate": 9.596324765362614e-06,
+      "loss": 1.154,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.539293914150005,
+      "learning_rate": 9.59581317325042e-06,
+      "loss": 0.8221,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.460131114780644,
+      "learning_rate": 9.595301270818411e-06,
+      "loss": 0.7284,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5304059527857843,
+      "learning_rate": 9.594789058101154e-06,
+      "loss": 0.8338,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4860589328315805,
+      "learning_rate": 9.594276535133232e-06,
+      "loss": 0.825,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8856197892352309,
+      "learning_rate": 9.593763701949253e-06,
+      "loss": 1.2051,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5130406344868286,
+      "learning_rate": 9.593250558583846e-06,
+      "loss": 0.8339,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4716043320582513,
+      "learning_rate": 9.592737105071657e-06,
+      "loss": 0.8468,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.578091370244503,
+      "learning_rate": 9.592223341447357e-06,
+      "loss": 0.8179,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.619685237875716,
+      "learning_rate": 9.591709267745635e-06,
+      "loss": 0.8001,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.46787499085914,
+      "learning_rate": 9.591194884001205e-06,
+      "loss": 0.8854,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4478472053033615,
+      "learning_rate": 9.590680190248797e-06,
+      "loss": 0.8025,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8779506792836145,
+      "learning_rate": 9.590165186523166e-06,
+      "loss": 1.159,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6402958016195528,
+      "learning_rate": 9.589649872859086e-06,
+      "loss": 0.8218,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6222334347202898,
+      "learning_rate": 9.589134249291352e-06,
+      "loss": 0.879,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5688963707710817,
+      "learning_rate": 9.588618315854779e-06,
+      "loss": 0.7228,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.444387335333108,
+      "learning_rate": 9.588102072584204e-06,
+      "loss": 0.7721,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5416620701577475,
+      "learning_rate": 9.587585519514487e-06,
+      "loss": 0.8052,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.535094422655125,
+      "learning_rate": 9.587068656680506e-06,
+      "loss": 0.8476,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5486971614978873,
+      "learning_rate": 9.58655148411716e-06,
+      "loss": 0.8101,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9407853303983144,
+      "learning_rate": 9.586034001859368e-06,
+      "loss": 1.1375,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8473184264588529,
+      "learning_rate": 9.585516209942077e-06,
+      "loss": 1.1524,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4607168546118343,
+      "learning_rate": 9.584998108400243e-06,
+      "loss": 0.7828,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.451335229150967,
+      "learning_rate": 9.584479697268854e-06,
+      "loss": 0.7445,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6259967175486456,
+      "learning_rate": 9.583960976582914e-06,
+      "loss": 0.8593,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4611492870796858,
+      "learning_rate": 9.583441946377445e-06,
+      "loss": 0.827,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5320261995963909,
+      "learning_rate": 9.582922606687495e-06,
+      "loss": 0.7932,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6663199891705658,
+      "learning_rate": 9.582402957548132e-06,
+      "loss": 0.8263,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5106437468927638,
+      "learning_rate": 9.581882998994442e-06,
+      "loss": 0.8054,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4488080696130805,
+      "learning_rate": 9.581362731061537e-06,
+      "loss": 0.864,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3518759642735945,
+      "learning_rate": 9.580842153784542e-06,
+      "loss": 1.1979,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6557396991620612,
+      "learning_rate": 9.580321267198611e-06,
+      "loss": 0.7622,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5576491340894745,
+      "learning_rate": 9.579800071338915e-06,
+      "loss": 0.8304,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6421353511099428,
+      "learning_rate": 9.579278566240646e-06,
+      "loss": 0.865,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5950115560258147,
+      "learning_rate": 9.578756751939017e-06,
+      "loss": 0.8593,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4879709247103965,
+      "learning_rate": 9.57823462846926e-06,
+      "loss": 0.7878,
+      "step": 1966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4132387057288935,
+      "learning_rate": 9.577712195866634e-06,
+      "loss": 0.7245,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8303345242631137,
+      "learning_rate": 9.577189454166414e-06,
+      "loss": 0.8199,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.504998426988508,
+      "learning_rate": 9.576666403403894e-06,
+      "loss": 0.7702,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4171771564609978,
+      "learning_rate": 9.576143043614393e-06,
+      "loss": 0.8041,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.018158099347651,
+      "learning_rate": 9.57561937483325e-06,
+      "loss": 1.1838,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6354530380802565,
+      "learning_rate": 9.575095397095824e-06,
+      "loss": 0.7886,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4081521493242066,
+      "learning_rate": 9.574571110437496e-06,
+      "loss": 0.8637,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4592120599050626,
+      "learning_rate": 9.574046514893667e-06,
+      "loss": 0.8088,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5242720845036777,
+      "learning_rate": 9.573521610499756e-06,
+      "loss": 0.876,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8648406607296378,
+      "learning_rate": 9.572996397291209e-06,
+      "loss": 1.1459,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8023884409772494,
+      "learning_rate": 9.572470875303488e-06,
+      "loss": 1.1524,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6096258897212612,
+      "learning_rate": 9.571945044572079e-06,
+      "loss": 0.8636,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.524697864588973,
+      "learning_rate": 9.571418905132486e-06,
+      "loss": 0.8322,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6275365199180214,
+      "learning_rate": 9.570892457020233e-06,
+      "loss": 0.7787,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5489770770693743,
+      "learning_rate": 9.570365700270872e-06,
+      "loss": 0.688,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8687087857902865,
+      "learning_rate": 9.569838634919968e-06,
+      "loss": 1.1414,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3907902823670681,
+      "learning_rate": 9.569311261003108e-06,
+      "loss": 0.7978,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4801077999409677,
+      "learning_rate": 9.568783578555904e-06,
+      "loss": 0.7986,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8247994134928776,
+      "learning_rate": 9.568255587613986e-06,
+      "loss": 0.9336,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5290343716768693,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.7561,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8962036807717707,
+      "learning_rate": 9.567198680388632e-06,
+      "loss": 1.1802,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.554517624306963,
+      "learning_rate": 9.566669764176562e-06,
+      "loss": 0.8261,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4407902648746544,
+      "learning_rate": 9.566140539612506e-06,
+      "loss": 0.8345,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.478552519173747,
+      "learning_rate": 9.565611006732201e-06,
+      "loss": 0.8206,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8269263857066178,
+      "learning_rate": 9.5650811655714e-06,
+      "loss": 1.1757,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.584076308274061,
+      "learning_rate": 9.564551016165879e-06,
+      "loss": 0.8713,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6177442423537147,
+      "learning_rate": 9.564020558551437e-06,
+      "loss": 0.8274,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.564769845079321,
+      "learning_rate": 9.56348979276389e-06,
+      "loss": 0.7795,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4469394368930757,
+      "learning_rate": 9.562958718839078e-06,
+      "loss": 0.7564,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.41775881739794,
+      "learning_rate": 9.562427336812859e-06,
+      "loss": 0.8132,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8340506300875027,
+      "learning_rate": 9.561895646721113e-06,
+      "loss": 1.1485,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4314700834740577,
+      "learning_rate": 9.561363648599742e-06,
+      "loss": 0.8653,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5556179449093055,
+      "learning_rate": 9.560831342484668e-06,
+      "loss": 0.906,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.695931216928343,
+      "learning_rate": 9.560298728411833e-06,
+      "loss": 0.7533,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5548177138660582,
+      "learning_rate": 9.559765806417198e-06,
+      "loss": 0.9132,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4834800453069483,
+      "learning_rate": 9.55923257653675e-06,
+      "loss": 0.8037,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.842534846303849,
+      "learning_rate": 9.558699038806494e-06,
+      "loss": 1.1661,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8147928577499463,
+      "learning_rate": 9.558165193262455e-06,
+      "loss": 1.1491,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4651154921627292,
+      "learning_rate": 9.557631039940678e-06,
+      "loss": 0.7976,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4775401430952735,
+      "learning_rate": 9.557096578877232e-06,
+      "loss": 0.8029,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.442471322137238,
+      "learning_rate": 9.556561810108205e-06,
+      "loss": 0.8657,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.440087798122106,
+      "learning_rate": 9.556026733669706e-06,
+      "loss": 0.7912,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4522236870456824,
+      "learning_rate": 9.555491349597862e-06,
+      "loss": 0.7075,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4086375758751186,
+      "learning_rate": 9.554955657928828e-06,
+      "loss": 0.8965,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5526463637832908,
+      "learning_rate": 9.55441965869877e-06,
+      "loss": 0.8198,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5405722404386109,
+      "learning_rate": 9.553883351943882e-06,
+      "loss": 0.8009,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4334982610963374,
+      "learning_rate": 9.55334673770038e-06,
+      "loss": 0.8151,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.597102999347615,
+      "learning_rate": 9.552809816004491e-06,
+      "loss": 0.8327,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5404945020064367,
+      "learning_rate": 9.552272586892475e-06,
+      "loss": 0.7972,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4999958359113896,
+      "learning_rate": 9.551735050400603e-06,
+      "loss": 0.8994,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6005243026581775,
+      "learning_rate": 9.551197206565174e-06,
+      "loss": 0.8585,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1101390736515815,
+      "learning_rate": 9.550659055422502e-06,
+      "loss": 1.1717,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5256614815794431,
+      "learning_rate": 9.550120597008925e-06,
+      "loss": 0.8467,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5635654397022658,
+      "learning_rate": 9.549581831360799e-06,
+      "loss": 0.8884,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5124055114417434,
+      "learning_rate": 9.549042758514505e-06,
+      "loss": 0.7685,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.515363090117947,
+      "learning_rate": 9.548503378506444e-06,
+      "loss": 0.8301,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6609574982501256,
+      "learning_rate": 9.547963691373033e-06,
+      "loss": 0.7951,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5170720984309314,
+      "learning_rate": 9.547423697150714e-06,
+      "loss": 0.914,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5850238289351597,
+      "learning_rate": 9.546883395875947e-06,
+      "loss": 0.8434,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4772370305066855,
+      "learning_rate": 9.54634278758522e-06,
+      "loss": 0.8794,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5916768671297898,
+      "learning_rate": 9.545801872315028e-06,
+      "loss": 0.8886,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5298054603189897,
+      "learning_rate": 9.545260650101902e-06,
+      "loss": 0.7431,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4609699030287253,
+      "learning_rate": 9.544719120982382e-06,
+      "loss": 0.7705,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6164060796216215,
+      "learning_rate": 9.544177284993035e-06,
+      "loss": 0.8011,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5087461083376432,
+      "learning_rate": 9.543635142170447e-06,
+      "loss": 0.8729,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6154532817564018,
+      "learning_rate": 9.543092692551224e-06,
+      "loss": 0.8245,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1211807465923624,
+      "learning_rate": 9.542549936171994e-06,
+      "loss": 0.8555,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7826255612234376,
+      "learning_rate": 9.542006873069404e-06,
+      "loss": 0.8081,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6566557841116214,
+      "learning_rate": 9.541463503280127e-06,
+      "loss": 0.8851,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6309681368643612,
+      "learning_rate": 9.540919826840848e-06,
+      "loss": 0.8091,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5875453047972214,
+      "learning_rate": 9.540375843788278e-06,
+      "loss": 0.7335,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5200056144487872,
+      "learning_rate": 9.539831554159152e-06,
+      "loss": 0.8098,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5398000046571634,
+      "learning_rate": 9.539286957990215e-06,
+      "loss": 0.921,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5068932575198775,
+      "learning_rate": 9.538742055318243e-06,
+      "loss": 0.8468,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4437664856269383,
+      "learning_rate": 9.538196846180033e-06,
+      "loss": 0.8117,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.507050775408361,
+      "learning_rate": 9.53765133061239e-06,
+      "loss": 0.8139,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4289315139188263,
+      "learning_rate": 9.537105508652156e-06,
+      "loss": 0.8306,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4556253493228337,
+      "learning_rate": 9.536559380336183e-06,
+      "loss": 0.8621,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.410097251607193,
+      "learning_rate": 9.536012945701345e-06,
+      "loss": 0.7413,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1795139289825596,
+      "learning_rate": 9.535466204784542e-06,
+      "loss": 1.1694,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5666830304945485,
+      "learning_rate": 9.53491915762269e-06,
+      "loss": 0.7922,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8816786053977199,
+      "learning_rate": 9.534371804252727e-06,
+      "loss": 1.1548,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6087610094690001,
+      "learning_rate": 9.533824144711612e-06,
+      "loss": 0.8347,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5539672774751516,
+      "learning_rate": 9.533276179036324e-06,
+      "loss": 0.7507,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5538271987534389,
+      "learning_rate": 9.532727907263861e-06,
+      "loss": 0.8741,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4663255725235147,
+      "learning_rate": 9.532179329431243e-06,
+      "loss": 0.7856,
+      "step": 2052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5870658117724739,
+      "learning_rate": 9.531630445575516e-06,
+      "loss": 0.803,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.538162675417869,
+      "learning_rate": 9.53108125573374e-06,
+      "loss": 0.7601,
+      "step": 2054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4335290445611213,
+      "learning_rate": 9.530531759942994e-06,
+      "loss": 0.8054,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.567707543091647,
+      "learning_rate": 9.529981958240386e-06,
+      "loss": 0.8285,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5614285894465612,
+      "learning_rate": 9.529431850663036e-06,
+      "loss": 0.8436,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5078563648487966,
+      "learning_rate": 9.528881437248092e-06,
+      "loss": 0.7465,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.548778924405887,
+      "learning_rate": 9.528330718032716e-06,
+      "loss": 0.8465,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.569953279471227,
+      "learning_rate": 9.527779693054095e-06,
+      "loss": 0.7959,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6111020901195088,
+      "learning_rate": 9.527228362349437e-06,
+      "loss": 0.8646,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5756714611813065,
+      "learning_rate": 9.526676725955968e-06,
+      "loss": 0.8022,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5240010988885164,
+      "learning_rate": 9.526124783910935e-06,
+      "loss": 0.8453,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4088776288386655,
+      "learning_rate": 9.525572536251608e-06,
+      "loss": 0.7902,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4395932589334097,
+      "learning_rate": 9.525019983015274e-06,
+      "loss": 0.8042,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4844657044914202,
+      "learning_rate": 9.524467124239243e-06,
+      "loss": 0.8279,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6001718981232678,
+      "learning_rate": 9.523913959960846e-06,
+      "loss": 0.8858,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.459166915834659,
+      "learning_rate": 9.523360490217435e-06,
+      "loss": 0.8044,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5885640927619404,
+      "learning_rate": 9.52280671504638e-06,
+      "loss": 0.8866,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.498360637011439,
+      "learning_rate": 9.522252634485071e-06,
+      "loss": 0.8446,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6628163404921328,
+      "learning_rate": 9.521698248570928e-06,
+      "loss": 0.7474,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.383755394028347,
+      "learning_rate": 9.521143557341378e-06,
+      "loss": 0.7574,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5464371876326106,
+      "learning_rate": 9.520588560833876e-06,
+      "loss": 0.7907,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4848428431731366,
+      "learning_rate": 9.520033259085897e-06,
+      "loss": 0.752,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4959677903249786,
+      "learning_rate": 9.519477652134938e-06,
+      "loss": 0.7615,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.54307098278177,
+      "learning_rate": 9.518921740018512e-06,
+      "loss": 0.9456,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4165815111712976,
+      "learning_rate": 9.518365522774157e-06,
+      "loss": 1.1971,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.298176375964562,
+      "learning_rate": 9.517809000439432e-06,
+      "loss": 1.1785,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5319971429975636,
+      "learning_rate": 9.517252173051912e-06,
+      "loss": 0.9188,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5799540359746305,
+      "learning_rate": 9.516695040649195e-06,
+      "loss": 0.7559,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5601980254238543,
+      "learning_rate": 9.516137603268903e-06,
+      "loss": 0.8696,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.558500343844507,
+      "learning_rate": 9.515579860948672e-06,
+      "loss": 0.8403,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.531770581325229,
+      "learning_rate": 9.515021813726162e-06,
+      "loss": 0.8491,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4204828569392018,
+      "learning_rate": 9.514463461639055e-06,
+      "loss": 0.7953,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6420364718889713,
+      "learning_rate": 9.513904804725054e-06,
+      "loss": 0.832,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5693719784151614,
+      "learning_rate": 9.513345843021878e-06,
+      "loss": 0.8973,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5859173024931585,
+      "learning_rate": 9.51278657656727e-06,
+      "loss": 0.7858,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.200025805776269,
+      "learning_rate": 9.512227005398992e-06,
+      "loss": 1.1956,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4535402868958767,
+      "learning_rate": 9.511667129554832e-06,
+      "loss": 0.8228,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.484412985829688,
+      "learning_rate": 9.511106949072588e-06,
+      "loss": 0.8294,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5227825115611187,
+      "learning_rate": 9.510546463990089e-06,
+      "loss": 0.9095,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4716656227651383,
+      "learning_rate": 9.509985674345179e-06,
+      "loss": 0.7868,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.458081447545481,
+      "learning_rate": 9.509424580175724e-06,
+      "loss": 0.8232,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1001736302014453,
+      "learning_rate": 9.508863181519608e-06,
+      "loss": 1.1835,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.561968626215087,
+      "learning_rate": 9.50830147841474e-06,
+      "loss": 0.7872,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5122608393218788,
+      "learning_rate": 9.507739470899048e-06,
+      "loss": 0.8592,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6332424973760133,
+      "learning_rate": 9.50717715901048e-06,
+      "loss": 0.8814,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5522742721355947,
+      "learning_rate": 9.506614542787003e-06,
+      "loss": 0.8092,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5900342227438522,
+      "learning_rate": 9.506051622266608e-06,
+      "loss": 0.8388,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5877402561046388,
+      "learning_rate": 9.505488397487303e-06,
+      "loss": 0.8431,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1499759375621594,
+      "learning_rate": 9.504924868487118e-06,
+      "loss": 1.1687,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.546693746661753,
+      "learning_rate": 9.504361035304106e-06,
+      "loss": 0.8611,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0350998333260615,
+      "learning_rate": 9.503796897976339e-06,
+      "loss": 1.1279,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.878201323343734,
+      "learning_rate": 9.503232456541904e-06,
+      "loss": 1.1592,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5677990365056014,
+      "learning_rate": 9.502667711038917e-06,
+      "loss": 0.7505,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.654728209652386,
+      "learning_rate": 9.50210266150551e-06,
+      "loss": 0.8595,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5488854348199774,
+      "learning_rate": 9.501537307979836e-06,
+      "loss": 0.868,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.180044400332189,
+      "learning_rate": 9.500971650500072e-06,
+      "loss": 1.1492,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5940741885667802,
+      "learning_rate": 9.500405689104408e-06,
+      "loss": 0.8597,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5110415421957293,
+      "learning_rate": 9.499839423831062e-06,
+      "loss": 0.8315,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5161124359375824,
+      "learning_rate": 9.499272854718268e-06,
+      "loss": 0.793,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0760527603217243,
+      "learning_rate": 9.498705981804283e-06,
+      "loss": 1.132,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5528258855269335,
+      "learning_rate": 9.498138805127383e-06,
+      "loss": 0.8544,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5572580710714743,
+      "learning_rate": 9.497571324725865e-06,
+      "loss": 0.814,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4474290273521566,
+      "learning_rate": 9.497003540638047e-06,
+      "loss": 0.8112,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5296451017598425,
+      "learning_rate": 9.496435452902268e-06,
+      "loss": 0.8103,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4298819709449386,
+      "learning_rate": 9.495867061556884e-06,
+      "loss": 0.8095,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4829398011880368,
+      "learning_rate": 9.495298366640276e-06,
+      "loss": 0.8085,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5206479717954375,
+      "learning_rate": 9.494729368190843e-06,
+      "loss": 0.7782,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.873404289918305,
+      "learning_rate": 9.494160066247006e-06,
+      "loss": 0.7789,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0268484387428616,
+      "learning_rate": 9.493590460847204e-06,
+      "loss": 1.162,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4513800902525658,
+      "learning_rate": 9.4930205520299e-06,
+      "loss": 0.744,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5568742926042145,
+      "learning_rate": 9.492450339833573e-06,
+      "loss": 0.7672,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8566521929923023,
+      "learning_rate": 9.491879824296729e-06,
+      "loss": 1.1548,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5281890094016113,
+      "learning_rate": 9.491309005457885e-06,
+      "loss": 0.8647,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4806339144556646,
+      "learning_rate": 9.490737883355587e-06,
+      "loss": 0.7954,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5837364064120136,
+      "learning_rate": 9.4901664580284e-06,
+      "loss": 0.8453,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5867057799872515,
+      "learning_rate": 9.489594729514907e-06,
+      "loss": 0.8119,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5329971605249757,
+      "learning_rate": 9.48902269785371e-06,
+      "loss": 0.7895,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.631845253540148,
+      "learning_rate": 9.488450363083435e-06,
+      "loss": 0.8152,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4279116263224285,
+      "learning_rate": 9.48787772524273e-06,
+      "loss": 0.8523,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4580836107684407,
+      "learning_rate": 9.487304784370257e-06,
+      "loss": 0.7321,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5188237211342017,
+      "learning_rate": 9.486731540504705e-06,
+      "loss": 0.759,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.676518276728722,
+      "learning_rate": 9.48615799368478e-06,
+      "loss": 0.8169,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5291736490255858,
+      "learning_rate": 9.48558414394921e-06,
+      "loss": 0.8203,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.670853457846532,
+      "learning_rate": 9.48500999133674e-06,
+      "loss": 0.8245,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4306486237015201,
+      "learning_rate": 9.484435535886142e-06,
+      "loss": 0.7851,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.413913750685564,
+      "learning_rate": 9.4838607776362e-06,
+      "loss": 0.682,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3887091694760194,
+      "learning_rate": 9.483285716625727e-06,
+      "loss": 0.7608,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2378729349356958,
+      "learning_rate": 9.482710352893549e-06,
+      "loss": 1.1491,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.329785060480567,
+      "learning_rate": 9.48213468647852e-06,
+      "loss": 0.8265,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5982615918880068,
+      "learning_rate": 9.481558717419506e-06,
+      "loss": 0.8724,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.481082657944046,
+      "learning_rate": 9.4809824457554e-06,
+      "loss": 0.8105,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.488119328039974,
+      "learning_rate": 9.480405871525114e-06,
+      "loss": 0.7929,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5183815142154768,
+      "learning_rate": 9.479828994767577e-06,
+      "loss": 0.8464,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9162720706906529,
+      "learning_rate": 9.479251815521745e-06,
+      "loss": 1.1637,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.849114843188252,
+      "learning_rate": 9.478674333826586e-06,
+      "loss": 1.1692,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3414491824864267,
+      "learning_rate": 9.478096549721094e-06,
+      "loss": 0.6869,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4093935836935414,
+      "learning_rate": 9.477518463244284e-06,
+      "loss": 0.773,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4876030850139856,
+      "learning_rate": 9.476940074435189e-06,
+      "loss": 0.7885,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.616464725075642,
+      "learning_rate": 9.476361383332864e-06,
+      "loss": 0.8187,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6001166081300626,
+      "learning_rate": 9.475782389976382e-06,
+      "loss": 0.6992,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8570326225769496,
+      "learning_rate": 9.475203094404836e-06,
+      "loss": 0.7656,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4808079056633383,
+      "learning_rate": 9.474623496657347e-06,
+      "loss": 0.8221,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.061787492683681,
+      "learning_rate": 9.474043596773048e-06,
+      "loss": 1.1292,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6177951181536774,
+      "learning_rate": 9.473463394791093e-06,
+      "loss": 0.8721,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5480323673435714,
+      "learning_rate": 9.472882890750662e-06,
+      "loss": 0.8089,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.437392033160396,
+      "learning_rate": 9.472302084690948e-06,
+      "loss": 0.826,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5559288800446636,
+      "learning_rate": 9.471720976651173e-06,
+      "loss": 0.7628,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.521667628788659,
+      "learning_rate": 9.471139566670571e-06,
+      "loss": 0.7587,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5129018389789217,
+      "learning_rate": 9.470557854788402e-06,
+      "loss": 0.8344,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4934008841497588,
+      "learning_rate": 9.469975841043946e-06,
+      "loss": 0.71,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5490401054743221,
+      "learning_rate": 9.469393525476498e-06,
+      "loss": 0.837,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5786207173715618,
+      "learning_rate": 9.468810908125379e-06,
+      "loss": 0.8031,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4702982481981928,
+      "learning_rate": 9.468227989029929e-06,
+      "loss": 0.7858,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5124586106922184,
+      "learning_rate": 9.467644768229509e-06,
+      "loss": 0.8133,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9101090939486413,
+      "learning_rate": 9.467061245763499e-06,
+      "loss": 1.1778,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5839117102355542,
+      "learning_rate": 9.466477421671296e-06,
+      "loss": 0.8423,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5303065091084644,
+      "learning_rate": 9.465893295992326e-06,
+      "loss": 0.8137,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4725048707969537,
+      "learning_rate": 9.46530886876603e-06,
+      "loss": 0.7229,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4621103243510853,
+      "learning_rate": 9.464724140031866e-06,
+      "loss": 0.7663,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.475005015398539,
+      "learning_rate": 9.46413910982932e-06,
+      "loss": 0.8533,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3699921744937011,
+      "learning_rate": 9.463553778197897e-06,
+      "loss": 0.8637,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8282297120994768,
+      "learning_rate": 9.462968145177112e-06,
+      "loss": 0.7457,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.518068126232067,
+      "learning_rate": 9.462382210806514e-06,
+      "loss": 0.8473,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3962732002835698,
+      "learning_rate": 9.461795975125665e-06,
+      "loss": 0.7941,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5919861851279307,
+      "learning_rate": 9.461209438174148e-06,
+      "loss": 0.8638,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9683919122304938,
+      "learning_rate": 9.46062259999157e-06,
+      "loss": 1.1776,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.613195879580568,
+      "learning_rate": 9.460035460617555e-06,
+      "loss": 0.8854,
+      "step": 2179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8331130749215876,
+      "learning_rate": 9.459448020091746e-06,
+      "loss": 1.157,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4717283642964227,
+      "learning_rate": 9.45886027845381e-06,
+      "loss": 0.7812,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5023710772068761,
+      "learning_rate": 9.458272235743434e-06,
+      "loss": 0.7815,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4486567946514675,
+      "learning_rate": 9.457683892000318e-06,
+      "loss": 0.8181,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6591626242786128,
+      "learning_rate": 9.457095247264197e-06,
+      "loss": 0.9463,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.045601299762233,
+      "learning_rate": 9.45650630157481e-06,
+      "loss": 1.1608,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5799418955524047,
+      "learning_rate": 9.455917054971929e-06,
+      "loss": 0.8853,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6725159178427151,
+      "learning_rate": 9.455327507495338e-06,
+      "loss": 0.7993,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6892802921167822,
+      "learning_rate": 9.454737659184845e-06,
+      "loss": 0.7906,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5480470618414361,
+      "learning_rate": 9.45414751008028e-06,
+      "loss": 0.855,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5052535639480646,
+      "learning_rate": 9.45355706022149e-06,
+      "loss": 0.7726,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6103094544191743,
+      "learning_rate": 9.452966309648347e-06,
+      "loss": 0.8683,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0160254135878146,
+      "learning_rate": 9.452375258400732e-06,
+      "loss": 1.1234,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5585288104324018,
+      "learning_rate": 9.451783906518558e-06,
+      "loss": 0.7273,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4594836811495433,
+      "learning_rate": 9.451192254041759e-06,
+      "loss": 0.7998,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4877333153410297,
+      "learning_rate": 9.450600301010279e-06,
+      "loss": 0.8391,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5242492907421654,
+      "learning_rate": 9.45000804746409e-06,
+      "loss": 0.8078,
+      "step": 2196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.514317118877034,
+      "learning_rate": 9.449415493443181e-06,
+      "loss": 0.8065,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8835591538195738,
+      "learning_rate": 9.448822638987564e-06,
+      "loss": 1.1597,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3962471594885688,
+      "learning_rate": 9.44822948413727e-06,
+      "loss": 0.8879,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.489889452005428,
+      "learning_rate": 9.44763602893235e-06,
+      "loss": 0.8837,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4156134482077132,
+      "learning_rate": 9.447042273412873e-06,
+      "loss": 0.7768,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.462114683557008,
+      "learning_rate": 9.446448217618935e-06,
+      "loss": 0.8187,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.693005438792916,
+      "learning_rate": 9.445853861590647e-06,
+      "loss": 0.8404,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5564941247757622,
+      "learning_rate": 9.445259205368138e-06,
+      "loss": 0.7549,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8587664991133772,
+      "learning_rate": 9.444664248991563e-06,
+      "loss": 1.1196,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.523464610621261,
+      "learning_rate": 9.444068992501097e-06,
+      "loss": 0.7609,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8041221973951086,
+      "learning_rate": 9.44347343593693e-06,
+      "loss": 1.148,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5137151896952397,
+      "learning_rate": 9.442877579339273e-06,
+      "loss": 0.771,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5062186662026789,
+      "learning_rate": 9.442281422748365e-06,
+      "loss": 0.8607,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5244668676406488,
+      "learning_rate": 9.441684966204456e-06,
+      "loss": 0.7913,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4977607566378373,
+      "learning_rate": 9.441088209747823e-06,
+      "loss": 0.8155,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.521428901776055,
+      "learning_rate": 9.440491153418759e-06,
+      "loss": 0.8538,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.589509863167939,
+      "learning_rate": 9.439893797257578e-06,
+      "loss": 0.8432,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4668184795811325,
+      "learning_rate": 9.439296141304615e-06,
+      "loss": 0.7854,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.532414283691502,
+      "learning_rate": 9.438698185600226e-06,
+      "loss": 0.8819,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5229161899686,
+      "learning_rate": 9.438099930184783e-06,
+      "loss": 0.9098,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5428311306161095,
+      "learning_rate": 9.437501375098688e-06,
+      "loss": 0.7601,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5097778559357995,
+      "learning_rate": 9.436902520382352e-06,
+      "loss": 0.8351,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5314460971829593,
+      "learning_rate": 9.436303366076213e-06,
+      "loss": 0.8206,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5923034782512633,
+      "learning_rate": 9.435703912220727e-06,
+      "loss": 0.9379,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.531339849717247,
+      "learning_rate": 9.435104158856367e-06,
+      "loss": 0.8075,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5220375979160192,
+      "learning_rate": 9.434504106023634e-06,
+      "loss": 0.8673,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6157550333426582,
+      "learning_rate": 9.433903753763045e-06,
+      "loss": 0.7937,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.528030876968063,
+      "learning_rate": 9.433303102115136e-06,
+      "loss": 0.7806,
+      "step": 2224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5008505563884982,
+      "learning_rate": 9.432702151120464e-06,
+      "loss": 0.8813,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5978982993623998,
+      "learning_rate": 9.432100900819604e-06,
+      "loss": 0.8872,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4225618312662536,
+      "learning_rate": 9.431499351253159e-06,
+      "loss": 0.8822,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5752846964676948,
+      "learning_rate": 9.430897502461745e-06,
+      "loss": 0.8144,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9712858789887392,
+      "learning_rate": 9.430295354485999e-06,
+      "loss": 1.1575,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5510150067171729,
+      "learning_rate": 9.42969290736658e-06,
+      "loss": 0.8738,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4580628567400207,
+      "learning_rate": 9.429090161144166e-06,
+      "loss": 0.8383,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4257216096123133,
+      "learning_rate": 9.428487115859458e-06,
+      "loss": 0.7807,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.492338976508945,
+      "learning_rate": 9.427883771553172e-06,
+      "loss": 0.8373,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9378096728603565,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 1.1601,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.50116159122152,
+      "learning_rate": 9.42667618603885e-06,
+      "loss": 0.8255,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5522154223309075,
+      "learning_rate": 9.426071944912351e-06,
+      "loss": 0.8468,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8231893312153317,
+      "learning_rate": 9.425467404927356e-06,
+      "loss": 1.1747,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.496687113134294,
+      "learning_rate": 9.42486256612468e-06,
+      "loss": 0.8029,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4609103786613773,
+      "learning_rate": 9.424257428545166e-06,
+      "loss": 0.7729,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5120034068362511,
+      "learning_rate": 9.423651992229673e-06,
+      "loss": 0.8157,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9048209561668592,
+      "learning_rate": 9.423046257219083e-06,
+      "loss": 1.1542,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8710660185885328,
+      "learning_rate": 9.422440223554296e-06,
+      "loss": 1.1207,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5456547325887937,
+      "learning_rate": 9.421833891276233e-06,
+      "loss": 0.8568,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8183570069267132,
+      "learning_rate": 9.421227260425834e-06,
+      "loss": 1.1735,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4798376334199836,
+      "learning_rate": 9.42062033104406e-06,
+      "loss": 0.8513,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5676482686965014,
+      "learning_rate": 9.420013103171893e-06,
+      "loss": 0.8467,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6402454840549179,
+      "learning_rate": 9.419405576850334e-06,
+      "loss": 0.8313,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6494450064767665,
+      "learning_rate": 9.418797752120406e-06,
+      "loss": 0.7398,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.504794867567575,
+      "learning_rate": 9.418189629023149e-06,
+      "loss": 0.8776,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.504432324814278,
+      "learning_rate": 9.417581207599626e-06,
+      "loss": 0.7632,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4885288995658281,
+      "learning_rate": 9.416972487890918e-06,
+      "loss": 0.826,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6147520129410498,
+      "learning_rate": 9.416363469938128e-06,
+      "loss": 0.8256,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.054392646446437,
+      "learning_rate": 9.415754153782377e-06,
+      "loss": 1.1384,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.591742564460529,
+      "learning_rate": 9.415144539464809e-06,
+      "loss": 0.8422,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4047579033837214,
+      "learning_rate": 9.414534627026586e-06,
+      "loss": 0.8815,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7994289258791244,
+      "learning_rate": 9.413924416508891e-06,
+      "loss": 1.1246,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6376112406253958,
+      "learning_rate": 9.413313907952925e-06,
+      "loss": 0.8004,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7872891826227746,
+      "learning_rate": 9.412703101399912e-06,
+      "loss": 1.174,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4623619136243506,
+      "learning_rate": 9.412091996891097e-06,
+      "loss": 0.8262,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8332610993397026,
+      "learning_rate": 9.41148059446774e-06,
+      "loss": 1.1363,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7931089506990561,
+      "learning_rate": 9.410868894171126e-06,
+      "loss": 1.1443,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5058739310972282,
+      "learning_rate": 9.410256896042558e-06,
+      "loss": 0.8147,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5064991891967874,
+      "learning_rate": 9.409644600123362e-06,
+      "loss": 0.8847,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8471921859416185,
+      "learning_rate": 9.409032006454877e-06,
+      "loss": 0.8282,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5030017921694547,
+      "learning_rate": 9.40841911507847e-06,
+      "loss": 0.8305,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.513720679097505,
+      "learning_rate": 9.407805926035524e-06,
+      "loss": 0.8168,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5353803914990405,
+      "learning_rate": 9.407192439367443e-06,
+      "loss": 0.8694,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.609652548386237,
+      "learning_rate": 9.40657865511565e-06,
+      "loss": 0.8194,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3829395790122831,
+      "learning_rate": 9.40596457332159e-06,
+      "loss": 0.7911,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5197321795379435,
+      "learning_rate": 9.405350194026728e-06,
+      "loss": 0.9147,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5624644065103857,
+      "learning_rate": 9.404735517272547e-06,
+      "loss": 0.9,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.468519599234418,
+      "learning_rate": 9.404120543100553e-06,
+      "loss": 0.7932,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0451410344294818,
+      "learning_rate": 9.403505271552267e-06,
+      "loss": 1.159,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5336142217449362,
+      "learning_rate": 9.402889702669235e-06,
+      "loss": 0.9369,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5617754734150564,
+      "learning_rate": 9.402273836493026e-06,
+      "loss": 0.7928,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.651424888578387,
+      "learning_rate": 9.401657673065218e-06,
+      "loss": 0.7777,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4146233181646815,
+      "learning_rate": 9.401041212427422e-06,
+      "loss": 0.86,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5130278283894125,
+      "learning_rate": 9.400424454621258e-06,
+      "loss": 0.8358,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.589797501295723,
+      "learning_rate": 9.399807399688371e-06,
+      "loss": 0.9157,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5172480715960457,
+      "learning_rate": 9.39919004767043e-06,
+      "loss": 0.8279,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.436918572201185,
+      "learning_rate": 9.398572398609118e-06,
+      "loss": 0.8488,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9002967569532419,
+      "learning_rate": 9.397954452546139e-06,
+      "loss": 0.9042,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0579833345622311,
+      "learning_rate": 9.397336209523218e-06,
+      "loss": 1.1243,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.940718754472445,
+      "learning_rate": 9.396717669582102e-06,
+      "loss": 1.1522,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4047347674541666,
+      "learning_rate": 9.396098832764555e-06,
+      "loss": 0.8598,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.525875454032734,
+      "learning_rate": 9.395479699112363e-06,
+      "loss": 0.7459,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4609041845155042,
+      "learning_rate": 9.394860268667329e-06,
+      "loss": 0.7814,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6864513463658968,
+      "learning_rate": 9.394240541471282e-06,
+      "loss": 0.8137,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3718549209765878,
+      "learning_rate": 9.393620517566066e-06,
+      "loss": 0.7469,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4887363232867643,
+      "learning_rate": 9.393000196993544e-06,
+      "loss": 0.7297,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.493873494780687,
+      "learning_rate": 9.392379579795605e-06,
+      "loss": 1.1464,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6129748487171538,
+      "learning_rate": 9.391758666014152e-06,
+      "loss": 0.9141,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5571412809063239,
+      "learning_rate": 9.391137455691113e-06,
+      "loss": 0.8315,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4928072914762776,
+      "learning_rate": 9.39051594886843e-06,
+      "loss": 0.8963,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.505424809303369,
+      "learning_rate": 9.389894145588072e-06,
+      "loss": 0.8277,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4809810596085007,
+      "learning_rate": 9.389272045892023e-06,
+      "loss": 0.8328,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4832405102914936,
+      "learning_rate": 9.388649649822289e-06,
+      "loss": 0.8408,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.49164874286212,
+      "learning_rate": 9.388026957420895e-06,
+      "loss": 0.8173,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6237531092560462,
+      "learning_rate": 9.387403968729887e-06,
+      "loss": 0.7891,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5936813063153645,
+      "learning_rate": 9.386780683791331e-06,
+      "loss": 0.81,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.763657237633359,
+      "learning_rate": 9.386157102647312e-06,
+      "loss": 0.8504,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6730909591583305,
+      "learning_rate": 9.385533225339936e-06,
+      "loss": 0.8159,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6353347214690444,
+      "learning_rate": 9.384909051911329e-06,
+      "loss": 0.914,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4862383479040204,
+      "learning_rate": 9.384284582403636e-06,
+      "loss": 0.7922,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4605269964234635,
+      "learning_rate": 9.38365981685902e-06,
+      "loss": 0.8063,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5238915804119315,
+      "learning_rate": 9.383034755319673e-06,
+      "loss": 0.7339,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.470695900606466,
+      "learning_rate": 9.382409397827794e-06,
+      "loss": 0.7251,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0386695209242822,
+      "learning_rate": 9.381783744425615e-06,
+      "loss": 1.1441,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7039258012960135,
+      "learning_rate": 9.381157795155374e-06,
+      "loss": 0.8552,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8229847245522032,
+      "learning_rate": 9.380531550059345e-06,
+      "loss": 1.1296,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8054768888921484,
+      "learning_rate": 9.379905009179804e-06,
+      "loss": 1.135,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4479348511614942,
+      "learning_rate": 9.379278172559065e-06,
+      "loss": 0.8206,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.604122916752245,
+      "learning_rate": 9.378651040239449e-06,
+      "loss": 0.7975,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5896132157466434,
+      "learning_rate": 9.378023612263302e-06,
+      "loss": 0.822,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5120229839977333,
+      "learning_rate": 9.37739588867299e-06,
+      "loss": 0.857,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.245497066603633,
+      "learning_rate": 9.376767869510899e-06,
+      "loss": 1.1649,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.688712130060152,
+      "learning_rate": 9.376139554819432e-06,
+      "loss": 0.7953,
+      "step": 2317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5435441696900718,
+      "learning_rate": 9.375510944641017e-06,
+      "loss": 0.8395,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5082908244345896,
+      "learning_rate": 9.374882039018096e-06,
+      "loss": 0.7597,
+      "step": 2319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5175740095678978,
+      "learning_rate": 9.374252837993137e-06,
+      "loss": 0.8859,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5428567372431767,
+      "learning_rate": 9.373623341608624e-06,
+      "loss": 0.974,
+      "step": 2321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5227664673621937,
+      "learning_rate": 9.372993549907063e-06,
+      "loss": 0.8223,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5509384902668868,
+      "learning_rate": 9.372363462930976e-06,
+      "loss": 0.7721,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4596386820978262,
+      "learning_rate": 9.371733080722911e-06,
+      "loss": 0.8031,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5212226482078324,
+      "learning_rate": 9.371102403325432e-06,
+      "loss": 0.8203,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6525396996155555,
+      "learning_rate": 9.370471430781123e-06,
+      "loss": 0.8624,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6535655101983044,
+      "learning_rate": 9.36984016313259e-06,
+      "loss": 0.8495,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6378609275448528,
+      "learning_rate": 9.369208600422458e-06,
+      "loss": 0.852,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0669061825894104,
+      "learning_rate": 9.368576742693369e-06,
+      "loss": 1.1588,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6436487702130977,
+      "learning_rate": 9.36794458998799e-06,
+      "loss": 0.8615,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.658725649119393,
+      "learning_rate": 9.367312142349003e-06,
+      "loss": 0.8523,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4229937765685026,
+      "learning_rate": 9.366679399819115e-06,
+      "loss": 0.8424,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.451585656628338,
+      "learning_rate": 9.366046362441047e-06,
+      "loss": 0.8944,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8821858459976993,
+      "learning_rate": 9.365413030257546e-06,
+      "loss": 1.1295,
+      "step": 2334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4959518534594562,
+      "learning_rate": 9.364779403311375e-06,
+      "loss": 0.8526,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5124655267061424,
+      "learning_rate": 9.36414548164532e-06,
+      "loss": 0.7985,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5384832908460226,
+      "learning_rate": 9.36351126530218e-06,
+      "loss": 0.7325,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8086932569663726,
+      "learning_rate": 9.362876754324784e-06,
+      "loss": 1.17,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6517227970793549,
+      "learning_rate": 9.36224194875597e-06,
+      "loss": 0.818,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7307418044484764,
+      "learning_rate": 9.361606848638607e-06,
+      "loss": 0.7954,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4615989301131884,
+      "learning_rate": 9.360971454015577e-06,
+      "loss": 0.875,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3607220998225291,
+      "learning_rate": 9.360335764929781e-06,
+      "loss": 0.7927,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4229704985421556,
+      "learning_rate": 9.359699781424144e-06,
+      "loss": 0.8321,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.533270345653547,
+      "learning_rate": 9.359063503541609e-06,
+      "loss": 0.7553,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5394674834929325,
+      "learning_rate": 9.358426931325137e-06,
+      "loss": 0.9304,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5560860093133215,
+      "learning_rate": 9.357790064817715e-06,
+      "loss": 0.8497,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4672742368659368,
+      "learning_rate": 9.357152904062342e-06,
+      "loss": 0.7511,
+      "step": 2347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5284672390833025,
+      "learning_rate": 9.356515449102041e-06,
+      "loss": 0.8416,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.423839998102108,
+      "learning_rate": 9.355877699979856e-06,
+      "loss": 0.8061,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5454660096995154,
+      "learning_rate": 9.355239656738849e-06,
+      "loss": 0.8088,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4643021111025343,
+      "learning_rate": 9.354601319422099e-06,
+      "loss": 0.7818,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5083261911109935,
+      "learning_rate": 9.353962688072713e-06,
+      "loss": 0.8328,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4342587858732563,
+      "learning_rate": 9.35332376273381e-06,
+      "loss": 0.7935,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5032997842952949,
+      "learning_rate": 9.352684543448532e-06,
+      "loss": 0.7813,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5906871926804294,
+      "learning_rate": 9.35204503026004e-06,
+      "loss": 0.8072,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3883537644422235,
+      "learning_rate": 9.351405223211517e-06,
+      "loss": 0.741,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6477209592182354,
+      "learning_rate": 9.350765122346162e-06,
+      "loss": 0.8621,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9100279066147571,
+      "learning_rate": 9.350124727707197e-06,
+      "loss": 1.1618,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5167882830320787,
+      "learning_rate": 9.349484039337864e-06,
+      "loss": 0.8254,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8438362803754844,
+      "learning_rate": 9.348843057281423e-06,
+      "loss": 1.1377,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5818584461513054,
+      "learning_rate": 9.348201781581154e-06,
+      "loss": 0.817,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.524839122679668,
+      "learning_rate": 9.347560212280359e-06,
+      "loss": 0.8146,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.480583008177718,
+      "learning_rate": 9.346918349422356e-06,
+      "loss": 0.8449,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5257778025003537,
+      "learning_rate": 9.346276193050488e-06,
+      "loss": 0.7951,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.505932257298453,
+      "learning_rate": 9.345633743208112e-06,
+      "loss": 0.8322,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5057411829192455,
+      "learning_rate": 9.344990999938609e-06,
+      "loss": 0.7622,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4843200979777433,
+      "learning_rate": 9.344347963285376e-06,
+      "loss": 0.8292,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5398148645388707,
+      "learning_rate": 9.343704633291836e-06,
+      "loss": 0.8079,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4276558757653097,
+      "learning_rate": 9.343061010001428e-06,
+      "loss": 0.8099,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0748682044265536,
+      "learning_rate": 9.34241709345761e-06,
+      "loss": 1.1512,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4111492860890478,
+      "learning_rate": 9.341772883703859e-06,
+      "loss": 0.6905,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.57202516713628,
+      "learning_rate": 9.341128380783674e-06,
+      "loss": 0.8622,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5025490337411262,
+      "learning_rate": 9.340483584740576e-06,
+      "loss": 0.7899,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.590149078433891,
+      "learning_rate": 9.3398384956181e-06,
+      "loss": 0.8923,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5599930091379783,
+      "learning_rate": 9.339193113459805e-06,
+      "loss": 0.8719,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5609913830512618,
+      "learning_rate": 9.33854743830927e-06,
+      "loss": 0.8209,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5585473280546345,
+      "learning_rate": 9.33790147021009e-06,
+      "loss": 0.8199,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.506499313653081,
+      "learning_rate": 9.337255209205884e-06,
+      "loss": 0.7942,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9669577121380751,
+      "learning_rate": 9.336608655340289e-06,
+      "loss": 1.163,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.516845257474133,
+      "learning_rate": 9.335961808656961e-06,
+      "loss": 0.8396,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5110682536259774,
+      "learning_rate": 9.335314669199576e-06,
+      "loss": 0.736,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8013545142156087,
+      "learning_rate": 9.334667237011832e-06,
+      "loss": 1.1651,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8016399957230045,
+      "learning_rate": 9.334019512137444e-06,
+      "loss": 1.1514,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6806599893604186,
+      "learning_rate": 9.33337149462015e-06,
+      "loss": 0.8402,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6376706679120825,
+      "learning_rate": 9.332723184503702e-06,
+      "loss": 0.857,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4435748357656826,
+      "learning_rate": 9.332074581831879e-06,
+      "loss": 0.8358,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6541935567071493,
+      "learning_rate": 9.331425686648472e-06,
+      "loss": 0.7909,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4425880030565568,
+      "learning_rate": 9.330776498997299e-06,
+      "loss": 0.8275,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.455407351325235,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.8746,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3703728211259008,
+      "learning_rate": 9.32947724646701e-06,
+      "loss": 0.8046,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5227572964435723,
+      "learning_rate": 9.328827181675626e-06,
+      "loss": 0.791,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5154751568080274,
+      "learning_rate": 9.328176824591928e-06,
+      "loss": 0.846,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.559916459077352,
+      "learning_rate": 9.327526175259837e-06,
+      "loss": 0.8182,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6595893841742924,
+      "learning_rate": 9.326875233723282e-06,
+      "loss": 0.8672,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.433388852378212,
+      "learning_rate": 9.326224000026217e-06,
+      "loss": 0.8516,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5409636474799226,
+      "learning_rate": 9.325572474212615e-06,
+      "loss": 0.8178,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4928274800591503,
+      "learning_rate": 9.324920656326468e-06,
+      "loss": 0.8737,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.532936774849093,
+      "learning_rate": 9.32426854641179e-06,
+      "loss": 0.9262,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4693532939019567,
+      "learning_rate": 9.323616144512612e-06,
+      "loss": 0.8259,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4717795455295621,
+      "learning_rate": 9.322963450672984e-06,
+      "loss": 0.7841,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4355582691699873,
+      "learning_rate": 9.322310464936979e-06,
+      "loss": 0.7734,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5155085427752697,
+      "learning_rate": 9.321657187348689e-06,
+      "loss": 0.8711,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.389644166843602,
+      "learning_rate": 9.321003617952222e-06,
+      "loss": 0.7172,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0418753545268298,
+      "learning_rate": 9.32034975679171e-06,
+      "loss": 1.1317,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4722193976993903,
+      "learning_rate": 9.319695603911306e-06,
+      "loss": 0.8532,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4569090508411682,
+      "learning_rate": 9.319041159355175e-06,
+      "loss": 0.8036,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8756320403946444,
+      "learning_rate": 9.318386423167508e-06,
+      "loss": 1.1455,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8003882782038089,
+      "learning_rate": 9.317731395392517e-06,
+      "loss": 1.1308,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5484745029500007,
+      "learning_rate": 9.31707607607443e-06,
+      "loss": 0.8801,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5662333647510236,
+      "learning_rate": 9.316420465257492e-06,
+      "loss": 0.7722,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9511981292383505,
+      "learning_rate": 9.315764562985976e-06,
+      "loss": 1.0938,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8798321803541076,
+      "learning_rate": 9.315108369304168e-06,
+      "loss": 0.8748,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6020128456646812,
+      "learning_rate": 9.314451884256376e-06,
+      "loss": 0.8416,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5082439629937063,
+      "learning_rate": 9.313795107886925e-06,
+      "loss": 0.8357,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4855274676809238,
+      "learning_rate": 9.313138040240167e-06,
+      "loss": 0.7872,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.494226939904471,
+      "learning_rate": 9.312480681360465e-06,
+      "loss": 0.8576,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.444996065038447,
+      "learning_rate": 9.311823031292205e-06,
+      "loss": 0.872,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4986834951295838,
+      "learning_rate": 9.311165090079795e-06,
+      "loss": 0.7465,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5255595701778242,
+      "learning_rate": 9.310506857767662e-06,
+      "loss": 0.7477,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9497616453770169,
+      "learning_rate": 9.309848334400247e-06,
+      "loss": 1.1239,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.941344763717746,
+      "learning_rate": 9.309189520022018e-06,
+      "loss": 1.1396,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5229017025372245,
+      "learning_rate": 9.308530414677459e-06,
+      "loss": 0.8217,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.65285160590843,
+      "learning_rate": 9.307871018411074e-06,
+      "loss": 0.7857,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.525867278169634,
+      "learning_rate": 9.307211331267389e-06,
+      "loss": 0.7443,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6742388987345733,
+      "learning_rate": 9.306551353290945e-06,
+      "loss": 0.7903,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4293922445708995,
+      "learning_rate": 9.305891084526306e-06,
+      "loss": 0.791,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5527002548530668,
+      "learning_rate": 9.305230525018054e-06,
+      "loss": 0.7775,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5547744785593949,
+      "learning_rate": 9.304569674810794e-06,
+      "loss": 0.8584,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3939679049128477,
+      "learning_rate": 9.303908533949146e-06,
+      "loss": 1.1587,
+      "step": 2429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5065333760076325,
+      "learning_rate": 9.303247102477752e-06,
+      "loss": 0.7313,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.507308843062549,
+      "learning_rate": 9.302585380441274e-06,
+      "loss": 0.8194,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4244533122215144,
+      "learning_rate": 9.301923367884393e-06,
+      "loss": 0.8452,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4971976434654353,
+      "learning_rate": 9.301261064851807e-06,
+      "loss": 0.8444,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5536367082200695,
+      "learning_rate": 9.30059847138824e-06,
+      "loss": 0.765,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4015697339177768,
+      "learning_rate": 9.299935587538432e-06,
+      "loss": 0.7678,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6265026704126035,
+      "learning_rate": 9.29927241334714e-06,
+      "loss": 0.878,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5580150395581103,
+      "learning_rate": 9.298608948859141e-06,
+      "loss": 0.7204,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6060996141275004,
+      "learning_rate": 9.29794519411924e-06,
+      "loss": 0.8664,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4138747150557152,
+      "learning_rate": 9.29728114917225e-06,
+      "loss": 0.7184,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.549837754651771,
+      "learning_rate": 9.29661681406301e-06,
+      "loss": 0.8208,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4570257426225541,
+      "learning_rate": 9.29595218883638e-06,
+      "loss": 0.9175,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7359700828978222,
+      "learning_rate": 9.295287273537232e-06,
+      "loss": 0.7988,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.569880101367236,
+      "learning_rate": 9.294622068210466e-06,
+      "loss": 0.7769,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.972130423103631,
+      "learning_rate": 9.293956572900999e-06,
+      "loss": 1.1301,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3656588943902683,
+      "learning_rate": 9.293290787653766e-06,
+      "loss": 0.8039,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.604079322581345,
+      "learning_rate": 9.292624712513721e-06,
+      "loss": 0.8844,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4362205432912685,
+      "learning_rate": 9.29195834752584e-06,
+      "loss": 0.8596,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8702274536498518,
+      "learning_rate": 9.291291692735116e-06,
+      "loss": 0.8681,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5005170281505649,
+      "learning_rate": 9.290624748186565e-06,
+      "loss": 0.8316,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5889078648588482,
+      "learning_rate": 9.28995751392522e-06,
+      "loss": 0.8171,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.491605272585685,
+      "learning_rate": 9.289289989996133e-06,
+      "loss": 0.8172,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9057442138624477,
+      "learning_rate": 9.28862217644438e-06,
+      "loss": 1.1538,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4946484508778342,
+      "learning_rate": 9.28795407331505e-06,
+      "loss": 0.8334,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4814272815155147,
+      "learning_rate": 9.287285680653254e-06,
+      "loss": 0.8159,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.081182927799621,
+      "learning_rate": 9.28661699850413e-06,
+      "loss": 0.8244,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3895171644865245,
+      "learning_rate": 9.285948026912822e-06,
+      "loss": 0.6916,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4543243338088105,
+      "learning_rate": 9.285278765924502e-06,
+      "loss": 0.7757,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.696546952911893,
+      "learning_rate": 9.284609215584361e-06,
+      "loss": 0.8494,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4302752099200533,
+      "learning_rate": 9.283939375937609e-06,
+      "loss": 0.7587,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5597042744311742,
+      "learning_rate": 9.283269247029475e-06,
+      "loss": 0.8437,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4637461480282516,
+      "learning_rate": 9.282598828905205e-06,
+      "loss": 0.8102,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0546368114545133,
+      "learning_rate": 9.28192812161007e-06,
+      "loss": 1.163,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9489094753040603,
+      "learning_rate": 9.281257125189358e-06,
+      "loss": 1.1457,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5191850039413246,
+      "learning_rate": 9.280585839688377e-06,
+      "loss": 0.7515,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5345111822003539,
+      "learning_rate": 9.279914265152448e-06,
+      "loss": 0.8311,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7245765289943977,
+      "learning_rate": 9.279242401626924e-06,
+      "loss": 0.8596,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0403186654912064,
+      "learning_rate": 9.278570249157166e-06,
+      "loss": 1.1062,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7356703020813822,
+      "learning_rate": 9.277897807788562e-06,
+      "loss": 0.7534,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6331958969814673,
+      "learning_rate": 9.277225077566519e-06,
+      "loss": 0.7942,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5899155566869414,
+      "learning_rate": 9.276552058536454e-06,
+      "loss": 0.8181,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5586433126987576,
+      "learning_rate": 9.275878750743818e-06,
+      "loss": 0.7934,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3533408180703748,
+      "learning_rate": 9.275205154234069e-06,
+      "loss": 0.6761,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4255730194063503,
+      "learning_rate": 9.274531269052693e-06,
+      "loss": 0.7641,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5481276272250761,
+      "learning_rate": 9.273857095245192e-06,
+      "loss": 0.8269,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4969669216469579,
+      "learning_rate": 9.273182632857087e-06,
+      "loss": 0.8026,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5024078287475118,
+      "learning_rate": 9.272507881933919e-06,
+      "loss": 0.823,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9481250189586695,
+      "learning_rate": 9.271832842521249e-06,
+      "loss": 1.1387,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5587062328058077,
+      "learning_rate": 9.271157514664658e-06,
+      "loss": 0.7834,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5398695977260521,
+      "learning_rate": 9.270481898409744e-06,
+      "loss": 0.8485,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6116459211264786,
+      "learning_rate": 9.26980599380213e-06,
+      "loss": 0.7642,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4334102439634464,
+      "learning_rate": 9.269129800887448e-06,
+      "loss": 0.8102,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4429847173717756,
+      "learning_rate": 9.268453319711362e-06,
+      "loss": 0.758,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5549458653127683,
+      "learning_rate": 9.267776550319548e-06,
+      "loss": 0.8215,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5858123314893178,
+      "learning_rate": 9.2670994927577e-06,
+      "loss": 0.8247,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5471318979962474,
+      "learning_rate": 9.266422147071539e-06,
+      "loss": 0.8356,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9450752237004824,
+      "learning_rate": 9.265744513306798e-06,
+      "loss": 1.1276,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.47768196589061,
+      "learning_rate": 9.265066591509234e-06,
+      "loss": 0.7348,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4795328445863316,
+      "learning_rate": 9.264388381724621e-06,
+      "loss": 0.8471,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4816607367332844,
+      "learning_rate": 9.263709883998753e-06,
+      "loss": 0.8351,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.778375577697499,
+      "learning_rate": 9.263031098377445e-06,
+      "loss": 1.1364,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7721528459276297,
+      "learning_rate": 9.262352024906526e-06,
+      "loss": 1.1687,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.477379878079317,
+      "learning_rate": 9.261672663631854e-06,
+      "loss": 0.8056,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5180091211707516,
+      "learning_rate": 9.2609930145993e-06,
+      "loss": 0.7896,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4564436144952326,
+      "learning_rate": 9.260313077854753e-06,
+      "loss": 0.8115,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8089669398112698,
+      "learning_rate": 9.259632853444126e-06,
+      "loss": 1.1157,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3642970029582884,
+      "learning_rate": 9.258952341413347e-06,
+      "loss": 0.8287,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3972094397470123,
+      "learning_rate": 9.258271541808368e-06,
+      "loss": 0.8342,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4330029809673153,
+      "learning_rate": 9.257590454675159e-06,
+      "loss": 0.7664,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.600136785486778,
+      "learning_rate": 9.256909080059703e-06,
+      "loss": 0.8799,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7180306898502098,
+      "learning_rate": 9.256227418008015e-06,
+      "loss": 0.9033,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.54831355082679,
+      "learning_rate": 9.255545468566119e-06,
+      "loss": 0.7862,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.895986414167438,
+      "learning_rate": 9.254863231780062e-06,
+      "loss": 1.1261,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6182679046285227,
+      "learning_rate": 9.25418070769591e-06,
+      "loss": 0.8103,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4988999942857322,
+      "learning_rate": 9.253497896359749e-06,
+      "loss": 0.7803,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3400201909480869,
+      "learning_rate": 9.252814797817682e-06,
+      "loss": 0.7023,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8258353887860432,
+      "learning_rate": 9.252131412115838e-06,
+      "loss": 1.1591,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6029732018166505,
+      "learning_rate": 9.251447739300356e-06,
+      "loss": 0.8763,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5666731684174307,
+      "learning_rate": 9.250763779417402e-06,
+      "loss": 0.8323,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5369663265896087,
+      "learning_rate": 9.250079532513158e-06,
+      "loss": 0.8142,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4847669923366615,
+      "learning_rate": 9.249394998633825e-06,
+      "loss": 0.8802,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.436761575791087,
+      "learning_rate": 9.248710177825627e-06,
+      "loss": 0.7944,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6367701116655093,
+      "learning_rate": 9.2480250701348e-06,
+      "loss": 0.8334,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6306418273280565,
+      "learning_rate": 9.247339675607606e-06,
+      "loss": 0.7679,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5620956372105916,
+      "learning_rate": 9.246653994290327e-06,
+      "loss": 0.8324,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5526424229480875,
+      "learning_rate": 9.245968026229258e-06,
+      "loss": 0.7437,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7658279961660834,
+      "learning_rate": 9.24528177147072e-06,
+      "loss": 0.7855,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5150501861297674,
+      "learning_rate": 9.24459523006105e-06,
+      "loss": 0.7918,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.556825071416571,
+      "learning_rate": 9.243908402046602e-06,
+      "loss": 0.819,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4696200469352885,
+      "learning_rate": 9.243221287473755e-06,
+      "loss": 0.7696,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.479881402281857,
+      "learning_rate": 9.242533886388905e-06,
+      "loss": 0.8913,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5494408116995453,
+      "learning_rate": 9.241846198838466e-06,
+      "loss": 0.9769,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9717623171969201,
+      "learning_rate": 9.241158224868871e-06,
+      "loss": 1.1252,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6354236682458738,
+      "learning_rate": 9.240469964526576e-06,
+      "loss": 0.9319,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8371376312444543,
+      "learning_rate": 9.239781417858052e-06,
+      "loss": 1.1401,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.42056073824133,
+      "learning_rate": 9.23909258490979e-06,
+      "loss": 0.8026,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5642446989624226,
+      "learning_rate": 9.238403465728306e-06,
+      "loss": 0.7566,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5364295275921966,
+      "learning_rate": 9.237714060360128e-06,
+      "loss": 0.7784,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5327505734388687,
+      "learning_rate": 9.237024368851805e-06,
+      "loss": 0.8669,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6389460059777625,
+      "learning_rate": 9.236334391249909e-06,
+      "loss": 0.8271,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.561251447729545,
+      "learning_rate": 9.235644127601028e-06,
+      "loss": 0.8893,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4484755766900925,
+      "learning_rate": 9.23495357795177e-06,
+      "loss": 0.888,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5256584642615125,
+      "learning_rate": 9.234262742348764e-06,
+      "loss": 0.8083,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4356220132502138,
+      "learning_rate": 9.233571620838653e-06,
+      "loss": 0.7838,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5272841763595624,
+      "learning_rate": 9.232880213468106e-06,
+      "loss": 0.7579,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5261167968787952,
+      "learning_rate": 9.23218852028381e-06,
+      "loss": 0.7463,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4413707977197021,
+      "learning_rate": 9.231496541332465e-06,
+      "loss": 0.8054,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.247575903128527,
+      "learning_rate": 9.230804276660799e-06,
+      "loss": 1.166,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5321147167756957,
+      "learning_rate": 9.230111726315553e-06,
+      "loss": 0.8438,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9412893935866841,
+      "learning_rate": 9.229418890343491e-06,
+      "loss": 1.1398,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.482744276685692,
+      "learning_rate": 9.228725768791394e-06,
+      "loss": 0.833,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5658094804955902,
+      "learning_rate": 9.228032361706065e-06,
+      "loss": 0.8442,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5069577774286516,
+      "learning_rate": 9.227338669134322e-06,
+      "loss": 0.8715,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5247796765269208,
+      "learning_rate": 9.226644691123006e-06,
+      "loss": 0.7684,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2572329616769877,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 1.1534,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.63323559261434,
+      "learning_rate": 9.225255878969108e-06,
+      "loss": 0.8586,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5332105776010034,
+      "learning_rate": 9.224561044920303e-06,
+      "loss": 0.8846,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5755331550683576,
+      "learning_rate": 9.223865925619476e-06,
+      "loss": 0.8072,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9185212800110313,
+      "learning_rate": 9.223170521113563e-06,
+      "loss": 1.1449,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.615931671617186,
+      "learning_rate": 9.222474831449519e-06,
+      "loss": 0.8364,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8208561158773751,
+      "learning_rate": 9.221778856674319e-06,
+      "loss": 1.1287,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4641742592212923,
+      "learning_rate": 9.221082596834959e-06,
+      "loss": 0.8177,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.458968604948614,
+      "learning_rate": 9.220386051978449e-06,
+      "loss": 0.8455,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9015409137708176,
+      "learning_rate": 9.219689222151821e-06,
+      "loss": 1.1376,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5384815928545281,
+      "learning_rate": 9.21899210740213e-06,
+      "loss": 0.8761,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4867110105159658,
+      "learning_rate": 9.218294707776441e-06,
+      "loss": 0.868,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5274993247221518,
+      "learning_rate": 9.217597023321851e-06,
+      "loss": 0.8214,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5432830782903058,
+      "learning_rate": 9.216899054085465e-06,
+      "loss": 0.8044,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.87359682339141,
+      "learning_rate": 9.216200800114412e-06,
+      "loss": 1.1544,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5828680272876856,
+      "learning_rate": 9.215502261455839e-06,
+      "loss": 0.7806,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4330020550342948,
+      "learning_rate": 9.214803438156916e-06,
+      "loss": 0.8955,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.837724005233504,
+      "learning_rate": 9.214104330264826e-06,
+      "loss": 1.12,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8050881256311607,
+      "learning_rate": 9.213404937826775e-06,
+      "loss": 1.1214,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5824611672829865,
+      "learning_rate": 9.212705260889991e-06,
+      "loss": 0.8794,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.664006373466688,
+      "learning_rate": 9.212005299501712e-06,
+      "loss": 0.8225,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4814192913195985,
+      "learning_rate": 9.211305053709204e-06,
+      "loss": 0.8337,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5152630589211757,
+      "learning_rate": 9.210604523559749e-06,
+      "loss": 0.7804,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5000665793720065,
+      "learning_rate": 9.20990370910065e-06,
+      "loss": 0.8156,
+      "step": 2567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.606592282992881,
+      "learning_rate": 9.209202610379225e-06,
+      "loss": 0.7436,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5115778571313527,
+      "learning_rate": 9.208501227442815e-06,
+      "loss": 0.8822,
+      "step": 2569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6505058735566853,
+      "learning_rate": 9.207799560338779e-06,
+      "loss": 0.8475,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5266554406780122,
+      "learning_rate": 9.207097609114495e-06,
+      "loss": 0.7133,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3930525103010445,
+      "learning_rate": 9.206395373817358e-06,
+      "loss": 0.8466,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.049417228015875,
+      "learning_rate": 9.20569285449479e-06,
+      "loss": 1.1714,
+      "step": 2573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5605174121424572,
+      "learning_rate": 9.204990051194223e-06,
+      "loss": 0.8447,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4839626542236661,
+      "learning_rate": 9.204286963963112e-06,
+      "loss": 0.7874,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5589991407822532,
+      "learning_rate": 9.20358359284893e-06,
+      "loss": 0.8202,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.571370009741804,
+      "learning_rate": 9.202879937899175e-06,
+      "loss": 0.8047,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.61653770393215,
+      "learning_rate": 9.202175999161353e-06,
+      "loss": 0.7918,
+      "step": 2578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.503546534742903,
+      "learning_rate": 9.201471776682999e-06,
+      "loss": 0.7798,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4977353880812723,
+      "learning_rate": 9.200767270511666e-06,
+      "loss": 0.7605,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9138957659664549,
+      "learning_rate": 9.200062480694919e-06,
+      "loss": 1.1572,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6765468127964391,
+      "learning_rate": 9.199357407280349e-06,
+      "loss": 0.8251,
+      "step": 2582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.47209785524806,
+      "learning_rate": 9.198652050315566e-06,
+      "loss": 0.8069,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4681897455342794,
+      "learning_rate": 9.197946409848196e-06,
+      "loss": 0.838,
+      "step": 2584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6703315928522064,
+      "learning_rate": 9.197240485925883e-06,
+      "loss": 0.7673,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8268171043618279,
+      "learning_rate": 9.196534278596296e-06,
+      "loss": 1.1283,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.473064043645414,
+      "learning_rate": 9.195827787907118e-06,
+      "loss": 0.8215,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5165929848954347,
+      "learning_rate": 9.195121013906055e-06,
+      "loss": 0.8348,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7521464642457938,
+      "learning_rate": 9.194413956640827e-06,
+      "loss": 0.8401,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6759232551961154,
+      "learning_rate": 9.193706616159179e-06,
+      "loss": 0.815,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5025848188094852,
+      "learning_rate": 9.19299899250887e-06,
+      "loss": 0.8319,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5434808291464792,
+      "learning_rate": 9.192291085737682e-06,
+      "loss": 0.8714,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4012653954295182,
+      "learning_rate": 9.191582895893412e-06,
+      "loss": 0.7851,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4532579440452962,
+      "learning_rate": 9.19087442302388e-06,
+      "loss": 0.7893,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4869734084702937,
+      "learning_rate": 9.190165667176924e-06,
+      "loss": 0.8197,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.553291644752851,
+      "learning_rate": 9.189456628400403e-06,
+      "loss": 0.8394,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5304013902106708,
+      "learning_rate": 9.18874730674219e-06,
+      "loss": 0.806,
+      "step": 2597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7561513496702543,
+      "learning_rate": 9.188037702250179e-06,
+      "loss": 0.7906,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4673145029558168,
+      "learning_rate": 9.187327814972286e-06,
+      "loss": 0.8431,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4463770074354507,
+      "learning_rate": 9.186617644956445e-06,
+      "loss": 0.7968,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8709882522778485,
+      "learning_rate": 9.185907192250608e-06,
+      "loss": 1.1098,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.524867052480939,
+      "learning_rate": 9.185196456902744e-06,
+      "loss": 0.8664,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6742710632392814,
+      "learning_rate": 9.184485438960846e-06,
+      "loss": 0.8441,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4722461553669444,
+      "learning_rate": 9.183774138472923e-06,
+      "loss": 0.7387,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5621579332120756,
+      "learning_rate": 9.183062555487003e-06,
+      "loss": 0.7752,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4166504171625893,
+      "learning_rate": 9.182350690051134e-06,
+      "loss": 0.7216,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.447639711506031,
+      "learning_rate": 9.181638542213383e-06,
+      "loss": 0.7652,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7370137799100938,
+      "learning_rate": 9.180926112021837e-06,
+      "loss": 0.8311,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6534942037393068,
+      "learning_rate": 9.180213399524599e-06,
+      "loss": 0.7587,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5315085597071736,
+      "learning_rate": 9.179500404769792e-06,
+      "loss": 0.9,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4532660850305146,
+      "learning_rate": 9.178787127805561e-06,
+      "loss": 0.7569,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5743370989835592,
+      "learning_rate": 9.178073568680071e-06,
+      "loss": 0.8462,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4789297279405655,
+      "learning_rate": 9.177359727441498e-06,
+      "loss": 0.7965,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.720299041971694,
+      "learning_rate": 9.176645604138041e-06,
+      "loss": 0.8639,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5265942872703935,
+      "learning_rate": 9.175931198817926e-06,
+      "loss": 0.8965,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5045024903079554,
+      "learning_rate": 9.175216511529387e-06,
+      "loss": 0.8825,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8526720037831905,
+      "learning_rate": 9.17450154232068e-06,
+      "loss": 1.1497,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6601239101603003,
+      "learning_rate": 9.173786291240085e-06,
+      "loss": 0.7962,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6026089673236867,
+      "learning_rate": 9.173070758335895e-06,
+      "loss": 0.7874,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4312355288495462,
+      "learning_rate": 9.172354943656428e-06,
+      "loss": 0.7483,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5114226882317068,
+      "learning_rate": 9.17163884725001e-06,
+      "loss": 0.7307,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3240011011299948,
+      "learning_rate": 9.170922469165001e-06,
+      "loss": 0.769,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5657328032770272,
+      "learning_rate": 9.170205809449768e-06,
+      "loss": 0.9282,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8525883100972355,
+      "learning_rate": 9.169488868152704e-06,
+      "loss": 1.172,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9080880147917914,
+      "learning_rate": 9.168771645322217e-06,
+      "loss": 1.1842,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5139054882992735,
+      "learning_rate": 9.168054141006737e-06,
+      "loss": 0.7665,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5023827998544188,
+      "learning_rate": 9.16733635525471e-06,
+      "loss": 0.7744,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6251510502624804,
+      "learning_rate": 9.166618288114602e-06,
+      "loss": 0.7657,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5021508625957891,
+      "learning_rate": 9.1658999396349e-06,
+      "loss": 0.7771,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1106787003922318,
+      "learning_rate": 9.165181309864108e-06,
+      "loss": 1.1394,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4899257011392206,
+      "learning_rate": 9.16446239885075e-06,
+      "loss": 0.797,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8618978411886932,
+      "learning_rate": 9.16374320664337e-06,
+      "loss": 1.1425,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8128061374458699,
+      "learning_rate": 9.163023733290525e-06,
+      "loss": 1.1801,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5031824838996164,
+      "learning_rate": 9.162303978840801e-06,
+      "loss": 0.8299,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.50529889858848,
+      "learning_rate": 9.161583943342793e-06,
+      "loss": 0.7856,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1406115887477155,
+      "learning_rate": 9.16086362684512e-06,
+      "loss": 1.1417,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4999525213190121,
+      "learning_rate": 9.160143029396422e-06,
+      "loss": 0.738,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.536523850494388,
+      "learning_rate": 9.159422151045354e-06,
+      "loss": 0.7652,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5588538970599959,
+      "learning_rate": 9.15870099184059e-06,
+      "loss": 0.8002,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.387876477153082,
+      "learning_rate": 9.157979551830827e-06,
+      "loss": 0.8043,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9106835827118681,
+      "learning_rate": 9.157257831064776e-06,
+      "loss": 1.1672,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6131456691463888,
+      "learning_rate": 9.15653582959117e-06,
+      "loss": 0.8273,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5174240954541203,
+      "learning_rate": 9.155813547458761e-06,
+      "loss": 0.8258,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5010847854201181,
+      "learning_rate": 9.155090984716319e-06,
+      "loss": 0.7515,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.005303931720966,
+      "learning_rate": 9.154368141412632e-06,
+      "loss": 1.1716,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4092357534148454,
+      "learning_rate": 9.153645017596508e-06,
+      "loss": 0.8775,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.572211652166216,
+      "learning_rate": 9.152921613316775e-06,
+      "loss": 0.8204,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.609865365479956,
+      "learning_rate": 9.152197928622278e-06,
+      "loss": 0.8481,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8507665366491778,
+      "learning_rate": 9.151473963561884e-06,
+      "loss": 1.1741,
+      "step": 2649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4476063491119602,
+      "learning_rate": 9.150749718184473e-06,
+      "loss": 0.7695,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5812445317457728,
+      "learning_rate": 9.15002519253895e-06,
+      "loss": 0.8531,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.61084006384248,
+      "learning_rate": 9.149300386674236e-06,
+      "loss": 0.7527,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4069163506952242,
+      "learning_rate": 9.148575300639273e-06,
+      "loss": 0.8479,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5519375410644023,
+      "learning_rate": 9.147849934483019e-06,
+      "loss": 0.8238,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5411847790140933,
+      "learning_rate": 9.147124288254453e-06,
+      "loss": 0.8032,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9304813232897968,
+      "learning_rate": 9.146398362002572e-06,
+      "loss": 1.1472,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3883197491904644,
+      "learning_rate": 9.145672155776392e-06,
+      "loss": 0.8326,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8874458543774099,
+      "learning_rate": 9.144945669624948e-06,
+      "loss": 1.1624,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4258527114866246,
+      "learning_rate": 9.144218903597295e-06,
+      "loss": 0.8421,
+      "step": 2659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.534751876827594,
+      "learning_rate": 9.143491857742505e-06,
+      "loss": 0.8222,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9072853566008295,
+      "learning_rate": 9.142764532109672e-06,
+      "loss": 1.1205,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9354698862017317,
+      "learning_rate": 9.142036926747904e-06,
+      "loss": 1.1468,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.556081593839327,
+      "learning_rate": 9.14130904170633e-06,
+      "loss": 0.781,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4851954122712183,
+      "learning_rate": 9.1405808770341e-06,
+      "loss": 0.8597,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7756899977019704,
+      "learning_rate": 9.13985243278038e-06,
+      "loss": 1.1323,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5854120986171591,
+      "learning_rate": 9.13912370899436e-06,
+      "loss": 0.7843,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5975845807215767,
+      "learning_rate": 9.138394705725242e-06,
+      "loss": 0.8409,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.691545023556888,
+      "learning_rate": 9.13766542302225e-06,
+      "loss": 0.9271,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9748404320540639,
+      "learning_rate": 9.136935860934628e-06,
+      "loss": 1.1442,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6656328330941927,
+      "learning_rate": 9.136206019511635e-06,
+      "loss": 0.8624,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4950492401787432,
+      "learning_rate": 9.135475898802555e-06,
+      "loss": 0.7505,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5281006017056877,
+      "learning_rate": 9.134745498856685e-06,
+      "loss": 0.8405,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7885699592985145,
+      "learning_rate": 9.134014819723346e-06,
+      "loss": 1.1411,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6455930353893764,
+      "learning_rate": 9.133283861451872e-06,
+      "loss": 0.7689,
+      "step": 2674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5538032286478747,
+      "learning_rate": 9.13255262409162e-06,
+      "loss": 0.8422,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5466189222323465,
+      "learning_rate": 9.131821107691967e-06,
+      "loss": 0.8052,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5068723598874263,
+      "learning_rate": 9.131089312302304e-06,
+      "loss": 0.8001,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6343859394886489,
+      "learning_rate": 9.130357237972044e-06,
+      "loss": 0.7871,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.545503926580848,
+      "learning_rate": 9.129624884750618e-06,
+      "loss": 0.7395,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5212960381638376,
+      "learning_rate": 9.128892252687478e-06,
+      "loss": 0.879,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.604478981136443,
+      "learning_rate": 9.128159341832092e-06,
+      "loss": 0.86,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5792169595204804,
+      "learning_rate": 9.127426152233946e-06,
+      "loss": 0.7914,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4344271960325858,
+      "learning_rate": 9.126692683942549e-06,
+      "loss": 0.7376,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.418394646871754,
+      "learning_rate": 9.125958937007427e-06,
+      "loss": 0.7976,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5240792451892742,
+      "learning_rate": 9.125224911478122e-06,
+      "loss": 0.8043,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.184244534521629,
+      "learning_rate": 9.124490607404197e-06,
+      "loss": 0.8942,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6554362341733437,
+      "learning_rate": 9.123756024835237e-06,
+      "loss": 0.8902,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.526072381036271,
+      "learning_rate": 9.123021163820839e-06,
+      "loss": 0.8028,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4732700580607097,
+      "learning_rate": 9.122286024410625e-06,
+      "loss": 0.7848,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8843371575063754,
+      "learning_rate": 9.121550606654232e-06,
+      "loss": 1.1591,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.499571768179961,
+      "learning_rate": 9.120814910601319e-06,
+      "loss": 0.7329,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8395161108973446,
+      "learning_rate": 9.120078936301559e-06,
+      "loss": 1.1559,
+      "step": 2692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5377599892200695,
+      "learning_rate": 9.119342683804649e-06,
+      "loss": 0.88,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5202991547056606,
+      "learning_rate": 9.1186061531603e-06,
+      "loss": 0.8434,
+      "step": 2694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.412368540945013,
+      "learning_rate": 9.117869344418246e-06,
+      "loss": 0.8059,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4858519166017816,
+      "learning_rate": 9.11713225762824e-06,
+      "loss": 0.7971,
+      "step": 2696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5428875257765227,
+      "learning_rate": 9.116394892840047e-06,
+      "loss": 0.8496,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5891228521929235,
+      "learning_rate": 9.11565725010346e-06,
+      "loss": 0.807,
+      "step": 2698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.536870788394737,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.8729,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4233762481997725,
+      "learning_rate": 9.114181130984343e-06,
+      "loss": 0.7604,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7011575231470912,
+      "learning_rate": 9.113442654701487e-06,
+      "loss": 0.8198,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0504204986841759,
+      "learning_rate": 9.112703900669577e-06,
+      "loss": 1.123,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9705936042465975,
+      "learning_rate": 9.111964868938495e-06,
+      "loss": 1.1377,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6697688602186724,
+      "learning_rate": 9.111225559558144e-06,
+      "loss": 0.762,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6243873198550438,
+      "learning_rate": 9.110485972578439e-06,
+      "loss": 0.8388,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.610735627888926,
+      "learning_rate": 9.109746108049325e-06,
+      "loss": 0.771,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6186053775134144,
+      "learning_rate": 9.109005966020753e-06,
+      "loss": 0.7712,
+      "step": 2707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.55696185568793,
+      "learning_rate": 9.108265546542705e-06,
+      "loss": 0.9213,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2605773016443498,
+      "learning_rate": 9.107524849665173e-06,
+      "loss": 1.134,
+      "step": 2709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4937501577886223,
+      "learning_rate": 9.106783875438169e-06,
+      "loss": 0.7583,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5898569633880812,
+      "learning_rate": 9.106042623911728e-06,
+      "loss": 0.8185,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4949295501407474,
+      "learning_rate": 9.105301095135902e-06,
+      "loss": 0.7081,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4667325696705478,
+      "learning_rate": 9.104559289160757e-06,
+      "loss": 0.7485,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6501943417609812,
+      "learning_rate": 9.103817206036383e-06,
+      "loss": 0.8839,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5072453247770954,
+      "learning_rate": 9.103074845812888e-06,
+      "loss": 0.8237,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.567478283694221,
+      "learning_rate": 9.102332208540396e-06,
+      "loss": 0.8534,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5703982114168584,
+      "learning_rate": 9.101589294269054e-06,
+      "loss": 0.8295,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4073780037559471,
+      "learning_rate": 9.10084610304902e-06,
+      "loss": 0.7395,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5791200969886008,
+      "learning_rate": 9.100102634930485e-06,
+      "loss": 0.8351,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8477809643021568,
+      "learning_rate": 9.099358889963643e-06,
+      "loss": 1.1753,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.577844627014302,
+      "learning_rate": 9.098614868198715e-06,
+      "loss": 0.856,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5109881445815694,
+      "learning_rate": 9.097870569685938e-06,
+      "loss": 0.8493,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5545392213081684,
+      "learning_rate": 9.097125994475572e-06,
+      "loss": 0.8081,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4936266788204047,
+      "learning_rate": 9.09638114261789e-06,
+      "loss": 0.7001,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5772067670533552,
+      "learning_rate": 9.095636014163184e-06,
+      "loss": 0.853,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4780604983786856,
+      "learning_rate": 9.09489060916177e-06,
+      "loss": 0.7789,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8261035263870105,
+      "learning_rate": 9.094144927663979e-06,
+      "loss": 1.1818,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5523003697500033,
+      "learning_rate": 9.09339896972016e-06,
+      "loss": 0.7064,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5800918336608898,
+      "learning_rate": 9.092652735380683e-06,
+      "loss": 0.8111,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5029000690236245,
+      "learning_rate": 9.091906224695935e-06,
+      "loss": 0.8594,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8518533135066468,
+      "learning_rate": 9.091159437716322e-06,
+      "loss": 1.185,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.523137945364994,
+      "learning_rate": 9.09041237449227e-06,
+      "loss": 0.8605,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5742739742087455,
+      "learning_rate": 9.08966503507422e-06,
+      "loss": 0.8582,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5193055484808926,
+      "learning_rate": 9.088917419512634e-06,
+      "loss": 0.8924,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5503271668595082,
+      "learning_rate": 9.088169527857996e-06,
+      "loss": 0.8458,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6985238254937525,
+      "learning_rate": 9.087421360160802e-06,
+      "loss": 0.9147,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4508042973157518,
+      "learning_rate": 9.08667291647157e-06,
+      "loss": 0.7582,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5508237751729161,
+      "learning_rate": 9.085924196840841e-06,
+      "loss": 0.7826,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4920365984626915,
+      "learning_rate": 9.085175201319165e-06,
+      "loss": 0.9011,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9637838530382876,
+      "learning_rate": 9.08442592995712e-06,
+      "loss": 1.1117,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.68719857680009,
+      "learning_rate": 9.083676382805295e-06,
+      "loss": 0.8822,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9266589414355086,
+      "learning_rate": 9.082926559914303e-06,
+      "loss": 0.8671,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5819406201864188,
+      "learning_rate": 9.082176461334774e-06,
+      "loss": 0.7545,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6768617501146295,
+      "learning_rate": 9.081426087117356e-06,
+      "loss": 0.8052,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5216843153131245,
+      "learning_rate": 9.080675437312715e-06,
+      "loss": 0.8654,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5506642948520322,
+      "learning_rate": 9.079924511971538e-06,
+      "loss": 0.8448,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6671674533515648,
+      "learning_rate": 9.07917331114453e-06,
+      "loss": 0.809,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5004197565331499,
+      "learning_rate": 9.078421834882409e-06,
+      "loss": 0.7968,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6253511820513293,
+      "learning_rate": 9.077670083235923e-06,
+      "loss": 0.7796,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4706554560400404,
+      "learning_rate": 9.076918056255827e-06,
+      "loss": 0.8917,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5238577876431205,
+      "learning_rate": 9.076165753992902e-06,
+      "loss": 0.8016,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4734811399930205,
+      "learning_rate": 9.075413176497944e-06,
+      "loss": 0.8063,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4929912296668069,
+      "learning_rate": 9.074660323821772e-06,
+      "loss": 0.8342,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.786005563672545,
+      "learning_rate": 9.073907196015215e-06,
+      "loss": 0.9254,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4251010957232777,
+      "learning_rate": 9.073153793129129e-06,
+      "loss": 0.8005,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9833946655020197,
+      "learning_rate": 9.072400115214385e-06,
+      "loss": 1.1498,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5834666741707812,
+      "learning_rate": 9.071646162321873e-06,
+      "loss": 0.8074,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.641944664236657,
+      "learning_rate": 9.070891934502501e-06,
+      "loss": 0.8183,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5916665907363245,
+      "learning_rate": 9.0701374318072e-06,
+      "loss": 0.8237,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8862095453467964,
+      "learning_rate": 9.06938265428691e-06,
+      "loss": 1.1363,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5361331791082247,
+      "learning_rate": 9.0686276019926e-06,
+      "loss": 0.7357,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8594920604811607,
+      "learning_rate": 9.067872274975248e-06,
+      "loss": 1.1305,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8455582108439683,
+      "learning_rate": 9.06711667328586e-06,
+      "loss": 1.1043,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8006841629645659,
+      "learning_rate": 9.066360796975452e-06,
+      "loss": 1.1573,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.523255231724165,
+      "learning_rate": 9.065604646095068e-06,
+      "loss": 0.7987,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9106503123825963,
+      "learning_rate": 9.06484822069576e-06,
+      "loss": 1.1462,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.503029594319182,
+      "learning_rate": 9.064091520828606e-06,
+      "loss": 0.7939,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5342127976171993,
+      "learning_rate": 9.0633345465447e-06,
+      "loss": 0.7906,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.424522801702873,
+      "learning_rate": 9.062577297895152e-06,
+      "loss": 0.7775,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.60410655153158,
+      "learning_rate": 9.061819774931096e-06,
+      "loss": 0.8274,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5034169546769118,
+      "learning_rate": 9.061061977703682e-06,
+      "loss": 0.7303,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6483470073625308,
+      "learning_rate": 9.060303906264076e-06,
+      "loss": 0.8671,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0371819446776256,
+      "learning_rate": 9.059545560663466e-06,
+      "loss": 1.1219,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5397995538869336,
+      "learning_rate": 9.058786940953057e-06,
+      "loss": 0.8353,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4777902407079404,
+      "learning_rate": 9.058028047184074e-06,
+      "loss": 0.7805,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4873016376334518,
+      "learning_rate": 9.057268879407757e-06,
+      "loss": 0.7861,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9273665144315468,
+      "learning_rate": 9.056509437675366e-06,
+      "loss": 1.1367,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4166990250748213,
+      "learning_rate": 9.055749722038185e-06,
+      "loss": 0.8189,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6563975138090987,
+      "learning_rate": 9.054989732547507e-06,
+      "loss": 0.7594,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5453638986035552,
+      "learning_rate": 9.05422946925465e-06,
+      "loss": 0.83,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6354881853133996,
+      "learning_rate": 9.05346893221095e-06,
+      "loss": 0.9247,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5792386151586097,
+      "learning_rate": 9.052708121467758e-06,
+      "loss": 0.817,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4924725733558986,
+      "learning_rate": 9.051947037076446e-06,
+      "loss": 0.8549,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.434354880167259,
+      "learning_rate": 9.051185679088406e-06,
+      "loss": 0.8304,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4347623792977258,
+      "learning_rate": 9.050424047555043e-06,
+      "loss": 0.7923,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4491918514006743,
+      "learning_rate": 9.049662142527788e-06,
+      "loss": 0.7308,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5805661674267073,
+      "learning_rate": 9.048899964058085e-06,
+      "loss": 0.824,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0079574171061683,
+      "learning_rate": 9.048137512197398e-06,
+      "loss": 1.1206,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8503934112479651,
+      "learning_rate": 9.04737478699721e-06,
+      "loss": 1.1266,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5003092339728685,
+      "learning_rate": 9.046611788509021e-06,
+      "loss": 0.7199,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5503689750202785,
+      "learning_rate": 9.045848516784351e-06,
+      "loss": 0.8783,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5725833674792489,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.7636,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4924882390102854,
+      "learning_rate": 9.044321153831737e-06,
+      "loss": 0.8774,
+      "step": 2793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5222432347149883,
+      "learning_rate": 9.043557062706925e-06,
+      "loss": 0.7498,
+      "step": 2794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5099066324551467,
+      "learning_rate": 9.042792698551894e-06,
+      "loss": 0.7429,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8401561015311043,
+      "learning_rate": 9.042028061418255e-06,
+      "loss": 0.8719,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.393227328475668,
+      "learning_rate": 9.04126315135764e-06,
+      "loss": 1.1803,
+      "step": 2797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4886164324953508,
+      "learning_rate": 9.040497968421694e-06,
+      "loss": 0.7996,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5694665022741647,
+      "learning_rate": 9.039732512662087e-06,
+      "loss": 0.7398,
+      "step": 2799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5709534175348423,
+      "learning_rate": 9.038966784130503e-06,
+      "loss": 0.7835,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5761764161876064,
+      "learning_rate": 9.038200782878647e-06,
+      "loss": 0.8545,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6559938013876296,
+      "learning_rate": 9.03743450895824e-06,
+      "loss": 0.7597,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4082235982383455,
+      "learning_rate": 9.036667962421023e-06,
+      "loss": 0.8656,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8770938118209625,
+      "learning_rate": 9.035901143318753e-06,
+      "loss": 1.1542,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8438130286138171,
+      "learning_rate": 9.03513405170321e-06,
+      "loss": 1.1356,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.421949868076069,
+      "learning_rate": 9.03436668762619e-06,
+      "loss": 0.8452,
+      "step": 2806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4138358421767636,
+      "learning_rate": 9.033599051139506e-06,
+      "loss": 0.7841,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.509266047202581,
+      "learning_rate": 9.032831142294988e-06,
+      "loss": 0.8081,
+      "step": 2808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4680689089062113,
+      "learning_rate": 9.032062961144493e-06,
+      "loss": 0.8288,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.557240307702731,
+      "learning_rate": 9.031294507739885e-06,
+      "loss": 0.772,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.569641847168124,
+      "learning_rate": 9.030525782133055e-06,
+      "loss": 0.7387,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4286286595317244,
+      "learning_rate": 9.029756784375907e-06,
+      "loss": 0.7392,
+      "step": 2812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5568313993386487,
+      "learning_rate": 9.028987514520366e-06,
+      "loss": 0.7944,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5391399248488427,
+      "learning_rate": 9.028217972618376e-06,
+      "loss": 0.86,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5781342163944243,
+      "learning_rate": 9.027448158721898e-06,
+      "loss": 0.8631,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3964123084789426,
+      "learning_rate": 9.02667807288291e-06,
+      "loss": 0.7256,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5049282969885376,
+      "learning_rate": 9.02590771515341e-06,
+      "loss": 0.8073,
+      "step": 2817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.587664784886817,
+      "learning_rate": 9.025137085585417e-06,
+      "loss": 0.7709,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7011584202925614,
+      "learning_rate": 9.024366184230964e-06,
+      "loss": 0.8018,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6061428859183073,
+      "learning_rate": 9.023595011142103e-06,
+      "loss": 0.7664,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4977607515454034,
+      "learning_rate": 9.022823566370907e-06,
+      "loss": 0.766,
+      "step": 2821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1587008142601938,
+      "learning_rate": 9.022051849969465e-06,
+      "loss": 1.1481,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6819304422761396,
+      "learning_rate": 9.021279861989884e-06,
+      "loss": 0.7757,
+      "step": 2823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5162379417284817,
+      "learning_rate": 9.020507602484293e-06,
+      "loss": 0.8464,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5127299946980448,
+      "learning_rate": 9.019735071504836e-06,
+      "loss": 0.7935,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.573910478049113,
+      "learning_rate": 9.018962269103672e-06,
+      "loss": 0.8129,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.670076779943516,
+      "learning_rate": 9.01818919533299e-06,
+      "loss": 0.8751,
+      "step": 2827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9085267294951515,
+      "learning_rate": 9.017415850244981e-06,
+      "loss": 1.1164,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5336147489826537,
+      "learning_rate": 9.016642233891871e-06,
+      "loss": 0.7139,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4593788940364583,
+      "learning_rate": 9.01586834632589e-06,
+      "loss": 0.78,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4471877723709619,
+      "learning_rate": 9.015094187599297e-06,
+      "loss": 0.7492,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5116541442908384,
+      "learning_rate": 9.014319757764363e-06,
+      "loss": 0.7715,
+      "step": 2832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5104326982508283,
+      "learning_rate": 9.01354505687338e-06,
+      "loss": 0.8167,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.42495266300129,
+      "learning_rate": 9.012770084978657e-06,
+      "loss": 0.8549,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.502935100063998,
+      "learning_rate": 9.01199484213252e-06,
+      "loss": 0.8248,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5144688366466426,
+      "learning_rate": 9.011219328387321e-06,
+      "loss": 0.8722,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5362925103899885,
+      "learning_rate": 9.01044354379542e-06,
+      "loss": 0.8159,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4732109085517446,
+      "learning_rate": 9.009667488409197e-06,
+      "loss": 0.8432,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4754796042898999,
+      "learning_rate": 9.00889116228106e-06,
+      "loss": 0.7628,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.526808344550726,
+      "learning_rate": 9.008114565463423e-06,
+      "loss": 0.7684,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.522599896003508,
+      "learning_rate": 9.007337698008727e-06,
+      "loss": 0.823,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5056311162585918,
+      "learning_rate": 9.006560559969426e-06,
+      "loss": 0.816,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6964257928803927,
+      "learning_rate": 9.005783151397994e-06,
+      "loss": 0.8421,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4981445357809546,
+      "learning_rate": 9.005005472346923e-06,
+      "loss": 0.8344,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3884495515980424,
+      "learning_rate": 9.004227522868727e-06,
+      "loss": 0.8147,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8843444148268373,
+      "learning_rate": 9.00344930301593e-06,
+      "loss": 0.8407,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5678728420608044,
+      "learning_rate": 9.002670812841082e-06,
+      "loss": 0.7449,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9752374374918128,
+      "learning_rate": 9.001892052396749e-06,
+      "loss": 1.1303,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5124312773089488,
+      "learning_rate": 9.001113021735512e-06,
+      "loss": 0.7948,
+      "step": 2849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4996656947373783,
+      "learning_rate": 9.000333720909978e-06,
+      "loss": 0.8367,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4339501536365045,
+      "learning_rate": 8.99955414997276e-06,
+      "loss": 0.8263,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.532967561910912,
+      "learning_rate": 8.998774308976504e-06,
+      "loss": 0.7928,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5456648842661505,
+      "learning_rate": 8.997994197973861e-06,
+      "loss": 0.7563,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4952120070202044,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.8512,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6375715260436081,
+      "learning_rate": 8.996433166160137e-06,
+      "loss": 0.739,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0284091228677428,
+      "learning_rate": 8.99565224545446e-06,
+      "loss": 1.1398,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5105661536045198,
+      "learning_rate": 8.994871054953207e-06,
+      "loss": 0.7857,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8167332722287997,
+      "learning_rate": 8.994089594709126e-06,
+      "loss": 1.1583,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.43041516565146,
+      "learning_rate": 8.993307864774982e-06,
+      "loss": 0.8466,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4617727041295017,
+      "learning_rate": 8.992525865203561e-06,
+      "loss": 0.7674,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5319426321814078,
+      "learning_rate": 8.991743596047664e-06,
+      "loss": 0.8592,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5188445137738937,
+      "learning_rate": 8.990961057360111e-06,
+      "loss": 0.8245,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4475697338145361,
+      "learning_rate": 8.990178249193741e-06,
+      "loss": 0.7811,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6282364554483362,
+      "learning_rate": 8.989395171601413e-06,
+      "loss": 0.8335,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.47295817454312,
+      "learning_rate": 8.988611824636e-06,
+      "loss": 0.7977,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6249577332321916,
+      "learning_rate": 8.987828208350397e-06,
+      "loss": 0.7008,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7083680048264143,
+      "learning_rate": 8.987044322797513e-06,
+      "loss": 0.8047,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4503355637356607,
+      "learning_rate": 8.98626016803028e-06,
+      "loss": 0.7975,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4796865573458986,
+      "learning_rate": 8.985475744101646e-06,
+      "loss": 0.7856,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5363781622816894,
+      "learning_rate": 8.984691051064576e-06,
+      "loss": 0.8611,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4535221825374731,
+      "learning_rate": 8.983906088972055e-06,
+      "loss": 0.8144,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4936031781988626,
+      "learning_rate": 8.983120857877083e-06,
+      "loss": 0.9549,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5637116771563628,
+      "learning_rate": 8.982335357832685e-06,
+      "loss": 0.7642,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5081338973871137,
+      "learning_rate": 8.981549588891899e-06,
+      "loss": 0.8396,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.244200230628726,
+      "learning_rate": 8.980763551107777e-06,
+      "loss": 1.1364,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4450153120398683,
+      "learning_rate": 8.979977244533398e-06,
+      "loss": 0.825,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5323367957733631,
+      "learning_rate": 8.979190669221856e-06,
+      "loss": 0.7297,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8206695022137738,
+      "learning_rate": 8.97840382522626e-06,
+      "loss": 1.1206,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5183910460479892,
+      "learning_rate": 8.97761671259974e-06,
+      "loss": 0.8279,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5451305866588778,
+      "learning_rate": 8.976829331395447e-06,
+      "loss": 0.7513,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5366094745026992,
+      "learning_rate": 8.976041681666541e-06,
+      "loss": 0.8564,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1146734014226087,
+      "learning_rate": 8.97525376346621e-06,
+      "loss": 1.1141,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.460531143910605,
+      "learning_rate": 8.974465576847655e-06,
+      "loss": 0.8093,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4695241875788463,
+      "learning_rate": 8.973677121864098e-06,
+      "loss": 0.8166,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4542630723282235,
+      "learning_rate": 8.972888398568772e-06,
+      "loss": 0.7475,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5662744054442388,
+      "learning_rate": 8.972099407014938e-06,
+      "loss": 0.7669,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4332049763724766,
+      "learning_rate": 8.971310147255869e-06,
+      "loss": 0.7681,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8729195579071725,
+      "learning_rate": 8.97052061934486e-06,
+      "loss": 1.131,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6763431100550428,
+      "learning_rate": 8.969730823335217e-06,
+      "loss": 0.8844,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5336994145934675,
+      "learning_rate": 8.968940759280274e-06,
+      "loss": 0.8486,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.552722258259549,
+      "learning_rate": 8.968150427233373e-06,
+      "loss": 0.8232,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4514139487568405,
+      "learning_rate": 8.967359827247882e-06,
+      "loss": 0.7118,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5612137065154816,
+      "learning_rate": 8.966568959377184e-06,
+      "loss": 0.7848,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5741098855247713,
+      "learning_rate": 8.965777823674679e-06,
+      "loss": 0.743,
+      "step": 2894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8646999289568675,
+      "learning_rate": 8.964986420193788e-06,
+      "loss": 1.1224,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.519257630771146,
+      "learning_rate": 8.964194748987948e-06,
+      "loss": 0.7588,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5431681216140927,
+      "learning_rate": 8.963402810110613e-06,
+      "loss": 0.7913,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4554978127172866,
+      "learning_rate": 8.962610603615257e-06,
+      "loss": 0.8418,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4950723359975566,
+      "learning_rate": 8.961818129555372e-06,
+      "loss": 0.823,
+      "step": 2899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8241285507316574,
+      "learning_rate": 8.961025387984468e-06,
+      "loss": 1.1289,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5030712649545925,
+      "learning_rate": 8.960232378956073e-06,
+      "loss": 0.783,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5134209638333844,
+      "learning_rate": 8.95943910252373e-06,
+      "loss": 0.7661,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5249034090377314,
+      "learning_rate": 8.958645558741009e-06,
+      "loss": 0.8948,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8364807925788422,
+      "learning_rate": 8.957851747661483e-06,
+      "loss": 1.1207,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.522809924844445,
+      "learning_rate": 8.95705766933876e-06,
+      "loss": 0.8443,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6197373339341499,
+      "learning_rate": 8.956263323826455e-06,
+      "loss": 0.8336,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5360875766041366,
+      "learning_rate": 8.955468711178202e-06,
+      "loss": 0.839,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5846915170075357,
+      "learning_rate": 8.954673831447658e-06,
+      "loss": 0.7979,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8302922267436699,
+      "learning_rate": 8.953878684688492e-06,
+      "loss": 0.7767,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8526377718667729,
+      "learning_rate": 8.953083270954399e-06,
+      "loss": 1.1601,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5285748971247237,
+      "learning_rate": 8.952287590299084e-06,
+      "loss": 0.8046,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5640481003979552,
+      "learning_rate": 8.951491642776274e-06,
+      "loss": 0.6982,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4980919380346966,
+      "learning_rate": 8.950695428439709e-06,
+      "loss": 0.8115,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4303460991428953,
+      "learning_rate": 8.949898947343158e-06,
+      "loss": 0.7528,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4198484606347106,
+      "learning_rate": 8.949102199540397e-06,
+      "loss": 0.7828,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5273463727456664,
+      "learning_rate": 8.948305185085226e-06,
+      "loss": 0.8576,
+      "step": 2916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.420574115458076,
+      "learning_rate": 8.947507904031459e-06,
+      "loss": 0.7548,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5549575061802778,
+      "learning_rate": 8.946710356432932e-06,
+      "loss": 0.8176,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.528149485400961,
+      "learning_rate": 8.9459125423435e-06,
+      "loss": 0.7949,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3593274795279662,
+      "learning_rate": 8.945114461817028e-06,
+      "loss": 0.7559,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.647780961918954,
+      "learning_rate": 8.944316114907406e-06,
+      "loss": 0.7971,
+      "step": 2921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4032123810216546,
+      "learning_rate": 8.943517501668541e-06,
+      "loss": 0.7622,
+      "step": 2922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9326260243073812,
+      "learning_rate": 8.942718622154359e-06,
+      "loss": 1.12,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.483494740193894,
+      "learning_rate": 8.941919476418797e-06,
+      "loss": 0.7865,
+      "step": 2924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5146962673431117,
+      "learning_rate": 8.94112006451582e-06,
+      "loss": 0.7533,
+      "step": 2925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.608767387774408,
+      "learning_rate": 8.940320386499404e-06,
+      "loss": 0.7943,
+      "step": 2926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4248648006933904,
+      "learning_rate": 8.939520442423544e-06,
+      "loss": 0.7743,
+      "step": 2927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7610036982612236,
+      "learning_rate": 8.938720232342257e-06,
+      "loss": 0.837,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5322794494100196,
+      "learning_rate": 8.937919756309574e-06,
+      "loss": 0.8103,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4591870420357749,
+      "learning_rate": 8.937119014379543e-06,
+      "loss": 0.7395,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5493358479552397,
+      "learning_rate": 8.936318006606236e-06,
+      "loss": 0.8654,
+      "step": 2931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4813590181065424,
+      "learning_rate": 8.935516733043732e-06,
+      "loss": 0.8018,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5402623223914294,
+      "learning_rate": 8.934715193746143e-06,
+      "loss": 0.8155,
+      "step": 2933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4625065491451386,
+      "learning_rate": 8.933913388767584e-06,
+      "loss": 0.8122,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9390063800795907,
+      "learning_rate": 8.9331113181622e-06,
+      "loss": 1.1334,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4362266395687735,
+      "learning_rate": 8.932308981984145e-06,
+      "loss": 0.7872,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4077486756476278,
+      "learning_rate": 8.931506380287595e-06,
+      "loss": 0.8364,
+      "step": 2937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5533559626382858,
+      "learning_rate": 8.930703513126744e-06,
+      "loss": 0.823,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3683578251864288,
+      "learning_rate": 8.929900380555805e-06,
+      "loss": 0.7642,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5279528084297416,
+      "learning_rate": 8.929096982629007e-06,
+      "loss": 0.9554,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8643991625879426,
+      "learning_rate": 8.928293319400595e-06,
+      "loss": 1.1272,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5566570355205518,
+      "learning_rate": 8.927489390924835e-06,
+      "loss": 0.7354,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5306002908298693,
+      "learning_rate": 8.92668519725601e-06,
+      "loss": 0.8362,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6670288179510317,
+      "learning_rate": 8.925880738448425e-06,
+      "loss": 0.8175,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.572457259434509,
+      "learning_rate": 8.925076014556393e-06,
+      "loss": 0.758,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4719993362319428,
+      "learning_rate": 8.924271025634252e-06,
+      "loss": 0.7984,
+      "step": 2946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.506690809600484,
+      "learning_rate": 8.923465771736361e-06,
+      "loss": 0.7719,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9433627297380383,
+      "learning_rate": 8.922660252917088e-06,
+      "loss": 1.1364,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4496464620569396,
+      "learning_rate": 8.921854469230824e-06,
+      "loss": 0.906,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4920173013338027,
+      "learning_rate": 8.92104842073198e-06,
+      "loss": 0.8444,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5143440268447341,
+      "learning_rate": 8.920242107474979e-06,
+      "loss": 0.7679,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6081072525971154,
+      "learning_rate": 8.919435529514269e-06,
+      "loss": 0.9018,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.524108512813696,
+      "learning_rate": 8.918628686904307e-06,
+      "loss": 0.7228,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4601149071338422,
+      "learning_rate": 8.917821579699578e-06,
+      "loss": 0.7925,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4541345272351975,
+      "learning_rate": 8.917014207954576e-06,
+      "loss": 0.8047,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5613969044588754,
+      "learning_rate": 8.916206571723818e-06,
+      "loss": 0.9001,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6230390029193713,
+      "learning_rate": 8.915398671061838e-06,
+      "loss": 0.8646,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6051765761335404,
+      "learning_rate": 8.914590506023187e-06,
+      "loss": 0.9032,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5071683258832949,
+      "learning_rate": 8.913782076662431e-06,
+      "loss": 0.8155,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9246142011389854,
+      "learning_rate": 8.912973383034162e-06,
+      "loss": 1.1546,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.538619690773534,
+      "learning_rate": 8.912164425192983e-06,
+      "loss": 0.9053,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8183437175185185,
+      "learning_rate": 8.911355203193515e-06,
+      "loss": 1.1144,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.56053114487121,
+      "learning_rate": 8.9105457170904e-06,
+      "loss": 0.8362,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.50517743658222,
+      "learning_rate": 8.909735966938297e-06,
+      "loss": 0.7812,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7399138316328289,
+      "learning_rate": 8.908925952791882e-06,
+      "loss": 0.806,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.590748582848573,
+      "learning_rate": 8.908115674705847e-06,
+      "loss": 0.7389,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4896297912536705,
+      "learning_rate": 8.907305132734905e-06,
+      "loss": 0.7729,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6133944724696818,
+      "learning_rate": 8.906494326933788e-06,
+      "loss": 0.8317,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6749848041542281,
+      "learning_rate": 8.905683257357238e-06,
+      "loss": 0.756,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5372234769772013,
+      "learning_rate": 8.904871924060025e-06,
+      "loss": 0.8431,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.058042023031189,
+      "learning_rate": 8.904060327096933e-06,
+      "loss": 1.1399,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.483329686337272,
+      "learning_rate": 8.903248466522758e-06,
+      "loss": 0.7649,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4516533020690223,
+      "learning_rate": 8.902436342392322e-06,
+      "loss": 0.8427,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.513261485303323,
+      "learning_rate": 8.90162395476046e-06,
+      "loss": 0.8222,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5857353270809185,
+      "learning_rate": 8.900811303682028e-06,
+      "loss": 0.8325,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8968545753413047,
+      "learning_rate": 8.899998389211896e-06,
+      "loss": 1.1292,
+      "step": 2976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6459249712875532,
+      "learning_rate": 8.899185211404955e-06,
+      "loss": 0.8091,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8422790087258277,
+      "learning_rate": 8.898371770316113e-06,
+      "loss": 1.0999,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.464973303570124,
+      "learning_rate": 8.897558066000293e-06,
+      "loss": 0.7728,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4976982914844355,
+      "learning_rate": 8.896744098512443e-06,
+      "loss": 0.8615,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.889006453209654,
+      "learning_rate": 8.89592986790752e-06,
+      "loss": 1.1135,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5488523715809996,
+      "learning_rate": 8.895115374240505e-06,
+      "loss": 0.856,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.438919596609975,
+      "learning_rate": 8.894300617566391e-06,
+      "loss": 0.7499,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8228071477693027,
+      "learning_rate": 8.893485597940195e-06,
+      "loss": 1.1247,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8403746411770103,
+      "learning_rate": 8.892670315416948e-06,
+      "loss": 1.1381,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4334007749909583,
+      "learning_rate": 8.8918547700517e-06,
+      "loss": 0.7347,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5359216626745498,
+      "learning_rate": 8.891038961899521e-06,
+      "loss": 0.8595,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.55755050109655,
+      "learning_rate": 8.890222891015492e-06,
+      "loss": 0.7341,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4372568486525628,
+      "learning_rate": 8.889406557454719e-06,
+      "loss": 0.7678,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4425042080151422,
+      "learning_rate": 8.88858996127232e-06,
+      "loss": 0.8285,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6412357001142646,
+      "learning_rate": 8.887773102523436e-06,
+      "loss": 0.8403,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5491729873391995,
+      "learning_rate": 8.88695598126322e-06,
+      "loss": 0.6921,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4173814230040522,
+      "learning_rate": 8.886138597546852e-06,
+      "loss": 0.8,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4708363390605785,
+      "learning_rate": 8.885320951429518e-06,
+      "loss": 0.7975,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4427564359821285,
+      "learning_rate": 8.884503042966428e-06,
+      "loss": 0.8008,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3005155957035575,
+      "learning_rate": 8.883684872212811e-06,
+      "loss": 1.1161,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7032888367121148,
+      "learning_rate": 8.882866439223911e-06,
+      "loss": 0.904,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5012818010571198,
+      "learning_rate": 8.88204774405499e-06,
+      "loss": 0.7822,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5473490477483385,
+      "learning_rate": 8.881228786761329e-06,
+      "loss": 0.8287,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5390447766650186,
+      "learning_rate": 8.880409567398225e-06,
+      "loss": 0.8502,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5919106269453307,
+      "learning_rate": 8.879590086020993e-06,
+      "loss": 0.7651,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.489650344079886,
+      "learning_rate": 8.878770342684968e-06,
+      "loss": 0.7285,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4802837618100906,
+      "learning_rate": 8.8779503374455e-06,
+      "loss": 0.8841,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.532261752868713,
+      "learning_rate": 8.877130070357959e-06,
+      "loss": 0.8358,
+      "step": 3004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4701679081488315,
+      "learning_rate": 8.876309541477729e-06,
+      "loss": 0.6946,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4587358151462773,
+      "learning_rate": 8.875488750860215e-06,
+      "loss": 0.8322,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5035401721848534,
+      "learning_rate": 8.874667698560838e-06,
+      "loss": 0.849,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7040655798194564,
+      "learning_rate": 8.87384638463504e-06,
+      "loss": 0.8313,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6148731969817678,
+      "learning_rate": 8.873024809138272e-06,
+      "loss": 0.7672,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5825485614261268,
+      "learning_rate": 8.872202972126017e-06,
+      "loss": 0.7878,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4177329442997801,
+      "learning_rate": 8.871380873653761e-06,
+      "loss": 0.7676,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5506413058671813,
+      "learning_rate": 8.870558513777017e-06,
+      "loss": 0.7579,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1747591140190825,
+      "learning_rate": 8.869735892551312e-06,
+      "loss": 1.1411,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.486904591611567,
+      "learning_rate": 8.868913010032188e-06,
+      "loss": 0.7397,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.922671595108368,
+      "learning_rate": 8.868089866275214e-06,
+      "loss": 1.1328,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5190575304587202,
+      "learning_rate": 8.867266461335965e-06,
+      "loss": 0.7974,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8318131077061892,
+      "learning_rate": 8.866442795270042e-06,
+      "loss": 1.1451,
+      "step": 3017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5193479836339012,
+      "learning_rate": 8.865618868133061e-06,
+      "loss": 0.8397,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4805724627642425,
+      "learning_rate": 8.864794679980654e-06,
+      "loss": 0.7638,
+      "step": 3019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5588367616511347,
+      "learning_rate": 8.863970230868474e-06,
+      "loss": 0.7872,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5561187498343676,
+      "learning_rate": 8.863145520852187e-06,
+      "loss": 0.879,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.453245090005874,
+      "learning_rate": 8.862320549987483e-06,
+      "loss": 0.8315,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.570835518006856,
+      "learning_rate": 8.861495318330062e-06,
+      "loss": 0.8479,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5504410554323658,
+      "learning_rate": 8.860669825935647e-06,
+      "loss": 0.7705,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5199753891489487,
+      "learning_rate": 8.859844072859978e-06,
+      "loss": 0.8559,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.467600833375324,
+      "learning_rate": 8.85901805915881e-06,
+      "loss": 0.7986,
+      "step": 3026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5560118993196896,
+      "learning_rate": 8.85819178488792e-06,
+      "loss": 0.8421,
+      "step": 3027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5937484811145721,
+      "learning_rate": 8.857365250103098e-06,
+      "loss": 0.7733,
+      "step": 3028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5892924545819431,
+      "learning_rate": 8.856538454860155e-06,
+      "loss": 0.8141,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5374475815983144,
+      "learning_rate": 8.855711399214914e-06,
+      "loss": 0.76,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3147612029278255,
+      "learning_rate": 8.854884083223225e-06,
+      "loss": 0.7527,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2302772796785721,
+      "learning_rate": 8.854056506940949e-06,
+      "loss": 1.0772,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4634001149428653,
+      "learning_rate": 8.853228670423964e-06,
+      "loss": 0.8004,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5847290564817398,
+      "learning_rate": 8.852400573728167e-06,
+      "loss": 0.8523,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6239657056058256,
+      "learning_rate": 8.851572216909475e-06,
+      "loss": 0.7989,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5957565322089509,
+      "learning_rate": 8.850743600023819e-06,
+      "loss": 0.7908,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5944874269421445,
+      "learning_rate": 8.849914723127151e-06,
+      "loss": 0.8908,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5760806610457292,
+      "learning_rate": 8.849085586275437e-06,
+      "loss": 0.8105,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0790843371896945,
+      "learning_rate": 8.848256189524661e-06,
+      "loss": 1.177,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4726795384664406,
+      "learning_rate": 8.84742653293083e-06,
+      "loss": 0.7552,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5225260124999893,
+      "learning_rate": 8.846596616549961e-06,
+      "loss": 0.7989,
+      "step": 3041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.584347284364022,
+      "learning_rate": 8.845766440438093e-06,
+      "loss": 0.7916,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5695851252189608,
+      "learning_rate": 8.84493600465128e-06,
+      "loss": 0.7719,
+      "step": 3043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6307835665556678,
+      "learning_rate": 8.8441053092456e-06,
+      "loss": 0.7825,
+      "step": 3044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5776015152889598,
+      "learning_rate": 8.843274354277134e-06,
+      "loss": 0.7599,
+      "step": 3045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0024409473035796,
+      "learning_rate": 8.842443139801998e-06,
+      "loss": 1.1341,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6633893145129788,
+      "learning_rate": 8.841611665876315e-06,
+      "loss": 0.8601,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.388028630116776,
+      "learning_rate": 8.840779932556227e-06,
+      "loss": 0.795,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1565168002877426,
+      "learning_rate": 8.839947939897896e-06,
+      "loss": 0.7257,
+      "step": 3049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.501958649904521,
+      "learning_rate": 8.839115687957501e-06,
+      "loss": 0.8142,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6605799456087253,
+      "learning_rate": 8.838283176791234e-06,
+      "loss": 0.8389,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5035271508486334,
+      "learning_rate": 8.83745040645531e-06,
+      "loss": 0.8125,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4363544307207956,
+      "learning_rate": 8.83661737700596e-06,
+      "loss": 0.8195,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5461857329576576,
+      "learning_rate": 8.835784088499433e-06,
+      "loss": 0.8445,
+      "step": 3054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.408979375978495,
+      "learning_rate": 8.834950540991992e-06,
+      "loss": 0.731,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4980069818997153,
+      "learning_rate": 8.834116734539922e-06,
+      "loss": 0.8133,
+      "step": 3056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5538525547047286,
+      "learning_rate": 8.833282669199523e-06,
+      "loss": 0.8694,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.475315729472042,
+      "learning_rate": 8.832448345027113e-06,
+      "loss": 0.7662,
+      "step": 3058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6490662389481463,
+      "learning_rate": 8.831613762079026e-06,
+      "loss": 0.9098,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5151682728918048,
+      "learning_rate": 8.830778920411618e-06,
+      "loss": 0.757,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6828258983911344,
+      "learning_rate": 8.829943820081258e-06,
+      "loss": 0.8903,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.953367937432381,
+      "learning_rate": 8.829108461144332e-06,
+      "loss": 1.1628,
+      "step": 3062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.388860181344939,
+      "learning_rate": 8.828272843657246e-06,
+      "loss": 0.784,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4499347514253744,
+      "learning_rate": 8.827436967676426e-06,
+      "loss": 0.7633,
+      "step": 3064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9151956737167994,
+      "learning_rate": 8.826600833258307e-06,
+      "loss": 0.839,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6438766916970986,
+      "learning_rate": 8.825764440459353e-06,
+      "loss": 0.852,
+      "step": 3066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4815984368640527,
+      "learning_rate": 8.824927789336034e-06,
+      "loss": 0.8329,
+      "step": 3067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4581093802469367,
+      "learning_rate": 8.824090879944844e-06,
+      "loss": 0.7263,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.483330024955782,
+      "learning_rate": 8.823253712342295e-06,
+      "loss": 0.8176,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.519211978312961,
+      "learning_rate": 8.822416286584913e-06,
+      "loss": 0.7323,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4531148202258481,
+      "learning_rate": 8.821578602729242e-06,
+      "loss": 0.7102,
+      "step": 3071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9167369172003292,
+      "learning_rate": 8.820740660831845e-06,
+      "loss": 0.8102,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.577211218735741,
+      "learning_rate": 8.819902460949303e-06,
+      "loss": 0.7761,
+      "step": 3073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5224795671610412,
+      "learning_rate": 8.819064003138211e-06,
+      "loss": 0.8299,
+      "step": 3074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5645539076245956,
+      "learning_rate": 8.818225287455186e-06,
+      "loss": 0.8964,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5339144843636363,
+      "learning_rate": 8.817386313956858e-06,
+      "loss": 0.7954,
+      "step": 3076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4418946714374445,
+      "learning_rate": 8.816547082699877e-06,
+      "loss": 0.8264,
+      "step": 3077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5746300795472294,
+      "learning_rate": 8.815707593740909e-06,
+      "loss": 0.8746,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5165452582710521,
+      "learning_rate": 8.81486784713664e-06,
+      "loss": 0.8388,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6990914402266366,
+      "learning_rate": 8.814027842943772e-06,
+      "loss": 0.7952,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.530252697877969,
+      "learning_rate": 8.813187581219021e-06,
+      "loss": 0.8778,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5064189923856053,
+      "learning_rate": 8.812347062019124e-06,
+      "loss": 0.7338,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5038453113582162,
+      "learning_rate": 8.81150628540084e-06,
+      "loss": 0.7733,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4786528947152273,
+      "learning_rate": 8.810665251420932e-06,
+      "loss": 0.7563,
+      "step": 3084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.456033965314379,
+      "learning_rate": 8.809823960136194e-06,
+      "loss": 0.7121,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5447317411971586,
+      "learning_rate": 8.80898241160343e-06,
+      "loss": 0.861,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5719431302333535,
+      "learning_rate": 8.808140605879464e-06,
+      "loss": 0.7639,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5566190109433007,
+      "learning_rate": 8.80729854302114e-06,
+      "loss": 0.77,
+      "step": 3088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5078953340592898,
+      "learning_rate": 8.806456223085308e-06,
+      "loss": 0.8043,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5398601806810954,
+      "learning_rate": 8.80561364612885e-06,
+      "loss": 0.7593,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.002166350760614,
+      "learning_rate": 8.804770812208655e-06,
+      "loss": 1.1402,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5277240104488827,
+      "learning_rate": 8.803927721381635e-06,
+      "loss": 0.7284,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7160364004749231,
+      "learning_rate": 8.803084373704717e-06,
+      "loss": 0.7661,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5762730641103861,
+      "learning_rate": 8.802240769234846e-06,
+      "loss": 0.8729,
+      "step": 3094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8292016282678623,
+      "learning_rate": 8.801396908028985e-06,
+      "loss": 1.1334,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.580406384576486,
+      "learning_rate": 8.800552790144113e-06,
+      "loss": 0.8061,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4374534526434397,
+      "learning_rate": 8.799708415637224e-06,
+      "loss": 0.7546,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.428930016606241,
+      "learning_rate": 8.798863784565335e-06,
+      "loss": 0.7551,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5728350060254503,
+      "learning_rate": 8.798018896985476e-06,
+      "loss": 0.7663,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.466300973178292,
+      "learning_rate": 8.797173752954698e-06,
+      "loss": 0.7085,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4194336338661122,
+      "learning_rate": 8.796328352530063e-06,
+      "loss": 0.7777,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5208385993387847,
+      "learning_rate": 8.795482695768658e-06,
+      "loss": 0.7354,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6558786623908743,
+      "learning_rate": 8.794636782727584e-06,
+      "loss": 0.7354,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.614059411276489,
+      "learning_rate": 8.793790613463956e-06,
+      "loss": 0.7078,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5197712334389075,
+      "learning_rate": 8.792944188034911e-06,
+      "loss": 0.8748,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9062791738171584,
+      "learning_rate": 8.7920975064976e-06,
+      "loss": 1.107,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5613542689771573,
+      "learning_rate": 8.791250568909196e-06,
+      "loss": 0.7876,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5332545434759974,
+      "learning_rate": 8.790403375326883e-06,
+      "loss": 0.7865,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6248168790085138,
+      "learning_rate": 8.789555925807868e-06,
+      "loss": 0.7474,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4877003513099791,
+      "learning_rate": 8.78870822040937e-06,
+      "loss": 0.7364,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4262439894787198,
+      "learning_rate": 8.787860259188633e-06,
+      "loss": 0.8131,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4220815932089819,
+      "learning_rate": 8.787012042202907e-06,
+      "loss": 0.8666,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.915290199089687,
+      "learning_rate": 8.786163569509468e-06,
+      "loss": 0.8077,
+      "step": 3113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5940290283863148,
+      "learning_rate": 8.785314841165608e-06,
+      "loss": 0.8163,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6472773368774871,
+      "learning_rate": 8.784465857228635e-06,
+      "loss": 0.8714,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.598673219845349,
+      "learning_rate": 8.783616617755872e-06,
+      "loss": 0.7387,
+      "step": 3116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.454492943891432,
+      "learning_rate": 8.782767122804664e-06,
+      "loss": 0.722,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.396045599365883,
+      "learning_rate": 8.781917372432372e-06,
+      "loss": 0.7535,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.605447466359176,
+      "learning_rate": 8.781067366696368e-06,
+      "loss": 0.8544,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4289127922821865,
+      "learning_rate": 8.780217105654053e-06,
+      "loss": 0.773,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9410023276919095,
+      "learning_rate": 8.779366589362832e-06,
+      "loss": 1.0767,
+      "step": 3121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4740690962377336,
+      "learning_rate": 8.778515817880138e-06,
+      "loss": 0.829,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7967490513297111,
+      "learning_rate": 8.777664791263414e-06,
+      "loss": 1.1135,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4140678156614672,
+      "learning_rate": 8.776813509570128e-06,
+      "loss": 0.7277,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4730568218062938,
+      "learning_rate": 8.775961972857756e-06,
+      "loss": 0.7255,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.500131241714859,
+      "learning_rate": 8.775110181183797e-06,
+      "loss": 0.84,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9183382393714884,
+      "learning_rate": 8.774258134605768e-06,
+      "loss": 1.14,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3513378619113963,
+      "learning_rate": 8.773405833181199e-06,
+      "loss": 0.7812,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8618208668025693,
+      "learning_rate": 8.772553276967639e-06,
+      "loss": 1.1474,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5008291443943045,
+      "learning_rate": 8.771700466022655e-06,
+      "loss": 0.7719,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.627120206523081,
+      "learning_rate": 8.77084740040383e-06,
+      "loss": 0.8561,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.577153124859275,
+      "learning_rate": 8.76999408016877e-06,
+      "loss": 0.7629,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5762420491818898,
+      "learning_rate": 8.769140505375084e-06,
+      "loss": 0.8378,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.52622794783532,
+      "learning_rate": 8.768286676080415e-06,
+      "loss": 0.8286,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4620932149657226,
+      "learning_rate": 8.767432592342413e-06,
+      "loss": 0.8005,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.476515698824882,
+      "learning_rate": 8.766578254218746e-06,
+      "loss": 0.7857,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.582162614342305,
+      "learning_rate": 8.765723661767105e-06,
+      "loss": 0.7861,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4356998058437276,
+      "learning_rate": 8.76486881504519e-06,
+      "loss": 0.7535,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.619427941183286,
+      "learning_rate": 8.764013714110727e-06,
+      "loss": 0.8608,
+      "step": 3139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4884048255448308,
+      "learning_rate": 8.763158359021447e-06,
+      "loss": 0.7166,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4229080578026914,
+      "learning_rate": 8.76230274983511e-06,
+      "loss": 0.7775,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6010770236954148,
+      "learning_rate": 8.761446886609489e-06,
+      "loss": 0.77,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.506324294220358,
+      "learning_rate": 8.760590769402372e-06,
+      "loss": 0.8111,
+      "step": 3143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.662117393319966,
+      "learning_rate": 8.759734398271571e-06,
+      "loss": 0.8316,
+      "step": 3144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4724432755027153,
+      "learning_rate": 8.758877773274903e-06,
+      "loss": 0.7609,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5795335256474001,
+      "learning_rate": 8.758020894470214e-06,
+      "loss": 0.7494,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6171500027536925,
+      "learning_rate": 8.757163761915359e-06,
+      "loss": 0.8917,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3039661982121495,
+      "learning_rate": 8.756306375668217e-06,
+      "loss": 1.1474,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6125911142586806,
+      "learning_rate": 8.75544873578668e-06,
+      "loss": 0.829,
+      "step": 3149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4968126295687911,
+      "learning_rate": 8.754590842328655e-06,
+      "loss": 0.7336,
+      "step": 3150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4179705414085084,
+      "learning_rate": 8.753732695352072e-06,
+      "loss": 0.7244,
+      "step": 3151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5889144611713966,
+      "learning_rate": 8.752874294914874e-06,
+      "loss": 0.7575,
+      "step": 3152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5657062189493105,
+      "learning_rate": 8.75201564107502e-06,
+      "loss": 0.7586,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4289395697778238,
+      "learning_rate": 8.751156733890494e-06,
+      "loss": 0.8244,
+      "step": 3154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4432469869815945,
+      "learning_rate": 8.750297573419286e-06,
+      "loss": 0.7347,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5267339704425509,
+      "learning_rate": 8.74943815971941e-06,
+      "loss": 0.8223,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4963257222084592,
+      "learning_rate": 8.748578492848896e-06,
+      "loss": 0.8414,
+      "step": 3157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5275194786985582,
+      "learning_rate": 8.74771857286579e-06,
+      "loss": 0.768,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5558219075185953,
+      "learning_rate": 8.746858399828158e-06,
+      "loss": 0.804,
+      "step": 3159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6540195989395183,
+      "learning_rate": 8.74599797379408e-06,
+      "loss": 0.7821,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4523496964118205,
+      "learning_rate": 8.74513729482165e-06,
+      "loss": 0.8151,
+      "step": 3161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7423097177071774,
+      "learning_rate": 8.744276362968988e-06,
+      "loss": 0.815,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6984400558775399,
+      "learning_rate": 8.743415178294223e-06,
+      "loss": 0.7092,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.738712537058898,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.8743,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4587834351917808,
+      "learning_rate": 8.741692050711003e-06,
+      "loss": 0.8256,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4955772014900865,
+      "learning_rate": 8.740830107918896e-06,
+      "loss": 0.7875,
+      "step": 3166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2942240755673318,
+      "learning_rate": 8.739967912537385e-06,
+      "loss": 1.1428,
+      "step": 3167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5986785478802823,
+      "learning_rate": 8.739105464624691e-06,
+      "loss": 0.8616,
+      "step": 3168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4262295009261339,
+      "learning_rate": 8.738242764239046e-06,
+      "loss": 0.7875,
+      "step": 3169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4636930831957609,
+      "learning_rate": 8.7373798114387e-06,
+      "loss": 0.8415,
+      "step": 3170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.511452584195006,
+      "learning_rate": 8.736516606281923e-06,
+      "loss": 0.7394,
+      "step": 3171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3697482038855815,
+      "learning_rate": 8.735653148827002e-06,
+      "loss": 0.812,
+      "step": 3172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9471873617273203,
+      "learning_rate": 8.734789439132239e-06,
+      "loss": 1.1085,
+      "step": 3173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5430876954324562,
+      "learning_rate": 8.733925477255952e-06,
+      "loss": 0.7055,
+      "step": 3174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8641466236667026,
+      "learning_rate": 8.733061263256477e-06,
+      "loss": 1.1472,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4672447911631943,
+      "learning_rate": 8.73219679719217e-06,
+      "loss": 0.7309,
+      "step": 3176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6129592790063643,
+      "learning_rate": 8.731332079121403e-06,
+      "loss": 0.8509,
+      "step": 3177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4505144681869895,
+      "learning_rate": 8.73046710910256e-06,
+      "loss": 0.7993,
+      "step": 3178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3671435331109323,
+      "learning_rate": 8.729601887194048e-06,
+      "loss": 0.7271,
+      "step": 3179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4734354803717087,
+      "learning_rate": 8.728736413454288e-06,
+      "loss": 0.7661,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7007357656330926,
+      "learning_rate": 8.727870687941721e-06,
+      "loss": 0.7661,
+      "step": 3181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1917305184937865,
+      "learning_rate": 8.7270047107148e-06,
+      "loss": 1.1249,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5311581719983118,
+      "learning_rate": 8.726138481831997e-06,
+      "loss": 0.8312,
+      "step": 3183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4587146670958966,
+      "learning_rate": 8.725272001351805e-06,
+      "loss": 0.8079,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5218678729933361,
+      "learning_rate": 8.724405269332727e-06,
+      "loss": 0.841,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6471763189073796,
+      "learning_rate": 8.72353828583329e-06,
+      "loss": 0.8412,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6378130392734012,
+      "learning_rate": 8.722671050912034e-06,
+      "loss": 0.7951,
+      "step": 3187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4748806114775252,
+      "learning_rate": 8.721803564627517e-06,
+      "loss": 0.7773,
+      "step": 3188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.493333067803711,
+      "learning_rate": 8.720935827038313e-06,
+      "loss": 0.7301,
+      "step": 3189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6108300477060489,
+      "learning_rate": 8.720067838203013e-06,
+      "loss": 0.838,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.673565279852178,
+      "learning_rate": 8.719199598180224e-06,
+      "loss": 0.8306,
+      "step": 3191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6081104497356187,
+      "learning_rate": 8.718331107028577e-06,
+      "loss": 0.7437,
+      "step": 3192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.431411754960613,
+      "learning_rate": 8.717462364806708e-06,
+      "loss": 0.8064,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5241899220090658,
+      "learning_rate": 8.71659337157328e-06,
+      "loss": 0.8154,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4933202231887361,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.8286,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5393260131463324,
+      "learning_rate": 8.714854632306473e-06,
+      "loss": 0.7991,
+      "step": 3196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5095728143233038,
+      "learning_rate": 8.713984886390494e-06,
+      "loss": 0.7872,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4702508794916604,
+      "learning_rate": 8.713114889697764e-06,
+      "loss": 0.7141,
+      "step": 3198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0300625371145722,
+      "learning_rate": 8.712244642287025e-06,
+      "loss": 1.17,
+      "step": 3199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.558518663911173,
+      "learning_rate": 8.711374144217039e-06,
+      "loss": 0.8305,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5617763900491402,
+      "learning_rate": 8.710503395546584e-06,
+      "loss": 0.8174,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.635960020140099,
+      "learning_rate": 8.709632396334458e-06,
+      "loss": 0.8229,
+      "step": 3202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5516639991039758,
+      "learning_rate": 8.708761146639466e-06,
+      "loss": 0.78,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5653001465380068,
+      "learning_rate": 8.707889646520443e-06,
+      "loss": 0.7724,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.59117217440984,
+      "learning_rate": 8.707017896036232e-06,
+      "loss": 0.8172,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5273871899522644,
+      "learning_rate": 8.706145895245696e-06,
+      "loss": 0.7794,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4838534552581528,
+      "learning_rate": 8.705273644207715e-06,
+      "loss": 0.8388,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.498031921364573,
+      "learning_rate": 8.704401142981184e-06,
+      "loss": 0.8272,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4741911663102372,
+      "learning_rate": 8.70352839162502e-06,
+      "loss": 0.8025,
+      "step": 3209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4906753127407333,
+      "learning_rate": 8.702655390198149e-06,
+      "loss": 0.7577,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8248897409871006,
+      "learning_rate": 8.701782138759517e-06,
+      "loss": 0.8979,
+      "step": 3211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.508039567055376,
+      "learning_rate": 8.700908637368093e-06,
+      "loss": 0.7871,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6189980039812384,
+      "learning_rate": 8.700034886082853e-06,
+      "loss": 0.8996,
+      "step": 3213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5364015859862488,
+      "learning_rate": 8.699160884962798e-06,
+      "loss": 0.8821,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.037103335572283,
+      "learning_rate": 8.69828663406694e-06,
+      "loss": 1.1652,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6090865892499293,
+      "learning_rate": 8.697412133454315e-06,
+      "loss": 0.6963,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8441008215431169,
+      "learning_rate": 8.696537383183967e-06,
+      "loss": 1.1325,
+      "step": 3217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.773011928556298,
+      "learning_rate": 8.695662383314962e-06,
+      "loss": 1.1507,
+      "step": 3218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6681593154390442,
+      "learning_rate": 8.694787133906381e-06,
+      "loss": 0.7869,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.493865781145631,
+      "learning_rate": 8.693911635017324e-06,
+      "loss": 0.798,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6162064367577786,
+      "learning_rate": 8.693035886706909e-06,
+      "loss": 0.8476,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5423695744509542,
+      "learning_rate": 8.692159889034264e-06,
+      "loss": 0.7316,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4207159941020306,
+      "learning_rate": 8.691283642058543e-06,
+      "loss": 0.7706,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6641163092071336,
+      "learning_rate": 8.690407145838909e-06,
+      "loss": 0.7993,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6420973222144537,
+      "learning_rate": 8.689530400434545e-06,
+      "loss": 0.7752,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.508623944742499,
+      "learning_rate": 8.688653405904653e-06,
+      "loss": 0.7778,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5120302182830403,
+      "learning_rate": 8.687776162308446e-06,
+      "loss": 0.7865,
+      "step": 3227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4806291251170662,
+      "learning_rate": 8.68689866970516e-06,
+      "loss": 0.7787,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3786908080292355,
+      "learning_rate": 8.686020928154049e-06,
+      "loss": 0.6899,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.525832885340039,
+      "learning_rate": 8.685142937714374e-06,
+      "loss": 0.7991,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.432934754694863,
+      "learning_rate": 8.68426469844542e-06,
+      "loss": 0.8296,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5074054848302954,
+      "learning_rate": 8.683386210406491e-06,
+      "loss": 0.7621,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4955758186378982,
+      "learning_rate": 8.682507473656902e-06,
+      "loss": 0.8077,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3934161470577267,
+      "learning_rate": 8.681628488255986e-06,
+      "loss": 0.8703,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5902481856323938,
+      "learning_rate": 8.6807492542631e-06,
+      "loss": 0.8324,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.475185764466494,
+      "learning_rate": 8.679869771737603e-06,
+      "loss": 0.8177,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6149222399852938,
+      "learning_rate": 8.678990040738889e-06,
+      "loss": 1.1315,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4361954373475614,
+      "learning_rate": 8.678110061326352e-06,
+      "loss": 0.8192,
+      "step": 3238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4976554128923416,
+      "learning_rate": 8.677229833559413e-06,
+      "loss": 0.7844,
+      "step": 3239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.386714041517101,
+      "learning_rate": 8.676349357497509e-06,
+      "loss": 0.7348,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.449819636447611,
+      "learning_rate": 8.675468633200089e-06,
+      "loss": 0.7878,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.579395293652694,
+      "learning_rate": 8.674587660726622e-06,
+      "loss": 0.8447,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5419875854975602,
+      "learning_rate": 8.673706440136594e-06,
+      "loss": 0.8783,
+      "step": 3243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.637591092284609,
+      "learning_rate": 8.672824971489506e-06,
+      "loss": 0.8235,
+      "step": 3244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9659991722211284,
+      "learning_rate": 8.671943254844878e-06,
+      "loss": 1.1071,
+      "step": 3245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5421289501177977,
+      "learning_rate": 8.671061290262245e-06,
+      "loss": 0.796,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4912361274572705,
+      "learning_rate": 8.67017907780116e-06,
+      "loss": 0.8056,
+      "step": 3247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4846199526850645,
+      "learning_rate": 8.669296617521192e-06,
+      "loss": 0.8961,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5866265827751722,
+      "learning_rate": 8.668413909481926e-06,
+      "loss": 0.7841,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4252790469194703,
+      "learning_rate": 8.667530953742963e-06,
+      "loss": 0.8585,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6286266203177167,
+      "learning_rate": 8.666647750363924e-06,
+      "loss": 0.7844,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9166121246289527,
+      "learning_rate": 8.665764299404445e-06,
+      "loss": 1.0988,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4103099433099304,
+      "learning_rate": 8.664880600924176e-06,
+      "loss": 0.6941,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5743431752403882,
+      "learning_rate": 8.663996654982791e-06,
+      "loss": 0.8149,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4967822484595512,
+      "learning_rate": 8.663112461639973e-06,
+      "loss": 0.7461,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4234346143247456,
+      "learning_rate": 8.662228020955425e-06,
+      "loss": 0.8249,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.534514075312154,
+      "learning_rate": 8.661343332988869e-06,
+      "loss": 0.7728,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4281755740677766,
+      "learning_rate": 8.660458397800036e-06,
+      "loss": 0.7924,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6291331625770813,
+      "learning_rate": 8.659573215448685e-06,
+      "loss": 0.7514,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4290980487148393,
+      "learning_rate": 8.658687785994579e-06,
+      "loss": 0.7435,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4448832758614776,
+      "learning_rate": 8.65780210949751e-06,
+      "loss": 0.754,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6130430765891297,
+      "learning_rate": 8.656916186017277e-06,
+      "loss": 0.8313,
+      "step": 3262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5186555830329866,
+      "learning_rate": 8.6560300156137e-06,
+      "loss": 0.7451,
+      "step": 3263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4975999552448578,
+      "learning_rate": 8.65514359834662e-06,
+      "loss": 0.8002,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5020589058439096,
+      "learning_rate": 8.654256934275885e-06,
+      "loss": 0.8492,
+      "step": 3265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4192390282247898,
+      "learning_rate": 8.653370023461365e-06,
+      "loss": 0.7147,
+      "step": 3266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5980151963507347,
+      "learning_rate": 8.652482865962947e-06,
+      "loss": 0.7904,
+      "step": 3267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4830254593857364,
+      "learning_rate": 8.651595461840533e-06,
+      "loss": 0.7688,
+      "step": 3268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6090668563157975,
+      "learning_rate": 8.650707811154045e-06,
+      "loss": 0.7963,
+      "step": 3269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4473405594764457,
+      "learning_rate": 8.649819913963417e-06,
+      "loss": 0.8255,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9740902067347231,
+      "learning_rate": 8.648931770328604e-06,
+      "loss": 1.1647,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4740074637914191,
+      "learning_rate": 8.648043380309574e-06,
+      "loss": 0.7504,
+      "step": 3272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5596307573395274,
+      "learning_rate": 8.64715474396631e-06,
+      "loss": 0.8563,
+      "step": 3273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.544592782802664,
+      "learning_rate": 8.646265861358822e-06,
+      "loss": 0.8511,
+      "step": 3274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5331966829051096,
+      "learning_rate": 8.645376732547123e-06,
+      "loss": 0.8366,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5414292735602828,
+      "learning_rate": 8.644487357591252e-06,
+      "loss": 0.8417,
+      "step": 3276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5528264214801235,
+      "learning_rate": 8.643597736551262e-06,
+      "loss": 0.7625,
+      "step": 3277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.638042925537915,
+      "learning_rate": 8.642707869487218e-06,
+      "loss": 0.8559,
+      "step": 3278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6497572837647958,
+      "learning_rate": 8.641817756459212e-06,
+      "loss": 0.8811,
+      "step": 3279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.539790742782902,
+      "learning_rate": 8.640927397527344e-06,
+      "loss": 0.7133,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.470068350701348,
+      "learning_rate": 8.64003679275173e-06,
+      "loss": 0.7815,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5574091335189058,
+      "learning_rate": 8.639145942192511e-06,
+      "loss": 0.8128,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6386956816701577,
+      "learning_rate": 8.638254845909837e-06,
+      "loss": 0.8262,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8422347761236193,
+      "learning_rate": 8.637363503963873e-06,
+      "loss": 0.7746,
+      "step": 3284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4562055914037269,
+      "learning_rate": 8.63647191641481e-06,
+      "loss": 0.7768,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5456377326106159,
+      "learning_rate": 8.635580083322847e-06,
+      "loss": 0.7835,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5623340144891305,
+      "learning_rate": 8.634688004748205e-06,
+      "loss": 0.815,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4584696798781007,
+      "learning_rate": 8.633795680751116e-06,
+      "loss": 0.7796,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7236563241736043,
+      "learning_rate": 8.632903111391836e-06,
+      "loss": 0.8918,
+      "step": 3289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.532946154438956,
+      "learning_rate": 8.63201029673063e-06,
+      "loss": 0.8634,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5557272501066854,
+      "learning_rate": 8.631117236827782e-06,
+      "loss": 0.7815,
+      "step": 3291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9830316941327243,
+      "learning_rate": 8.630223931743595e-06,
+      "loss": 1.1333,
+      "step": 3292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.459604712604364,
+      "learning_rate": 8.629330381538387e-06,
+      "loss": 0.8041,
+      "step": 3293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5574552516033986,
+      "learning_rate": 8.628436586272495e-06,
+      "loss": 0.7798,
+      "step": 3294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6841527603693505,
+      "learning_rate": 8.627542546006267e-06,
+      "loss": 0.7965,
+      "step": 3295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7241445619099565,
+      "learning_rate": 8.62664826080007e-06,
+      "loss": 0.8206,
+      "step": 3296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.510372082434077,
+      "learning_rate": 8.62575373071429e-06,
+      "loss": 0.7973,
+      "step": 3297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.969191801869946,
+      "learning_rate": 8.624858955809328e-06,
+      "loss": 1.1461,
+      "step": 3298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.418181611038574,
+      "learning_rate": 8.6239639361456e-06,
+      "loss": 0.7936,
+      "step": 3299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5776373093021492,
+      "learning_rate": 8.623068671783541e-06,
+      "loss": 0.784,
+      "step": 3300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8206280854290395,
+      "learning_rate": 8.6221731627836e-06,
+      "loss": 0.8269,
+      "step": 3301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4997731941423187,
+      "learning_rate": 8.621277409206245e-06,
+      "loss": 0.8351,
+      "step": 3302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7141788296684661,
+      "learning_rate": 8.620381411111958e-06,
+      "loss": 0.8194,
+      "step": 3303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.466053773915491,
+      "learning_rate": 8.619485168561242e-06,
+      "loss": 0.8282,
+      "step": 3304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.516533910548092,
+      "learning_rate": 8.618588681614609e-06,
+      "loss": 0.7827,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.67325491738838,
+      "learning_rate": 8.617691950332592e-06,
+      "loss": 0.8475,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4403451844931572,
+      "learning_rate": 8.616794974775747e-06,
+      "loss": 0.7821,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5001917106981668,
+      "learning_rate": 8.61589775500463e-06,
+      "loss": 0.8034,
+      "step": 3308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5463256348331216,
+      "learning_rate": 8.615000291079831e-06,
+      "loss": 0.8082,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8959006218315122,
+      "learning_rate": 8.614102583061944e-06,
+      "loss": 1.135,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4623630899341653,
+      "learning_rate": 8.613204631011589e-06,
+      "loss": 0.8329,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.559672996872607,
+      "learning_rate": 8.612306434989395e-06,
+      "loss": 0.7971,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6591986758736,
+      "learning_rate": 8.61140799505601e-06,
+      "loss": 0.9211,
+      "step": 3313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5673557376589486,
+      "learning_rate": 8.610509311272099e-06,
+      "loss": 0.7973,
+      "step": 3314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8417875708278847,
+      "learning_rate": 8.609610383698343e-06,
+      "loss": 1.1247,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8744058656477753,
+      "learning_rate": 8.608711212395439e-06,
+      "loss": 1.1002,
+      "step": 3316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.487392471507692,
+      "learning_rate": 8.607811797424104e-06,
+      "loss": 0.7715,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4668020218861684,
+      "learning_rate": 8.606912138845066e-06,
+      "loss": 0.7071,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5033294157885893,
+      "learning_rate": 8.606012236719073e-06,
+      "loss": 0.8406,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.530815028663454,
+      "learning_rate": 8.605112091106889e-06,
+      "loss": 0.7378,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4524292550628113,
+      "learning_rate": 8.604211702069292e-06,
+      "loss": 0.7956,
+      "step": 3321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.529002899155621,
+      "learning_rate": 8.603311069667079e-06,
+      "loss": 0.8017,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4287958175045858,
+      "learning_rate": 8.602410193961063e-06,
+      "loss": 0.7947,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.449101707758444,
+      "learning_rate": 8.601509075012074e-06,
+      "loss": 0.7566,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.557832933409542,
+      "learning_rate": 8.600607712880956e-06,
+      "loss": 0.7678,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.544886823648102,
+      "learning_rate": 8.599706107628573e-06,
+      "loss": 0.7955,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.524679372208119,
+      "learning_rate": 8.598804259315802e-06,
+      "loss": 0.7735,
+      "step": 3327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4340673842201674,
+      "learning_rate": 8.597902168003539e-06,
+      "loss": 0.8678,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5045437336409346,
+      "learning_rate": 8.596999833752694e-06,
+      "loss": 0.8518,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5387139189310655,
+      "learning_rate": 8.596097256624194e-06,
+      "loss": 0.6806,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.022464173225089,
+      "learning_rate": 8.595194436678983e-06,
+      "loss": 1.1212,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4811281559434066,
+      "learning_rate": 8.594291373978028e-06,
+      "loss": 0.7287,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6015363750156646,
+      "learning_rate": 8.593388068582296e-06,
+      "loss": 0.784,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.409733960272055,
+      "learning_rate": 8.592484520552786e-06,
+      "loss": 0.7992,
+      "step": 3334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6275956443650899,
+      "learning_rate": 8.591580729950506e-06,
+      "loss": 0.8182,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6036253434041168,
+      "learning_rate": 8.590676696836484e-06,
+      "loss": 0.8481,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.489228781790211,
+      "learning_rate": 8.58977242127176e-06,
+      "loss": 0.8007,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.47290304889642,
+      "learning_rate": 8.588867903317395e-06,
+      "loss": 0.848,
+      "step": 3338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5907501429595872,
+      "learning_rate": 8.587963143034461e-06,
+      "loss": 0.789,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.532781583802185,
+      "learning_rate": 8.587058140484051e-06,
+      "loss": 0.7716,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5574053591718162,
+      "learning_rate": 8.586152895727273e-06,
+      "loss": 0.787,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9373462971721662,
+      "learning_rate": 8.585247408825252e-06,
+      "loss": 1.1224,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4838439580441407,
+      "learning_rate": 8.584341679839129e-06,
+      "loss": 0.8445,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5951994290541198,
+      "learning_rate": 8.583435708830058e-06,
+      "loss": 0.8087,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.820250645930716,
+      "learning_rate": 8.582529495859214e-06,
+      "loss": 1.1737,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5576241353849183,
+      "learning_rate": 8.581623040987788e-06,
+      "loss": 0.7524,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4917543966721278,
+      "learning_rate": 8.580716344276983e-06,
+      "loss": 0.7369,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3848183813559427,
+      "learning_rate": 8.579809405788022e-06,
+      "loss": 0.7311,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.957568144174073,
+      "learning_rate": 8.578902225582145e-06,
+      "loss": 1.1309,
+      "step": 3349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5721491671945418,
+      "learning_rate": 8.577994803720605e-06,
+      "loss": 0.8221,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.645263961177014,
+      "learning_rate": 8.577087140264677e-06,
+      "loss": 0.7473,
+      "step": 3351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8124226806758676,
+      "learning_rate": 8.576179235275643e-06,
+      "loss": 1.164,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4471767431748168,
+      "learning_rate": 8.575271088814811e-06,
+      "loss": 0.8124,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5610849132602493,
+      "learning_rate": 8.574362700943501e-06,
+      "loss": 0.766,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8707505313602819,
+      "learning_rate": 8.573454071723046e-06,
+      "loss": 1.1145,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4848332186577673,
+      "learning_rate": 8.572545201214802e-06,
+      "loss": 0.7562,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4578253786173254,
+      "learning_rate": 8.571636089480135e-06,
+      "loss": 0.7685,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5857738878691185,
+      "learning_rate": 8.570726736580434e-06,
+      "loss": 0.8397,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5672089498113106,
+      "learning_rate": 8.569817142577099e-06,
+      "loss": 0.8157,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4951836297050192,
+      "learning_rate": 8.568907307531547e-06,
+      "loss": 0.7282,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5300229562060224,
+      "learning_rate": 8.567997231505213e-06,
+      "loss": 0.84,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5370789855626543,
+      "learning_rate": 8.567086914559545e-06,
+      "loss": 0.7204,
+      "step": 3362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6068788413868187,
+      "learning_rate": 8.566176356756015e-06,
+      "loss": 0.8576,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3226262288558737,
+      "learning_rate": 8.565265558156101e-06,
+      "loss": 0.7503,
+      "step": 3364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7318498795543795,
+      "learning_rate": 8.564354518821307e-06,
+      "loss": 0.8925,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.458387703398481,
+      "learning_rate": 8.563443238813143e-06,
+      "loss": 0.8151,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.476384808159295,
+      "learning_rate": 8.562531718193144e-06,
+      "loss": 0.798,
+      "step": 3367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4200510801235173,
+      "learning_rate": 8.561619957022855e-06,
+      "loss": 0.8354,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6055330033170263,
+      "learning_rate": 8.560707955363845e-06,
+      "loss": 0.6985,
+      "step": 3369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5712442771062076,
+      "learning_rate": 8.55979571327769e-06,
+      "loss": 0.8828,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8773848608317171,
+      "learning_rate": 8.55888323082599e-06,
+      "loss": 1.153,
+      "step": 3371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4559647381473184,
+      "learning_rate": 8.557970508070356e-06,
+      "loss": 0.7741,
+      "step": 3372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.540014556802582,
+      "learning_rate": 8.557057545072417e-06,
+      "loss": 0.7992,
+      "step": 3373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4765466430303391,
+      "learning_rate": 8.556144341893819e-06,
+      "loss": 0.7934,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5634188876914148,
+      "learning_rate": 8.555230898596223e-06,
+      "loss": 0.8551,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3976782931050098,
+      "learning_rate": 8.554317215241308e-06,
+      "loss": 0.8072,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.855948445387043,
+      "learning_rate": 8.553403291890767e-06,
+      "loss": 1.1528,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4738211328070994,
+      "learning_rate": 8.55248912860631e-06,
+      "loss": 0.7172,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5241101804538542,
+      "learning_rate": 8.551574725449665e-06,
+      "loss": 0.8492,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7740462955708514,
+      "learning_rate": 8.55066008248257e-06,
+      "loss": 1.0972,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.525710534585981,
+      "learning_rate": 8.549745199766792e-06,
+      "loss": 0.8703,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4703983261637346,
+      "learning_rate": 8.548830077364099e-06,
+      "loss": 0.7253,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.537241642957234,
+      "learning_rate": 8.547914715336283e-06,
+      "loss": 0.7437,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6349451210094343,
+      "learning_rate": 8.546999113745153e-06,
+      "loss": 0.8519,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4759516785145512,
+      "learning_rate": 8.546083272652534e-06,
+      "loss": 0.7501,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.39187232412472,
+      "learning_rate": 8.545167192120263e-06,
+      "loss": 0.8106,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.417444057460987,
+      "learning_rate": 8.544250872210196e-06,
+      "loss": 0.7704,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4433882925185795,
+      "learning_rate": 8.543334312984207e-06,
+      "loss": 0.7445,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5801291977494123,
+      "learning_rate": 8.54241751450418e-06,
+      "loss": 0.7383,
+      "step": 3389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7393058092144527,
+      "learning_rate": 8.541500476832025e-06,
+      "loss": 0.7884,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.505755027127155,
+      "learning_rate": 8.540583200029657e-06,
+      "loss": 0.7872,
+      "step": 3391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5296802366700433,
+      "learning_rate": 8.539665684159018e-06,
+      "loss": 0.8106,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5593493525762068,
+      "learning_rate": 8.538747929282058e-06,
+      "loss": 0.8221,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5196211507813802,
+      "learning_rate": 8.537829935460745e-06,
+      "loss": 0.8476,
+      "step": 3394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4180977852685681,
+      "learning_rate": 8.536911702757064e-06,
+      "loss": 0.7786,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5622299541382703,
+      "learning_rate": 8.53599323123302e-06,
+      "loss": 0.8938,
+      "step": 3396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5124413984623433,
+      "learning_rate": 8.535074520950624e-06,
+      "loss": 0.7222,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.464792259427628,
+      "learning_rate": 8.534155571971916e-06,
+      "loss": 0.7424,
+      "step": 3398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4594214473155425,
+      "learning_rate": 8.53323638435894e-06,
+      "loss": 0.8133,
+      "step": 3399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4437549586336162,
+      "learning_rate": 8.532316958173765e-06,
+      "loss": 0.8153,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4276620149813757,
+      "learning_rate": 8.531397293478472e-06,
+      "loss": 0.7581,
+      "step": 3401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3655580442562667,
+      "learning_rate": 8.530477390335158e-06,
+      "loss": 0.7369,
+      "step": 3402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.587041806840878,
+      "learning_rate": 8.52955724880594e-06,
+      "loss": 0.7686,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4730932730858157,
+      "learning_rate": 8.528636868952944e-06,
+      "loss": 0.7589,
+      "step": 3404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5002166227243023,
+      "learning_rate": 8.527716250838318e-06,
+      "loss": 0.807,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.631903669745231,
+      "learning_rate": 8.526795394524224e-06,
+      "loss": 0.8439,
+      "step": 3406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5062431862496068,
+      "learning_rate": 8.525874300072841e-06,
+      "loss": 0.8029,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4256710367421443,
+      "learning_rate": 8.524952967546363e-06,
+      "loss": 0.8186,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.951457407786043,
+      "learning_rate": 8.524031397007e-06,
+      "loss": 1.1074,
+      "step": 3409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8843161066462978,
+      "learning_rate": 8.523109588516978e-06,
+      "loss": 1.1271,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5088167767572076,
+      "learning_rate": 8.522187542138541e-06,
+      "loss": 0.667,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.380443180521383,
+      "learning_rate": 8.521265257933948e-06,
+      "loss": 0.7679,
+      "step": 3412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.406970901721928,
+      "learning_rate": 8.520342735965473e-06,
+      "loss": 0.7952,
+      "step": 3413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5451141772027912,
+      "learning_rate": 8.519419976295405e-06,
+      "loss": 0.8545,
+      "step": 3414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5123450844757678,
+      "learning_rate": 8.518496978986054e-06,
+      "loss": 0.7831,
+      "step": 3415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4847687123430433,
+      "learning_rate": 8.51757374409974e-06,
+      "loss": 0.7988,
+      "step": 3416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4587887083993443,
+      "learning_rate": 8.516650271698805e-06,
+      "loss": 0.7418,
+      "step": 3417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4945521603228082,
+      "learning_rate": 8.515726561845602e-06,
+      "loss": 0.7315,
+      "step": 3418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8212989471174916,
+      "learning_rate": 8.514802614602503e-06,
+      "loss": 0.7959,
+      "step": 3419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4408393407702618,
+      "learning_rate": 8.513878430031891e-06,
+      "loss": 0.7948,
+      "step": 3420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.531375938828622,
+      "learning_rate": 8.512954008196178e-06,
+      "loss": 0.7797,
+      "step": 3421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.441120882442314,
+      "learning_rate": 8.512029349157774e-06,
+      "loss": 0.7907,
+      "step": 3422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5701273432760572,
+      "learning_rate": 8.511104452979117e-06,
+      "loss": 0.8191,
+      "step": 3423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.476361827817332,
+      "learning_rate": 8.51017931972266e-06,
+      "loss": 0.798,
+      "step": 3424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.48300995102003,
+      "learning_rate": 8.509253949450869e-06,
+      "loss": 0.7975,
+      "step": 3425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.506361942406405,
+      "learning_rate": 8.508328342226226e-06,
+      "loss": 0.7637,
+      "step": 3426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5092718903679876,
+      "learning_rate": 8.507402498111233e-06,
+      "loss": 0.7495,
+      "step": 3427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5446469225809485,
+      "learning_rate": 8.5064764171684e-06,
+      "loss": 0.7864,
+      "step": 3428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2295115963659597,
+      "learning_rate": 8.505550099460264e-06,
+      "loss": 1.1197,
+      "step": 3429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.500921275855427,
+      "learning_rate": 8.504623545049369e-06,
+      "loss": 0.7815,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8162125818291612,
+      "learning_rate": 8.503696753998277e-06,
+      "loss": 0.7315,
+      "step": 3431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8681609389355085,
+      "learning_rate": 8.50276972636957e-06,
+      "loss": 1.1213,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.525897148028584,
+      "learning_rate": 8.50184246222584e-06,
+      "loss": 0.8357,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4638296565314717,
+      "learning_rate": 8.5009149616297e-06,
+      "loss": 0.722,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5685386288975545,
+      "learning_rate": 8.499987224643777e-06,
+      "loss": 0.7839,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5725060553161119,
+      "learning_rate": 8.499059251330714e-06,
+      "loss": 0.7893,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0302495374238567,
+      "learning_rate": 8.498131041753168e-06,
+      "loss": 1.111,
+      "step": 3437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5633207609663466,
+      "learning_rate": 8.497202595973818e-06,
+      "loss": 0.771,
+      "step": 3438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5138126748100817,
+      "learning_rate": 8.496273914055347e-06,
+      "loss": 0.7697,
+      "step": 3439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4737362054429548,
+      "learning_rate": 8.495344996060471e-06,
+      "loss": 0.7434,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.505879076918591,
+      "learning_rate": 8.494415842051905e-06,
+      "loss": 0.7607,
+      "step": 3441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5678052691514186,
+      "learning_rate": 8.493486452092391e-06,
+      "loss": 0.7781,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.39825468555444,
+      "learning_rate": 8.492556826244687e-06,
+      "loss": 0.7937,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6111135686172926,
+      "learning_rate": 8.491626964571555e-06,
+      "loss": 0.8077,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6753891428526013,
+      "learning_rate": 8.490696867135791e-06,
+      "loss": 0.8171,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9493230391663534,
+      "learning_rate": 8.48976653400019e-06,
+      "loss": 1.1369,
+      "step": 3446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.454294987375585,
+      "learning_rate": 8.488835965227572e-06,
+      "loss": 0.811,
+      "step": 3447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.432085862134217,
+      "learning_rate": 8.487905160880773e-06,
+      "loss": 0.8602,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.56887584892502,
+      "learning_rate": 8.486974121022642e-06,
+      "loss": 0.8661,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6628760286502084,
+      "learning_rate": 8.486042845716046e-06,
+      "loss": 0.8263,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5241911127270555,
+      "learning_rate": 8.485111335023865e-06,
+      "loss": 0.8099,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4524584119119777,
+      "learning_rate": 8.484179589008997e-06,
+      "loss": 0.7805,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5605734684782737,
+      "learning_rate": 8.483247607734355e-06,
+      "loss": 0.7973,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.592705945556069,
+      "learning_rate": 8.482315391262871e-06,
+      "loss": 0.7875,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.710011480061702,
+      "learning_rate": 8.48138293965749e-06,
+      "loss": 0.814,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4298347304993668,
+      "learning_rate": 8.48045025298117e-06,
+      "loss": 0.7317,
+      "step": 3456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6073738243183575,
+      "learning_rate": 8.479517331296892e-06,
+      "loss": 0.8571,
+      "step": 3457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6231324311064255,
+      "learning_rate": 8.478584174667647e-06,
+      "loss": 0.8098,
+      "step": 3458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6106405716226526,
+      "learning_rate": 8.477650783156443e-06,
+      "loss": 0.7767,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.546614468783573,
+      "learning_rate": 8.476717156826308e-06,
+      "loss": 0.8482,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.464839041891109,
+      "learning_rate": 8.475783295740279e-06,
+      "loss": 0.8585,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5322583130892928,
+      "learning_rate": 8.474849199961415e-06,
+      "loss": 0.8599,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.633006437146197,
+      "learning_rate": 8.473914869552787e-06,
+      "loss": 0.8074,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5182978885782994,
+      "learning_rate": 8.472980304577483e-06,
+      "loss": 0.7563,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5402854249575322,
+      "learning_rate": 8.472045505098609e-06,
+      "loss": 0.8253,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5906717967773234,
+      "learning_rate": 8.471110471179282e-06,
+      "loss": 0.8463,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.43541290733252,
+      "learning_rate": 8.470175202882638e-06,
+      "loss": 0.7544,
+      "step": 3467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4574526576487707,
+      "learning_rate": 8.46923970027183e-06,
+      "loss": 0.8165,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4824488000759777,
+      "learning_rate": 8.468303963410026e-06,
+      "loss": 0.7668,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5487400852361024,
+      "learning_rate": 8.467367992360405e-06,
+      "loss": 0.8247,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4864660435796813,
+      "learning_rate": 8.466431787186169e-06,
+      "loss": 0.7931,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.633947319527433,
+      "learning_rate": 8.465495347950533e-06,
+      "loss": 0.7432,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4657784071170146,
+      "learning_rate": 8.464558674716727e-06,
+      "loss": 0.7303,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4198597163112592,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.8256,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.458079065485533,
+      "learning_rate": 8.462684626507605e-06,
+      "loss": 0.7275,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8303178944336082,
+      "learning_rate": 8.46174725165883e-06,
+      "loss": 1.1589,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4735962862355394,
+      "learning_rate": 8.460809643064964e-06,
+      "loss": 0.8573,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4957794112675198,
+      "learning_rate": 8.459871800789318e-06,
+      "loss": 0.8366,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4504628591376099,
+      "learning_rate": 8.458933724895216e-06,
+      "loss": 0.6875,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6075967174580625,
+      "learning_rate": 8.457995415445999e-06,
+      "loss": 0.8055,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.513486256775826,
+      "learning_rate": 8.457056872505024e-06,
+      "loss": 0.7449,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4758283156072165,
+      "learning_rate": 8.456118096135666e-06,
+      "loss": 0.724,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5223218280837805,
+      "learning_rate": 8.455179086401309e-06,
+      "loss": 0.8106,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.485995796894997,
+      "learning_rate": 8.45423984336536e-06,
+      "loss": 0.778,
+      "step": 3484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8419018710485615,
+      "learning_rate": 8.45330036709124e-06,
+      "loss": 1.1431,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6026072859450107,
+      "learning_rate": 8.45236065764238e-06,
+      "loss": 0.8896,
+      "step": 3486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5634998589283264,
+      "learning_rate": 8.451420715082236e-06,
+      "loss": 0.7729,
+      "step": 3487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4947106466170899,
+      "learning_rate": 8.450480539474271e-06,
+      "loss": 0.8232,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8979939694605681,
+      "learning_rate": 8.449540130881973e-06,
+      "loss": 1.1267,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8296454263871952,
+      "learning_rate": 8.448599489368836e-06,
+      "loss": 1.0815,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4215036315216873,
+      "learning_rate": 8.447658614998375e-06,
+      "loss": 0.7776,
+      "step": 3491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6191229442275317,
+      "learning_rate": 8.44671750783412e-06,
+      "loss": 0.7631,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.506951805456231,
+      "learning_rate": 8.44577616793962e-06,
+      "loss": 0.8294,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3079219756502056,
+      "learning_rate": 8.444834595378434e-06,
+      "loss": 0.8114,
+      "step": 3494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6064680685914556,
+      "learning_rate": 8.443892790214138e-06,
+      "loss": 0.8644,
+      "step": 3495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.506293576796927,
+      "learning_rate": 8.442950752510327e-06,
+      "loss": 0.7577,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.580818464088887,
+      "learning_rate": 8.442008482330606e-06,
+      "loss": 0.7479,
+      "step": 3497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.540480897160856,
+      "learning_rate": 8.441065979738602e-06,
+      "loss": 0.7803,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.131553039788583,
+      "learning_rate": 8.440123244797955e-06,
+      "loss": 1.15,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4701165735778428,
+      "learning_rate": 8.439180277572321e-06,
+      "loss": 0.9115,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4628975921374012,
+      "learning_rate": 8.43823707812537e-06,
+      "loss": 0.8291,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4760332594761176,
+      "learning_rate": 8.43729364652079e-06,
+      "loss": 0.813,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4294463676007427,
+      "learning_rate": 8.436349982822283e-06,
+      "loss": 0.703,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4729042395678342,
+      "learning_rate": 8.435406087093568e-06,
+      "loss": 0.8349,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5006937057029421,
+      "learning_rate": 8.434461959398377e-06,
+      "loss": 0.718,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.581317936486802,
+      "learning_rate": 8.433517599800462e-06,
+      "loss": 0.7641,
+      "step": 3506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.476405171792754,
+      "learning_rate": 8.432573008363587e-06,
+      "loss": 0.7607,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4515141713591286,
+      "learning_rate": 8.431628185151535e-06,
+      "loss": 0.7731,
+      "step": 3508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9724372226062912,
+      "learning_rate": 8.4306831302281e-06,
+      "loss": 1.1081,
+      "step": 3509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.643807464200815,
+      "learning_rate": 8.429737843657094e-06,
+      "loss": 0.8588,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5014252190347916,
+      "learning_rate": 8.428792325502347e-06,
+      "loss": 0.7444,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5397738032041766,
+      "learning_rate": 8.427846575827702e-06,
+      "loss": 0.881,
+      "step": 3512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.596683411830558,
+      "learning_rate": 8.426900594697018e-06,
+      "loss": 0.8654,
+      "step": 3513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4053044391216072,
+      "learning_rate": 8.425954382174169e-06,
+      "loss": 0.7386,
+      "step": 3514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.904155826425117,
+      "learning_rate": 8.425007938323049e-06,
+      "loss": 1.1322,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.479788420301487,
+      "learning_rate": 8.424061263207558e-06,
+      "loss": 0.7539,
+      "step": 3516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6236486592876873,
+      "learning_rate": 8.423114356891622e-06,
+      "loss": 0.8188,
+      "step": 3517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.50736812975999,
+      "learning_rate": 8.422167219439177e-06,
+      "loss": 0.8024,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7812072948088002,
+      "learning_rate": 8.421219850914176e-06,
+      "loss": 1.1122,
+      "step": 3519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4550556706801905,
+      "learning_rate": 8.42027225138059e-06,
+      "loss": 0.808,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5363364084480264,
+      "learning_rate": 8.419324420902398e-06,
+      "loss": 0.7876,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5596946058702839,
+      "learning_rate": 8.418376359543604e-06,
+      "loss": 0.8099,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8901033578691886,
+      "learning_rate": 8.417428067368218e-06,
+      "loss": 1.167,
+      "step": 3523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8687796713851912,
+      "learning_rate": 8.416479544440279e-06,
+      "loss": 1.1057,
+      "step": 3524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4373183884333895,
+      "learning_rate": 8.415530790823825e-06,
+      "loss": 0.8549,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5706648415540552,
+      "learning_rate": 8.414581806582925e-06,
+      "loss": 0.7867,
+      "step": 3526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4303687074474138,
+      "learning_rate": 8.413632591781653e-06,
+      "loss": 0.704,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4347332774150747,
+      "learning_rate": 8.412683146484103e-06,
+      "loss": 0.8072,
+      "step": 3528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5063191964151044,
+      "learning_rate": 8.411733470754381e-06,
+      "loss": 0.7593,
+      "step": 3529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4980236286929907,
+      "learning_rate": 8.410783564656614e-06,
+      "loss": 0.7748,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.853791949028391,
+      "learning_rate": 8.409833428254943e-06,
+      "loss": 0.8032,
+      "step": 3531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.564060017551178,
+      "learning_rate": 8.408883061613522e-06,
+      "loss": 0.7603,
+      "step": 3532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5522010164054605,
+      "learning_rate": 8.407932464796521e-06,
+      "loss": 0.7276,
+      "step": 3533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4426414634124267,
+      "learning_rate": 8.406981637868128e-06,
+      "loss": 0.7877,
+      "step": 3534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1141531685475987,
+      "learning_rate": 8.406030580892543e-06,
+      "loss": 1.1123,
+      "step": 3535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0179349867686365,
+      "learning_rate": 8.405079293933986e-06,
+      "loss": 1.1394,
+      "step": 3536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4106271735580664,
+      "learning_rate": 8.40412777705669e-06,
+      "loss": 0.8019,
+      "step": 3537
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5175659904839938,
+      "learning_rate": 8.4031760303249e-06,
+      "loss": 0.8614,
+      "step": 3538
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5886226852418635,
+      "learning_rate": 8.402224053802884e-06,
+      "loss": 0.931,
+      "step": 3539
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.643329232088793,
+      "learning_rate": 8.401271847554919e-06,
+      "loss": 0.7948,
+      "step": 3540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5140155631933534,
+      "learning_rate": 8.400319411645302e-06,
+      "loss": 0.7446,
+      "step": 3541
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4345181748704043,
+      "learning_rate": 8.399366746138345e-06,
+      "loss": 0.7616,
+      "step": 3542
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2308407134282107,
+      "learning_rate": 8.39841385109837e-06,
+      "loss": 1.1507,
+      "step": 3543
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4310635574820163,
+      "learning_rate": 8.397460726589722e-06,
+      "loss": 0.8172,
+      "step": 3544
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4181192013006847,
+      "learning_rate": 8.396507372676754e-06,
+      "loss": 0.8172,
+      "step": 3545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7316412510703956,
+      "learning_rate": 8.395553789423844e-06,
+      "loss": 0.8404,
+      "step": 3546
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5309765868261809,
+      "learning_rate": 8.394599976895378e-06,
+      "loss": 0.7667,
+      "step": 3547
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5015501199279628,
+      "learning_rate": 8.393645935155758e-06,
+      "loss": 0.8442,
+      "step": 3548
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.559935933169414,
+      "learning_rate": 8.392691664269406e-06,
+      "loss": 0.7742,
+      "step": 3549
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3508052585719088,
+      "learning_rate": 8.391737164300755e-06,
+      "loss": 0.8243,
+      "step": 3550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.365643594713128,
+      "learning_rate": 8.390782435314254e-06,
+      "loss": 0.7736,
+      "step": 3551
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5088400479716837,
+      "learning_rate": 8.38982747737437e-06,
+      "loss": 0.8819,
+      "step": 3552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5669348942204682,
+      "learning_rate": 8.388872290545583e-06,
+      "loss": 0.816,
+      "step": 3553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4119942179202158,
+      "learning_rate": 8.38791687489239e-06,
+      "loss": 0.7617,
+      "step": 3554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4970504931134134,
+      "learning_rate": 8.386961230479303e-06,
+      "loss": 0.8679,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4954023228736286,
+      "learning_rate": 8.386005357370848e-06,
+      "loss": 0.8353,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4851758261980823,
+      "learning_rate": 8.38504925563157e-06,
+      "loss": 0.7723,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5461433668204432,
+      "learning_rate": 8.384092925326025e-06,
+      "loss": 0.8318,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9426589337693875,
+      "learning_rate": 8.383136366518788e-06,
+      "loss": 1.1237,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.572001906941129,
+      "learning_rate": 8.382179579274447e-06,
+      "loss": 0.8276,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4903960315900653,
+      "learning_rate": 8.381222563657608e-06,
+      "loss": 0.8106,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.599690232210152,
+      "learning_rate": 8.380265319732887e-06,
+      "loss": 0.8789,
+      "step": 3562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8603297711654769,
+      "learning_rate": 8.379307847564925e-06,
+      "loss": 1.123,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8328562856294598,
+      "learning_rate": 8.378350147218369e-06,
+      "loss": 1.1373,
+      "step": 3564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6198609011793015,
+      "learning_rate": 8.377392218757887e-06,
+      "loss": 0.9007,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3831740250682023,
+      "learning_rate": 8.376434062248158e-06,
+      "loss": 0.7445,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4644786949229776,
+      "learning_rate": 8.375475677753882e-06,
+      "loss": 0.8395,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6201882089361996,
+      "learning_rate": 8.374517065339768e-06,
+      "loss": 0.8287,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8616476228510223,
+      "learning_rate": 8.373558225070546e-06,
+      "loss": 1.1476,
+      "step": 3569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5459815502416623,
+      "learning_rate": 8.37259915701096e-06,
+      "loss": 0.7435,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.514167588862741,
+      "learning_rate": 8.371639861225765e-06,
+      "loss": 0.7614,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4851243519043094,
+      "learning_rate": 8.370680337779737e-06,
+      "loss": 0.7432,
+      "step": 3572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4961695646764002,
+      "learning_rate": 8.369720586737666e-06,
+      "loss": 0.8334,
+      "step": 3573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8450236183428267,
+      "learning_rate": 8.368760608164356e-06,
+      "loss": 1.1196,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8461803096920018,
+      "learning_rate": 8.367800402124626e-06,
+      "loss": 1.1004,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4602588952633153,
+      "learning_rate": 8.366839968683312e-06,
+      "loss": 0.7647,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.596052072987529,
+      "learning_rate": 8.365879307905263e-06,
+      "loss": 0.8549,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5816345499693734,
+      "learning_rate": 8.36491841985535e-06,
+      "loss": 0.8124,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4416977365339427,
+      "learning_rate": 8.363957304598447e-06,
+      "loss": 0.7397,
+      "step": 3579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.68055063469111,
+      "learning_rate": 8.362995962199459e-06,
+      "loss": 0.8005,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5109318470112802,
+      "learning_rate": 8.36203439272329e-06,
+      "loss": 0.8067,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.511182446942989,
+      "learning_rate": 8.36107259623487e-06,
+      "loss": 0.8212,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5953739171847179,
+      "learning_rate": 8.360110572799146e-06,
+      "loss": 0.7268,
+      "step": 3583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4971741027242267,
+      "learning_rate": 8.359148322481073e-06,
+      "loss": 0.7145,
+      "step": 3584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5464124386523794,
+      "learning_rate": 8.358185845345623e-06,
+      "loss": 0.8064,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4304522878117933,
+      "learning_rate": 8.357223141457787e-06,
+      "loss": 0.8149,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5299836144057333,
+      "learning_rate": 8.356260210882565e-06,
+      "loss": 0.8036,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5917946362073667,
+      "learning_rate": 8.355297053684982e-06,
+      "loss": 0.7834,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6361821340320137,
+      "learning_rate": 8.354333669930067e-06,
+      "loss": 0.8206,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5627545794231181,
+      "learning_rate": 8.353370059682873e-06,
+      "loss": 0.8306,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6681872277547753,
+      "learning_rate": 8.352406223008465e-06,
+      "loss": 0.8069,
+      "step": 3591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5213964168428444,
+      "learning_rate": 8.351442159971922e-06,
+      "loss": 0.888,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3666305647624115,
+      "learning_rate": 8.350477870638346e-06,
+      "loss": 0.6951,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9113522660832283,
+      "learning_rate": 8.349513355072836e-06,
+      "loss": 0.7474,
+      "step": 3594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.428919935725367,
+      "learning_rate": 8.348548613340529e-06,
+      "loss": 0.7646,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5463819740131208,
+      "learning_rate": 8.347583645506561e-06,
+      "loss": 0.8222,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9635301694117145,
+      "learning_rate": 8.346618451636092e-06,
+      "loss": 1.1399,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8888381221245981,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 1.1152,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8072268162936169,
+      "learning_rate": 8.344687386046348e-06,
+      "loss": 1.0938,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8359251190535291,
+      "learning_rate": 8.343721514457465e-06,
+      "loss": 1.0949,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9315328323164076,
+      "learning_rate": 8.34275541709286e-06,
+      "loss": 1.116,
+      "step": 3601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.837743510381232,
+      "learning_rate": 8.341789094017766e-06,
+      "loss": 1.1151,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5196189521007437,
+      "learning_rate": 8.340822545297426e-06,
+      "loss": 0.8833,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5240922651798163,
+      "learning_rate": 8.339855770997113e-06,
+      "loss": 0.8213,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6095805330489794,
+      "learning_rate": 8.3388887711821e-06,
+      "loss": 0.8299,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9931913176022844,
+      "learning_rate": 8.337921545917684e-06,
+      "loss": 1.1295,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5530389141451728,
+      "learning_rate": 8.336954095269171e-06,
+      "loss": 0.862,
+      "step": 3607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.521089332276859,
+      "learning_rate": 8.335986419301886e-06,
+      "loss": 0.7409,
+      "step": 3608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9593272193954987,
+      "learning_rate": 8.335018518081171e-06,
+      "loss": 1.1291,
+      "step": 3609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4686296168525876,
+      "learning_rate": 8.33405039167238e-06,
+      "loss": 0.8159,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4283410628535451,
+      "learning_rate": 8.333082040140884e-06,
+      "loss": 0.8233,
+      "step": 3611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4805013166565821,
+      "learning_rate": 8.332113463552065e-06,
+      "loss": 0.7855,
+      "step": 3612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8659744583502094,
+      "learning_rate": 8.331144661971325e-06,
+      "loss": 1.0775,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4191932013812378,
+      "learning_rate": 8.330175635464082e-06,
+      "loss": 0.7432,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5539851337791017,
+      "learning_rate": 8.329206384095765e-06,
+      "loss": 0.7615,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5836001916084503,
+      "learning_rate": 8.328236907931819e-06,
+      "loss": 0.7891,
+      "step": 3616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5472185156655707,
+      "learning_rate": 8.327267207037707e-06,
+      "loss": 0.7804,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.529773598300593,
+      "learning_rate": 8.326297281478906e-06,
+      "loss": 0.8066,
+      "step": 3618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4851421910833338,
+      "learning_rate": 8.325327131320907e-06,
+      "loss": 0.8167,
+      "step": 3619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.572705418517364,
+      "learning_rate": 8.324356756629215e-06,
+      "loss": 0.7637,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6081858377060583,
+      "learning_rate": 8.323386157469353e-06,
+      "loss": 0.7926,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4000884633593909,
+      "learning_rate": 8.322415333906859e-06,
+      "loss": 0.6416,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5117813133396107,
+      "learning_rate": 8.321444286007283e-06,
+      "loss": 0.7818,
+      "step": 3623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9386306354677698,
+      "learning_rate": 8.320473013836197e-06,
+      "loss": 0.7324,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.503345526855877,
+      "learning_rate": 8.319501517459178e-06,
+      "loss": 0.8259,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4238231424554582,
+      "learning_rate": 8.318529796941825e-06,
+      "loss": 0.8363,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4710546563521905,
+      "learning_rate": 8.317557852349753e-06,
+      "loss": 0.7564,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4765150245971121,
+      "learning_rate": 8.31658568374859e-06,
+      "loss": 0.8249,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.523716677530175,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.8568,
+      "step": 3629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5630988721985417,
+      "learning_rate": 8.314640674781572e-06,
+      "loss": 0.8259,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4591595108802864,
+      "learning_rate": 8.31366783454705e-06,
+      "loss": 0.785,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5597784194332103,
+      "learning_rate": 8.312694770566099e-06,
+      "loss": 0.758,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.527395684505565,
+      "learning_rate": 8.311721482904423e-06,
+      "loss": 0.7684,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5321497390807093,
+      "learning_rate": 8.310747971627736e-06,
+      "loss": 0.8293,
+      "step": 3634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4336090120498481,
+      "learning_rate": 8.309774236801779e-06,
+      "loss": 0.8172,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5941264517496108,
+      "learning_rate": 8.308800278492298e-06,
+      "loss": 0.8842,
+      "step": 3636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4262656887077885,
+      "learning_rate": 8.307826096765054e-06,
+      "loss": 0.7364,
+      "step": 3637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.22061669374943,
+      "learning_rate": 8.306851691685828e-06,
+      "loss": 0.7089,
+      "step": 3638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3484767790273897,
+      "learning_rate": 8.305877063320415e-06,
+      "loss": 0.7653,
+      "step": 3639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3980156399540455,
+      "learning_rate": 8.304902211734623e-06,
+      "loss": 0.7436,
+      "step": 3640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.466159099878649,
+      "learning_rate": 8.303927136994278e-06,
+      "loss": 0.7404,
+      "step": 3641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5701449236394625,
+      "learning_rate": 8.302951839165217e-06,
+      "loss": 0.7577,
+      "step": 3642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4661738535056066,
+      "learning_rate": 8.301976318313295e-06,
+      "loss": 0.7635,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4637220261097987,
+      "learning_rate": 8.30100057450438e-06,
+      "loss": 0.8165,
+      "step": 3644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9610922162589698,
+      "learning_rate": 8.300024607804359e-06,
+      "loss": 1.1358,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4723588716795462,
+      "learning_rate": 8.299048418279133e-06,
+      "loss": 0.7368,
+      "step": 3646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8486091620765244,
+      "learning_rate": 8.298072005994611e-06,
+      "loss": 1.075,
+      "step": 3647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4832652627077432,
+      "learning_rate": 8.297095371016726e-06,
+      "loss": 0.7338,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5847340772523881,
+      "learning_rate": 8.296118513411422e-06,
+      "loss": 0.7499,
+      "step": 3649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6176907450218756,
+      "learning_rate": 8.29514143324466e-06,
+      "loss": 0.7289,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5553312838498012,
+      "learning_rate": 8.294164130582413e-06,
+      "loss": 0.818,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9192958158564914,
+      "learning_rate": 8.293186605490673e-06,
+      "loss": 1.1329,
+      "step": 3652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.578034747205365,
+      "learning_rate": 8.292208858035441e-06,
+      "loss": 0.8171,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5654318499895448,
+      "learning_rate": 8.29123088828274e-06,
+      "loss": 0.7741,
+      "step": 3654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5754295223670007,
+      "learning_rate": 8.290252696298604e-06,
+      "loss": 0.8479,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4656544874062276,
+      "learning_rate": 8.28927428214908e-06,
+      "loss": 0.8207,
+      "step": 3656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4598342522150534,
+      "learning_rate": 8.288295645900237e-06,
+      "loss": 0.7514,
+      "step": 3657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5477924208552247,
+      "learning_rate": 8.287316787618153e-06,
+      "loss": 0.8253,
+      "step": 3658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.52410443810361,
+      "learning_rate": 8.286337707368922e-06,
+      "loss": 0.8346,
+      "step": 3659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9675623714475131,
+      "learning_rate": 8.285358405218655e-06,
+      "loss": 1.1069,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7818552332769975,
+      "learning_rate": 8.284378881233474e-06,
+      "loss": 0.7733,
+      "step": 3661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8264771739143096,
+      "learning_rate": 8.283399135479523e-06,
+      "loss": 1.1344,
+      "step": 3662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8297807961899168,
+      "learning_rate": 8.282419168022953e-06,
+      "loss": 1.1201,
+      "step": 3663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.491049669976652,
+      "learning_rate": 8.281438978929937e-06,
+      "loss": 0.8348,
+      "step": 3664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6267413764207723,
+      "learning_rate": 8.280458568266656e-06,
+      "loss": 0.761,
+      "step": 3665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4919180644454502,
+      "learning_rate": 8.279477936099312e-06,
+      "loss": 0.8624,
+      "step": 3666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.46996093908195,
+      "learning_rate": 8.27849708249412e-06,
+      "loss": 0.8523,
+      "step": 3667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7409776208033274,
+      "learning_rate": 8.277516007517306e-06,
+      "loss": 0.7843,
+      "step": 3668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5220752575857301,
+      "learning_rate": 8.276534711235117e-06,
+      "loss": 0.8422,
+      "step": 3669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1337312671261135,
+      "learning_rate": 8.275553193713812e-06,
+      "loss": 1.1251,
+      "step": 3670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6814279936876195,
+      "learning_rate": 8.274571455019665e-06,
+      "loss": 0.784,
+      "step": 3671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6516102712233145,
+      "learning_rate": 8.273589495218966e-06,
+      "loss": 0.7107,
+      "step": 3672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6525951258644387,
+      "learning_rate": 8.27260731437802e-06,
+      "loss": 0.8266,
+      "step": 3673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.516670064968148,
+      "learning_rate": 8.271624912563143e-06,
+      "loss": 0.8129,
+      "step": 3674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5649738579532149,
+      "learning_rate": 8.270642289840673e-06,
+      "loss": 0.8436,
+      "step": 3675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.459511381277704,
+      "learning_rate": 8.269659446276955e-06,
+      "loss": 0.8439,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3666627441030532,
+      "learning_rate": 8.268676381938356e-06,
+      "loss": 0.7474,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9620365104151546,
+      "learning_rate": 8.267693096891253e-06,
+      "loss": 1.1109,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7794060770429951,
+      "learning_rate": 8.266709591202039e-06,
+      "loss": 0.7939,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8232326273459333,
+      "learning_rate": 8.265725864937124e-06,
+      "loss": 1.092,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5847028958484406,
+      "learning_rate": 8.264741918162933e-06,
+      "loss": 0.7788,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9885877524555804,
+      "learning_rate": 8.2637577509459e-06,
+      "loss": 0.8181,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5636140232589018,
+      "learning_rate": 8.262773363352482e-06,
+      "loss": 0.8773,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5322197240413016,
+      "learning_rate": 8.261788755449145e-06,
+      "loss": 0.8159,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6039068759489672,
+      "learning_rate": 8.260803927302372e-06,
+      "loss": 0.8286,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.007325652423658,
+      "learning_rate": 8.259818878978662e-06,
+      "loss": 1.1182,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4162184983187196,
+      "learning_rate": 8.25883361054453e-06,
+      "loss": 0.7065,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5289544948721676,
+      "learning_rate": 8.257848122066498e-06,
+      "loss": 0.9228,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5822096067741045,
+      "learning_rate": 8.256862413611113e-06,
+      "loss": 0.8038,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4420114914349513,
+      "learning_rate": 8.255876485244927e-06,
+      "loss": 0.8016,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.538905630628463,
+      "learning_rate": 8.25489033703452e-06,
+      "loss": 0.7534,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4937399020306241,
+      "learning_rate": 8.253903969046473e-06,
+      "loss": 0.7573,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5781882040795854,
+      "learning_rate": 8.252917381347389e-06,
+      "loss": 0.7993,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6569071117480108,
+      "learning_rate": 8.251930574003886e-06,
+      "loss": 0.764,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5205082328085566,
+      "learning_rate": 8.250943547082592e-06,
+      "loss": 0.8047,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5461981662483404,
+      "learning_rate": 8.249956300650159e-06,
+      "loss": 0.8179,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8421898963856037,
+      "learning_rate": 8.248968834773246e-06,
+      "loss": 1.1171,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6692837290794227,
+      "learning_rate": 8.247981149518525e-06,
+      "loss": 0.8149,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4887124029496397,
+      "learning_rate": 8.24699324495269e-06,
+      "loss": 0.8349,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5220247267260225,
+      "learning_rate": 8.246005121142448e-06,
+      "loss": 0.7925,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4747789859773595,
+      "learning_rate": 8.245016778154519e-06,
+      "loss": 0.7028,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4576761039548471,
+      "learning_rate": 8.244028216055636e-06,
+      "loss": 0.7689,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8419338015585344,
+      "learning_rate": 8.243039434912547e-06,
+      "loss": 1.152,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8240030479286782,
+      "learning_rate": 8.242050434792022e-06,
+      "loss": 1.1003,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4602738618256044,
+      "learning_rate": 8.24106121576084e-06,
+      "loss": 0.7115,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6235031110181317,
+      "learning_rate": 8.24007177788579e-06,
+      "loss": 0.7587,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4304300010602238,
+      "learning_rate": 8.239082121233687e-06,
+      "loss": 0.8004,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5433043268103428,
+      "learning_rate": 8.238092245871352e-06,
+      "loss": 0.7667,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5402911024148342,
+      "learning_rate": 8.237102151865625e-06,
+      "loss": 0.8272,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4748910148196346,
+      "learning_rate": 8.236111839283355e-06,
+      "loss": 0.8931,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3816695488617081,
+      "learning_rate": 8.23512130819142e-06,
+      "loss": 0.691,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3535917767919141,
+      "learning_rate": 8.234130558656693e-06,
+      "loss": 0.7338,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4823528517561166,
+      "learning_rate": 8.233139590746076e-06,
+      "loss": 0.7646,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.53471570466257,
+      "learning_rate": 8.23214840452648e-06,
+      "loss": 0.7852,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.441458710572957,
+      "learning_rate": 8.231157000064833e-06,
+      "loss": 0.7895,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3935945549390185,
+      "learning_rate": 8.230165377428078e-06,
+      "loss": 0.6701,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.454242128022412,
+      "learning_rate": 8.229173536683169e-06,
+      "loss": 0.7493,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.504515241434594,
+      "learning_rate": 8.22818147789708e-06,
+      "loss": 0.8199,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5473342784488457,
+      "learning_rate": 8.227189201136796e-06,
+      "loss": 0.7917,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4447779263507392,
+      "learning_rate": 8.226196706469315e-06,
+      "loss": 0.7246,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4685502558086694,
+      "learning_rate": 8.22520399396166e-06,
+      "loss": 0.8244,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.497781444278884,
+      "learning_rate": 8.224211063680854e-06,
+      "loss": 0.8015,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7830922293341724,
+      "learning_rate": 8.223217915693944e-06,
+      "loss": 0.8017,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6598816043274716,
+      "learning_rate": 8.22222455006799e-06,
+      "loss": 0.7467,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5134208142926044,
+      "learning_rate": 8.221230966870068e-06,
+      "loss": 0.7741,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5235333212244835,
+      "learning_rate": 8.220237166167264e-06,
+      "loss": 0.8242,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5098455282775667,
+      "learning_rate": 8.219243148026683e-06,
+      "loss": 0.7085,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.440153981879797,
+      "learning_rate": 8.218248912515443e-06,
+      "loss": 0.7333,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6002551382802694,
+      "learning_rate": 8.217254459700679e-06,
+      "loss": 0.8191,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4508402114343613,
+      "learning_rate": 8.216259789649536e-06,
+      "loss": 0.8478,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.815799532687122,
+      "learning_rate": 8.215264902429177e-06,
+      "loss": 0.8376,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6342800399226038,
+      "learning_rate": 8.21426979810678e-06,
+      "loss": 0.8772,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5016387010413945,
+      "learning_rate": 8.213274476749537e-06,
+      "loss": 0.786,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.579277310982556,
+      "learning_rate": 8.212278938424654e-06,
+      "loss": 0.8516,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.502594269467104,
+      "learning_rate": 8.211283183199353e-06,
+      "loss": 0.779,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5880978154192384,
+      "learning_rate": 8.210287211140864e-06,
+      "loss": 0.7874,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5255841949050248,
+      "learning_rate": 8.209291022316445e-06,
+      "loss": 0.786,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.535895465453568,
+      "learning_rate": 8.208294616793357e-06,
+      "loss": 0.8235,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6296658802408148,
+      "learning_rate": 8.20729799463888e-06,
+      "loss": 0.8158,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.560846904942272,
+      "learning_rate": 8.20630115592031e-06,
+      "loss": 0.785,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.650451617029216,
+      "learning_rate": 8.205304100704953e-06,
+      "loss": 0.8163,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5020728105883276,
+      "learning_rate": 8.204306829060133e-06,
+      "loss": 0.756,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.590602830309445,
+      "learning_rate": 8.203309341053191e-06,
+      "loss": 0.7697,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5011316153175087,
+      "learning_rate": 8.202311636751476e-06,
+      "loss": 0.8496,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3903652483703992,
+      "learning_rate": 8.201313716222357e-06,
+      "loss": 0.7736,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5917948373487967,
+      "learning_rate": 8.200315579533217e-06,
+      "loss": 0.7876,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.646637601885048,
+      "learning_rate": 8.19931722675145e-06,
+      "loss": 0.8394,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.554126868557104,
+      "learning_rate": 8.198318657944466e-06,
+      "loss": 0.8789,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4218246823484486,
+      "learning_rate": 8.197319873179694e-06,
+      "loss": 0.737,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9418594207579307,
+      "learning_rate": 8.196320872524574e-06,
+      "loss": 1.1392,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6975148357684215,
+      "learning_rate": 8.19532165604656e-06,
+      "loss": 0.7794,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4792893053976437,
+      "learning_rate": 8.19432222381312e-06,
+      "loss": 0.7572,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5300165095405902,
+      "learning_rate": 8.19332257589174e-06,
+      "loss": 0.8972,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5351479279027318,
+      "learning_rate": 8.192322712349917e-06,
+      "loss": 0.8186,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4286072679340858,
+      "learning_rate": 8.191322633255166e-06,
+      "loss": 0.8809,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6030101223458852,
+      "learning_rate": 8.190322338675015e-06,
+      "loss": 0.8155,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9690597426693164,
+      "learning_rate": 8.189321828677002e-06,
+      "loss": 1.1307,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4574411251427775,
+      "learning_rate": 8.188321103328685e-06,
+      "loss": 0.7583,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6644053400273908,
+      "learning_rate": 8.18732016269764e-06,
+      "loss": 0.8636,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5665260766105897,
+      "learning_rate": 8.186319006851446e-06,
+      "loss": 0.7698,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4535727250840884,
+      "learning_rate": 8.185317635857709e-06,
+      "loss": 0.7784,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5175547824455082,
+      "learning_rate": 8.18431604978404e-06,
+      "loss": 0.8994,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5803185786194645,
+      "learning_rate": 8.183314248698072e-06,
+      "loss": 0.8206,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5454968380607725,
+      "learning_rate": 8.182312232667446e-06,
+      "loss": 0.7982,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5352119106569502,
+      "learning_rate": 8.18131000175982e-06,
+      "loss": 0.8117,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.656401416344771,
+      "learning_rate": 8.18030755604287e-06,
+      "loss": 0.7844,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1652032986188399,
+      "learning_rate": 8.179304895584282e-06,
+      "loss": 1.1349,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9698988631470631,
+      "learning_rate": 8.178302020451754e-06,
+      "loss": 1.1165,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6976265807926911,
+      "learning_rate": 8.17729893071301e-06,
+      "loss": 0.8226,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8271160645959916,
+      "learning_rate": 8.176295626435776e-06,
+      "loss": 0.7614,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9143021059372802,
+      "learning_rate": 8.175292107687796e-06,
+      "loss": 1.1392,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6852365904727538,
+      "learning_rate": 8.174288374536834e-06,
+      "loss": 0.8647,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6884990672839613,
+      "learning_rate": 8.17328442705066e-06,
+      "loss": 0.7427,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5634508092897763,
+      "learning_rate": 8.172280265297068e-06,
+      "loss": 0.8249,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3938030404699415,
+      "learning_rate": 8.17127588934386e-06,
+      "loss": 0.8709,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3944505307849302,
+      "learning_rate": 8.170271299258849e-06,
+      "loss": 0.7559,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5072856740948022,
+      "learning_rate": 8.169266495109872e-06,
+      "loss": 0.7591,
+      "step": 3777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5302284626618314,
+      "learning_rate": 8.168261476964774e-06,
+      "loss": 0.8332,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.646421707866915,
+      "learning_rate": 8.167256244891416e-06,
+      "loss": 0.8894,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3908120607088021,
+      "learning_rate": 8.166250798957676e-06,
+      "loss": 0.8027,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5748914418017717,
+      "learning_rate": 8.165245139231441e-06,
+      "loss": 0.8037,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5668879701571503,
+      "learning_rate": 8.164239265780616e-06,
+      "loss": 0.8598,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4808885196152388,
+      "learning_rate": 8.16323317867312e-06,
+      "loss": 0.7874,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3681648258540557,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.7975,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9543092178766449,
+      "learning_rate": 8.161220363759865e-06,
+      "loss": 1.1422,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4976535676063727,
+      "learning_rate": 8.160213636090014e-06,
+      "loss": 0.8143,
+      "step": 3786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5342618771005647,
+      "learning_rate": 8.159206695035314e-06,
+      "loss": 0.8479,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.505531143908182,
+      "learning_rate": 8.15819954066375e-06,
+      "loss": 0.8263,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6120737348004686,
+      "learning_rate": 8.157192173043336e-06,
+      "loss": 0.7893,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5321389236967746,
+      "learning_rate": 8.156184592242085e-06,
+      "loss": 0.7724,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3897670368135453,
+      "learning_rate": 8.155176798328033e-06,
+      "loss": 0.8432,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4196144256306165,
+      "learning_rate": 8.15416879136923e-06,
+      "loss": 0.7268,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.563232251447117,
+      "learning_rate": 8.153160571433738e-06,
+      "loss": 0.7697,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.746483957083267,
+      "learning_rate": 8.152152138589633e-06,
+      "loss": 0.7688,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.50875533742132,
+      "learning_rate": 8.151143492905008e-06,
+      "loss": 0.7751,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4557661417280336,
+      "learning_rate": 8.150134634447969e-06,
+      "loss": 0.7651,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0659002442138317,
+      "learning_rate": 8.149125563286635e-06,
+      "loss": 1.1427,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6735708120160868,
+      "learning_rate": 8.148116279489144e-06,
+      "loss": 0.7828,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4317118555511164,
+      "learning_rate": 8.147106783123642e-06,
+      "loss": 0.7984,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4369565546771987,
+      "learning_rate": 8.146097074258294e-06,
+      "loss": 0.7761,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8249258462412429,
+      "learning_rate": 8.145087152961278e-06,
+      "loss": 1.0918,
+      "step": 3801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5325023316493578,
+      "learning_rate": 8.144077019300785e-06,
+      "loss": 0.7056,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.453736520667085,
+      "learning_rate": 8.143066673345023e-06,
+      "loss": 0.779,
+      "step": 3803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5625239653088754,
+      "learning_rate": 8.14205611516221e-06,
+      "loss": 0.7194,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.527708630441371,
+      "learning_rate": 8.141045344820586e-06,
+      "loss": 0.7945,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5543797517872724,
+      "learning_rate": 8.140034362388398e-06,
+      "loss": 0.7201,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8646173606919456,
+      "learning_rate": 8.139023167933908e-06,
+      "loss": 1.0832,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.555878182691804,
+      "learning_rate": 8.138011761525397e-06,
+      "loss": 0.7481,
+      "step": 3808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4369293845774085,
+      "learning_rate": 8.137000143231156e-06,
+      "loss": 0.8275,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.470640667305319,
+      "learning_rate": 8.135988313119493e-06,
+      "loss": 0.8074,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.495055429142874,
+      "learning_rate": 8.134976271258727e-06,
+      "loss": 0.742,
+      "step": 3811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5747092597799497,
+      "learning_rate": 8.133964017717195e-06,
+      "loss": 0.8375,
+      "step": 3812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5031204172443617,
+      "learning_rate": 8.132951552563247e-06,
+      "loss": 0.8658,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.517225300082931,
+      "learning_rate": 8.131938875865246e-06,
+      "loss": 0.8629,
+      "step": 3814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4782154497001123,
+      "learning_rate": 8.13092598769157e-06,
+      "loss": 0.8337,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4590981810605412,
+      "learning_rate": 8.12991288811061e-06,
+      "loss": 0.7192,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5711796235150852,
+      "learning_rate": 8.128899577190778e-06,
+      "loss": 0.8182,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.417457526530165,
+      "learning_rate": 8.127886055000491e-06,
+      "loss": 0.7989,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5197320250105462,
+      "learning_rate": 8.126872321608185e-06,
+      "loss": 0.8358,
+      "step": 3819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.68673994512127,
+      "learning_rate": 8.12585837708231e-06,
+      "loss": 0.7517,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6699721715673357,
+      "learning_rate": 8.124844221491327e-06,
+      "loss": 0.7934,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4494430187694314,
+      "learning_rate": 8.123829854903722e-06,
+      "loss": 0.8534,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3859474274381696,
+      "learning_rate": 8.122815277387978e-06,
+      "loss": 0.7909,
+      "step": 3823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9332965048509334,
+      "learning_rate": 8.121800489012608e-06,
+      "loss": 1.1154,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.409940431908979,
+      "learning_rate": 8.12078548984613e-06,
+      "loss": 0.7289,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4304083912971546,
+      "learning_rate": 8.119770279957079e-06,
+      "loss": 0.7839,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4677266108101017,
+      "learning_rate": 8.118754859414006e-06,
+      "loss": 0.7568,
+      "step": 3827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5530649546064237,
+      "learning_rate": 8.117739228285471e-06,
+      "loss": 0.7867,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6659948830070217,
+      "learning_rate": 8.116723386640057e-06,
+      "loss": 0.7192,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5122200071820289,
+      "learning_rate": 8.115707334546352e-06,
+      "loss": 0.7892,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5288178933601648,
+      "learning_rate": 8.114691072072962e-06,
+      "loss": 0.8215,
+      "step": 3831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5026383673556092,
+      "learning_rate": 8.11367459928851e-06,
+      "loss": 0.7931,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6046061121231503,
+      "learning_rate": 8.112657916261631e-06,
+      "loss": 0.7624,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4888653767173121,
+      "learning_rate": 8.11164102306097e-06,
+      "loss": 0.768,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5057079829488431,
+      "learning_rate": 8.11062391975519e-06,
+      "loss": 0.8051,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4173125578615495,
+      "learning_rate": 8.109606606412972e-06,
+      "loss": 0.7585,
+      "step": 3836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.515333250481367,
+      "learning_rate": 8.108589083103006e-06,
+      "loss": 0.8022,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4595418745971636,
+      "learning_rate": 8.107571349893997e-06,
+      "loss": 0.8191,
+      "step": 3838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4235478911545925,
+      "learning_rate": 8.106553406854664e-06,
+      "loss": 0.7653,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5900959071913308,
+      "learning_rate": 8.10553525405374e-06,
+      "loss": 0.83,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4092762504099623,
+      "learning_rate": 8.104516891559977e-06,
+      "loss": 0.7133,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5692655958992905,
+      "learning_rate": 8.103498319442133e-06,
+      "loss": 0.7976,
+      "step": 3842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5511298237813889,
+      "learning_rate": 8.102479537768985e-06,
+      "loss": 0.8779,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4965989545924243,
+      "learning_rate": 8.101460546609327e-06,
+      "loss": 0.8089,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.607638871790108,
+      "learning_rate": 8.100441346031958e-06,
+      "loss": 0.8427,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4797124340015562,
+      "learning_rate": 8.099421936105702e-06,
+      "loss": 0.9141,
+      "step": 3846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9760770383415995,
+      "learning_rate": 8.098402316899389e-06,
+      "loss": 1.1339,
+      "step": 3847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5526191328045458,
+      "learning_rate": 8.097382488481867e-06,
+      "loss": 0.742,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5453979701493838,
+      "learning_rate": 8.096362450921995e-06,
+      "loss": 0.8594,
+      "step": 3849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4374558724845192,
+      "learning_rate": 8.095342204288651e-06,
+      "loss": 0.7258,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4980734851951187,
+      "learning_rate": 8.094321748650725e-06,
+      "loss": 0.7945,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5357532763995854,
+      "learning_rate": 8.093301084077116e-06,
+      "loss": 0.8638,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4497691010249782,
+      "learning_rate": 8.092280210636747e-06,
+      "loss": 0.7821,
+      "step": 3853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4674682640507208,
+      "learning_rate": 8.091259128398548e-06,
+      "loss": 0.7743,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6029868491514643,
+      "learning_rate": 8.09023783743146e-06,
+      "loss": 0.7881,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4536523567780455,
+      "learning_rate": 8.089216337804452e-06,
+      "loss": 0.7498,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5803589661548947,
+      "learning_rate": 8.08819462958649e-06,
+      "loss": 0.7167,
+      "step": 3857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.427776863249755,
+      "learning_rate": 8.087172712846565e-06,
+      "loss": 0.7956,
+      "step": 3858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4406870027362026,
+      "learning_rate": 8.08615058765368e-06,
+      "loss": 0.8,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0592910744920072,
+      "learning_rate": 8.08512825407685e-06,
+      "loss": 1.1002,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.959155808902995,
+      "learning_rate": 8.084105712185105e-06,
+      "loss": 1.082,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4944425388728546,
+      "learning_rate": 8.08308296204749e-06,
+      "loss": 0.7748,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4769353439466872,
+      "learning_rate": 8.082060003733065e-06,
+      "loss": 0.798,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.568008258479058,
+      "learning_rate": 8.0810368373109e-06,
+      "loss": 0.8098,
+      "step": 3864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5682484320596064,
+      "learning_rate": 8.080013462850083e-06,
+      "loss": 0.8465,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5531170905391214,
+      "learning_rate": 8.078989880419715e-06,
+      "loss": 0.7951,
+      "step": 3866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4437436400358405,
+      "learning_rate": 8.07796609008891e-06,
+      "loss": 0.7958,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.851946981784591,
+      "learning_rate": 8.076942091926794e-06,
+      "loss": 1.1192,
+      "step": 3868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4902794002262763,
+      "learning_rate": 8.075917886002514e-06,
+      "loss": 0.7888,
+      "step": 3869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5142626053545107,
+      "learning_rate": 8.074893472385223e-06,
+      "loss": 0.7871,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.624297612978066,
+      "learning_rate": 8.073868851144094e-06,
+      "loss": 0.8043,
+      "step": 3871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5482468964235694,
+      "learning_rate": 8.072844022348312e-06,
+      "loss": 0.7358,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0558270817923883,
+      "learning_rate": 8.071818986067075e-06,
+      "loss": 1.11,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5961546176017365,
+      "learning_rate": 8.070793742369595e-06,
+      "loss": 0.7534,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5190197568140746,
+      "learning_rate": 8.069768291325103e-06,
+      "loss": 0.7875,
+      "step": 3875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6049338444511745,
+      "learning_rate": 8.068742633002834e-06,
+      "loss": 0.7808,
+      "step": 3876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.526258122809751,
+      "learning_rate": 8.067716767472045e-06,
+      "loss": 0.8175,
+      "step": 3877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4975571767294351,
+      "learning_rate": 8.066690694802007e-06,
+      "loss": 0.7865,
+      "step": 3878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6032305116303007,
+      "learning_rate": 8.065664415061998e-06,
+      "loss": 0.9079,
+      "step": 3879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4272513408256953,
+      "learning_rate": 8.064637928321319e-06,
+      "loss": 0.8322,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.507469362960976,
+      "learning_rate": 8.06361123464928e-06,
+      "loss": 0.7513,
+      "step": 3881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4606703926712012,
+      "learning_rate": 8.062584334115205e-06,
+      "loss": 0.8167,
+      "step": 3882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.512377878868531,
+      "learning_rate": 8.061557226788433e-06,
+      "loss": 0.8154,
+      "step": 3883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.274152062813424,
+      "learning_rate": 8.060529912738316e-06,
+      "loss": 1.1046,
+      "step": 3884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5353433255701492,
+      "learning_rate": 8.059502392034219e-06,
+      "loss": 0.8182,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5325766863458323,
+      "learning_rate": 8.058474664745527e-06,
+      "loss": 0.797,
+      "step": 3886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5491816845647797,
+      "learning_rate": 8.057446730941631e-06,
+      "loss": 0.7352,
+      "step": 3887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6204196710056848,
+      "learning_rate": 8.056418590691942e-06,
+      "loss": 0.8068,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4963253745689755,
+      "learning_rate": 8.055390244065878e-06,
+      "loss": 0.7683,
+      "step": 3889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8502328116301726,
+      "learning_rate": 8.05436169113288e-06,
+      "loss": 1.0924,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5034118476661062,
+      "learning_rate": 8.053332931962397e-06,
+      "loss": 0.8689,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6894781389035394,
+      "learning_rate": 8.052303966623892e-06,
+      "loss": 0.8307,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.615911901695192,
+      "learning_rate": 8.051274795186842e-06,
+      "loss": 0.7851,
+      "step": 3893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6016333391138815,
+      "learning_rate": 8.050245417720742e-06,
+      "loss": 0.7695,
+      "step": 3894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5421956835095123,
+      "learning_rate": 8.049215834295097e-06,
+      "loss": 0.7585,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6039131085541198,
+      "learning_rate": 8.048186044979425e-06,
+      "loss": 0.7706,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8645859848822454,
+      "learning_rate": 8.047156049843264e-06,
+      "loss": 1.1456,
+      "step": 3897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.825894519966886,
+      "learning_rate": 8.046125848956155e-06,
+      "loss": 1.1067,
+      "step": 3898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7435986848376734,
+      "learning_rate": 8.045095442387666e-06,
+      "loss": 0.8175,
+      "step": 3899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.442719297918469,
+      "learning_rate": 8.04406483020737e-06,
+      "loss": 0.7432,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.484868110570449,
+      "learning_rate": 8.043034012484853e-06,
+      "loss": 0.837,
+      "step": 3901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5040116327475295,
+      "learning_rate": 8.042002989289723e-06,
+      "loss": 0.6894,
+      "step": 3902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.635191258745165,
+      "learning_rate": 8.040971760691596e-06,
+      "loss": 0.8409,
+      "step": 3903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4931201052380514,
+      "learning_rate": 8.0399403267601e-06,
+      "loss": 0.7305,
+      "step": 3904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6160006047228046,
+      "learning_rate": 8.038908687564884e-06,
+      "loss": 0.7625,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5789898914179807,
+      "learning_rate": 8.037876843175602e-06,
+      "loss": 0.7334,
+      "step": 3906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5161423110216306,
+      "learning_rate": 8.036844793661933e-06,
+      "loss": 0.7481,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5279660145993619,
+      "learning_rate": 8.035812539093557e-06,
+      "loss": 0.8463,
+      "step": 3908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5729013542724508,
+      "learning_rate": 8.034780079540177e-06,
+      "loss": 0.8958,
+      "step": 3909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4780183930755386,
+      "learning_rate": 8.033747415071507e-06,
+      "loss": 0.7521,
+      "step": 3910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.595416949058809,
+      "learning_rate": 8.032714545757274e-06,
+      "loss": 0.7413,
+      "step": 3911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5768405698393624,
+      "learning_rate": 8.031681471667218e-06,
+      "loss": 0.7642,
+      "step": 3912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2855672214473255,
+      "learning_rate": 8.030648192871098e-06,
+      "loss": 1.1237,
+      "step": 3913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6288616836578869,
+      "learning_rate": 8.029614709438684e-06,
+      "loss": 0.7922,
+      "step": 3914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4917122284767028,
+      "learning_rate": 8.028581021439755e-06,
+      "loss": 0.8448,
+      "step": 3915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4978350686115647,
+      "learning_rate": 8.027547128944111e-06,
+      "loss": 0.756,
+      "step": 3916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.600167008675974,
+      "learning_rate": 8.026513032021563e-06,
+      "loss": 0.807,
+      "step": 3917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.533081932848928,
+      "learning_rate": 8.025478730741932e-06,
+      "loss": 0.8281,
+      "step": 3918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.528962051245787,
+      "learning_rate": 8.02444422517506e-06,
+      "loss": 0.7881,
+      "step": 3919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5880010832412603,
+      "learning_rate": 8.023409515390798e-06,
+      "loss": 0.8686,
+      "step": 3920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5772079374608643,
+      "learning_rate": 8.022374601459012e-06,
+      "loss": 0.8091,
+      "step": 3921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4950076215100974,
+      "learning_rate": 8.021339483449585e-06,
+      "loss": 0.7054,
+      "step": 3922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5981906239735482,
+      "learning_rate": 8.020304161432404e-06,
+      "loss": 0.8258,
+      "step": 3923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.582871466044518,
+      "learning_rate": 8.01926863547738e-06,
+      "loss": 0.7819,
+      "step": 3924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5111313561514277,
+      "learning_rate": 8.018232905654435e-06,
+      "loss": 0.7971,
+      "step": 3925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4606330208308904,
+      "learning_rate": 8.017196972033502e-06,
+      "loss": 0.7065,
+      "step": 3926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5981221335128113,
+      "learning_rate": 8.01616083468453e-06,
+      "loss": 0.8025,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5010692513893884,
+      "learning_rate": 8.015124493677483e-06,
+      "loss": 0.7582,
+      "step": 3928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5093800921341247,
+      "learning_rate": 8.014087949082333e-06,
+      "loss": 0.8315,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5224157662854603,
+      "learning_rate": 8.013051200969074e-06,
+      "loss": 0.7838,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3965478340878315,
+      "learning_rate": 8.012014249407707e-06,
+      "loss": 0.7161,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.522353712739441,
+      "learning_rate": 8.01097709446825e-06,
+      "loss": 0.822,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6572966323217033,
+      "learning_rate": 8.009939736220737e-06,
+      "loss": 0.8421,
+      "step": 3933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5463011077422244,
+      "learning_rate": 8.008902174735209e-06,
+      "loss": 0.9301,
+      "step": 3934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5258653522267716,
+      "learning_rate": 8.007864410081726e-06,
+      "loss": 0.8304,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9508047777912927,
+      "learning_rate": 8.006826442330362e-06,
+      "loss": 1.1158,
+      "step": 3936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4941367142638013,
+      "learning_rate": 8.005788271551198e-06,
+      "loss": 0.8637,
+      "step": 3937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5717426933810503,
+      "learning_rate": 8.004749897814338e-06,
+      "loss": 0.7602,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.41273927472715,
+      "learning_rate": 8.003711321189895e-06,
+      "loss": 0.8006,
+      "step": 3939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8022454838430284,
+      "learning_rate": 8.002672541747996e-06,
+      "loss": 1.1078,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.397537296827498,
+      "learning_rate": 8.00163355955878e-06,
+      "loss": 0.7286,
+      "step": 3941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6035521833292534,
+      "learning_rate": 8.0005943746924e-06,
+      "loss": 0.8873,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5815275763886332,
+      "learning_rate": 7.999554987219029e-06,
+      "loss": 0.7995,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4728916686088978,
+      "learning_rate": 7.998515397208846e-06,
+      "loss": 0.7566,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4588430282845442,
+      "learning_rate": 7.997475604732047e-06,
+      "loss": 0.7871,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4199227971928787,
+      "learning_rate": 7.99643560985884e-06,
+      "loss": 0.7435,
+      "step": 3946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.511175635275945,
+      "learning_rate": 7.995395412659449e-06,
+      "loss": 0.7961,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5800139440966647,
+      "learning_rate": 7.994355013204111e-06,
+      "loss": 0.7533,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.497130382038133,
+      "learning_rate": 7.993314411563075e-06,
+      "loss": 0.7005,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.523062216750589,
+      "learning_rate": 7.992273607806607e-06,
+      "loss": 0.8628,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0552400363427563,
+      "learning_rate": 7.99123260200498e-06,
+      "loss": 1.1456,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4838964707276467,
+      "learning_rate": 7.99019139422849e-06,
+      "loss": 0.8016,
+      "step": 3952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4721447584061125,
+      "learning_rate": 7.98914998454744e-06,
+      "loss": 0.8448,
+      "step": 3953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5045001981265207,
+      "learning_rate": 7.988108373032147e-06,
+      "loss": 0.8349,
+      "step": 3954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4764427517327932,
+      "learning_rate": 7.987066559752943e-06,
+      "loss": 0.8238,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8068766089902887,
+      "learning_rate": 7.986024544780175e-06,
+      "loss": 1.1134,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4506128140720853,
+      "learning_rate": 7.984982328184203e-06,
+      "loss": 0.8638,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5971101122615332,
+      "learning_rate": 7.983939910035398e-06,
+      "loss": 0.8183,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.430903795450547,
+      "learning_rate": 7.982897290404146e-06,
+      "loss": 0.8311,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5335505344391536,
+      "learning_rate": 7.981854469360851e-06,
+      "loss": 0.7507,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6263028626172162,
+      "learning_rate": 7.98081144697592e-06,
+      "loss": 0.8323,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.55845023430453,
+      "learning_rate": 7.979768223319786e-06,
+      "loss": 0.8229,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.482748764866512,
+      "learning_rate": 7.978724798462886e-06,
+      "loss": 0.8022,
+      "step": 3963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.493735008454351,
+      "learning_rate": 7.977681172475679e-06,
+      "loss": 0.8549,
+      "step": 3964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.482532742297793,
+      "learning_rate": 7.97663734542863e-06,
+      "loss": 0.8127,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7513453208855454,
+      "learning_rate": 7.97559331739222e-06,
+      "loss": 0.841,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4847733527778437,
+      "learning_rate": 7.974549088436945e-06,
+      "loss": 0.766,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9435898956883498,
+      "learning_rate": 7.973504658633316e-06,
+      "loss": 1.0978,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4863797702954973,
+      "learning_rate": 7.972460028051852e-06,
+      "loss": 0.8088,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4353803053852876,
+      "learning_rate": 7.971415196763088e-06,
+      "loss": 0.7848,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5241599716294403,
+      "learning_rate": 7.970370164837577e-06,
+      "loss": 0.8083,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.469412444802688,
+      "learning_rate": 7.96932493234588e-06,
+      "loss": 0.7511,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.520296326018375,
+      "learning_rate": 7.968279499358573e-06,
+      "loss": 0.7745,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8838530225665533,
+      "learning_rate": 7.967233865946249e-06,
+      "loss": 0.7725,
+      "step": 3974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4907306864964074,
+      "learning_rate": 7.966188032179507e-06,
+      "loss": 0.7788,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4041968878156443,
+      "learning_rate": 7.965141998128968e-06,
+      "loss": 0.7796,
+      "step": 3976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.469675791506989,
+      "learning_rate": 7.96409576386526e-06,
+      "loss": 0.7511,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4097788653134227,
+      "learning_rate": 7.963049329459029e-06,
+      "loss": 0.7424,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4844135232211404,
+      "learning_rate": 7.962002694980933e-06,
+      "loss": 0.8286,
+      "step": 3979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4362714088611868,
+      "learning_rate": 7.960955860501641e-06,
+      "loss": 0.7923,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6764891449864034,
+      "learning_rate": 7.959908826091838e-06,
+      "loss": 0.8033,
+      "step": 3981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.922305893030689,
+      "learning_rate": 7.958861591822223e-06,
+      "loss": 1.1002,
+      "step": 3982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1487739174794696,
+      "learning_rate": 7.957814157763505e-06,
+      "loss": 0.808,
+      "step": 3983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5184797715214313,
+      "learning_rate": 7.956766523986416e-06,
+      "loss": 0.8008,
+      "step": 3984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.538374290753456,
+      "learning_rate": 7.95571869056169e-06,
+      "loss": 0.8075,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4921750956302897,
+      "learning_rate": 7.954670657560078e-06,
+      "loss": 0.7502,
+      "step": 3986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7767683722944754,
+      "learning_rate": 7.953622425052346e-06,
+      "loss": 1.0855,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8113986052820087,
+      "learning_rate": 7.952573993109273e-06,
+      "loss": 1.1276,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6856254695787238,
+      "learning_rate": 7.951525361801655e-06,
+      "loss": 0.7938,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4442232547313745,
+      "learning_rate": 7.950476531200295e-06,
+      "loss": 0.7076,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7882897148337747,
+      "learning_rate": 7.949427501376014e-06,
+      "loss": 1.1051,
+      "step": 3991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.42989455433651,
+      "learning_rate": 7.948378272399641e-06,
+      "loss": 0.8489,
+      "step": 3992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4323906353812812,
+      "learning_rate": 7.947328844342028e-06,
+      "loss": 0.7633,
+      "step": 3993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.423544031476372,
+      "learning_rate": 7.94627921727403e-06,
+      "loss": 0.7676,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6779422803489272,
+      "learning_rate": 7.945229391266522e-06,
+      "loss": 0.7558,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4034077794978006,
+      "learning_rate": 7.944179366390392e-06,
+      "loss": 0.7166,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4439694649847845,
+      "learning_rate": 7.943129142716538e-06,
+      "loss": 0.7891,
+      "step": 3997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9293026565166922,
+      "learning_rate": 7.942078720315876e-06,
+      "loss": 1.1202,
+      "step": 3998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6122789818754983,
+      "learning_rate": 7.941028099259331e-06,
+      "loss": 0.8726,
+      "step": 3999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6053021430983,
+      "learning_rate": 7.939977279617843e-06,
+      "loss": 0.7149,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4264230849097115,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.7008,
+      "step": 4001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3770758150469997,
+      "learning_rate": 7.937875044863868e-06,
+      "loss": 0.6878,
+      "step": 4002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5716848223825886,
+      "learning_rate": 7.93682362989333e-06,
+      "loss": 0.7607,
+      "step": 4003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4644607383001038,
+      "learning_rate": 7.935772016621744e-06,
+      "loss": 0.8052,
+      "step": 4004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4080722193835142,
+      "learning_rate": 7.93472020512012e-06,
+      "loss": 0.7979,
+      "step": 4005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5861018672425293,
+      "learning_rate": 7.933668195459474e-06,
+      "loss": 0.7805,
+      "step": 4006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5487384721554986,
+      "learning_rate": 7.932615987710846e-06,
+      "loss": 0.7512,
+      "step": 4007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6335414516069713,
+      "learning_rate": 7.931563581945278e-06,
+      "loss": 0.8429,
+      "step": 4008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4901232154264255,
+      "learning_rate": 7.930510978233837e-06,
+      "loss": 0.7712,
+      "step": 4009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6244171023945388,
+      "learning_rate": 7.92945817664759e-06,
+      "loss": 0.7726,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0244116488773327,
+      "learning_rate": 7.928405177257632e-06,
+      "loss": 1.0874,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.589382661929631,
+      "learning_rate": 7.927351980135056e-06,
+      "loss": 0.8932,
+      "step": 4012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4551717560896589,
+      "learning_rate": 7.926298585350985e-06,
+      "loss": 0.7523,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8086252023680871,
+      "learning_rate": 7.925244992976538e-06,
+      "loss": 1.137,
+      "step": 4014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4313569773399977,
+      "learning_rate": 7.924191203082863e-06,
+      "loss": 0.7442,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.542888586651162,
+      "learning_rate": 7.92313721574111e-06,
+      "loss": 0.866,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5041287099370164,
+      "learning_rate": 7.922083031022448e-06,
+      "loss": 0.7606,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5654488246795153,
+      "learning_rate": 7.92102864899806e-06,
+      "loss": 0.781,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.409758508701643,
+      "learning_rate": 7.919974069739136e-06,
+      "loss": 0.7864,
+      "step": 4019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9756787816220674,
+      "learning_rate": 7.918919293316886e-06,
+      "loss": 1.1427,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5737861390962202,
+      "learning_rate": 7.917864319802533e-06,
+      "loss": 0.6556,
+      "step": 4021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5410880969005443,
+      "learning_rate": 7.916809149267307e-06,
+      "loss": 0.7605,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5377615528700967,
+      "learning_rate": 7.915753781782458e-06,
+      "loss": 0.7204,
+      "step": 4023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5184454524875124,
+      "learning_rate": 7.914698217419246e-06,
+      "loss": 0.7666,
+      "step": 4024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4538770627168731,
+      "learning_rate": 7.913642456248947e-06,
+      "loss": 0.7906,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6910433230368835,
+      "learning_rate": 7.912586498342845e-06,
+      "loss": 0.7678,
+      "step": 4026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4585120888757963,
+      "learning_rate": 7.911530343772244e-06,
+      "loss": 0.877,
+      "step": 4027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.516307564575182,
+      "learning_rate": 7.910473992608456e-06,
+      "loss": 0.7885,
+      "step": 4028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.614751566727202,
+      "learning_rate": 7.90941744492281e-06,
+      "loss": 0.7929,
+      "step": 4029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3386850160481112,
+      "learning_rate": 7.908360700786643e-06,
+      "loss": 0.7608,
+      "step": 4030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4692965236456101,
+      "learning_rate": 7.907303760271313e-06,
+      "loss": 0.7737,
+      "step": 4031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.434564167102992,
+      "learning_rate": 7.906246623448184e-06,
+      "loss": 0.753,
+      "step": 4032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5189727344698256,
+      "learning_rate": 7.905189290388637e-06,
+      "loss": 0.7387,
+      "step": 4033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6775843680061555,
+      "learning_rate": 7.904131761164068e-06,
+      "loss": 0.8349,
+      "step": 4034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4518383636582608,
+      "learning_rate": 7.903074035845882e-06,
+      "loss": 0.7412,
+      "step": 4035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4645602391130947,
+      "learning_rate": 7.902016114505495e-06,
+      "loss": 0.7586,
+      "step": 4036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.586925272838685,
+      "learning_rate": 7.900957997214349e-06,
+      "loss": 0.8,
+      "step": 4037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7772556984880807,
+      "learning_rate": 7.899899684043882e-06,
+      "loss": 0.7893,
+      "step": 4038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4200462696909866,
+      "learning_rate": 7.898841175065559e-06,
+      "loss": 0.7463,
+      "step": 4039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0361127140023738,
+      "learning_rate": 7.89778247035085e-06,
+      "loss": 1.1175,
+      "step": 4040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4970211209021385,
+      "learning_rate": 7.896723569971243e-06,
+      "loss": 0.758,
+      "step": 4041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3944338380812138,
+      "learning_rate": 7.895664473998237e-06,
+      "loss": 0.7544,
+      "step": 4042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4572424437143552,
+      "learning_rate": 7.894605182503346e-06,
+      "loss": 0.805,
+      "step": 4043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8340840137789861,
+      "learning_rate": 7.89354569555809e-06,
+      "loss": 1.1268,
+      "step": 4044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.409674244037888,
+      "learning_rate": 7.892486013234015e-06,
+      "loss": 0.7817,
+      "step": 4045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5416391090270438,
+      "learning_rate": 7.891426135602672e-06,
+      "loss": 0.779,
+      "step": 4046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4389902976456188,
+      "learning_rate": 7.89036606273562e-06,
+      "loss": 0.7497,
+      "step": 4047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4342151363392264,
+      "learning_rate": 7.889305794704446e-06,
+      "loss": 0.7498,
+      "step": 4048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4615083143292265,
+      "learning_rate": 7.888245331580737e-06,
+      "loss": 0.7344,
+      "step": 4049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1191123232837668,
+      "learning_rate": 7.887184673436099e-06,
+      "loss": 1.1355,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.507052673709932,
+      "learning_rate": 7.88612382034215e-06,
+      "loss": 0.7615,
+      "step": 4051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7046166682342563,
+      "learning_rate": 7.88506277237052e-06,
+      "loss": 0.9015,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4480657340788563,
+      "learning_rate": 7.884001529592855e-06,
+      "loss": 0.776,
+      "step": 4053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.451168725773406,
+      "learning_rate": 7.882940092080813e-06,
+      "loss": 0.7548,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4824560484265543,
+      "learning_rate": 7.881878459906062e-06,
+      "loss": 0.7171,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8727166542680257,
+      "learning_rate": 7.880816633140289e-06,
+      "loss": 1.1119,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4786151724210466,
+      "learning_rate": 7.879754611855191e-06,
+      "loss": 0.7979,
+      "step": 4057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.43418046278535,
+      "learning_rate": 7.878692396122474e-06,
+      "loss": 0.8375,
+      "step": 4058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.597385724479353,
+      "learning_rate": 7.877629986013864e-06,
+      "loss": 0.7918,
+      "step": 4059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.580689772398066,
+      "learning_rate": 7.876567381601097e-06,
+      "loss": 0.6873,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8228734625899167,
+      "learning_rate": 7.875504582955925e-06,
+      "loss": 1.146,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5864897145889505,
+      "learning_rate": 7.874441590150105e-06,
+      "loss": 0.8292,
+      "step": 4062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6022631171520454,
+      "learning_rate": 7.87337840325542e-06,
+      "loss": 0.7522,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5566406270484123,
+      "learning_rate": 7.872315022343654e-06,
+      "loss": 0.8353,
+      "step": 4064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.516582647511497,
+      "learning_rate": 7.871251447486608e-06,
+      "loss": 0.7415,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5479071464902212,
+      "learning_rate": 7.870187678756099e-06,
+      "loss": 0.7962,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4371510181474694,
+      "learning_rate": 7.869123716223954e-06,
+      "loss": 0.8238,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6477147155495877,
+      "learning_rate": 7.868059559962017e-06,
+      "loss": 0.7683,
+      "step": 4068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8643213532316336,
+      "learning_rate": 7.866995210042139e-06,
+      "loss": 1.0929,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4538245747162126,
+      "learning_rate": 7.865930666536188e-06,
+      "loss": 0.797,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4427271612613777,
+      "learning_rate": 7.864865929516047e-06,
+      "loss": 0.6874,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7917307904056428,
+      "learning_rate": 7.863800999053609e-06,
+      "loss": 1.0872,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5068676435797244,
+      "learning_rate": 7.862735875220775e-06,
+      "loss": 0.799,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3920598739037855,
+      "learning_rate": 7.861670558089471e-06,
+      "loss": 0.7388,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4556337555821652,
+      "learning_rate": 7.860605047731627e-06,
+      "loss": 0.6964,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4783271325254737,
+      "learning_rate": 7.859539344219189e-06,
+      "loss": 0.722,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.601602194872592,
+      "learning_rate": 7.858473447624116e-06,
+      "loss": 0.7748,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8806870330300263,
+      "learning_rate": 7.857407358018378e-06,
+      "loss": 1.1427,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.506532956427876,
+      "learning_rate": 7.856341075473963e-06,
+      "loss": 0.9009,
+      "step": 4079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4226336392883272,
+      "learning_rate": 7.855274600062866e-06,
+      "loss": 0.74,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4438381699499918,
+      "learning_rate": 7.8542079318571e-06,
+      "loss": 0.7447,
+      "step": 4081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.579882102746191,
+      "learning_rate": 7.853141070928687e-06,
+      "loss": 0.8575,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4538322322652721,
+      "learning_rate": 7.852074017349665e-06,
+      "loss": 0.7674,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.402468000645661,
+      "learning_rate": 7.851006771192083e-06,
+      "loss": 0.7969,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6529370566615584,
+      "learning_rate": 7.849939332528007e-06,
+      "loss": 0.7627,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5695476636550638,
+      "learning_rate": 7.848871701429508e-06,
+      "loss": 0.8248,
+      "step": 4086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6294974980001364,
+      "learning_rate": 7.847803877968679e-06,
+      "loss": 0.7705,
+      "step": 4087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.571814779154909,
+      "learning_rate": 7.84673586221762e-06,
+      "loss": 0.7596,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9084715371598262,
+      "learning_rate": 7.845667654248445e-06,
+      "loss": 1.1068,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8207476228538237,
+      "learning_rate": 7.844599254133284e-06,
+      "loss": 0.7604,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4508272775647215,
+      "learning_rate": 7.843530661944277e-06,
+      "loss": 0.726,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5907223890401145,
+      "learning_rate": 7.842461877753575e-06,
+      "loss": 0.8023,
+      "step": 4092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5669366383091048,
+      "learning_rate": 7.84139290163335e-06,
+      "loss": 0.7904,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8564035950884357,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 1.1278,
+      "step": 4094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.556462657097795,
+      "learning_rate": 7.839254373893056e-06,
+      "loss": 0.7652,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5299307383643663,
+      "learning_rate": 7.838184822417382e-06,
+      "loss": 0.8047,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3197477452136857,
+      "learning_rate": 7.83711507930098e-06,
+      "loss": 0.6328,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.441556170905288,
+      "learning_rate": 7.836045144616082e-06,
+      "loss": 0.8265,
+      "step": 4098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7949569619280575,
+      "learning_rate": 7.834975018434929e-06,
+      "loss": 1.0918,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.521381015762904,
+      "learning_rate": 7.833904700829782e-06,
+      "loss": 0.7295,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8054600486016754,
+      "learning_rate": 7.832834191872907e-06,
+      "loss": 1.1056,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5923153208522647,
+      "learning_rate": 7.831763491636592e-06,
+      "loss": 0.7895,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6630896943771885,
+      "learning_rate": 7.830692600193129e-06,
+      "loss": 0.8648,
+      "step": 4103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8160574882308241,
+      "learning_rate": 7.829621517614829e-06,
+      "loss": 1.1085,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.54162733514036,
+      "learning_rate": 7.828550243974015e-06,
+      "loss": 0.8517,
+      "step": 4105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.638883667788359,
+      "learning_rate": 7.827478779343021e-06,
+      "loss": 0.7748,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6654975842578723,
+      "learning_rate": 7.826407123794195e-06,
+      "loss": 0.8303,
+      "step": 4107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6193716780900804,
+      "learning_rate": 7.825335277399896e-06,
+      "loss": 0.815,
+      "step": 4108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5303807132423113,
+      "learning_rate": 7.824263240232497e-06,
+      "loss": 0.6822,
+      "step": 4109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6187999543342606,
+      "learning_rate": 7.823191012364386e-06,
+      "loss": 0.8476,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3816735424447701,
+      "learning_rate": 7.822118593867964e-06,
+      "loss": 0.8098,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.578213594403489,
+      "learning_rate": 7.821045984815641e-06,
+      "loss": 0.7366,
+      "step": 4112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.526516972990694,
+      "learning_rate": 7.81997318527984e-06,
+      "loss": 0.8391,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5594017964384999,
+      "learning_rate": 7.818900195333007e-06,
+      "loss": 0.8134,
+      "step": 4114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9490098654622884,
+      "learning_rate": 7.817827015047581e-06,
+      "loss": 1.1202,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6347637286981944,
+      "learning_rate": 7.816753644496034e-06,
+      "loss": 0.7623,
+      "step": 4116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4179584170073853,
+      "learning_rate": 7.81568008375084e-06,
+      "loss": 0.776,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.015602844713071,
+      "learning_rate": 7.81460633288449e-06,
+      "loss": 0.8749,
+      "step": 4118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5811573162590287,
+      "learning_rate": 7.813532391969482e-06,
+      "loss": 0.7697,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8242715783884929,
+      "learning_rate": 7.812458261078333e-06,
+      "loss": 1.1056,
+      "step": 4120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7907055640814653,
+      "learning_rate": 7.811383940283571e-06,
+      "loss": 1.1142,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4068079353304708,
+      "learning_rate": 7.81030942965774e-06,
+      "loss": 0.8015,
+      "step": 4122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5446237816410053,
+      "learning_rate": 7.809234729273386e-06,
+      "loss": 0.7982,
+      "step": 4123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4930966998179445,
+      "learning_rate": 7.808159839203085e-06,
+      "loss": 0.753,
+      "step": 4124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8500370876162396,
+      "learning_rate": 7.807084759519405e-06,
+      "loss": 1.1089,
+      "step": 4125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5481342382308017,
+      "learning_rate": 7.806009490294946e-06,
+      "loss": 0.7734,
+      "step": 4126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6396289498717669,
+      "learning_rate": 7.80493403160231e-06,
+      "loss": 0.7621,
+      "step": 4127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4884550053728478,
+      "learning_rate": 7.803858383514111e-06,
+      "loss": 0.7916,
+      "step": 4128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6710335838552137,
+      "learning_rate": 7.802782546102987e-06,
+      "loss": 0.7253,
+      "step": 4129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.51243507440311,
+      "learning_rate": 7.801706519441572e-06,
+      "loss": 0.8137,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4713141545070838,
+      "learning_rate": 7.800630303602529e-06,
+      "loss": 0.7663,
+      "step": 4131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7024140399457812,
+      "learning_rate": 7.799553898658525e-06,
+      "loss": 0.8451,
+      "step": 4132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4937818726418948,
+      "learning_rate": 7.798477304682237e-06,
+      "loss": 0.774,
+      "step": 4133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.392286372853869,
+      "learning_rate": 7.797400521746365e-06,
+      "loss": 0.7338,
+      "step": 4134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8315094396665093,
+      "learning_rate": 7.796323549923611e-06,
+      "loss": 0.7982,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5703303950496914,
+      "learning_rate": 7.7952463892867e-06,
+      "loss": 0.7972,
+      "step": 4136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4370505945353111,
+      "learning_rate": 7.79416903990836e-06,
+      "loss": 0.8176,
+      "step": 4137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5315378188189168,
+      "learning_rate": 7.793091501861336e-06,
+      "loss": 0.8113,
+      "step": 4138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.493297826204814,
+      "learning_rate": 7.792013775218385e-06,
+      "loss": 0.7883,
+      "step": 4139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5103652741010474,
+      "learning_rate": 7.790935860052283e-06,
+      "loss": 0.8237,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.60878437066434,
+      "learning_rate": 7.78985775643581e-06,
+      "loss": 0.7883,
+      "step": 4141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6303379230767572,
+      "learning_rate": 7.78877946444176e-06,
+      "loss": 0.7849,
+      "step": 4142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4521125867053326,
+      "learning_rate": 7.787700984142945e-06,
+      "loss": 0.7383,
+      "step": 4143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5215535166000425,
+      "learning_rate": 7.786622315612182e-06,
+      "loss": 0.8376,
+      "step": 4144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4834470221984573,
+      "learning_rate": 7.785543458922311e-06,
+      "loss": 0.7714,
+      "step": 4145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4113437731746563,
+      "learning_rate": 7.784464414146176e-06,
+      "loss": 0.831,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4426812193711636,
+      "learning_rate": 7.783385181356637e-06,
+      "loss": 0.7565,
+      "step": 4147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5369042303573446,
+      "learning_rate": 7.782305760626564e-06,
+      "loss": 0.8149,
+      "step": 4148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3885004308701723,
+      "learning_rate": 7.781226152028845e-06,
+      "loss": 0.8012,
+      "step": 4149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4805343010848515,
+      "learning_rate": 7.780146355636378e-06,
+      "loss": 0.8468,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.461350739339513,
+      "learning_rate": 7.779066371522071e-06,
+      "loss": 0.819,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5010365283091356,
+      "learning_rate": 7.777986199758847e-06,
+      "loss": 0.7936,
+      "step": 4152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6005476932629914,
+      "learning_rate": 7.776905840419643e-06,
+      "loss": 0.6725,
+      "step": 4153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5268487675121112,
+      "learning_rate": 7.775825293577407e-06,
+      "loss": 0.8111,
+      "step": 4154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7348004092040583,
+      "learning_rate": 7.7747445593051e-06,
+      "loss": 0.8825,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.521061782528016,
+      "learning_rate": 7.773663637675695e-06,
+      "loss": 0.7844,
+      "step": 4156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5080029549405463,
+      "learning_rate": 7.772582528762179e-06,
+      "loss": 0.7922,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5267580668560983,
+      "learning_rate": 7.771501232637551e-06,
+      "loss": 0.7863,
+      "step": 4158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4504849781900764,
+      "learning_rate": 7.770419749374822e-06,
+      "loss": 0.7597,
+      "step": 4159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4215401641975258,
+      "learning_rate": 7.769338079047018e-06,
+      "loss": 0.7474,
+      "step": 4160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4740358823908783,
+      "learning_rate": 7.768256221727174e-06,
+      "loss": 0.7466,
+      "step": 4161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.480892350314493,
+      "learning_rate": 7.767174177488337e-06,
+      "loss": 0.7742,
+      "step": 4162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5689751089711186,
+      "learning_rate": 7.766091946403575e-06,
+      "loss": 0.7378,
+      "step": 4163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4764941955455988,
+      "learning_rate": 7.76500952854596e-06,
+      "loss": 0.817,
+      "step": 4164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5208929577878219,
+      "learning_rate": 7.763926923988577e-06,
+      "loss": 0.7989,
+      "step": 4165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5255794407822512,
+      "learning_rate": 7.762844132804527e-06,
+      "loss": 0.7369,
+      "step": 4166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4059737456689585,
+      "learning_rate": 7.761761155066927e-06,
+      "loss": 0.817,
+      "step": 4167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3859612241865986,
+      "learning_rate": 7.760677990848896e-06,
+      "loss": 0.7259,
+      "step": 4168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.573993097685248,
+      "learning_rate": 7.759594640223576e-06,
+      "loss": 0.7785,
+      "step": 4169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5092795509331314,
+      "learning_rate": 7.758511103264116e-06,
+      "loss": 0.888,
+      "step": 4170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5571101154144542,
+      "learning_rate": 7.757427380043678e-06,
+      "loss": 0.8347,
+      "step": 4171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4738906182238791,
+      "learning_rate": 7.756343470635439e-06,
+      "loss": 0.9369,
+      "step": 4172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5614808897314527,
+      "learning_rate": 7.755259375112584e-06,
+      "loss": 0.7991,
+      "step": 4173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.344243084474159,
+      "learning_rate": 7.754175093548317e-06,
+      "loss": 0.7198,
+      "step": 4174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.482746674485004,
+      "learning_rate": 7.753090626015849e-06,
+      "loss": 0.8319,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6475920421551085,
+      "learning_rate": 7.752005972588407e-06,
+      "loss": 0.7721,
+      "step": 4176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5715240201339817,
+      "learning_rate": 7.75092113333923e-06,
+      "loss": 0.8921,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.579177309058003,
+      "learning_rate": 7.749836108341567e-06,
+      "loss": 0.7415,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8657553044010098,
+      "learning_rate": 7.748750897668683e-06,
+      "loss": 1.0925,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4854591447709586,
+      "learning_rate": 7.747665501393851e-06,
+      "loss": 0.79,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6709120749312794,
+      "learning_rate": 7.746579919590361e-06,
+      "loss": 0.9027,
+      "step": 4181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5488594144211054,
+      "learning_rate": 7.745494152331516e-06,
+      "loss": 0.8503,
+      "step": 4182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4598032600445816,
+      "learning_rate": 7.744408199690628e-06,
+      "loss": 0.7489,
+      "step": 4183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4956423355160344,
+      "learning_rate": 7.743322061741024e-06,
+      "loss": 0.7445,
+      "step": 4184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6692359573244475,
+      "learning_rate": 7.742235738556039e-06,
+      "loss": 0.8656,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4944136664385241,
+      "learning_rate": 7.741149230209027e-06,
+      "loss": 0.7515,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4329411554988916,
+      "learning_rate": 7.740062536773352e-06,
+      "loss": 0.8521,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9262220855397046,
+      "learning_rate": 7.738975658322387e-06,
+      "loss": 1.0981,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8486505636447852,
+      "learning_rate": 7.737888594929523e-06,
+      "loss": 1.1131,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5269352771980167,
+      "learning_rate": 7.73680134666816e-06,
+      "loss": 0.8515,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6269869950257942,
+      "learning_rate": 7.735713913611716e-06,
+      "loss": 0.8207,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5697695565775103,
+      "learning_rate": 7.734626295833609e-06,
+      "loss": 0.7819,
+      "step": 4192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8993608569996664,
+      "learning_rate": 7.733538493407283e-06,
+      "loss": 1.1325,
+      "step": 4193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5062565891862143,
+      "learning_rate": 7.732450506406187e-06,
+      "loss": 0.8299,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.560781833527426,
+      "learning_rate": 7.731362334903784e-06,
+      "loss": 0.7587,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4762869223912052,
+      "learning_rate": 7.730273978973552e-06,
+      "loss": 0.7801,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.620528820247918,
+      "learning_rate": 7.729185438688978e-06,
+      "loss": 0.7637,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5120754040670334,
+      "learning_rate": 7.728096714123561e-06,
+      "loss": 0.8028,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.47642578481397,
+      "learning_rate": 7.727007805350815e-06,
+      "loss": 0.7648,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9337473362751902,
+      "learning_rate": 7.725918712444266e-06,
+      "loss": 1.1137,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8491196234026714,
+      "learning_rate": 7.724829435477455e-06,
+      "loss": 1.0872,
+      "step": 4201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4885911911670133,
+      "learning_rate": 7.723739974523929e-06,
+      "loss": 0.7191,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.532010853104647,
+      "learning_rate": 7.72265032965725e-06,
+      "loss": 0.8507,
+      "step": 4203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5116661490788128,
+      "learning_rate": 7.721560500950997e-06,
+      "loss": 0.8456,
+      "step": 4204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5542730504734836,
+      "learning_rate": 7.720470488478755e-06,
+      "loss": 0.847,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4754077371952712,
+      "learning_rate": 7.719380292314126e-06,
+      "loss": 0.797,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4867416314680901,
+      "learning_rate": 7.71828991253072e-06,
+      "loss": 0.7553,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5585481286202898,
+      "learning_rate": 7.717199349202165e-06,
+      "loss": 0.7899,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2332870615249518,
+      "learning_rate": 7.716108602402094e-06,
+      "loss": 1.1336,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.384018840270826,
+      "learning_rate": 7.71501767220416e-06,
+      "loss": 0.7228,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0001086061011353,
+      "learning_rate": 7.713926558682027e-06,
+      "loss": 1.1181,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5205406437322297,
+      "learning_rate": 7.712835261909366e-06,
+      "loss": 0.7845,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5061567898121624,
+      "learning_rate": 7.711743781959863e-06,
+      "loss": 0.8497,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5229451963354952,
+      "learning_rate": 7.71065211890722e-06,
+      "loss": 0.7751,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5503241927131728,
+      "learning_rate": 7.709560272825149e-06,
+      "loss": 0.774,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4753823556613102,
+      "learning_rate": 7.708468243787371e-06,
+      "loss": 0.7805,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.573694264778314,
+      "learning_rate": 7.707376031867625e-06,
+      "loss": 0.8028,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.400754269374335,
+      "learning_rate": 7.706283637139658e-06,
+      "loss": 0.7149,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4270743058664341,
+      "learning_rate": 7.705191059677231e-06,
+      "loss": 0.7731,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.534769115221182,
+      "learning_rate": 7.704098299554119e-06,
+      "loss": 0.7238,
+      "step": 4220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4886995187871348,
+      "learning_rate": 7.703005356844106e-06,
+      "loss": 0.7922,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3610081203008522,
+      "learning_rate": 7.70191223162099e-06,
+      "loss": 0.7748,
+      "step": 4222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4548267989979342,
+      "learning_rate": 7.700818923958582e-06,
+      "loss": 0.8084,
+      "step": 4223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2400816528957908,
+      "learning_rate": 7.699725433930705e-06,
+      "loss": 1.1251,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5358913792434428,
+      "learning_rate": 7.698631761611193e-06,
+      "loss": 0.8301,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.438210203840085,
+      "learning_rate": 7.697537907073893e-06,
+      "loss": 0.7406,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9842504321736207,
+      "learning_rate": 7.696443870392666e-06,
+      "loss": 1.1299,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6360067375423002,
+      "learning_rate": 7.695349651641384e-06,
+      "loss": 0.7787,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4504943693263903,
+      "learning_rate": 7.694255250893927e-06,
+      "loss": 0.7882,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6371987991567813,
+      "learning_rate": 7.693160668224197e-06,
+      "loss": 0.7731,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.544152245441889,
+      "learning_rate": 7.692065903706104e-06,
+      "loss": 0.7121,
+      "step": 4231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5505410199279106,
+      "learning_rate": 7.690970957413559e-06,
+      "loss": 0.8486,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5337651113506334,
+      "learning_rate": 7.689875829420505e-06,
+      "loss": 0.7536,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5138418905910365,
+      "learning_rate": 7.688780519800882e-06,
+      "loss": 0.6508,
+      "step": 4234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.546396844943002,
+      "learning_rate": 7.687685028628653e-06,
+      "loss": 0.7727,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4193403776120561,
+      "learning_rate": 7.686589355977785e-06,
+      "loss": 0.7582,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5449389287458746,
+      "learning_rate": 7.685493501922258e-06,
+      "loss": 0.7575,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5512966988031274,
+      "learning_rate": 7.684397466536071e-06,
+      "loss": 0.79,
+      "step": 4238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4362023933909442,
+      "learning_rate": 7.683301249893226e-06,
+      "loss": 0.8598,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7004927446082503,
+      "learning_rate": 7.682204852067748e-06,
+      "loss": 0.8044,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5156650422296631,
+      "learning_rate": 7.681108273133665e-06,
+      "loss": 0.8103,
+      "step": 4241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4996818030672985,
+      "learning_rate": 7.68001151316502e-06,
+      "loss": 0.8051,
+      "step": 4242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4823611199318365,
+      "learning_rate": 7.678914572235868e-06,
+      "loss": 1.125,
+      "step": 4243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4681732840674302,
+      "learning_rate": 7.677817450420279e-06,
+      "loss": 0.8094,
+      "step": 4244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.643339784082716,
+      "learning_rate": 7.676720147792333e-06,
+      "loss": 0.8536,
+      "step": 4245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5190489252386372,
+      "learning_rate": 7.675622664426124e-06,
+      "loss": 0.7592,
+      "step": 4246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5887328276269221,
+      "learning_rate": 7.674525000395752e-06,
+      "loss": 0.7962,
+      "step": 4247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6014947241379347,
+      "learning_rate": 7.673427155775336e-06,
+      "loss": 0.8027,
+      "step": 4248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.498226002584108,
+      "learning_rate": 7.672329130639007e-06,
+      "loss": 0.7619,
+      "step": 4249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7112751540062408,
+      "learning_rate": 7.671230925060903e-06,
+      "loss": 0.7711,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4313764021937572,
+      "learning_rate": 7.670132539115179e-06,
+      "loss": 0.8063,
+      "step": 4251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5333841356848545,
+      "learning_rate": 7.669033972876001e-06,
+      "loss": 0.7923,
+      "step": 4252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5204686882144647,
+      "learning_rate": 7.667935226417545e-06,
+      "loss": 0.7532,
+      "step": 4253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4893916451312839,
+      "learning_rate": 7.666836299814003e-06,
+      "loss": 0.7112,
+      "step": 4254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4594904915078954,
+      "learning_rate": 7.665737193139575e-06,
+      "loss": 0.7341,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5167015008115199,
+      "learning_rate": 7.664637906468477e-06,
+      "loss": 0.7587,
+      "step": 4256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5330451058012113,
+      "learning_rate": 7.663538439874934e-06,
+      "loss": 0.7473,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5886322712991043,
+      "learning_rate": 7.662438793433185e-06,
+      "loss": 0.8275,
+      "step": 4258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7863566802981898,
+      "learning_rate": 7.661338967217483e-06,
+      "loss": 0.7989,
+      "step": 4259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5713131040663302,
+      "learning_rate": 7.660238961302085e-06,
+      "loss": 0.821,
+      "step": 4260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1108849386792157,
+      "learning_rate": 7.65913877576127e-06,
+      "loss": 1.0498,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5963217384674315,
+      "learning_rate": 7.658038410669326e-06,
+      "loss": 0.8797,
+      "step": 4262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4907705443218633,
+      "learning_rate": 7.656937866100549e-06,
+      "loss": 0.7793,
+      "step": 4263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4749402754291656,
+      "learning_rate": 7.655837142129252e-06,
+      "loss": 0.7628,
+      "step": 4264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4881702695950063,
+      "learning_rate": 7.654736238829759e-06,
+      "loss": 0.7763,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5398864243399293,
+      "learning_rate": 7.653635156276405e-06,
+      "loss": 0.7818,
+      "step": 4266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5385809698618045,
+      "learning_rate": 7.652533894543534e-06,
+      "loss": 0.8133,
+      "step": 4267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4626000468635845,
+      "learning_rate": 7.651432453705514e-06,
+      "loss": 0.7307,
+      "step": 4268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5276105220949034,
+      "learning_rate": 7.65033083383671e-06,
+      "loss": 0.7797,
+      "step": 4269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5237831994498543,
+      "learning_rate": 7.649229035011504e-06,
+      "loss": 0.7812,
+      "step": 4270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.469165776406536,
+      "learning_rate": 7.648127057304302e-06,
+      "loss": 0.8344,
+      "step": 4271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4143305900997072,
+      "learning_rate": 7.6470249007895e-06,
+      "loss": 0.6502,
+      "step": 4272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5022938013445049,
+      "learning_rate": 7.645922565541528e-06,
+      "loss": 0.7883,
+      "step": 4273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5182796070949698,
+      "learning_rate": 7.644820051634813e-06,
+      "loss": 0.8199,
+      "step": 4274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7024858698920895,
+      "learning_rate": 7.6437173591438e-06,
+      "loss": 0.8742,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0923416916447097,
+      "learning_rate": 7.642614488142946e-06,
+      "loss": 0.7372,
+      "step": 4276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.455838645607585,
+      "learning_rate": 7.641511438706718e-06,
+      "loss": 0.7721,
+      "step": 4277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.46508158027866,
+      "learning_rate": 7.640408210909599e-06,
+      "loss": 0.774,
+      "step": 4278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0715073211515973,
+      "learning_rate": 7.63930480482608e-06,
+      "loss": 1.1039,
+      "step": 4279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4824786695416268,
+      "learning_rate": 7.638201220530664e-06,
+      "loss": 0.793,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5541555329890155,
+      "learning_rate": 7.637097458097871e-06,
+      "loss": 0.7648,
+      "step": 4281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4715581830376807,
+      "learning_rate": 7.635993517602226e-06,
+      "loss": 0.8416,
+      "step": 4282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4705097981144006,
+      "learning_rate": 7.634889399118271e-06,
+      "loss": 0.778,
+      "step": 4283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4463876868230718,
+      "learning_rate": 7.633785102720558e-06,
+      "loss": 0.7733,
+      "step": 4284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8639601504419201,
+      "learning_rate": 7.632680628483655e-06,
+      "loss": 1.103,
+      "step": 4285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4571909782070105,
+      "learning_rate": 7.631575976482135e-06,
+      "loss": 0.8408,
+      "step": 4286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.517812879755212,
+      "learning_rate": 7.630471146790586e-06,
+      "loss": 0.789,
+      "step": 4287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5462482534105548,
+      "learning_rate": 7.629366139483611e-06,
+      "loss": 0.7502,
+      "step": 4288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5847112411135258,
+      "learning_rate": 7.628260954635822e-06,
+      "loss": 0.8321,
+      "step": 4289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.626924496872272,
+      "learning_rate": 7.627155592321844e-06,
+      "loss": 0.7551,
+      "step": 4290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8772560374759438,
+      "learning_rate": 7.626050052616314e-06,
+      "loss": 1.087,
+      "step": 4291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5511450638842579,
+      "learning_rate": 7.624944335593878e-06,
+      "loss": 0.8056,
+      "step": 4292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4983994197298636,
+      "learning_rate": 7.623838441329197e-06,
+      "loss": 0.8079,
+      "step": 4293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4373888864610964,
+      "learning_rate": 7.622732369896946e-06,
+      "loss": 0.7368,
+      "step": 4294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.684183011001249,
+      "learning_rate": 7.621626121371809e-06,
+      "loss": 0.864,
+      "step": 4295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.440889101203594,
+      "learning_rate": 7.620519695828481e-06,
+      "loss": 0.7506,
+      "step": 4296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4845237042277035,
+      "learning_rate": 7.61941309334167e-06,
+      "loss": 0.7719,
+      "step": 4297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4923881144684834,
+      "learning_rate": 7.618306313986099e-06,
+      "loss": 0.8285,
+      "step": 4298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.449345623249864,
+      "learning_rate": 7.617199357836498e-06,
+      "loss": 0.8164,
+      "step": 4299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.895653352120223,
+      "learning_rate": 7.61609222496761e-06,
+      "loss": 1.1347,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4499179531961914,
+      "learning_rate": 7.6149849154541955e-06,
+      "loss": 0.7333,
+      "step": 4301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5156602284064744,
+      "learning_rate": 7.613877429371019e-06,
+      "loss": 0.8273,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5202399665063493,
+      "learning_rate": 7.612769766792861e-06,
+      "loss": 0.7777,
+      "step": 4303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4178538454410083,
+      "learning_rate": 7.6116619277945135e-06,
+      "loss": 0.7463,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5226165345090021,
+      "learning_rate": 7.610553912450782e-06,
+      "loss": 0.693,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5169762742940291,
+      "learning_rate": 7.60944572083648e-06,
+      "loss": 0.8253,
+      "step": 4306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4233474818748284,
+      "learning_rate": 7.608337353026437e-06,
+      "loss": 0.7646,
+      "step": 4307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6090349201743785,
+      "learning_rate": 7.607228809095491e-06,
+      "loss": 0.8129,
+      "step": 4308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.382930286371338,
+      "learning_rate": 7.606120089118494e-06,
+      "loss": 0.6996,
+      "step": 4309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4954050559958356,
+      "learning_rate": 7.60501119317031e-06,
+      "loss": 0.814,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6677357418047905,
+      "learning_rate": 7.603902121325813e-06,
+      "loss": 0.8031,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5535510126596617,
+      "learning_rate": 7.60279287365989e-06,
+      "loss": 0.7414,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.498769976494804,
+      "learning_rate": 7.6016834502474415e-06,
+      "loss": 0.7587,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.978610783976176,
+      "learning_rate": 7.6005738511633755e-06,
+      "loss": 1.1219,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7269511289022488,
+      "learning_rate": 7.599464076482619e-06,
+      "loss": 0.8417,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7060346774858497,
+      "learning_rate": 7.598354126280102e-06,
+      "loss": 0.7448,
+      "step": 4316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4753743345545858,
+      "learning_rate": 7.597244000630775e-06,
+      "loss": 0.7531,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4280563318368944,
+      "learning_rate": 7.596133699609591e-06,
+      "loss": 0.8283,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4649503477157357,
+      "learning_rate": 7.595023223291525e-06,
+      "loss": 0.8026,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4984128177447051,
+      "learning_rate": 7.593912571751558e-06,
+      "loss": 0.7637,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5527304287500094,
+      "learning_rate": 7.592801745064682e-06,
+      "loss": 0.728,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4780724706426427,
+      "learning_rate": 7.591690743305904e-06,
+      "loss": 0.7869,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6462237263694135,
+      "learning_rate": 7.590579566550238e-06,
+      "loss": 0.8134,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6449242963562214,
+      "learning_rate": 7.589468214872719e-06,
+      "loss": 0.7595,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6011153254890655,
+      "learning_rate": 7.588356688348384e-06,
+      "loss": 0.8581,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6177060397970395,
+      "learning_rate": 7.587244987052287e-06,
+      "loss": 0.8288,
+      "step": 4326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8949060822975573,
+      "learning_rate": 7.586133111059493e-06,
+      "loss": 1.1403,
+      "step": 4327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.447444160786925,
+      "learning_rate": 7.585021060445074e-06,
+      "loss": 0.7068,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.57448977320994,
+      "learning_rate": 7.5839088352841265e-06,
+      "loss": 0.7661,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.499405515443895,
+      "learning_rate": 7.582796435651745e-06,
+      "loss": 0.8352,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5916389890347067,
+      "learning_rate": 7.581683861623041e-06,
+      "loss": 0.8205,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5176762065767715,
+      "learning_rate": 7.580571113273141e-06,
+      "loss": 0.8708,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4280023457206201,
+      "learning_rate": 7.579458190677176e-06,
+      "loss": 0.6761,
+      "step": 4333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5270159083107353,
+      "learning_rate": 7.578345093910298e-06,
+      "loss": 0.7619,
+      "step": 4334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.520632281119033,
+      "learning_rate": 7.577231823047664e-06,
+      "loss": 0.8307,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4538646097903793,
+      "learning_rate": 7.576118378164442e-06,
+      "loss": 0.7529,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4675163782524028,
+      "learning_rate": 7.575004759335817e-06,
+      "loss": 0.717,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5084054387201304,
+      "learning_rate": 7.5738909666369816e-06,
+      "loss": 0.758,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9296587741729297,
+      "learning_rate": 7.572777000143145e-06,
+      "loss": 1.0737,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8391554317580222,
+      "learning_rate": 7.57166285992952e-06,
+      "loss": 1.1121,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4933739961715538,
+      "learning_rate": 7.570548546071342e-06,
+      "loss": 0.7411,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4487064852882023,
+      "learning_rate": 7.5694340586438446e-06,
+      "loss": 0.7399,
+      "step": 4342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5604343512892715,
+      "learning_rate": 7.568319397722284e-06,
+      "loss": 0.7918,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5442962835629153,
+      "learning_rate": 7.567204563381927e-06,
+      "loss": 0.8073,
+      "step": 4344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6528658285051003,
+      "learning_rate": 7.566089555698046e-06,
+      "loss": 0.7736,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6325014598952858,
+      "learning_rate": 7.564974374745931e-06,
+      "loss": 0.6921,
+      "step": 4346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4975678958744811,
+      "learning_rate": 7.563859020600882e-06,
+      "loss": 0.8246,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4249088103674723,
+      "learning_rate": 7.562743493338207e-06,
+      "loss": 0.7158,
+      "step": 4348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.402532592590064,
+      "learning_rate": 7.561627793033233e-06,
+      "loss": 0.7845,
+      "step": 4349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4681373641464144,
+      "learning_rate": 7.560511919761293e-06,
+      "loss": 0.7661,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4998228344296594,
+      "learning_rate": 7.559395873597733e-06,
+      "loss": 0.7771,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5046530568252192,
+      "learning_rate": 7.5582796546179125e-06,
+      "loss": 0.741,
+      "step": 4352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1864870786322053,
+      "learning_rate": 7.557163262897198e-06,
+      "loss": 1.1282,
+      "step": 4353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6146223862429492,
+      "learning_rate": 7.5560466985109726e-06,
+      "loss": 0.8195,
+      "step": 4354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4153723594844139,
+      "learning_rate": 7.5549299615346294e-06,
+      "loss": 0.7825,
+      "step": 4355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4395470460852111,
+      "learning_rate": 7.553813052043575e-06,
+      "loss": 0.7692,
+      "step": 4356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3705871238030503,
+      "learning_rate": 7.552695970113222e-06,
+      "loss": 0.8619,
+      "step": 4357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5993427687030761,
+      "learning_rate": 7.551578715819e-06,
+      "loss": 0.8234,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8463752470139627,
+      "learning_rate": 7.550461289236348e-06,
+      "loss": 1.1407,
+      "step": 4359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4827854245845675,
+      "learning_rate": 7.549343690440718e-06,
+      "loss": 0.7472,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6147954901108643,
+      "learning_rate": 7.548225919507575e-06,
+      "loss": 0.8017,
+      "step": 4361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5413710275147965,
+      "learning_rate": 7.547107976512387e-06,
+      "loss": 0.804,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8223464145809496,
+      "learning_rate": 7.545989861530648e-06,
+      "loss": 1.0896,
+      "step": 4363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4971003494653998,
+      "learning_rate": 7.54487157463785e-06,
+      "loss": 0.8084,
+      "step": 4364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5355669182765073,
+      "learning_rate": 7.543753115909504e-06,
+      "loss": 0.8691,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4444675090143833,
+      "learning_rate": 7.5426344854211304e-06,
+      "loss": 0.7142,
+      "step": 4366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.529455095851079,
+      "learning_rate": 7.541515683248263e-06,
+      "loss": 0.8909,
+      "step": 4367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6087136288377117,
+      "learning_rate": 7.5403967094664454e-06,
+      "loss": 0.804,
+      "step": 4368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5501334457764158,
+      "learning_rate": 7.539277564151233e-06,
+      "loss": 0.7987,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8446806320154354,
+      "learning_rate": 7.538158247378193e-06,
+      "loss": 1.108,
+      "step": 4370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.450893278301483,
+      "learning_rate": 7.537038759222903e-06,
+      "loss": 0.7438,
+      "step": 4371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4835860747238647,
+      "learning_rate": 7.535919099760956e-06,
+      "loss": 0.8105,
+      "step": 4372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7976284157809693,
+      "learning_rate": 7.534799269067952e-06,
+      "loss": 1.1268,
+      "step": 4373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5178645785045264,
+      "learning_rate": 7.533679267219507e-06,
+      "loss": 0.8134,
+      "step": 4374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5339237007045197,
+      "learning_rate": 7.532559094291243e-06,
+      "loss": 0.836,
+      "step": 4375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4191639501819395,
+      "learning_rate": 7.531438750358797e-06,
+      "loss": 0.8009,
+      "step": 4376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.073404249264496,
+      "learning_rate": 7.530318235497818e-06,
+      "loss": 0.8453,
+      "step": 4377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.544258890705297,
+      "learning_rate": 7.529197549783967e-06,
+      "loss": 0.7596,
+      "step": 4378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5847716319897944,
+      "learning_rate": 7.528076693292914e-06,
+      "loss": 0.8175,
+      "step": 4379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5634322928059634,
+      "learning_rate": 7.526955666100343e-06,
+      "loss": 0.8157,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5309875894993068,
+      "learning_rate": 7.525834468281945e-06,
+      "loss": 0.8265,
+      "step": 4381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4992099689723153,
+      "learning_rate": 7.52471309991343e-06,
+      "loss": 0.8632,
+      "step": 4382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5229507569762466,
+      "learning_rate": 7.523591561070511e-06,
+      "loss": 0.8605,
+      "step": 4383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4936596960148676,
+      "learning_rate": 7.5224698518289196e-06,
+      "loss": 0.7816,
+      "step": 4384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.57904237450513,
+      "learning_rate": 7.521347972264399e-06,
+      "loss": 0.808,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4224433317540848,
+      "learning_rate": 7.5202259224526945e-06,
+      "loss": 0.8491,
+      "step": 4386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5678934524067225,
+      "learning_rate": 7.5191037024695745e-06,
+      "loss": 0.7717,
+      "step": 4387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4955881464980407,
+      "learning_rate": 7.517981312390811e-06,
+      "loss": 0.7346,
+      "step": 4388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5287352712131965,
+      "learning_rate": 7.516858752292191e-06,
+      "loss": 0.8329,
+      "step": 4389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4760573917619335,
+      "learning_rate": 7.515736022249516e-06,
+      "loss": 0.7144,
+      "step": 4390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5908775484915763,
+      "learning_rate": 7.5146131223385895e-06,
+      "loss": 0.9125,
+      "step": 4391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4446357512279138,
+      "learning_rate": 7.5134900526352375e-06,
+      "loss": 0.7508,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5482528554724346,
+      "learning_rate": 7.512366813215286e-06,
+      "loss": 0.7382,
+      "step": 4393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3876184210655609,
+      "learning_rate": 7.511243404154586e-06,
+      "loss": 0.7232,
+      "step": 4394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4442582284719454,
+      "learning_rate": 7.510119825528986e-06,
+      "loss": 0.7791,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5393808941433358,
+      "learning_rate": 7.508996077414359e-06,
+      "loss": 0.857,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3776948031403315,
+      "learning_rate": 7.507872159886578e-06,
+      "loss": 0.7386,
+      "step": 4397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6032398901900127,
+      "learning_rate": 7.506748073021532e-06,
+      "loss": 0.7618,
+      "step": 4398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5171875576112623,
+      "learning_rate": 7.505623816895126e-06,
+      "loss": 0.7374,
+      "step": 4399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5651690176117776,
+      "learning_rate": 7.504499391583271e-06,
+      "loss": 0.7324,
+      "step": 4400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4565345845617848,
+      "learning_rate": 7.503374797161891e-06,
+      "loss": 0.7012,
+      "step": 4401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4037754607174209,
+      "learning_rate": 7.502250033706919e-06,
+      "loss": 0.7759,
+      "step": 4402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4495364046868457,
+      "learning_rate": 7.501125101294302e-06,
+      "loss": 0.8297,
+      "step": 4403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5264249502714267,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.8069,
+      "step": 4404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5478318214426088,
+      "learning_rate": 7.498874729899982e-06,
+      "loss": 0.7798,
+      "step": 4405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4691252815554494,
+      "learning_rate": 7.497749291070226e-06,
+      "loss": 0.803,
+      "step": 4406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.923281081026553,
+      "learning_rate": 7.49662368358673e-06,
+      "loss": 0.7212,
+      "step": 4407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9067664229865748,
+      "learning_rate": 7.495497907525492e-06,
+      "loss": 1.1349,
+      "step": 4408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4700144285102168,
+      "learning_rate": 7.49437196296253e-06,
+      "loss": 0.8083,
+      "step": 4409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4812391974820225,
+      "learning_rate": 7.493245849973868e-06,
+      "loss": 0.7616,
+      "step": 4410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8369749140059285,
+      "learning_rate": 7.492119568635545e-06,
+      "loss": 1.0596,
+      "step": 4411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7938021944908413,
+      "learning_rate": 7.490993119023611e-06,
+      "loss": 1.113,
+      "step": 4412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4209759499328414,
+      "learning_rate": 7.489866501214124e-06,
+      "loss": 0.8308,
+      "step": 4413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5860663149442193,
+      "learning_rate": 7.48873971528316e-06,
+      "loss": 0.7837,
+      "step": 4414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4184322378109762,
+      "learning_rate": 7.487612761306798e-06,
+      "loss": 0.6778,
+      "step": 4415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4808637069724075,
+      "learning_rate": 7.486485639361135e-06,
+      "loss": 0.7912,
+      "step": 4416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3302913483851104,
+      "learning_rate": 7.4853583495222745e-06,
+      "loss": 0.7175,
+      "step": 4417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.534302163497541,
+      "learning_rate": 7.484230891866337e-06,
+      "loss": 0.775,
+      "step": 4418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3908339473546543,
+      "learning_rate": 7.483103266469448e-06,
+      "loss": 0.7386,
+      "step": 4419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6856892358510989,
+      "learning_rate": 7.481975473407748e-06,
+      "loss": 0.9158,
+      "step": 4420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5243205134949183,
+      "learning_rate": 7.48084751275739e-06,
+      "loss": 0.7314,
+      "step": 4421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6164006810785032,
+      "learning_rate": 7.479719384594533e-06,
+      "loss": 0.8176,
+      "step": 4422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9683032367136145,
+      "learning_rate": 7.478591088995355e-06,
+      "loss": 1.126,
+      "step": 4423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5916230685098063,
+      "learning_rate": 7.4774626260360384e-06,
+      "loss": 0.705,
+      "step": 4424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8542153996657541,
+      "learning_rate": 7.476333995792779e-06,
+      "loss": 1.1145,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6812178844583427,
+      "learning_rate": 7.475205198341788e-06,
+      "loss": 0.8239,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4743410205209384,
+      "learning_rate": 7.47407623375928e-06,
+      "loss": 0.752,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.479955406980183,
+      "learning_rate": 7.472947102121489e-06,
+      "loss": 0.7278,
+      "step": 4428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4907345608657643,
+      "learning_rate": 7.471817803504655e-06,
+      "loss": 0.8304,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9486845595597211,
+      "learning_rate": 7.470688337985029e-06,
+      "loss": 1.1356,
+      "step": 4430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5589725171143547,
+      "learning_rate": 7.469558705638879e-06,
+      "loss": 0.8287,
+      "step": 4431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5094618436847094,
+      "learning_rate": 7.468428906542476e-06,
+      "loss": 0.8714,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.480691476626689,
+      "learning_rate": 7.46729894077211e-06,
+      "loss": 0.7521,
+      "step": 4433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9057490845002483,
+      "learning_rate": 7.466168808404077e-06,
+      "loss": 1.1213,
+      "step": 4434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5192021096777966,
+      "learning_rate": 7.465038509514688e-06,
+      "loss": 0.7599,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.452669260275029,
+      "learning_rate": 7.463908044180263e-06,
+      "loss": 0.7493,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5791738425488053,
+      "learning_rate": 7.46277741247713e-06,
+      "loss": 0.7953,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4802366438930668,
+      "learning_rate": 7.461646614481637e-06,
+      "loss": 0.8104,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4529672225795658,
+      "learning_rate": 7.460515650270134e-06,
+      "loss": 0.7314,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5513619906625926,
+      "learning_rate": 7.459384519918989e-06,
+      "loss": 0.856,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5407113701238604,
+      "learning_rate": 7.458253223504577e-06,
+      "loss": 0.8,
+      "step": 4441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5912154029641719,
+      "learning_rate": 7.457121761103286e-06,
+      "loss": 0.7282,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5751070182298923,
+      "learning_rate": 7.455990132791516e-06,
+      "loss": 0.8197,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.502449450366959,
+      "learning_rate": 7.454858338645675e-06,
+      "loss": 0.758,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.587884959320326,
+      "learning_rate": 7.453726378742187e-06,
+      "loss": 0.8548,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.270947997604239,
+      "learning_rate": 7.4525942531574836e-06,
+      "loss": 1.1157,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9882953318827534,
+      "learning_rate": 7.451461961968006e-06,
+      "loss": 1.1053,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8137494513368353,
+      "learning_rate": 7.450329505250212e-06,
+      "loss": 1.0708,
+      "step": 4448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.835422930949959,
+      "learning_rate": 7.449196883080567e-06,
+      "loss": 1.0709,
+      "step": 4449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6190029994898856,
+      "learning_rate": 7.448064095535547e-06,
+      "loss": 0.7643,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6579467268678785,
+      "learning_rate": 7.446931142691644e-06,
+      "loss": 0.8615,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5093710239680909,
+      "learning_rate": 7.445798024625354e-06,
+      "loss": 0.7787,
+      "step": 4452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5379650287917215,
+      "learning_rate": 7.444664741413188e-06,
+      "loss": 0.7675,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5848515483771741,
+      "learning_rate": 7.443531293131667e-06,
+      "loss": 0.7581,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6704589357492818,
+      "learning_rate": 7.4423976798573285e-06,
+      "loss": 0.8263,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5631170523382516,
+      "learning_rate": 7.441263901666711e-06,
+      "loss": 0.8442,
+      "step": 4456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6217389237251711,
+      "learning_rate": 7.440129958636375e-06,
+      "loss": 0.7995,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5626161828355551,
+      "learning_rate": 7.438995850842884e-06,
+      "loss": 0.7935,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6020995854827766,
+      "learning_rate": 7.437861578362815e-06,
+      "loss": 0.8315,
+      "step": 4459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6582491822006413,
+      "learning_rate": 7.436727141272759e-06,
+      "loss": 0.7549,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.435160035167625,
+      "learning_rate": 7.435592539649313e-06,
+      "loss": 0.8647,
+      "step": 4461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5905219312794014,
+      "learning_rate": 7.4344577735690915e-06,
+      "loss": 0.8327,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5433677509765416,
+      "learning_rate": 7.433322843108714e-06,
+      "loss": 0.7848,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6767344127329553,
+      "learning_rate": 7.4321877483448125e-06,
+      "loss": 1.106,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.533550614717265,
+      "learning_rate": 7.431052489354033e-06,
+      "loss": 0.8358,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4621292186024115,
+      "learning_rate": 7.42991706621303e-06,
+      "loss": 0.7932,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0760036401895108,
+      "learning_rate": 7.428781478998472e-06,
+      "loss": 1.1319,
+      "step": 4467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.59575383688077,
+      "learning_rate": 7.427645727787035e-06,
+      "loss": 0.7335,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8557642023623613,
+      "learning_rate": 7.4265098126554065e-06,
+      "loss": 1.1122,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.509300032844583,
+      "learning_rate": 7.425373733680286e-06,
+      "loss": 0.7834,
+      "step": 4470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.561025565333014,
+      "learning_rate": 7.424237490938385e-06,
+      "loss": 0.7641,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.446727324134945,
+      "learning_rate": 7.423101084506427e-06,
+      "loss": 0.7707,
+      "step": 4472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.518145623590487,
+      "learning_rate": 7.421964514461142e-06,
+      "loss": 0.874,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4931113269508136,
+      "learning_rate": 7.420827780879276e-06,
+      "loss": 0.7993,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6705298487902591,
+      "learning_rate": 7.4196908838375804e-06,
+      "loss": 0.8045,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.567935852419217,
+      "learning_rate": 7.418553823412824e-06,
+      "loss": 0.7696,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4108153137643755,
+      "learning_rate": 7.4174165996817845e-06,
+      "loss": 0.7291,
+      "step": 4477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5441984109733717,
+      "learning_rate": 7.416279212721247e-06,
+      "loss": 0.7615,
+      "step": 4478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.571409104433138,
+      "learning_rate": 7.415141662608013e-06,
+      "loss": 1.1001,
+      "step": 4479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4674510068689417,
+      "learning_rate": 7.414003949418891e-06,
+      "loss": 0.7972,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.559559161589207,
+      "learning_rate": 7.412866073230702e-06,
+      "loss": 0.7933,
+      "step": 4481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5124636325568186,
+      "learning_rate": 7.411728034120279e-06,
+      "loss": 0.8364,
+      "step": 4482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4764521174429057,
+      "learning_rate": 7.4105898321644635e-06,
+      "loss": 0.7999,
+      "step": 4483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9994380360928251,
+      "learning_rate": 7.409451467440111e-06,
+      "loss": 1.1061,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4280754616379234,
+      "learning_rate": 7.408312940024086e-06,
+      "loss": 0.817,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.564637307707769,
+      "learning_rate": 7.407174249993266e-06,
+      "loss": 0.8172,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.862540568225697,
+      "learning_rate": 7.406035397424532e-06,
+      "loss": 1.1264,
+      "step": 4487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5694149438684317,
+      "learning_rate": 7.404896382394788e-06,
+      "loss": 0.7129,
+      "step": 4488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5989724019661358,
+      "learning_rate": 7.403757204980943e-06,
+      "loss": 0.8007,
+      "step": 4489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5846469964759329,
+      "learning_rate": 7.4026178652599146e-06,
+      "loss": 0.717,
+      "step": 4490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9394319606599356,
+      "learning_rate": 7.401478363308633e-06,
+      "loss": 1.0998,
+      "step": 4491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5236728555917405,
+      "learning_rate": 7.400338699204042e-06,
+      "loss": 0.8036,
+      "step": 4492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.773832535934365,
+      "learning_rate": 7.3991988730230925e-06,
+      "loss": 0.7195,
+      "step": 4493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4775360301054485,
+      "learning_rate": 7.3980588848427485e-06,
+      "loss": 0.8552,
+      "step": 4494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5374459487790948,
+      "learning_rate": 7.396918734739985e-06,
+      "loss": 0.7039,
+      "step": 4495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5569974631329133,
+      "learning_rate": 7.395778422791788e-06,
+      "loss": 0.7892,
+      "step": 4496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4311045343933464,
+      "learning_rate": 7.3946379490751545e-06,
+      "loss": 0.7424,
+      "step": 4497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6136030205219534,
+      "learning_rate": 7.3934973136670905e-06,
+      "loss": 0.8124,
+      "step": 4498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9137278403717024,
+      "learning_rate": 7.392356516644614e-06,
+      "loss": 1.1334,
+      "step": 4499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5618897397664098,
+      "learning_rate": 7.391215558084755e-06,
+      "loss": 0.7721,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8346859955591993,
+      "learning_rate": 7.390074438064555e-06,
+      "loss": 1.0942,
+      "step": 4501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4709962071679026,
+      "learning_rate": 7.388933156661064e-06,
+      "loss": 0.8021,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5227932686653578,
+      "learning_rate": 7.387791713951343e-06,
+      "loss": 0.7822,
+      "step": 4503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.511214386707306,
+      "learning_rate": 7.386650110012465e-06,
+      "loss": 0.7607,
+      "step": 4504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5717853346110455,
+      "learning_rate": 7.385508344921514e-06,
+      "loss": 0.8101,
+      "step": 4505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5527214377975584,
+      "learning_rate": 7.3843664187555855e-06,
+      "loss": 0.827,
+      "step": 4506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4372600497301877,
+      "learning_rate": 7.383224331591784e-06,
+      "loss": 0.7536,
+      "step": 4507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5170466113504606,
+      "learning_rate": 7.382082083507226e-06,
+      "loss": 0.8055,
+      "step": 4508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4703107889201252,
+      "learning_rate": 7.380939674579038e-06,
+      "loss": 0.7668,
+      "step": 4509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5217976796363797,
+      "learning_rate": 7.3797971048843606e-06,
+      "loss": 0.7973,
+      "step": 4510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5646557771711247,
+      "learning_rate": 7.378654374500339e-06,
+      "loss": 0.8751,
+      "step": 4511
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.648499216136936,
+      "learning_rate": 7.3775114835041366e-06,
+      "loss": 0.763,
+      "step": 4512
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4468493278035663,
+      "learning_rate": 7.376368431972921e-06,
+      "loss": 0.7567,
+      "step": 4513
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.534880464602087,
+      "learning_rate": 7.375225219983876e-06,
+      "loss": 0.7824,
+      "step": 4514
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.417214334223045,
+      "learning_rate": 7.374081847614191e-06,
+      "loss": 0.8012,
+      "step": 4515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5169733049776772,
+      "learning_rate": 7.372938314941073e-06,
+      "loss": 0.856,
+      "step": 4516
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.471238646938426,
+      "learning_rate": 7.371794622041734e-06,
+      "loss": 0.7251,
+      "step": 4517
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.478420950869702,
+      "learning_rate": 7.370650768993398e-06,
+      "loss": 0.8114,
+      "step": 4518
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.56609965284195,
+      "learning_rate": 7.3695067558733015e-06,
+      "loss": 0.7349,
+      "step": 4519
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.501973499341524,
+      "learning_rate": 7.36836258275869e-06,
+      "loss": 0.7551,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3593224998388382,
+      "learning_rate": 7.367218249726821e-06,
+      "loss": 0.7322,
+      "step": 4521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.96299982494295,
+      "learning_rate": 7.3660737568549635e-06,
+      "loss": 1.1156,
+      "step": 4522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.479178383605749,
+      "learning_rate": 7.3649291042203955e-06,
+      "loss": 0.861,
+      "step": 4523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6239217981423475,
+      "learning_rate": 7.363784291900407e-06,
+      "loss": 0.8625,
+      "step": 4524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5466267012072479,
+      "learning_rate": 7.362639319972298e-06,
+      "loss": 0.7372,
+      "step": 4525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8297573882565058,
+      "learning_rate": 7.3614941885133785e-06,
+      "loss": 1.125,
+      "step": 4526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5747516875943632,
+      "learning_rate": 7.360348897600973e-06,
+      "loss": 0.7438,
+      "step": 4527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5776705484578561,
+      "learning_rate": 7.35920344731241e-06,
+      "loss": 0.8328,
+      "step": 4528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5133240724694084,
+      "learning_rate": 7.358057837725039e-06,
+      "loss": 0.7571,
+      "step": 4529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9634985790222286,
+      "learning_rate": 7.35691206891621e-06,
+      "loss": 0.8292,
+      "step": 4530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6446186441207828,
+      "learning_rate": 7.355766140963288e-06,
+      "loss": 0.785,
+      "step": 4531
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4869081828584967,
+      "learning_rate": 7.354620053943649e-06,
+      "loss": 0.7866,
+      "step": 4532
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.513116775560579,
+      "learning_rate": 7.35347380793468e-06,
+      "loss": 0.7672,
+      "step": 4533
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4576860551614095,
+      "learning_rate": 7.352327403013779e-06,
+      "loss": 0.7815,
+      "step": 4534
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8783875179570352,
+      "learning_rate": 7.3511808392583515e-06,
+      "loss": 1.1358,
+      "step": 4535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5373897843430362,
+      "learning_rate": 7.350034116745818e-06,
+      "loss": 0.7773,
+      "step": 4536
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5238170824036747,
+      "learning_rate": 7.348887235553608e-06,
+      "loss": 0.7779,
+      "step": 4537
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4576988423339814,
+      "learning_rate": 7.3477401957591586e-06,
+      "loss": 0.8563,
+      "step": 4538
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4689268014676848,
+      "learning_rate": 7.346592997439925e-06,
+      "loss": 0.8447,
+      "step": 4539
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.797290334326206,
+      "learning_rate": 7.345445640673365e-06,
+      "loss": 1.1263,
+      "step": 4540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5412287354675323,
+      "learning_rate": 7.344298125536953e-06,
+      "loss": 0.7407,
+      "step": 4541
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3765144291168856,
+      "learning_rate": 7.343150452108171e-06,
+      "loss": 0.7819,
+      "step": 4542
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4170475858166487,
+      "learning_rate": 7.342002620464512e-06,
+      "loss": 0.7877,
+      "step": 4543
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6153469470951902,
+      "learning_rate": 7.340854630683481e-06,
+      "loss": 0.8111,
+      "step": 4544
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4853307034715892,
+      "learning_rate": 7.339706482842593e-06,
+      "loss": 0.6871,
+      "step": 4545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8018667633294743,
+      "learning_rate": 7.338558177019372e-06,
+      "loss": 1.078,
+      "step": 4546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6132677679633705,
+      "learning_rate": 7.337409713291357e-06,
+      "loss": 0.8763,
+      "step": 4547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5599499378976402,
+      "learning_rate": 7.336261091736092e-06,
+      "loss": 0.7218,
+      "step": 4548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5133871959801573,
+      "learning_rate": 7.335112312431137e-06,
+      "loss": 0.806,
+      "step": 4549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4359670567053873,
+      "learning_rate": 7.333963375454058e-06,
+      "loss": 0.729,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7163139114865082,
+      "learning_rate": 7.332814280882437e-06,
+      "loss": 0.808,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7813993052819308,
+      "learning_rate": 7.33166502879386e-06,
+      "loss": 1.102,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.610011836222946,
+      "learning_rate": 7.330515619265928e-06,
+      "loss": 0.7423,
+      "step": 4553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.44057529650174,
+      "learning_rate": 7.3293660523762535e-06,
+      "loss": 0.7548,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.493230618739604,
+      "learning_rate": 7.328216328202456e-06,
+      "loss": 0.8319,
+      "step": 4555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7211274838524104,
+      "learning_rate": 7.3270664468221685e-06,
+      "loss": 0.8386,
+      "step": 4556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4734333494349543,
+      "learning_rate": 7.325916408313034e-06,
+      "loss": 0.7887,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8706362076510563,
+      "learning_rate": 7.324766212752703e-06,
+      "loss": 1.1077,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.500314755079102,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.6582,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4359050462423344,
+      "learning_rate": 7.322465350789126e-06,
+      "loss": 0.7437,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.452246924245335,
+      "learning_rate": 7.32131468454124e-06,
+      "loss": 0.6863,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7425143771132683,
+      "learning_rate": 7.320163861552877e-06,
+      "loss": 1.1132,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6528053666337676,
+      "learning_rate": 7.3190128819017455e-06,
+      "loss": 0.735,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7838750572933093,
+      "learning_rate": 7.31786174566556e-06,
+      "loss": 1.0975,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8006898540404943,
+      "learning_rate": 7.316710452922049e-06,
+      "loss": 1.1369,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.520959534455294,
+      "learning_rate": 7.315559003748952e-06,
+      "loss": 0.8038,
+      "step": 4566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7915753251655056,
+      "learning_rate": 7.314407398224016e-06,
+      "loss": 1.1097,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5205197459164397,
+      "learning_rate": 7.313255636425002e-06,
+      "loss": 0.807,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.835679507813991,
+      "learning_rate": 7.312103718429675e-06,
+      "loss": 1.0988,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4102329720132356,
+      "learning_rate": 7.310951644315818e-06,
+      "loss": 0.8494,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4050092065510982,
+      "learning_rate": 7.309799414161224e-06,
+      "loss": 0.7956,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5269907826275169,
+      "learning_rate": 7.30864702804369e-06,
+      "loss": 0.7739,
+      "step": 4572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4851631654013626,
+      "learning_rate": 7.30749448604103e-06,
+      "loss": 0.7154,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.554502823323659,
+      "learning_rate": 7.306341788231067e-06,
+      "loss": 0.7708,
+      "step": 4574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5306898368352797,
+      "learning_rate": 7.305188934691632e-06,
+      "loss": 0.8476,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5058027216673626,
+      "learning_rate": 7.304035925500567e-06,
+      "loss": 0.8892,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4261896175346152,
+      "learning_rate": 7.30288276073573e-06,
+      "loss": 0.745,
+      "step": 4577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5146869291857377,
+      "learning_rate": 7.301729440474984e-06,
+      "loss": 0.8121,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4675755543713302,
+      "learning_rate": 7.300575964796201e-06,
+      "loss": 0.7123,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5417926167405183,
+      "learning_rate": 7.299422333777271e-06,
+      "loss": 0.8558,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5599664311581074,
+      "learning_rate": 7.298268547496084e-06,
+      "loss": 0.7736,
+      "step": 4581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6341064090453004,
+      "learning_rate": 7.29711460603055e-06,
+      "loss": 0.7506,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.558610488855587,
+      "learning_rate": 7.295960509458586e-06,
+      "loss": 0.7454,
+      "step": 4583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5630480440594088,
+      "learning_rate": 7.294806257858118e-06,
+      "loss": 0.8568,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3882603179005861,
+      "learning_rate": 7.293651851307084e-06,
+      "loss": 0.7046,
+      "step": 4585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4869904026901715,
+      "learning_rate": 7.292497289883432e-06,
+      "loss": 0.8207,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9222619709560712,
+      "learning_rate": 7.291342573665121e-06,
+      "loss": 1.0963,
+      "step": 4587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5635459795793658,
+      "learning_rate": 7.29018770273012e-06,
+      "loss": 0.8578,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4499581092029958,
+      "learning_rate": 7.289032677156408e-06,
+      "loss": 0.7989,
+      "step": 4589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5739923097791575,
+      "learning_rate": 7.287877497021978e-06,
+      "loss": 0.8267,
+      "step": 4590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5774278575228435,
+      "learning_rate": 7.286722162404825e-06,
+      "loss": 0.7308,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5367162063957502,
+      "learning_rate": 7.285566673382965e-06,
+      "loss": 0.7739,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4919534070041502,
+      "learning_rate": 7.284411030034414e-06,
+      "loss": 0.781,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4695548868980683,
+      "learning_rate": 7.2832552324372075e-06,
+      "loss": 0.8481,
+      "step": 4594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.408539649559492,
+      "learning_rate": 7.282099280669388e-06,
+      "loss": 0.7927,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4535218871447468,
+      "learning_rate": 7.280943174809006e-06,
+      "loss": 0.7999,
+      "step": 4596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4497711259524442,
+      "learning_rate": 7.279786914934126e-06,
+      "loss": 0.7823,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4143340308817791,
+      "learning_rate": 7.278630501122819e-06,
+      "loss": 0.7972,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.49317668349312,
+      "learning_rate": 7.27747393345317e-06,
+      "loss": 0.8239,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5089484522728334,
+      "learning_rate": 7.276317212003274e-06,
+      "loss": 0.7534,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.656626757135166,
+      "learning_rate": 7.2751603368512354e-06,
+      "loss": 0.8086,
+      "step": 4601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3684730138091772,
+      "learning_rate": 7.274003308075169e-06,
+      "loss": 0.8026,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.58249112439259,
+      "learning_rate": 7.272846125753198e-06,
+      "loss": 0.8248,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5605885966638997,
+      "learning_rate": 7.271688789963458e-06,
+      "loss": 0.8327,
+      "step": 4604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5680393457938742,
+      "learning_rate": 7.2705313007841e-06,
+      "loss": 0.7864,
+      "step": 4605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.554642227587047,
+      "learning_rate": 7.269373658293275e-06,
+      "loss": 0.6971,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9079736589620343,
+      "learning_rate": 7.268215862569151e-06,
+      "loss": 1.124,
+      "step": 4607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5559281402441132,
+      "learning_rate": 7.2670579136899045e-06,
+      "loss": 0.8009,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4522046219291458,
+      "learning_rate": 7.265899811733726e-06,
+      "loss": 0.7476,
+      "step": 4609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7147423107046722,
+      "learning_rate": 7.264741556778808e-06,
+      "loss": 0.7651,
+      "step": 4610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5432222771012145,
+      "learning_rate": 7.263583148903363e-06,
+      "loss": 0.8001,
+      "step": 4611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.527247636635427,
+      "learning_rate": 7.2624245881856094e-06,
+      "loss": 0.8456,
+      "step": 4612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5882101945762195,
+      "learning_rate": 7.261265874703771e-06,
+      "loss": 0.7857,
+      "step": 4613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6618429721523837,
+      "learning_rate": 7.260107008536092e-06,
+      "loss": 0.8391,
+      "step": 4614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7271491426580903,
+      "learning_rate": 7.25894798976082e-06,
+      "loss": 0.7935,
+      "step": 4615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4909041368048042,
+      "learning_rate": 7.257788818456213e-06,
+      "loss": 0.7526,
+      "step": 4616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5760311477886415,
+      "learning_rate": 7.2566294947005434e-06,
+      "loss": 0.8035,
+      "step": 4617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5100025415169613,
+      "learning_rate": 7.255470018572091e-06,
+      "loss": 0.7892,
+      "step": 4618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5494231019424556,
+      "learning_rate": 7.254310390149144e-06,
+      "loss": 0.7834,
+      "step": 4619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6053484764909907,
+      "learning_rate": 7.253150609510005e-06,
+      "loss": 0.7834,
+      "step": 4620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5422690019823806,
+      "learning_rate": 7.251990676732985e-06,
+      "loss": 0.7564,
+      "step": 4621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4555659452523377,
+      "learning_rate": 7.250830591896404e-06,
+      "loss": 0.7828,
+      "step": 4622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.356230663872062,
+      "learning_rate": 7.249670355078595e-06,
+      "loss": 0.7126,
+      "step": 4623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5054456934583529,
+      "learning_rate": 7.2485099663579e-06,
+      "loss": 0.7813,
+      "step": 4624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4944373577851044,
+      "learning_rate": 7.247349425812671e-06,
+      "loss": 0.8102,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5655681630023026,
+      "learning_rate": 7.246188733521269e-06,
+      "loss": 0.7098,
+      "step": 4626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.556136235873271,
+      "learning_rate": 7.245027889562067e-06,
+      "loss": 0.8211,
+      "step": 4627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5530729986278262,
+      "learning_rate": 7.243866894013449e-06,
+      "loss": 0.8293,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4844219835216177,
+      "learning_rate": 7.242705746953809e-06,
+      "loss": 0.7922,
+      "step": 4629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5030128696748513,
+      "learning_rate": 7.241544448461546e-06,
+      "loss": 0.7481,
+      "step": 4630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4868687273969419,
+      "learning_rate": 7.240382998615079e-06,
+      "loss": 0.8063,
+      "step": 4631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5311249362731447,
+      "learning_rate": 7.239221397492826e-06,
+      "loss": 0.7758,
+      "step": 4632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.581873105053724,
+      "learning_rate": 7.238059645173225e-06,
+      "loss": 0.8057,
+      "step": 4633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5001216846599679,
+      "learning_rate": 7.236897741734721e-06,
+      "loss": 0.8261,
+      "step": 4634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9017323113833672,
+      "learning_rate": 7.2357356872557646e-06,
+      "loss": 1.0761,
+      "step": 4635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4638189923872649,
+      "learning_rate": 7.234573481814823e-06,
+      "loss": 0.7376,
+      "step": 4636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8284602880687917,
+      "learning_rate": 7.233411125490369e-06,
+      "loss": 1.1378,
+      "step": 4637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5145340481943748,
+      "learning_rate": 7.232248618360889e-06,
+      "loss": 0.7478,
+      "step": 4638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4341885715769642,
+      "learning_rate": 7.231085960504879e-06,
+      "loss": 0.7209,
+      "step": 4639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5434596172304318,
+      "learning_rate": 7.229923152000844e-06,
+      "loss": 0.7461,
+      "step": 4640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5921242351850915,
+      "learning_rate": 7.2287601929272975e-06,
+      "loss": 0.7693,
+      "step": 4641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4552397050096042,
+      "learning_rate": 7.227597083362766e-06,
+      "loss": 0.7114,
+      "step": 4642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7778863229514374,
+      "learning_rate": 7.226433823385787e-06,
+      "loss": 0.7765,
+      "step": 4643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4298980811750102,
+      "learning_rate": 7.225270413074904e-06,
+      "loss": 0.6778,
+      "step": 4644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9191843047887203,
+      "learning_rate": 7.2241068525086745e-06,
+      "loss": 0.7763,
+      "step": 4645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4958884764427443,
+      "learning_rate": 7.222943141765666e-06,
+      "loss": 0.8022,
+      "step": 4646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5353935157918404,
+      "learning_rate": 7.221779280924451e-06,
+      "loss": 0.805,
+      "step": 4647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5672524438746964,
+      "learning_rate": 7.220615270063621e-06,
+      "loss": 0.8547,
+      "step": 4648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4791295649184775,
+      "learning_rate": 7.219451109261768e-06,
+      "loss": 0.729,
+      "step": 4649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5072207574863103,
+      "learning_rate": 7.2182867985975036e-06,
+      "loss": 0.7533,
+      "step": 4650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2563532099145733,
+      "learning_rate": 7.217122338149441e-06,
+      "loss": 1.111,
+      "step": 4651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4966573008385011,
+      "learning_rate": 7.215957727996208e-06,
+      "loss": 0.8433,
+      "step": 4652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4573497273656213,
+      "learning_rate": 7.214792968216442e-06,
+      "loss": 0.8106,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5201068717350101,
+      "learning_rate": 7.21362805888879e-06,
+      "loss": 0.8007,
+      "step": 4654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4448079159780887,
+      "learning_rate": 7.21246300009191e-06,
+      "loss": 0.7464,
+      "step": 4655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4595710020956827,
+      "learning_rate": 7.21129779190447e-06,
+      "loss": 0.7993,
+      "step": 4656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7455999304862324,
+      "learning_rate": 7.210132434405146e-06,
+      "loss": 0.776,
+      "step": 4657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.802064195901956,
+      "learning_rate": 7.208966927672627e-06,
+      "loss": 0.8088,
+      "step": 4658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9532426347209004,
+      "learning_rate": 7.2078012717856086e-06,
+      "loss": 1.1044,
+      "step": 4659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5330124560751197,
+      "learning_rate": 7.2066354668228e-06,
+      "loss": 0.8845,
+      "step": 4660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7920123859247055,
+      "learning_rate": 7.205469512862919e-06,
+      "loss": 0.7932,
+      "step": 4661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.52795734465919,
+      "learning_rate": 7.204303409984694e-06,
+      "loss": 0.8356,
+      "step": 4662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7955861053176866,
+      "learning_rate": 7.203137158266863e-06,
+      "loss": 1.1176,
+      "step": 4663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5111570276163415,
+      "learning_rate": 7.201970757788172e-06,
+      "loss": 0.7861,
+      "step": 4664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7976404007993863,
+      "learning_rate": 7.200804208627381e-06,
+      "loss": 1.109,
+      "step": 4665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4971517058325816,
+      "learning_rate": 7.1996375108632566e-06,
+      "loss": 0.7925,
+      "step": 4666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.819389475096111,
+      "learning_rate": 7.19847066457458e-06,
+      "loss": 1.1012,
+      "step": 4667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8180190806499755,
+      "learning_rate": 7.197303669840134e-06,
+      "loss": 1.1038,
+      "step": 4668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7267538332590993,
+      "learning_rate": 7.1961365267387205e-06,
+      "loss": 0.8255,
+      "step": 4669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4519627778288022,
+      "learning_rate": 7.194969235349149e-06,
+      "loss": 0.755,
+      "step": 4670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4739521326190854,
+      "learning_rate": 7.193801795750233e-06,
+      "loss": 0.7597,
+      "step": 4671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5952054460417515,
+      "learning_rate": 7.192634208020805e-06,
+      "loss": 0.807,
+      "step": 4672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5972105363609925,
+      "learning_rate": 7.191466472239701e-06,
+      "loss": 0.7917,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5456909904471943,
+      "learning_rate": 7.190298588485769e-06,
+      "loss": 0.7943,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9832806310442592,
+      "learning_rate": 7.189130556837869e-06,
+      "loss": 1.1255,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8739278074580219,
+      "learning_rate": 7.187962377374866e-06,
+      "loss": 1.079,
+      "step": 4676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6167959229171456,
+      "learning_rate": 7.186794050175643e-06,
+      "loss": 0.7271,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5380461353563224,
+      "learning_rate": 7.185625575319085e-06,
+      "loss": 0.8499,
+      "step": 4678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4375321687959584,
+      "learning_rate": 7.184456952884089e-06,
+      "loss": 0.6912,
+      "step": 4679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6065665810045027,
+      "learning_rate": 7.183288182949565e-06,
+      "loss": 0.7215,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9295714590017,
+      "learning_rate": 7.182119265594429e-06,
+      "loss": 1.1174,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.584268276821332,
+      "learning_rate": 7.180950200897614e-06,
+      "loss": 0.8308,
+      "step": 4682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8701558071867386,
+      "learning_rate": 7.179780988938051e-06,
+      "loss": 1.0828,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5861893550133337,
+      "learning_rate": 7.178611629794693e-06,
+      "loss": 0.7777,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5547645048160008,
+      "learning_rate": 7.177442123546496e-06,
+      "loss": 0.777,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6277743617012475,
+      "learning_rate": 7.1762724702724275e-06,
+      "loss": 0.7349,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4786975759617262,
+      "learning_rate": 7.175102670051466e-06,
+      "loss": 0.7861,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9087873213334678,
+      "learning_rate": 7.173932722962597e-06,
+      "loss": 1.124,
+      "step": 4688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5132928646127968,
+      "learning_rate": 7.172762629084821e-06,
+      "loss": 0.8025,
+      "step": 4689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5042329740151752,
+      "learning_rate": 7.171592388497144e-06,
+      "loss": 0.8413,
+      "step": 4690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5426032993217735,
+      "learning_rate": 7.170422001278583e-06,
+      "loss": 0.8084,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5470401683051946,
+      "learning_rate": 7.169251467508165e-06,
+      "loss": 0.867,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4718186752424467,
+      "learning_rate": 7.168080787264927e-06,
+      "loss": 0.7206,
+      "step": 4693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4615703133144988,
+      "learning_rate": 7.166909960627918e-06,
+      "loss": 0.7467,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6107050589118383,
+      "learning_rate": 7.165738987676193e-06,
+      "loss": 0.8347,
+      "step": 4695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4203285280556759,
+      "learning_rate": 7.16456786848882e-06,
+      "loss": 0.7266,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6780348232938431,
+      "learning_rate": 7.163396603144874e-06,
+      "loss": 0.8072,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.553866768747694,
+      "learning_rate": 7.162225191723442e-06,
+      "loss": 0.8635,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4633860668805754,
+      "learning_rate": 7.161053634303622e-06,
+      "loss": 0.8355,
+      "step": 4699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5695228071891096,
+      "learning_rate": 7.159881930964518e-06,
+      "loss": 0.768,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5521824893609644,
+      "learning_rate": 7.15871008178525e-06,
+      "loss": 0.791,
+      "step": 4701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5709941451027039,
+      "learning_rate": 7.157538086844937e-06,
+      "loss": 0.7757,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9355185611724404,
+      "learning_rate": 7.156365946222721e-06,
+      "loss": 1.0996,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8949272571348935,
+      "learning_rate": 7.155193659997746e-06,
+      "loss": 1.1632,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4633614874554774,
+      "learning_rate": 7.154021228249165e-06,
+      "loss": 0.7305,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4716449478530491,
+      "learning_rate": 7.1528486510561455e-06,
+      "loss": 0.8039,
+      "step": 4706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4075663371076896,
+      "learning_rate": 7.151675928497864e-06,
+      "loss": 0.7332,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4367493760369496,
+      "learning_rate": 7.150503060653504e-06,
+      "loss": 0.6884,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4724479010673097,
+      "learning_rate": 7.149330047602258e-06,
+      "loss": 0.8272,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.537058225943324,
+      "learning_rate": 7.148156889423332e-06,
+      "loss": 0.7532,
+      "step": 4710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5530247571541198,
+      "learning_rate": 7.146983586195942e-06,
+      "loss": 0.7302,
+      "step": 4711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3649103684693964,
+      "learning_rate": 7.145810137999312e-06,
+      "loss": 0.795,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.389736103084445,
+      "learning_rate": 7.144636544912674e-06,
+      "loss": 0.7768,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5723482335448258,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.8262,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.446910255274307,
+      "learning_rate": 7.142288924386359e-06,
+      "loss": 0.6704,
+      "step": 4715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1355781753960117,
+      "learning_rate": 7.141114897105202e-06,
+      "loss": 1.0961,
+      "step": 4716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4805296464591282,
+      "learning_rate": 7.1399407252510685e-06,
+      "loss": 0.7922,
+      "step": 4717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4235596442475063,
+      "learning_rate": 7.138766408903246e-06,
+      "loss": 0.7017,
+      "step": 4718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4978706236664283,
+      "learning_rate": 7.137591948141023e-06,
+      "loss": 0.813,
+      "step": 4719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.506171555570857,
+      "learning_rate": 7.136417343043704e-06,
+      "loss": 0.8295,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.434289035553792,
+      "learning_rate": 7.135242593690601e-06,
+      "loss": 0.7568,
+      "step": 4721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8670395928025799,
+      "learning_rate": 7.134067700161037e-06,
+      "loss": 1.1006,
+      "step": 4722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6809299928443586,
+      "learning_rate": 7.132892662534342e-06,
+      "loss": 0.8345,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4685430165054532,
+      "learning_rate": 7.131717480889854e-06,
+      "loss": 0.7915,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8272451609205067,
+      "learning_rate": 7.130542155306931e-06,
+      "loss": 1.1014,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8212032911436201,
+      "learning_rate": 7.129366685864928e-06,
+      "loss": 1.0921,
+      "step": 4726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.457361089919343,
+      "learning_rate": 7.128191072643217e-06,
+      "loss": 0.7161,
+      "step": 4727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3906867190451795,
+      "learning_rate": 7.127015315721179e-06,
+      "loss": 0.7546,
+      "step": 4728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5614937817137227,
+      "learning_rate": 7.125839415178204e-06,
+      "loss": 0.7939,
+      "step": 4729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6030349819770577,
+      "learning_rate": 7.124663371093691e-06,
+      "loss": 0.7045,
+      "step": 4730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5466153661628774,
+      "learning_rate": 7.123487183547046e-06,
+      "loss": 0.8282,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1866171752776085,
+      "learning_rate": 7.122310852617693e-06,
+      "loss": 1.1273,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5341601744034048,
+      "learning_rate": 7.121134378385057e-06,
+      "loss": 0.7352,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4835260697088204,
+      "learning_rate": 7.11995776092858e-06,
+      "loss": 0.8545,
+      "step": 4734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.525620594644031,
+      "learning_rate": 7.118781000327706e-06,
+      "loss": 0.8009,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4810109557536222,
+      "learning_rate": 7.1176040966618934e-06,
+      "loss": 0.7055,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8000234904671825,
+      "learning_rate": 7.116427050010611e-06,
+      "loss": 1.1064,
+      "step": 4737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.470231486794984,
+      "learning_rate": 7.115249860453333e-06,
+      "loss": 0.8462,
+      "step": 4738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6912256651903552,
+      "learning_rate": 7.114072528069549e-06,
+      "loss": 0.8138,
+      "step": 4739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4814240492003237,
+      "learning_rate": 7.1128950529387534e-06,
+      "loss": 0.7689,
+      "step": 4740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3926488617658959,
+      "learning_rate": 7.1117174351404515e-06,
+      "loss": 0.6873,
+      "step": 4741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9041495388596472,
+      "learning_rate": 7.11053967475416e-06,
+      "loss": 0.8173,
+      "step": 4742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4946878942453439,
+      "learning_rate": 7.109361771859404e-06,
+      "loss": 0.7981,
+      "step": 4743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6257482112000994,
+      "learning_rate": 7.1081837265357174e-06,
+      "loss": 0.7647,
+      "step": 4744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4471693455954073,
+      "learning_rate": 7.107005538862647e-06,
+      "loss": 0.7424,
+      "step": 4745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8689533635635018,
+      "learning_rate": 7.1058272089197425e-06,
+      "loss": 1.0694,
+      "step": 4746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5399021731342382,
+      "learning_rate": 7.10464873678657e-06,
+      "loss": 0.7775,
+      "step": 4747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4894470323079931,
+      "learning_rate": 7.103470122542701e-06,
+      "loss": 0.7631,
+      "step": 4748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5982741505583355,
+      "learning_rate": 7.1022913662677225e-06,
+      "loss": 0.8138,
+      "step": 4749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4693654107056158,
+      "learning_rate": 7.101112468041221e-06,
+      "loss": 0.7953,
+      "step": 4750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5090567344523407,
+      "learning_rate": 7.099933427942803e-06,
+      "loss": 0.7381,
+      "step": 4751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4597634931933907,
+      "learning_rate": 7.098754246052077e-06,
+      "loss": 0.7365,
+      "step": 4752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3822243525695932,
+      "learning_rate": 7.0975749224486665e-06,
+      "loss": 0.7496,
+      "step": 4753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.428731981452327,
+      "learning_rate": 7.0963954572122e-06,
+      "loss": 0.7087,
+      "step": 4754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6130815642702456,
+      "learning_rate": 7.095215850422318e-06,
+      "loss": 0.745,
+      "step": 4755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5174966208222809,
+      "learning_rate": 7.094036102158672e-06,
+      "loss": 0.8015,
+      "step": 4756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5382952469387174,
+      "learning_rate": 7.0928562125009195e-06,
+      "loss": 0.7654,
+      "step": 4757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5587395342791164,
+      "learning_rate": 7.091676181528729e-06,
+      "loss": 0.7727,
+      "step": 4758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5500282452413885,
+      "learning_rate": 7.090496009321781e-06,
+      "loss": 0.7891,
+      "step": 4759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5577171672752854,
+      "learning_rate": 7.089315695959762e-06,
+      "loss": 0.886,
+      "step": 4760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5881840640509954,
+      "learning_rate": 7.088135241522369e-06,
+      "loss": 0.8257,
+      "step": 4761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5868128760290876,
+      "learning_rate": 7.086954646089311e-06,
+      "loss": 0.8228,
+      "step": 4762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8968101850521635,
+      "learning_rate": 7.085773909740302e-06,
+      "loss": 1.1076,
+      "step": 4763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4900700589273637,
+      "learning_rate": 7.084593032555071e-06,
+      "loss": 0.6934,
+      "step": 4764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4506193549529383,
+      "learning_rate": 7.083412014613349e-06,
+      "loss": 0.7996,
+      "step": 4765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3377442677604336,
+      "learning_rate": 7.082230855994885e-06,
+      "loss": 0.7271,
+      "step": 4766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5374098790903197,
+      "learning_rate": 7.081049556779433e-06,
+      "loss": 0.8265,
+      "step": 4767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.474164789341235,
+      "learning_rate": 7.079868117046755e-06,
+      "loss": 0.7387,
+      "step": 4768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8279594969063385,
+      "learning_rate": 7.078686536876627e-06,
+      "loss": 1.0865,
+      "step": 4769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.512197022132002,
+      "learning_rate": 7.07750481634883e-06,
+      "loss": 0.7947,
+      "step": 4770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3514133208397847,
+      "learning_rate": 7.076322955543158e-06,
+      "loss": 0.7407,
+      "step": 4771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8299497638856991,
+      "learning_rate": 7.075140954539412e-06,
+      "loss": 1.1444,
+      "step": 4772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6560337512722119,
+      "learning_rate": 7.073958813417404e-06,
+      "loss": 0.8352,
+      "step": 4773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5703790226769188,
+      "learning_rate": 7.0727765322569545e-06,
+      "loss": 0.8296,
+      "step": 4774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5669459823343077,
+      "learning_rate": 7.071594111137892e-06,
+      "loss": 0.8121,
+      "step": 4775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8153017556867063,
+      "learning_rate": 7.07041155014006e-06,
+      "loss": 1.0916,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4907896201975301,
+      "learning_rate": 7.069228849343306e-06,
+      "loss": 0.8128,
+      "step": 4777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6162074965184496,
+      "learning_rate": 7.0680460088274885e-06,
+      "loss": 0.7233,
+      "step": 4778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4605572775121975,
+      "learning_rate": 7.066863028672475e-06,
+      "loss": 0.7819,
+      "step": 4779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7987014830552988,
+      "learning_rate": 7.065679908958143e-06,
+      "loss": 1.1055,
+      "step": 4780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5975690420650885,
+      "learning_rate": 7.064496649764381e-06,
+      "loss": 0.7966,
+      "step": 4781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3845048007180782,
+      "learning_rate": 7.063313251171084e-06,
+      "loss": 0.7921,
+      "step": 4782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9081352704701624,
+      "learning_rate": 7.062129713258159e-06,
+      "loss": 1.1293,
+      "step": 4783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5430158381729853,
+      "learning_rate": 7.06094603610552e-06,
+      "loss": 0.764,
+      "step": 4784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4742317028522816,
+      "learning_rate": 7.059762219793091e-06,
+      "loss": 0.8065,
+      "step": 4785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.471919419726245,
+      "learning_rate": 7.05857826440081e-06,
+      "loss": 0.7715,
+      "step": 4786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.449370585713447,
+      "learning_rate": 7.057394170008614e-06,
+      "loss": 0.8474,
+      "step": 4787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4181539216523533,
+      "learning_rate": 7.056209936696461e-06,
+      "loss": 0.8036,
+      "step": 4788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.963546690739823,
+      "learning_rate": 7.055025564544311e-06,
+      "loss": 1.1111,
+      "step": 4789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.454988992424918,
+      "learning_rate": 7.053841053632135e-06,
+      "loss": 0.8206,
+      "step": 4790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4717532136633846,
+      "learning_rate": 7.052656404039915e-06,
+      "loss": 0.7579,
+      "step": 4791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5259938487117404,
+      "learning_rate": 7.0514716158476405e-06,
+      "loss": 0.8525,
+      "step": 4792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4337201588967021,
+      "learning_rate": 7.050286689135313e-06,
+      "loss": 0.8067,
+      "step": 4793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.769858358906952,
+      "learning_rate": 7.049101623982938e-06,
+      "loss": 0.7919,
+      "step": 4794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8595700059558524,
+      "learning_rate": 7.0479164204705376e-06,
+      "loss": 0.7871,
+      "step": 4795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5925553079726997,
+      "learning_rate": 7.046731078678137e-06,
+      "loss": 0.8546,
+      "step": 4796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4912729092977195,
+      "learning_rate": 7.0455455986857724e-06,
+      "loss": 0.7855,
+      "step": 4797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.478574203483191,
+      "learning_rate": 7.044359980573494e-06,
+      "loss": 0.739,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5488038520271574,
+      "learning_rate": 7.043174224421353e-06,
+      "loss": 0.8579,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4901876969334749,
+      "learning_rate": 7.041988330309417e-06,
+      "loss": 0.7675,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3958974175486691,
+      "learning_rate": 7.040802298317762e-06,
+      "loss": 0.8259,
+      "step": 4801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5901434476883838,
+      "learning_rate": 7.039616128526465e-06,
+      "loss": 0.744,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4864627747289088,
+      "learning_rate": 7.038429821015627e-06,
+      "loss": 0.7929,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4634583622816204,
+      "learning_rate": 7.037243375865344e-06,
+      "loss": 0.7793,
+      "step": 4804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.498476364278755,
+      "learning_rate": 7.03605679315573e-06,
+      "loss": 0.8505,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5023445431357645,
+      "learning_rate": 7.034870072966906e-06,
+      "loss": 0.7958,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5428843966554227,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.7326,
+      "step": 4807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.846428854483477,
+      "learning_rate": 7.032496220472157e-06,
+      "loss": 1.0755,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5647307369161931,
+      "learning_rate": 7.031309088326519e-06,
+      "loss": 0.8288,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5753474452757965,
+      "learning_rate": 7.030121819022247e-06,
+      "loss": 0.8789,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4888670742947328,
+      "learning_rate": 7.028934412639508e-06,
+      "loss": 0.7963,
+      "step": 4811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5211996431023533,
+      "learning_rate": 7.027746869258477e-06,
+      "loss": 0.8492,
+      "step": 4812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.513409449787587,
+      "learning_rate": 7.026559188959341e-06,
+      "loss": 0.8311,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9484344368646307,
+      "learning_rate": 7.025371371822294e-06,
+      "loss": 0.8402,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.463577453332081,
+      "learning_rate": 7.024183417927542e-06,
+      "loss": 0.7955,
+      "step": 4815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5854670313746038,
+      "learning_rate": 7.022995327355296e-06,
+      "loss": 0.7536,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3542936232000855,
+      "learning_rate": 7.02180710018578e-06,
+      "loss": 0.7971,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6415891605027393,
+      "learning_rate": 7.0206187364992255e-06,
+      "loss": 0.7249,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.494184725537653,
+      "learning_rate": 7.0194302363758735e-06,
+      "loss": 0.7803,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8446516574572622,
+      "learning_rate": 7.018241599895974e-06,
+      "loss": 1.1093,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5311126234558692,
+      "learning_rate": 7.017052827139786e-06,
+      "loss": 0.6977,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5457996914169605,
+      "learning_rate": 7.0158639181875795e-06,
+      "loss": 0.7429,
+      "step": 4822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8032721028539408,
+      "learning_rate": 7.014674873119634e-06,
+      "loss": 1.1343,
+      "step": 4823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5619961832735656,
+      "learning_rate": 7.013485692016232e-06,
+      "loss": 0.8616,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5480958162366456,
+      "learning_rate": 7.012296374957671e-06,
+      "loss": 0.7769,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4724327746348307,
+      "learning_rate": 7.011106922024258e-06,
+      "loss": 0.6519,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8745714462995694,
+      "learning_rate": 7.009917333296308e-06,
+      "loss": 1.132,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4435004302095549,
+      "learning_rate": 7.0087276088541435e-06,
+      "loss": 0.8217,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.462227384167135,
+      "learning_rate": 7.007537748778097e-06,
+      "loss": 0.7958,
+      "step": 4829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4893170218551093,
+      "learning_rate": 7.006347753148511e-06,
+      "loss": 0.7683,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1703738121094354,
+      "learning_rate": 7.005157622045737e-06,
+      "loss": 0.7549,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6987051860533304,
+      "learning_rate": 7.0039673555501365e-06,
+      "loss": 0.6992,
+      "step": 4832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4696003345374853,
+      "learning_rate": 7.002776953742078e-06,
+      "loss": 0.7719,
+      "step": 4833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4905628923933092,
+      "learning_rate": 7.001586416701939e-06,
+      "loss": 0.7616,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6590880696017936,
+      "learning_rate": 7.000395744510107e-06,
+      "loss": 0.718,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5732705361025423,
+      "learning_rate": 6.9992049372469815e-06,
+      "loss": 0.8023,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.455824718157807,
+      "learning_rate": 6.998013994992967e-06,
+      "loss": 0.7086,
+      "step": 4837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8571724559128869,
+      "learning_rate": 6.9968229178284775e-06,
+      "loss": 1.1274,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8136378075900033,
+      "learning_rate": 6.995631705833942e-06,
+      "loss": 1.113,
+      "step": 4839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5825325181630443,
+      "learning_rate": 6.994440359089787e-06,
+      "loss": 0.8577,
+      "step": 4840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8041988510822367,
+      "learning_rate": 6.99324887767646e-06,
+      "loss": 0.7573,
+      "step": 4841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7881952416451836,
+      "learning_rate": 6.9920572616744096e-06,
+      "loss": 1.0923,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7929672627501072,
+      "learning_rate": 6.9908655111640984e-06,
+      "loss": 1.095,
+      "step": 4843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4642194256795575,
+      "learning_rate": 6.989673626225997e-06,
+      "loss": 0.8323,
+      "step": 4844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5139854054672122,
+      "learning_rate": 6.988481606940582e-06,
+      "loss": 0.7803,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5765629689516845,
+      "learning_rate": 6.9872894533883415e-06,
+      "loss": 0.8097,
+      "step": 4846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5488066037830295,
+      "learning_rate": 6.986097165649772e-06,
+      "loss": 0.8177,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4722389476083872,
+      "learning_rate": 6.984904743805383e-06,
+      "loss": 0.7389,
+      "step": 4848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4324636595180675,
+      "learning_rate": 6.983712187935684e-06,
+      "loss": 0.7645,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.478334349905141,
+      "learning_rate": 6.982519498121204e-06,
+      "loss": 0.7917,
+      "step": 4850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5258947025611473,
+      "learning_rate": 6.981326674442474e-06,
+      "loss": 0.8413,
+      "step": 4851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.631648229609166,
+      "learning_rate": 6.980133716980035e-06,
+      "loss": 0.7732,
+      "step": 4852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5769998757385868,
+      "learning_rate": 6.978940625814441e-06,
+      "loss": 0.6912,
+      "step": 4853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5251288646524548,
+      "learning_rate": 6.977747401026249e-06,
+      "loss": 0.8208,
+      "step": 4854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5092084157227337,
+      "learning_rate": 6.9765540426960334e-06,
+      "loss": 0.8071,
+      "step": 4855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4912611488464829,
+      "learning_rate": 6.9753605509043665e-06,
+      "loss": 0.7391,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6416057697280744,
+      "learning_rate": 6.974166925731839e-06,
+      "loss": 0.8117,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5284424000546486,
+      "learning_rate": 6.972973167259046e-06,
+      "loss": 0.7248,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4648419973114,
+      "learning_rate": 6.971779275566593e-06,
+      "loss": 0.7482,
+      "step": 4859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4496584683804032,
+      "learning_rate": 6.9705852507350945e-06,
+      "loss": 0.8385,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5411292158803716,
+      "learning_rate": 6.969391092845177e-06,
+      "loss": 0.8319,
+      "step": 4861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9505968468371127,
+      "learning_rate": 6.968196801977466e-06,
+      "loss": 1.0884,
+      "step": 4862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6866597226454823,
+      "learning_rate": 6.967002378212608e-06,
+      "loss": 0.7212,
+      "step": 4863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.432594992913711,
+      "learning_rate": 6.965807821631251e-06,
+      "loss": 0.7808,
+      "step": 4864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4827826555361951,
+      "learning_rate": 6.964613132314055e-06,
+      "loss": 0.774,
+      "step": 4865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.532949788111831,
+      "learning_rate": 6.963418310341688e-06,
+      "loss": 0.837,
+      "step": 4866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4303334914568375,
+      "learning_rate": 6.962223355794827e-06,
+      "loss": 0.7445,
+      "step": 4867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6820797250385324,
+      "learning_rate": 6.961028268754159e-06,
+      "loss": 0.7739,
+      "step": 4868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8017476268988318,
+      "learning_rate": 6.959833049300376e-06,
+      "loss": 1.1255,
+      "step": 4869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.420187307305575,
+      "learning_rate": 6.958637697514186e-06,
+      "loss": 0.7809,
+      "step": 4870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.522283390003779,
+      "learning_rate": 6.957442213476299e-06,
+      "loss": 0.7384,
+      "step": 4871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5408893494416438,
+      "learning_rate": 6.956246597267438e-06,
+      "loss": 0.7645,
+      "step": 4872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7948083027528613,
+      "learning_rate": 6.955050848968334e-06,
+      "loss": 1.126,
+      "step": 4873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4391041997251297,
+      "learning_rate": 6.953854968659726e-06,
+      "loss": 0.7539,
+      "step": 4874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5595289857553398,
+      "learning_rate": 6.952658956422362e-06,
+      "loss": 0.7451,
+      "step": 4875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5370544228553753,
+      "learning_rate": 6.951462812337e-06,
+      "loss": 0.8079,
+      "step": 4876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5324439334635023,
+      "learning_rate": 6.950266536484408e-06,
+      "loss": 0.8161,
+      "step": 4877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8177139320477639,
+      "learning_rate": 6.94907012894536e-06,
+      "loss": 1.1288,
+      "step": 4878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5959775910861793,
+      "learning_rate": 6.947873589800637e-06,
+      "loss": 0.6745,
+      "step": 4879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6021370765210248,
+      "learning_rate": 6.946676919131039e-06,
+      "loss": 0.8935,
+      "step": 4880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4972463458560592,
+      "learning_rate": 6.945480117017362e-06,
+      "loss": 0.8267,
+      "step": 4881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7789943336075468,
+      "learning_rate": 6.94428318354042e-06,
+      "loss": 1.0786,
+      "step": 4882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4531248816760898,
+      "learning_rate": 6.943086118781032e-06,
+      "loss": 0.7368,
+      "step": 4883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7913345565827008,
+      "learning_rate": 6.941888922820023e-06,
+      "loss": 1.0832,
+      "step": 4884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5252456853733112,
+      "learning_rate": 6.940691595738237e-06,
+      "loss": 0.7606,
+      "step": 4885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.414901232185457,
+      "learning_rate": 6.939494137616515e-06,
+      "loss": 0.8687,
+      "step": 4886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4060458739046064,
+      "learning_rate": 6.938296548535714e-06,
+      "loss": 0.7735,
+      "step": 4887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4155585347914468,
+      "learning_rate": 6.937098828576699e-06,
+      "loss": 0.7896,
+      "step": 4888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.557774313037611,
+      "learning_rate": 6.935900977820341e-06,
+      "loss": 0.7633,
+      "step": 4889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3852216430455788,
+      "learning_rate": 6.934702996347522e-06,
+      "loss": 0.7178,
+      "step": 4890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8327437670664632,
+      "learning_rate": 6.933504884239133e-06,
+      "loss": 1.0711,
+      "step": 4891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4490893978345991,
+      "learning_rate": 6.932306641576073e-06,
+      "loss": 0.7439,
+      "step": 4892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8018367696285715,
+      "learning_rate": 6.93110826843925e-06,
+      "loss": 1.0943,
+      "step": 4893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7591132884052368,
+      "learning_rate": 6.929909764909582e-06,
+      "loss": 1.076,
+      "step": 4894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.524573127851954,
+      "learning_rate": 6.928711131067992e-06,
+      "loss": 0.7505,
+      "step": 4895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7619258065974707,
+      "learning_rate": 6.927512366995416e-06,
+      "loss": 1.1199,
+      "step": 4896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5051605691761225,
+      "learning_rate": 6.926313472772799e-06,
+      "loss": 0.8038,
+      "step": 4897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6118386413258803,
+      "learning_rate": 6.925114448481089e-06,
+      "loss": 0.8371,
+      "step": 4898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5718790716829274,
+      "learning_rate": 6.923915294201252e-06,
+      "loss": 0.8059,
+      "step": 4899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5676217912142556,
+      "learning_rate": 6.922716010014256e-06,
+      "loss": 0.7753,
+      "step": 4900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5272250925703863,
+      "learning_rate": 6.921516596001075e-06,
+      "loss": 0.7547,
+      "step": 4901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7900271050921532,
+      "learning_rate": 6.920317052242702e-06,
+      "loss": 0.8723,
+      "step": 4902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5166296826673427,
+      "learning_rate": 6.919117378820129e-06,
+      "loss": 0.6612,
+      "step": 4903
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5228404032072853,
+      "learning_rate": 6.917917575814364e-06,
+      "loss": 0.8532,
+      "step": 4904
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.553375936272835,
+      "learning_rate": 6.9167176433064175e-06,
+      "loss": 0.8056,
+      "step": 4905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4465161832955624,
+      "learning_rate": 6.915517581377314e-06,
+      "loss": 0.8077,
+      "step": 4906
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6624152013665454,
+      "learning_rate": 6.914317390108082e-06,
+      "loss": 0.7941,
+      "step": 4907
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.477955459907076,
+      "learning_rate": 6.913117069579763e-06,
+      "loss": 0.8156,
+      "step": 4908
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4649758545419018,
+      "learning_rate": 6.9119166198734046e-06,
+      "loss": 0.7713,
+      "step": 4909
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5019665174134,
+      "learning_rate": 6.910716041070064e-06,
+      "loss": 0.6512,
+      "step": 4910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5352783441953315,
+      "learning_rate": 6.909515333250809e-06,
+      "loss": 0.8194,
+      "step": 4911
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6799420995401015,
+      "learning_rate": 6.90831449649671e-06,
+      "loss": 0.7191,
+      "step": 4912
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5334128424109752,
+      "learning_rate": 6.907113530888853e-06,
+      "loss": 0.7335,
+      "step": 4913
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.507613981875418,
+      "learning_rate": 6.905912436508331e-06,
+      "loss": 0.8172,
+      "step": 4914
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9586626567819123,
+      "learning_rate": 6.904711213436241e-06,
+      "loss": 1.1071,
+      "step": 4915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.532772455292424,
+      "learning_rate": 6.903509861753695e-06,
+      "loss": 0.8972,
+      "step": 4916
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4785724245233158,
+      "learning_rate": 6.902308381541812e-06,
+      "loss": 0.7494,
+      "step": 4917
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5755638357587924,
+      "learning_rate": 6.901106772881716e-06,
+      "loss": 0.9011,
+      "step": 4918
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5502723445917514,
+      "learning_rate": 6.899905035854544e-06,
+      "loss": 0.8711,
+      "step": 4919
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5695534979964207,
+      "learning_rate": 6.898703170541439e-06,
+      "loss": 0.8299,
+      "step": 4920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5049259045245127,
+      "learning_rate": 6.897501177023556e-06,
+      "loss": 0.6952,
+      "step": 4921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4392622483846549,
+      "learning_rate": 6.896299055382053e-06,
+      "loss": 0.7493,
+      "step": 4922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6892054460585515,
+      "learning_rate": 6.895096805698103e-06,
+      "loss": 0.7811,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6538163559410586,
+      "learning_rate": 6.893894428052881e-06,
+      "loss": 0.7638,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.868621555216517,
+      "learning_rate": 6.892691922527576e-06,
+      "loss": 1.1282,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.460657709351587,
+      "learning_rate": 6.891489289203388e-06,
+      "loss": 0.7836,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5780388159636856,
+      "learning_rate": 6.890286528161516e-06,
+      "loss": 0.7835,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.404558738521302,
+      "learning_rate": 6.889083639483176e-06,
+      "loss": 0.8381,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8160851761888349,
+      "learning_rate": 6.887880623249589e-06,
+      "loss": 1.0954,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5647038494330388,
+      "learning_rate": 6.886677479541984e-06,
+      "loss": 0.7225,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7952371249376197,
+      "learning_rate": 6.885474208441602e-06,
+      "loss": 1.1073,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7840895137168205,
+      "learning_rate": 6.88427081002969e-06,
+      "loss": 1.1003,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4466921777254216,
+      "learning_rate": 6.883067284387505e-06,
+      "loss": 0.6985,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7905802827730208,
+      "learning_rate": 6.881863631596313e-06,
+      "loss": 1.0844,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6754842274060233,
+      "learning_rate": 6.880659851737384e-06,
+      "loss": 0.8007,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5027688591855028,
+      "learning_rate": 6.879455944892e-06,
+      "loss": 0.7553,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5257269544408887,
+      "learning_rate": 6.8782519111414515e-06,
+      "loss": 0.7812,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.512371522133477,
+      "learning_rate": 6.877047750567042e-06,
+      "loss": 0.8606,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4979082648413515,
+      "learning_rate": 6.8758434632500756e-06,
+      "loss": 0.8904,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4704267951599195,
+      "learning_rate": 6.8746390492718695e-06,
+      "loss": 0.689,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4972714208887041,
+      "learning_rate": 6.873434508713748e-06,
+      "loss": 0.7085,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8640609018953209,
+      "learning_rate": 6.872229841657043e-06,
+      "loss": 1.1167,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4223380690795144,
+      "learning_rate": 6.8710250481831e-06,
+      "loss": 0.7326,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4298978336561585,
+      "learning_rate": 6.869820128373267e-06,
+      "loss": 0.7181,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4308609741654195,
+      "learning_rate": 6.868615082308904e-06,
+      "loss": 0.7658,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4487943382723723,
+      "learning_rate": 6.867409910071376e-06,
+      "loss": 0.7761,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4537968564964712,
+      "learning_rate": 6.866204611742062e-06,
+      "loss": 0.7442,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4652358873007205,
+      "learning_rate": 6.864999187402343e-06,
+      "loss": 0.7698,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5910199613327107,
+      "learning_rate": 6.863793637133618e-06,
+      "loss": 0.7808,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4286914552163246,
+      "learning_rate": 6.862587961017283e-06,
+      "loss": 0.6706,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.569757434781543,
+      "learning_rate": 6.86138215913475e-06,
+      "loss": 0.7655,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.506536875878353,
+      "learning_rate": 6.860176231567437e-06,
+      "loss": 0.7916,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5274987368730564,
+      "learning_rate": 6.858970178396771e-06,
+      "loss": 0.8464,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5170822114946045,
+      "learning_rate": 6.857763999704188e-06,
+      "loss": 0.7187,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4906290747063256,
+      "learning_rate": 6.856557695571131e-06,
+      "loss": 0.7635,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4693273909686388,
+      "learning_rate": 6.855351266079056e-06,
+      "loss": 0.8361,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4352476252446822,
+      "learning_rate": 6.854144711309418e-06,
+      "loss": 0.823,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4138985772503154,
+      "learning_rate": 6.85293803134369e-06,
+      "loss": 0.7002,
+      "step": 4958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6037479745766512,
+      "learning_rate": 6.851731226263348e-06,
+      "loss": 0.8248,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4569830268142927,
+      "learning_rate": 6.8505242961498816e-06,
+      "loss": 0.7894,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5489351476247422,
+      "learning_rate": 6.849317241084783e-06,
+      "loss": 0.8321,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5029234600317602,
+      "learning_rate": 6.848110061149555e-06,
+      "loss": 0.8559,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3631777728324188,
+      "learning_rate": 6.846902756425709e-06,
+      "loss": 0.728,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5879121380838839,
+      "learning_rate": 6.845695326994768e-06,
+      "loss": 0.8068,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8552289082094887,
+      "learning_rate": 6.844487772938255e-06,
+      "loss": 1.0985,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8034991595012907,
+      "learning_rate": 6.843280094337712e-06,
+      "loss": 1.1268,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5755477126193738,
+      "learning_rate": 6.842072291274681e-06,
+      "loss": 0.8266,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4959580774978334,
+      "learning_rate": 6.840864363830718e-06,
+      "loss": 0.8185,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8336096238418472,
+      "learning_rate": 6.839656312087384e-06,
+      "loss": 1.0631,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4493704299463315,
+      "learning_rate": 6.838448136126247e-06,
+      "loss": 0.8427,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5561195223171072,
+      "learning_rate": 6.837239836028889e-06,
+      "loss": 0.7969,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4970766903639408,
+      "learning_rate": 6.836031411876898e-06,
+      "loss": 0.7874,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5607115528740678,
+      "learning_rate": 6.834822863751864e-06,
+      "loss": 0.8213,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8745393421365627,
+      "learning_rate": 6.833614191735398e-06,
+      "loss": 1.0678,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.968332827726,
+      "learning_rate": 6.832405395909107e-06,
+      "loss": 0.8423,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4197418592926685,
+      "learning_rate": 6.831196476354615e-06,
+      "loss": 0.7379,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5020517325316227,
+      "learning_rate": 6.829987433153549e-06,
+      "loss": 0.728,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.466772375003132,
+      "learning_rate": 6.828778266387547e-06,
+      "loss": 0.7998,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4056696695988935,
+      "learning_rate": 6.827568976138255e-06,
+      "loss": 0.7544,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5391505366589904,
+      "learning_rate": 6.826359562487326e-06,
+      "loss": 0.7788,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5405905775449409,
+      "learning_rate": 6.825150025516423e-06,
+      "loss": 0.8438,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.454214656151131,
+      "learning_rate": 6.823940365307217e-06,
+      "loss": 0.823,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5229918925797552,
+      "learning_rate": 6.822730581941388e-06,
+      "loss": 0.8193,
+      "step": 4983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4828640557193797,
+      "learning_rate": 6.8215206755006214e-06,
+      "loss": 0.7979,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.006148628177163,
+      "learning_rate": 6.820310646066613e-06,
+      "loss": 1.0742,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.476194304258027,
+      "learning_rate": 6.819100493721068e-06,
+      "loss": 0.7482,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.554574162394729,
+      "learning_rate": 6.817890218545697e-06,
+      "loss": 0.8601,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4343588507991312,
+      "learning_rate": 6.816679820622223e-06,
+      "loss": 0.7558,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5545987578806304,
+      "learning_rate": 6.815469300032374e-06,
+      "loss": 0.7503,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3877945325380683,
+      "learning_rate": 6.814258656857885e-06,
+      "loss": 0.8107,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3741605303519229,
+      "learning_rate": 6.8130478911805044e-06,
+      "loss": 0.7256,
+      "step": 4991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8447756237278693,
+      "learning_rate": 6.811837003081983e-06,
+      "loss": 1.0847,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8507802574544433,
+      "learning_rate": 6.810625992644085e-06,
+      "loss": 1.1196,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3929417299987816,
+      "learning_rate": 6.809414859948579e-06,
+      "loss": 0.7948,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5415217409864312,
+      "learning_rate": 6.808203605077244e-06,
+      "loss": 0.8411,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8037984952101191,
+      "learning_rate": 6.806992228111868e-06,
+      "loss": 1.088,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5826994220395743,
+      "learning_rate": 6.805780729134244e-06,
+      "loss": 0.7394,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.495436723031971,
+      "learning_rate": 6.804569108226176e-06,
+      "loss": 0.7921,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4014224918813896,
+      "learning_rate": 6.803357365469475e-06,
+      "loss": 0.7566,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.491679971724398,
+      "learning_rate": 6.802145500945962e-06,
+      "loss": 0.7684,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4480147662112097,
+      "learning_rate": 6.800933514737465e-06,
+      "loss": 0.7929,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4591330477025146,
+      "learning_rate": 6.7997214069258166e-06,
+      "loss": 0.724,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4047769858337935,
+      "learning_rate": 6.7985091775928646e-06,
+      "loss": 0.752,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9396517858083783,
+      "learning_rate": 6.79729682682046e-06,
+      "loss": 1.11,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5423142191357428,
+      "learning_rate": 6.796084354690465e-06,
+      "loss": 0.7244,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5397871889947181,
+      "learning_rate": 6.794871761284747e-06,
+      "loss": 0.8116,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6258564458620668,
+      "learning_rate": 6.793659046685182e-06,
+      "loss": 0.7877,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5813999579260771,
+      "learning_rate": 6.792446210973658e-06,
+      "loss": 0.7602,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4805218712136188,
+      "learning_rate": 6.791233254232066e-06,
+      "loss": 0.7801,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4773935576623607,
+      "learning_rate": 6.79002017654231e-06,
+      "loss": 0.7622,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7277716737894064,
+      "learning_rate": 6.7888069779863e-06,
+      "loss": 0.8064,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4792453256601847,
+      "learning_rate": 6.787593658645949e-06,
+      "loss": 0.7778,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5144915658159812,
+      "learning_rate": 6.786380218603189e-06,
+      "loss": 0.7453,
+      "step": 5013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8926871605833002,
+      "learning_rate": 6.78516665793995e-06,
+      "loss": 1.0838,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.407776356341224,
+      "learning_rate": 6.7839529767381785e-06,
+      "loss": 0.8013,
+      "step": 5015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4287730840964246,
+      "learning_rate": 6.7827391750798225e-06,
+      "loss": 0.8443,
+      "step": 5016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5261468235859288,
+      "learning_rate": 6.781525253046839e-06,
+      "loss": 0.8063,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7038505933217547,
+      "learning_rate": 6.780311210721198e-06,
+      "loss": 0.8744,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3902740918030598,
+      "learning_rate": 6.779097048184873e-06,
+      "loss": 0.7427,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5540586308532773,
+      "learning_rate": 6.777882765519846e-06,
+      "loss": 0.7996,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8541748684459625,
+      "learning_rate": 6.776668362808111e-06,
+      "loss": 1.1154,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5063528573963116,
+      "learning_rate": 6.775453840131666e-06,
+      "loss": 0.7627,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5020249872902989,
+      "learning_rate": 6.774239197572516e-06,
+      "loss": 0.7419,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4502692488082431,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.6662,
+      "step": 5024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.396223679869627,
+      "learning_rate": 6.771809553134178e-06,
+      "loss": 0.7312,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4738823954155023,
+      "learning_rate": 6.770594551419044e-06,
+      "loss": 0.7917,
+      "step": 5026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5978441721322767,
+      "learning_rate": 6.769379430149318e-06,
+      "loss": 0.7563,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4705464045679253,
+      "learning_rate": 6.768164189407047e-06,
+      "loss": 0.8256,
+      "step": 5028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6296459695063903,
+      "learning_rate": 6.766948829274286e-06,
+      "loss": 0.8454,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5154801134495768,
+      "learning_rate": 6.7657333498331e-06,
+      "loss": 0.7218,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4101027917915399,
+      "learning_rate": 6.76451775116556e-06,
+      "loss": 0.8047,
+      "step": 5031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5404248899563413,
+      "learning_rate": 6.763302033353748e-06,
+      "loss": 0.7869,
+      "step": 5032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5326165653874584,
+      "learning_rate": 6.7620861964797505e-06,
+      "loss": 0.8152,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5470521736685614,
+      "learning_rate": 6.760870240625663e-06,
+      "loss": 0.7296,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6429712866818738,
+      "learning_rate": 6.75965416587359e-06,
+      "loss": 0.8103,
+      "step": 5035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9001982270018848,
+      "learning_rate": 6.758437972305645e-06,
+      "loss": 1.0904,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6081751851887238,
+      "learning_rate": 6.757221660003947e-06,
+      "loss": 0.8092,
+      "step": 5037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4627321351088385,
+      "learning_rate": 6.756005229050624e-06,
+      "loss": 0.8273,
+      "step": 5038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.463084579768749,
+      "learning_rate": 6.7547886795278136e-06,
+      "loss": 0.7042,
+      "step": 5039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.485218244602575,
+      "learning_rate": 6.753572011517658e-06,
+      "loss": 0.8188,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.79903068672677,
+      "learning_rate": 6.752355225102309e-06,
+      "loss": 1.1297,
+      "step": 5041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5331218414089514,
+      "learning_rate": 6.75113832036393e-06,
+      "loss": 0.7544,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7945505539034713,
+      "learning_rate": 6.749921297384688e-06,
+      "loss": 1.0935,
+      "step": 5043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.535504296883687,
+      "learning_rate": 6.748704156246759e-06,
+      "loss": 0.8205,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4376724213713956,
+      "learning_rate": 6.747486897032325e-06,
+      "loss": 0.8611,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.374361158816089,
+      "learning_rate": 6.74626951982358e-06,
+      "loss": 0.745,
+      "step": 5046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4749142152124315,
+      "learning_rate": 6.745052024702724e-06,
+      "loss": 0.7402,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8261629194669423,
+      "learning_rate": 6.743834411751964e-06,
+      "loss": 1.1295,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.567634925873744,
+      "learning_rate": 6.742616681053518e-06,
+      "loss": 0.6677,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4493852941895933,
+      "learning_rate": 6.7413988326896106e-06,
+      "loss": 0.743,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.572896648088167,
+      "learning_rate": 6.740180866742472e-06,
+      "loss": 0.7349,
+      "step": 5051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4435258523313497,
+      "learning_rate": 6.738962783294339e-06,
+      "loss": 0.7944,
+      "step": 5052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.785160021783493,
+      "learning_rate": 6.737744582427464e-06,
+      "loss": 1.0889,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5948482406248132,
+      "learning_rate": 6.736526264224101e-06,
+      "loss": 0.7932,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6944631745282728,
+      "learning_rate": 6.735307828766515e-06,
+      "loss": 0.8387,
+      "step": 5055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4503711240281798,
+      "learning_rate": 6.734089276136977e-06,
+      "loss": 0.7067,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7968476304282831,
+      "learning_rate": 6.732870606417764e-06,
+      "loss": 1.0947,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4521058090156567,
+      "learning_rate": 6.7316518196911654e-06,
+      "loss": 0.7915,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5457251846000761,
+      "learning_rate": 6.730432916039476e-06,
+      "loss": 0.781,
+      "step": 5059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.48439726042228,
+      "learning_rate": 6.7292138955450005e-06,
+      "loss": 0.7692,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5960466298966511,
+      "learning_rate": 6.727994758290048e-06,
+      "loss": 0.8215,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5011927950091615,
+      "learning_rate": 6.726775504356939e-06,
+      "loss": 0.7675,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5902706011676242,
+      "learning_rate": 6.725556133827998e-06,
+      "loss": 0.648,
+      "step": 5063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4987346409577067,
+      "learning_rate": 6.724336646785561e-06,
+      "loss": 0.7903,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4854535693793278,
+      "learning_rate": 6.723117043311971e-06,
+      "loss": 0.7689,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5700939019860534,
+      "learning_rate": 6.7218973234895805e-06,
+      "loss": 0.7494,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6870096465506468,
+      "learning_rate": 6.7206774874007415e-06,
+      "loss": 0.6651,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4582543455797539,
+      "learning_rate": 6.719457535127827e-06,
+      "loss": 0.7589,
+      "step": 5068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4964858133449725,
+      "learning_rate": 6.718237466753206e-06,
+      "loss": 0.7601,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.440818931827845,
+      "learning_rate": 6.717017282359263e-06,
+      "loss": 0.7123,
+      "step": 5070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8923344600308764,
+      "learning_rate": 6.715796982028386e-06,
+      "loss": 1.1276,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8495259493276508,
+      "learning_rate": 6.714576565842976e-06,
+      "loss": 1.1155,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7779454828982353,
+      "learning_rate": 6.713356033885434e-06,
+      "loss": 1.073,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5374235379354897,
+      "learning_rate": 6.7121353862381746e-06,
+      "loss": 0.8581,
+      "step": 5074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4550850967900268,
+      "learning_rate": 6.710914622983619e-06,
+      "loss": 0.7091,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.977205876697971,
+      "learning_rate": 6.7096937442041956e-06,
+      "loss": 1.1335,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3945927607609554,
+      "learning_rate": 6.708472749982341e-06,
+      "loss": 0.7045,
+      "step": 5077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5620170742633996,
+      "learning_rate": 6.707251640400501e-06,
+      "loss": 0.8065,
+      "step": 5078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4289914203518963,
+      "learning_rate": 6.706030415541125e-06,
+      "loss": 0.7928,
+      "step": 5079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5835502971605004,
+      "learning_rate": 6.704809075486674e-06,
+      "loss": 0.7148,
+      "step": 5080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3973673551905534,
+      "learning_rate": 6.703587620319616e-06,
+      "loss": 0.7474,
+      "step": 5081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5370550876446254,
+      "learning_rate": 6.702366050122428e-06,
+      "loss": 0.8092,
+      "step": 5082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5290032699688738,
+      "learning_rate": 6.701144364977591e-06,
+      "loss": 0.7435,
+      "step": 5083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.511879756661044,
+      "learning_rate": 6.6999225649675955e-06,
+      "loss": 0.7817,
+      "step": 5084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4754122675053818,
+      "learning_rate": 6.698700650174943e-06,
+      "loss": 0.8127,
+      "step": 5085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.489005543125916,
+      "learning_rate": 6.697478620682137e-06,
+      "loss": 0.8221,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4169914567587139,
+      "learning_rate": 6.696256476571692e-06,
+      "loss": 0.6785,
+      "step": 5087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4480620542066576,
+      "learning_rate": 6.695034217926133e-06,
+      "loss": 0.7187,
+      "step": 5088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4307090609999533,
+      "learning_rate": 6.693811844827987e-06,
+      "loss": 0.7619,
+      "step": 5089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5500147102983772,
+      "learning_rate": 6.692589357359792e-06,
+      "loss": 0.7921,
+      "step": 5090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4627049078015082,
+      "learning_rate": 6.691366755604093e-06,
+      "loss": 0.7047,
+      "step": 5091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4636887343590725,
+      "learning_rate": 6.690144039643443e-06,
+      "loss": 0.8126,
+      "step": 5092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4962124938115384,
+      "learning_rate": 6.6889212095604036e-06,
+      "loss": 0.8002,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.567767800125546,
+      "learning_rate": 6.687698265437542e-06,
+      "loss": 0.7422,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6489763409068139,
+      "learning_rate": 6.686475207357435e-06,
+      "loss": 0.7398,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5145787523676202,
+      "learning_rate": 6.6852520354026625e-06,
+      "loss": 0.7133,
+      "step": 5096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.955426655707722,
+      "learning_rate": 6.684028749655822e-06,
+      "loss": 1.1258,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8799552242984676,
+      "learning_rate": 6.682805350199508e-06,
+      "loss": 1.1338,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5092073210863823,
+      "learning_rate": 6.681581837116331e-06,
+      "loss": 0.7238,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.772200211048803,
+      "learning_rate": 6.680358210488902e-06,
+      "loss": 1.1036,
+      "step": 5100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8424215179596961,
+      "learning_rate": 6.679134470399843e-06,
+      "loss": 1.077,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5138503210315437,
+      "learning_rate": 6.677910616931787e-06,
+      "loss": 0.7778,
+      "step": 5102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5536808988422206,
+      "learning_rate": 6.676686650167367e-06,
+      "loss": 0.7322,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8506095396258891,
+      "learning_rate": 6.6754625701892325e-06,
+      "loss": 1.1135,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5438816161775908,
+      "learning_rate": 6.674238377080034e-06,
+      "loss": 0.791,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4251306881380594,
+      "learning_rate": 6.67301407092243e-06,
+      "loss": 0.8141,
+      "step": 5106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5255984603090358,
+      "learning_rate": 6.671789651799092e-06,
+      "loss": 0.7204,
+      "step": 5107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.572330300068239,
+      "learning_rate": 6.670565119792694e-06,
+      "loss": 0.7839,
+      "step": 5108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5161347506869107,
+      "learning_rate": 6.669340474985918e-06,
+      "loss": 0.8893,
+      "step": 5109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6198426630045863,
+      "learning_rate": 6.6681157174614575e-06,
+      "loss": 0.7937,
+      "step": 5110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5229523205137783,
+      "learning_rate": 6.666890847302008e-06,
+      "loss": 0.7341,
+      "step": 5111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.509117549877118,
+      "learning_rate": 6.665665864590277e-06,
+      "loss": 0.7757,
+      "step": 5112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9794985704911618,
+      "learning_rate": 6.664440769408977e-06,
+      "loss": 0.775,
+      "step": 5113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6023062656879932,
+      "learning_rate": 6.6632155618408335e-06,
+      "loss": 0.8046,
+      "step": 5114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4492366153321568,
+      "learning_rate": 6.66199024196857e-06,
+      "loss": 0.8473,
+      "step": 5115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5590187531630764,
+      "learning_rate": 6.6607648098749244e-06,
+      "loss": 0.8577,
+      "step": 5116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6070371460118378,
+      "learning_rate": 6.659539265642643e-06,
+      "loss": 0.8596,
+      "step": 5117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.695580628428527,
+      "learning_rate": 6.658313609354474e-06,
+      "loss": 0.8882,
+      "step": 5118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.408800420256203,
+      "learning_rate": 6.657087841093179e-06,
+      "loss": 0.7806,
+      "step": 5119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4218064884357873,
+      "learning_rate": 6.655861960941524e-06,
+      "loss": 0.8707,
+      "step": 5120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5146069956536823,
+      "learning_rate": 6.654635968982284e-06,
+      "loss": 0.8131,
+      "step": 5121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4824627392713794,
+      "learning_rate": 6.653409865298238e-06,
+      "loss": 0.7709,
+      "step": 5122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.584130625118169,
+      "learning_rate": 6.652183649972177e-06,
+      "loss": 0.8291,
+      "step": 5123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.593476420095539,
+      "learning_rate": 6.6509573230868995e-06,
+      "loss": 0.7683,
+      "step": 5124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.558735792463535,
+      "learning_rate": 6.6497308847252074e-06,
+      "loss": 0.7646,
+      "step": 5125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.420576389298931,
+      "learning_rate": 6.648504334969914e-06,
+      "loss": 0.7846,
+      "step": 5126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8893763894071135,
+      "learning_rate": 6.647277673903838e-06,
+      "loss": 1.1154,
+      "step": 5127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4724727808839415,
+      "learning_rate": 6.646050901609806e-06,
+      "loss": 0.6825,
+      "step": 5128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4581446933854199,
+      "learning_rate": 6.644824018170655e-06,
+      "loss": 0.8903,
+      "step": 5129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4670498149691078,
+      "learning_rate": 6.643597023669224e-06,
+      "loss": 0.7702,
+      "step": 5130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5645346601026624,
+      "learning_rate": 6.642369918188365e-06,
+      "loss": 0.6753,
+      "step": 5131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7769710743553795,
+      "learning_rate": 6.641142701810932e-06,
+      "loss": 0.7957,
+      "step": 5132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5494821060273092,
+      "learning_rate": 6.639915374619793e-06,
+      "loss": 0.756,
+      "step": 5133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5338875636229348,
+      "learning_rate": 6.638687936697816e-06,
+      "loss": 0.7786,
+      "step": 5134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4983787649310027,
+      "learning_rate": 6.637460388127882e-06,
+      "loss": 0.7771,
+      "step": 5135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8539983648978486,
+      "learning_rate": 6.6362327289928795e-06,
+      "loss": 1.104,
+      "step": 5136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.840596336146245,
+      "learning_rate": 6.635004959375701e-06,
+      "loss": 1.0664,
+      "step": 5137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3252389216348521,
+      "learning_rate": 6.6337770793592515e-06,
+      "loss": 0.7437,
+      "step": 5138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5390346101157333,
+      "learning_rate": 6.632549089026435e-06,
+      "loss": 0.7795,
+      "step": 5139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6717524875268783,
+      "learning_rate": 6.631320988460172e-06,
+      "loss": 0.7855,
+      "step": 5140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4760544431432925,
+      "learning_rate": 6.6300927777433856e-06,
+      "loss": 0.8559,
+      "step": 5141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5653357090723692,
+      "learning_rate": 6.6288644569590065e-06,
+      "loss": 0.764,
+      "step": 5142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4308848622185135,
+      "learning_rate": 6.627636026189975e-06,
+      "loss": 0.7637,
+      "step": 5143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3685992404361451,
+      "learning_rate": 6.6264074855192385e-06,
+      "loss": 0.7791,
+      "step": 5144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4991518368616403,
+      "learning_rate": 6.625178835029749e-06,
+      "loss": 0.757,
+      "step": 5145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9641363787914339,
+      "learning_rate": 6.623950074804468e-06,
+      "loss": 1.0753,
+      "step": 5146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.456606393289079,
+      "learning_rate": 6.622721204926363e-06,
+      "loss": 0.8207,
+      "step": 5147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4734584338466337,
+      "learning_rate": 6.6214922254784145e-06,
+      "loss": 0.765,
+      "step": 5148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7983631399469822,
+      "learning_rate": 6.620263136543602e-06,
+      "loss": 1.0865,
+      "step": 5149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6012191451540079,
+      "learning_rate": 6.619033938204917e-06,
+      "loss": 0.7889,
+      "step": 5150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.602507798884412,
+      "learning_rate": 6.617804630545359e-06,
+      "loss": 0.7615,
+      "step": 5151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4327616304972914,
+      "learning_rate": 6.616575213647932e-06,
+      "loss": 0.7766,
+      "step": 5152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5601012898187485,
+      "learning_rate": 6.615345687595652e-06,
+      "loss": 0.7794,
+      "step": 5153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8458255060409743,
+      "learning_rate": 6.614116052471537e-06,
+      "loss": 1.0786,
+      "step": 5154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4700300341889976,
+      "learning_rate": 6.612886308358615e-06,
+      "loss": 0.7635,
+      "step": 5155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.551623399036214,
+      "learning_rate": 6.61165645533992e-06,
+      "loss": 0.7971,
+      "step": 5156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5364775948795695,
+      "learning_rate": 6.610426493498496e-06,
+      "loss": 0.8218,
+      "step": 5157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4052003778608038,
+      "learning_rate": 6.609196422917394e-06,
+      "loss": 0.7381,
+      "step": 5158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.556766391257282,
+      "learning_rate": 6.607966243679669e-06,
+      "loss": 0.7958,
+      "step": 5159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5796470453708265,
+      "learning_rate": 6.606735955868387e-06,
+      "loss": 0.7808,
+      "step": 5160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5797931148574027,
+      "learning_rate": 6.605505559566619e-06,
+      "loss": 0.8879,
+      "step": 5161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4268791314212117,
+      "learning_rate": 6.6042750548574455e-06,
+      "loss": 0.7412,
+      "step": 5162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4746829057806619,
+      "learning_rate": 6.6030444418239495e-06,
+      "loss": 0.7214,
+      "step": 5163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4375031428668945,
+      "learning_rate": 6.601813720549229e-06,
+      "loss": 0.7641,
+      "step": 5164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.511179755174562,
+      "learning_rate": 6.600582891116383e-06,
+      "loss": 0.7948,
+      "step": 5165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5348156923438556,
+      "learning_rate": 6.599351953608519e-06,
+      "loss": 0.8116,
+      "step": 5166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5039512406587,
+      "learning_rate": 6.598120908108756e-06,
+      "loss": 0.7429,
+      "step": 5167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5924894209712466,
+      "learning_rate": 6.596889754700213e-06,
+      "loss": 0.8116,
+      "step": 5168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5251115803419217,
+      "learning_rate": 6.595658493466024e-06,
+      "loss": 0.7482,
+      "step": 5169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5849991948936215,
+      "learning_rate": 6.594427124489325e-06,
+      "loss": 0.8505,
+      "step": 5170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.417780668352426,
+      "learning_rate": 6.5931956478532585e-06,
+      "loss": 0.7258,
+      "step": 5171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8703729671655517,
+      "learning_rate": 6.591964063640981e-06,
+      "loss": 1.0598,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3890323111777436,
+      "learning_rate": 6.590732371935649e-06,
+      "loss": 0.7876,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8174734888178268,
+      "learning_rate": 6.589500572820428e-06,
+      "loss": 1.0945,
+      "step": 5174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8132054278186829,
+      "learning_rate": 6.5882686663784955e-06,
+      "loss": 1.0857,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4553071516205818,
+      "learning_rate": 6.587036652693031e-06,
+      "loss": 0.832,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6029592563547859,
+      "learning_rate": 6.585804531847223e-06,
+      "loss": 0.804,
+      "step": 5177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5450025744395206,
+      "learning_rate": 6.584572303924266e-06,
+      "loss": 0.7707,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3923966540439565,
+      "learning_rate": 6.583339969007364e-06,
+      "loss": 0.7885,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5078869105885953,
+      "learning_rate": 6.582107527179726e-06,
+      "loss": 0.7597,
+      "step": 5180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.896617557544874,
+      "learning_rate": 6.58087497852457e-06,
+      "loss": 1.0477,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8249240460012384,
+      "learning_rate": 6.579642323125123e-06,
+      "loss": 0.878,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5048926217001046,
+      "learning_rate": 6.5784095610646115e-06,
+      "loss": 0.7826,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5510802406580388,
+      "learning_rate": 6.5771766924262795e-06,
+      "loss": 0.8751,
+      "step": 5184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6257504568888683,
+      "learning_rate": 6.575943717293368e-06,
+      "loss": 0.8428,
+      "step": 5185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5806136248560991,
+      "learning_rate": 6.574710635749134e-06,
+      "loss": 0.8112,
+      "step": 5186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4005448024960188,
+      "learning_rate": 6.573477447876838e-06,
+      "loss": 0.7214,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4501427343156157,
+      "learning_rate": 6.572244153759747e-06,
+      "loss": 0.7962,
+      "step": 5188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4383394430649246,
+      "learning_rate": 6.571010753481135e-06,
+      "loss": 0.7874,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.39161976603999,
+      "learning_rate": 6.569777247124285e-06,
+      "loss": 0.7453,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4878496536596488,
+      "learning_rate": 6.568543634772485e-06,
+      "loss": 0.8082,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4090363984372747,
+      "learning_rate": 6.567309916509033e-06,
+      "loss": 0.7249,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9092238473112751,
+      "learning_rate": 6.5660760924172304e-06,
+      "loss": 1.1009,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9281437165493032,
+      "learning_rate": 6.56484216258039e-06,
+      "loss": 0.7729,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4644267617008637,
+      "learning_rate": 6.563608127081827e-06,
+      "loss": 0.8013,
+      "step": 5195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5283021595307977,
+      "learning_rate": 6.562373986004871e-06,
+      "loss": 0.7289,
+      "step": 5196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7148511513980296,
+      "learning_rate": 6.5611397394328465e-06,
+      "loss": 0.771,
+      "step": 5197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5959857301858007,
+      "learning_rate": 6.5599053874491e-06,
+      "loss": 0.7692,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.437633961481371,
+      "learning_rate": 6.558670930136975e-06,
+      "loss": 0.6625,
+      "step": 5199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5109525411461984,
+      "learning_rate": 6.557436367579823e-06,
+      "loss": 0.8966,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.485055052782625,
+      "learning_rate": 6.556201699861008e-06,
+      "loss": 0.758,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.349609233105596,
+      "learning_rate": 6.554966927063895e-06,
+      "loss": 0.6523,
+      "step": 5202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5062272927607419,
+      "learning_rate": 6.55373204927186e-06,
+      "loss": 0.7185,
+      "step": 5203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3966728842588125,
+      "learning_rate": 6.552497066568282e-06,
+      "loss": 0.7576,
+      "step": 5204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.499573535320154,
+      "learning_rate": 6.551261979036554e-06,
+      "loss": 0.6838,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4620457172275592,
+      "learning_rate": 6.55002678676007e-06,
+      "loss": 0.7419,
+      "step": 5206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.48944356128856,
+      "learning_rate": 6.548791489822232e-06,
+      "loss": 0.8102,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.436678881564857,
+      "learning_rate": 6.547556088306453e-06,
+      "loss": 0.7478,
+      "step": 5208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4375720239595842,
+      "learning_rate": 6.546320582296145e-06,
+      "loss": 0.6943,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4586176293922488,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.7885,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5794478989789005,
+      "learning_rate": 6.543849257125661e-06,
+      "loss": 0.7313,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.528726653900992,
+      "learning_rate": 6.542613438132349e-06,
+      "loss": 0.8158,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.496523166353207,
+      "learning_rate": 6.541377514978253e-06,
+      "loss": 0.6831,
+      "step": 5213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9874270594359373,
+      "learning_rate": 6.54014148774682e-06,
+      "loss": 1.0966,
+      "step": 5214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8567491297679107,
+      "learning_rate": 6.538905356521515e-06,
+      "loss": 1.1355,
+      "step": 5215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.599359279026421,
+      "learning_rate": 6.537669121385801e-06,
+      "loss": 0.8165,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5347982026986946,
+      "learning_rate": 6.53643278242315e-06,
+      "loss": 0.7704,
+      "step": 5217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5925121857957893,
+      "learning_rate": 6.535196339717046e-06,
+      "loss": 0.8009,
+      "step": 5218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4858693246143697,
+      "learning_rate": 6.533959793350974e-06,
+      "loss": 0.8262,
+      "step": 5219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4574767012029153,
+      "learning_rate": 6.532723143408428e-06,
+      "loss": 0.7158,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1637011275501026,
+      "learning_rate": 6.531486389972913e-06,
+      "loss": 1.0776,
+      "step": 5221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.550112350985245,
+      "learning_rate": 6.530249533127932e-06,
+      "loss": 0.7893,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4967564947691119,
+      "learning_rate": 6.5290125729570066e-06,
+      "loss": 0.8214,
+      "step": 5223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4121715877769112,
+      "learning_rate": 6.527775509543653e-06,
+      "loss": 0.734,
+      "step": 5224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4261991779203285,
+      "learning_rate": 6.526538342971406e-06,
+      "loss": 0.7186,
+      "step": 5225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.459626269826093,
+      "learning_rate": 6.525301073323798e-06,
+      "loss": 0.7568,
+      "step": 5226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8547828232623059,
+      "learning_rate": 6.524063700684375e-06,
+      "loss": 1.0745,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4707010946419086,
+      "learning_rate": 6.522826225136685e-06,
+      "loss": 0.7694,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4557131832517571,
+      "learning_rate": 6.5215886467642855e-06,
+      "loss": 0.6333,
+      "step": 5229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7712295070343915,
+      "learning_rate": 6.520350965650742e-06,
+      "loss": 1.0863,
+      "step": 5230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4369618043305292,
+      "learning_rate": 6.519113181879624e-06,
+      "loss": 0.7632,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4049921404386776,
+      "learning_rate": 6.517875295534511e-06,
+      "loss": 0.8069,
+      "step": 5232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6985905692974774,
+      "learning_rate": 6.5166373066989885e-06,
+      "loss": 0.8274,
+      "step": 5233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6612872054521255,
+      "learning_rate": 6.5153992154566445e-06,
+      "loss": 0.794,
+      "step": 5234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.559978684126204,
+      "learning_rate": 6.514161021891082e-06,
+      "loss": 0.7599,
+      "step": 5235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4999739183267309,
+      "learning_rate": 6.512922726085904e-06,
+      "loss": 0.7276,
+      "step": 5236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4849937991500837,
+      "learning_rate": 6.511684328124725e-06,
+      "loss": 0.7094,
+      "step": 5237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5279540910222877,
+      "learning_rate": 6.510445828091164e-06,
+      "loss": 0.7954,
+      "step": 5238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9951782115026421,
+      "learning_rate": 6.509207226068845e-06,
+      "loss": 1.1128,
+      "step": 5239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5104256185754006,
+      "learning_rate": 6.507968522141405e-06,
+      "loss": 0.7559,
+      "step": 5240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4310896263894965,
+      "learning_rate": 6.50672971639248e-06,
+      "loss": 0.771,
+      "step": 5241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5793364432785386,
+      "learning_rate": 6.505490808905721e-06,
+      "loss": 0.7956,
+      "step": 5242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.806615855347152,
+      "learning_rate": 6.50425179976478e-06,
+      "loss": 1.0759,
+      "step": 5243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7742593467529842,
+      "learning_rate": 6.5030126890533165e-06,
+      "loss": 1.0867,
+      "step": 5244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8065194775254936,
+      "learning_rate": 6.501773476855e-06,
+      "loss": 1.0647,
+      "step": 5245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.540288359073978,
+      "learning_rate": 6.5005341632535045e-06,
+      "loss": 0.7725,
+      "step": 5246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7789658303300508,
+      "learning_rate": 6.499294748332512e-06,
+      "loss": 0.7726,
+      "step": 5247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.852017610816155,
+      "learning_rate": 6.498055232175708e-06,
+      "loss": 0.8574,
+      "step": 5248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.487844482988714,
+      "learning_rate": 6.496815614866792e-06,
+      "loss": 0.7859,
+      "step": 5249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8634870110044907,
+      "learning_rate": 6.49557589648946e-06,
+      "loss": 1.1476,
+      "step": 5250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4670722260763087,
+      "learning_rate": 6.4943360771274235e-06,
+      "loss": 0.802,
+      "step": 5251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8219069665444414,
+      "learning_rate": 6.4930961568644e-06,
+      "loss": 1.087,
+      "step": 5252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5159137378009837,
+      "learning_rate": 6.491856135784109e-06,
+      "loss": 0.8721,
+      "step": 5253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4188785100072878,
+      "learning_rate": 6.490616013970281e-06,
+      "loss": 0.7953,
+      "step": 5254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4380403964902955,
+      "learning_rate": 6.489375791506651e-06,
+      "loss": 0.7783,
+      "step": 5255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8258157029777003,
+      "learning_rate": 6.48813546847696e-06,
+      "loss": 1.1079,
+      "step": 5256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4168873158216369,
+      "learning_rate": 6.486895044964963e-06,
+      "loss": 0.7948,
+      "step": 5257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5052840745076674,
+      "learning_rate": 6.485654521054408e-06,
+      "loss": 0.7526,
+      "step": 5258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5066756195793956,
+      "learning_rate": 6.484413896829067e-06,
+      "loss": 0.7546,
+      "step": 5259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6006447154078212,
+      "learning_rate": 6.4831731723727035e-06,
+      "loss": 0.7806,
+      "step": 5260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8630537367780872,
+      "learning_rate": 6.481932347769097e-06,
+      "loss": 1.0906,
+      "step": 5261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.507401652754367,
+      "learning_rate": 6.480691423102028e-06,
+      "loss": 0.7909,
+      "step": 5262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4855505832547082,
+      "learning_rate": 6.479450398455287e-06,
+      "loss": 0.7958,
+      "step": 5263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4156618333042317,
+      "learning_rate": 6.478209273912675e-06,
+      "loss": 0.8299,
+      "step": 5264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5044052714540666,
+      "learning_rate": 6.476968049557993e-06,
+      "loss": 0.7844,
+      "step": 5265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.481148259038955,
+      "learning_rate": 6.475726725475049e-06,
+      "loss": 0.7787,
+      "step": 5266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5599739993062285,
+      "learning_rate": 6.474485301747663e-06,
+      "loss": 0.7952,
+      "step": 5267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6091669508372386,
+      "learning_rate": 6.473243778459657e-06,
+      "loss": 0.8427,
+      "step": 5268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5044858875186025,
+      "learning_rate": 6.472002155694863e-06,
+      "loss": 0.7513,
+      "step": 5269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4282311467118167,
+      "learning_rate": 6.470760433537116e-06,
+      "loss": 0.6772,
+      "step": 5270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5739984277706165,
+      "learning_rate": 6.469518612070265e-06,
+      "loss": 0.8621,
+      "step": 5271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5096571256541809,
+      "learning_rate": 6.468276691378155e-06,
+      "loss": 0.7788,
+      "step": 5272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.430132676171242,
+      "learning_rate": 6.467034671544644e-06,
+      "loss": 0.7634,
+      "step": 5273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5277401212199342,
+      "learning_rate": 6.4657925526535995e-06,
+      "loss": 0.8105,
+      "step": 5274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5373378947787077,
+      "learning_rate": 6.464550334788888e-06,
+      "loss": 0.8413,
+      "step": 5275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4523577237646559,
+      "learning_rate": 6.463308018034391e-06,
+      "loss": 0.8318,
+      "step": 5276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4948652324714429,
+      "learning_rate": 6.46206560247399e-06,
+      "loss": 0.8731,
+      "step": 5277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5845131775561614,
+      "learning_rate": 6.460823088191577e-06,
+      "loss": 0.8073,
+      "step": 5278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5395145556534044,
+      "learning_rate": 6.4595804752710475e-06,
+      "loss": 0.7508,
+      "step": 5279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4984431359385824,
+      "learning_rate": 6.458337763796306e-06,
+      "loss": 0.8454,
+      "step": 5280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8819296806898511,
+      "learning_rate": 6.457094953851266e-06,
+      "loss": 0.8309,
+      "step": 5281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5214378575283345,
+      "learning_rate": 6.455852045519843e-06,
+      "loss": 0.7265,
+      "step": 5282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8903688142471105,
+      "learning_rate": 6.454609038885959e-06,
+      "loss": 1.1128,
+      "step": 5283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8235192477404827,
+      "learning_rate": 6.453365934033548e-06,
+      "loss": 1.1065,
+      "step": 5284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4719376572840779,
+      "learning_rate": 6.452122731046544e-06,
+      "loss": 0.7281,
+      "step": 5285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4953144851234383,
+      "learning_rate": 6.450879430008895e-06,
+      "loss": 0.7349,
+      "step": 5286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7378750273183305,
+      "learning_rate": 6.449636031004548e-06,
+      "loss": 0.8772,
+      "step": 5287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5683923217629299,
+      "learning_rate": 6.4483925341174625e-06,
+      "loss": 0.854,
+      "step": 5288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.432686483084503,
+      "learning_rate": 6.4471489394316e-06,
+      "loss": 0.7757,
+      "step": 5289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.451056603739025,
+      "learning_rate": 6.4459052470309324e-06,
+      "loss": 0.7772,
+      "step": 5290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.90596064847353,
+      "learning_rate": 6.444661456999435e-06,
+      "loss": 1.0833,
+      "step": 5291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8609914777939631,
+      "learning_rate": 6.443417569421093e-06,
+      "loss": 1.1188,
+      "step": 5292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6317522249117764,
+      "learning_rate": 6.442173584379898e-06,
+      "loss": 0.8119,
+      "step": 5293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.492164679637194,
+      "learning_rate": 6.440929501959844e-06,
+      "loss": 0.7717,
+      "step": 5294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.850892898296131,
+      "learning_rate": 6.439685322244935e-06,
+      "loss": 1.1042,
+      "step": 5295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5534527895717405,
+      "learning_rate": 6.43844104531918e-06,
+      "loss": 0.8116,
+      "step": 5296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5015576121800887,
+      "learning_rate": 6.437196671266597e-06,
+      "loss": 0.8203,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4322675531428575,
+      "learning_rate": 6.435952200171209e-06,
+      "loss": 0.8335,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.49500522110894,
+      "learning_rate": 6.434707632117046e-06,
+      "loss": 0.6922,
+      "step": 5299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5008745865828224,
+      "learning_rate": 6.4334629671881425e-06,
+      "loss": 0.7739,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4689707691327347,
+      "learning_rate": 6.432218205468539e-06,
+      "loss": 0.7873,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5660354672698713,
+      "learning_rate": 6.430973347042289e-06,
+      "loss": 0.8623,
+      "step": 5302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4245259342070176,
+      "learning_rate": 6.429728391993446e-06,
+      "loss": 0.7679,
+      "step": 5303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6841374778575604,
+      "learning_rate": 6.428483340406074e-06,
+      "loss": 0.8218,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5850225382158867,
+      "learning_rate": 6.42723819236424e-06,
+      "loss": 0.7831,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4401851269816714,
+      "learning_rate": 6.42599294795202e-06,
+      "loss": 0.6947,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5635028122293435,
+      "learning_rate": 6.424747607253494e-06,
+      "loss": 0.7772,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0031186748345458,
+      "learning_rate": 6.423502170352752e-06,
+      "loss": 1.0475,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8849167020330146,
+      "learning_rate": 6.42225663733389e-06,
+      "loss": 1.0783,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.74501474209221,
+      "learning_rate": 6.4210110082810076e-06,
+      "loss": 0.7476,
+      "step": 5310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6241108858250781,
+      "learning_rate": 6.41976528327821e-06,
+      "loss": 0.8459,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4048208009988643,
+      "learning_rate": 6.418519462409616e-06,
+      "loss": 0.6686,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4492779302498808,
+      "learning_rate": 6.4172735457593435e-06,
+      "loss": 0.7703,
+      "step": 5313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5057880756249267,
+      "learning_rate": 6.41602753341152e-06,
+      "loss": 0.702,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.407104835889254,
+      "learning_rate": 6.414781425450282e-06,
+      "loss": 0.7783,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2280257485360622,
+      "learning_rate": 6.413535221959765e-06,
+      "loss": 1.096,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5530811519591503,
+      "learning_rate": 6.412288923024118e-06,
+      "loss": 0.8153,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6268409014393808,
+      "learning_rate": 6.411042528727492e-06,
+      "loss": 0.8463,
+      "step": 5318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0005156558194832,
+      "learning_rate": 6.4097960391540505e-06,
+      "loss": 1.0893,
+      "step": 5319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5189762256756192,
+      "learning_rate": 6.408549454387954e-06,
+      "loss": 0.7862,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.53639011463977,
+      "learning_rate": 6.40730277451338e-06,
+      "loss": 0.7521,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5430263105159505,
+      "learning_rate": 6.406055999614504e-06,
+      "loss": 0.7904,
+      "step": 5322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5434220777723873,
+      "learning_rate": 6.404809129775511e-06,
+      "loss": 0.7888,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.871134837480756,
+      "learning_rate": 6.403562165080594e-06,
+      "loss": 0.7186,
+      "step": 5324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4452661384284828,
+      "learning_rate": 6.4023151056139495e-06,
+      "loss": 0.7594,
+      "step": 5325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1087797049967127,
+      "learning_rate": 6.401067951459783e-06,
+      "loss": 1.0993,
+      "step": 5326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0562632902191567,
+      "learning_rate": 6.3998207027023056e-06,
+      "loss": 1.1241,
+      "step": 5327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6405202212246943,
+      "learning_rate": 6.398573359425732e-06,
+      "loss": 0.8116,
+      "step": 5328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6237800864190006,
+      "learning_rate": 6.397325921714288e-06,
+      "loss": 0.8108,
+      "step": 5329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.557928162185094,
+      "learning_rate": 6.396078389652201e-06,
+      "loss": 0.7979,
+      "step": 5330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.486344372118869,
+      "learning_rate": 6.394830763323711e-06,
+      "loss": 0.6812,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0423822030650176,
+      "learning_rate": 6.393583042813058e-06,
+      "loss": 1.0886,
+      "step": 5332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5066837348423356,
+      "learning_rate": 6.392335228204489e-06,
+      "loss": 0.7758,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.500828606972683,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.742,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9480205099603639,
+      "learning_rate": 6.389839317030642e-06,
+      "loss": 1.0778,
+      "step": 5335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8433307176323245,
+      "learning_rate": 6.388591220633891e-06,
+      "loss": 1.0834,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5011799882531291,
+      "learning_rate": 6.387343030476285e-06,
+      "loss": 0.807,
+      "step": 5337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4303518314255452,
+      "learning_rate": 6.386094746642105e-06,
+      "loss": 0.7564,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5741579910284678,
+      "learning_rate": 6.3848463692156396e-06,
+      "loss": 0.8296,
+      "step": 5339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4704493185583964,
+      "learning_rate": 6.383597898281179e-06,
+      "loss": 0.807,
+      "step": 5340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5783176542888324,
+      "learning_rate": 6.382349333923026e-06,
+      "loss": 0.8305,
+      "step": 5341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5145323393812957,
+      "learning_rate": 6.3811006762254845e-06,
+      "loss": 0.8426,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5384758390645246,
+      "learning_rate": 6.379851925272867e-06,
+      "loss": 0.7776,
+      "step": 5343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6005859975629086,
+      "learning_rate": 6.3786030811494935e-06,
+      "loss": 0.7349,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5708403110485372,
+      "learning_rate": 6.377354143939686e-06,
+      "loss": 0.8812,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4941344981749325,
+      "learning_rate": 6.376105113727778e-06,
+      "loss": 0.7335,
+      "step": 5346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4040914038721555,
+      "learning_rate": 6.374855990598106e-06,
+      "loss": 0.8081,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3906902034589184,
+      "learning_rate": 6.3736067746350135e-06,
+      "loss": 0.7892,
+      "step": 5348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5533349467866688,
+      "learning_rate": 6.372357465922851e-06,
+      "loss": 0.656,
+      "step": 5349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5143468171940913,
+      "learning_rate": 6.371108064545974e-06,
+      "loss": 0.7623,
+      "step": 5350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5109184872793924,
+      "learning_rate": 6.369858570588745e-06,
+      "loss": 0.8606,
+      "step": 5351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4834064476914546,
+      "learning_rate": 6.368608984135534e-06,
+      "loss": 0.7124,
+      "step": 5352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.360531245082571,
+      "learning_rate": 6.367359305270714e-06,
+      "loss": 1.0811,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.469737752250483,
+      "learning_rate": 6.366109534078667e-06,
+      "loss": 0.7693,
+      "step": 5354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.492891086326922,
+      "learning_rate": 6.36485967064378e-06,
+      "loss": 0.739,
+      "step": 5355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5032027491740074,
+      "learning_rate": 6.363609715050449e-06,
+      "loss": 0.7538,
+      "step": 5356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5648596322115182,
+      "learning_rate": 6.36235966738307e-06,
+      "loss": 0.7838,
+      "step": 5357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4848532760086242,
+      "learning_rate": 6.361109527726052e-06,
+      "loss": 0.7417,
+      "step": 5358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5640871417333742,
+      "learning_rate": 6.3598592961638065e-06,
+      "loss": 0.8714,
+      "step": 5359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4973083032988224,
+      "learning_rate": 6.358608972780752e-06,
+      "loss": 0.8038,
+      "step": 5360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8094706590089262,
+      "learning_rate": 6.3573585576613115e-06,
+      "loss": 1.1199,
+      "step": 5361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.490765376706642,
+      "learning_rate": 6.356108050889918e-06,
+      "loss": 0.7659,
+      "step": 5362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5471169950442245,
+      "learning_rate": 6.354857452551009e-06,
+      "loss": 0.7897,
+      "step": 5363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.428211377421227,
+      "learning_rate": 6.353606762729025e-06,
+      "loss": 0.7642,
+      "step": 5364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8250701724102718,
+      "learning_rate": 6.35235598150842e-06,
+      "loss": 1.1082,
+      "step": 5365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8869276971802629,
+      "learning_rate": 6.351105108973644e-06,
+      "loss": 1.1015,
+      "step": 5366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5247901050010868,
+      "learning_rate": 6.349854145209162e-06,
+      "loss": 0.8135,
+      "step": 5367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4789837187161965,
+      "learning_rate": 6.348603090299442e-06,
+      "loss": 0.8451,
+      "step": 5368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8155726006074527,
+      "learning_rate": 6.347351944328958e-06,
+      "loss": 1.1285,
+      "step": 5369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7652460094477846,
+      "learning_rate": 6.346100707382189e-06,
+      "loss": 0.7804,
+      "step": 5370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.422822455002983,
+      "learning_rate": 6.344849379543623e-06,
+      "loss": 0.7176,
+      "step": 5371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4312015242790952,
+      "learning_rate": 6.3435979608977515e-06,
+      "loss": 0.8002,
+      "step": 5372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8722686912644217,
+      "learning_rate": 6.342346451529073e-06,
+      "loss": 1.1037,
+      "step": 5373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4678717194011006,
+      "learning_rate": 6.341094851522093e-06,
+      "loss": 0.7994,
+      "step": 5374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5052718083893966,
+      "learning_rate": 6.339843160961321e-06,
+      "loss": 0.8363,
+      "step": 5375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8808169447180074,
+      "learning_rate": 6.338591379931277e-06,
+      "loss": 0.7709,
+      "step": 5376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6200282130665045,
+      "learning_rate": 6.337339508516481e-06,
+      "loss": 0.864,
+      "step": 5377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.418687446531194,
+      "learning_rate": 6.336087546801464e-06,
+      "loss": 0.7275,
+      "step": 5378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4966220671446786,
+      "learning_rate": 6.334835494870759e-06,
+      "loss": 0.7685,
+      "step": 5379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5859332949647846,
+      "learning_rate": 6.33358335280891e-06,
+      "loss": 0.7941,
+      "step": 5380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.700895736693925,
+      "learning_rate": 6.332331120700465e-06,
+      "loss": 0.7966,
+      "step": 5381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4063372006106443,
+      "learning_rate": 6.331078798629975e-06,
+      "loss": 0.8023,
+      "step": 5382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.969135685469333,
+      "learning_rate": 6.329826386682e-06,
+      "loss": 0.8289,
+      "step": 5383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.655925897083058,
+      "learning_rate": 6.328573884941107e-06,
+      "loss": 0.7308,
+      "step": 5384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5208108060122223,
+      "learning_rate": 6.327321293491868e-06,
+      "loss": 0.8282,
+      "step": 5385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6485806240675949,
+      "learning_rate": 6.326068612418859e-06,
+      "loss": 0.8586,
+      "step": 5386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5968333937978891,
+      "learning_rate": 6.324815841806668e-06,
+      "loss": 0.7899,
+      "step": 5387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5334832002052323,
+      "learning_rate": 6.323562981739878e-06,
+      "loss": 0.6899,
+      "step": 5388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8720925399208505,
+      "learning_rate": 6.322310032303092e-06,
+      "loss": 1.0891,
+      "step": 5389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6615435238378606,
+      "learning_rate": 6.3210569935809076e-06,
+      "loss": 0.8083,
+      "step": 5390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4334545926458857,
+      "learning_rate": 6.319803865657933e-06,
+      "loss": 0.8002,
+      "step": 5391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5659303893462666,
+      "learning_rate": 6.318550648618785e-06,
+      "loss": 0.7595,
+      "step": 5392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.386954472438171,
+      "learning_rate": 6.317297342548083e-06,
+      "loss": 0.7803,
+      "step": 5393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8645651950690447,
+      "learning_rate": 6.3160439475304515e-06,
+      "loss": 1.0884,
+      "step": 5394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5010219794307362,
+      "learning_rate": 6.314790463650522e-06,
+      "loss": 0.7613,
+      "step": 5395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5371773700821099,
+      "learning_rate": 6.313536890992935e-06,
+      "loss": 0.8612,
+      "step": 5396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5718175196200712,
+      "learning_rate": 6.312283229642333e-06,
+      "loss": 0.7411,
+      "step": 5397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7950852671335584,
+      "learning_rate": 6.3110294796833685e-06,
+      "loss": 0.782,
+      "step": 5398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5131312464469449,
+      "learning_rate": 6.309775641200695e-06,
+      "loss": 0.8505,
+      "step": 5399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5088457349755084,
+      "learning_rate": 6.308521714278973e-06,
+      "loss": 0.8246,
+      "step": 5400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4992540107373507,
+      "learning_rate": 6.307267699002874e-06,
+      "loss": 0.7985,
+      "step": 5401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.562086029763132,
+      "learning_rate": 6.306013595457072e-06,
+      "loss": 0.8156,
+      "step": 5402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6973329738520702,
+      "learning_rate": 6.304759403726245e-06,
+      "loss": 0.7775,
+      "step": 5403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8584774214732819,
+      "learning_rate": 6.303505123895079e-06,
+      "loss": 1.1366,
+      "step": 5404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5135967721215173,
+      "learning_rate": 6.302250756048267e-06,
+      "loss": 0.7829,
+      "step": 5405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5016862737743029,
+      "learning_rate": 6.300996300270505e-06,
+      "loss": 0.7169,
+      "step": 5406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4742189834747996,
+      "learning_rate": 6.299741756646499e-06,
+      "loss": 0.7993,
+      "step": 5407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5389975462690597,
+      "learning_rate": 6.298487125260957e-06,
+      "loss": 0.7871,
+      "step": 5408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4227004133206302,
+      "learning_rate": 6.297232406198597e-06,
+      "loss": 0.7838,
+      "step": 5409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7893124859044485,
+      "learning_rate": 6.295977599544137e-06,
+      "loss": 1.0856,
+      "step": 5410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4701241116350714,
+      "learning_rate": 6.2947227053823055e-06,
+      "loss": 0.8232,
+      "step": 5411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5737045005292047,
+      "learning_rate": 6.293467723797837e-06,
+      "loss": 0.8197,
+      "step": 5412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.542679663774191,
+      "learning_rate": 6.2922126548754696e-06,
+      "loss": 0.7851,
+      "step": 5413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7989419771611457,
+      "learning_rate": 6.290957498699949e-06,
+      "loss": 1.1112,
+      "step": 5414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4752573130822828,
+      "learning_rate": 6.289702255356027e-06,
+      "loss": 0.8467,
+      "step": 5415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5150567158880677,
+      "learning_rate": 6.288446924928459e-06,
+      "loss": 0.7914,
+      "step": 5416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7912596392762239,
+      "learning_rate": 6.287191507502009e-06,
+      "loss": 1.0817,
+      "step": 5417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6293098617223778,
+      "learning_rate": 6.285936003161445e-06,
+      "loss": 0.8351,
+      "step": 5418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7887330318299309,
+      "learning_rate": 6.2846804119915405e-06,
+      "loss": 1.0767,
+      "step": 5419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5155755324763134,
+      "learning_rate": 6.283424734077078e-06,
+      "loss": 0.8251,
+      "step": 5420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7867052919372468,
+      "learning_rate": 6.282168969502843e-06,
+      "loss": 1.1088,
+      "step": 5421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5179695081207245,
+      "learning_rate": 6.280913118353627e-06,
+      "loss": 0.7166,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4487264724458715,
+      "learning_rate": 6.279657180714227e-06,
+      "loss": 0.7547,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5195469849755592,
+      "learning_rate": 6.278401156669449e-06,
+      "loss": 0.7795,
+      "step": 5424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.272245231206262,
+      "learning_rate": 6.2771450463041015e-06,
+      "loss": 0.8092,
+      "step": 5425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4005666108108732,
+      "learning_rate": 6.275888849703001e-06,
+      "loss": 0.7948,
+      "step": 5426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4864242039133861,
+      "learning_rate": 6.274632566950967e-06,
+      "loss": 0.8035,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4236139898651898,
+      "learning_rate": 6.273376198132828e-06,
+      "loss": 0.7952,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.717549775192585,
+      "learning_rate": 6.272119743333416e-06,
+      "loss": 0.8176,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.565628430878305,
+      "learning_rate": 6.270863202637569e-06,
+      "loss": 0.7198,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5062731928807451,
+      "learning_rate": 6.269606576130135e-06,
+      "loss": 0.8672,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5522639693257139,
+      "learning_rate": 6.268349863895959e-06,
+      "loss": 0.7177,
+      "step": 5432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5697288331426367,
+      "learning_rate": 6.2670930660199025e-06,
+      "loss": 0.7857,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4780266572910774,
+      "learning_rate": 6.265836182586823e-06,
+      "loss": 0.7931,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5441673387737247,
+      "learning_rate": 6.264579213681592e-06,
+      "loss": 0.7781,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4730138771942984,
+      "learning_rate": 6.263322159389078e-06,
+      "loss": 0.7246,
+      "step": 5436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.658339698149235,
+      "learning_rate": 6.262065019794165e-06,
+      "loss": 0.8579,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4682722853919492,
+      "learning_rate": 6.260807794981736e-06,
+      "loss": 0.8046,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9080079513423435,
+      "learning_rate": 6.259550485036681e-06,
+      "loss": 1.0977,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6034843319809537,
+      "learning_rate": 6.2582930900438975e-06,
+      "loss": 0.7421,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8207860063121404,
+      "learning_rate": 6.257035610088287e-06,
+      "loss": 1.0766,
+      "step": 5441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5562070891453748,
+      "learning_rate": 6.255778045254758e-06,
+      "loss": 0.7825,
+      "step": 5442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8198184254046909,
+      "learning_rate": 6.254520395628224e-06,
+      "loss": 1.0962,
+      "step": 5443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8246407650671811,
+      "learning_rate": 6.2532626612936035e-06,
+      "loss": 1.0878,
+      "step": 5444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5159315743523272,
+      "learning_rate": 6.252004842335824e-06,
+      "loss": 0.8269,
+      "step": 5445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8121480318247445,
+      "learning_rate": 6.250746938839811e-06,
+      "loss": 1.0714,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5219204678732232,
+      "learning_rate": 6.249488950890509e-06,
+      "loss": 0.7544,
+      "step": 5447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.605321370470892,
+      "learning_rate": 6.248230878572854e-06,
+      "loss": 0.7475,
+      "step": 5448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4780324446307072,
+      "learning_rate": 6.246972721971796e-06,
+      "loss": 0.8557,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5378015129041276,
+      "learning_rate": 6.245714481172288e-06,
+      "loss": 0.6901,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4226277785703438,
+      "learning_rate": 6.24445615625929e-06,
+      "loss": 0.6884,
+      "step": 5451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4905823589889238,
+      "learning_rate": 6.243197747317766e-06,
+      "loss": 0.7551,
+      "step": 5452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3833002841437545,
+      "learning_rate": 6.24193925443269e-06,
+      "loss": 0.8042,
+      "step": 5453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5221016434073646,
+      "learning_rate": 6.240680677689033e-06,
+      "loss": 0.7531,
+      "step": 5454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9568446557218008,
+      "learning_rate": 6.2394220171717805e-06,
+      "loss": 1.07,
+      "step": 5455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5169510171456861,
+      "learning_rate": 6.238163272965918e-06,
+      "loss": 0.804,
+      "step": 5456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6034476659208345,
+      "learning_rate": 6.236904445156442e-06,
+      "loss": 0.867,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.540845701915718,
+      "learning_rate": 6.235645533828348e-06,
+      "loss": 0.8733,
+      "step": 5458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5236112078899997,
+      "learning_rate": 6.234386539066643e-06,
+      "loss": 0.7613,
+      "step": 5459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7793286555034444,
+      "learning_rate": 6.233127460956337e-06,
+      "loss": 0.7042,
+      "step": 5460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5919274978252669,
+      "learning_rate": 6.231868299582444e-06,
+      "loss": 0.6878,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8631552062489987,
+      "learning_rate": 6.230609055029987e-06,
+      "loss": 1.1271,
+      "step": 5462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5268155917378479,
+      "learning_rate": 6.229349727383992e-06,
+      "loss": 0.7975,
+      "step": 5463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.618781817005855,
+      "learning_rate": 6.228090316729493e-06,
+      "loss": 0.7134,
+      "step": 5464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4722772107996318,
+      "learning_rate": 6.226830823151528e-06,
+      "loss": 0.7715,
+      "step": 5465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4761464362424526,
+      "learning_rate": 6.22557124673514e-06,
+      "loss": 0.8188,
+      "step": 5466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6714219066089482,
+      "learning_rate": 6.224311587565379e-06,
+      "loss": 0.7785,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5071692497754334,
+      "learning_rate": 6.223051845727299e-06,
+      "loss": 0.8151,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5764130057880015,
+      "learning_rate": 6.221792021305964e-06,
+      "loss": 0.7856,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8807057436212978,
+      "learning_rate": 6.220532114386437e-06,
+      "loss": 1.0866,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5240455551862604,
+      "learning_rate": 6.21927212505379e-06,
+      "loss": 0.7511,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8297996185513958,
+      "learning_rate": 6.218012053393101e-06,
+      "loss": 1.0808,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4171828096522558,
+      "learning_rate": 6.216751899489454e-06,
+      "loss": 0.726,
+      "step": 5473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5126088564320275,
+      "learning_rate": 6.215491663427936e-06,
+      "loss": 0.7379,
+      "step": 5474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4770132312672042,
+      "learning_rate": 6.214231345293641e-06,
+      "loss": 0.8145,
+      "step": 5475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4604208677512673,
+      "learning_rate": 6.21297094517167e-06,
+      "loss": 0.8136,
+      "step": 5476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5079490522727337,
+      "learning_rate": 6.211710463147127e-06,
+      "loss": 0.7736,
+      "step": 5477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5011891920907696,
+      "learning_rate": 6.210449899305121e-06,
+      "loss": 0.7279,
+      "step": 5478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6481383781541319,
+      "learning_rate": 6.209189253730772e-06,
+      "loss": 0.7833,
+      "step": 5479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6731111948885569,
+      "learning_rate": 6.207928526509198e-06,
+      "loss": 0.7765,
+      "step": 5480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7038795674985152,
+      "learning_rate": 6.206667717725529e-06,
+      "loss": 0.7674,
+      "step": 5481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4988646318436027,
+      "learning_rate": 6.205406827464897e-06,
+      "loss": 0.8714,
+      "step": 5482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5094695545345682,
+      "learning_rate": 6.204145855812439e-06,
+      "loss": 0.7562,
+      "step": 5483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.382218882544167,
+      "learning_rate": 6.202884802853299e-06,
+      "loss": 0.7441,
+      "step": 5484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4728695916855619,
+      "learning_rate": 6.201623668672627e-06,
+      "loss": 0.7311,
+      "step": 5485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3737410527556522,
+      "learning_rate": 6.200362453355578e-06,
+      "loss": 0.7437,
+      "step": 5486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4643724326947702,
+      "learning_rate": 6.19910115698731e-06,
+      "loss": 0.7836,
+      "step": 5487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.477599177649633,
+      "learning_rate": 6.197839779652991e-06,
+      "loss": 0.807,
+      "step": 5488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.775395684686117,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.7791,
+      "step": 5489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.492123924927538,
+      "learning_rate": 6.195316782426884e-06,
+      "loss": 0.7618,
+      "step": 5490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.567247686616791,
+      "learning_rate": 6.194055162705457e-06,
+      "loss": 0.7902,
+      "step": 5491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5444411107796612,
+      "learning_rate": 6.192793462358695e-06,
+      "loss": 0.7651,
+      "step": 5492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4674369671918859,
+      "learning_rate": 6.191531681471792e-06,
+      "loss": 0.7291,
+      "step": 5493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5718302599513225,
+      "learning_rate": 6.1902698201299425e-06,
+      "loss": 0.7819,
+      "step": 5494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9416648752255887,
+      "learning_rate": 6.189007878418354e-06,
+      "loss": 1.1107,
+      "step": 5495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6305591640464028,
+      "learning_rate": 6.187745856422236e-06,
+      "loss": 0.7657,
+      "step": 5496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4954947775141298,
+      "learning_rate": 6.1864837542268e-06,
+      "loss": 0.8048,
+      "step": 5497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.530189247898666,
+      "learning_rate": 6.185221571917271e-06,
+      "loss": 0.8644,
+      "step": 5498
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5658013030075886,
+      "learning_rate": 6.1839593095788675e-06,
+      "loss": 0.7989,
+      "step": 5499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5576844021068639,
+      "learning_rate": 6.182696967296825e-06,
+      "loss": 0.7853,
+      "step": 5500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5178709036485662,
+      "learning_rate": 6.181434545156379e-06,
+      "loss": 0.7978,
+      "step": 5501
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6681298000627816,
+      "learning_rate": 6.180172043242772e-06,
+      "loss": 0.8353,
+      "step": 5502
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.621568790014267,
+      "learning_rate": 6.17890946164125e-06,
+      "loss": 0.7352,
+      "step": 5503
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6062805616744213,
+      "learning_rate": 6.177646800437066e-06,
+      "loss": 0.7327,
+      "step": 5504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7291237998637405,
+      "learning_rate": 6.176384059715477e-06,
+      "loss": 0.7475,
+      "step": 5505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.562937862613243,
+      "learning_rate": 6.175121239561745e-06,
+      "loss": 0.7731,
+      "step": 5506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8820085519300088,
+      "learning_rate": 6.173858340061138e-06,
+      "loss": 1.1217,
+      "step": 5507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5033407823816525,
+      "learning_rate": 6.172595361298935e-06,
+      "loss": 0.7506,
+      "step": 5508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.800638521062958,
+      "learning_rate": 6.171332303360411e-06,
+      "loss": 1.1081,
+      "step": 5509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5014196397592523,
+      "learning_rate": 6.170069166330852e-06,
+      "loss": 0.7662,
+      "step": 5510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4862517772029635,
+      "learning_rate": 6.168805950295547e-06,
+      "loss": 0.6755,
+      "step": 5511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4870704372687238,
+      "learning_rate": 6.167542655339791e-06,
+      "loss": 0.7667,
+      "step": 5512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6418751680938126,
+      "learning_rate": 6.166279281548886e-06,
+      "loss": 0.8638,
+      "step": 5513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6038263458979407,
+      "learning_rate": 6.165015829008137e-06,
+      "loss": 0.8035,
+      "step": 5514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4569311182060263,
+      "learning_rate": 6.1637522978028545e-06,
+      "loss": 0.7309,
+      "step": 5515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.422291285646642,
+      "learning_rate": 6.1624886880183555e-06,
+      "loss": 0.769,
+      "step": 5516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9660891929188207,
+      "learning_rate": 6.161224999739963e-06,
+      "loss": 1.0942,
+      "step": 5517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5949811276579677,
+      "learning_rate": 6.159961233053002e-06,
+      "loss": 0.73,
+      "step": 5518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6248135231866787,
+      "learning_rate": 6.158697388042807e-06,
+      "loss": 0.7777,
+      "step": 5519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8076602535159536,
+      "learning_rate": 6.157433464794717e-06,
+      "loss": 1.1087,
+      "step": 5520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4792580752168412,
+      "learning_rate": 6.15616946339407e-06,
+      "loss": 0.7546,
+      "step": 5521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4323087527715166,
+      "learning_rate": 6.154905383926218e-06,
+      "loss": 0.751,
+      "step": 5522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8492462697722462,
+      "learning_rate": 6.153641226476512e-06,
+      "loss": 1.1035,
+      "step": 5523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5082179536207279,
+      "learning_rate": 6.152376991130313e-06,
+      "loss": 0.7956,
+      "step": 5524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6128802965916764,
+      "learning_rate": 6.151112677972987e-06,
+      "loss": 0.7954,
+      "step": 5525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4211362726795298,
+      "learning_rate": 6.149848287089899e-06,
+      "loss": 0.7191,
+      "step": 5526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.569961725303403,
+      "learning_rate": 6.148583818566426e-06,
+      "loss": 0.7789,
+      "step": 5527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5241827231041294,
+      "learning_rate": 6.147319272487946e-06,
+      "loss": 0.7761,
+      "step": 5528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4395342122586952,
+      "learning_rate": 6.146054648939845e-06,
+      "loss": 0.7692,
+      "step": 5529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5127711941161888,
+      "learning_rate": 6.144789948007516e-06,
+      "loss": 0.7397,
+      "step": 5530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5006750947809413,
+      "learning_rate": 6.14352516977635e-06,
+      "loss": 0.7947,
+      "step": 5531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6278917518202463,
+      "learning_rate": 6.142260314331751e-06,
+      "loss": 0.6689,
+      "step": 5532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4843556926873132,
+      "learning_rate": 6.140995381759121e-06,
+      "loss": 0.8571,
+      "step": 5533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5062605893425838,
+      "learning_rate": 6.139730372143877e-06,
+      "loss": 0.7303,
+      "step": 5534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6026004885527438,
+      "learning_rate": 6.1384652855714295e-06,
+      "loss": 0.8356,
+      "step": 5535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8520758762043584,
+      "learning_rate": 6.1372001221272045e-06,
+      "loss": 1.0947,
+      "step": 5536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.430515282655726,
+      "learning_rate": 6.1359348818966265e-06,
+      "loss": 0.7422,
+      "step": 5537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5081100186969714,
+      "learning_rate": 6.134669564965128e-06,
+      "loss": 0.7766,
+      "step": 5538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5508750650234953,
+      "learning_rate": 6.133404171418145e-06,
+      "loss": 0.7957,
+      "step": 5539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8601603479761468,
+      "learning_rate": 6.13213870134112e-06,
+      "loss": 0.7683,
+      "step": 5540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5426365835035472,
+      "learning_rate": 6.1308731548195025e-06,
+      "loss": 0.8316,
+      "step": 5541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5965920622988348,
+      "learning_rate": 6.129607531938744e-06,
+      "loss": 0.8433,
+      "step": 5542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.483010589431316,
+      "learning_rate": 6.128341832784301e-06,
+      "loss": 0.7276,
+      "step": 5543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.472950805673831,
+      "learning_rate": 6.127076057441638e-06,
+      "loss": 0.8164,
+      "step": 5544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6298862761138326,
+      "learning_rate": 6.125810205996221e-06,
+      "loss": 0.7912,
+      "step": 5545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5003767966940265,
+      "learning_rate": 6.124544278533526e-06,
+      "loss": 0.8195,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5028119841929566,
+      "learning_rate": 6.12327827513903e-06,
+      "loss": 0.7714,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4321050544791523,
+      "learning_rate": 6.122012195898216e-06,
+      "loss": 0.7605,
+      "step": 5548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8511037934342788,
+      "learning_rate": 6.120746040896572e-06,
+      "loss": 1.1185,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.536054342996016,
+      "learning_rate": 6.119479810219593e-06,
+      "loss": 0.8502,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5510074210539517,
+      "learning_rate": 6.118213503952779e-06,
+      "loss": 0.7806,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5212532338471112,
+      "learning_rate": 6.11694712218163e-06,
+      "loss": 0.7551,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.573339575859948,
+      "learning_rate": 6.115680664991658e-06,
+      "loss": 0.8611,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8207003076126247,
+      "learning_rate": 6.114414132468377e-06,
+      "loss": 1.1008,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5101746958483362,
+      "learning_rate": 6.113147524697305e-06,
+      "loss": 0.7543,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4474904978496517,
+      "learning_rate": 6.111880841763966e-06,
+      "loss": 0.7546,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8103151684072666,
+      "learning_rate": 6.110614083753891e-06,
+      "loss": 1.1013,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6485552567099258,
+      "learning_rate": 6.109347250752614e-06,
+      "loss": 0.8573,
+      "step": 5558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4065521205048261,
+      "learning_rate": 6.1080803428456735e-06,
+      "loss": 0.7182,
+      "step": 5559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5373697658467078,
+      "learning_rate": 6.106813360118614e-06,
+      "loss": 0.7306,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3751684246128415,
+      "learning_rate": 6.105546302656986e-06,
+      "loss": 0.7514,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5026263110434799,
+      "learning_rate": 6.104279170546344e-06,
+      "loss": 0.7797,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8602072999023384,
+      "learning_rate": 6.103011963872246e-06,
+      "loss": 1.1022,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5592814840559779,
+      "learning_rate": 6.10174468272026e-06,
+      "loss": 0.7896,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6421771581641462,
+      "learning_rate": 6.100477327175951e-06,
+      "loss": 0.7751,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3676666686111933,
+      "learning_rate": 6.0992098973249e-06,
+      "loss": 0.6513,
+      "step": 5566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4876123858431767,
+      "learning_rate": 6.09794239325268e-06,
+      "loss": 0.687,
+      "step": 5567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3912679043820255,
+      "learning_rate": 6.09667481504488e-06,
+      "loss": 0.6871,
+      "step": 5568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6443611072099442,
+      "learning_rate": 6.095407162787088e-06,
+      "loss": 0.7504,
+      "step": 5569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5331678881960737,
+      "learning_rate": 6.094139436564902e-06,
+      "loss": 0.7454,
+      "step": 5570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5778201683310515,
+      "learning_rate": 6.092871636463919e-06,
+      "loss": 0.8287,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4162117527184854,
+      "learning_rate": 6.0916037625697425e-06,
+      "loss": 0.7904,
+      "step": 5572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5103676347298858,
+      "learning_rate": 6.090335814967984e-06,
+      "loss": 0.8383,
+      "step": 5573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4415038126314506,
+      "learning_rate": 6.089067793744258e-06,
+      "loss": 0.7411,
+      "step": 5574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5519048847931312,
+      "learning_rate": 6.0877996989841845e-06,
+      "loss": 0.7356,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6495779749869393,
+      "learning_rate": 6.086531530773389e-06,
+      "loss": 0.7407,
+      "step": 5576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5202203242824701,
+      "learning_rate": 6.0852632891974986e-06,
+      "loss": 0.8504,
+      "step": 5577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4787871729956095,
+      "learning_rate": 6.083994974342151e-06,
+      "loss": 0.7985,
+      "step": 5578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5504304657665704,
+      "learning_rate": 6.082726586292982e-06,
+      "loss": 0.7828,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8978247823049745,
+      "learning_rate": 6.081458125135639e-06,
+      "loss": 1.1133,
+      "step": 5580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4942877754364783,
+      "learning_rate": 6.080189590955772e-06,
+      "loss": 0.7998,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.536676737204576,
+      "learning_rate": 6.078920983839032e-06,
+      "loss": 0.7578,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4855675171969795,
+      "learning_rate": 6.07765230387108e-06,
+      "loss": 0.7457,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6567148601544681,
+      "learning_rate": 6.07638355113758e-06,
+      "loss": 0.7723,
+      "step": 5584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5382448980473618,
+      "learning_rate": 6.075114725724203e-06,
+      "loss": 0.8082,
+      "step": 5585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.490408788516059,
+      "learning_rate": 6.07384582771662e-06,
+      "loss": 0.851,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5117484718138865,
+      "learning_rate": 6.072576857200512e-06,
+      "loss": 0.8081,
+      "step": 5587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.472805435279507,
+      "learning_rate": 6.071307814261561e-06,
+      "loss": 0.7661,
+      "step": 5588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5540061508366736,
+      "learning_rate": 6.070038698985457e-06,
+      "loss": 0.8585,
+      "step": 5589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3545397713683067,
+      "learning_rate": 6.068769511457894e-06,
+      "loss": 0.6834,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5024707412832095,
+      "learning_rate": 6.0675002517645685e-06,
+      "loss": 0.7546,
+      "step": 5591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.818118551471803,
+      "learning_rate": 6.0662309199911874e-06,
+      "loss": 1.1065,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.868781219719909,
+      "learning_rate": 6.064961516223453e-06,
+      "loss": 1.1333,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6369505096857158,
+      "learning_rate": 6.063692040547083e-06,
+      "loss": 0.7694,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.403317968425861,
+      "learning_rate": 6.062422493047796e-06,
+      "loss": 0.7115,
+      "step": 5595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.475295784011105,
+      "learning_rate": 6.061152873811311e-06,
+      "loss": 0.73,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4427827065806427,
+      "learning_rate": 6.059883182923359e-06,
+      "loss": 0.7404,
+      "step": 5597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.69121367146256,
+      "learning_rate": 6.05861342046967e-06,
+      "loss": 0.7316,
+      "step": 5598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5629393375192662,
+      "learning_rate": 6.057343586535982e-06,
+      "loss": 0.8191,
+      "step": 5599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.426722983580572,
+      "learning_rate": 6.056073681208038e-06,
+      "loss": 0.7803,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.470334387422037,
+      "learning_rate": 6.054803704571582e-06,
+      "loss": 0.8007,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4462786215810641,
+      "learning_rate": 6.0535336567123715e-06,
+      "loss": 0.8064,
+      "step": 5602
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9555436772052028,
+      "learning_rate": 6.052263537716158e-06,
+      "loss": 1.091,
+      "step": 5603
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5104333040917042,
+      "learning_rate": 6.0509933476687036e-06,
+      "loss": 0.7518,
+      "step": 5604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5978506557201515,
+      "learning_rate": 6.0497230866557746e-06,
+      "loss": 0.8114,
+      "step": 5605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.469934088289661,
+      "learning_rate": 6.048452754763143e-06,
+      "loss": 0.7422,
+      "step": 5606
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.459429660226803,
+      "learning_rate": 6.047182352076585e-06,
+      "loss": 0.773,
+      "step": 5607
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5518583239230312,
+      "learning_rate": 6.045911878681879e-06,
+      "loss": 0.8757,
+      "step": 5608
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5486071124802223,
+      "learning_rate": 6.044641334664812e-06,
+      "loss": 0.8106,
+      "step": 5609
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5918920727015184,
+      "learning_rate": 6.043370720111172e-06,
+      "loss": 0.8088,
+      "step": 5610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8318583848826501,
+      "learning_rate": 6.042100035106756e-06,
+      "loss": 1.0843,
+      "step": 5611
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4896308953045287,
+      "learning_rate": 6.040829279737363e-06,
+      "loss": 0.6898,
+      "step": 5612
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.435376309898992,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.7784,
+      "step": 5613
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.433405440349099,
+      "learning_rate": 6.038287558246868e-06,
+      "loss": 0.6801,
+      "step": 5614
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5667811839552435,
+      "learning_rate": 6.037016592297388e-06,
+      "loss": 0.7946,
+      "step": 5615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3032381274246556,
+      "learning_rate": 6.035745556326176e-06,
+      "loss": 0.8212,
+      "step": 5616
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5779041016914368,
+      "learning_rate": 6.034474450419056e-06,
+      "loss": 0.7203,
+      "step": 5617
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.813191361041576,
+      "learning_rate": 6.033203274661854e-06,
+      "loss": 1.079,
+      "step": 5618
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8215572772447268,
+      "learning_rate": 6.031932029140407e-06,
+      "loss": 1.0935,
+      "step": 5619
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5535889067523356,
+      "learning_rate": 6.030660713940549e-06,
+      "loss": 0.8094,
+      "step": 5620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7340064976542419,
+      "learning_rate": 6.029389329148123e-06,
+      "loss": 1.0866,
+      "step": 5621
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.481239515753897,
+      "learning_rate": 6.0281178748489745e-06,
+      "loss": 0.7314,
+      "step": 5622
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8063177529147812,
+      "learning_rate": 6.026846351128955e-06,
+      "loss": 1.108,
+      "step": 5623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4603019031357471,
+      "learning_rate": 6.025574758073925e-06,
+      "loss": 0.6998,
+      "step": 5624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5665352387640574,
+      "learning_rate": 6.024303095769741e-06,
+      "loss": 0.7475,
+      "step": 5625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7919445066477342,
+      "learning_rate": 6.02303136430227e-06,
+      "loss": 1.116,
+      "step": 5626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4077201347659456,
+      "learning_rate": 6.021759563757381e-06,
+      "loss": 0.7817,
+      "step": 5627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5346187768313124,
+      "learning_rate": 6.02048769422095e-06,
+      "loss": 0.7108,
+      "step": 5628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4366248203901046,
+      "learning_rate": 6.019215755778857e-06,
+      "loss": 0.7652,
+      "step": 5629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.576415782847079,
+      "learning_rate": 6.017943748516987e-06,
+      "loss": 0.8228,
+      "step": 5630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5775103902934209,
+      "learning_rate": 6.016671672521226e-06,
+      "loss": 0.7539,
+      "step": 5631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3751257650760778,
+      "learning_rate": 6.015399527877468e-06,
+      "loss": 0.7824,
+      "step": 5632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.505940762259155,
+      "learning_rate": 6.014127314671613e-06,
+      "loss": 0.8309,
+      "step": 5633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8349901020570976,
+      "learning_rate": 6.0128550329895615e-06,
+      "loss": 1.1306,
+      "step": 5634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.695651013117236,
+      "learning_rate": 6.011582682917223e-06,
+      "loss": 0.8057,
+      "step": 5635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5041700747473987,
+      "learning_rate": 6.010310264540511e-06,
+      "loss": 0.7831,
+      "step": 5636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.535048492717353,
+      "learning_rate": 6.009037777945337e-06,
+      "loss": 0.8563,
+      "step": 5637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.523477171557345,
+      "learning_rate": 6.007765223217626e-06,
+      "loss": 0.8115,
+      "step": 5638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5398949547576108,
+      "learning_rate": 6.006492600443301e-06,
+      "loss": 0.7932,
+      "step": 5639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7815058738985868,
+      "learning_rate": 6.0052199097082955e-06,
+      "loss": 1.0339,
+      "step": 5640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5159170873872643,
+      "learning_rate": 6.003947151098543e-06,
+      "loss": 0.7593,
+      "step": 5641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4409890915083312,
+      "learning_rate": 6.002674324699983e-06,
+      "loss": 0.7713,
+      "step": 5642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4965347827141842,
+      "learning_rate": 6.001401430598561e-06,
+      "loss": 0.7966,
+      "step": 5643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7098542854061902,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.7718,
+      "step": 5644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.524801070568037,
+      "learning_rate": 5.998855439630925e-06,
+      "loss": 0.8508,
+      "step": 5645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5111958297374621,
+      "learning_rate": 5.997582342936622e-06,
+      "loss": 0.8444,
+      "step": 5646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.439839326648737,
+      "learning_rate": 5.99630917888328e-06,
+      "loss": 0.8335,
+      "step": 5647
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8348560747127013,
+      "learning_rate": 5.9950359475568634e-06,
+      "loss": 1.0826,
+      "step": 5648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5984925713805744,
+      "learning_rate": 5.993762649043344e-06,
+      "loss": 0.759,
+      "step": 5649
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4774590515903818,
+      "learning_rate": 5.992489283428699e-06,
+      "loss": 0.7978,
+      "step": 5650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7722649379562184,
+      "learning_rate": 5.991215850798906e-06,
+      "loss": 1.0801,
+      "step": 5651
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6057534380497958,
+      "learning_rate": 5.989942351239954e-06,
+      "loss": 0.8025,
+      "step": 5652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9095355562917857,
+      "learning_rate": 5.988668784837831e-06,
+      "loss": 0.8037,
+      "step": 5653
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.526420885177167,
+      "learning_rate": 5.98739515167853e-06,
+      "loss": 0.7798,
+      "step": 5654
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4834909886340653,
+      "learning_rate": 5.986121451848051e-06,
+      "loss": 0.7411,
+      "step": 5655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5226607522669677,
+      "learning_rate": 5.984847685432397e-06,
+      "loss": 0.8023,
+      "step": 5656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.526595001380897,
+      "learning_rate": 5.983573852517575e-06,
+      "loss": 0.7506,
+      "step": 5657
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4268422495093869,
+      "learning_rate": 5.982299953189598e-06,
+      "loss": 0.7688,
+      "step": 5658
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5672927211415975,
+      "learning_rate": 5.9810259875344815e-06,
+      "loss": 0.7405,
+      "step": 5659
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3996835807789478,
+      "learning_rate": 5.979751955638246e-06,
+      "loss": 0.7606,
+      "step": 5660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6051727635080095,
+      "learning_rate": 5.9784778575869185e-06,
+      "loss": 0.8566,
+      "step": 5661
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6809194025368093,
+      "learning_rate": 5.97720369346653e-06,
+      "loss": 0.757,
+      "step": 5662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9438614456665771,
+      "learning_rate": 5.975929463363112e-06,
+      "loss": 1.0925,
+      "step": 5663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5273297585187247,
+      "learning_rate": 5.9746551673627065e-06,
+      "loss": 0.7874,
+      "step": 5664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.828386765177613,
+      "learning_rate": 5.973380805551354e-06,
+      "loss": 1.0894,
+      "step": 5665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4896897373247366,
+      "learning_rate": 5.972106378015104e-06,
+      "loss": 0.8328,
+      "step": 5666
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5720608127532691,
+      "learning_rate": 5.970831884840011e-06,
+      "loss": 0.7886,
+      "step": 5667
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4444822089161087,
+      "learning_rate": 5.969557326112125e-06,
+      "loss": 0.809,
+      "step": 5668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4218702735152222,
+      "learning_rate": 5.968282701917514e-06,
+      "loss": 0.7674,
+      "step": 5669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.466692819167025,
+      "learning_rate": 5.967008012342242e-06,
+      "loss": 0.74,
+      "step": 5670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.421509572885203,
+      "learning_rate": 5.965733257472374e-06,
+      "loss": 0.6694,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4922104890042747,
+      "learning_rate": 5.964458437393992e-06,
+      "loss": 0.7952,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5225312861199365,
+      "learning_rate": 5.963183552193168e-06,
+      "loss": 0.7426,
+      "step": 5673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9021362671752503,
+      "learning_rate": 5.96190860195599e-06,
+      "loss": 1.075,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4305369146787803,
+      "learning_rate": 5.9606335867685424e-06,
+      "loss": 0.748,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.458889798144112,
+      "learning_rate": 5.9593585067169195e-06,
+      "loss": 0.8969,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8543215626568816,
+      "learning_rate": 5.958083361887216e-06,
+      "loss": 1.1012,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8110222326096646,
+      "learning_rate": 5.956808152365532e-06,
+      "loss": 1.1232,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7940939975266098,
+      "learning_rate": 5.9555328782379765e-06,
+      "loss": 1.1021,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8076291784401028,
+      "learning_rate": 5.954257539590654e-06,
+      "loss": 1.0965,
+      "step": 5680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6929953635629338,
+      "learning_rate": 5.952982136509681e-06,
+      "loss": 0.7546,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4011541543137216,
+      "learning_rate": 5.951706669081174e-06,
+      "loss": 0.7925,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4129079000263332,
+      "learning_rate": 5.950431137391257e-06,
+      "loss": 0.6795,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6460306387291959,
+      "learning_rate": 5.949155541526057e-06,
+      "loss": 0.8413,
+      "step": 5684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4282296325226402,
+      "learning_rate": 5.947879881571703e-06,
+      "loss": 0.8495,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6397249859688259,
+      "learning_rate": 5.946604157614334e-06,
+      "loss": 0.8121,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6162467452259364,
+      "learning_rate": 5.945328369740088e-06,
+      "loss": 0.8288,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4478603521480466,
+      "learning_rate": 5.9440525180351064e-06,
+      "loss": 0.8153,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.497948519969562,
+      "learning_rate": 5.942776602585542e-06,
+      "loss": 0.7353,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6204420054989939,
+      "learning_rate": 5.9415006234775445e-06,
+      "loss": 0.7318,
+      "step": 5690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4135909404663896,
+      "learning_rate": 5.940224580797272e-06,
+      "loss": 0.743,
+      "step": 5691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.539401407735776,
+      "learning_rate": 5.9389484746308865e-06,
+      "loss": 0.7527,
+      "step": 5692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0395637017567574,
+      "learning_rate": 5.937672305064552e-06,
+      "loss": 0.6611,
+      "step": 5693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6370687733939733,
+      "learning_rate": 5.93639607218444e-06,
+      "loss": 0.6842,
+      "step": 5694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5938905685071598,
+      "learning_rate": 5.935119776076724e-06,
+      "loss": 0.7207,
+      "step": 5695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6148030260303214,
+      "learning_rate": 5.933843416827584e-06,
+      "loss": 0.8154,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5115429137134893,
+      "learning_rate": 5.932566994523199e-06,
+      "loss": 0.7641,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5557822956375593,
+      "learning_rate": 5.931290509249758e-06,
+      "loss": 0.8198,
+      "step": 5698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.54314141781923,
+      "learning_rate": 5.930013961093454e-06,
+      "loss": 0.7995,
+      "step": 5699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5116949930892296,
+      "learning_rate": 5.9287373501404786e-06,
+      "loss": 0.7689,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7167777262377522,
+      "learning_rate": 5.927460676477036e-06,
+      "loss": 0.8324,
+      "step": 5701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.954125251160797,
+      "learning_rate": 5.926183940189327e-06,
+      "loss": 1.0747,
+      "step": 5702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5178798232387414,
+      "learning_rate": 5.92490714136356e-06,
+      "loss": 0.7605,
+      "step": 5703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5597335039867501,
+      "learning_rate": 5.923630280085948e-06,
+      "loss": 0.8045,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5186027824687112,
+      "learning_rate": 5.922353356442706e-06,
+      "loss": 0.7547,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5648297348217117,
+      "learning_rate": 5.921076370520058e-06,
+      "loss": 0.8297,
+      "step": 5706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.459351140188791,
+      "learning_rate": 5.919799322404227e-06,
+      "loss": 0.7997,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5459568145277383,
+      "learning_rate": 5.918522212181444e-06,
+      "loss": 0.7875,
+      "step": 5708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4689232867288686,
+      "learning_rate": 5.9172450399379385e-06,
+      "loss": 0.7791,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.61753931111374,
+      "learning_rate": 5.915967805759951e-06,
+      "loss": 0.8298,
+      "step": 5710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7768477622530853,
+      "learning_rate": 5.914690509733723e-06,
+      "loss": 1.0899,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5952476727372327,
+      "learning_rate": 5.9134131519455005e-06,
+      "loss": 0.856,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8026911749478612,
+      "learning_rate": 5.912135732481533e-06,
+      "loss": 1.0508,
+      "step": 5713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5840419090575626,
+      "learning_rate": 5.910858251428077e-06,
+      "loss": 0.7421,
+      "step": 5714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4937544193307006,
+      "learning_rate": 5.909580708871388e-06,
+      "loss": 0.7678,
+      "step": 5715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.778227386729819,
+      "learning_rate": 5.908303104897728e-06,
+      "loss": 1.095,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5654589721766001,
+      "learning_rate": 5.907025439593366e-06,
+      "loss": 0.8198,
+      "step": 5717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7887550927219074,
+      "learning_rate": 5.905747713044573e-06,
+      "loss": 1.0722,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6008553739828169,
+      "learning_rate": 5.904469925337624e-06,
+      "loss": 0.7557,
+      "step": 5719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5775484063354812,
+      "learning_rate": 5.903192076558795e-06,
+      "loss": 0.7345,
+      "step": 5720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4949331650301871,
+      "learning_rate": 5.901914166794374e-06,
+      "loss": 0.8018,
+      "step": 5721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5508956187707397,
+      "learning_rate": 5.900636196130645e-06,
+      "loss": 0.7628,
+      "step": 5722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.568251182010219,
+      "learning_rate": 5.899358164653901e-06,
+      "loss": 0.7937,
+      "step": 5723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4155553890709534,
+      "learning_rate": 5.898080072450437e-06,
+      "loss": 0.7352,
+      "step": 5724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5461200075109482,
+      "learning_rate": 5.896801919606554e-06,
+      "loss": 0.7535,
+      "step": 5725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5370931150473581,
+      "learning_rate": 5.895523706208552e-06,
+      "loss": 0.7555,
+      "step": 5726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8453028187430656,
+      "learning_rate": 5.894245432342743e-06,
+      "loss": 1.0878,
+      "step": 5727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8178326000527647,
+      "learning_rate": 5.892967098095439e-06,
+      "loss": 1.0712,
+      "step": 5728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.551627418309578,
+      "learning_rate": 5.891688703552953e-06,
+      "loss": 0.8184,
+      "step": 5729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.773952446798454,
+      "learning_rate": 5.890410248801608e-06,
+      "loss": 1.0728,
+      "step": 5730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7022689027087656,
+      "learning_rate": 5.889131733927726e-06,
+      "loss": 0.7897,
+      "step": 5731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4912416571839358,
+      "learning_rate": 5.887853159017638e-06,
+      "loss": 0.727,
+      "step": 5732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4980657988746622,
+      "learning_rate": 5.886574524157672e-06,
+      "loss": 0.7159,
+      "step": 5733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5333745264976293,
+      "learning_rate": 5.885295829434168e-06,
+      "loss": 0.7615,
+      "step": 5734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.48301351597407,
+      "learning_rate": 5.884017074933466e-06,
+      "loss": 0.773,
+      "step": 5735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4909182725753882,
+      "learning_rate": 5.8827382607419084e-06,
+      "loss": 0.6996,
+      "step": 5736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6268247693364153,
+      "learning_rate": 5.8814593869458455e-06,
+      "loss": 0.7597,
+      "step": 5737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5599869043112304,
+      "learning_rate": 5.880180453631628e-06,
+      "loss": 0.8173,
+      "step": 5738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4267192239917286,
+      "learning_rate": 5.878901460885616e-06,
+      "loss": 0.7975,
+      "step": 5739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5230330742961529,
+      "learning_rate": 5.877622408794166e-06,
+      "loss": 0.8192,
+      "step": 5740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3938911660760969,
+      "learning_rate": 5.876343297443645e-06,
+      "loss": 0.7536,
+      "step": 5741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4691804421664285,
+      "learning_rate": 5.87506412692042e-06,
+      "loss": 0.7638,
+      "step": 5742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4933921322777197,
+      "learning_rate": 5.873784897310864e-06,
+      "loss": 0.796,
+      "step": 5743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5429335377287474,
+      "learning_rate": 5.872505608701354e-06,
+      "loss": 0.8118,
+      "step": 5744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4337045354895883,
+      "learning_rate": 5.87122626117827e-06,
+      "loss": 0.7163,
+      "step": 5745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3980506077269867,
+      "learning_rate": 5.869946854827996e-06,
+      "loss": 0.7991,
+      "step": 5746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5373122358805782,
+      "learning_rate": 5.868667389736924e-06,
+      "loss": 0.7672,
+      "step": 5747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5772908897266074,
+      "learning_rate": 5.8673878659914405e-06,
+      "loss": 0.7159,
+      "step": 5748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5049659914457099,
+      "learning_rate": 5.866108283677947e-06,
+      "loss": 0.8005,
+      "step": 5749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0746114102377629,
+      "learning_rate": 5.864828642882839e-06,
+      "loss": 1.097,
+      "step": 5750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4832307584955144,
+      "learning_rate": 5.863548943692525e-06,
+      "loss": 0.804,
+      "step": 5751
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.459852551796001,
+      "learning_rate": 5.862269186193412e-06,
+      "loss": 0.7892,
+      "step": 5752
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4014312083579374,
+      "learning_rate": 5.860989370471912e-06,
+      "loss": 0.7448,
+      "step": 5753
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4734907190814381,
+      "learning_rate": 5.859709496614442e-06,
+      "loss": 0.7862,
+      "step": 5754
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5761024659784348,
+      "learning_rate": 5.858429564707419e-06,
+      "loss": 0.8225,
+      "step": 5755
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5611611426032175,
+      "learning_rate": 5.857149574837269e-06,
+      "loss": 0.8523,
+      "step": 5756
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.819837240186687,
+      "learning_rate": 5.85586952709042e-06,
+      "loss": 1.108,
+      "step": 5757
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5287523348703764,
+      "learning_rate": 5.854589421553304e-06,
+      "loss": 0.7085,
+      "step": 5758
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4916046645291923,
+      "learning_rate": 5.853309258312356e-06,
+      "loss": 0.7768,
+      "step": 5759
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4800492909611327,
+      "learning_rate": 5.852029037454014e-06,
+      "loss": 0.7517,
+      "step": 5760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8400630704317602,
+      "learning_rate": 5.850748759064726e-06,
+      "loss": 1.0763,
+      "step": 5761
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5147587332498131,
+      "learning_rate": 5.849468423230934e-06,
+      "loss": 0.7944,
+      "step": 5762
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4493766481656487,
+      "learning_rate": 5.848188030039093e-06,
+      "loss": 0.8653,
+      "step": 5763
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5471727076722313,
+      "learning_rate": 5.846907579575657e-06,
+      "loss": 0.7546,
+      "step": 5764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7829873502007327,
+      "learning_rate": 5.8456270719270835e-06,
+      "loss": 1.0988,
+      "step": 5765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5366256312020363,
+      "learning_rate": 5.8443465071798365e-06,
+      "loss": 0.6816,
+      "step": 5766
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4195496763490516,
+      "learning_rate": 5.843065885420382e-06,
+      "loss": 0.6935,
+      "step": 5767
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8481618002860285,
+      "learning_rate": 5.841785206735192e-06,
+      "loss": 1.0897,
+      "step": 5768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4802867012676408,
+      "learning_rate": 5.840504471210742e-06,
+      "loss": 0.818,
+      "step": 5769
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.501874456244142,
+      "learning_rate": 5.839223678933505e-06,
+      "loss": 0.7886,
+      "step": 5770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7738976921437907,
+      "learning_rate": 5.837942829989969e-06,
+      "loss": 1.1067,
+      "step": 5771
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4535084052209677,
+      "learning_rate": 5.836661924466614e-06,
+      "loss": 0.7251,
+      "step": 5772
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5280900206679697,
+      "learning_rate": 5.835380962449936e-06,
+      "loss": 0.8349,
+      "step": 5773
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5793060713483904,
+      "learning_rate": 5.834099944026422e-06,
+      "loss": 0.7871,
+      "step": 5774
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4253156786211705,
+      "learning_rate": 5.832818869282575e-06,
+      "loss": 0.7868,
+      "step": 5775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8244166392555515,
+      "learning_rate": 5.831537738304893e-06,
+      "loss": 1.069,
+      "step": 5776
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.539197233090516,
+      "learning_rate": 5.8302565511798805e-06,
+      "loss": 0.7491,
+      "step": 5777
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3891245800445329,
+      "learning_rate": 5.828975307994048e-06,
+      "loss": 0.7212,
+      "step": 5778
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.535944895882464,
+      "learning_rate": 5.827694008833906e-06,
+      "loss": 0.8067,
+      "step": 5779
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4852202194422794,
+      "learning_rate": 5.826412653785974e-06,
+      "loss": 0.7948,
+      "step": 5780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4701036338306253,
+      "learning_rate": 5.825131242936768e-06,
+      "loss": 0.7581,
+      "step": 5781
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5637533344934782,
+      "learning_rate": 5.823849776372814e-06,
+      "loss": 0.7299,
+      "step": 5782
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6298064591796844,
+      "learning_rate": 5.82256825418064e-06,
+      "loss": 0.8367,
+      "step": 5783
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5563926281469649,
+      "learning_rate": 5.821286676446776e-06,
+      "loss": 0.8282,
+      "step": 5784
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3784518230332832,
+      "learning_rate": 5.820005043257758e-06,
+      "loss": 0.7092,
+      "step": 5785
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.575577794914877,
+      "learning_rate": 5.818723354700124e-06,
+      "loss": 0.81,
+      "step": 5786
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5200625037567643,
+      "learning_rate": 5.817441610860417e-06,
+      "loss": 0.7574,
+      "step": 5787
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4372474016304093,
+      "learning_rate": 5.816159811825184e-06,
+      "loss": 0.712,
+      "step": 5788
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.582400819450848,
+      "learning_rate": 5.814877957680973e-06,
+      "loss": 0.7774,
+      "step": 5789
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8066280226728584,
+      "learning_rate": 5.81359604851434e-06,
+      "loss": 1.0984,
+      "step": 5790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.513530275943973,
+      "learning_rate": 5.812314084411842e-06,
+      "loss": 0.7394,
+      "step": 5791
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7802058464359282,
+      "learning_rate": 5.811032065460037e-06,
+      "loss": 1.0883,
+      "step": 5792
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5913157082885896,
+      "learning_rate": 5.809749991745495e-06,
+      "loss": 0.7676,
+      "step": 5793
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5019773432266628,
+      "learning_rate": 5.808467863354781e-06,
+      "loss": 0.741,
+      "step": 5794
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.470996814837823,
+      "learning_rate": 5.807185680374467e-06,
+      "loss": 0.7939,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8065435699176291,
+      "learning_rate": 5.805903442891132e-06,
+      "loss": 1.1002,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5390604418206384,
+      "learning_rate": 5.804621150991353e-06,
+      "loss": 0.7494,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5957927341076568,
+      "learning_rate": 5.803338804761714e-06,
+      "loss": 0.8009,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.588620312225338,
+      "learning_rate": 5.8020564042888015e-06,
+      "loss": 0.7788,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5169071768935753,
+      "learning_rate": 5.8007739496592075e-06,
+      "loss": 0.7146,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4721410937612827,
+      "learning_rate": 5.7994914409595236e-06,
+      "loss": 0.8114,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.433077801767643,
+      "learning_rate": 5.798208878276352e-06,
+      "loss": 0.794,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5455097180833473,
+      "learning_rate": 5.7969262616962905e-06,
+      "loss": 0.7979,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.570103272976438,
+      "learning_rate": 5.795643591305945e-06,
+      "loss": 0.8431,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.797097769088055,
+      "learning_rate": 5.794360867191926e-06,
+      "loss": 1.0817,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8128357728704698,
+      "learning_rate": 5.7930780894408435e-06,
+      "loss": 1.0477,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5949078281655433,
+      "learning_rate": 5.7917952581393155e-06,
+      "loss": 0.8799,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.516985498254989,
+      "learning_rate": 5.790512373373962e-06,
+      "loss": 0.7622,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7489844835485704,
+      "learning_rate": 5.789229435231404e-06,
+      "loss": 1.0504,
+      "step": 5809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4866152908634782,
+      "learning_rate": 5.787946443798271e-06,
+      "loss": 0.7616,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4340081603790742,
+      "learning_rate": 5.786663399161191e-06,
+      "loss": 0.7242,
+      "step": 5811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4760383766182872,
+      "learning_rate": 5.785380301406801e-06,
+      "loss": 0.7569,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6335384722939872,
+      "learning_rate": 5.784097150621737e-06,
+      "loss": 0.7062,
+      "step": 5813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.487750567543447,
+      "learning_rate": 5.782813946892639e-06,
+      "loss": 0.7709,
+      "step": 5814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5574743909657494,
+      "learning_rate": 5.781530690306156e-06,
+      "loss": 0.7884,
+      "step": 5815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5996529275333928,
+      "learning_rate": 5.78024738094893e-06,
+      "loss": 0.8216,
+      "step": 5816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8544291642391503,
+      "learning_rate": 5.778964018907619e-06,
+      "loss": 1.1223,
+      "step": 5817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8169870089796534,
+      "learning_rate": 5.777680604268876e-06,
+      "loss": 1.0772,
+      "step": 5818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.449767058791186,
+      "learning_rate": 5.776397137119362e-06,
+      "loss": 0.7769,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5185048417343046,
+      "learning_rate": 5.775113617545735e-06,
+      "loss": 0.8004,
+      "step": 5820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4564064452685923,
+      "learning_rate": 5.773830045634664e-06,
+      "loss": 0.7787,
+      "step": 5821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.599756428510654,
+      "learning_rate": 5.772546421472821e-06,
+      "loss": 0.7891,
+      "step": 5822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5164228984188264,
+      "learning_rate": 5.771262745146876e-06,
+      "loss": 0.7594,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5034470824594237,
+      "learning_rate": 5.769979016743508e-06,
+      "loss": 0.743,
+      "step": 5824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7454844252835044,
+      "learning_rate": 5.768695236349396e-06,
+      "loss": 0.7202,
+      "step": 5825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4946602878420343,
+      "learning_rate": 5.767411404051222e-06,
+      "loss": 0.8346,
+      "step": 5826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9335828328616237,
+      "learning_rate": 5.766127519935676e-06,
+      "loss": 1.0634,
+      "step": 5827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4205622027225868,
+      "learning_rate": 5.7648435840894475e-06,
+      "loss": 0.7832,
+      "step": 5828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8380583209855147,
+      "learning_rate": 5.763559596599233e-06,
+      "loss": 1.1054,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8501707813433197,
+      "learning_rate": 5.762275557551728e-06,
+      "loss": 0.7687,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5985793843080427,
+      "learning_rate": 5.760991467033634e-06,
+      "loss": 0.7126,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5647886047898651,
+      "learning_rate": 5.759707325131656e-06,
+      "loss": 0.7593,
+      "step": 5832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5474184707853607,
+      "learning_rate": 5.758423131932501e-06,
+      "loss": 0.7409,
+      "step": 5833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4768617684505911,
+      "learning_rate": 5.757138887522884e-06,
+      "loss": 0.7363,
+      "step": 5834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4760777722118392,
+      "learning_rate": 5.755854591989518e-06,
+      "loss": 0.7675,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.439785343754657,
+      "learning_rate": 5.754570245419121e-06,
+      "loss": 0.8068,
+      "step": 5836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.640506322433333,
+      "learning_rate": 5.7532858478984144e-06,
+      "loss": 0.7746,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.562413802742019,
+      "learning_rate": 5.752001399514125e-06,
+      "loss": 0.7913,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1005026651926941,
+      "learning_rate": 5.750716900352983e-06,
+      "loss": 1.088,
+      "step": 5839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5284037257601057,
+      "learning_rate": 5.749432350501718e-06,
+      "loss": 0.8084,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5111665818828572,
+      "learning_rate": 5.7481477500470695e-06,
+      "loss": 0.7495,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6688530806542463,
+      "learning_rate": 5.746863099075771e-06,
+      "loss": 0.8173,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7953212401547314,
+      "learning_rate": 5.74557839767457e-06,
+      "loss": 1.0993,
+      "step": 5843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8436862181802415,
+      "learning_rate": 5.74429364593021e-06,
+      "loss": 1.0853,
+      "step": 5844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6110560579734319,
+      "learning_rate": 5.743008843929441e-06,
+      "loss": 0.765,
+      "step": 5845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5554662933301058,
+      "learning_rate": 5.741723991759016e-06,
+      "loss": 0.786,
+      "step": 5846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4614745592493747,
+      "learning_rate": 5.740439089505691e-06,
+      "loss": 0.6362,
+      "step": 5847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1711884828118513,
+      "learning_rate": 5.739154137256227e-06,
+      "loss": 0.7614,
+      "step": 5848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9381442857265868,
+      "learning_rate": 5.7378691350973835e-06,
+      "loss": 1.0614,
+      "step": 5849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4422838664769437,
+      "learning_rate": 5.736584083115929e-06,
+      "loss": 0.6935,
+      "step": 5850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4289419271289645,
+      "learning_rate": 5.735298981398634e-06,
+      "loss": 0.7549,
+      "step": 5851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.46133656806265,
+      "learning_rate": 5.73401383003227e-06,
+      "loss": 0.742,
+      "step": 5852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.608701101132847,
+      "learning_rate": 5.732728629103615e-06,
+      "loss": 0.7362,
+      "step": 5853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4837608511146556,
+      "learning_rate": 5.731443378699445e-06,
+      "loss": 0.7248,
+      "step": 5854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5114090157954212,
+      "learning_rate": 5.730158078906546e-06,
+      "loss": 0.8165,
+      "step": 5855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6099166435456116,
+      "learning_rate": 5.728872729811705e-06,
+      "loss": 0.795,
+      "step": 5856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5491046098025547,
+      "learning_rate": 5.72758733150171e-06,
+      "loss": 0.7501,
+      "step": 5857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5750484340442703,
+      "learning_rate": 5.726301884063356e-06,
+      "loss": 0.781,
+      "step": 5858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.496490261892033,
+      "learning_rate": 5.725016387583435e-06,
+      "loss": 0.786,
+      "step": 5859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9315131779074122,
+      "learning_rate": 5.723730842148752e-06,
+      "loss": 1.0825,
+      "step": 5860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8808526810077502,
+      "learning_rate": 5.722445247846107e-06,
+      "loss": 1.1027,
+      "step": 5861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5362984546215928,
+      "learning_rate": 5.721159604762307e-06,
+      "loss": 0.8489,
+      "step": 5862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5994619417088973,
+      "learning_rate": 5.719873912984163e-06,
+      "loss": 0.8417,
+      "step": 5863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6334479676120663,
+      "learning_rate": 5.7185881725984835e-06,
+      "loss": 0.7365,
+      "step": 5864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5895892406269867,
+      "learning_rate": 5.71730238369209e-06,
+      "loss": 0.7966,
+      "step": 5865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5888211138816024,
+      "learning_rate": 5.716016546351797e-06,
+      "loss": 0.7566,
+      "step": 5866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4516495878645554,
+      "learning_rate": 5.714730660664429e-06,
+      "loss": 0.7459,
+      "step": 5867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5262962520940098,
+      "learning_rate": 5.713444726716814e-06,
+      "loss": 0.7556,
+      "step": 5868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5779654073220646,
+      "learning_rate": 5.712158744595781e-06,
+      "loss": 0.7889,
+      "step": 5869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.607065773107172,
+      "learning_rate": 5.71087271438816e-06,
+      "loss": 0.7714,
+      "step": 5870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6125608877174562,
+      "learning_rate": 5.709586636180787e-06,
+      "loss": 0.7576,
+      "step": 5871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4907071487516352,
+      "learning_rate": 5.708300510060502e-06,
+      "loss": 0.7055,
+      "step": 5872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.608952818578593,
+      "learning_rate": 5.707014336114147e-06,
+      "loss": 0.8863,
+      "step": 5873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5624774241030055,
+      "learning_rate": 5.705728114428568e-06,
+      "loss": 0.775,
+      "step": 5874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4638403840208445,
+      "learning_rate": 5.704441845090614e-06,
+      "loss": 0.7494,
+      "step": 5875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4980128401365091,
+      "learning_rate": 5.703155528187133e-06,
+      "loss": 0.7738,
+      "step": 5876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.027160937238546,
+      "learning_rate": 5.701869163804985e-06,
+      "loss": 1.0507,
+      "step": 5877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5049704567862197,
+      "learning_rate": 5.700582752031025e-06,
+      "loss": 0.7648,
+      "step": 5878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.512652132554367,
+      "learning_rate": 5.699296292952117e-06,
+      "loss": 0.7476,
+      "step": 5879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5847666308782775,
+      "learning_rate": 5.698009786655123e-06,
+      "loss": 0.7543,
+      "step": 5880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5143370777107985,
+      "learning_rate": 5.6967232332269116e-06,
+      "loss": 0.6776,
+      "step": 5881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4209244589937435,
+      "learning_rate": 5.695436632754356e-06,
+      "loss": 0.7523,
+      "step": 5882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6920724998689949,
+      "learning_rate": 5.694149985324326e-06,
+      "loss": 0.744,
+      "step": 5883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4988884922405947,
+      "learning_rate": 5.6928632910237035e-06,
+      "loss": 0.7717,
+      "step": 5884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4814773030328379,
+      "learning_rate": 5.691576549939369e-06,
+      "loss": 0.7758,
+      "step": 5885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6157679904086464,
+      "learning_rate": 5.690289762158203e-06,
+      "loss": 0.7836,
+      "step": 5886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.588512627107001,
+      "learning_rate": 5.689002927767094e-06,
+      "loss": 0.7802,
+      "step": 5887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4468213962473515,
+      "learning_rate": 5.687716046852931e-06,
+      "loss": 0.7751,
+      "step": 5888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.623802354204206,
+      "learning_rate": 5.686429119502608e-06,
+      "loss": 0.8245,
+      "step": 5889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5124256001021952,
+      "learning_rate": 5.685142145803021e-06,
+      "loss": 0.6293,
+      "step": 5890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4070623924393106,
+      "learning_rate": 5.683855125841071e-06,
+      "loss": 0.7372,
+      "step": 5891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9991227072035811,
+      "learning_rate": 5.682568059703659e-06,
+      "loss": 1.0781,
+      "step": 5892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.407788834418454,
+      "learning_rate": 5.68128094747769e-06,
+      "loss": 0.724,
+      "step": 5893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8206247133721507,
+      "learning_rate": 5.679993789250075e-06,
+      "loss": 1.066,
+      "step": 5894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4889730438300162,
+      "learning_rate": 5.678706585107721e-06,
+      "loss": 0.791,
+      "step": 5895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5357798763881876,
+      "learning_rate": 5.677419335137549e-06,
+      "loss": 0.7585,
+      "step": 5896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6044705950576097,
+      "learning_rate": 5.676132039426475e-06,
+      "loss": 0.8208,
+      "step": 5897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5864313745933896,
+      "learning_rate": 5.674844698061419e-06,
+      "loss": 0.8241,
+      "step": 5898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8708834939219458,
+      "learning_rate": 5.673557311129306e-06,
+      "loss": 1.0945,
+      "step": 5899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5348855680610956,
+      "learning_rate": 5.672269878717063e-06,
+      "loss": 0.8096,
+      "step": 5900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4673275804687378,
+      "learning_rate": 5.67098240091162e-06,
+      "loss": 0.7726,
+      "step": 5901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6026638838786798,
+      "learning_rate": 5.669694877799912e-06,
+      "loss": 0.7184,
+      "step": 5902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4020098717446732,
+      "learning_rate": 5.668407309468873e-06,
+      "loss": 0.7547,
+      "step": 5903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7965778819140131,
+      "learning_rate": 5.667119696005445e-06,
+      "loss": 1.0741,
+      "step": 5904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5136555928844029,
+      "learning_rate": 5.665832037496569e-06,
+      "loss": 0.6591,
+      "step": 5905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.525527203728997,
+      "learning_rate": 5.664544334029193e-06,
+      "loss": 0.7488,
+      "step": 5906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5670808951194715,
+      "learning_rate": 5.663256585690263e-06,
+      "loss": 0.8026,
+      "step": 5907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5789505974292397,
+      "learning_rate": 5.661968792566731e-06,
+      "loss": 0.7475,
+      "step": 5908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5566919617571968,
+      "learning_rate": 5.660680954745554e-06,
+      "loss": 0.7586,
+      "step": 5909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.657428890143161,
+      "learning_rate": 5.659393072313687e-06,
+      "loss": 0.81,
+      "step": 5910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8462584332209059,
+      "learning_rate": 5.658105145358093e-06,
+      "loss": 1.075,
+      "step": 5911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7963206079158476,
+      "learning_rate": 5.656817173965733e-06,
+      "loss": 1.0993,
+      "step": 5912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.50766264428043,
+      "learning_rate": 5.655529158223577e-06,
+      "loss": 0.7685,
+      "step": 5913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5245860187361033,
+      "learning_rate": 5.654241098218594e-06,
+      "loss": 0.7412,
+      "step": 5914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4493064630388122,
+      "learning_rate": 5.6529529940377526e-06,
+      "loss": 0.7498,
+      "step": 5915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4824137286193177,
+      "learning_rate": 5.651664845768036e-06,
+      "loss": 0.7086,
+      "step": 5916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8019281354706871,
+      "learning_rate": 5.6503766534964156e-06,
+      "loss": 1.0756,
+      "step": 5917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4536695958373542,
+      "learning_rate": 5.649088417309878e-06,
+      "loss": 0.7049,
+      "step": 5918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5041696370498356,
+      "learning_rate": 5.647800137295407e-06,
+      "loss": 0.7508,
+      "step": 5919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.444701034643854,
+      "learning_rate": 5.646511813539987e-06,
+      "loss": 0.7732,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6699440848371692,
+      "learning_rate": 5.6452234461306145e-06,
+      "loss": 0.7844,
+      "step": 5921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8523935507607763,
+      "learning_rate": 5.6439350351542765e-06,
+      "loss": 1.0795,
+      "step": 5922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5493872203455221,
+      "learning_rate": 5.642646580697974e-06,
+      "loss": 0.7777,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.463241233707051,
+      "learning_rate": 5.641358082848705e-06,
+      "loss": 0.7354,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.347545773481136,
+      "learning_rate": 5.64006954169347e-06,
+      "loss": 0.7244,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3936973973131739,
+      "learning_rate": 5.638780957319278e-06,
+      "loss": 0.7214,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4891159645152297,
+      "learning_rate": 5.637492329813133e-06,
+      "loss": 0.7874,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5555941579779757,
+      "learning_rate": 5.636203659262049e-06,
+      "loss": 0.7314,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4047199627721416,
+      "learning_rate": 5.634914945753041e-06,
+      "loss": 0.7842,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8298289933968395,
+      "learning_rate": 5.633626189373123e-06,
+      "loss": 1.038,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5751175922925589,
+      "learning_rate": 5.632337390209315e-06,
+      "loss": 0.7756,
+      "step": 5931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4201299297873249,
+      "learning_rate": 5.63104854834864e-06,
+      "loss": 0.7292,
+      "step": 5932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6040853553982801,
+      "learning_rate": 5.629759663878125e-06,
+      "loss": 0.8243,
+      "step": 5933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5352924023310548,
+      "learning_rate": 5.628470736884797e-06,
+      "loss": 0.7589,
+      "step": 5934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3445162260845536,
+      "learning_rate": 5.627181767455688e-06,
+      "loss": 0.7096,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5821437914019594,
+      "learning_rate": 5.625892755677833e-06,
+      "loss": 0.8604,
+      "step": 5936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.473341032007916,
+      "learning_rate": 5.624603701638266e-06,
+      "loss": 0.7553,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6007093509528745,
+      "learning_rate": 5.623314605424031e-06,
+      "loss": 0.7745,
+      "step": 5938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.507218267562719,
+      "learning_rate": 5.622025467122167e-06,
+      "loss": 0.8203,
+      "step": 5939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5314831656870473,
+      "learning_rate": 5.620736286819721e-06,
+      "loss": 0.8062,
+      "step": 5940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4992307863405472,
+      "learning_rate": 5.619447064603743e-06,
+      "loss": 0.7829,
+      "step": 5941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.564854248079923,
+      "learning_rate": 5.6181578005612805e-06,
+      "loss": 0.8141,
+      "step": 5942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5333252733700806,
+      "learning_rate": 5.616868494779391e-06,
+      "loss": 0.8169,
+      "step": 5943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4499270751445121,
+      "learning_rate": 5.61557914734513e-06,
+      "loss": 0.7379,
+      "step": 5944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4100564604428272,
+      "learning_rate": 5.614289758345558e-06,
+      "loss": 0.8083,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5580319585459659,
+      "learning_rate": 5.613000327867737e-06,
+      "loss": 0.7592,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.601481866364937,
+      "learning_rate": 5.611710855998732e-06,
+      "loss": 0.778,
+      "step": 5947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8199154944599473,
+      "learning_rate": 5.610421342825611e-06,
+      "loss": 1.095,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8079723764775566,
+      "learning_rate": 5.6091317884354435e-06,
+      "loss": 1.0976,
+      "step": 5949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6121274992897798,
+      "learning_rate": 5.607842192915307e-06,
+      "loss": 0.7935,
+      "step": 5950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.930476984581987,
+      "learning_rate": 5.606552556352275e-06,
+      "loss": 0.7101,
+      "step": 5951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.415745758595045,
+      "learning_rate": 5.6052628788334285e-06,
+      "loss": 0.7505,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6596705883597394,
+      "learning_rate": 5.603973160445846e-06,
+      "loss": 0.7445,
+      "step": 5953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0779013623514633,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.8563,
+      "step": 5954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9553806677230609,
+      "learning_rate": 5.601393601412825e-06,
+      "loss": 1.1122,
+      "step": 5955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8319055871420634,
+      "learning_rate": 5.600103760941561e-06,
+      "loss": 1.0582,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5860280972272478,
+      "learning_rate": 5.598813879949922e-06,
+      "loss": 0.7836,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4875453002150707,
+      "learning_rate": 5.597523958524999e-06,
+      "loss": 0.8224,
+      "step": 5958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4599186957439978,
+      "learning_rate": 5.5962339967538915e-06,
+      "loss": 0.7772,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4302685095143886,
+      "learning_rate": 5.5949439947237004e-06,
+      "loss": 0.81,
+      "step": 5960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.84489409005719,
+      "learning_rate": 5.593653952521532e-06,
+      "loss": 1.0758,
+      "step": 5961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4123792497786907,
+      "learning_rate": 5.5923638702344905e-06,
+      "loss": 0.7303,
+      "step": 5962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.406677621001775,
+      "learning_rate": 5.591073747949687e-06,
+      "loss": 0.7461,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8485569446136199,
+      "learning_rate": 5.5897835857542315e-06,
+      "loss": 0.8364,
+      "step": 5964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4700971802809797,
+      "learning_rate": 5.588493383735239e-06,
+      "loss": 0.7785,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5219937783052218,
+      "learning_rate": 5.587203141979828e-06,
+      "loss": 0.7831,
+      "step": 5966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5962996102716793,
+      "learning_rate": 5.585912860575119e-06,
+      "loss": 0.7269,
+      "step": 5967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4927877264214104,
+      "learning_rate": 5.584622539608234e-06,
+      "loss": 0.7713,
+      "step": 5968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.565317310355898,
+      "learning_rate": 5.5833321791662975e-06,
+      "loss": 0.8127,
+      "step": 5969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0281768717100572,
+      "learning_rate": 5.582041779336438e-06,
+      "loss": 0.8077,
+      "step": 5970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5184433380331699,
+      "learning_rate": 5.580751340205788e-06,
+      "loss": 0.8087,
+      "step": 5971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7930397032455248,
+      "learning_rate": 5.579460861861477e-06,
+      "loss": 1.0772,
+      "step": 5972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.445560748985988,
+      "learning_rate": 5.578170344390647e-06,
+      "loss": 0.8262,
+      "step": 5973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7424757517189902,
+      "learning_rate": 5.576879787880432e-06,
+      "loss": 0.7891,
+      "step": 5974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.574812538181466,
+      "learning_rate": 5.575589192417973e-06,
+      "loss": 0.8085,
+      "step": 5975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5688656199310413,
+      "learning_rate": 5.574298558090418e-06,
+      "loss": 0.7845,
+      "step": 5976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5534326478579819,
+      "learning_rate": 5.573007884984909e-06,
+      "loss": 0.7844,
+      "step": 5977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7929497828212151,
+      "learning_rate": 5.571717173188597e-06,
+      "loss": 1.0588,
+      "step": 5978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.646593233246233,
+      "learning_rate": 5.570426422788636e-06,
+      "loss": 0.7443,
+      "step": 5979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5418093913835174,
+      "learning_rate": 5.569135633872178e-06,
+      "loss": 0.7411,
+      "step": 5980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5499448781662335,
+      "learning_rate": 5.567844806526381e-06,
+      "loss": 0.7305,
+      "step": 5981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5329073609701704,
+      "learning_rate": 5.566553940838404e-06,
+      "loss": 0.7944,
+      "step": 5982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5794353471686298,
+      "learning_rate": 5.565263036895409e-06,
+      "loss": 0.8752,
+      "step": 5983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4822037241526296,
+      "learning_rate": 5.563972094784561e-06,
+      "loss": 0.728,
+      "step": 5984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.48288199500195,
+      "learning_rate": 5.562681114593028e-06,
+      "loss": 0.8409,
+      "step": 5985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5708098794069498,
+      "learning_rate": 5.5613900964079805e-06,
+      "loss": 0.7635,
+      "step": 5986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4902254962709587,
+      "learning_rate": 5.560099040316588e-06,
+      "loss": 0.8317,
+      "step": 5987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5802122031892938,
+      "learning_rate": 5.5588079464060285e-06,
+      "loss": 0.78,
+      "step": 5988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4117193943684558,
+      "learning_rate": 5.557516814763478e-06,
+      "loss": 0.6466,
+      "step": 5989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6150346105788806,
+      "learning_rate": 5.556225645476119e-06,
+      "loss": 0.7792,
+      "step": 5990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4755418146474861,
+      "learning_rate": 5.5549344386311325e-06,
+      "loss": 0.755,
+      "step": 5991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8332869822352872,
+      "learning_rate": 5.553643194315702e-06,
+      "loss": 1.1144,
+      "step": 5992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5628245970710906,
+      "learning_rate": 5.552351912617017e-06,
+      "loss": 0.8442,
+      "step": 5993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7902071376801086,
+      "learning_rate": 5.551060593622269e-06,
+      "loss": 1.0962,
+      "step": 5994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5274324628015281,
+      "learning_rate": 5.549769237418649e-06,
+      "loss": 0.7402,
+      "step": 5995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6032663134152718,
+      "learning_rate": 5.548477844093354e-06,
+      "loss": 0.7946,
+      "step": 5996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6166058572186293,
+      "learning_rate": 5.547186413733579e-06,
+      "loss": 0.8306,
+      "step": 5997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.565688609490214,
+      "learning_rate": 5.545894946426529e-06,
+      "loss": 0.7431,
+      "step": 5998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4712657595278058,
+      "learning_rate": 5.544603442259401e-06,
+      "loss": 0.7215,
+      "step": 5999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3837086071775688,
+      "learning_rate": 5.543311901319405e-06,
+      "loss": 0.7301,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4271384867310195,
+      "learning_rate": 5.542020323693745e-06,
+      "loss": 0.7714,
+      "step": 6001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1842659078429074,
+      "learning_rate": 5.540728709469636e-06,
+      "loss": 0.7552,
+      "step": 6002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.855122248399551,
+      "learning_rate": 5.539437058734287e-06,
+      "loss": 1.0891,
+      "step": 6003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.580855109539842,
+      "learning_rate": 5.538145371574913e-06,
+      "loss": 0.7704,
+      "step": 6004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.464140866988004,
+      "learning_rate": 5.536853648078735e-06,
+      "loss": 0.8278,
+      "step": 6005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4618854422445564,
+      "learning_rate": 5.53556188833297e-06,
+      "loss": 0.7636,
+      "step": 6006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6363193168488492,
+      "learning_rate": 5.534270092424843e-06,
+      "loss": 0.751,
+      "step": 6007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5727034029029818,
+      "learning_rate": 5.532978260441576e-06,
+      "loss": 0.8132,
+      "step": 6008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7747358686722856,
+      "learning_rate": 5.5316863924703986e-06,
+      "loss": 0.7117,
+      "step": 6009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0564444029218185,
+      "learning_rate": 5.5303944885985405e-06,
+      "loss": 0.8769,
+      "step": 6010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7045341414835142,
+      "learning_rate": 5.529102548913233e-06,
+      "loss": 0.7454,
+      "step": 6011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5579177349516862,
+      "learning_rate": 5.527810573501713e-06,
+      "loss": 0.7409,
+      "step": 6012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.422780670899157,
+      "learning_rate": 5.526518562451215e-06,
+      "loss": 0.7363,
+      "step": 6013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5106972941620882,
+      "learning_rate": 5.525226515848979e-06,
+      "loss": 0.7387,
+      "step": 6014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6265852660676492,
+      "learning_rate": 5.52393443378225e-06,
+      "loss": 0.7616,
+      "step": 6015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6410639198854176,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.7269,
+      "step": 6016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5610594468328525,
+      "learning_rate": 5.521350163604282e-06,
+      "loss": 0.7979,
+      "step": 6017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4629038144503204,
+      "learning_rate": 5.520057975667542e-06,
+      "loss": 0.7269,
+      "step": 6018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4288792448484384,
+      "learning_rate": 5.518765752615297e-06,
+      "loss": 0.7394,
+      "step": 6019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5063641645832095,
+      "learning_rate": 5.517473494534803e-06,
+      "loss": 0.7594,
+      "step": 6020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5038044975225866,
+      "learning_rate": 5.516181201513314e-06,
+      "loss": 0.7605,
+      "step": 6021
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3836185044801907,
+      "learning_rate": 5.514888873638091e-06,
+      "loss": 0.7421,
+      "step": 6022
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8421231505930131,
+      "learning_rate": 5.513596510996393e-06,
+      "loss": 1.0979,
+      "step": 6023
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.553669247391853,
+      "learning_rate": 5.5123041136754865e-06,
+      "loss": 0.888,
+      "step": 6024
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4512625940203483,
+      "learning_rate": 5.5110116817626335e-06,
+      "loss": 0.7492,
+      "step": 6025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4862261204081997,
+      "learning_rate": 5.5097192153451014e-06,
+      "loss": 0.7676,
+      "step": 6026
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5556693695866262,
+      "learning_rate": 5.508426714510164e-06,
+      "loss": 0.761,
+      "step": 6027
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.442847396954611,
+      "learning_rate": 5.507134179345093e-06,
+      "loss": 0.7573,
+      "step": 6028
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8330335174866577,
+      "learning_rate": 5.505841609937162e-06,
+      "loss": 1.1024,
+      "step": 6029
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6866642327377033,
+      "learning_rate": 5.504549006373649e-06,
+      "loss": 0.83,
+      "step": 6030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.570547416682292,
+      "learning_rate": 5.503256368741832e-06,
+      "loss": 0.828,
+      "step": 6031
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7814164692416823,
+      "learning_rate": 5.501963697128995e-06,
+      "loss": 1.0829,
+      "step": 6032
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5516604901744246,
+      "learning_rate": 5.500670991622421e-06,
+      "loss": 0.8551,
+      "step": 6033
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4275974473526831,
+      "learning_rate": 5.499378252309397e-06,
+      "loss": 0.8281,
+      "step": 6034
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4586287409128862,
+      "learning_rate": 5.498085479277213e-06,
+      "loss": 0.8356,
+      "step": 6035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4996655813804465,
+      "learning_rate": 5.496792672613157e-06,
+      "loss": 0.7862,
+      "step": 6036
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8486346400343505,
+      "learning_rate": 5.495499832404525e-06,
+      "loss": 0.7251,
+      "step": 6037
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.564745757573283,
+      "learning_rate": 5.494206958738609e-06,
+      "loss": 0.7719,
+      "step": 6038
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8354563046202366,
+      "learning_rate": 5.492914051702711e-06,
+      "loss": 1.0724,
+      "step": 6039
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.541223225414002,
+      "learning_rate": 5.491621111384129e-06,
+      "loss": 0.8607,
+      "step": 6040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5908053992177627,
+      "learning_rate": 5.490328137870164e-06,
+      "loss": 0.7671,
+      "step": 6041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7886766426600219,
+      "learning_rate": 5.489035131248124e-06,
+      "loss": 1.078,
+      "step": 6042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8007623465262053,
+      "learning_rate": 5.487742091605311e-06,
+      "loss": 1.0837,
+      "step": 6043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6578550735102104,
+      "learning_rate": 5.4864490190290386e-06,
+      "loss": 0.7808,
+      "step": 6044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5148744820641664,
+      "learning_rate": 5.4851559136066154e-06,
+      "loss": 0.6848,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3716273320120578,
+      "learning_rate": 5.483862775425358e-06,
+      "loss": 0.7081,
+      "step": 6046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5378673225917214,
+      "learning_rate": 5.482569604572577e-06,
+      "loss": 0.7926,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4797577436293656,
+      "learning_rate": 5.481276401135592e-06,
+      "loss": 0.8041,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5336177656956764,
+      "learning_rate": 5.479983165201726e-06,
+      "loss": 0.7295,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3875361825796522,
+      "learning_rate": 5.478689896858298e-06,
+      "loss": 0.7232,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8449530839207916,
+      "learning_rate": 5.477396596192633e-06,
+      "loss": 1.0746,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4499606630082549,
+      "learning_rate": 5.476103263292061e-06,
+      "loss": 0.8577,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4614611194311344,
+      "learning_rate": 5.474809898243905e-06,
+      "loss": 0.8287,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.525757170750006,
+      "learning_rate": 5.4735165011355005e-06,
+      "loss": 0.8469,
+      "step": 6054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4898577419330654,
+      "learning_rate": 5.472223072054178e-06,
+      "loss": 0.7924,
+      "step": 6055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3858086720299816,
+      "learning_rate": 5.470929611087274e-06,
+      "loss": 0.7982,
+      "step": 6056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.523584581811622,
+      "learning_rate": 5.469636118322128e-06,
+      "loss": 0.7631,
+      "step": 6057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4794027085902963,
+      "learning_rate": 5.468342593846075e-06,
+      "loss": 0.7928,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4930709201544579,
+      "learning_rate": 5.46704903774646e-06,
+      "loss": 0.7905,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6941601014354108,
+      "learning_rate": 5.465755450110624e-06,
+      "loss": 0.7748,
+      "step": 6060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5269076739738683,
+      "learning_rate": 5.464461831025918e-06,
+      "loss": 0.6925,
+      "step": 6061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3731826313223685,
+      "learning_rate": 5.463168180579686e-06,
+      "loss": 0.7844,
+      "step": 6062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5827418051772875,
+      "learning_rate": 5.461874498859281e-06,
+      "loss": 0.8311,
+      "step": 6063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7952820429839367,
+      "learning_rate": 5.4605807859520506e-06,
+      "loss": 1.0773,
+      "step": 6064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7079853422440503,
+      "learning_rate": 5.4592870419453534e-06,
+      "loss": 0.8215,
+      "step": 6065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4577486148060619,
+      "learning_rate": 5.457993266926546e-06,
+      "loss": 0.8119,
+      "step": 6066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7888077510417876,
+      "learning_rate": 5.456699460982983e-06,
+      "loss": 1.0781,
+      "step": 6067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5937897025592462,
+      "learning_rate": 5.455405624202032e-06,
+      "loss": 0.7633,
+      "step": 6068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4554191009533146,
+      "learning_rate": 5.45411175667105e-06,
+      "loss": 0.677,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.791576231802028,
+      "learning_rate": 5.452817858477404e-06,
+      "loss": 1.0731,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7966236637647214,
+      "learning_rate": 5.451523929708461e-06,
+      "loss": 1.1007,
+      "step": 6071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4841474244410227,
+      "learning_rate": 5.45022997045159e-06,
+      "loss": 0.7929,
+      "step": 6072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.500995350369392,
+      "learning_rate": 5.448935980794161e-06,
+      "loss": 0.7996,
+      "step": 6073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5378250893131635,
+      "learning_rate": 5.447641960823549e-06,
+      "loss": 0.7429,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4551455566952733,
+      "learning_rate": 5.446347910627128e-06,
+      "loss": 0.7279,
+      "step": 6075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.592300245576374,
+      "learning_rate": 5.445053830292274e-06,
+      "loss": 0.7401,
+      "step": 6076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5037374430158446,
+      "learning_rate": 5.443759719906369e-06,
+      "loss": 0.7925,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.534221426740775,
+      "learning_rate": 5.442465579556793e-06,
+      "loss": 0.7847,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.828382904565253,
+      "learning_rate": 5.4411714093309295e-06,
+      "loss": 1.0995,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5234031233884195,
+      "learning_rate": 5.4398772093161635e-06,
+      "loss": 0.775,
+      "step": 6080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4966825722778214,
+      "learning_rate": 5.4385829795998815e-06,
+      "loss": 0.7703,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5006025877376132,
+      "learning_rate": 5.4372887202694735e-06,
+      "loss": 0.7246,
+      "step": 6082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7611038898183065,
+      "learning_rate": 5.435994431412334e-06,
+      "loss": 1.0626,
+      "step": 6083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5312711852464498,
+      "learning_rate": 5.434700113115852e-06,
+      "loss": 0.7951,
+      "step": 6084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5152597241907526,
+      "learning_rate": 5.433405765467424e-06,
+      "loss": 0.8271,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5183001346406908,
+      "learning_rate": 5.432111388554448e-06,
+      "loss": 0.7933,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4737755240970125,
+      "learning_rate": 5.430816982464322e-06,
+      "loss": 0.8196,
+      "step": 6087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5043825152904817,
+      "learning_rate": 5.429522547284449e-06,
+      "loss": 0.7877,
+      "step": 6088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5971530205552558,
+      "learning_rate": 5.42822808310223e-06,
+      "loss": 0.733,
+      "step": 6089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.673784769977393,
+      "learning_rate": 5.426933590005076e-06,
+      "loss": 0.8405,
+      "step": 6090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5460038854258638,
+      "learning_rate": 5.425639068080386e-06,
+      "loss": 0.8078,
+      "step": 6091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6909644001045452,
+      "learning_rate": 5.424344517415574e-06,
+      "loss": 0.7843,
+      "step": 6092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.612815367268557,
+      "learning_rate": 5.423049938098048e-06,
+      "loss": 0.801,
+      "step": 6093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4592798146681034,
+      "learning_rate": 5.421755330215223e-06,
+      "loss": 0.7615,
+      "step": 6094
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.434752420565882,
+      "learning_rate": 5.420460693854517e-06,
+      "loss": 0.7936,
+      "step": 6095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5426359385359758,
+      "learning_rate": 5.419166029103342e-06,
+      "loss": 0.7573,
+      "step": 6096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5516441126522518,
+      "learning_rate": 5.417871336049119e-06,
+      "loss": 0.8267,
+      "step": 6097
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8718892416100053,
+      "learning_rate": 5.416576614779265e-06,
+      "loss": 1.0703,
+      "step": 6098
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.702384028174739,
+      "learning_rate": 5.415281865381207e-06,
+      "loss": 0.8367,
+      "step": 6099
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7920825337805188,
+      "learning_rate": 5.413987087942369e-06,
+      "loss": 1.0711,
+      "step": 6100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4040724025534994,
+      "learning_rate": 5.412692282550175e-06,
+      "loss": 0.6389,
+      "step": 6101
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.632530556748743,
+      "learning_rate": 5.411397449292054e-06,
+      "loss": 0.8328,
+      "step": 6102
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3902769735570726,
+      "learning_rate": 5.410102588255437e-06,
+      "loss": 0.8013,
+      "step": 6103
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5398472466172348,
+      "learning_rate": 5.4088076995277564e-06,
+      "loss": 0.7995,
+      "step": 6104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6350555583317152,
+      "learning_rate": 5.407512783196443e-06,
+      "loss": 0.7636,
+      "step": 6105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6663666022064754,
+      "learning_rate": 5.406217839348936e-06,
+      "loss": 0.7771,
+      "step": 6106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6400131034928693,
+      "learning_rate": 5.404922868072673e-06,
+      "loss": 0.7562,
+      "step": 6107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2447378219478407,
+      "learning_rate": 5.403627869455089e-06,
+      "loss": 1.0713,
+      "step": 6108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3918762533679103,
+      "learning_rate": 5.402332843583631e-06,
+      "loss": 0.7488,
+      "step": 6109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8477741799867561,
+      "learning_rate": 5.401037790545737e-06,
+      "loss": 1.1029,
+      "step": 6110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4835920265437306,
+      "learning_rate": 5.399742710428855e-06,
+      "loss": 0.7269,
+      "step": 6111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5338598153774472,
+      "learning_rate": 5.398447603320433e-06,
+      "loss": 0.7006,
+      "step": 6112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5027443722809748,
+      "learning_rate": 5.397152469307916e-06,
+      "loss": 0.7093,
+      "step": 6113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5820138990129178,
+      "learning_rate": 5.395857308478757e-06,
+      "loss": 0.6186,
+      "step": 6114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.482190458505023,
+      "learning_rate": 5.394562120920407e-06,
+      "loss": 0.7594,
+      "step": 6115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4613754207850298,
+      "learning_rate": 5.39326690672032e-06,
+      "loss": 0.8068,
+      "step": 6116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4781941154565186,
+      "learning_rate": 5.3919716659659515e-06,
+      "loss": 0.7615,
+      "step": 6117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5085945361103859,
+      "learning_rate": 5.390676398744762e-06,
+      "loss": 0.8003,
+      "step": 6118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4974122857455892,
+      "learning_rate": 5.389381105144208e-06,
+      "loss": 0.77,
+      "step": 6119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.656554571389215,
+      "learning_rate": 5.38808578525175e-06,
+      "loss": 0.7482,
+      "step": 6120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.635294938167728,
+      "learning_rate": 5.386790439154854e-06,
+      "loss": 0.7551,
+      "step": 6121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4292380946156966,
+      "learning_rate": 5.3854950669409825e-06,
+      "loss": 0.6948,
+      "step": 6122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.490584988066094,
+      "learning_rate": 5.384199668697602e-06,
+      "loss": 0.7984,
+      "step": 6123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.496409333953262,
+      "learning_rate": 5.3829042445121825e-06,
+      "loss": 0.8403,
+      "step": 6124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.531335377402628,
+      "learning_rate": 5.381608794472192e-06,
+      "loss": 0.827,
+      "step": 6125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4531957694380173,
+      "learning_rate": 5.380313318665103e-06,
+      "loss": 0.7984,
+      "step": 6126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4278714069304286,
+      "learning_rate": 5.379017817178389e-06,
+      "loss": 0.7817,
+      "step": 6127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5572813012510762,
+      "learning_rate": 5.377722290099526e-06,
+      "loss": 0.7135,
+      "step": 6128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5391938288480635,
+      "learning_rate": 5.37642673751599e-06,
+      "loss": 0.766,
+      "step": 6129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5430414446150635,
+      "learning_rate": 5.37513115951526e-06,
+      "loss": 0.7054,
+      "step": 6130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6244777525499756,
+      "learning_rate": 5.373835556184817e-06,
+      "loss": 0.8931,
+      "step": 6131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5196654266250602,
+      "learning_rate": 5.37253992761214e-06,
+      "loss": 0.7817,
+      "step": 6132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6153873714957685,
+      "learning_rate": 5.371244273884718e-06,
+      "loss": 0.7855,
+      "step": 6133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0082471817337435,
+      "learning_rate": 5.369948595090033e-06,
+      "loss": 1.0878,
+      "step": 6134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5451451206590685,
+      "learning_rate": 5.368652891315571e-06,
+      "loss": 0.7485,
+      "step": 6135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8741249094930904,
+      "learning_rate": 5.3673571626488254e-06,
+      "loss": 1.0283,
+      "step": 6136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.378494873608061,
+      "learning_rate": 5.3660614091772826e-06,
+      "loss": 0.7324,
+      "step": 6137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6350155457402777,
+      "learning_rate": 5.3647656309884365e-06,
+      "loss": 0.7822,
+      "step": 6138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4566786766397117,
+      "learning_rate": 5.363469828169782e-06,
+      "loss": 0.7019,
+      "step": 6139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4413445666294156,
+      "learning_rate": 5.362174000808813e-06,
+      "loss": 0.8039,
+      "step": 6140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.466224689543455,
+      "learning_rate": 5.360878148993027e-06,
+      "loss": 0.7871,
+      "step": 6141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0772611602818163,
+      "learning_rate": 5.359582272809922e-06,
+      "loss": 1.0855,
+      "step": 6142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.529564707849868,
+      "learning_rate": 5.358286372347002e-06,
+      "loss": 0.7655,
+      "step": 6143
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.541330666300403,
+      "learning_rate": 5.356990447691765e-06,
+      "loss": 0.7838,
+      "step": 6144
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5903178148146007,
+      "learning_rate": 5.355694498931718e-06,
+      "loss": 0.8331,
+      "step": 6145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4850842033569462,
+      "learning_rate": 5.354398526154365e-06,
+      "loss": 0.7853,
+      "step": 6146
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5073178084707142,
+      "learning_rate": 5.353102529447213e-06,
+      "loss": 0.7854,
+      "step": 6147
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5816679685000012,
+      "learning_rate": 5.351806508897771e-06,
+      "loss": 0.76,
+      "step": 6148
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.418381058901958,
+      "learning_rate": 5.350510464593548e-06,
+      "loss": 0.7104,
+      "step": 6149
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6113929285438835,
+      "learning_rate": 5.349214396622058e-06,
+      "loss": 0.84,
+      "step": 6150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.583936523510558,
+      "learning_rate": 5.347918305070813e-06,
+      "loss": 0.7816,
+      "step": 6151
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5221054852777243,
+      "learning_rate": 5.346622190027329e-06,
+      "loss": 0.7551,
+      "step": 6152
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7980098339808834,
+      "learning_rate": 5.3453260515791216e-06,
+      "loss": 1.0953,
+      "step": 6153
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.554246521776231,
+      "learning_rate": 5.3440298898137084e-06,
+      "loss": 0.8085,
+      "step": 6154
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4549040998466947,
+      "learning_rate": 5.3427337048186124e-06,
+      "loss": 0.7355,
+      "step": 6155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4698179408359375,
+      "learning_rate": 5.341437496681352e-06,
+      "loss": 0.7503,
+      "step": 6156
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5312447022158004,
+      "learning_rate": 5.340141265489451e-06,
+      "loss": 0.7929,
+      "step": 6157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5381990604151663,
+      "learning_rate": 5.338845011330435e-06,
+      "loss": 0.8005,
+      "step": 6158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5433112818826693,
+      "learning_rate": 5.337548734291827e-06,
+      "loss": 0.7997,
+      "step": 6159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.810840244039006,
+      "learning_rate": 5.336252434461158e-06,
+      "loss": 1.0592,
+      "step": 6160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.592939409200986,
+      "learning_rate": 5.3349561119259555e-06,
+      "loss": 0.8555,
+      "step": 6161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.541202972602141,
+      "learning_rate": 5.33365976677375e-06,
+      "loss": 0.8041,
+      "step": 6162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7996924278983366,
+      "learning_rate": 5.332363399092076e-06,
+      "loss": 1.1015,
+      "step": 6163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5142760393107522,
+      "learning_rate": 5.331067008968462e-06,
+      "loss": 0.8158,
+      "step": 6164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.481805833712582,
+      "learning_rate": 5.329770596490449e-06,
+      "loss": 0.7794,
+      "step": 6165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4994044699031979,
+      "learning_rate": 5.328474161745571e-06,
+      "loss": 0.6773,
+      "step": 6166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5419100892792639,
+      "learning_rate": 5.327177704821366e-06,
+      "loss": 0.7793,
+      "step": 6167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.533158550656812,
+      "learning_rate": 5.325881225805373e-06,
+      "loss": 0.8013,
+      "step": 6168
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5295844200043938,
+      "learning_rate": 5.324584724785137e-06,
+      "loss": 0.7692,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4522664998350614,
+      "learning_rate": 5.323288201848197e-06,
+      "loss": 0.7406,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5884272705708322,
+      "learning_rate": 5.3219916570820976e-06,
+      "loss": 0.8644,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.410326554784251,
+      "learning_rate": 5.320695090574386e-06,
+      "loss": 0.6637,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5747870133251298,
+      "learning_rate": 5.319398502412609e-06,
+      "loss": 0.7521,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5645256145372708,
+      "learning_rate": 5.318101892684315e-06,
+      "loss": 0.822,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4569519390039,
+      "learning_rate": 5.316805261477052e-06,
+      "loss": 0.7591,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5783260109934827,
+      "learning_rate": 5.315508608878375e-06,
+      "loss": 0.708,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5682445085522791,
+      "learning_rate": 5.314211934975835e-06,
+      "loss": 0.7619,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5671645778327399,
+      "learning_rate": 5.312915239856986e-06,
+      "loss": 0.7714,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8714206214517676,
+      "learning_rate": 5.311618523609386e-06,
+      "loss": 1.0774,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3927122495351656,
+      "learning_rate": 5.310321786320588e-06,
+      "loss": 0.7807,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5985220415854617,
+      "learning_rate": 5.309025028078155e-06,
+      "loss": 0.8846,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.46899464382845,
+      "learning_rate": 5.307728248969646e-06,
+      "loss": 0.7032,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5128566782425388,
+      "learning_rate": 5.306431449082621e-06,
+      "loss": 0.7722,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.400965973592896,
+      "learning_rate": 5.3051346285046445e-06,
+      "loss": 0.7942,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4547480601478833,
+      "learning_rate": 5.30383778732328e-06,
+      "loss": 0.7191,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5908646063455092,
+      "learning_rate": 5.302540925626094e-06,
+      "loss": 0.8474,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7489266843177886,
+      "learning_rate": 5.301244043500651e-06,
+      "loss": 0.7549,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5658987608567778,
+      "learning_rate": 5.299947141034521e-06,
+      "loss": 0.8284,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4484850789891373,
+      "learning_rate": 5.298650218315277e-06,
+      "loss": 0.7353,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6286243928904662,
+      "learning_rate": 5.297353275430487e-06,
+      "loss": 0.8623,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5104690239109555,
+      "learning_rate": 5.296056312467723e-06,
+      "loss": 0.7441,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9034051548329042,
+      "learning_rate": 5.29475932951456e-06,
+      "loss": 1.0686,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8961401342661431,
+      "learning_rate": 5.293462326658572e-06,
+      "loss": 1.0518,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4309438637927183,
+      "learning_rate": 5.292165303987336e-06,
+      "loss": 0.6935,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.512715868398929,
+      "learning_rate": 5.290868261588433e-06,
+      "loss": 0.6855,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5007440179323912,
+      "learning_rate": 5.28957119954944e-06,
+      "loss": 0.8193,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4424621682927343,
+      "learning_rate": 5.288274117957936e-06,
+      "loss": 0.8322,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.561701661660001,
+      "learning_rate": 5.286977016901503e-06,
+      "loss": 0.6842,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6708980385910963,
+      "learning_rate": 5.285679896467729e-06,
+      "loss": 0.7983,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4369534492140668,
+      "learning_rate": 5.284382756744194e-06,
+      "loss": 0.7119,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6830686851561298,
+      "learning_rate": 5.283085597818485e-06,
+      "loss": 0.8005,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.484900894403216,
+      "learning_rate": 5.281788419778187e-06,
+      "loss": 0.8058,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5114127391963805,
+      "learning_rate": 5.280491222710893e-06,
+      "loss": 0.7907,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5084261404031585,
+      "learning_rate": 5.279194006704189e-06,
+      "loss": 0.7445,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5003009157600762,
+      "learning_rate": 5.277896771845668e-06,
+      "loss": 0.7848,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5616288779906593,
+      "learning_rate": 5.27659951822292e-06,
+      "loss": 0.7609,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4854748643269604,
+      "learning_rate": 5.275302245923543e-06,
+      "loss": 0.7471,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4991651236280947,
+      "learning_rate": 5.2740049550351266e-06,
+      "loss": 0.756,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9779216431791927,
+      "learning_rate": 5.27270764564527e-06,
+      "loss": 1.1001,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4724426143611455,
+      "learning_rate": 5.271410317841568e-06,
+      "loss": 0.6746,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5257229090314113,
+      "learning_rate": 5.2701129717116215e-06,
+      "loss": 0.8869,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4874201051517941,
+      "learning_rate": 5.26881560734303e-06,
+      "loss": 0.8118,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4992883257663459,
+      "learning_rate": 5.267518224823395e-06,
+      "loss": 0.7743,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5045865352606573,
+      "learning_rate": 5.266220824240316e-06,
+      "loss": 0.7756,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6468691718427197,
+      "learning_rate": 5.264923405681399e-06,
+      "loss": 0.791,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3672834532321414,
+      "learning_rate": 5.263625969234247e-06,
+      "loss": 0.7074,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5252195707331477,
+      "learning_rate": 5.262328514986468e-06,
+      "loss": 0.8207,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.453961311876116,
+      "learning_rate": 5.261031043025669e-06,
+      "loss": 0.6847,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4810590671863757,
+      "learning_rate": 5.259733553439453e-06,
+      "loss": 0.8028,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5803318118291212,
+      "learning_rate": 5.258436046315437e-06,
+      "loss": 0.7664,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5123040499463003,
+      "learning_rate": 5.257138521741226e-06,
+      "loss": 0.7067,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.729258935668166,
+      "learning_rate": 5.255840979804436e-06,
+      "loss": 0.7802,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8243649301993219,
+      "learning_rate": 5.254543420592677e-06,
+      "loss": 0.7732,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4521535339958287,
+      "learning_rate": 5.253245844193564e-06,
+      "loss": 0.7393,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4852544705772817,
+      "learning_rate": 5.2519482506947135e-06,
+      "loss": 0.7899,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3879269507442464,
+      "learning_rate": 5.25065064018374e-06,
+      "loss": 0.8284,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5569475949901304,
+      "learning_rate": 5.2493530127482624e-06,
+      "loss": 0.7264,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3638706482737752,
+      "learning_rate": 5.248055368475899e-06,
+      "loss": 0.7721,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8463547639684167,
+      "learning_rate": 5.246757707454271e-06,
+      "loss": 1.0576,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4700921487954783,
+      "learning_rate": 5.245460029770998e-06,
+      "loss": 0.767,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5000611219789426,
+      "learning_rate": 5.244162335513701e-06,
+      "loss": 0.797,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.416677568241714,
+      "learning_rate": 5.242864624770007e-06,
+      "loss": 0.7819,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4689530775184194,
+      "learning_rate": 5.241566897627536e-06,
+      "loss": 0.8362,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3741261745692872,
+      "learning_rate": 5.240269154173917e-06,
+      "loss": 0.7145,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7735346046973679,
+      "learning_rate": 5.238971394496776e-06,
+      "loss": 1.0754,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3907519883119903,
+      "learning_rate": 5.237673618683737e-06,
+      "loss": 0.7581,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8092572074832497,
+      "learning_rate": 5.236375826822435e-06,
+      "loss": 1.0802,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.564227758128665,
+      "learning_rate": 5.235078019000495e-06,
+      "loss": 0.8006,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6283526178546928,
+      "learning_rate": 5.23378019530555e-06,
+      "loss": 0.75,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5230778898536241,
+      "learning_rate": 5.232482355825233e-06,
+      "loss": 0.7897,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8051200030726031,
+      "learning_rate": 5.231184500647173e-06,
+      "loss": 1.112,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4907451576605,
+      "learning_rate": 5.229886629859009e-06,
+      "loss": 0.7798,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4477576365288327,
+      "learning_rate": 5.228588743548373e-06,
+      "loss": 0.77,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.42725829228119,
+      "learning_rate": 5.227290841802903e-06,
+      "loss": 0.8002,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4605560020704147,
+      "learning_rate": 5.225992924710236e-06,
+      "loss": 0.6835,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3778136800964071,
+      "learning_rate": 5.224694992358009e-06,
+      "loss": 0.7018,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.446889355460579,
+      "learning_rate": 5.223397044833863e-06,
+      "loss": 0.6648,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.436088098732948,
+      "learning_rate": 5.222099082225437e-06,
+      "loss": 0.7799,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4704426557874466,
+      "learning_rate": 5.2208011046203735e-06,
+      "loss": 0.6749,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4802288567188224,
+      "learning_rate": 5.2195031121063145e-06,
+      "loss": 0.7463,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4050833510604857,
+      "learning_rate": 5.2182051047709035e-06,
+      "loss": 0.7849,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8451818055126844,
+      "learning_rate": 5.2169070827017855e-06,
+      "loss": 1.0693,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5830983878856357,
+      "learning_rate": 5.215609045986604e-06,
+      "loss": 0.7405,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.547912743420679,
+      "learning_rate": 5.214310994713008e-06,
+      "loss": 0.8108,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8070808826537955,
+      "learning_rate": 5.213012928968642e-06,
+      "loss": 1.0768,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8150422292930273,
+      "learning_rate": 5.211714848841157e-06,
+      "loss": 1.0579,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5243946443021381,
+      "learning_rate": 5.210416754418202e-06,
+      "loss": 0.8226,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.498951266337785,
+      "learning_rate": 5.209118645787425e-06,
+      "loss": 0.7314,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.543515035118257,
+      "learning_rate": 5.2078205230364795e-06,
+      "loss": 0.7619,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8208229037548771,
+      "learning_rate": 5.206522386253017e-06,
+      "loss": 1.0976,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.532222719468825,
+      "learning_rate": 5.205224235524692e-06,
+      "loss": 0.7399,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3776954098364917,
+      "learning_rate": 5.203926070939156e-06,
+      "loss": 0.7549,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.610515748635863,
+      "learning_rate": 5.2026278925840656e-06,
+      "loss": 0.8061,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3471024878310238,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.7441,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4872570384346513,
+      "learning_rate": 5.2000314949158445e-06,
+      "loss": 0.7436,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4658078972471462,
+      "learning_rate": 5.198733275778031e-06,
+      "loss": 0.7997,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5012307342720812,
+      "learning_rate": 5.197435043221291e-06,
+      "loss": 0.7275,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.539153282332061,
+      "learning_rate": 5.196136797333285e-06,
+      "loss": 0.733,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6338612016312521,
+      "learning_rate": 5.194838538201676e-06,
+      "loss": 0.776,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.498938571411505,
+      "learning_rate": 5.193540265914121e-06,
+      "loss": 0.6951,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4989844858971688,
+      "learning_rate": 5.192241980558286e-06,
+      "loss": 0.7822,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6202017559003221,
+      "learning_rate": 5.1909436822218316e-06,
+      "loss": 0.7284,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5909837443241308,
+      "learning_rate": 5.189645370992426e-06,
+      "loss": 0.7786,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5732940045044745,
+      "learning_rate": 5.188347046957728e-06,
+      "loss": 0.7136,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5030399260076257,
+      "learning_rate": 5.187048710205407e-06,
+      "loss": 0.7658,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5076804418392957,
+      "learning_rate": 5.18575036082313e-06,
+      "loss": 0.7279,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5980246452668478,
+      "learning_rate": 5.184451998898565e-06,
+      "loss": 0.8237,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8480911893618831,
+      "learning_rate": 5.1831536245193795e-06,
+      "loss": 1.0797,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.597148773607799,
+      "learning_rate": 5.181855237773242e-06,
+      "loss": 0.86,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.634317052136927,
+      "learning_rate": 5.180556838747821e-06,
+      "loss": 0.7782,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5478313643937323,
+      "learning_rate": 5.179258427530791e-06,
+      "loss": 0.7518,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.341627850169706,
+      "learning_rate": 5.177960004209822e-06,
+      "loss": 0.7667,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.466622320972047,
+      "learning_rate": 5.1766615688725865e-06,
+      "loss": 0.8626,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4233475731692977,
+      "learning_rate": 5.175363121606759e-06,
+      "loss": 0.7703,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5357054352525195,
+      "learning_rate": 5.174064662500011e-06,
+      "loss": 0.7055,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5791718042416973,
+      "learning_rate": 5.1727661916400195e-06,
+      "loss": 0.7831,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4349237759769478,
+      "learning_rate": 5.171467709114458e-06,
+      "loss": 0.7848,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5755525187639967,
+      "learning_rate": 5.170169215011007e-06,
+      "loss": 0.8502,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5565650422455792,
+      "learning_rate": 5.168870709417342e-06,
+      "loss": 0.8066,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5082161257822573,
+      "learning_rate": 5.16757219242114e-06,
+      "loss": 0.6827,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4002100339848378,
+      "learning_rate": 5.166273664110079e-06,
+      "loss": 0.7315,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3792027610106983,
+      "learning_rate": 5.16497512457184e-06,
+      "loss": 0.7131,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8479786058355041,
+      "learning_rate": 5.163676573894104e-06,
+      "loss": 1.1027,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4984035639360112,
+      "learning_rate": 5.162378012164552e-06,
+      "loss": 0.7994,
+      "step": 6294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6558855934789058,
+      "learning_rate": 5.1610794394708665e-06,
+      "loss": 0.7824,
+      "step": 6295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6890517177535387,
+      "learning_rate": 5.159780855900725e-06,
+      "loss": 0.79,
+      "step": 6296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5318677375742684,
+      "learning_rate": 5.158482261541817e-06,
+      "loss": 0.8145,
+      "step": 6297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.439137895224054,
+      "learning_rate": 5.157183656481826e-06,
+      "loss": 0.6736,
+      "step": 6298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3980259286074601,
+      "learning_rate": 5.155885040808432e-06,
+      "loss": 0.7377,
+      "step": 6299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8093400694190801,
+      "learning_rate": 5.154586414609326e-06,
+      "loss": 1.0899,
+      "step": 6300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4856023835425332,
+      "learning_rate": 5.153287777972192e-06,
+      "loss": 0.6975,
+      "step": 6301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6179470723165938,
+      "learning_rate": 5.151989130984715e-06,
+      "loss": 0.8213,
+      "step": 6302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.47335456747346,
+      "learning_rate": 5.150690473734584e-06,
+      "loss": 0.8258,
+      "step": 6303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6263330133461826,
+      "learning_rate": 5.149391806309488e-06,
+      "loss": 0.7669,
+      "step": 6304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5019543844458962,
+      "learning_rate": 5.148093128797117e-06,
+      "loss": 0.8397,
+      "step": 6305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4809467184384197,
+      "learning_rate": 5.146794441285159e-06,
+      "loss": 0.6856,
+      "step": 6306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6253536015432246,
+      "learning_rate": 5.145495743861304e-06,
+      "loss": 0.8301,
+      "step": 6307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8038636453113016,
+      "learning_rate": 5.144197036613243e-06,
+      "loss": 1.0945,
+      "step": 6308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5330312498696057,
+      "learning_rate": 5.1428983196286686e-06,
+      "loss": 0.9082,
+      "step": 6309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4984920602458138,
+      "learning_rate": 5.141599592995274e-06,
+      "loss": 0.7644,
+      "step": 6310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5829593380869562,
+      "learning_rate": 5.1403008568007505e-06,
+      "loss": 0.7534,
+      "step": 6311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.504188723090126,
+      "learning_rate": 5.1390021111327936e-06,
+      "loss": 0.7858,
+      "step": 6312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4828443407424783,
+      "learning_rate": 5.137703356079095e-06,
+      "loss": 0.809,
+      "step": 6313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.590355347944815,
+      "learning_rate": 5.1364045917273505e-06,
+      "loss": 0.8366,
+      "step": 6314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5293534689011408,
+      "learning_rate": 5.135105818165256e-06,
+      "loss": 0.8159,
+      "step": 6315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7846283051512121,
+      "learning_rate": 5.133807035480508e-06,
+      "loss": 1.061,
+      "step": 6316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5373527013718988,
+      "learning_rate": 5.132508243760806e-06,
+      "loss": 0.7902,
+      "step": 6317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5338596721113498,
+      "learning_rate": 5.13120944309384e-06,
+      "loss": 0.7358,
+      "step": 6318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5616794078188856,
+      "learning_rate": 5.1299106335673144e-06,
+      "loss": 0.7798,
+      "step": 6319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5792581936431642,
+      "learning_rate": 5.128611815268925e-06,
+      "loss": 0.8413,
+      "step": 6320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4424436700636187,
+      "learning_rate": 5.127312988286372e-06,
+      "loss": 0.7856,
+      "step": 6321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7841208530391793,
+      "learning_rate": 5.126014152707355e-06,
+      "loss": 1.0787,
+      "step": 6322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6034464136476305,
+      "learning_rate": 5.124715308619574e-06,
+      "loss": 0.8095,
+      "step": 6323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.914177150048868,
+      "learning_rate": 5.123416456110731e-06,
+      "loss": 0.7562,
+      "step": 6324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5674434510431392,
+      "learning_rate": 5.122117595268526e-06,
+      "loss": 0.7148,
+      "step": 6325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.604859048898425,
+      "learning_rate": 5.120818726180662e-06,
+      "loss": 0.9276,
+      "step": 6326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5296910767538998,
+      "learning_rate": 5.1195198489348405e-06,
+      "loss": 0.6923,
+      "step": 6327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3836411107291229,
+      "learning_rate": 5.118220963618767e-06,
+      "loss": 0.7472,
+      "step": 6328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.552679806726167,
+      "learning_rate": 5.116922070320144e-06,
+      "loss": 0.7667,
+      "step": 6329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5622838632213487,
+      "learning_rate": 5.115623169126673e-06,
+      "loss": 0.8165,
+      "step": 6330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8507202984399388,
+      "learning_rate": 5.114324260126064e-06,
+      "loss": 1.0845,
+      "step": 6331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.468214840008952,
+      "learning_rate": 5.113025343406017e-06,
+      "loss": 0.7392,
+      "step": 6332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6308985279254928,
+      "learning_rate": 5.111726419054242e-06,
+      "loss": 0.8039,
+      "step": 6333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4530759579821584,
+      "learning_rate": 5.110427487158444e-06,
+      "loss": 0.7428,
+      "step": 6334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4202822869210754,
+      "learning_rate": 5.109128547806328e-06,
+      "loss": 0.72,
+      "step": 6335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.559381639892466,
+      "learning_rate": 5.107829601085604e-06,
+      "loss": 0.8204,
+      "step": 6336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.625347603541948,
+      "learning_rate": 5.106530647083978e-06,
+      "loss": 0.7856,
+      "step": 6337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.554905981476767,
+      "learning_rate": 5.10523168588916e-06,
+      "loss": 0.7889,
+      "step": 6338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5198566646427836,
+      "learning_rate": 5.1039327175888585e-06,
+      "loss": 0.8241,
+      "step": 6339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.385689419156612,
+      "learning_rate": 5.10263374227078e-06,
+      "loss": 0.7428,
+      "step": 6340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4938588711925693,
+      "learning_rate": 5.101334760022639e-06,
+      "loss": 0.7629,
+      "step": 6341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5898394665264874,
+      "learning_rate": 5.100035770932141e-06,
+      "loss": 0.7857,
+      "step": 6342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.429099673467969,
+      "learning_rate": 5.0987367750870005e-06,
+      "loss": 0.7813,
+      "step": 6343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4774877246599243,
+      "learning_rate": 5.097437772574927e-06,
+      "loss": 0.7703,
+      "step": 6344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5408308207461083,
+      "learning_rate": 5.0961387634836324e-06,
+      "loss": 0.8715,
+      "step": 6345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7682416502229155,
+      "learning_rate": 5.094839747900828e-06,
+      "loss": 0.7207,
+      "step": 6346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8858592722579279,
+      "learning_rate": 5.093540725914227e-06,
+      "loss": 1.0874,
+      "step": 6347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.531501842613457,
+      "learning_rate": 5.092241697611543e-06,
+      "loss": 0.7867,
+      "step": 6348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7234342824434714,
+      "learning_rate": 5.090942663080488e-06,
+      "loss": 0.7466,
+      "step": 6349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5729967492639005,
+      "learning_rate": 5.089643622408778e-06,
+      "loss": 0.7776,
+      "step": 6350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5015747903395165,
+      "learning_rate": 5.0883445756841244e-06,
+      "loss": 0.8023,
+      "step": 6351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4958402633714873,
+      "learning_rate": 5.087045522994242e-06,
+      "loss": 0.8147,
+      "step": 6352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5883149036764286,
+      "learning_rate": 5.085746464426848e-06,
+      "loss": 0.7238,
+      "step": 6353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8135767148113574,
+      "learning_rate": 5.084447400069656e-06,
+      "loss": 1.0873,
+      "step": 6354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8966465564032653,
+      "learning_rate": 5.083148330010383e-06,
+      "loss": 0.7108,
+      "step": 6355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6419829852724594,
+      "learning_rate": 5.081849254336745e-06,
+      "loss": 0.7841,
+      "step": 6356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8303448188712281,
+      "learning_rate": 5.080550173136457e-06,
+      "loss": 1.0727,
+      "step": 6357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5195709333018521,
+      "learning_rate": 5.0792510864972384e-06,
+      "loss": 0.801,
+      "step": 6358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5353032152237684,
+      "learning_rate": 5.077951994506805e-06,
+      "loss": 0.7642,
+      "step": 6359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7614957714211149,
+      "learning_rate": 5.076652897252874e-06,
+      "loss": 1.0531,
+      "step": 6360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5753872024648066,
+      "learning_rate": 5.075353794823165e-06,
+      "loss": 0.7606,
+      "step": 6361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5769985310783925,
+      "learning_rate": 5.074054687305394e-06,
+      "loss": 0.7463,
+      "step": 6362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4779834692586884,
+      "learning_rate": 5.072755574787282e-06,
+      "loss": 0.6905,
+      "step": 6363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6272880194156245,
+      "learning_rate": 5.071456457356547e-06,
+      "loss": 0.8267,
+      "step": 6364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3544468644296723,
+      "learning_rate": 5.0701573351009105e-06,
+      "loss": 0.7809,
+      "step": 6365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.527362411624617,
+      "learning_rate": 5.068858208108087e-06,
+      "loss": 0.7284,
+      "step": 6366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5937254363767703,
+      "learning_rate": 5.067559076465803e-06,
+      "loss": 0.7465,
+      "step": 6367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.534774681394336,
+      "learning_rate": 5.066259940261774e-06,
+      "loss": 0.742,
+      "step": 6368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5850870979047744,
+      "learning_rate": 5.064960799583722e-06,
+      "loss": 0.7882,
+      "step": 6369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5680986240909527,
+      "learning_rate": 5.06366165451937e-06,
+      "loss": 0.7611,
+      "step": 6370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8702628794300165,
+      "learning_rate": 5.062362505156435e-06,
+      "loss": 1.083,
+      "step": 6371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4658764015925108,
+      "learning_rate": 5.061063351582642e-06,
+      "loss": 0.7924,
+      "step": 6372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.429584741878266,
+      "learning_rate": 5.059764193885713e-06,
+      "loss": 0.8118,
+      "step": 6373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7622081421603959,
+      "learning_rate": 5.058465032153368e-06,
+      "loss": 1.0623,
+      "step": 6374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5040150115713196,
+      "learning_rate": 5.0571658664733314e-06,
+      "loss": 0.7748,
+      "step": 6375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4945317693496563,
+      "learning_rate": 5.055866696933324e-06,
+      "loss": 0.76,
+      "step": 6376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4716468271220315,
+      "learning_rate": 5.054567523621069e-06,
+      "loss": 0.7768,
+      "step": 6377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4944777603568808,
+      "learning_rate": 5.05326834662429e-06,
+      "loss": 0.7679,
+      "step": 6378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4585884389243409,
+      "learning_rate": 5.051969166030711e-06,
+      "loss": 0.7831,
+      "step": 6379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8436007164477644,
+      "learning_rate": 5.050669981928056e-06,
+      "loss": 1.0759,
+      "step": 6380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4459808975680575,
+      "learning_rate": 5.049370794404046e-06,
+      "loss": 0.6919,
+      "step": 6381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4843535144244686,
+      "learning_rate": 5.048071603546409e-06,
+      "loss": 0.7285,
+      "step": 6382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6233102113993472,
+      "learning_rate": 5.046772409442866e-06,
+      "loss": 0.731,
+      "step": 6383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5898849085363784,
+      "learning_rate": 5.045473212181145e-06,
+      "loss": 0.7084,
+      "step": 6384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5719279533844586,
+      "learning_rate": 5.044174011848966e-06,
+      "loss": 0.8415,
+      "step": 6385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5729072149969299,
+      "learning_rate": 5.0428748085340565e-06,
+      "loss": 0.7244,
+      "step": 6386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.590232042453039,
+      "learning_rate": 5.041575602324144e-06,
+      "loss": 0.8272,
+      "step": 6387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4691885332548085,
+      "learning_rate": 5.0402763933069496e-06,
+      "loss": 0.758,
+      "step": 6388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.558791481962644,
+      "learning_rate": 5.038977181570204e-06,
+      "loss": 0.7495,
+      "step": 6389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.648121755013111,
+      "learning_rate": 5.037677967201629e-06,
+      "loss": 0.741,
+      "step": 6390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5216222837170945,
+      "learning_rate": 5.036378750288949e-06,
+      "loss": 0.8026,
+      "step": 6391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8403329930315464,
+      "learning_rate": 5.035079530919895e-06,
+      "loss": 1.0758,
+      "step": 6392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5093342943380332,
+      "learning_rate": 5.0337803091821905e-06,
+      "loss": 0.7862,
+      "step": 6393
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5528835208864047,
+      "learning_rate": 5.032481085163562e-06,
+      "loss": 0.7089,
+      "step": 6394
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.460099150259266,
+      "learning_rate": 5.031181858951737e-06,
+      "loss": 0.7927,
+      "step": 6395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4695033236469097,
+      "learning_rate": 5.029882630634441e-06,
+      "loss": 0.7267,
+      "step": 6396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6073685610750217,
+      "learning_rate": 5.028583400299402e-06,
+      "loss": 0.7544,
+      "step": 6397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4886265473622153,
+      "learning_rate": 5.027284168034344e-06,
+      "loss": 0.8229,
+      "step": 6398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4648639091058069,
+      "learning_rate": 5.025984933927e-06,
+      "loss": 0.7515,
+      "step": 6399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5335129972263561,
+      "learning_rate": 5.024685698065093e-06,
+      "loss": 0.8724,
+      "step": 6400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4028259466929902,
+      "learning_rate": 5.02338646053635e-06,
+      "loss": 0.651,
+      "step": 6401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.653342939596997,
+      "learning_rate": 5.0220872214285e-06,
+      "loss": 0.7757,
+      "step": 6402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8498046992252434,
+      "learning_rate": 5.02078798082927e-06,
+      "loss": 1.0417,
+      "step": 6403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.433729468480321,
+      "learning_rate": 5.0194887388263895e-06,
+      "loss": 0.8026,
+      "step": 6404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4374156885205223,
+      "learning_rate": 5.018189495507584e-06,
+      "loss": 0.7691,
+      "step": 6405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5013101089067444,
+      "learning_rate": 5.016890250960582e-06,
+      "loss": 0.8012,
+      "step": 6406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9077921372843638,
+      "learning_rate": 5.0155910052731116e-06,
+      "loss": 0.6503,
+      "step": 6407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5132613150844731,
+      "learning_rate": 5.0142917585329e-06,
+      "loss": 0.7665,
+      "step": 6408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4194916087658804,
+      "learning_rate": 5.012992510827678e-06,
+      "loss": 0.7717,
+      "step": 6409
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.553717623276083,
+      "learning_rate": 5.01169326224517e-06,
+      "loss": 0.7442,
+      "step": 6410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4710493304345436,
+      "learning_rate": 5.010394012873107e-06,
+      "loss": 0.7425,
+      "step": 6411
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5840795676703907,
+      "learning_rate": 5.009094762799218e-06,
+      "loss": 0.81,
+      "step": 6412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4886871770917565,
+      "learning_rate": 5.0077955121112285e-06,
+      "loss": 0.776,
+      "step": 6413
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4991076060960011,
+      "learning_rate": 5.006496260896868e-06,
+      "loss": 0.791,
+      "step": 6414
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8336997986581897,
+      "learning_rate": 5.0051970092438655e-06,
+      "loss": 1.0739,
+      "step": 6415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5294706846268395,
+      "learning_rate": 5.003897757239949e-06,
+      "loss": 0.6606,
+      "step": 6416
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5886204267215926,
+      "learning_rate": 5.002598504972848e-06,
+      "loss": 0.7312,
+      "step": 6417
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.543066086084952,
+      "learning_rate": 5.0012992525302885e-06,
+      "loss": 0.7258,
+      "step": 6418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6073384626028389,
+      "learning_rate": 5e-06,
+      "loss": 0.8081,
+      "step": 6419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.545236148980762,
+      "learning_rate": 4.998700747469713e-06,
+      "loss": 0.8214,
+      "step": 6420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7892700117370993,
+      "learning_rate": 4.997401495027154e-06,
+      "loss": 1.0722,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5323551721387514,
+      "learning_rate": 4.996102242760053e-06,
+      "loss": 0.7485,
+      "step": 6422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4635867370863458,
+      "learning_rate": 4.994802990756136e-06,
+      "loss": 0.7849,
+      "step": 6423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8497808674492808,
+      "learning_rate": 4.9935037391031346e-06,
+      "loss": 0.7867,
+      "step": 6424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.628240885918961,
+      "learning_rate": 4.992204487888772e-06,
+      "loss": 0.8272,
+      "step": 6425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4648870900713535,
+      "learning_rate": 4.9909052372007834e-06,
+      "loss": 0.8433,
+      "step": 6426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.753603882304953,
+      "learning_rate": 4.9896059871268934e-06,
+      "loss": 0.7618,
+      "step": 6427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4434484660624407,
+      "learning_rate": 4.98830673775483e-06,
+      "loss": 0.6977,
+      "step": 6428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5401244116421686,
+      "learning_rate": 4.987007489172323e-06,
+      "loss": 0.8304,
+      "step": 6429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.39287256694849,
+      "learning_rate": 4.9857082414671015e-06,
+      "loss": 0.7178,
+      "step": 6430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8505393133893913,
+      "learning_rate": 4.984408994726889e-06,
+      "loss": 1.084,
+      "step": 6431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3981005307893708,
+      "learning_rate": 4.9831097490394195e-06,
+      "loss": 0.7084,
+      "step": 6432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8209200825415741,
+      "learning_rate": 4.981810504492418e-06,
+      "loss": 1.1023,
+      "step": 6433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5500416468655098,
+      "learning_rate": 4.980511261173613e-06,
+      "loss": 0.7412,
+      "step": 6434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4428367219170273,
+      "learning_rate": 4.979212019170731e-06,
+      "loss": 0.829,
+      "step": 6435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5994690143916275,
+      "learning_rate": 4.977912778571501e-06,
+      "loss": 0.7845,
+      "step": 6436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4698746203639657,
+      "learning_rate": 4.976613539463652e-06,
+      "loss": 0.7885,
+      "step": 6437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4901436956164573,
+      "learning_rate": 4.975314301934909e-06,
+      "loss": 0.827,
+      "step": 6438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.513267267325777,
+      "learning_rate": 4.974015066073002e-06,
+      "loss": 0.7124,
+      "step": 6439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8326463811820981,
+      "learning_rate": 4.972715831965657e-06,
+      "loss": 1.1035,
+      "step": 6440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4966616094063074,
+      "learning_rate": 4.971416599700601e-06,
+      "loss": 0.8403,
+      "step": 6441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.600038443020443,
+      "learning_rate": 4.97011736936556e-06,
+      "loss": 0.8186,
+      "step": 6442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8004497456864529,
+      "learning_rate": 4.968818141048264e-06,
+      "loss": 1.0671,
+      "step": 6443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4736674124986657,
+      "learning_rate": 4.967518914836439e-06,
+      "loss": 0.7743,
+      "step": 6444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6120093520659209,
+      "learning_rate": 4.96621969081781e-06,
+      "loss": 0.8447,
+      "step": 6445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5142589748478543,
+      "learning_rate": 4.964920469080107e-06,
+      "loss": 0.8229,
+      "step": 6446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7585248264790571,
+      "learning_rate": 4.963621249711052e-06,
+      "loss": 1.0734,
+      "step": 6447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.536379949646187,
+      "learning_rate": 4.9623220327983745e-06,
+      "loss": 0.7324,
+      "step": 6448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4701693102670197,
+      "learning_rate": 4.961022818429798e-06,
+      "loss": 0.7885,
+      "step": 6449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7583910439565497,
+      "learning_rate": 4.959723606693051e-06,
+      "loss": 1.0724,
+      "step": 6450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.112456729763788,
+      "learning_rate": 4.958424397675859e-06,
+      "loss": 0.7197,
+      "step": 6451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.480326143205103,
+      "learning_rate": 4.9571251914659435e-06,
+      "loss": 0.8343,
+      "step": 6452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5408453259952195,
+      "learning_rate": 4.955825988151036e-06,
+      "loss": 0.8417,
+      "step": 6453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5958346627945268,
+      "learning_rate": 4.9545267878188585e-06,
+      "loss": 0.7745,
+      "step": 6454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6089433669065925,
+      "learning_rate": 4.953227590557136e-06,
+      "loss": 0.7606,
+      "step": 6455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.654648816503151,
+      "learning_rate": 4.951928396453593e-06,
+      "loss": 0.8564,
+      "step": 6456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7326092986383894,
+      "learning_rate": 4.950629205595955e-06,
+      "loss": 0.7881,
+      "step": 6457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.43304510259923,
+      "learning_rate": 4.949330018071947e-06,
+      "loss": 0.781,
+      "step": 6458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4726382085397431,
+      "learning_rate": 4.948030833969289e-06,
+      "loss": 0.6617,
+      "step": 6459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8473671837899007,
+      "learning_rate": 4.946731653375711e-06,
+      "loss": 1.0935,
+      "step": 6460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4080591134682687,
+      "learning_rate": 4.945432476378933e-06,
+      "loss": 0.7394,
+      "step": 6461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8013557886990337,
+      "learning_rate": 4.944133303066677e-06,
+      "loss": 1.0512,
+      "step": 6462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5449880110854264,
+      "learning_rate": 4.94283413352667e-06,
+      "loss": 0.7629,
+      "step": 6463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4889939833507277,
+      "learning_rate": 4.9415349678466335e-06,
+      "loss": 0.8013,
+      "step": 6464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5810427920849335,
+      "learning_rate": 4.940235806114289e-06,
+      "loss": 0.8321,
+      "step": 6465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5402087370488495,
+      "learning_rate": 4.938936648417359e-06,
+      "loss": 0.8168,
+      "step": 6466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.528993746268958,
+      "learning_rate": 4.937637494843566e-06,
+      "loss": 0.8974,
+      "step": 6467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6206708296205121,
+      "learning_rate": 4.936338345480633e-06,
+      "loss": 0.7658,
+      "step": 6468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4459501312953147,
+      "learning_rate": 4.935039200416279e-06,
+      "loss": 0.7242,
+      "step": 6469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9083455540425515,
+      "learning_rate": 4.933740059738227e-06,
+      "loss": 1.1002,
+      "step": 6470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5135210114614037,
+      "learning_rate": 4.932440923534199e-06,
+      "loss": 0.7984,
+      "step": 6471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3949659506944094,
+      "learning_rate": 4.931141791891913e-06,
+      "loss": 0.7469,
+      "step": 6472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4223973909458398,
+      "learning_rate": 4.929842664899092e-06,
+      "loss": 0.7895,
+      "step": 6473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4874815384217368,
+      "learning_rate": 4.928543542643454e-06,
+      "loss": 0.7694,
+      "step": 6474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5752708452894981,
+      "learning_rate": 4.92724442521272e-06,
+      "loss": 0.7733,
+      "step": 6475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.560095584104035,
+      "learning_rate": 4.925945312694606e-06,
+      "loss": 0.7367,
+      "step": 6476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5517567295584944,
+      "learning_rate": 4.924646205176836e-06,
+      "loss": 0.6818,
+      "step": 6477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.581536756637996,
+      "learning_rate": 4.923347102747129e-06,
+      "loss": 0.7659,
+      "step": 6478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4952480089883156,
+      "learning_rate": 4.922048005493196e-06,
+      "loss": 0.8041,
+      "step": 6479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7937958434535268,
+      "learning_rate": 4.920748913502763e-06,
+      "loss": 1.0632,
+      "step": 6480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5143591022040594,
+      "learning_rate": 4.919449826863544e-06,
+      "loss": 0.7789,
+      "step": 6481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7650382053721767,
+      "learning_rate": 4.9181507456632574e-06,
+      "loss": 1.0765,
+      "step": 6482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4838996972233147,
+      "learning_rate": 4.9168516699896185e-06,
+      "loss": 0.7727,
+      "step": 6483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4814017018105492,
+      "learning_rate": 4.915552599930345e-06,
+      "loss": 0.8288,
+      "step": 6484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4958479633315331,
+      "learning_rate": 4.914253535573154e-06,
+      "loss": 0.7299,
+      "step": 6485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5254318313121316,
+      "learning_rate": 4.912954477005758e-06,
+      "loss": 0.8068,
+      "step": 6486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5583065834647298,
+      "learning_rate": 4.911655424315877e-06,
+      "loss": 0.8065,
+      "step": 6487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5029390966784038,
+      "learning_rate": 4.910356377591224e-06,
+      "loss": 0.7343,
+      "step": 6488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5030473924130376,
+      "learning_rate": 4.909057336919513e-06,
+      "loss": 0.8166,
+      "step": 6489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7721701774058407,
+      "learning_rate": 4.907758302388458e-06,
+      "loss": 0.7579,
+      "step": 6490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5892450038549366,
+      "learning_rate": 4.906459274085774e-06,
+      "loss": 0.8303,
+      "step": 6491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.457407435370129,
+      "learning_rate": 4.905160252099174e-06,
+      "loss": 0.7807,
+      "step": 6492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5168357037440405,
+      "learning_rate": 4.903861236516369e-06,
+      "loss": 0.7567,
+      "step": 6493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4856418847093935,
+      "learning_rate": 4.902562227425075e-06,
+      "loss": 0.7529,
+      "step": 6494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7761629502836265,
+      "learning_rate": 4.901263224913001e-06,
+      "loss": 0.6403,
+      "step": 6495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5850458293113319,
+      "learning_rate": 4.899964229067859e-06,
+      "loss": 0.7808,
+      "step": 6496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.510097095601767,
+      "learning_rate": 4.8986652399773625e-06,
+      "loss": 0.7779,
+      "step": 6497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6499328812896474,
+      "learning_rate": 4.897366257729221e-06,
+      "loss": 0.8087,
+      "step": 6498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5950171976264655,
+      "learning_rate": 4.896067282411144e-06,
+      "loss": 0.7711,
+      "step": 6499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.541596010190185,
+      "learning_rate": 4.894768314110841e-06,
+      "loss": 0.7226,
+      "step": 6500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4682725118606523,
+      "learning_rate": 4.893469352916023e-06,
+      "loss": 0.7334,
+      "step": 6501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5529400185493567,
+      "learning_rate": 4.892170398914398e-06,
+      "loss": 0.8329,
+      "step": 6502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.683749494991563,
+      "learning_rate": 4.890871452193673e-06,
+      "loss": 0.7923,
+      "step": 6503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.513640823612606,
+      "learning_rate": 4.889572512841557e-06,
+      "loss": 0.799,
+      "step": 6504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5157349898236623,
+      "learning_rate": 4.8882735809457594e-06,
+      "loss": 0.7884,
+      "step": 6505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6013235478292636,
+      "learning_rate": 4.886974656593986e-06,
+      "loss": 0.7134,
+      "step": 6506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3466769949823978,
+      "learning_rate": 4.885675739873938e-06,
+      "loss": 0.6386,
+      "step": 6507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6083519080816646,
+      "learning_rate": 4.8843768308733285e-06,
+      "loss": 0.7664,
+      "step": 6508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.855208814150592,
+      "learning_rate": 4.883077929679859e-06,
+      "loss": 1.109,
+      "step": 6509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.587428529222084,
+      "learning_rate": 4.881779036381234e-06,
+      "loss": 0.7355,
+      "step": 6510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.786876803078171,
+      "learning_rate": 4.88048015106516e-06,
+      "loss": 1.0711,
+      "step": 6511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4901684854756414,
+      "learning_rate": 4.87918127381934e-06,
+      "loss": 0.7869,
+      "step": 6512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5176874694101439,
+      "learning_rate": 4.877882404731474e-06,
+      "loss": 0.7193,
+      "step": 6513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5817594973821638,
+      "learning_rate": 4.87658354388927e-06,
+      "loss": 0.7906,
+      "step": 6514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4679859826111616,
+      "learning_rate": 4.875284691380427e-06,
+      "loss": 0.7983,
+      "step": 6515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.51493781893582,
+      "learning_rate": 4.873985847292647e-06,
+      "loss": 0.6716,
+      "step": 6516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6487592101352864,
+      "learning_rate": 4.872687011713629e-06,
+      "loss": 0.764,
+      "step": 6517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7723148776336929,
+      "learning_rate": 4.871388184731077e-06,
+      "loss": 1.1063,
+      "step": 6518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8898810051464694,
+      "learning_rate": 4.870089366432688e-06,
+      "loss": 0.9017,
+      "step": 6519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4574292176576356,
+      "learning_rate": 4.868790556906161e-06,
+      "loss": 0.6978,
+      "step": 6520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4497915377880644,
+      "learning_rate": 4.867491756239197e-06,
+      "loss": 0.8322,
+      "step": 6521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5643138437015354,
+      "learning_rate": 4.866192964519493e-06,
+      "loss": 0.7335,
+      "step": 6522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.459124303262117,
+      "learning_rate": 4.8648941818347465e-06,
+      "loss": 0.7189,
+      "step": 6523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5768394452809897,
+      "learning_rate": 4.86359540827265e-06,
+      "loss": 0.7792,
+      "step": 6524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6226952554231497,
+      "learning_rate": 4.862296643920907e-06,
+      "loss": 0.8023,
+      "step": 6525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5795545153804817,
+      "learning_rate": 4.860997888867209e-06,
+      "loss": 0.8735,
+      "step": 6526
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4891096834423039,
+      "learning_rate": 4.85969914319925e-06,
+      "loss": 0.8081,
+      "step": 6527
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.461571651889349,
+      "learning_rate": 4.8584004070047275e-06,
+      "loss": 0.8013,
+      "step": 6528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5774315639224543,
+      "learning_rate": 4.857101680371333e-06,
+      "loss": 0.7547,
+      "step": 6529
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4853511122903726,
+      "learning_rate": 4.855802963386757e-06,
+      "loss": 0.8448,
+      "step": 6530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8819504396452725,
+      "learning_rate": 4.8545042561386975e-06,
+      "loss": 1.0534,
+      "step": 6531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8360991669671466,
+      "learning_rate": 4.853205558714843e-06,
+      "loss": 1.115,
+      "step": 6532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5964781297985573,
+      "learning_rate": 4.851906871202885e-06,
+      "loss": 0.8192,
+      "step": 6533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5270888157884512,
+      "learning_rate": 4.8506081936905124e-06,
+      "loss": 0.6651,
+      "step": 6534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.795856466773291,
+      "learning_rate": 4.849309526265417e-06,
+      "loss": 1.0788,
+      "step": 6535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4791727051628718,
+      "learning_rate": 4.848010869015288e-06,
+      "loss": 0.7942,
+      "step": 6536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6964711065071454,
+      "learning_rate": 4.846712222027811e-06,
+      "loss": 0.7051,
+      "step": 6537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8408982701451679,
+      "learning_rate": 4.845413585390676e-06,
+      "loss": 1.0865,
+      "step": 6538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5695823285103387,
+      "learning_rate": 4.844114959191569e-06,
+      "loss": 0.8251,
+      "step": 6539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8068832269713099,
+      "learning_rate": 4.842816343518178e-06,
+      "loss": 1.1215,
+      "step": 6540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6567438701114299,
+      "learning_rate": 4.841517738458183e-06,
+      "loss": 0.7997,
+      "step": 6541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.558436974948719,
+      "learning_rate": 4.8402191440992755e-06,
+      "loss": 0.8806,
+      "step": 6542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6005547111940115,
+      "learning_rate": 4.838920560529137e-06,
+      "loss": 0.7358,
+      "step": 6543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.447219648912299,
+      "learning_rate": 4.837621987835449e-06,
+      "loss": 0.6843,
+      "step": 6544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5275517450200786,
+      "learning_rate": 4.836323426105897e-06,
+      "loss": 0.7065,
+      "step": 6545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6046749707827714,
+      "learning_rate": 4.835024875428162e-06,
+      "loss": 0.7428,
+      "step": 6546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.450644595828523,
+      "learning_rate": 4.833726335889922e-06,
+      "loss": 0.7833,
+      "step": 6547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4265858344694498,
+      "learning_rate": 4.832427807578862e-06,
+      "loss": 0.6018,
+      "step": 6548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4650526391866825,
+      "learning_rate": 4.83112929058266e-06,
+      "loss": 0.67,
+      "step": 6549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4116807058159213,
+      "learning_rate": 4.829830784988995e-06,
+      "loss": 0.7114,
+      "step": 6550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5155028453601673,
+      "learning_rate": 4.828532290885541e-06,
+      "loss": 0.727,
+      "step": 6551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4730596879557711,
+      "learning_rate": 4.827233808359982e-06,
+      "loss": 0.7682,
+      "step": 6552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6789068550740676,
+      "learning_rate": 4.825935337499991e-06,
+      "loss": 0.7937,
+      "step": 6553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.524247026722769,
+      "learning_rate": 4.824636878393243e-06,
+      "loss": 0.758,
+      "step": 6554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6160458163196136,
+      "learning_rate": 4.823338431127414e-06,
+      "loss": 0.7635,
+      "step": 6555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9287572237546142,
+      "learning_rate": 4.82203999579018e-06,
+      "loss": 0.769,
+      "step": 6556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5451878340476455,
+      "learning_rate": 4.820741572469211e-06,
+      "loss": 0.7867,
+      "step": 6557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.634336163996332,
+      "learning_rate": 4.819443161252179e-06,
+      "loss": 0.7258,
+      "step": 6558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5855406592570065,
+      "learning_rate": 4.81814476222676e-06,
+      "loss": 0.7431,
+      "step": 6559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4849348566967928,
+      "learning_rate": 4.816846375480623e-06,
+      "loss": 0.8275,
+      "step": 6560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5604195180300142,
+      "learning_rate": 4.8155480011014354e-06,
+      "loss": 0.7768,
+      "step": 6561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3982437802804302,
+      "learning_rate": 4.81424963917687e-06,
+      "loss": 0.6944,
+      "step": 6562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4706612670790948,
+      "learning_rate": 4.812951289794594e-06,
+      "loss": 0.7543,
+      "step": 6563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.473060152420395,
+      "learning_rate": 4.8116529530422745e-06,
+      "loss": 0.7794,
+      "step": 6564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4739685231525774,
+      "learning_rate": 4.810354629007576e-06,
+      "loss": 0.7409,
+      "step": 6565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5522301016839524,
+      "learning_rate": 4.80905631777817e-06,
+      "loss": 0.8364,
+      "step": 6566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3985813543188996,
+      "learning_rate": 4.807758019441717e-06,
+      "loss": 0.8074,
+      "step": 6567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4350509556647661,
+      "learning_rate": 4.80645973408588e-06,
+      "loss": 0.7596,
+      "step": 6568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6133734009431189,
+      "learning_rate": 4.805161461798326e-06,
+      "loss": 0.6666,
+      "step": 6569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6890135242823225,
+      "learning_rate": 4.803863202666716e-06,
+      "loss": 0.7603,
+      "step": 6570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.423206687258015,
+      "learning_rate": 4.8025649567787095e-06,
+      "loss": 0.7557,
+      "step": 6571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5713056267825107,
+      "learning_rate": 4.80126672422197e-06,
+      "loss": 0.7547,
+      "step": 6572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9055934158401004,
+      "learning_rate": 4.799968505084156e-06,
+      "loss": 1.0475,
+      "step": 6573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.537480410269998,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.7665,
+      "step": 6574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5558199405836304,
+      "learning_rate": 4.797372107415935e-06,
+      "loss": 0.7477,
+      "step": 6575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.477980826324111,
+      "learning_rate": 4.796073929060845e-06,
+      "loss": 0.7734,
+      "step": 6576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8358758564906432,
+      "learning_rate": 4.79477576447531e-06,
+      "loss": 1.1205,
+      "step": 6577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4751800684559153,
+      "learning_rate": 4.793477613746984e-06,
+      "loss": 0.7461,
+      "step": 6578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7848411182580356,
+      "learning_rate": 4.792179476963521e-06,
+      "loss": 1.1015,
+      "step": 6579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4898781789431677,
+      "learning_rate": 4.7908813542125765e-06,
+      "loss": 0.7582,
+      "step": 6580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.549735385242299,
+      "learning_rate": 4.789583245581801e-06,
+      "loss": 0.8243,
+      "step": 6581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4603272206459428,
+      "learning_rate": 4.788285151158844e-06,
+      "loss": 0.7331,
+      "step": 6582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.22606750674253,
+      "learning_rate": 4.786987071031359e-06,
+      "loss": 0.716,
+      "step": 6583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5144539851123386,
+      "learning_rate": 4.785689005286995e-06,
+      "loss": 0.8057,
+      "step": 6584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4946502857559942,
+      "learning_rate": 4.784390954013396e-06,
+      "loss": 0.7978,
+      "step": 6585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5295029981199595,
+      "learning_rate": 4.783092917298216e-06,
+      "loss": 0.8071,
+      "step": 6586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4831591224238339,
+      "learning_rate": 4.781794895229097e-06,
+      "loss": 0.7285,
+      "step": 6587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.592945357557965,
+      "learning_rate": 4.780496887893686e-06,
+      "loss": 0.7339,
+      "step": 6588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.478260215958859,
+      "learning_rate": 4.779198895379627e-06,
+      "loss": 0.7624,
+      "step": 6589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5758860468213067,
+      "learning_rate": 4.7779009177745645e-06,
+      "loss": 0.809,
+      "step": 6590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5029767172651876,
+      "learning_rate": 4.7766029551661395e-06,
+      "loss": 0.8688,
+      "step": 6591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5017567519992474,
+      "learning_rate": 4.7753050076419916e-06,
+      "loss": 0.8281,
+      "step": 6592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5476474533874085,
+      "learning_rate": 4.774007075289766e-06,
+      "loss": 0.773,
+      "step": 6593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6972990946699873,
+      "learning_rate": 4.772709158197098e-06,
+      "loss": 0.7139,
+      "step": 6594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6436218683370412,
+      "learning_rate": 4.771411256451628e-06,
+      "loss": 0.7867,
+      "step": 6595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5401917620855783,
+      "learning_rate": 4.770113370140992e-06,
+      "loss": 0.7726,
+      "step": 6596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.54804670125832,
+      "learning_rate": 4.768815499352828e-06,
+      "loss": 0.8179,
+      "step": 6597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5303551405090337,
+      "learning_rate": 4.76751764417477e-06,
+      "loss": 0.8369,
+      "step": 6598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.46431807576495,
+      "learning_rate": 4.766219804694451e-06,
+      "loss": 0.8001,
+      "step": 6599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9709887101807445,
+      "learning_rate": 4.764921980999507e-06,
+      "loss": 1.0667,
+      "step": 6600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5151494384730637,
+      "learning_rate": 4.763624173177568e-06,
+      "loss": 0.7344,
+      "step": 6601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4684602563638292,
+      "learning_rate": 4.762326381316263e-06,
+      "loss": 0.6811,
+      "step": 6602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9315114926891812,
+      "learning_rate": 4.761028605503226e-06,
+      "loss": 0.7966,
+      "step": 6603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5611713414289543,
+      "learning_rate": 4.7597308458260845e-06,
+      "loss": 0.8578,
+      "step": 6604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.512347446742377,
+      "learning_rate": 4.758433102372466e-06,
+      "loss": 0.784,
+      "step": 6605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8842981610607556,
+      "learning_rate": 4.7571353752299955e-06,
+      "loss": 1.1221,
+      "step": 6606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8607453401366636,
+      "learning_rate": 4.7558376644863e-06,
+      "loss": 1.1085,
+      "step": 6607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6030697077930964,
+      "learning_rate": 4.754539970229005e-06,
+      "loss": 0.8012,
+      "step": 6608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5573501820373252,
+      "learning_rate": 4.75324229254573e-06,
+      "loss": 0.7428,
+      "step": 6609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5920673672094048,
+      "learning_rate": 4.7519446315241025e-06,
+      "loss": 0.7747,
+      "step": 6610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4065127768083898,
+      "learning_rate": 4.75064698725174e-06,
+      "loss": 0.8176,
+      "step": 6611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8256040648198439,
+      "learning_rate": 4.749349359816261e-06,
+      "loss": 1.11,
+      "step": 6612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5204605439929173,
+      "learning_rate": 4.748051749305288e-06,
+      "loss": 0.837,
+      "step": 6613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8001378820079399,
+      "learning_rate": 4.746754155806437e-06,
+      "loss": 1.0689,
+      "step": 6614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5075738639922274,
+      "learning_rate": 4.7454565794073244e-06,
+      "loss": 0.7958,
+      "step": 6615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5241153675449426,
+      "learning_rate": 4.744159020195566e-06,
+      "loss": 0.7396,
+      "step": 6616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4688654206279637,
+      "learning_rate": 4.742861478258775e-06,
+      "loss": 0.6956,
+      "step": 6617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4931990326876936,
+      "learning_rate": 4.741563953684566e-06,
+      "loss": 0.7882,
+      "step": 6618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.806189694085349,
+      "learning_rate": 4.740266446560547e-06,
+      "loss": 1.1117,
+      "step": 6619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7870570703100621,
+      "learning_rate": 4.738968956974334e-06,
+      "loss": 1.1343,
+      "step": 6620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9967818814112137,
+      "learning_rate": 4.737671485013533e-06,
+      "loss": 0.7473,
+      "step": 6621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.573910938302091,
+      "learning_rate": 4.736374030765754e-06,
+      "loss": 0.8197,
+      "step": 6622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5808421440003961,
+      "learning_rate": 4.735076594318602e-06,
+      "loss": 0.7474,
+      "step": 6623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7922501525137619,
+      "learning_rate": 4.733779175759685e-06,
+      "loss": 1.1262,
+      "step": 6624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5901335743839164,
+      "learning_rate": 4.732481775176607e-06,
+      "loss": 0.8136,
+      "step": 6625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8138392900959909,
+      "learning_rate": 4.7311843926569704e-06,
+      "loss": 1.0774,
+      "step": 6626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6722990043666195,
+      "learning_rate": 4.729887028288379e-06,
+      "loss": 0.8477,
+      "step": 6627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4525515458959124,
+      "learning_rate": 4.728589682158434e-06,
+      "loss": 0.7842,
+      "step": 6628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5248037366533527,
+      "learning_rate": 4.727292354354731e-06,
+      "loss": 0.7462,
+      "step": 6629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6017422043822407,
+      "learning_rate": 4.725995044964874e-06,
+      "loss": 0.8397,
+      "step": 6630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7841607134810539,
+      "learning_rate": 4.724697754076459e-06,
+      "loss": 1.0739,
+      "step": 6631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5029753538437,
+      "learning_rate": 4.723400481777081e-06,
+      "loss": 0.7911,
+      "step": 6632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.538012276920181,
+      "learning_rate": 4.722103228154333e-06,
+      "loss": 0.7769,
+      "step": 6633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4976018561001678,
+      "learning_rate": 4.7208059932958125e-06,
+      "loss": 0.7673,
+      "step": 6634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4790421783976868,
+      "learning_rate": 4.7195087772891096e-06,
+      "loss": 0.7318,
+      "step": 6635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.606771523702515,
+      "learning_rate": 4.718211580221813e-06,
+      "loss": 0.8078,
+      "step": 6636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7898932202575697,
+      "learning_rate": 4.716914402181517e-06,
+      "loss": 1.082,
+      "step": 6637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4350198102498026,
+      "learning_rate": 4.7156172432558075e-06,
+      "loss": 0.7396,
+      "step": 6638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4583377218324256,
+      "learning_rate": 4.7143201035322735e-06,
+      "loss": 0.7377,
+      "step": 6639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5104258124647387,
+      "learning_rate": 4.713022983098496e-06,
+      "loss": 0.7698,
+      "step": 6640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4191909123602346,
+      "learning_rate": 4.711725882042066e-06,
+      "loss": 0.7738,
+      "step": 6641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4676470109201796,
+      "learning_rate": 4.710428800450562e-06,
+      "loss": 0.7384,
+      "step": 6642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.596183862664256,
+      "learning_rate": 4.7091317384115675e-06,
+      "loss": 0.7628,
+      "step": 6643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4655418121031003,
+      "learning_rate": 4.7078346960126645e-06,
+      "loss": 0.8198,
+      "step": 6644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.479252919465189,
+      "learning_rate": 4.70653767334143e-06,
+      "loss": 0.7871,
+      "step": 6645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5370994950701813,
+      "learning_rate": 4.705240670485441e-06,
+      "loss": 0.6917,
+      "step": 6646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8581882244882423,
+      "learning_rate": 4.703943687532279e-06,
+      "loss": 1.0672,
+      "step": 6647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4835967447244494,
+      "learning_rate": 4.7026467245695155e-06,
+      "loss": 0.8103,
+      "step": 6648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5687801600978966,
+      "learning_rate": 4.701349781684724e-06,
+      "loss": 0.7876,
+      "step": 6649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5663925726685595,
+      "learning_rate": 4.700052858965478e-06,
+      "loss": 0.8471,
+      "step": 6650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5506410608758596,
+      "learning_rate": 4.69875595649935e-06,
+      "loss": 0.8428,
+      "step": 6651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4200634579855331,
+      "learning_rate": 4.697459074373909e-06,
+      "loss": 0.7736,
+      "step": 6652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8038886612715305,
+      "learning_rate": 4.696162212676721e-06,
+      "loss": 1.086,
+      "step": 6653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5144043076476261,
+      "learning_rate": 4.694865371495357e-06,
+      "loss": 0.7628,
+      "step": 6654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.560584275191394,
+      "learning_rate": 4.6935685509173815e-06,
+      "loss": 0.7415,
+      "step": 6655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.564844451295664,
+      "learning_rate": 4.6922717510303565e-06,
+      "loss": 0.7083,
+      "step": 6656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6757220491581997,
+      "learning_rate": 4.690974971921846e-06,
+      "loss": 0.7848,
+      "step": 6657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7746224590638965,
+      "learning_rate": 4.6896782136794126e-06,
+      "loss": 1.1006,
+      "step": 6658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6090707877798003,
+      "learning_rate": 4.688381476390617e-06,
+      "loss": 0.7793,
+      "step": 6659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.498298428451919,
+      "learning_rate": 4.687084760143015e-06,
+      "loss": 0.7936,
+      "step": 6660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4635141617602387,
+      "learning_rate": 4.685788065024167e-06,
+      "loss": 0.7034,
+      "step": 6661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5398026829545355,
+      "learning_rate": 4.684491391121628e-06,
+      "loss": 0.8135,
+      "step": 6662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5833776514688946,
+      "learning_rate": 4.68319473852295e-06,
+      "loss": 0.8264,
+      "step": 6663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7604121489925689,
+      "learning_rate": 4.681898107315687e-06,
+      "loss": 1.0501,
+      "step": 6664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2173088905073084,
+      "learning_rate": 4.680601497587392e-06,
+      "loss": 0.7406,
+      "step": 6665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5345188888686891,
+      "learning_rate": 4.679304909425615e-06,
+      "loss": 0.7522,
+      "step": 6666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4288529756206767,
+      "learning_rate": 4.678008342917903e-06,
+      "loss": 0.6677,
+      "step": 6667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.394454320906747,
+      "learning_rate": 4.676711798151805e-06,
+      "loss": 0.7554,
+      "step": 6668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4339047722445055,
+      "learning_rate": 4.675415275214865e-06,
+      "loss": 0.7756,
+      "step": 6669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.367358724300926,
+      "learning_rate": 4.674118774194627e-06,
+      "loss": 0.6872,
+      "step": 6670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8111319230339418,
+      "learning_rate": 4.672822295178636e-06,
+      "loss": 1.0796,
+      "step": 6671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5157160066564959,
+      "learning_rate": 4.671525838254432e-06,
+      "loss": 0.8594,
+      "step": 6672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.570599103074908,
+      "learning_rate": 4.670229403509554e-06,
+      "loss": 0.7373,
+      "step": 6673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6292486363137393,
+      "learning_rate": 4.668932991031538e-06,
+      "loss": 0.8171,
+      "step": 6674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6925484597669624,
+      "learning_rate": 4.667636600907926e-06,
+      "loss": 0.7265,
+      "step": 6675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5343612639345021,
+      "learning_rate": 4.666340233226251e-06,
+      "loss": 0.7719,
+      "step": 6676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5317379202004135,
+      "learning_rate": 4.665043888074045e-06,
+      "loss": 0.8363,
+      "step": 6677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.78170101378087,
+      "learning_rate": 4.663747565538843e-06,
+      "loss": 1.0703,
+      "step": 6678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4259998089535644,
+      "learning_rate": 4.662451265708174e-06,
+      "loss": 0.6877,
+      "step": 6679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4984130877000916,
+      "learning_rate": 4.661154988669569e-06,
+      "loss": 0.6923,
+      "step": 6680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.463762241120263,
+      "learning_rate": 4.65985873451055e-06,
+      "loss": 0.8314,
+      "step": 6681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5525144411041807,
+      "learning_rate": 4.658562503318649e-06,
+      "loss": 0.8054,
+      "step": 6682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4240767501890976,
+      "learning_rate": 4.657266295181391e-06,
+      "loss": 0.7769,
+      "step": 6683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4783289076551482,
+      "learning_rate": 4.655970110186292e-06,
+      "loss": 0.7553,
+      "step": 6684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3379860833532575,
+      "learning_rate": 4.65467394842088e-06,
+      "loss": 0.6806,
+      "step": 6685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.514312800849696,
+      "learning_rate": 4.653377809972673e-06,
+      "loss": 0.8235,
+      "step": 6686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4768252689611825,
+      "learning_rate": 4.652081694929188e-06,
+      "loss": 0.7278,
+      "step": 6687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.589736341131553,
+      "learning_rate": 4.6507856033779435e-06,
+      "loss": 0.8412,
+      "step": 6688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.823976052903158,
+      "learning_rate": 4.649489535406454e-06,
+      "loss": 1.0525,
+      "step": 6689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5354660524114252,
+      "learning_rate": 4.648193491102232e-06,
+      "loss": 0.84,
+      "step": 6690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6244741935801714,
+      "learning_rate": 4.646897470552787e-06,
+      "loss": 0.7628,
+      "step": 6691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.770626869681101,
+      "learning_rate": 4.645601473845636e-06,
+      "loss": 0.8245,
+      "step": 6692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6251600825371855,
+      "learning_rate": 4.644305501068283e-06,
+      "loss": 0.8183,
+      "step": 6693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5718121726417718,
+      "learning_rate": 4.643009552308235e-06,
+      "loss": 0.7205,
+      "step": 6694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7952940710994577,
+      "learning_rate": 4.641713627652999e-06,
+      "loss": 1.0761,
+      "step": 6695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4311785122973297,
+      "learning_rate": 4.6404177271900785e-06,
+      "loss": 0.7139,
+      "step": 6696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4215280386417883,
+      "learning_rate": 4.639121851006976e-06,
+      "loss": 0.7649,
+      "step": 6697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4802841135858886,
+      "learning_rate": 4.637825999191189e-06,
+      "loss": 0.7783,
+      "step": 6698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5564792881890266,
+      "learning_rate": 4.63653017183022e-06,
+      "loss": 0.7016,
+      "step": 6699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.547901424012035,
+      "learning_rate": 4.635234369011565e-06,
+      "loss": 0.7907,
+      "step": 6700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5468952830747797,
+      "learning_rate": 4.633938590822718e-06,
+      "loss": 0.7879,
+      "step": 6701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6453643022066053,
+      "learning_rate": 4.632642837351176e-06,
+      "loss": 0.7652,
+      "step": 6702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.531767100423295,
+      "learning_rate": 4.63134710868443e-06,
+      "loss": 0.7335,
+      "step": 6703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7732116298994447,
+      "learning_rate": 4.6300514049099694e-06,
+      "loss": 1.0656,
+      "step": 6704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5176588417586472,
+      "learning_rate": 4.628755726115284e-06,
+      "loss": 0.9211,
+      "step": 6705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5950946284854635,
+      "learning_rate": 4.627460072387861e-06,
+      "loss": 0.6642,
+      "step": 6706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4651150744273913,
+      "learning_rate": 4.626164443815186e-06,
+      "loss": 0.7216,
+      "step": 6707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.642438497613779,
+      "learning_rate": 4.62486884048474e-06,
+      "loss": 0.7551,
+      "step": 6708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5932013396363436,
+      "learning_rate": 4.623573262484011e-06,
+      "loss": 0.749,
+      "step": 6709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4579368862463598,
+      "learning_rate": 4.622277709900475e-06,
+      "loss": 0.8149,
+      "step": 6710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4552237990496306,
+      "learning_rate": 4.620982182821611e-06,
+      "loss": 0.7842,
+      "step": 6711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4373146392670582,
+      "learning_rate": 4.6196866813348985e-06,
+      "loss": 0.7655,
+      "step": 6712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.754321596951422,
+      "learning_rate": 4.61839120552781e-06,
+      "loss": 1.0464,
+      "step": 6713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8243537298753041,
+      "learning_rate": 4.61709575548782e-06,
+      "loss": 1.062,
+      "step": 6714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4478369578708192,
+      "learning_rate": 4.615800331302399e-06,
+      "loss": 0.7067,
+      "step": 6715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7899929874187903,
+      "learning_rate": 4.61450493305902e-06,
+      "loss": 1.1153,
+      "step": 6716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4875675701222748,
+      "learning_rate": 4.613209560845148e-06,
+      "loss": 0.7452,
+      "step": 6717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5550751965291805,
+      "learning_rate": 4.61191421474825e-06,
+      "loss": 0.7919,
+      "step": 6718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8034537556591744,
+      "learning_rate": 4.6106188948557935e-06,
+      "loss": 1.1145,
+      "step": 6719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.647604902067715,
+      "learning_rate": 4.6093236012552394e-06,
+      "loss": 0.734,
+      "step": 6720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4318057570190756,
+      "learning_rate": 4.608028334034049e-06,
+      "loss": 0.7742,
+      "step": 6721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4718741665836244,
+      "learning_rate": 4.606733093279681e-06,
+      "loss": 0.7239,
+      "step": 6722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5676067548902917,
+      "learning_rate": 4.605437879079595e-06,
+      "loss": 0.733,
+      "step": 6723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5594260972025653,
+      "learning_rate": 4.604142691521245e-06,
+      "loss": 0.7478,
+      "step": 6724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.575096258415649,
+      "learning_rate": 4.602847530692084e-06,
+      "loss": 0.806,
+      "step": 6725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.795613161803088,
+      "learning_rate": 4.601552396679568e-06,
+      "loss": 1.0592,
+      "step": 6726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.543929949722361,
+      "learning_rate": 4.600257289571146e-06,
+      "loss": 0.7689,
+      "step": 6727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6078631085204615,
+      "learning_rate": 4.598962209454263e-06,
+      "loss": 0.7019,
+      "step": 6728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5021400011444175,
+      "learning_rate": 4.597667156416371e-06,
+      "loss": 0.697,
+      "step": 6729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.468244648144051,
+      "learning_rate": 4.5963721305449125e-06,
+      "loss": 0.7837,
+      "step": 6730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5081647617230125,
+      "learning_rate": 4.59507713192733e-06,
+      "loss": 0.7281,
+      "step": 6731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5319842812390967,
+      "learning_rate": 4.5937821606510656e-06,
+      "loss": 0.7644,
+      "step": 6732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.496813507738929,
+      "learning_rate": 4.592487216803558e-06,
+      "loss": 0.7571,
+      "step": 6733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5574620071011724,
+      "learning_rate": 4.591192300472247e-06,
+      "loss": 0.8295,
+      "step": 6734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7133051998494504,
+      "learning_rate": 4.589897411744563e-06,
+      "loss": 0.8317,
+      "step": 6735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5166679878226084,
+      "learning_rate": 4.5886025507079465e-06,
+      "loss": 0.7132,
+      "step": 6736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4906586720634227,
+      "learning_rate": 4.587307717449827e-06,
+      "loss": 0.6493,
+      "step": 6737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.437571371418521,
+      "learning_rate": 4.586012912057634e-06,
+      "loss": 0.6981,
+      "step": 6738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5305128465394449,
+      "learning_rate": 4.584718134618793e-06,
+      "loss": 0.7461,
+      "step": 6739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4664683633004731,
+      "learning_rate": 4.583423385220736e-06,
+      "loss": 0.7419,
+      "step": 6740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5316537848328047,
+      "learning_rate": 4.582128663950884e-06,
+      "loss": 0.8022,
+      "step": 6741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4631190526919073,
+      "learning_rate": 4.58083397089666e-06,
+      "loss": 0.7128,
+      "step": 6742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.533318776570989,
+      "learning_rate": 4.579539306145485e-06,
+      "loss": 0.7611,
+      "step": 6743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4369447996994582,
+      "learning_rate": 4.5782446697847775e-06,
+      "loss": 0.8259,
+      "step": 6744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4803991955468436,
+      "learning_rate": 4.5769500619019516e-06,
+      "loss": 0.7081,
+      "step": 6745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8516004550012849,
+      "learning_rate": 4.575655482584428e-06,
+      "loss": 1.0909,
+      "step": 6746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6489730746279472,
+      "learning_rate": 4.574360931919616e-06,
+      "loss": 0.8371,
+      "step": 6747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4775898985832043,
+      "learning_rate": 4.573066409994927e-06,
+      "loss": 0.8301,
+      "step": 6748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4464101695235616,
+      "learning_rate": 4.5717719168977696e-06,
+      "loss": 0.6815,
+      "step": 6749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3881902265020774,
+      "learning_rate": 4.570477452715553e-06,
+      "loss": 0.8009,
+      "step": 6750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.485082877876475,
+      "learning_rate": 4.56918301753568e-06,
+      "loss": 0.7971,
+      "step": 6751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5469149809237595,
+      "learning_rate": 4.567888611445553e-06,
+      "loss": 0.7328,
+      "step": 6752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5075360125769597,
+      "learning_rate": 4.5665942345325776e-06,
+      "loss": 0.779,
+      "step": 6753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5420553817821978,
+      "learning_rate": 4.56529988688415e-06,
+      "loss": 0.8087,
+      "step": 6754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8241051131335586,
+      "learning_rate": 4.5640055685876695e-06,
+      "loss": 1.0852,
+      "step": 6755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5289716567793399,
+      "learning_rate": 4.562711279730526e-06,
+      "loss": 0.7025,
+      "step": 6756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5251346371663899,
+      "learning_rate": 4.561417020400119e-06,
+      "loss": 0.7689,
+      "step": 6757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4357136554797612,
+      "learning_rate": 4.560122790683839e-06,
+      "loss": 0.7106,
+      "step": 6758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.329206321389052,
+      "learning_rate": 4.558828590669072e-06,
+      "loss": 0.6608,
+      "step": 6759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4967299395728209,
+      "learning_rate": 4.557534420443209e-06,
+      "loss": 0.8004,
+      "step": 6760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8183998303087598,
+      "learning_rate": 4.556240280093633e-06,
+      "loss": 1.1025,
+      "step": 6761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5034617970270898,
+      "learning_rate": 4.554946169707728e-06,
+      "loss": 0.7609,
+      "step": 6762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7153262492485235,
+      "learning_rate": 4.5536520893728735e-06,
+      "loss": 0.738,
+      "step": 6763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4252224006931138,
+      "learning_rate": 4.552358039176453e-06,
+      "loss": 0.8231,
+      "step": 6764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5104147479047756,
+      "learning_rate": 4.551064019205841e-06,
+      "loss": 0.7451,
+      "step": 6765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.472387305723786,
+      "learning_rate": 4.549770029548411e-06,
+      "loss": 0.7371,
+      "step": 6766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5169088508942212,
+      "learning_rate": 4.548476070291541e-06,
+      "loss": 0.776,
+      "step": 6767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4477797162462265,
+      "learning_rate": 4.547182141522598e-06,
+      "loss": 0.7409,
+      "step": 6768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4456352776386698,
+      "learning_rate": 4.54588824332895e-06,
+      "loss": 0.7651,
+      "step": 6769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5538468128265064,
+      "learning_rate": 4.544594375797969e-06,
+      "loss": 0.8117,
+      "step": 6770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5246530488803955,
+      "learning_rate": 4.5433005390170174e-06,
+      "loss": 0.727,
+      "step": 6771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5097774995946178,
+      "learning_rate": 4.542006733073457e-06,
+      "loss": 0.7329,
+      "step": 6772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.460901067941363,
+      "learning_rate": 4.540712958054647e-06,
+      "loss": 0.7599,
+      "step": 6773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.677362165291922,
+      "learning_rate": 4.53941921404795e-06,
+      "loss": 0.793,
+      "step": 6774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.546771481604218,
+      "learning_rate": 4.5381255011407225e-06,
+      "loss": 0.742,
+      "step": 6775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5594299280006725,
+      "learning_rate": 4.5368318194203145e-06,
+      "loss": 0.6976,
+      "step": 6776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4846287433488095,
+      "learning_rate": 4.535538168974083e-06,
+      "loss": 0.7257,
+      "step": 6777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8533699156102753,
+      "learning_rate": 4.534244549889376e-06,
+      "loss": 1.0611,
+      "step": 6778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1710059510526634,
+      "learning_rate": 4.532950962253543e-06,
+      "loss": 0.7825,
+      "step": 6779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4619918530137257,
+      "learning_rate": 4.531657406153926e-06,
+      "loss": 0.8072,
+      "step": 6780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7656421734625081,
+      "learning_rate": 4.530363881677874e-06,
+      "loss": 1.0531,
+      "step": 6781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5744699718255917,
+      "learning_rate": 4.529070388912727e-06,
+      "loss": 0.7794,
+      "step": 6782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7671498696004015,
+      "learning_rate": 4.527776927945823e-06,
+      "loss": 1.0708,
+      "step": 6783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5388952735090162,
+      "learning_rate": 4.526483498864501e-06,
+      "loss": 0.8377,
+      "step": 6784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7803529163336377,
+      "learning_rate": 4.525190101756097e-06,
+      "loss": 1.1138,
+      "step": 6785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.534583703032621,
+      "learning_rate": 4.523896736707941e-06,
+      "loss": 0.7754,
+      "step": 6786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4803642475633196,
+      "learning_rate": 4.5226034038073675e-06,
+      "loss": 0.7295,
+      "step": 6787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5412103038710805,
+      "learning_rate": 4.521310103141704e-06,
+      "loss": 0.698,
+      "step": 6788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4889724304111946,
+      "learning_rate": 4.520016834798277e-06,
+      "loss": 0.7737,
+      "step": 6789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6293551368593064,
+      "learning_rate": 4.518723598864408e-06,
+      "loss": 0.814,
+      "step": 6790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.526822594562841,
+      "learning_rate": 4.517430395427424e-06,
+      "loss": 0.7998,
+      "step": 6791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8317215622899846,
+      "learning_rate": 4.516137224574645e-06,
+      "loss": 1.1073,
+      "step": 6792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3779015406401227,
+      "learning_rate": 4.5148440863933845e-06,
+      "loss": 0.7987,
+      "step": 6793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.591676456570048,
+      "learning_rate": 4.513550980970962e-06,
+      "loss": 0.8106,
+      "step": 6794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.469022891554638,
+      "learning_rate": 4.51225790839469e-06,
+      "loss": 0.7631,
+      "step": 6795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3530491833764033,
+      "learning_rate": 4.510964868751879e-06,
+      "loss": 0.7212,
+      "step": 6796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3702536841853044,
+      "learning_rate": 4.509671862129837e-06,
+      "loss": 0.6864,
+      "step": 6797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4569782211077371,
+      "learning_rate": 4.508378888615872e-06,
+      "loss": 0.8171,
+      "step": 6798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0067336125626194,
+      "learning_rate": 4.5070859482972915e-06,
+      "loss": 0.8205,
+      "step": 6799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4790912918879442,
+      "learning_rate": 4.505793041261391e-06,
+      "loss": 0.7102,
+      "step": 6800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4889596548626505,
+      "learning_rate": 4.504500167595477e-06,
+      "loss": 0.7772,
+      "step": 6801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5312732414528996,
+      "learning_rate": 4.5032073273868445e-06,
+      "loss": 0.8229,
+      "step": 6802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8174747478934806,
+      "learning_rate": 4.501914520722788e-06,
+      "loss": 1.0636,
+      "step": 6803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6080743361931922,
+      "learning_rate": 4.500621747690604e-06,
+      "loss": 0.8181,
+      "step": 6804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5343800880222516,
+      "learning_rate": 4.499329008377581e-06,
+      "loss": 0.7295,
+      "step": 6805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6019152815234603,
+      "learning_rate": 4.498036302871007e-06,
+      "loss": 0.7587,
+      "step": 6806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.482259681491783,
+      "learning_rate": 4.496743631258169e-06,
+      "loss": 0.7672,
+      "step": 6807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4286051518796945,
+      "learning_rate": 4.495450993626353e-06,
+      "loss": 0.7804,
+      "step": 6808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5064271914105436,
+      "learning_rate": 4.49415839006284e-06,
+      "loss": 0.7615,
+      "step": 6809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5735823783458043,
+      "learning_rate": 4.492865820654908e-06,
+      "loss": 0.7235,
+      "step": 6810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8481495383951919,
+      "learning_rate": 4.4915732854898365e-06,
+      "loss": 1.0824,
+      "step": 6811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.581371367471879,
+      "learning_rate": 4.490280784654899e-06,
+      "loss": 0.8587,
+      "step": 6812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5775968935822104,
+      "learning_rate": 4.48898831823737e-06,
+      "loss": 0.7906,
+      "step": 6813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8127386339365587,
+      "learning_rate": 4.487695886324514e-06,
+      "loss": 1.0742,
+      "step": 6814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7955865719098081,
+      "learning_rate": 4.486403489003608e-06,
+      "loss": 1.1129,
+      "step": 6815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5065994240678835,
+      "learning_rate": 4.485111126361911e-06,
+      "loss": 0.7565,
+      "step": 6816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8171745746729149,
+      "learning_rate": 4.4838187984866865e-06,
+      "loss": 1.0816,
+      "step": 6817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4392642299641591,
+      "learning_rate": 4.482526505465199e-06,
+      "loss": 0.7791,
+      "step": 6818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5144736251670543,
+      "learning_rate": 4.481234247384705e-06,
+      "loss": 0.7552,
+      "step": 6819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5553505809898935,
+      "learning_rate": 4.4799420243324605e-06,
+      "loss": 0.8053,
+      "step": 6820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5110060358236679,
+      "learning_rate": 4.478649836395719e-06,
+      "loss": 0.8194,
+      "step": 6821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6714828521175757,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.7787,
+      "step": 6822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5798657477435094,
+      "learning_rate": 4.476065566217753e-06,
+      "loss": 0.7194,
+      "step": 6823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4557559539125728,
+      "learning_rate": 4.474773484151021e-06,
+      "loss": 0.7453,
+      "step": 6824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7990492611754859,
+      "learning_rate": 4.473481437548786e-06,
+      "loss": 1.0847,
+      "step": 6825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4928686871888033,
+      "learning_rate": 4.472189426498289e-06,
+      "loss": 0.7229,
+      "step": 6826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4366613209559382,
+      "learning_rate": 4.470897451086767e-06,
+      "loss": 0.6576,
+      "step": 6827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7463674199635106,
+      "learning_rate": 4.469605511401461e-06,
+      "loss": 0.8391,
+      "step": 6828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5735965916233225,
+      "learning_rate": 4.468313607529603e-06,
+      "loss": 0.8312,
+      "step": 6829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4658736325721764,
+      "learning_rate": 4.467021739558426e-06,
+      "loss": 0.796,
+      "step": 6830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.559203935588992,
+      "learning_rate": 4.4657299075751596e-06,
+      "loss": 0.8094,
+      "step": 6831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7442613403578175,
+      "learning_rate": 4.464438111667031e-06,
+      "loss": 0.7753,
+      "step": 6832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8484585453808401,
+      "learning_rate": 4.463146351921267e-06,
+      "loss": 1.1134,
+      "step": 6833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5871504871585123,
+      "learning_rate": 4.461854628425087e-06,
+      "loss": 0.7893,
+      "step": 6834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5863662854629188,
+      "learning_rate": 4.4605629412657145e-06,
+      "loss": 0.8106,
+      "step": 6835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3995744465396254,
+      "learning_rate": 4.459271290530365e-06,
+      "loss": 0.7633,
+      "step": 6836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6132150650393045,
+      "learning_rate": 4.457979676306256e-06,
+      "loss": 0.7676,
+      "step": 6837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.572213405310079,
+      "learning_rate": 4.456688098680597e-06,
+      "loss": 0.7734,
+      "step": 6838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5445084477246462,
+      "learning_rate": 4.4553965577406006e-06,
+      "loss": 0.8467,
+      "step": 6839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.822237315842328,
+      "learning_rate": 4.454105053573474e-06,
+      "loss": 1.1095,
+      "step": 6840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4632640660036722,
+      "learning_rate": 4.4528135862664206e-06,
+      "loss": 0.7737,
+      "step": 6841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5004340984924578,
+      "learning_rate": 4.451522155906647e-06,
+      "loss": 0.7785,
+      "step": 6842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.602922472570306,
+      "learning_rate": 4.4502307625813515e-06,
+      "loss": 0.7143,
+      "step": 6843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.628518624088369,
+      "learning_rate": 4.448939406377732e-06,
+      "loss": 0.8366,
+      "step": 6844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.551950771814103,
+      "learning_rate": 4.4476480873829834e-06,
+      "loss": 0.778,
+      "step": 6845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5587565547229294,
+      "learning_rate": 4.4463568056842995e-06,
+      "loss": 0.7627,
+      "step": 6846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.41804815579874,
+      "learning_rate": 4.44506556136887e-06,
+      "loss": 0.7066,
+      "step": 6847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5650562127947183,
+      "learning_rate": 4.443774354523883e-06,
+      "loss": 0.7421,
+      "step": 6848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5881646035115946,
+      "learning_rate": 4.442483185236523e-06,
+      "loss": 0.8756,
+      "step": 6849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5877180095434507,
+      "learning_rate": 4.441192053593973e-06,
+      "loss": 0.7431,
+      "step": 6850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5628029742277691,
+      "learning_rate": 4.439900959683412e-06,
+      "loss": 0.7619,
+      "step": 6851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6823109328360657,
+      "learning_rate": 4.438609903592021e-06,
+      "loss": 0.7013,
+      "step": 6852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8106627077664682,
+      "learning_rate": 4.4373188854069736e-06,
+      "loss": 1.1043,
+      "step": 6853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.78660753323529,
+      "learning_rate": 4.4360279052154406e-06,
+      "loss": 1.0485,
+      "step": 6854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4861430833492006,
+      "learning_rate": 4.434736963104592e-06,
+      "loss": 0.7971,
+      "step": 6855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5599612476771783,
+      "learning_rate": 4.433446059161598e-06,
+      "loss": 0.7865,
+      "step": 6856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5782015687109299,
+      "learning_rate": 4.432155193473621e-06,
+      "loss": 0.754,
+      "step": 6857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.373247083168306,
+      "learning_rate": 4.430864366127821e-06,
+      "loss": 0.7023,
+      "step": 6858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4244475681101365,
+      "learning_rate": 4.429573577211365e-06,
+      "loss": 0.7233,
+      "step": 6859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5563731633318956,
+      "learning_rate": 4.428282826811404e-06,
+      "loss": 0.6454,
+      "step": 6860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5231249732247683,
+      "learning_rate": 4.426992115015094e-06,
+      "loss": 0.7635,
+      "step": 6861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6113718413426183,
+      "learning_rate": 4.425701441909584e-06,
+      "loss": 0.7714,
+      "step": 6862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5221013186223222,
+      "learning_rate": 4.424410807582029e-06,
+      "loss": 0.6973,
+      "step": 6863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5326281527469157,
+      "learning_rate": 4.423120212119571e-06,
+      "loss": 0.7012,
+      "step": 6864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4115488339092077,
+      "learning_rate": 4.421829655609355e-06,
+      "loss": 0.7309,
+      "step": 6865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6106645423358947,
+      "learning_rate": 4.4205391381385235e-06,
+      "loss": 0.8064,
+      "step": 6866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4136202897703214,
+      "learning_rate": 4.419248659794215e-06,
+      "loss": 0.7297,
+      "step": 6867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9110755975678478,
+      "learning_rate": 4.417958220663563e-06,
+      "loss": 1.0842,
+      "step": 6868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4892462545815033,
+      "learning_rate": 4.416667820833704e-06,
+      "loss": 0.7586,
+      "step": 6869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6092782770430807,
+      "learning_rate": 4.415377460391768e-06,
+      "loss": 0.7154,
+      "step": 6870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4767368090081745,
+      "learning_rate": 4.414087139424883e-06,
+      "loss": 0.7858,
+      "step": 6871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7227041638770426,
+      "learning_rate": 4.412796858020173e-06,
+      "loss": 0.7789,
+      "step": 6872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5746817532874446,
+      "learning_rate": 4.4115066162647625e-06,
+      "loss": 0.8315,
+      "step": 6873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7933377422207455,
+      "learning_rate": 4.410216414245771e-06,
+      "loss": 1.0791,
+      "step": 6874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6720518601149836,
+      "learning_rate": 4.408926252050315e-06,
+      "loss": 0.767,
+      "step": 6875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.592339115271695,
+      "learning_rate": 4.407636129765511e-06,
+      "loss": 0.8208,
+      "step": 6876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4224869843135501,
+      "learning_rate": 4.406346047478471e-06,
+      "loss": 0.7421,
+      "step": 6877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5082418924718668,
+      "learning_rate": 4.405056005276302e-06,
+      "loss": 0.8334,
+      "step": 6878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5676682238551272,
+      "learning_rate": 4.40376600324611e-06,
+      "loss": 0.7218,
+      "step": 6879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4171720635387464,
+      "learning_rate": 4.402476041475004e-06,
+      "loss": 0.8057,
+      "step": 6880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.629833326510161,
+      "learning_rate": 4.401186120050081e-06,
+      "loss": 0.8157,
+      "step": 6881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6027327796867108,
+      "learning_rate": 4.399896239058439e-06,
+      "loss": 0.82,
+      "step": 6882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5842443885304018,
+      "learning_rate": 4.398606398587177e-06,
+      "loss": 0.7632,
+      "step": 6883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5861413624681564,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.7398,
+      "step": 6884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5173027743671035,
+      "learning_rate": 4.396026839554154e-06,
+      "loss": 0.7533,
+      "step": 6885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5004790126881262,
+      "learning_rate": 4.394737121166573e-06,
+      "loss": 0.7174,
+      "step": 6886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5437648118791238,
+      "learning_rate": 4.393447443647726e-06,
+      "loss": 0.705,
+      "step": 6887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5522244124785598,
+      "learning_rate": 4.392157807084696e-06,
+      "loss": 0.7704,
+      "step": 6888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6022007338449897,
+      "learning_rate": 4.3908682115645565e-06,
+      "loss": 0.7716,
+      "step": 6889
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5616732588163402,
+      "learning_rate": 4.389578657174391e-06,
+      "loss": 0.7422,
+      "step": 6890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6171168510841951,
+      "learning_rate": 4.38828914400127e-06,
+      "loss": 0.7797,
+      "step": 6891
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.640118408610226,
+      "learning_rate": 4.386999672132264e-06,
+      "loss": 0.7773,
+      "step": 6892
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.476944075019624,
+      "learning_rate": 4.385710241654443e-06,
+      "loss": 0.7847,
+      "step": 6893
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4528551323397358,
+      "learning_rate": 4.3844208526548705e-06,
+      "loss": 0.794,
+      "step": 6894
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.610822339479896,
+      "learning_rate": 4.38313150522061e-06,
+      "loss": 0.7502,
+      "step": 6895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4554778106596893,
+      "learning_rate": 4.3818421994387194e-06,
+      "loss": 0.7536,
+      "step": 6896
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4687121775242085,
+      "learning_rate": 4.380552935396259e-06,
+      "loss": 0.6813,
+      "step": 6897
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8181727482761814,
+      "learning_rate": 4.3792637131802805e-06,
+      "loss": 1.0776,
+      "step": 6898
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3823714627796473,
+      "learning_rate": 4.377974532877834e-06,
+      "loss": 0.8764,
+      "step": 6899
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7953269306952987,
+      "learning_rate": 4.376685394575971e-06,
+      "loss": 1.0971,
+      "step": 6900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8007194022591374,
+      "learning_rate": 4.375396298361735e-06,
+      "loss": 1.0967,
+      "step": 6901
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7493237587248764,
+      "learning_rate": 4.374107244322167e-06,
+      "loss": 1.1089,
+      "step": 6902
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4121823042023827,
+      "learning_rate": 4.372818232544313e-06,
+      "loss": 0.7301,
+      "step": 6903
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5795239805200996,
+      "learning_rate": 4.3715292631152045e-06,
+      "loss": 0.7039,
+      "step": 6904
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.47498526910125,
+      "learning_rate": 4.370240336121877e-06,
+      "loss": 0.7258,
+      "step": 6905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8499374479124103,
+      "learning_rate": 4.36895145165136e-06,
+      "loss": 1.1284,
+      "step": 6906
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.533146721066811,
+      "learning_rate": 4.3676626097906865e-06,
+      "loss": 0.7324,
+      "step": 6907
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8523982157150912,
+      "learning_rate": 4.36637381062688e-06,
+      "loss": 1.0426,
+      "step": 6908
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4547962368763625,
+      "learning_rate": 4.36508505424696e-06,
+      "loss": 0.7518,
+      "step": 6909
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4919433299993383,
+      "learning_rate": 4.3637963407379515e-06,
+      "loss": 0.7194,
+      "step": 6910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.518883604910818,
+      "learning_rate": 4.362507670186868e-06,
+      "loss": 0.7686,
+      "step": 6911
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4195179663838116,
+      "learning_rate": 4.361219042680725e-06,
+      "loss": 0.7521,
+      "step": 6912
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8047798541250969,
+      "learning_rate": 4.35993045830653e-06,
+      "loss": 1.098,
+      "step": 6913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.547962436599547,
+      "learning_rate": 4.358641917151297e-06,
+      "loss": 0.7756,
+      "step": 6914
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6050091722781497,
+      "learning_rate": 4.357353419302028e-06,
+      "loss": 0.81,
+      "step": 6915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4500043309851394,
+      "learning_rate": 4.356064964845724e-06,
+      "loss": 0.7347,
+      "step": 6916
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8878544672012398,
+      "learning_rate": 4.354776553869387e-06,
+      "loss": 0.8111,
+      "step": 6917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6642289457875687,
+      "learning_rate": 4.3534881864600135e-06,
+      "loss": 0.8092,
+      "step": 6918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4723799994708773,
+      "learning_rate": 4.352199862704596e-06,
+      "loss": 0.8402,
+      "step": 6919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.561344852843685,
+      "learning_rate": 4.3509115826901235e-06,
+      "loss": 0.8075,
+      "step": 6920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5666981384899021,
+      "learning_rate": 4.349623346503586e-06,
+      "loss": 0.8227,
+      "step": 6921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7717022688955494,
+      "learning_rate": 4.348335154231967e-06,
+      "loss": 1.0715,
+      "step": 6922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8404357962483123,
+      "learning_rate": 4.347047005962247e-06,
+      "loss": 1.0741,
+      "step": 6923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4997388002546004,
+      "learning_rate": 4.345758901781408e-06,
+      "loss": 0.8208,
+      "step": 6924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7913310757988739,
+      "learning_rate": 4.344470841776424e-06,
+      "loss": 1.0882,
+      "step": 6925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4613413673643725,
+      "learning_rate": 4.343182826034268e-06,
+      "loss": 0.7903,
+      "step": 6926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5100074754962065,
+      "learning_rate": 4.34189485464191e-06,
+      "loss": 0.7588,
+      "step": 6927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5058973473258286,
+      "learning_rate": 4.340606927686315e-06,
+      "loss": 0.6989,
+      "step": 6928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4963407646823559,
+      "learning_rate": 4.339319045254448e-06,
+      "loss": 0.7585,
+      "step": 6929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4822583221026402,
+      "learning_rate": 4.338031207433268e-06,
+      "loss": 0.7716,
+      "step": 6930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4661476956748791,
+      "learning_rate": 4.336743414309738e-06,
+      "loss": 0.7061,
+      "step": 6931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.575986646681442,
+      "learning_rate": 4.33545566597081e-06,
+      "loss": 0.7497,
+      "step": 6932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6679979777982052,
+      "learning_rate": 4.33416796250343e-06,
+      "loss": 0.7469,
+      "step": 6933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4893703705346955,
+      "learning_rate": 4.3328803039945555e-06,
+      "loss": 0.7557,
+      "step": 6934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5335983244726297,
+      "learning_rate": 4.331592690531128e-06,
+      "loss": 0.7944,
+      "step": 6935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5554704593814996,
+      "learning_rate": 4.330305122200091e-06,
+      "loss": 0.7952,
+      "step": 6936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4166634751349931,
+      "learning_rate": 4.3290175990883815e-06,
+      "loss": 0.8234,
+      "step": 6937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8724591301767475,
+      "learning_rate": 4.327730121282939e-06,
+      "loss": 1.0937,
+      "step": 6938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7730911623411953,
+      "learning_rate": 4.326442688870697e-06,
+      "loss": 0.7686,
+      "step": 6939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4488985979335967,
+      "learning_rate": 4.325155301938582e-06,
+      "loss": 0.8032,
+      "step": 6940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5411220689957446,
+      "learning_rate": 4.323867960573526e-06,
+      "loss": 0.7719,
+      "step": 6941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7921506584053002,
+      "learning_rate": 4.3225806648624516e-06,
+      "loss": 1.1082,
+      "step": 6942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7910349842892567,
+      "learning_rate": 4.3212934148922785e-06,
+      "loss": 1.0549,
+      "step": 6943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5122529600131875,
+      "learning_rate": 4.320006210749928e-06,
+      "loss": 0.7701,
+      "step": 6944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5044272456969652,
+      "learning_rate": 4.318719052522312e-06,
+      "loss": 0.7584,
+      "step": 6945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4504793080707057,
+      "learning_rate": 4.3174319402963436e-06,
+      "loss": 0.7462,
+      "step": 6946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.474591126001228,
+      "learning_rate": 4.3161448741589305e-06,
+      "loss": 0.8647,
+      "step": 6947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5933982697412945,
+      "learning_rate": 4.31485785419698e-06,
+      "loss": 0.7856,
+      "step": 6948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8999830664151784,
+      "learning_rate": 4.313570880497394e-06,
+      "loss": 1.111,
+      "step": 6949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.540896401430285,
+      "learning_rate": 4.312283953147069e-06,
+      "loss": 0.7803,
+      "step": 6950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8168646182965901,
+      "learning_rate": 4.310997072232907e-06,
+      "loss": 1.0933,
+      "step": 6951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5886395803885651,
+      "learning_rate": 4.3097102378417985e-06,
+      "loss": 0.8706,
+      "step": 6952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5487830453006384,
+      "learning_rate": 4.3084234500606334e-06,
+      "loss": 0.8355,
+      "step": 6953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7578411591482668,
+      "learning_rate": 4.3071367089762965e-06,
+      "loss": 1.0814,
+      "step": 6954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4641970801517272,
+      "learning_rate": 4.305850014675675e-06,
+      "loss": 0.8014,
+      "step": 6955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7586025420028498,
+      "learning_rate": 4.304563367245646e-06,
+      "loss": 1.1198,
+      "step": 6956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6697543356987439,
+      "learning_rate": 4.303276766773088e-06,
+      "loss": 0.6382,
+      "step": 6957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6743491278819924,
+      "learning_rate": 4.301990213344878e-06,
+      "loss": 0.8116,
+      "step": 6958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7713396386967131,
+      "learning_rate": 4.3007037070478855e-06,
+      "loss": 1.0757,
+      "step": 6959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5203509873255936,
+      "learning_rate": 4.299417247968978e-06,
+      "loss": 0.7973,
+      "step": 6960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.647129247588905,
+      "learning_rate": 4.298130836195017e-06,
+      "loss": 0.6954,
+      "step": 6961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4668755552320825,
+      "learning_rate": 4.296844471812868e-06,
+      "loss": 0.7819,
+      "step": 6962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7106202133608355,
+      "learning_rate": 4.295558154909389e-06,
+      "loss": 0.7567,
+      "step": 6963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7827394005865854,
+      "learning_rate": 4.294271885571433e-06,
+      "loss": 1.0605,
+      "step": 6964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5445077440366493,
+      "learning_rate": 4.292985663885854e-06,
+      "loss": 0.7717,
+      "step": 6965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7926750189186063,
+      "learning_rate": 4.291699489939499e-06,
+      "loss": 1.0698,
+      "step": 6966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5417004091368554,
+      "learning_rate": 4.290413363819213e-06,
+      "loss": 0.7775,
+      "step": 6967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5838542939159044,
+      "learning_rate": 4.2891272856118415e-06,
+      "loss": 0.8278,
+      "step": 6968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4723974559251585,
+      "learning_rate": 4.28784125540422e-06,
+      "loss": 0.7631,
+      "step": 6969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8110665782725236,
+      "learning_rate": 4.2865552732831864e-06,
+      "loss": 1.0759,
+      "step": 6970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.458405662978379,
+      "learning_rate": 4.285269339335571e-06,
+      "loss": 0.7241,
+      "step": 6971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7882538328398603,
+      "learning_rate": 4.283983453648205e-06,
+      "loss": 1.0965,
+      "step": 6972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6873386588233177,
+      "learning_rate": 4.282697616307913e-06,
+      "loss": 0.7431,
+      "step": 6973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7411033990284964,
+      "learning_rate": 4.281411827401517e-06,
+      "loss": 1.0536,
+      "step": 6974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4888270837813484,
+      "learning_rate": 4.280126087015839e-06,
+      "loss": 0.7285,
+      "step": 6975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.774768537152064,
+      "learning_rate": 4.278840395237695e-06,
+      "loss": 1.0832,
+      "step": 6976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4552873350324782,
+      "learning_rate": 4.277554752153895e-06,
+      "loss": 0.7154,
+      "step": 6977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7632505453358764,
+      "learning_rate": 4.2762691578512485e-06,
+      "loss": 1.0882,
+      "step": 6978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6134987582781806,
+      "learning_rate": 4.274983612416566e-06,
+      "loss": 0.8242,
+      "step": 6979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7700126464417867,
+      "learning_rate": 4.273698115936647e-06,
+      "loss": 1.0606,
+      "step": 6980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.47960890502507,
+      "learning_rate": 4.272412668498291e-06,
+      "loss": 0.6921,
+      "step": 6981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8985859201454274,
+      "learning_rate": 4.271127270188297e-06,
+      "loss": 0.6952,
+      "step": 6982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5274719121498368,
+      "learning_rate": 4.269841921093456e-06,
+      "loss": 0.7485,
+      "step": 6983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5031179881657442,
+      "learning_rate": 4.268556621300555e-06,
+      "loss": 0.8557,
+      "step": 6984
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5398428784746296,
+      "learning_rate": 4.267271370896387e-06,
+      "loss": 0.7467,
+      "step": 6985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4947655459650928,
+      "learning_rate": 4.265986169967731e-06,
+      "loss": 0.7929,
+      "step": 6986
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.628079580521067,
+      "learning_rate": 4.264701018601367e-06,
+      "loss": 0.7485,
+      "step": 6987
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.509337783990883,
+      "learning_rate": 4.263415916884071e-06,
+      "loss": 0.7077,
+      "step": 6988
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8153642249167848,
+      "learning_rate": 4.262130864902617e-06,
+      "loss": 1.1147,
+      "step": 6989
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5214981222882862,
+      "learning_rate": 4.260845862743775e-06,
+      "loss": 0.7776,
+      "step": 6990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5446307601604488,
+      "learning_rate": 4.2595609104943095e-06,
+      "loss": 0.77,
+      "step": 6991
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5728453655360544,
+      "learning_rate": 4.258276008240985e-06,
+      "loss": 0.7544,
+      "step": 6992
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.59908816769836,
+      "learning_rate": 4.256991156070561e-06,
+      "loss": 0.7655,
+      "step": 6993
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5783748937893796,
+      "learning_rate": 4.255706354069793e-06,
+      "loss": 0.7476,
+      "step": 6994
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.508834813556522,
+      "learning_rate": 4.2544216023254314e-06,
+      "loss": 0.7728,
+      "step": 6995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5237966803246454,
+      "learning_rate": 4.25313690092423e-06,
+      "loss": 0.7689,
+      "step": 6996
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.456342898373902,
+      "learning_rate": 4.251852249952934e-06,
+      "loss": 0.7446,
+      "step": 6997
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3549555836380007,
+      "learning_rate": 4.250567649498283e-06,
+      "loss": 0.695,
+      "step": 6998
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8136640184360785,
+      "learning_rate": 4.249283099647019e-06,
+      "loss": 1.0903,
+      "step": 6999
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4900308069928334,
+      "learning_rate": 4.2479986004858766e-06,
+      "loss": 0.7796,
+      "step": 7000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5922725822319352,
+      "learning_rate": 4.2467141521015855e-06,
+      "loss": 0.8098,
+      "step": 7001
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7609079542638947,
+      "learning_rate": 4.245429754580881e-06,
+      "loss": 1.0564,
+      "step": 7002
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7412150750411018,
+      "learning_rate": 4.244145408010484e-06,
+      "loss": 1.0917,
+      "step": 7003
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.498145633574176,
+      "learning_rate": 4.2428611124771184e-06,
+      "loss": 0.7407,
+      "step": 7004
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7657978138650661,
+      "learning_rate": 4.241576868067499e-06,
+      "loss": 1.0622,
+      "step": 7005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5851893718000987,
+      "learning_rate": 4.240292674868346e-06,
+      "loss": 0.8088,
+      "step": 7006
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5951092461965612,
+      "learning_rate": 4.2390085329663685e-06,
+      "loss": 0.6999,
+      "step": 7007
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4204506282037903,
+      "learning_rate": 4.237724442448273e-06,
+      "loss": 0.7327,
+      "step": 7008
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5050806888376942,
+      "learning_rate": 4.2364404034007685e-06,
+      "loss": 0.744,
+      "step": 7009
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.588904689268676,
+      "learning_rate": 4.235156415910553e-06,
+      "loss": 0.7144,
+      "step": 7010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6311589989690745,
+      "learning_rate": 4.233872480064326e-06,
+      "loss": 0.822,
+      "step": 7011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5129676403662289,
+      "learning_rate": 4.232588595948779e-06,
+      "loss": 0.7731,
+      "step": 7012
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6153881229756593,
+      "learning_rate": 4.231304763650607e-06,
+      "loss": 0.8254,
+      "step": 7013
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5326578790820926,
+      "learning_rate": 4.230020983256494e-06,
+      "loss": 0.7526,
+      "step": 7014
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6830427670101706,
+      "learning_rate": 4.2287372548531245e-06,
+      "loss": 0.7418,
+      "step": 7015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.859381105343199,
+      "learning_rate": 4.22745357852718e-06,
+      "loss": 1.0806,
+      "step": 7016
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5803470729949507,
+      "learning_rate": 4.226169954365337e-06,
+      "loss": 0.7951,
+      "step": 7017
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4768766708041745,
+      "learning_rate": 4.224886382454267e-06,
+      "loss": 0.6884,
+      "step": 7018
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5598983693446669,
+      "learning_rate": 4.22360286288064e-06,
+      "loss": 0.8311,
+      "step": 7019
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.557215804039948,
+      "learning_rate": 4.222319395731126e-06,
+      "loss": 0.8038,
+      "step": 7020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4260877791722688,
+      "learning_rate": 4.221035981092383e-06,
+      "loss": 0.6871,
+      "step": 7021
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4865103488792937,
+      "learning_rate": 4.21975261905107e-06,
+      "loss": 0.7186,
+      "step": 7022
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5847607138771493,
+      "learning_rate": 4.218469309693847e-06,
+      "loss": 0.7653,
+      "step": 7023
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4244366139360238,
+      "learning_rate": 4.2171860531073624e-06,
+      "loss": 0.73,
+      "step": 7024
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.62391589702249,
+      "learning_rate": 4.215902849378265e-06,
+      "loss": 0.841,
+      "step": 7025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5325421561933692,
+      "learning_rate": 4.2146196985932e-06,
+      "loss": 0.8372,
+      "step": 7026
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.628463471836452,
+      "learning_rate": 4.21333660083881e-06,
+      "loss": 0.8097,
+      "step": 7027
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4640618983708393,
+      "learning_rate": 4.2120535562017314e-06,
+      "loss": 0.7433,
+      "step": 7028
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.520796711312939,
+      "learning_rate": 4.210770564768597e-06,
+      "loss": 0.7102,
+      "step": 7029
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4002023455669288,
+      "learning_rate": 4.209487626626039e-06,
+      "loss": 0.6997,
+      "step": 7030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6954458248045787,
+      "learning_rate": 4.208204741860685e-06,
+      "loss": 0.8332,
+      "step": 7031
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4576887262480682,
+      "learning_rate": 4.2069219105591565e-06,
+      "loss": 0.7248,
+      "step": 7032
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6260693807469435,
+      "learning_rate": 4.205639132808076e-06,
+      "loss": 0.8395,
+      "step": 7033
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.790259541690348,
+      "learning_rate": 4.2043564086940565e-06,
+      "loss": 1.0823,
+      "step": 7034
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7708139741847582,
+      "learning_rate": 4.203073738303712e-06,
+      "loss": 1.079,
+      "step": 7035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5079869990972794,
+      "learning_rate": 4.2017911217236495e-06,
+      "loss": 0.7572,
+      "step": 7036
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4936906135003922,
+      "learning_rate": 4.200508559040477e-06,
+      "loss": 0.823,
+      "step": 7037
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.562509591020829,
+      "learning_rate": 4.199226050340795e-06,
+      "loss": 0.7383,
+      "step": 7038
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6457173716580216,
+      "learning_rate": 4.1979435957111984e-06,
+      "loss": 0.7311,
+      "step": 7039
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.498161336457439,
+      "learning_rate": 4.196661195238287e-06,
+      "loss": 0.8183,
+      "step": 7040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5579801595343497,
+      "learning_rate": 4.1953788490086486e-06,
+      "loss": 0.7756,
+      "step": 7041
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5379747325824085,
+      "learning_rate": 4.194096557108869e-06,
+      "loss": 0.7261,
+      "step": 7042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4153470707217075,
+      "learning_rate": 4.192814319625534e-06,
+      "loss": 0.7058,
+      "step": 7043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5605623246040383,
+      "learning_rate": 4.191532136645221e-06,
+      "loss": 0.7261,
+      "step": 7044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4855251932528117,
+      "learning_rate": 4.190250008254507e-06,
+      "loss": 0.7666,
+      "step": 7045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8656495892936655,
+      "learning_rate": 4.188967934539962e-06,
+      "loss": 1.0962,
+      "step": 7046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.521638648883849,
+      "learning_rate": 4.18768591558816e-06,
+      "loss": 0.7501,
+      "step": 7047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8015441036999291,
+      "learning_rate": 4.186403951485662e-06,
+      "loss": 1.0585,
+      "step": 7048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4427445827182785,
+      "learning_rate": 4.185122042319027e-06,
+      "loss": 0.6736,
+      "step": 7049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5024840170398595,
+      "learning_rate": 4.183840188174818e-06,
+      "loss": 0.8035,
+      "step": 7050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5366742961891835,
+      "learning_rate": 4.182558389139584e-06,
+      "loss": 0.7646,
+      "step": 7051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6096158506389084,
+      "learning_rate": 4.181276645299878e-06,
+      "loss": 0.7317,
+      "step": 7052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4626238991871474,
+      "learning_rate": 4.179994956742244e-06,
+      "loss": 0.7298,
+      "step": 7053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6837791660950916,
+      "learning_rate": 4.178713323553226e-06,
+      "loss": 0.7543,
+      "step": 7054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3996684354699744,
+      "learning_rate": 4.177431745819362e-06,
+      "loss": 0.7572,
+      "step": 7055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.429596213021206,
+      "learning_rate": 4.176150223627186e-06,
+      "loss": 0.7634,
+      "step": 7056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.504858365311278,
+      "learning_rate": 4.174868757063233e-06,
+      "loss": 0.7439,
+      "step": 7057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8982884163587705,
+      "learning_rate": 4.173587346214028e-06,
+      "loss": 1.0581,
+      "step": 7058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.448651302094281,
+      "learning_rate": 4.172305991166094e-06,
+      "loss": 0.7716,
+      "step": 7059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.392905060304066,
+      "learning_rate": 4.1710246920059535e-06,
+      "loss": 0.7871,
+      "step": 7060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4036449295265534,
+      "learning_rate": 4.169743448820121e-06,
+      "loss": 0.7267,
+      "step": 7061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5812031997751954,
+      "learning_rate": 4.16846226169511e-06,
+      "loss": 0.7543,
+      "step": 7062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4617419129280593,
+      "learning_rate": 4.1671811307174255e-06,
+      "loss": 0.8186,
+      "step": 7063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.71415105127411,
+      "learning_rate": 4.165900055973579e-06,
+      "loss": 0.7659,
+      "step": 7064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5515863318389964,
+      "learning_rate": 4.164619037550067e-06,
+      "loss": 0.7325,
+      "step": 7065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4722172655048211,
+      "learning_rate": 4.163338075533385e-06,
+      "loss": 0.7374,
+      "step": 7066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6225362486119685,
+      "learning_rate": 4.162057170010033e-06,
+      "loss": 0.7705,
+      "step": 7067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5303833819523325,
+      "learning_rate": 4.1607763210664955e-06,
+      "loss": 0.8069,
+      "step": 7068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4746485196179766,
+      "learning_rate": 4.15949552878926e-06,
+      "loss": 0.8053,
+      "step": 7069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6048993047749815,
+      "learning_rate": 4.158214793264808e-06,
+      "loss": 0.8076,
+      "step": 7070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8685964959809596,
+      "learning_rate": 4.1569341145796185e-06,
+      "loss": 0.8002,
+      "step": 7071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.704742785930983,
+      "learning_rate": 4.155653492820165e-06,
+      "loss": 0.7721,
+      "step": 7072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5604548516654537,
+      "learning_rate": 4.154372928072917e-06,
+      "loss": 0.797,
+      "step": 7073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5219394785559033,
+      "learning_rate": 4.153092420424344e-06,
+      "loss": 0.7394,
+      "step": 7074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.567205663730935,
+      "learning_rate": 4.151811969960908e-06,
+      "loss": 0.7167,
+      "step": 7075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.569289973905083,
+      "learning_rate": 4.1505315767690675e-06,
+      "loss": 0.7335,
+      "step": 7076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4887889751178849,
+      "learning_rate": 4.1492512409352755e-06,
+      "loss": 0.709,
+      "step": 7077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5513860166690452,
+      "learning_rate": 4.147970962545987e-06,
+      "loss": 0.8265,
+      "step": 7078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6011317947612889,
+      "learning_rate": 4.1466907416876466e-06,
+      "loss": 0.7545,
+      "step": 7079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5216050097762117,
+      "learning_rate": 4.145410578446697e-06,
+      "loss": 0.7224,
+      "step": 7080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8386451667464562,
+      "learning_rate": 4.1441304729095815e-06,
+      "loss": 1.0935,
+      "step": 7081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7934165042921365,
+      "learning_rate": 4.1428504251627335e-06,
+      "loss": 1.103,
+      "step": 7082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5135287846568213,
+      "learning_rate": 4.141570435292582e-06,
+      "loss": 0.8418,
+      "step": 7083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.564432601813079,
+      "learning_rate": 4.14029050338556e-06,
+      "loss": 0.7662,
+      "step": 7084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.387086604553629,
+      "learning_rate": 4.139010629528089e-06,
+      "loss": 0.6856,
+      "step": 7085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5262350938140619,
+      "learning_rate": 4.137730813806589e-06,
+      "loss": 0.756,
+      "step": 7086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7430656879929525,
+      "learning_rate": 4.136451056307475e-06,
+      "loss": 0.7107,
+      "step": 7087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4981729085503597,
+      "learning_rate": 4.135171357117162e-06,
+      "loss": 0.7791,
+      "step": 7088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4365290621828308,
+      "learning_rate": 4.133891716322056e-06,
+      "loss": 0.7624,
+      "step": 7089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.526314980811003,
+      "learning_rate": 4.1326121340085595e-06,
+      "loss": 0.713,
+      "step": 7090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.494077388323042,
+      "learning_rate": 4.1313326102630775e-06,
+      "loss": 0.7278,
+      "step": 7091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.468433708559831,
+      "learning_rate": 4.130053145172005e-06,
+      "loss": 0.7733,
+      "step": 7092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6169622928077265,
+      "learning_rate": 4.1287737388217325e-06,
+      "loss": 0.718,
+      "step": 7093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8358845661212813,
+      "learning_rate": 4.127494391298647e-06,
+      "loss": 1.1133,
+      "step": 7094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.547959506487438,
+      "learning_rate": 4.126215102689137e-06,
+      "loss": 0.7541,
+      "step": 7095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5624047570482738,
+      "learning_rate": 4.124935873079582e-06,
+      "loss": 0.8297,
+      "step": 7096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8074977996225164,
+      "learning_rate": 4.1236567025563565e-06,
+      "loss": 1.0849,
+      "step": 7097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6408462949691376,
+      "learning_rate": 4.122377591205835e-06,
+      "loss": 0.8198,
+      "step": 7098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5271300867825561,
+      "learning_rate": 4.121098539114387e-06,
+      "loss": 0.7557,
+      "step": 7099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6176042525155987,
+      "learning_rate": 4.1198195463683716e-06,
+      "loss": 0.6571,
+      "step": 7100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5919570317737428,
+      "learning_rate": 4.118540613054155e-06,
+      "loss": 0.7298,
+      "step": 7101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5860673838245258,
+      "learning_rate": 4.117261739258092e-06,
+      "loss": 0.7613,
+      "step": 7102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7720359422638717,
+      "learning_rate": 4.115982925066536e-06,
+      "loss": 1.0755,
+      "step": 7103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6631120422585277,
+      "learning_rate": 4.114704170565833e-06,
+      "loss": 0.7826,
+      "step": 7104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.543269038483739,
+      "learning_rate": 4.113425475842329e-06,
+      "loss": 0.6782,
+      "step": 7105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5918814080879675,
+      "learning_rate": 4.112146840982365e-06,
+      "loss": 0.7164,
+      "step": 7106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6396402610293275,
+      "learning_rate": 4.110868266072273e-06,
+      "loss": 0.8204,
+      "step": 7107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6092472191273226,
+      "learning_rate": 4.109589751198393e-06,
+      "loss": 0.7221,
+      "step": 7108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.388014418878276,
+      "learning_rate": 4.108311296447048e-06,
+      "loss": 0.7559,
+      "step": 7109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5446557287113962,
+      "learning_rate": 4.107032901904564e-06,
+      "loss": 0.7103,
+      "step": 7110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8279707242164354,
+      "learning_rate": 4.105754567657257e-06,
+      "loss": 1.0756,
+      "step": 7111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4002401523237993,
+      "learning_rate": 4.104476293791449e-06,
+      "loss": 0.73,
+      "step": 7112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4691121826274531,
+      "learning_rate": 4.103198080393449e-06,
+      "loss": 0.8337,
+      "step": 7113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.754328160845997,
+      "learning_rate": 4.101919927549564e-06,
+      "loss": 1.0551,
+      "step": 7114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8024385858466563,
+      "learning_rate": 4.100641835346101e-06,
+      "loss": 1.0801,
+      "step": 7115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4729572008807656,
+      "learning_rate": 4.0993638038693575e-06,
+      "loss": 0.7158,
+      "step": 7116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5147121738263984,
+      "learning_rate": 4.098085833205629e-06,
+      "loss": 0.8078,
+      "step": 7117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5947104671273853,
+      "learning_rate": 4.0968079234412054e-06,
+      "loss": 0.7671,
+      "step": 7118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3856971064534735,
+      "learning_rate": 4.0955300746623785e-06,
+      "loss": 0.7249,
+      "step": 7119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5281177458021948,
+      "learning_rate": 4.094252286955429e-06,
+      "loss": 0.7894,
+      "step": 7120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4479367430242662,
+      "learning_rate": 4.092974560406635e-06,
+      "loss": 0.7044,
+      "step": 7121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.54243591201938,
+      "learning_rate": 4.091696895102274e-06,
+      "loss": 0.7868,
+      "step": 7122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.540491417062718,
+      "learning_rate": 4.090419291128616e-06,
+      "loss": 0.7339,
+      "step": 7123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.574562157142487,
+      "learning_rate": 4.089141748571926e-06,
+      "loss": 0.7537,
+      "step": 7124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4927405041267854,
+      "learning_rate": 4.0878642675184675e-06,
+      "loss": 0.7024,
+      "step": 7125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0882468373744474,
+      "learning_rate": 4.086586848054501e-06,
+      "loss": 0.7429,
+      "step": 7126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6335344463837769,
+      "learning_rate": 4.085309490266278e-06,
+      "loss": 0.8015,
+      "step": 7127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5072018266943266,
+      "learning_rate": 4.08403219424005e-06,
+      "loss": 0.7725,
+      "step": 7128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6210881311302356,
+      "learning_rate": 4.082754960062062e-06,
+      "loss": 0.8355,
+      "step": 7129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4632537188742267,
+      "learning_rate": 4.081477787818559e-06,
+      "loss": 0.7874,
+      "step": 7130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4712006850891277,
+      "learning_rate": 4.0802006775957735e-06,
+      "loss": 0.7807,
+      "step": 7131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.468836921041486,
+      "learning_rate": 4.0789236294799425e-06,
+      "loss": 0.7548,
+      "step": 7132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5308232808269413,
+      "learning_rate": 4.077646643557295e-06,
+      "loss": 0.8063,
+      "step": 7133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5666626068333973,
+      "learning_rate": 4.076369719914055e-06,
+      "loss": 0.7795,
+      "step": 7134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5377358265677132,
+      "learning_rate": 4.075092858636441e-06,
+      "loss": 0.8133,
+      "step": 7135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4571039319355938,
+      "learning_rate": 4.073816059810675e-06,
+      "loss": 0.7501,
+      "step": 7136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5473773043761014,
+      "learning_rate": 4.072539323522967e-06,
+      "loss": 0.8034,
+      "step": 7137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4905016995132188,
+      "learning_rate": 4.0712626498595206e-06,
+      "loss": 0.7857,
+      "step": 7138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4818524041242678,
+      "learning_rate": 4.069986038906547e-06,
+      "loss": 0.7469,
+      "step": 7139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4368787817807676,
+      "learning_rate": 4.0687094907502425e-06,
+      "loss": 0.7852,
+      "step": 7140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5433060516578183,
+      "learning_rate": 4.067433005476802e-06,
+      "loss": 0.7259,
+      "step": 7141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3948417656051415,
+      "learning_rate": 4.0661565831724185e-06,
+      "loss": 0.7725,
+      "step": 7142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.467304525198999,
+      "learning_rate": 4.064880223923277e-06,
+      "loss": 0.7411,
+      "step": 7143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7018906813780617,
+      "learning_rate": 4.063603927815561e-06,
+      "loss": 0.8394,
+      "step": 7144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.488822441960508,
+      "learning_rate": 4.062327694935448e-06,
+      "loss": 0.7173,
+      "step": 7145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8228050794284565,
+      "learning_rate": 4.061051525369114e-06,
+      "loss": 0.7253,
+      "step": 7146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.540225282192984,
+      "learning_rate": 4.059775419202729e-06,
+      "loss": 0.7666,
+      "step": 7147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3949528797560709,
+      "learning_rate": 4.058499376522456e-06,
+      "loss": 0.7436,
+      "step": 7148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5317956426315347,
+      "learning_rate": 4.05722339741446e-06,
+      "loss": 0.7675,
+      "step": 7149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4902789622961525,
+      "learning_rate": 4.055947481964895e-06,
+      "loss": 0.7456,
+      "step": 7150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4922505916641091,
+      "learning_rate": 4.0546716302599156e-06,
+      "loss": 0.7847,
+      "step": 7151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9157573078400717,
+      "learning_rate": 4.053395842385668e-06,
+      "loss": 1.0402,
+      "step": 7152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8669598843367065,
+      "learning_rate": 4.052120118428298e-06,
+      "loss": 1.0821,
+      "step": 7153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.496477479700937,
+      "learning_rate": 4.050844458473945e-06,
+      "loss": 0.8051,
+      "step": 7154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.572690363581884,
+      "learning_rate": 4.049568862608743e-06,
+      "loss": 0.8392,
+      "step": 7155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7810375592555158,
+      "learning_rate": 4.048293330918827e-06,
+      "loss": 1.0474,
+      "step": 7156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.49694165920929,
+      "learning_rate": 4.047017863490322e-06,
+      "loss": 0.7374,
+      "step": 7157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5580336522446159,
+      "learning_rate": 4.045742460409348e-06,
+      "loss": 0.8219,
+      "step": 7158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9045209913316393,
+      "learning_rate": 4.044467121762026e-06,
+      "loss": 1.0623,
+      "step": 7159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6309969299574245,
+      "learning_rate": 4.043191847634469e-06,
+      "loss": 0.8006,
+      "step": 7160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4166723294111496,
+      "learning_rate": 4.0419166381127865e-06,
+      "loss": 0.6948,
+      "step": 7161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7649859680725803,
+      "learning_rate": 4.040641493283081e-06,
+      "loss": 0.7796,
+      "step": 7162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5247404594546232,
+      "learning_rate": 4.039366413231458e-06,
+      "loss": 0.8558,
+      "step": 7163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3311251339927925,
+      "learning_rate": 4.038091398044012e-06,
+      "loss": 0.6425,
+      "step": 7164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.463494458220714,
+      "learning_rate": 4.036816447806832e-06,
+      "loss": 0.7241,
+      "step": 7165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4064974063687137,
+      "learning_rate": 4.03554156260601e-06,
+      "loss": 0.7029,
+      "step": 7166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.64264977071479,
+      "learning_rate": 4.0342667425276265e-06,
+      "loss": 0.7495,
+      "step": 7167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8355448835502453,
+      "learning_rate": 4.032991987657762e-06,
+      "loss": 1.0702,
+      "step": 7168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4976396585244225,
+      "learning_rate": 4.031717298082487e-06,
+      "loss": 0.7245,
+      "step": 7169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.522291765206997,
+      "learning_rate": 4.030442673887876e-06,
+      "loss": 0.8049,
+      "step": 7170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8028540636940037,
+      "learning_rate": 4.029168115159993e-06,
+      "loss": 1.0519,
+      "step": 7171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7799469344630628,
+      "learning_rate": 4.027893621984896e-06,
+      "loss": 1.0702,
+      "step": 7172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5155806798282636,
+      "learning_rate": 4.026619194448647e-06,
+      "loss": 0.7645,
+      "step": 7173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7918685161233011,
+      "learning_rate": 4.025344832637295e-06,
+      "loss": 1.0695,
+      "step": 7174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7875785102457681,
+      "learning_rate": 4.024070536636889e-06,
+      "loss": 1.0672,
+      "step": 7175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5275750469430311,
+      "learning_rate": 4.022796306533472e-06,
+      "loss": 0.7003,
+      "step": 7176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7011304514495424,
+      "learning_rate": 4.021522142413082e-06,
+      "loss": 0.8243,
+      "step": 7177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5928614131373477,
+      "learning_rate": 4.020248044361756e-06,
+      "loss": 0.7684,
+      "step": 7178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.640593197921435,
+      "learning_rate": 4.018974012465519e-06,
+      "loss": 0.8355,
+      "step": 7179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5924582890077175,
+      "learning_rate": 4.017700046810403e-06,
+      "loss": 0.6925,
+      "step": 7180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6120998079711137,
+      "learning_rate": 4.016426147482427e-06,
+      "loss": 0.6828,
+      "step": 7181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.491434826315475,
+      "learning_rate": 4.015152314567603e-06,
+      "loss": 0.8034,
+      "step": 7182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4402109082201846,
+      "learning_rate": 4.01387854815195e-06,
+      "loss": 0.7823,
+      "step": 7183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8672487359496766,
+      "learning_rate": 4.012604848321471e-06,
+      "loss": 1.0685,
+      "step": 7184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4447653310971225,
+      "learning_rate": 4.011331215162171e-06,
+      "loss": 0.7733,
+      "step": 7185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4886240948485858,
+      "learning_rate": 4.0100576487600465e-06,
+      "loss": 0.7523,
+      "step": 7186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8081762635676816,
+      "learning_rate": 4.0087841492010946e-06,
+      "loss": 1.0758,
+      "step": 7187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7902171925632017,
+      "learning_rate": 4.007510716571304e-06,
+      "loss": 1.0607,
+      "step": 7188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5008392014435952,
+      "learning_rate": 4.006237350956657e-06,
+      "loss": 0.7271,
+      "step": 7189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.56291557809456,
+      "learning_rate": 4.004964052443137e-06,
+      "loss": 0.7348,
+      "step": 7190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5482025357115359,
+      "learning_rate": 4.003690821116721e-06,
+      "loss": 0.8193,
+      "step": 7191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4489551284704332,
+      "learning_rate": 4.002417657063379e-06,
+      "loss": 0.701,
+      "step": 7192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8850259789319137,
+      "learning_rate": 4.001144560369077e-06,
+      "loss": 1.11,
+      "step": 7193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.562665657748395,
+      "learning_rate": 3.999871531119779e-06,
+      "loss": 0.7592,
+      "step": 7194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4764821556510002,
+      "learning_rate": 3.9985985694014414e-06,
+      "loss": 0.7855,
+      "step": 7195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.42402710793973,
+      "learning_rate": 3.997325675300018e-06,
+      "loss": 0.7294,
+      "step": 7196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4747840896582316,
+      "learning_rate": 3.996052848901459e-06,
+      "loss": 0.7424,
+      "step": 7197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6050726523879517,
+      "learning_rate": 3.994780090291707e-06,
+      "loss": 0.9039,
+      "step": 7198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.481324804103119,
+      "learning_rate": 3.993507399556699e-06,
+      "loss": 0.7522,
+      "step": 7199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.489211148366926,
+      "learning_rate": 3.992234776782376e-06,
+      "loss": 0.7794,
+      "step": 7200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5084992866168918,
+      "learning_rate": 3.990962222054665e-06,
+      "loss": 0.7749,
+      "step": 7201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7745093260620298,
+      "learning_rate": 3.989689735459492e-06,
+      "loss": 1.0571,
+      "step": 7202
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4285179785574975,
+      "learning_rate": 3.988417317082777e-06,
+      "loss": 0.7757,
+      "step": 7203
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7672536960448711,
+      "learning_rate": 3.987144967010439e-06,
+      "loss": 1.0768,
+      "step": 7204
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4535208576265604,
+      "learning_rate": 3.985872685328389e-06,
+      "loss": 0.8318,
+      "step": 7205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6911978542740556,
+      "learning_rate": 3.984600472122533e-06,
+      "loss": 0.8724,
+      "step": 7206
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5166547046459407,
+      "learning_rate": 3.983328327478776e-06,
+      "loss": 0.7843,
+      "step": 7207
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.774598094722898,
+      "learning_rate": 3.982056251483016e-06,
+      "loss": 1.0544,
+      "step": 7208
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7680335511216643,
+      "learning_rate": 3.980784244221145e-06,
+      "loss": 1.0984,
+      "step": 7209
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5189397268679563,
+      "learning_rate": 3.97951230577905e-06,
+      "loss": 0.8037,
+      "step": 7210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5115252510365207,
+      "learning_rate": 3.97824043624262e-06,
+      "loss": 0.7727,
+      "step": 7211
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4637110269781521,
+      "learning_rate": 3.976968635697732e-06,
+      "loss": 0.7508,
+      "step": 7212
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7696970144061797,
+      "learning_rate": 3.9756969042302605e-06,
+      "loss": 1.0681,
+      "step": 7213
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3377641952242485,
+      "learning_rate": 3.974425241926076e-06,
+      "loss": 0.73,
+      "step": 7214
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7651638855504689,
+      "learning_rate": 3.973153648871045e-06,
+      "loss": 1.0534,
+      "step": 7215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5942248006454918,
+      "learning_rate": 3.971882125151028e-06,
+      "loss": 0.6832,
+      "step": 7216
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5474020869809793,
+      "learning_rate": 3.9706106708518785e-06,
+      "loss": 0.7531,
+      "step": 7217
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7448428516691636,
+      "learning_rate": 3.969339286059452e-06,
+      "loss": 0.8421,
+      "step": 7218
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7638557319296726,
+      "learning_rate": 3.968067970859595e-06,
+      "loss": 1.0656,
+      "step": 7219
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6102123518751517,
+      "learning_rate": 3.9667967253381455e-06,
+      "loss": 0.7031,
+      "step": 7220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4903400877596606,
+      "learning_rate": 3.965525549580946e-06,
+      "loss": 0.7343,
+      "step": 7221
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.500826808133186,
+      "learning_rate": 3.964254443673826e-06,
+      "loss": 0.8153,
+      "step": 7222
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.521775024667985,
+      "learning_rate": 3.962983407702613e-06,
+      "loss": 0.8091,
+      "step": 7223
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7706669857353786,
+      "learning_rate": 3.961712441753134e-06,
+      "loss": 1.1003,
+      "step": 7224
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.612001896996319,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.8214,
+      "step": 7225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.465306544644467,
+      "learning_rate": 3.959170720262639e-06,
+      "loss": 0.6893,
+      "step": 7226
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.629457248808697,
+      "learning_rate": 3.957899964893245e-06,
+      "loss": 0.7579,
+      "step": 7227
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.606871026288243,
+      "learning_rate": 3.956629279888829e-06,
+      "loss": 0.8282,
+      "step": 7228
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5834350249503308,
+      "learning_rate": 3.95535866533519e-06,
+      "loss": 0.8092,
+      "step": 7229
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6944650576956282,
+      "learning_rate": 3.954088121318122e-06,
+      "loss": 0.7154,
+      "step": 7230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.909970700743625,
+      "learning_rate": 3.952817647923417e-06,
+      "loss": 0.7503,
+      "step": 7231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4593757728239447,
+      "learning_rate": 3.951547245236859e-06,
+      "loss": 0.8069,
+      "step": 7232
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5802182234298403,
+      "learning_rate": 3.950276913344228e-06,
+      "loss": 0.7361,
+      "step": 7233
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7900477374880661,
+      "learning_rate": 3.949006652331297e-06,
+      "loss": 1.0528,
+      "step": 7234
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.611913455115591,
+      "learning_rate": 3.947736462283844e-06,
+      "loss": 0.7637,
+      "step": 7235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5990893846695369,
+      "learning_rate": 3.94646634328763e-06,
+      "loss": 0.7433,
+      "step": 7236
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6338897741126617,
+      "learning_rate": 3.945196295428417e-06,
+      "loss": 0.7386,
+      "step": 7237
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4570129378017322,
+      "learning_rate": 3.9439263187919635e-06,
+      "loss": 0.8337,
+      "step": 7238
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7819692770130315,
+      "learning_rate": 3.94265641346402e-06,
+      "loss": 1.0865,
+      "step": 7239
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4714070156452312,
+      "learning_rate": 3.941386579530331e-06,
+      "loss": 0.7898,
+      "step": 7240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5076082529055248,
+      "learning_rate": 3.940116817076643e-06,
+      "loss": 0.8176,
+      "step": 7241
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4608453612371899,
+      "learning_rate": 3.9388471261886905e-06,
+      "loss": 0.8235,
+      "step": 7242
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4415209308338504,
+      "learning_rate": 3.937577506952206e-06,
+      "loss": 0.7928,
+      "step": 7243
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.57786500990223,
+      "learning_rate": 3.936307959452917e-06,
+      "loss": 0.7192,
+      "step": 7244
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5064447392020524,
+      "learning_rate": 3.9350384837765475e-06,
+      "loss": 0.7984,
+      "step": 7245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4603479484311506,
+      "learning_rate": 3.933769080008816e-06,
+      "loss": 0.7732,
+      "step": 7246
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4705219740445552,
+      "learning_rate": 3.932499748235432e-06,
+      "loss": 0.662,
+      "step": 7247
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6423931859978949,
+      "learning_rate": 3.931230488542107e-06,
+      "loss": 0.8321,
+      "step": 7248
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.596457386454085,
+      "learning_rate": 3.929961301014544e-06,
+      "loss": 0.8058,
+      "step": 7249
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.811745916970097,
+      "learning_rate": 3.928692185738442e-06,
+      "loss": 1.0619,
+      "step": 7250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4991928120466487,
+      "learning_rate": 3.927423142799489e-06,
+      "loss": 0.8787,
+      "step": 7251
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8357500037904331,
+      "learning_rate": 3.926154172283382e-06,
+      "loss": 1.0782,
+      "step": 7252
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6684433144819075,
+      "learning_rate": 3.9248852742758e-06,
+      "loss": 0.9182,
+      "step": 7253
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5029255778264479,
+      "learning_rate": 3.92361644886242e-06,
+      "loss": 0.7524,
+      "step": 7254
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4371308284618662,
+      "learning_rate": 3.922347696128922e-06,
+      "loss": 0.8178,
+      "step": 7255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4066092504500654,
+      "learning_rate": 3.92107901616097e-06,
+      "loss": 0.7023,
+      "step": 7256
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.430323907540675,
+      "learning_rate": 3.9198104090442305e-06,
+      "loss": 0.7535,
+      "step": 7257
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7953937174839691,
+      "learning_rate": 3.918541874864362e-06,
+      "loss": 1.127,
+      "step": 7258
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4827601836319217,
+      "learning_rate": 3.91727341370702e-06,
+      "loss": 0.7348,
+      "step": 7259
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.489728992173413,
+      "learning_rate": 3.916005025657852e-06,
+      "loss": 0.8275,
+      "step": 7260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4559106202284222,
+      "learning_rate": 3.914736710802501e-06,
+      "loss": 0.6996,
+      "step": 7261
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4018381218421112,
+      "learning_rate": 3.913468469226612e-06,
+      "loss": 0.6509,
+      "step": 7262
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5657357977665278,
+      "learning_rate": 3.912200301015816e-06,
+      "loss": 0.8539,
+      "step": 7263
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.527746666623484,
+      "learning_rate": 3.910932206255742e-06,
+      "loss": 0.7652,
+      "step": 7264
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5151709967475193,
+      "learning_rate": 3.909664185032017e-06,
+      "loss": 0.7109,
+      "step": 7265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4422402938074113,
+      "learning_rate": 3.90839623743026e-06,
+      "loss": 0.734,
+      "step": 7266
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4691318886615006,
+      "learning_rate": 3.907128363536084e-06,
+      "loss": 0.7702,
+      "step": 7267
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.466381231199717,
+      "learning_rate": 3.905860563435099e-06,
+      "loss": 0.6965,
+      "step": 7268
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5225750236345617,
+      "learning_rate": 3.904592837212913e-06,
+      "loss": 0.7092,
+      "step": 7269
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6807187248985447,
+      "learning_rate": 3.903325184955122e-06,
+      "loss": 0.6976,
+      "step": 7270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5298721728670528,
+      "learning_rate": 3.902057606747321e-06,
+      "loss": 0.7754,
+      "step": 7271
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3847728691913903,
+      "learning_rate": 3.900790102675103e-06,
+      "loss": 0.6332,
+      "step": 7272
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.60183246828662,
+      "learning_rate": 3.89952267282405e-06,
+      "loss": 0.7818,
+      "step": 7273
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4668105011430197,
+      "learning_rate": 3.898255317279744e-06,
+      "loss": 0.6872,
+      "step": 7274
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7112537134267514,
+      "learning_rate": 3.896988036127755e-06,
+      "loss": 0.8407,
+      "step": 7275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5342072574995707,
+      "learning_rate": 3.895720829453659e-06,
+      "loss": 0.7808,
+      "step": 7276
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6429465386039408,
+      "learning_rate": 3.894453697343016e-06,
+      "loss": 0.8413,
+      "step": 7277
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7659567168942542,
+      "learning_rate": 3.893186639881387e-06,
+      "loss": 1.0209,
+      "step": 7278
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5960865966893847,
+      "learning_rate": 3.891919657154328e-06,
+      "loss": 0.745,
+      "step": 7279
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7859465767957946,
+      "learning_rate": 3.890652749247388e-06,
+      "loss": 1.0599,
+      "step": 7280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4617661998974252,
+      "learning_rate": 3.889385916246109e-06,
+      "loss": 0.6816,
+      "step": 7281
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7882271712682697,
+      "learning_rate": 3.8881191582360345e-06,
+      "loss": 1.0704,
+      "step": 7282
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9900396175960753,
+      "learning_rate": 3.886852475302697e-06,
+      "loss": 0.786,
+      "step": 7283
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6297709174841188,
+      "learning_rate": 3.885585867531625e-06,
+      "loss": 0.7931,
+      "step": 7284
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5818358078510768,
+      "learning_rate": 3.884319335008343e-06,
+      "loss": 0.7772,
+      "step": 7285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6013702708466255,
+      "learning_rate": 3.883052877818372e-06,
+      "loss": 0.6957,
+      "step": 7286
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6107152889100762,
+      "learning_rate": 3.881786496047224e-06,
+      "loss": 0.7863,
+      "step": 7287
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5081901825096355,
+      "learning_rate": 3.880520189780407e-06,
+      "loss": 0.8017,
+      "step": 7288
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4333904460399978,
+      "learning_rate": 3.879253959103429e-06,
+      "loss": 0.739,
+      "step": 7289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.524853359880095,
+      "learning_rate": 3.877987804101786e-06,
+      "loss": 0.7132,
+      "step": 7290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5255864929187066,
+      "learning_rate": 3.876721724860973e-06,
+      "loss": 0.8256,
+      "step": 7291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4809916081614716,
+      "learning_rate": 3.875455721466475e-06,
+      "loss": 0.7967,
+      "step": 7292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5063175408707346,
+      "learning_rate": 3.87418979400378e-06,
+      "loss": 0.7666,
+      "step": 7293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3947333158682247,
+      "learning_rate": 3.872923942558365e-06,
+      "loss": 0.7267,
+      "step": 7294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5420871675481762,
+      "learning_rate": 3.8716581672157e-06,
+      "loss": 0.7522,
+      "step": 7295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1644049950088515,
+      "learning_rate": 3.870392468061257e-06,
+      "loss": 0.6944,
+      "step": 7296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5208831919183567,
+      "learning_rate": 3.8691268451805e-06,
+      "loss": 0.7724,
+      "step": 7297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4628982276672735,
+      "learning_rate": 3.86786129865888e-06,
+      "loss": 0.7781,
+      "step": 7298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5369249098757705,
+      "learning_rate": 3.866595828581856e-06,
+      "loss": 0.819,
+      "step": 7299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4691486200291137,
+      "learning_rate": 3.8653304350348745e-06,
+      "loss": 0.8088,
+      "step": 7300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.391961157299506,
+      "learning_rate": 3.864065118103376e-06,
+      "loss": 0.7804,
+      "step": 7301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5505624560736497,
+      "learning_rate": 3.862799877872796e-06,
+      "loss": 0.7596,
+      "step": 7302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.551283146732892,
+      "learning_rate": 3.861534714428571e-06,
+      "loss": 0.7651,
+      "step": 7303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8823664557875566,
+      "learning_rate": 3.860269627856126e-06,
+      "loss": 0.7587,
+      "step": 7304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7020252179726705,
+      "learning_rate": 3.859004618240879e-06,
+      "loss": 0.7152,
+      "step": 7305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.409754639235626,
+      "learning_rate": 3.85773968566825e-06,
+      "loss": 0.6968,
+      "step": 7306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4692027053334245,
+      "learning_rate": 3.856474830223651e-06,
+      "loss": 0.6956,
+      "step": 7307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.618271946960897,
+      "learning_rate": 3.855210051992486e-06,
+      "loss": 0.7458,
+      "step": 7308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.555962867037443,
+      "learning_rate": 3.853945351060155e-06,
+      "loss": 0.7985,
+      "step": 7309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6269011833685136,
+      "learning_rate": 3.852680727512056e-06,
+      "loss": 0.8567,
+      "step": 7310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6498263761361962,
+      "learning_rate": 3.851416181433576e-06,
+      "loss": 0.7745,
+      "step": 7311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4905189746794514,
+      "learning_rate": 3.8501517129101015e-06,
+      "loss": 0.7827,
+      "step": 7312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4469199766504233,
+      "learning_rate": 3.848887322027015e-06,
+      "loss": 0.6875,
+      "step": 7313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3758673718646328,
+      "learning_rate": 3.8476230088696875e-06,
+      "loss": 0.6676,
+      "step": 7314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.590790565075438,
+      "learning_rate": 3.846358773523488e-06,
+      "loss": 0.7632,
+      "step": 7315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5841692263929703,
+      "learning_rate": 3.845094616073783e-06,
+      "loss": 0.8482,
+      "step": 7316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6164802624724364,
+      "learning_rate": 3.843830536605932e-06,
+      "loss": 0.8091,
+      "step": 7317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5397880433194941,
+      "learning_rate": 3.842566535205286e-06,
+      "loss": 0.7441,
+      "step": 7318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4153542080469386,
+      "learning_rate": 3.841302611957193e-06,
+      "loss": 0.677,
+      "step": 7319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5292731818864507,
+      "learning_rate": 3.840038766946999e-06,
+      "loss": 0.7189,
+      "step": 7320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6504548822632614,
+      "learning_rate": 3.8387750002600395e-06,
+      "loss": 0.7586,
+      "step": 7321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.832586143342959,
+      "learning_rate": 3.8375113119816444e-06,
+      "loss": 1.0582,
+      "step": 7322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8253660075751077,
+      "learning_rate": 3.836247702197146e-06,
+      "loss": 1.0879,
+      "step": 7323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4641661586255243,
+      "learning_rate": 3.834984170991865e-06,
+      "loss": 0.7152,
+      "step": 7324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5009724049939883,
+      "learning_rate": 3.833720718451116e-06,
+      "loss": 0.6923,
+      "step": 7325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5183310870000626,
+      "learning_rate": 3.83245734466021e-06,
+      "loss": 0.7339,
+      "step": 7326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7832400776252512,
+      "learning_rate": 3.831194049704455e-06,
+      "loss": 1.1181,
+      "step": 7327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5619617677336655,
+      "learning_rate": 3.82993083366915e-06,
+      "loss": 0.7442,
+      "step": 7328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6059698170557184,
+      "learning_rate": 3.8286676966395895e-06,
+      "loss": 0.7455,
+      "step": 7329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5173039119208334,
+      "learning_rate": 3.827404638701066e-06,
+      "loss": 0.792,
+      "step": 7330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5130347418810082,
+      "learning_rate": 3.8261416599388625e-06,
+      "loss": 0.8407,
+      "step": 7331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4966770129808409,
+      "learning_rate": 3.824878760438259e-06,
+      "loss": 0.6693,
+      "step": 7332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7907774121554396,
+      "learning_rate": 3.823615940284525e-06,
+      "loss": 1.0254,
+      "step": 7333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4614747281160294,
+      "learning_rate": 3.822353199562936e-06,
+      "loss": 0.7143,
+      "step": 7334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8016914734912023,
+      "learning_rate": 3.821090538358751e-06,
+      "loss": 1.0544,
+      "step": 7335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5509233814899828,
+      "learning_rate": 3.819827956757228e-06,
+      "loss": 0.6739,
+      "step": 7336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5738399458052668,
+      "learning_rate": 3.8185654548436215e-06,
+      "loss": 0.8006,
+      "step": 7337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5121767171645342,
+      "learning_rate": 3.817303032703176e-06,
+      "loss": 0.808,
+      "step": 7338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.556457245322098,
+      "learning_rate": 3.8160406904211325e-06,
+      "loss": 0.7896,
+      "step": 7339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4690313190232334,
+      "learning_rate": 3.814778428082732e-06,
+      "loss": 0.8441,
+      "step": 7340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4812244762413158,
+      "learning_rate": 3.8135162457732017e-06,
+      "loss": 0.784,
+      "step": 7341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4653686194351543,
+      "learning_rate": 3.812254143577767e-06,
+      "loss": 0.6927,
+      "step": 7342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8059576722181655,
+      "learning_rate": 3.8109921215816466e-06,
+      "loss": 1.1038,
+      "step": 7343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4475618268190038,
+      "learning_rate": 3.8097301798700587e-06,
+      "loss": 0.8229,
+      "step": 7344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4763860867286513,
+      "learning_rate": 3.808468318528211e-06,
+      "loss": 0.7147,
+      "step": 7345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5561827068760734,
+      "learning_rate": 3.807206537641306e-06,
+      "loss": 0.8029,
+      "step": 7346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4858360555654384,
+      "learning_rate": 3.805944837294544e-06,
+      "loss": 0.7675,
+      "step": 7347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4600143183590961,
+      "learning_rate": 3.8046832175731175e-06,
+      "loss": 0.7755,
+      "step": 7348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5053323533686256,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.7511,
+      "step": 7349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7817044963104299,
+      "learning_rate": 3.8021602203470102e-06,
+      "loss": 1.0917,
+      "step": 7350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.473930409870113,
+      "learning_rate": 3.8008988430126916e-06,
+      "loss": 0.7964,
+      "step": 7351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5882573601170922,
+      "learning_rate": 3.799637546644424e-06,
+      "loss": 0.7583,
+      "step": 7352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.552596843494377,
+      "learning_rate": 3.7983763313273737e-06,
+      "loss": 0.8129,
+      "step": 7353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7893785600951179,
+      "learning_rate": 3.797115197146702e-06,
+      "loss": 1.0598,
+      "step": 7354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4336710165606827,
+      "learning_rate": 3.7958541441875628e-06,
+      "loss": 0.7283,
+      "step": 7355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5507858915597676,
+      "learning_rate": 3.7945931725351028e-06,
+      "loss": 0.7937,
+      "step": 7356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7599082928879205,
+      "learning_rate": 3.793332282274472e-06,
+      "loss": 1.0721,
+      "step": 7357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5298738186014904,
+      "learning_rate": 3.7920714734908025e-06,
+      "loss": 0.8233,
+      "step": 7358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.525855890796849,
+      "learning_rate": 3.7908107462692303e-06,
+      "loss": 0.819,
+      "step": 7359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6410259680649963,
+      "learning_rate": 3.7895501006948787e-06,
+      "loss": 0.8454,
+      "step": 7360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6146231268004767,
+      "learning_rate": 3.788289536852875e-06,
+      "loss": 0.8079,
+      "step": 7361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5296447053937747,
+      "learning_rate": 3.787029054828332e-06,
+      "loss": 0.714,
+      "step": 7362
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8332501096478866,
+      "learning_rate": 3.78576865470636e-06,
+      "loss": 0.7782,
+      "step": 7363
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6043405674147053,
+      "learning_rate": 3.784508336572066e-06,
+      "loss": 0.7574,
+      "step": 7364
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8093002599400496,
+      "learning_rate": 3.7832481005105483e-06,
+      "loss": 1.0902,
+      "step": 7365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.58074598069486,
+      "learning_rate": 3.781987946606901e-06,
+      "loss": 0.7326,
+      "step": 7366
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.535106978721027,
+      "learning_rate": 3.7807278749462105e-06,
+      "loss": 0.8655,
+      "step": 7367
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.453255371436978,
+      "learning_rate": 3.7794678856135647e-06,
+      "loss": 0.6712,
+      "step": 7368
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.800621606898038,
+      "learning_rate": 3.7782079786940372e-06,
+      "loss": 1.0949,
+      "step": 7369
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.489905992777797,
+      "learning_rate": 3.7769481542727003e-06,
+      "loss": 0.741,
+      "step": 7370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4852413468690266,
+      "learning_rate": 3.775688412434622e-06,
+      "loss": 0.7107,
+      "step": 7371
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7973512280662108,
+      "learning_rate": 3.7744287532648615e-06,
+      "loss": 1.0725,
+      "step": 7372
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.514581351722614,
+      "learning_rate": 3.773169176848474e-06,
+      "loss": 0.719,
+      "step": 7373
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.507294900990096,
+      "learning_rate": 3.7719096832705075e-06,
+      "loss": 0.6512,
+      "step": 7374
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.539479914073825,
+      "learning_rate": 3.7706502726160087e-06,
+      "loss": 0.7643,
+      "step": 7375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4602429916934363,
+      "learning_rate": 3.769390944970015e-06,
+      "loss": 0.8454,
+      "step": 7376
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3974324064067387,
+      "learning_rate": 3.7681317004175565e-06,
+      "loss": 0.7436,
+      "step": 7377
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6187271235495457,
+      "learning_rate": 3.766872539043664e-06,
+      "loss": 0.7088,
+      "step": 7378
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4891667363261964,
+      "learning_rate": 3.7656134609333576e-06,
+      "loss": 0.7714,
+      "step": 7379
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7674804633558427,
+      "learning_rate": 3.7643544661716518e-06,
+      "loss": 1.0402,
+      "step": 7380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4234147235527612,
+      "learning_rate": 3.7630955548435595e-06,
+      "loss": 0.663,
+      "step": 7381
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7955822923746192,
+      "learning_rate": 3.7618367270340825e-06,
+      "loss": 1.0627,
+      "step": 7382
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6212872339326139,
+      "learning_rate": 3.7605779828282225e-06,
+      "loss": 0.8007,
+      "step": 7383
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4818167617211282,
+      "learning_rate": 3.759319322310968e-06,
+      "loss": 0.731,
+      "step": 7384
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5498789048333175,
+      "learning_rate": 3.7580607455673125e-06,
+      "loss": 0.7093,
+      "step": 7385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5100507732795363,
+      "learning_rate": 3.756802252682236e-06,
+      "loss": 0.6817,
+      "step": 7386
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4810427564087147,
+      "learning_rate": 3.755543843740711e-06,
+      "loss": 0.7911,
+      "step": 7387
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6202274024307886,
+      "learning_rate": 3.7542855188277134e-06,
+      "loss": 0.7577,
+      "step": 7388
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6622735449146322,
+      "learning_rate": 3.753027278028206e-06,
+      "loss": 0.8221,
+      "step": 7389
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4163571684248732,
+      "learning_rate": 3.7517691214271485e-06,
+      "loss": 0.7396,
+      "step": 7390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.597486007782243,
+      "learning_rate": 3.750511049109493e-06,
+      "loss": 0.7301,
+      "step": 7391
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8155423970724165,
+      "learning_rate": 3.7492530611601897e-06,
+      "loss": 1.0752,
+      "step": 7392
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6642958448896712,
+      "learning_rate": 3.7479951576641793e-06,
+      "loss": 0.766,
+      "step": 7393
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.409533414988318,
+      "learning_rate": 3.7467373387063973e-06,
+      "loss": 0.7641,
+      "step": 7394
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7765479073115865,
+      "learning_rate": 3.7454796043717777e-06,
+      "loss": 1.0641,
+      "step": 7395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4830021631958124,
+      "learning_rate": 3.7442219547452436e-06,
+      "loss": 0.7464,
+      "step": 7396
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5817053298355144,
+      "learning_rate": 3.742964389911714e-06,
+      "loss": 0.8078,
+      "step": 7397
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7684778887267976,
+      "learning_rate": 3.7417069099561038e-06,
+      "loss": 1.0874,
+      "step": 7398
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4971847832416363,
+      "learning_rate": 3.74044951496332e-06,
+      "loss": 0.6988,
+      "step": 7399
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5027559398271524,
+      "learning_rate": 3.739192205018266e-06,
+      "loss": 0.7522,
+      "step": 7400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3450194841763936,
+      "learning_rate": 3.7379349802058363e-06,
+      "loss": 0.7766,
+      "step": 7401
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4745294786764742,
+      "learning_rate": 3.7366778406109228e-06,
+      "loss": 0.8139,
+      "step": 7402
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5872990671565737,
+      "learning_rate": 3.735420786318411e-06,
+      "loss": 0.8585,
+      "step": 7403
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.485961198404866,
+      "learning_rate": 3.734163817413177e-06,
+      "loss": 0.8167,
+      "step": 7404
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5008803204971517,
+      "learning_rate": 3.7329069339800984e-06,
+      "loss": 0.7545,
+      "step": 7405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5635454740994974,
+      "learning_rate": 3.7316501361040412e-06,
+      "loss": 0.7583,
+      "step": 7406
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.787959548858775,
+      "learning_rate": 3.7303934238698675e-06,
+      "loss": 1.0938,
+      "step": 7407
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6638546370108085,
+      "learning_rate": 3.7291367973624314e-06,
+      "loss": 0.7608,
+      "step": 7408
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7930433138419533,
+      "learning_rate": 3.727880256666586e-06,
+      "loss": 1.0948,
+      "step": 7409
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4661730351027815,
+      "learning_rate": 3.726623801867174e-06,
+      "loss": 0.7288,
+      "step": 7410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5418319307299484,
+      "learning_rate": 3.725367433049033e-06,
+      "loss": 0.707,
+      "step": 7411
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.52582904045691,
+      "learning_rate": 3.7241111502970003e-06,
+      "loss": 0.8268,
+      "step": 7412
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.61620839076135,
+      "learning_rate": 3.722854953695899e-06,
+      "loss": 0.7008,
+      "step": 7413
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6293365605025272,
+      "learning_rate": 3.721598843330552e-06,
+      "loss": 0.8251,
+      "step": 7414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.538634350645867,
+      "learning_rate": 3.720342819285774e-06,
+      "loss": 0.7262,
+      "step": 7415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8028660882328434,
+      "learning_rate": 3.7190868816463753e-06,
+      "loss": 1.0225,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4902581256384901,
+      "learning_rate": 3.71783103049716e-06,
+      "loss": 0.7409,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.442191482254818,
+      "learning_rate": 3.716575265922924e-06,
+      "loss": 0.7646,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.531356555045697,
+      "learning_rate": 3.7153195880084616e-06,
+      "loss": 0.7279,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.638918668571168,
+      "learning_rate": 3.714063996838558e-06,
+      "loss": 0.7801,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.609986561207907,
+      "learning_rate": 3.712808492497992e-06,
+      "loss": 0.8274,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.604873756549189,
+      "learning_rate": 3.7115530750715416e-06,
+      "loss": 0.6997,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5792596942921844,
+      "learning_rate": 3.7102977446439743e-06,
+      "loss": 0.7824,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4567298072558976,
+      "learning_rate": 3.709042501300052e-06,
+      "loss": 0.7585,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5044455599525024,
+      "learning_rate": 3.7077873451245317e-06,
+      "loss": 0.8378,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.543329414681131,
+      "learning_rate": 3.706532276202165e-06,
+      "loss": 0.7475,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4822820965374717,
+      "learning_rate": 3.705277294617697e-06,
+      "loss": 0.8105,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5915595274848706,
+      "learning_rate": 3.7040224004558646e-06,
+      "loss": 0.7157,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4901148608981263,
+      "learning_rate": 3.7027675938014046e-06,
+      "loss": 0.838,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4240608376326165,
+      "learning_rate": 3.701512874739045e-06,
+      "loss": 0.7563,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8780802672923584,
+      "learning_rate": 3.7002582433535035e-06,
+      "loss": 0.764,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.479349948993328,
+      "learning_rate": 3.6990036997294953e-06,
+      "loss": 0.8256,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5342389113323844,
+      "learning_rate": 3.697749243951735e-06,
+      "loss": 0.7805,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.601046986047214,
+      "learning_rate": 3.6964948761049225e-06,
+      "loss": 0.8128,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8247698010786512,
+      "learning_rate": 3.6952405962737565e-06,
+      "loss": 1.0913,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5534110593666266,
+      "learning_rate": 3.69398640454293e-06,
+      "loss": 0.8417,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5317333900049852,
+      "learning_rate": 3.6927323009971273e-06,
+      "loss": 0.7323,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.79144289539661,
+      "learning_rate": 3.6914782857210263e-06,
+      "loss": 1.0855,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5561843607522912,
+      "learning_rate": 3.6902243587993068e-06,
+      "loss": 0.7702,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4871913752751227,
+      "learning_rate": 3.6889705203166327e-06,
+      "loss": 0.8021,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.478124789051195,
+      "learning_rate": 3.6877167703576676e-06,
+      "loss": 0.8387,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.502879393457538,
+      "learning_rate": 3.6864631090070656e-06,
+      "loss": 0.7783,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4754543991489188,
+      "learning_rate": 3.6852095363494788e-06,
+      "loss": 0.7649,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3434870234784428,
+      "learning_rate": 3.683956052469551e-06,
+      "loss": 0.7454,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7796303859245608,
+      "learning_rate": 3.682702657451919e-06,
+      "loss": 0.7892,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4735209345789462,
+      "learning_rate": 3.6814493513812165e-06,
+      "loss": 0.7183,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4333034360833905,
+      "learning_rate": 3.680196134342069e-06,
+      "loss": 0.7697,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8212285741732758,
+      "learning_rate": 3.678943006419096e-06,
+      "loss": 1.0763,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5991135859803531,
+      "learning_rate": 3.6776899676969104e-06,
+      "loss": 0.8305,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4209175598954586,
+      "learning_rate": 3.676437018260123e-06,
+      "loss": 0.7585,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7693564731624345,
+      "learning_rate": 3.6751841581933356e-06,
+      "loss": 1.068,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4966292402736916,
+      "learning_rate": 3.673931387581142e-06,
+      "loss": 0.7539,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5901003647995444,
+      "learning_rate": 3.672678706508134e-06,
+      "loss": 0.6898,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.489348035385817,
+      "learning_rate": 3.6714261150588947e-06,
+      "loss": 0.7758,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.495575771224626,
+      "learning_rate": 3.6701736133180007e-06,
+      "loss": 0.7646,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.508689074723911,
+      "learning_rate": 3.668921201370027e-06,
+      "loss": 0.7496,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5867446683589597,
+      "learning_rate": 3.6676688792995375e-06,
+      "loss": 0.7608,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.622250690845507,
+      "learning_rate": 3.6664166471910924e-06,
+      "loss": 0.7735,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5436070501283474,
+      "learning_rate": 3.6651645051292415e-06,
+      "loss": 0.8228,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5205480888480039,
+      "learning_rate": 3.663912453198538e-06,
+      "loss": 0.7456,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4712395186948484,
+      "learning_rate": 3.662660491483521e-06,
+      "loss": 0.7433,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.647927048030299,
+      "learning_rate": 3.661408620068725e-06,
+      "loss": 0.7773,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4465968719848012,
+      "learning_rate": 3.6601568390386797e-06,
+      "loss": 0.7907,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8033852491291057,
+      "learning_rate": 3.6589051484779094e-06,
+      "loss": 1.0552,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.517223224277276,
+      "learning_rate": 3.6576535484709298e-06,
+      "loss": 0.7942,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5985065234193088,
+      "learning_rate": 3.6564020391022493e-06,
+      "loss": 0.7714,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7808844474651351,
+      "learning_rate": 3.655150620456378e-06,
+      "loss": 1.082,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.430752475521129,
+      "learning_rate": 3.6538992926178117e-06,
+      "loss": 0.7998,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.478408396577184,
+      "learning_rate": 3.652648055671043e-06,
+      "loss": 0.7465,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7165214602779304,
+      "learning_rate": 3.6513969097005585e-06,
+      "loss": 0.7936,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5893213405403046,
+      "learning_rate": 3.6501458547908396e-06,
+      "loss": 0.7432,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3558036662466073,
+      "learning_rate": 3.648894891026358e-06,
+      "loss": 0.7619,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4491292086726095,
+      "learning_rate": 3.6476440184915817e-06,
+      "loss": 0.7198,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5502836031649407,
+      "learning_rate": 3.6463932372709763e-06,
+      "loss": 0.7147,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.611648355810139,
+      "learning_rate": 3.645142547448994e-06,
+      "loss": 0.7352,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5605110910287165,
+      "learning_rate": 3.643891949110082e-06,
+      "loss": 0.6964,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5266792476351092,
+      "learning_rate": 3.6426414423386898e-06,
+      "loss": 0.729,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5356754900388923,
+      "learning_rate": 3.6413910272192504e-06,
+      "loss": 0.7914,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5691789932641294,
+      "learning_rate": 3.6401407038361948e-06,
+      "loss": 0.7635,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6550119612651897,
+      "learning_rate": 3.6388904722739493e-06,
+      "loss": 0.8254,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5627594822313946,
+      "learning_rate": 3.6376403326169317e-06,
+      "loss": 0.8637,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5556852287939968,
+      "learning_rate": 3.6363902849495535e-06,
+      "loss": 0.711,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7837348707008455,
+      "learning_rate": 3.63514032935622e-06,
+      "loss": 1.0432,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.521539874865236,
+      "learning_rate": 3.6338904659213335e-06,
+      "loss": 0.7743,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5269743497254862,
+      "learning_rate": 3.6326406947292875e-06,
+      "loss": 0.7468,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4778940461699037,
+      "learning_rate": 3.631391015864467e-06,
+      "loss": 0.7115,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.462959661288377,
+      "learning_rate": 3.6301414294112557e-06,
+      "loss": 0.6991,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8016299241323159,
+      "learning_rate": 3.6288919354540276e-06,
+      "loss": 1.0723,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6100458622934315,
+      "learning_rate": 3.6276425340771517e-06,
+      "loss": 0.7724,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.61142448095863,
+      "learning_rate": 3.626393225364988e-06,
+      "loss": 0.7882,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4781464947015273,
+      "learning_rate": 3.6251440094018956e-06,
+      "loss": 0.7763,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.716731936028379,
+      "learning_rate": 3.6238948862722246e-06,
+      "loss": 0.7976,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7106938853357843,
+      "learning_rate": 3.6226458560603144e-06,
+      "loss": 0.7327,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7700988432335054,
+      "learning_rate": 3.621396918850508e-06,
+      "loss": 0.8713,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4202617696068167,
+      "learning_rate": 3.6201480747271337e-06,
+      "loss": 0.7336,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8300240537283983,
+      "learning_rate": 3.6188993237745163e-06,
+      "loss": 0.7498,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.476991949259957,
+      "learning_rate": 3.617650666076975e-06,
+      "loss": 0.7295,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5483707270079279,
+      "learning_rate": 3.6164021017188223e-06,
+      "loss": 0.7721,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7196706532168888,
+      "learning_rate": 3.6151536307843625e-06,
+      "loss": 0.8331,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.86331977190653,
+      "learning_rate": 3.613905253357895e-06,
+      "loss": 1.0704,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5941748816392318,
+      "learning_rate": 3.6126569695237156e-06,
+      "loss": 0.747,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8298000897555168,
+      "learning_rate": 3.6114087793661122e-06,
+      "loss": 1.0533,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.487340547263413,
+      "learning_rate": 3.610160682969359e-06,
+      "loss": 0.7697,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7751541403436542,
+      "learning_rate": 3.6089126804177373e-06,
+      "loss": 1.0621,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7564094570667504,
+      "learning_rate": 3.6076647717955117e-06,
+      "loss": 1.0527,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7933724308586982,
+      "learning_rate": 3.606416957186945e-06,
+      "loss": 0.7564,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.801611577812531,
+      "learning_rate": 3.605169236676291e-06,
+      "loss": 1.0759,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5244166475878427,
+      "learning_rate": 3.6039216103478004e-06,
+      "loss": 0.6991,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5649380415732512,
+      "learning_rate": 3.602674078285715e-06,
+      "loss": 0.7252,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4831515997071558,
+      "learning_rate": 3.601426640574269e-06,
+      "loss": 0.7413,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4287722743220175,
+      "learning_rate": 3.6001792972976957e-06,
+      "loss": 0.6193,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3814408405853835,
+      "learning_rate": 3.598932048540218e-06,
+      "loss": 0.7982,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.699914717364874,
+      "learning_rate": 3.597684894386051e-06,
+      "loss": 0.7369,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.489058176755613,
+      "learning_rate": 3.5964378349194075e-06,
+      "loss": 0.7966,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6180755905238915,
+      "learning_rate": 3.5951908702244904e-06,
+      "loss": 0.7287,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8583177767237031,
+      "learning_rate": 3.593944000385498e-06,
+      "loss": 1.0842,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8087541550536842,
+      "learning_rate": 3.59269722548662e-06,
+      "loss": 1.0755,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7771135720806651,
+      "learning_rate": 3.591450545612047e-06,
+      "loss": 1.0463,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5412034124603604,
+      "learning_rate": 3.590203960845952e-06,
+      "loss": 0.7333,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6368178900131987,
+      "learning_rate": 3.5889574712725077e-06,
+      "loss": 0.7476,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4753898203920943,
+      "learning_rate": 3.587711076975884e-06,
+      "loss": 0.7476,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.452107939704967,
+      "learning_rate": 3.5864647780402373e-06,
+      "loss": 0.7941,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4897885005384408,
+      "learning_rate": 3.5852185745497204e-06,
+      "loss": 0.6941,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6473171693395638,
+      "learning_rate": 3.58397246658848e-06,
+      "loss": 0.7303,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6604806988330723,
+      "learning_rate": 3.582726454240658e-06,
+      "loss": 0.7696,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.52036219223496,
+      "learning_rate": 3.581480537590386e-06,
+      "loss": 0.7615,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4516889256215708,
+      "learning_rate": 3.58023471672179e-06,
+      "loss": 0.8393,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5730056634965792,
+      "learning_rate": 3.5789889917189945e-06,
+      "loss": 0.7963,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5374346807371195,
+      "learning_rate": 3.577743362666112e-06,
+      "loss": 0.7616,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6524823223745182,
+      "learning_rate": 3.5764978296472484e-06,
+      "loss": 0.8409,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.568299055258799,
+      "learning_rate": 3.5752523927465066e-06,
+      "loss": 0.8278,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8681136861287417,
+      "learning_rate": 3.574007052047982e-06,
+      "loss": 1.0817,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5149006600380364,
+      "learning_rate": 3.5727618076357617e-06,
+      "loss": 0.7423,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5867275558988427,
+      "learning_rate": 3.5715166595939264e-06,
+      "loss": 0.831,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4934625843318252,
+      "learning_rate": 3.5702716080065546e-06,
+      "loss": 0.6631,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5241148036864458,
+      "learning_rate": 3.569026652957713e-06,
+      "loss": 0.7055,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5989233142591985,
+      "learning_rate": 3.567781794531461e-06,
+      "loss": 0.7209,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6289210553927786,
+      "learning_rate": 3.5665370328118596e-06,
+      "loss": 0.8328,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6369144709877455,
+      "learning_rate": 3.565292367882956e-06,
+      "loss": 0.6957,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.484696636696164,
+      "learning_rate": 3.564047799828792e-06,
+      "loss": 0.7923,
+      "step": 7540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6015695636127298,
+      "learning_rate": 3.562803328733403e-06,
+      "loss": 0.8145,
+      "step": 7541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8061253415311265,
+      "learning_rate": 3.5615589546808204e-06,
+      "loss": 1.0849,
+      "step": 7542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8796163731777786,
+      "learning_rate": 3.560314677755067e-06,
+      "loss": 0.7891,
+      "step": 7543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7681973021111453,
+      "learning_rate": 3.5590704980401564e-06,
+      "loss": 1.0424,
+      "step": 7544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3845478399340694,
+      "learning_rate": 3.5578264156201025e-06,
+      "loss": 0.7625,
+      "step": 7545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4734337366465642,
+      "learning_rate": 3.5565824305789076e-06,
+      "loss": 0.7683,
+      "step": 7546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.457416807484069,
+      "learning_rate": 3.5553385430005673e-06,
+      "loss": 0.714,
+      "step": 7547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4870509254038446,
+      "learning_rate": 3.5540947529690697e-06,
+      "loss": 0.7353,
+      "step": 7548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8215963627332331,
+      "learning_rate": 3.5528510605684017e-06,
+      "loss": 1.0819,
+      "step": 7549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5744157608066103,
+      "learning_rate": 3.55160746588254e-06,
+      "loss": 0.7598,
+      "step": 7550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7733327613727949,
+      "learning_rate": 3.5503639689954527e-06,
+      "loss": 1.0753,
+      "step": 7551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.499491525207169,
+      "learning_rate": 3.549120569991107e-06,
+      "loss": 0.8035,
+      "step": 7552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4915282259412823,
+      "learning_rate": 3.5478772689534568e-06,
+      "loss": 0.7681,
+      "step": 7553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7989377416883774,
+      "learning_rate": 3.5466340659664526e-06,
+      "loss": 1.093,
+      "step": 7554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5185657070878948,
+      "learning_rate": 3.5453909611140412e-06,
+      "loss": 0.7752,
+      "step": 7555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7445344716337217,
+      "learning_rate": 3.5441479544801586e-06,
+      "loss": 1.0735,
+      "step": 7556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.470539343726246,
+      "learning_rate": 3.542905046148735e-06,
+      "loss": 0.7418,
+      "step": 7557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.774567932419869,
+      "learning_rate": 3.5416622362036938e-06,
+      "loss": 1.061,
+      "step": 7558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7544483835583607,
+      "learning_rate": 3.540419524728954e-06,
+      "loss": 1.0558,
+      "step": 7559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4029237126279235,
+      "learning_rate": 3.5391769118084253e-06,
+      "loss": 0.7613,
+      "step": 7560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4581773524895458,
+      "learning_rate": 3.5379343975260094e-06,
+      "loss": 0.7776,
+      "step": 7561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4357724312380797,
+      "learning_rate": 3.53669198196561e-06,
+      "loss": 0.7076,
+      "step": 7562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7616847033473215,
+      "learning_rate": 3.5354496652111125e-06,
+      "loss": 1.0399,
+      "step": 7563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5453126708220262,
+      "learning_rate": 3.5342074473464026e-06,
+      "loss": 0.8157,
+      "step": 7564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.509260882523467,
+      "learning_rate": 3.532965328455356e-06,
+      "loss": 0.7569,
+      "step": 7565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5486423292601774,
+      "learning_rate": 3.5317233086218474e-06,
+      "loss": 0.7774,
+      "step": 7566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7428943575483389,
+      "learning_rate": 3.530481387929737e-06,
+      "loss": 0.7459,
+      "step": 7567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7885684013757214,
+      "learning_rate": 3.529239566462883e-06,
+      "loss": 1.0732,
+      "step": 7568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.454810455713462,
+      "learning_rate": 3.5279978443051383e-06,
+      "loss": 0.7792,
+      "step": 7569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.570097770452429,
+      "learning_rate": 3.526756221540345e-06,
+      "loss": 0.7032,
+      "step": 7570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5880038118390614,
+      "learning_rate": 3.525514698252338e-06,
+      "loss": 0.8371,
+      "step": 7571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5446698929215528,
+      "learning_rate": 3.5242732745249517e-06,
+      "loss": 0.6809,
+      "step": 7572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6608266398879645,
+      "learning_rate": 3.523031950442009e-06,
+      "loss": 0.7884,
+      "step": 7573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.675521983055264,
+      "learning_rate": 3.5217907260873265e-06,
+      "loss": 0.7662,
+      "step": 7574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7723098713019468,
+      "learning_rate": 3.5205496015447127e-06,
+      "loss": 1.0647,
+      "step": 7575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.823783610305435,
+      "learning_rate": 3.519308576897974e-06,
+      "loss": 1.065,
+      "step": 7576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4836591389929885,
+      "learning_rate": 3.5180676522309065e-06,
+      "loss": 0.8087,
+      "step": 7577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.430716937095831,
+      "learning_rate": 3.5168268276272977e-06,
+      "loss": 0.7727,
+      "step": 7578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4828065664720884,
+      "learning_rate": 3.515586103170935e-06,
+      "loss": 0.7937,
+      "step": 7579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.566977695215505,
+      "learning_rate": 3.514345478945592e-06,
+      "loss": 0.8541,
+      "step": 7580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4828633240398457,
+      "learning_rate": 3.5131049550350406e-06,
+      "loss": 0.7858,
+      "step": 7581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5624635503204707,
+      "learning_rate": 3.5118645315230394e-06,
+      "loss": 0.8118,
+      "step": 7582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4749047157608806,
+      "learning_rate": 3.5106242084933506e-06,
+      "loss": 0.8176,
+      "step": 7583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4650249970158078,
+      "learning_rate": 3.5093839860297206e-06,
+      "loss": 0.7541,
+      "step": 7584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7748572106997424,
+      "learning_rate": 3.5081438642158916e-06,
+      "loss": 1.0653,
+      "step": 7585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3953204298126096,
+      "learning_rate": 3.506903843135601e-06,
+      "loss": 0.6865,
+      "step": 7586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5478324789891318,
+      "learning_rate": 3.5056639228725777e-06,
+      "loss": 0.7739,
+      "step": 7587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.771658950634736,
+      "learning_rate": 3.5044241035105425e-06,
+      "loss": 1.0609,
+      "step": 7588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5795353204911822,
+      "learning_rate": 3.5031843851332105e-06,
+      "loss": 0.8114,
+      "step": 7589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5073379427027271,
+      "learning_rate": 3.5019447678242937e-06,
+      "loss": 0.8032,
+      "step": 7590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7819841978906235,
+      "learning_rate": 3.500705251667491e-06,
+      "loss": 1.0752,
+      "step": 7591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3910005799341711,
+      "learning_rate": 3.4994658367464963e-06,
+      "loss": 0.7564,
+      "step": 7592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4503262325501858,
+      "learning_rate": 3.4982265231450006e-06,
+      "loss": 0.7208,
+      "step": 7593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4219440564400516,
+      "learning_rate": 3.4969873109466847e-06,
+      "loss": 0.6768,
+      "step": 7594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.46450191009171,
+      "learning_rate": 3.4957482002352217e-06,
+      "loss": 0.7686,
+      "step": 7595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.478720707549166,
+      "learning_rate": 3.494509191094281e-06,
+      "loss": 0.8281,
+      "step": 7596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4920312028005376,
+      "learning_rate": 3.4932702836075216e-06,
+      "loss": 0.7606,
+      "step": 7597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.840069302730572,
+      "learning_rate": 3.492031477858598e-06,
+      "loss": 1.1017,
+      "step": 7598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4721144333944405,
+      "learning_rate": 3.4907927739311552e-06,
+      "loss": 0.7967,
+      "step": 7599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3587453420023408,
+      "learning_rate": 3.489554171908838e-06,
+      "loss": 0.6401,
+      "step": 7600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4863406994264252,
+      "learning_rate": 3.4883156718752763e-06,
+      "loss": 0.7542,
+      "step": 7601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4844771439142408,
+      "learning_rate": 3.4870772739140956e-06,
+      "loss": 0.8047,
+      "step": 7602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5084742039483718,
+      "learning_rate": 3.485838978108919e-06,
+      "loss": 0.7619,
+      "step": 7603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.585076818293297,
+      "learning_rate": 3.4846007845433568e-06,
+      "loss": 0.7396,
+      "step": 7604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5026226174881314,
+      "learning_rate": 3.4833626933010144e-06,
+      "loss": 0.7819,
+      "step": 7605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.583018447862338,
+      "learning_rate": 3.48212470446549e-06,
+      "loss": 0.7334,
+      "step": 7606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5595444385402168,
+      "learning_rate": 3.480886818120377e-06,
+      "loss": 0.7825,
+      "step": 7607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4767457732705096,
+      "learning_rate": 3.479649034349261e-06,
+      "loss": 0.6989,
+      "step": 7608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5549833543543723,
+      "learning_rate": 3.4784113532357157e-06,
+      "loss": 0.7387,
+      "step": 7609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8669489798525798,
+      "learning_rate": 3.477173774863317e-06,
+      "loss": 0.6708,
+      "step": 7610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8039718090125483,
+      "learning_rate": 3.4759362993156275e-06,
+      "loss": 1.0201,
+      "step": 7611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5288075125817986,
+      "learning_rate": 3.4746989266762034e-06,
+      "loss": 0.7091,
+      "step": 7612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6499571452125519,
+      "learning_rate": 3.4734616570285954e-06,
+      "loss": 0.7636,
+      "step": 7613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4411691021126618,
+      "learning_rate": 3.472224490456348e-06,
+      "loss": 0.7729,
+      "step": 7614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7771783080317571,
+      "learning_rate": 3.4709874270429968e-06,
+      "loss": 1.0379,
+      "step": 7615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.461151971271799,
+      "learning_rate": 3.4697504668720677e-06,
+      "loss": 0.7222,
+      "step": 7616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5926403433595129,
+      "learning_rate": 3.468513610027089e-06,
+      "loss": 0.7963,
+      "step": 7617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.417366418816824,
+      "learning_rate": 3.4672768565915726e-06,
+      "loss": 0.6927,
+      "step": 7618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7590127825287106,
+      "learning_rate": 3.4660402066490274e-06,
+      "loss": 1.0861,
+      "step": 7619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5143751247375883,
+      "learning_rate": 3.4648036602829556e-06,
+      "loss": 0.8597,
+      "step": 7620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5288457361545285,
+      "learning_rate": 3.4635672175768508e-06,
+      "loss": 0.7764,
+      "step": 7621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6079295176480357,
+      "learning_rate": 3.4623308786142017e-06,
+      "loss": 0.7939,
+      "step": 7622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3784660733392804,
+      "learning_rate": 3.4610946434784863e-06,
+      "loss": 0.6504,
+      "step": 7623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7899033664898677,
+      "learning_rate": 3.4598585122531802e-06,
+      "loss": 1.0849,
+      "step": 7624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.39961761116411,
+      "learning_rate": 3.4586224850217496e-06,
+      "loss": 0.7904,
+      "step": 7625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7799462097533494,
+      "learning_rate": 3.4573865618676506e-06,
+      "loss": 1.0735,
+      "step": 7626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4978188186109243,
+      "learning_rate": 3.456150742874341e-06,
+      "loss": 0.8106,
+      "step": 7627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5522902485485648,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.7641,
+      "step": 7628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5330431516458938,
+      "learning_rate": 3.4536794177038563e-06,
+      "loss": 0.6971,
+      "step": 7629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.462077711702339,
+      "learning_rate": 3.4524439116935492e-06,
+      "loss": 0.678,
+      "step": 7630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4863998632214435,
+      "learning_rate": 3.451208510177769e-06,
+      "loss": 0.7913,
+      "step": 7631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4910919079796032,
+      "learning_rate": 3.4499732132399324e-06,
+      "loss": 0.8358,
+      "step": 7632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5962862536776972,
+      "learning_rate": 3.448738020963446e-06,
+      "loss": 0.7617,
+      "step": 7633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5972695337144773,
+      "learning_rate": 3.4475029334317195e-06,
+      "loss": 0.8818,
+      "step": 7634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6401047156026347,
+      "learning_rate": 3.4462679507281433e-06,
+      "loss": 0.8702,
+      "step": 7635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7781267734021727,
+      "learning_rate": 3.445033072936106e-06,
+      "loss": 1.05,
+      "step": 7636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7770603340051605,
+      "learning_rate": 3.443798300138993e-06,
+      "loss": 1.0744,
+      "step": 7637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4887356590030498,
+      "learning_rate": 3.442563632420178e-06,
+      "loss": 0.7125,
+      "step": 7638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4124087756568406,
+      "learning_rate": 3.441329069863027e-06,
+      "loss": 0.6993,
+      "step": 7639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5029109686764768,
+      "learning_rate": 3.4400946125509003e-06,
+      "loss": 0.7818,
+      "step": 7640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5130376208961214,
+      "learning_rate": 3.438860260567154e-06,
+      "loss": 0.7529,
+      "step": 7641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4186527064241272,
+      "learning_rate": 3.437626013995132e-06,
+      "loss": 0.7169,
+      "step": 7642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5761395931469055,
+      "learning_rate": 3.4363918729181727e-06,
+      "loss": 0.7378,
+      "step": 7643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5207188562832274,
+      "learning_rate": 3.4351578374196117e-06,
+      "loss": 0.8204,
+      "step": 7644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.502131748277624,
+      "learning_rate": 3.4339239075827712e-06,
+      "loss": 0.8147,
+      "step": 7645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4721518005481307,
+      "learning_rate": 3.4326900834909694e-06,
+      "loss": 0.7363,
+      "step": 7646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.602426479795486,
+      "learning_rate": 3.431456365227516e-06,
+      "loss": 0.7815,
+      "step": 7647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.594465387620532,
+      "learning_rate": 3.430222752875717e-06,
+      "loss": 0.738,
+      "step": 7648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4975553440173384,
+      "learning_rate": 3.428989246518867e-06,
+      "loss": 0.73,
+      "step": 7649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4495008242576943,
+      "learning_rate": 3.4277558462402543e-06,
+      "loss": 0.7078,
+      "step": 7650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8101999529385817,
+      "learning_rate": 3.426522552123163e-06,
+      "loss": 1.0779,
+      "step": 7651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.567449252221874,
+      "learning_rate": 3.425289364250868e-06,
+      "loss": 0.8002,
+      "step": 7652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5300880920876043,
+      "learning_rate": 3.4240562827066326e-06,
+      "loss": 0.7202,
+      "step": 7653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5558439324116484,
+      "learning_rate": 3.4228233075737225e-06,
+      "loss": 0.7656,
+      "step": 7654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4499843131434937,
+      "learning_rate": 3.4215904389353897e-06,
+      "loss": 0.7154,
+      "step": 7655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4473192865088527,
+      "learning_rate": 3.42035767687488e-06,
+      "loss": 0.7159,
+      "step": 7656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4832094540484464,
+      "learning_rate": 3.4191250214754303e-06,
+      "loss": 0.7324,
+      "step": 7657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4569206152080438,
+      "learning_rate": 3.4178924728202757e-06,
+      "loss": 0.7097,
+      "step": 7658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4958638828901487,
+      "learning_rate": 3.416660030992639e-06,
+      "loss": 0.7303,
+      "step": 7659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8042847746718376,
+      "learning_rate": 3.415427696075735e-06,
+      "loss": 1.0928,
+      "step": 7660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.420237026388972,
+      "learning_rate": 3.414195468152779e-06,
+      "loss": 0.789,
+      "step": 7661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4959250470330512,
+      "learning_rate": 3.41296334730697e-06,
+      "loss": 0.8103,
+      "step": 7662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5351756794881675,
+      "learning_rate": 3.411731333621507e-06,
+      "loss": 0.7734,
+      "step": 7663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5017176534061838,
+      "learning_rate": 3.410499427179572e-06,
+      "loss": 0.7037,
+      "step": 7664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4386512294183724,
+      "learning_rate": 3.4092676280643533e-06,
+      "loss": 0.7266,
+      "step": 7665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8005861480789737,
+      "learning_rate": 3.4080359363590214e-06,
+      "loss": 1.0715,
+      "step": 7666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7951726105276048,
+      "learning_rate": 3.4068043521467424e-06,
+      "loss": 1.0764,
+      "step": 7667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7776369426031778,
+      "learning_rate": 3.4055728755106775e-06,
+      "loss": 1.0877,
+      "step": 7668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4171632119984605,
+      "learning_rate": 3.404341506533978e-06,
+      "loss": 0.6302,
+      "step": 7669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5436326713679167,
+      "learning_rate": 3.4031102452997864e-06,
+      "loss": 0.7769,
+      "step": 7670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5267072059182556,
+      "learning_rate": 3.4018790918912447e-06,
+      "loss": 0.7832,
+      "step": 7671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.660475983200203,
+      "learning_rate": 3.4006480463914814e-06,
+      "loss": 0.7632,
+      "step": 7672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4554772779602787,
+      "learning_rate": 3.399417108883619e-06,
+      "loss": 0.8324,
+      "step": 7673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5073015954402174,
+      "learning_rate": 3.3981862794507725e-06,
+      "loss": 0.7638,
+      "step": 7674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.682368460434198,
+      "learning_rate": 3.3969555581760518e-06,
+      "loss": 0.7109,
+      "step": 7675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4223391935364513,
+      "learning_rate": 3.395724945142558e-06,
+      "loss": 0.7177,
+      "step": 7676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7709082799196232,
+      "learning_rate": 3.3944944404333815e-06,
+      "loss": 1.0832,
+      "step": 7677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6134872499856183,
+      "learning_rate": 3.3932640441316135e-06,
+      "loss": 0.8333,
+      "step": 7678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.567629533635683,
+      "learning_rate": 3.392033756320333e-06,
+      "loss": 0.7478,
+      "step": 7679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.787993585169824,
+      "learning_rate": 3.3908035770826085e-06,
+      "loss": 1.0631,
+      "step": 7680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4355988783278564,
+      "learning_rate": 3.3895735065015044e-06,
+      "loss": 0.6998,
+      "step": 7681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5305075968266346,
+      "learning_rate": 3.388343544660082e-06,
+      "loss": 0.6448,
+      "step": 7682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.538724295682777,
+      "learning_rate": 3.387113691641388e-06,
+      "loss": 0.78,
+      "step": 7683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4169611112825142,
+      "learning_rate": 3.385883947528465e-06,
+      "loss": 0.715,
+      "step": 7684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5461424102766514,
+      "learning_rate": 3.3846543124043496e-06,
+      "loss": 0.8918,
+      "step": 7685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.526251471047948,
+      "learning_rate": 3.3834247863520692e-06,
+      "loss": 0.7093,
+      "step": 7686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2231663207469103,
+      "learning_rate": 3.3821953694546433e-06,
+      "loss": 0.7586,
+      "step": 7687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4796068198311394,
+      "learning_rate": 3.3809660617950835e-06,
+      "loss": 0.6639,
+      "step": 7688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.387221048383456,
+      "learning_rate": 3.379736863456399e-06,
+      "loss": 0.6734,
+      "step": 7689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4544434933735053,
+      "learning_rate": 3.378507774521587e-06,
+      "loss": 0.7909,
+      "step": 7690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8139126230276331,
+      "learning_rate": 3.377278795073637e-06,
+      "loss": 1.0782,
+      "step": 7691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.815805088331301,
+      "learning_rate": 3.376049925195534e-06,
+      "loss": 1.0573,
+      "step": 7692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4802041281772744,
+      "learning_rate": 3.3748211649702533e-06,
+      "loss": 0.7688,
+      "step": 7693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5368633027442444,
+      "learning_rate": 3.3735925144807623e-06,
+      "loss": 0.7591,
+      "step": 7694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5188434840632612,
+      "learning_rate": 3.3723639738100254e-06,
+      "loss": 0.7691,
+      "step": 7695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.539671563677665,
+      "learning_rate": 3.371135543040995e-06,
+      "loss": 0.743,
+      "step": 7696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8183696661805621,
+      "learning_rate": 3.369907222256617e-06,
+      "loss": 1.0306,
+      "step": 7697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5890172940816591,
+      "learning_rate": 3.3686790115398287e-06,
+      "loss": 0.8589,
+      "step": 7698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6008422251802295,
+      "learning_rate": 3.367450910973566e-06,
+      "loss": 0.8146,
+      "step": 7699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7583910451795927,
+      "learning_rate": 3.3662229206407505e-06,
+      "loss": 1.0855,
+      "step": 7700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.449687015023182,
+      "learning_rate": 3.3649950406242986e-06,
+      "loss": 0.7717,
+      "step": 7701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3858920354432247,
+      "learning_rate": 3.3637672710071213e-06,
+      "loss": 0.7001,
+      "step": 7702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4828308821146254,
+      "learning_rate": 3.362539611872119e-06,
+      "loss": 0.728,
+      "step": 7703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6995548224690664,
+      "learning_rate": 3.3613120633021868e-06,
+      "loss": 0.7982,
+      "step": 7704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4315692739849697,
+      "learning_rate": 3.360084625380209e-06,
+      "loss": 0.7399,
+      "step": 7705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7906718874783281,
+      "learning_rate": 3.358857298189069e-06,
+      "loss": 1.0645,
+      "step": 7706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5135814217164067,
+      "learning_rate": 3.357630081811638e-06,
+      "loss": 0.7795,
+      "step": 7707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4453116393666199,
+      "learning_rate": 3.356402976330776e-06,
+      "loss": 0.7766,
+      "step": 7708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.431605980667538,
+      "learning_rate": 3.355175981829346e-06,
+      "loss": 0.6501,
+      "step": 7709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.485342674065422,
+      "learning_rate": 3.3539490983901944e-06,
+      "loss": 0.7708,
+      "step": 7710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.46952762901311,
+      "learning_rate": 3.352722326096163e-06,
+      "loss": 0.6751,
+      "step": 7711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.518457286606172,
+      "learning_rate": 3.3514956650300877e-06,
+      "loss": 0.7464,
+      "step": 7712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5119029407721312,
+      "learning_rate": 3.3502691152747947e-06,
+      "loss": 0.7959,
+      "step": 7713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6370664235011787,
+      "learning_rate": 3.3490426769131035e-06,
+      "loss": 0.7822,
+      "step": 7714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.579172632932745,
+      "learning_rate": 3.347816350027823e-06,
+      "loss": 0.7549,
+      "step": 7715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3609181530164969,
+      "learning_rate": 3.3465901347017633e-06,
+      "loss": 0.707,
+      "step": 7716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5420055954275518,
+      "learning_rate": 3.345364031017718e-06,
+      "loss": 0.738,
+      "step": 7717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4709976958593354,
+      "learning_rate": 3.3441380390584765e-06,
+      "loss": 0.7985,
+      "step": 7718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5875290870711958,
+      "learning_rate": 3.3429121589068213e-06,
+      "loss": 0.7934,
+      "step": 7719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7841357655265737,
+      "learning_rate": 3.3416863906455264e-06,
+      "loss": 1.0996,
+      "step": 7720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5275904119934207,
+      "learning_rate": 3.340460734357359e-06,
+      "loss": 0.765,
+      "step": 7721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.610217874995973,
+      "learning_rate": 3.339235190125075e-06,
+      "loss": 0.8262,
+      "step": 7722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.554263477493308,
+      "learning_rate": 3.338009758031432e-06,
+      "loss": 0.8038,
+      "step": 7723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4994560689654726,
+      "learning_rate": 3.33678443815917e-06,
+      "loss": 0.85,
+      "step": 7724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4973053402150438,
+      "learning_rate": 3.3355592305910223e-06,
+      "loss": 0.6942,
+      "step": 7725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.485354515757623,
+      "learning_rate": 3.334334135409724e-06,
+      "loss": 0.7558,
+      "step": 7726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3770813567425635,
+      "learning_rate": 3.333109152697994e-06,
+      "loss": 0.6486,
+      "step": 7727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5722437509053684,
+      "learning_rate": 3.3318842825385454e-06,
+      "loss": 0.7815,
+      "step": 7728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5339600119853851,
+      "learning_rate": 3.3306595250140834e-06,
+      "loss": 0.6599,
+      "step": 7729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4834201763270622,
+      "learning_rate": 3.3294348802073085e-06,
+      "loss": 0.7609,
+      "step": 7730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4978176997527324,
+      "learning_rate": 3.3282103482009103e-06,
+      "loss": 0.6741,
+      "step": 7731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.824194045892066,
+      "learning_rate": 3.32698592907757e-06,
+      "loss": 1.0197,
+      "step": 7732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5273626628165662,
+      "learning_rate": 3.325761622919968e-06,
+      "loss": 0.7249,
+      "step": 7733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4828409856405271,
+      "learning_rate": 3.324537429810769e-06,
+      "loss": 0.7917,
+      "step": 7734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6592878525449464,
+      "learning_rate": 3.323313349832633e-06,
+      "loss": 0.7266,
+      "step": 7735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4822094409757134,
+      "learning_rate": 3.322089383068215e-06,
+      "loss": 0.7565,
+      "step": 7736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7681489446935946,
+      "learning_rate": 3.3208655296001585e-06,
+      "loss": 1.087,
+      "step": 7737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5285037494250227,
+      "learning_rate": 3.319641789511101e-06,
+      "loss": 0.7539,
+      "step": 7738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6149893711683694,
+      "learning_rate": 3.318418162883671e-06,
+      "loss": 0.7737,
+      "step": 7739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5224002877201968,
+      "learning_rate": 3.3171946498004925e-06,
+      "loss": 0.7227,
+      "step": 7740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.584677866764195,
+      "learning_rate": 3.3159712503441798e-06,
+      "loss": 0.8583,
+      "step": 7741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4411263673331944,
+      "learning_rate": 3.3147479645973367e-06,
+      "loss": 0.7379,
+      "step": 7742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8403717632591616,
+      "learning_rate": 3.3135247926425675e-06,
+      "loss": 0.7616,
+      "step": 7743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6356992430942525,
+      "learning_rate": 3.3123017345624597e-06,
+      "loss": 0.7831,
+      "step": 7744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4778209267358549,
+      "learning_rate": 3.311078790439598e-06,
+      "loss": 0.7639,
+      "step": 7745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5545549543177637,
+      "learning_rate": 3.309855960356557e-06,
+      "loss": 0.7718,
+      "step": 7746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5218035112431112,
+      "learning_rate": 3.3086332443959086e-06,
+      "loss": 0.7888,
+      "step": 7747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.518894853807762,
+      "learning_rate": 3.3074106426402097e-06,
+      "loss": 0.7169,
+      "step": 7748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7304711389266134,
+      "learning_rate": 3.306188155172013e-06,
+      "loss": 0.7842,
+      "step": 7749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8062458996308914,
+      "learning_rate": 3.304965782073868e-06,
+      "loss": 1.0585,
+      "step": 7750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4321074124924713,
+      "learning_rate": 3.30374352342831e-06,
+      "loss": 0.7552,
+      "step": 7751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6647044158653015,
+      "learning_rate": 3.3025213793178647e-06,
+      "loss": 0.767,
+      "step": 7752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4927534116294736,
+      "learning_rate": 3.301299349825059e-06,
+      "loss": 0.7629,
+      "step": 7753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.481934945366105,
+      "learning_rate": 3.300077435032406e-06,
+      "loss": 0.8199,
+      "step": 7754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7112164574935533,
+      "learning_rate": 3.298855635022411e-06,
+      "loss": 0.7413,
+      "step": 7755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4002598564481645,
+      "learning_rate": 3.2976339498775734e-06,
+      "loss": 0.8137,
+      "step": 7756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5655197849824238,
+      "learning_rate": 3.2964123796803847e-06,
+      "loss": 0.7574,
+      "step": 7757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7819476296236746,
+      "learning_rate": 3.2951909245133277e-06,
+      "loss": 1.0698,
+      "step": 7758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.59718820577481,
+      "learning_rate": 3.2939695844588758e-06,
+      "loss": 0.7669,
+      "step": 7759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.474407811989907,
+      "learning_rate": 3.2927483595995003e-06,
+      "loss": 0.6877,
+      "step": 7760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5958094987160363,
+      "learning_rate": 3.29152725001766e-06,
+      "loss": 0.7557,
+      "step": 7761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5568552557305642,
+      "learning_rate": 3.2903062557958065e-06,
+      "loss": 0.7037,
+      "step": 7762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5217694279969063,
+      "learning_rate": 3.2890853770163822e-06,
+      "loss": 0.7809,
+      "step": 7763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.395850989073227,
+      "learning_rate": 3.2878646137618275e-06,
+      "loss": 0.7711,
+      "step": 7764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7582074679438263,
+      "learning_rate": 3.2866439661145684e-06,
+      "loss": 1.0672,
+      "step": 7765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7602438570790034,
+      "learning_rate": 3.2854234341570247e-06,
+      "loss": 1.0554,
+      "step": 7766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5053707260218123,
+      "learning_rate": 3.2842030179716146e-06,
+      "loss": 0.8551,
+      "step": 7767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.612079291211773,
+      "learning_rate": 3.282982717640739e-06,
+      "loss": 0.872,
+      "step": 7768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.68799657215577,
+      "learning_rate": 3.281762533246794e-06,
+      "loss": 0.7998,
+      "step": 7769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5926629914845578,
+      "learning_rate": 3.280542464872174e-06,
+      "loss": 0.8295,
+      "step": 7770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6075243593945032,
+      "learning_rate": 3.279322512599259e-06,
+      "loss": 0.7991,
+      "step": 7771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5687269004277071,
+      "learning_rate": 3.2781026765104224e-06,
+      "loss": 0.7594,
+      "step": 7772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5456648903242183,
+      "learning_rate": 3.276882956688029e-06,
+      "loss": 0.7649,
+      "step": 7773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4934650332761044,
+      "learning_rate": 3.27566335321444e-06,
+      "loss": 0.7762,
+      "step": 7774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.380947989710664,
+      "learning_rate": 3.274443866172004e-06,
+      "loss": 0.6833,
+      "step": 7775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.588025714191186,
+      "learning_rate": 3.273224495643062e-06,
+      "loss": 0.7275,
+      "step": 7776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5872474298199568,
+      "learning_rate": 3.2720052417099526e-06,
+      "loss": 0.8538,
+      "step": 7777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8018654235931406,
+      "learning_rate": 3.2707861044550003e-06,
+      "loss": 1.1054,
+      "step": 7778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5407507488515706,
+      "learning_rate": 3.269567083960525e-06,
+      "loss": 0.7393,
+      "step": 7779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4923556461713623,
+      "learning_rate": 3.268348180308836e-06,
+      "loss": 0.7688,
+      "step": 7780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8255574979458629,
+      "learning_rate": 3.267129393582238e-06,
+      "loss": 0.7432,
+      "step": 7781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5729847288555268,
+      "learning_rate": 3.2659107238630258e-06,
+      "loss": 0.7771,
+      "step": 7782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.477684613311222,
+      "learning_rate": 3.2646921712334854e-06,
+      "loss": 0.7615,
+      "step": 7783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4289188473259566,
+      "learning_rate": 3.2634737357758994e-06,
+      "loss": 0.7907,
+      "step": 7784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5324739243135617,
+      "learning_rate": 3.2622554175725376e-06,
+      "loss": 0.7928,
+      "step": 7785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8064895378409249,
+      "learning_rate": 3.2610372167056633e-06,
+      "loss": 1.0643,
+      "step": 7786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.431838820914917,
+      "learning_rate": 3.25981913325753e-06,
+      "loss": 0.8515,
+      "step": 7787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5877710159724996,
+      "learning_rate": 3.2586011673103907e-06,
+      "loss": 0.7957,
+      "step": 7788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.508441333666042,
+      "learning_rate": 3.257383318946482e-06,
+      "loss": 0.7767,
+      "step": 7789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5144968139461934,
+      "learning_rate": 3.2561655882480358e-06,
+      "loss": 0.7031,
+      "step": 7790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4689640532218047,
+      "learning_rate": 3.254947975297278e-06,
+      "loss": 0.7505,
+      "step": 7791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5962760730569854,
+      "learning_rate": 3.2537304801764225e-06,
+      "loss": 0.7876,
+      "step": 7792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.542190258321352,
+      "learning_rate": 3.252513102967676e-06,
+      "loss": 0.7785,
+      "step": 7793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5986558853403114,
+      "learning_rate": 3.2512958437532426e-06,
+      "loss": 0.8383,
+      "step": 7794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7611895230340678,
+      "learning_rate": 3.250078702615314e-06,
+      "loss": 1.0514,
+      "step": 7795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7459017170071067,
+      "learning_rate": 3.2488616796360717e-06,
+      "loss": 1.0741,
+      "step": 7796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4205475691246763,
+      "learning_rate": 3.2476447748976906e-06,
+      "loss": 0.8066,
+      "step": 7797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4317242432487347,
+      "learning_rate": 3.2464279884823436e-06,
+      "loss": 0.762,
+      "step": 7798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.431288608589671,
+      "learning_rate": 3.245211320472189e-06,
+      "loss": 0.7633,
+      "step": 7799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.745610564594737,
+      "learning_rate": 3.243994770949377e-06,
+      "loss": 1.0468,
+      "step": 7800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4392876081289605,
+      "learning_rate": 3.2427783399960544e-06,
+      "loss": 0.6468,
+      "step": 7801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4820993634074593,
+      "learning_rate": 3.241562027694357e-06,
+      "loss": 0.7681,
+      "step": 7802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7879525761160608,
+      "learning_rate": 3.240345834126412e-06,
+      "loss": 1.088,
+      "step": 7803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5406098101999683,
+      "learning_rate": 3.2391297593743374e-06,
+      "loss": 0.7691,
+      "step": 7804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5208319660428213,
+      "learning_rate": 3.237913803520251e-06,
+      "loss": 0.7007,
+      "step": 7805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.515692031791251,
+      "learning_rate": 3.236697966646254e-06,
+      "loss": 0.7572,
+      "step": 7806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6037855105784322,
+      "learning_rate": 3.2354822488344407e-06,
+      "loss": 0.7732,
+      "step": 7807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6381248309371017,
+      "learning_rate": 3.234266650166901e-06,
+      "loss": 0.7483,
+      "step": 7808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6277485437439911,
+      "learning_rate": 3.2330511707257164e-06,
+      "loss": 0.7798,
+      "step": 7809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4910964743656239,
+      "learning_rate": 3.2318358105929538e-06,
+      "loss": 0.7823,
+      "step": 7810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6390698215695276,
+      "learning_rate": 3.2306205698506832e-06,
+      "loss": 0.7913,
+      "step": 7811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.49303984132423,
+      "learning_rate": 3.2294054485809577e-06,
+      "loss": 0.7293,
+      "step": 7812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4496776496849337,
+      "learning_rate": 3.228190446865824e-06,
+      "loss": 0.6615,
+      "step": 7813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6369959696170864,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.7853,
+      "step": 7814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4345916948205992,
+      "learning_rate": 3.2257608024274857e-06,
+      "loss": 0.7468,
+      "step": 7815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5731383502344427,
+      "learning_rate": 3.224546159868337e-06,
+      "loss": 0.7698,
+      "step": 7816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3323647257859619,
+      "learning_rate": 3.22333163719189e-06,
+      "loss": 0.6246,
+      "step": 7817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5175248211031584,
+      "learning_rate": 3.2221172344801543e-06,
+      "loss": 0.791,
+      "step": 7818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.608666180317806,
+      "learning_rate": 3.220902951815129e-06,
+      "loss": 0.7575,
+      "step": 7819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.45656570559801,
+      "learning_rate": 3.2196887892788044e-06,
+      "loss": 0.7393,
+      "step": 7820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5421844317473503,
+      "learning_rate": 3.2184747469531618e-06,
+      "loss": 0.6887,
+      "step": 7821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4815147291390154,
+      "learning_rate": 3.2172608249201796e-06,
+      "loss": 0.7419,
+      "step": 7822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5464214112353742,
+      "learning_rate": 3.2160470232618228e-06,
+      "loss": 0.7742,
+      "step": 7823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5255883546452593,
+      "learning_rate": 3.2148333420600497e-06,
+      "loss": 0.7578,
+      "step": 7824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5474463719932279,
+      "learning_rate": 3.213619781396812e-06,
+      "loss": 0.7833,
+      "step": 7825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8001016406975772,
+      "learning_rate": 3.2124063413540517e-06,
+      "loss": 1.0612,
+      "step": 7826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4345510796415737,
+      "learning_rate": 3.211193022013702e-06,
+      "loss": 0.7405,
+      "step": 7827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6746786765458157,
+      "learning_rate": 3.209979823457691e-06,
+      "loss": 0.8003,
+      "step": 7828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.531881553409104,
+      "learning_rate": 3.208766745767935e-06,
+      "loss": 0.7546,
+      "step": 7829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5767603630566982,
+      "learning_rate": 3.207553789026344e-06,
+      "loss": 0.7386,
+      "step": 7830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6667129134491754,
+      "learning_rate": 3.2063409533148183e-06,
+      "loss": 0.8291,
+      "step": 7831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3602322789040886,
+      "learning_rate": 3.2051282387152547e-06,
+      "loss": 0.7391,
+      "step": 7832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6054953762475686,
+      "learning_rate": 3.2039156453095366e-06,
+      "loss": 0.7988,
+      "step": 7833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7969563553130192,
+      "learning_rate": 3.2027031731795403e-06,
+      "loss": 1.0536,
+      "step": 7834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4265000619854054,
+      "learning_rate": 3.2014908224071367e-06,
+      "loss": 0.7279,
+      "step": 7835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5072576293189832,
+      "learning_rate": 3.2002785930741855e-06,
+      "loss": 0.7444,
+      "step": 7836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6172904031052406,
+      "learning_rate": 3.199066485262538e-06,
+      "loss": 0.7759,
+      "step": 7837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6940830930648774,
+      "learning_rate": 3.1978544990540383e-06,
+      "loss": 0.7719,
+      "step": 7838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6121913255937244,
+      "learning_rate": 3.1966426345305263e-06,
+      "loss": 0.7425,
+      "step": 7839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4819006913463797,
+      "learning_rate": 3.1954308917738263e-06,
+      "loss": 0.7077,
+      "step": 7840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5377942764151096,
+      "learning_rate": 3.194219270865757e-06,
+      "loss": 0.7826,
+      "step": 7841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4526125359385222,
+      "learning_rate": 3.1930077718881336e-06,
+      "loss": 0.7534,
+      "step": 7842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4226338788066621,
+      "learning_rate": 3.191796394922757e-06,
+      "loss": 0.6863,
+      "step": 7843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.805818346712067,
+      "learning_rate": 3.1905851400514232e-06,
+      "loss": 1.012,
+      "step": 7844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5235302992685509,
+      "learning_rate": 3.189374007355917e-06,
+      "loss": 0.6967,
+      "step": 7845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4826994911575055,
+      "learning_rate": 3.1881629969180197e-06,
+      "loss": 0.7567,
+      "step": 7846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4138080693804158,
+      "learning_rate": 3.186952108819499e-06,
+      "loss": 0.687,
+      "step": 7847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5811347273231366,
+      "learning_rate": 3.1857413431421156e-06,
+      "loss": 0.8498,
+      "step": 7848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4763936263369029,
+      "learning_rate": 3.1845306999676274e-06,
+      "loss": 0.7799,
+      "step": 7849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5326458985211964,
+      "learning_rate": 3.183320179377778e-06,
+      "loss": 0.8035,
+      "step": 7850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4510757324269528,
+      "learning_rate": 3.182109781454303e-06,
+      "loss": 0.7455,
+      "step": 7851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7860472443347002,
+      "learning_rate": 3.1808995062789335e-06,
+      "loss": 0.7288,
+      "step": 7852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6015356039303732,
+      "learning_rate": 3.1796893539333884e-06,
+      "loss": 0.7434,
+      "step": 7853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.55957233374763,
+      "learning_rate": 3.178479324499381e-06,
+      "loss": 0.6916,
+      "step": 7854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5143179540356113,
+      "learning_rate": 3.1772694180586137e-06,
+      "loss": 0.6709,
+      "step": 7855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.573744039644747,
+      "learning_rate": 3.1760596346927843e-06,
+      "loss": 0.8341,
+      "step": 7856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.560791867875191,
+      "learning_rate": 3.174849974483579e-06,
+      "loss": 0.7907,
+      "step": 7857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7858211671402192,
+      "learning_rate": 3.173640437512675e-06,
+      "loss": 1.0323,
+      "step": 7858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7771064932734836,
+      "learning_rate": 3.1724310238617475e-06,
+      "loss": 1.1031,
+      "step": 7859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3921858659723871,
+      "learning_rate": 3.171221733612455e-06,
+      "loss": 0.7436,
+      "step": 7860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6202562348252048,
+      "learning_rate": 3.1700125668464534e-06,
+      "loss": 0.7748,
+      "step": 7861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5599780995953991,
+      "learning_rate": 3.168803523645387e-06,
+      "loss": 0.8009,
+      "step": 7862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6330109483338922,
+      "learning_rate": 3.1675946040908946e-06,
+      "loss": 0.8111,
+      "step": 7863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5607665833767694,
+      "learning_rate": 3.1663858082646047e-06,
+      "loss": 0.7864,
+      "step": 7864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4558880150071696,
+      "learning_rate": 3.165177136248135e-06,
+      "loss": 0.8103,
+      "step": 7865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5558702558662263,
+      "learning_rate": 3.163968588123104e-06,
+      "loss": 0.7726,
+      "step": 7866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5040355541588522,
+      "learning_rate": 3.162760163971112e-06,
+      "loss": 0.7358,
+      "step": 7867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5592771868012663,
+      "learning_rate": 3.1615518638737534e-06,
+      "loss": 0.7646,
+      "step": 7868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5921876318572783,
+      "learning_rate": 3.160343687912618e-06,
+      "loss": 0.7408,
+      "step": 7869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5942615996605947,
+      "learning_rate": 3.159135636169284e-06,
+      "loss": 0.789,
+      "step": 7870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3807661303790821,
+      "learning_rate": 3.1579277087253202e-06,
+      "loss": 0.7747,
+      "step": 7871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5653883352760136,
+      "learning_rate": 3.156719905662289e-06,
+      "loss": 0.7542,
+      "step": 7872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6116420343139433,
+      "learning_rate": 3.1555122270617454e-06,
+      "loss": 0.7987,
+      "step": 7873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5077281141312564,
+      "learning_rate": 3.154304673005235e-06,
+      "loss": 0.7908,
+      "step": 7874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8117271494343843,
+      "learning_rate": 3.1530972435742902e-06,
+      "loss": 0.7985,
+      "step": 7875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6668986322358865,
+      "learning_rate": 3.151889938850445e-06,
+      "loss": 0.8366,
+      "step": 7876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5163693572502344,
+      "learning_rate": 3.150682758915218e-06,
+      "loss": 0.7823,
+      "step": 7877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5055714672942124,
+      "learning_rate": 3.1494757038501197e-06,
+      "loss": 0.7441,
+      "step": 7878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4541763259905094,
+      "learning_rate": 3.148268773736651e-06,
+      "loss": 0.6931,
+      "step": 7879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.526275167849469,
+      "learning_rate": 3.147061968656311e-06,
+      "loss": 0.7477,
+      "step": 7880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.543715915156076,
+      "learning_rate": 3.145855288690584e-06,
+      "loss": 0.7206,
+      "step": 7881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5064686373171996,
+      "learning_rate": 3.1446487339209455e-06,
+      "loss": 0.6337,
+      "step": 7882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5321585797141481,
+      "learning_rate": 3.1434423044288697e-06,
+      "loss": 0.7783,
+      "step": 7883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4746995412644193,
+      "learning_rate": 3.1422360002958143e-06,
+      "loss": 0.7572,
+      "step": 7884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7868770871636039,
+      "learning_rate": 3.141029821603232e-06,
+      "loss": 1.0639,
+      "step": 7885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5998398502917683,
+      "learning_rate": 3.1398237684325643e-06,
+      "loss": 0.7507,
+      "step": 7886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8283856976092014,
+      "learning_rate": 3.1386178408652524e-06,
+      "loss": 1.0737,
+      "step": 7887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5081277011004612,
+      "learning_rate": 3.137412038982719e-06,
+      "loss": 0.7063,
+      "step": 7888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.357911624773277,
+      "learning_rate": 3.1362063628663836e-06,
+      "loss": 0.7501,
+      "step": 7889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5549456230796386,
+      "learning_rate": 3.135000812597657e-06,
+      "loss": 0.7721,
+      "step": 7890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5384880021201024,
+      "learning_rate": 3.1337953882579408e-06,
+      "loss": 0.7575,
+      "step": 7891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.524105626707267,
+      "learning_rate": 3.1325900899286245e-06,
+      "loss": 0.7484,
+      "step": 7892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5309776855411286,
+      "learning_rate": 3.131384917691098e-06,
+      "loss": 0.7722,
+      "step": 7893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6527576154837786,
+      "learning_rate": 3.130179871626734e-06,
+      "loss": 0.9127,
+      "step": 7894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5437672183578086,
+      "learning_rate": 3.128974951816901e-06,
+      "loss": 0.7409,
+      "step": 7895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9700821504100932,
+      "learning_rate": 3.127770158342957e-06,
+      "loss": 0.8076,
+      "step": 7896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5465587189969936,
+      "learning_rate": 3.126565491286254e-06,
+      "loss": 0.7507,
+      "step": 7897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4487049607015645,
+      "learning_rate": 3.1253609507281326e-06,
+      "loss": 0.7157,
+      "step": 7898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7926514709129854,
+      "learning_rate": 3.1241565367499257e-06,
+      "loss": 1.065,
+      "step": 7899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5782621971278104,
+      "learning_rate": 3.122952249432959e-06,
+      "loss": 0.7845,
+      "step": 7900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.471358667778139,
+      "learning_rate": 3.1217480888585493e-06,
+      "loss": 0.7428,
+      "step": 7901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7677107957170246,
+      "learning_rate": 3.1205440551080033e-06,
+      "loss": 1.0651,
+      "step": 7902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.463420217395165,
+      "learning_rate": 3.1193401482626186e-06,
+      "loss": 0.8024,
+      "step": 7903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7944412764710764,
+      "learning_rate": 3.118136368403689e-06,
+      "loss": 0.7416,
+      "step": 7904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5235932758352553,
+      "learning_rate": 3.116932715612495e-06,
+      "loss": 0.7429,
+      "step": 7905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5985752596012477,
+      "learning_rate": 3.1157291899703097e-06,
+      "loss": 0.7441,
+      "step": 7906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.526484283278482,
+      "learning_rate": 3.114525791558398e-06,
+      "loss": 0.6608,
+      "step": 7907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.412562300352828,
+      "learning_rate": 3.1133225204580177e-06,
+      "loss": 0.7673,
+      "step": 7908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5060136241277517,
+      "learning_rate": 3.1121193767504117e-06,
+      "loss": 0.8455,
+      "step": 7909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3517749979544447,
+      "learning_rate": 3.1109163605168246e-06,
+      "loss": 0.6647,
+      "step": 7910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6161456382372785,
+      "learning_rate": 3.1097134718384846e-06,
+      "loss": 0.7827,
+      "step": 7911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6266915093773635,
+      "learning_rate": 3.1085107107966146e-06,
+      "loss": 0.8048,
+      "step": 7912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.54981550185417,
+      "learning_rate": 3.1073080774724227e-06,
+      "loss": 0.7779,
+      "step": 7913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5224881589053962,
+      "learning_rate": 3.10610557194712e-06,
+      "loss": 0.7119,
+      "step": 7914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5319787841857366,
+      "learning_rate": 3.1049031943019004e-06,
+      "loss": 0.7334,
+      "step": 7915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5283507870583308,
+      "learning_rate": 3.1037009446179483e-06,
+      "loss": 0.8448,
+      "step": 7916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.460622164274087,
+      "learning_rate": 3.102498822976446e-06,
+      "loss": 0.7195,
+      "step": 7917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4613912704618879,
+      "learning_rate": 3.101296829458562e-06,
+      "loss": 0.7924,
+      "step": 7918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5049396548718912,
+      "learning_rate": 3.100094964145458e-06,
+      "loss": 0.8111,
+      "step": 7919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4902887471820265,
+      "learning_rate": 3.098893227118285e-06,
+      "loss": 0.7943,
+      "step": 7920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5499880363518441,
+      "learning_rate": 3.097691618458189e-06,
+      "loss": 0.7797,
+      "step": 7921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5118413602558276,
+      "learning_rate": 3.0964901382463052e-06,
+      "loss": 0.8541,
+      "step": 7922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.552292552176281,
+      "learning_rate": 3.0952887865637593e-06,
+      "loss": 0.6696,
+      "step": 7923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.332166227487138,
+      "learning_rate": 3.0940875634916713e-06,
+      "loss": 0.7338,
+      "step": 7924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.573430330136126,
+      "learning_rate": 3.092886469111149e-06,
+      "loss": 0.7506,
+      "step": 7925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.546445582200416,
+      "learning_rate": 3.0916855035032905e-06,
+      "loss": 0.7046,
+      "step": 7926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4407201801834755,
+      "learning_rate": 3.090484666749193e-06,
+      "loss": 0.7563,
+      "step": 7927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.426430457258376,
+      "learning_rate": 3.089283958929938e-06,
+      "loss": 0.7309,
+      "step": 7928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4707885951986526,
+      "learning_rate": 3.088083380126598e-06,
+      "loss": 0.7737,
+      "step": 7929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.448299746735942,
+      "learning_rate": 3.0868829304202386e-06,
+      "loss": 0.8123,
+      "step": 7930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.495935184487191,
+      "learning_rate": 3.0856826098919196e-06,
+      "loss": 0.7502,
+      "step": 7931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.522750990225869,
+      "learning_rate": 3.0844824186226885e-06,
+      "loss": 0.7328,
+      "step": 7932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.579226713579999,
+      "learning_rate": 3.0832823566935833e-06,
+      "loss": 0.7387,
+      "step": 7933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4672088725375034,
+      "learning_rate": 3.0820824241856377e-06,
+      "loss": 0.7566,
+      "step": 7934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5964219563312856,
+      "learning_rate": 3.0808826211798725e-06,
+      "loss": 0.7963,
+      "step": 7935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.646035001421369,
+      "learning_rate": 3.0796829477573004e-06,
+      "loss": 0.7293,
+      "step": 7936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.502327853053212,
+      "learning_rate": 3.0784834039989253e-06,
+      "loss": 0.7537,
+      "step": 7937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4393912934746202,
+      "learning_rate": 3.0772839899857465e-06,
+      "loss": 0.6871,
+      "step": 7938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4833373647805088,
+      "learning_rate": 3.0760847057987486e-06,
+      "loss": 0.7798,
+      "step": 7939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5492116933048001,
+      "learning_rate": 3.0748855515189104e-06,
+      "loss": 0.7394,
+      "step": 7940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.659942606350061,
+      "learning_rate": 3.0736865272272024e-06,
+      "loss": 0.7638,
+      "step": 7941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.507731271327927,
+      "learning_rate": 3.072487633004585e-06,
+      "loss": 0.6941,
+      "step": 7942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6236872096873152,
+      "learning_rate": 3.0712888689320107e-06,
+      "loss": 0.804,
+      "step": 7943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4744118491107836,
+      "learning_rate": 3.0700902350904207e-06,
+      "loss": 0.6945,
+      "step": 7944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.525229554007588,
+      "learning_rate": 3.068891731560751e-06,
+      "loss": 0.7221,
+      "step": 7945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3962622446850879,
+      "learning_rate": 3.0676933584239287e-06,
+      "loss": 0.7005,
+      "step": 7946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8230611332684528,
+      "learning_rate": 3.0664951157608676e-06,
+      "loss": 1.0604,
+      "step": 7947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7833110891588118,
+      "learning_rate": 3.0652970036524787e-06,
+      "loss": 1.0987,
+      "step": 7948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4116601381691856,
+      "learning_rate": 3.064099022179661e-06,
+      "loss": 0.7037,
+      "step": 7949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5727377663560629,
+      "learning_rate": 3.0629011714233014e-06,
+      "loss": 0.7379,
+      "step": 7950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4384872899127135,
+      "learning_rate": 3.0617034514642865e-06,
+      "loss": 0.6701,
+      "step": 7951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5428286051017324,
+      "learning_rate": 3.060505862383486e-06,
+      "loss": 0.7751,
+      "step": 7952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7745107110449517,
+      "learning_rate": 3.059308404261765e-06,
+      "loss": 1.0884,
+      "step": 7953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.553500752873653,
+      "learning_rate": 3.058111077179976e-06,
+      "loss": 0.7585,
+      "step": 7954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4710626893173846,
+      "learning_rate": 3.0569138812189696e-06,
+      "loss": 0.7564,
+      "step": 7955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.640461977181579,
+      "learning_rate": 3.0557168164595817e-06,
+      "loss": 0.796,
+      "step": 7956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7857773527466386,
+      "learning_rate": 3.0545198829826383e-06,
+      "loss": 1.0679,
+      "step": 7957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4531885836931626,
+      "learning_rate": 3.0533230808689617e-06,
+      "loss": 0.7679,
+      "step": 7958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5562255501840483,
+      "learning_rate": 3.052126410199363e-06,
+      "loss": 0.8269,
+      "step": 7959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4454588087544018,
+      "learning_rate": 3.0509298710546433e-06,
+      "loss": 0.7381,
+      "step": 7960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5665970027994078,
+      "learning_rate": 3.0497334635155933e-06,
+      "loss": 0.7586,
+      "step": 7961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.501899125504199,
+      "learning_rate": 3.048537187663001e-06,
+      "loss": 0.8081,
+      "step": 7962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.446606358272631,
+      "learning_rate": 3.047341043577641e-06,
+      "loss": 0.7886,
+      "step": 7963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4067109821814765,
+      "learning_rate": 3.046145031340275e-06,
+      "loss": 0.8269,
+      "step": 7964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5882473964709063,
+      "learning_rate": 3.0449491510316675e-06,
+      "loss": 0.8358,
+      "step": 7965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.576854360524838,
+      "learning_rate": 3.0437534027325634e-06,
+      "loss": 0.7708,
+      "step": 7966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6486312235370708,
+      "learning_rate": 3.042557786523702e-06,
+      "loss": 0.7835,
+      "step": 7967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4881331326129894,
+      "learning_rate": 3.041362302485816e-06,
+      "loss": 0.7252,
+      "step": 7968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4472713096002712,
+      "learning_rate": 3.040166950699626e-06,
+      "loss": 0.7317,
+      "step": 7969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7757253603079295,
+      "learning_rate": 3.0389717312458446e-06,
+      "loss": 1.0743,
+      "step": 7970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6038455468368156,
+      "learning_rate": 3.0377766442051738e-06,
+      "loss": 0.8285,
+      "step": 7971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4448625431881938,
+      "learning_rate": 3.036581689658314e-06,
+      "loss": 0.8025,
+      "step": 7972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.810402769205492,
+      "learning_rate": 3.0353868676859477e-06,
+      "loss": 1.07,
+      "step": 7973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7733993936148941,
+      "learning_rate": 3.03419217836875e-06,
+      "loss": 1.0657,
+      "step": 7974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6119926668131304,
+      "learning_rate": 3.0329976217873935e-06,
+      "loss": 0.7354,
+      "step": 7975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.799427800416381,
+      "learning_rate": 3.0318031980225348e-06,
+      "loss": 0.6981,
+      "step": 7976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5165070043751696,
+      "learning_rate": 3.0306089071548263e-06,
+      "loss": 0.7207,
+      "step": 7977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.668444328082406,
+      "learning_rate": 3.029414749264905e-06,
+      "loss": 0.7737,
+      "step": 7978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.760588730356039,
+      "learning_rate": 3.0282207244334084e-06,
+      "loss": 1.0687,
+      "step": 7979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5720996746761045,
+      "learning_rate": 3.027026832740956e-06,
+      "loss": 0.6732,
+      "step": 7980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4199344182943525,
+      "learning_rate": 3.025833074268162e-06,
+      "loss": 0.7457,
+      "step": 7981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.54652241029616,
+      "learning_rate": 3.0246394490956343e-06,
+      "loss": 0.7702,
+      "step": 7982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.588834779800411,
+      "learning_rate": 3.0234459573039687e-06,
+      "loss": 0.7424,
+      "step": 7983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6041640573449811,
+      "learning_rate": 3.0222525989737517e-06,
+      "loss": 0.7297,
+      "step": 7984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5020243379532323,
+      "learning_rate": 3.02105937418556e-06,
+      "loss": 0.7073,
+      "step": 7985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.415189778987335,
+      "learning_rate": 3.019866283019966e-06,
+      "loss": 0.7708,
+      "step": 7986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5287471209405323,
+      "learning_rate": 3.0186733255575286e-06,
+      "loss": 0.7631,
+      "step": 7987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.562657859394054,
+      "learning_rate": 3.0174805018787973e-06,
+      "loss": 0.7513,
+      "step": 7988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5489141501117913,
+      "learning_rate": 3.016287812064317e-06,
+      "loss": 0.8214,
+      "step": 7989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4745590905766002,
+      "learning_rate": 3.01509525619462e-06,
+      "loss": 0.7385,
+      "step": 7990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0232713081072506,
+      "learning_rate": 3.0139028343502273e-06,
+      "loss": 0.7447,
+      "step": 7991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.588019227384956,
+      "learning_rate": 3.012710546611659e-06,
+      "loss": 0.749,
+      "step": 7992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5036644984630219,
+      "learning_rate": 3.0115183930594194e-06,
+      "loss": 0.7523,
+      "step": 7993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5591825829868644,
+      "learning_rate": 3.010326373774004e-06,
+      "loss": 0.7508,
+      "step": 7994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7143242545343365,
+      "learning_rate": 3.0091344888359015e-06,
+      "loss": 0.8813,
+      "step": 7995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4488289057126014,
+      "learning_rate": 3.007942738325591e-06,
+      "loss": 0.8276,
+      "step": 7996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.589684805111971,
+      "learning_rate": 3.0067511223235425e-06,
+      "loss": 0.7971,
+      "step": 7997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8131730114975364,
+      "learning_rate": 3.005559640910213e-06,
+      "loss": 1.0739,
+      "step": 7998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5280893490062593,
+      "learning_rate": 3.0043682941660603e-06,
+      "loss": 0.7926,
+      "step": 7999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5108552453737845,
+      "learning_rate": 3.0031770821715233e-06,
+      "loss": 0.7158,
+      "step": 8000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.623882738392162,
+      "learning_rate": 3.001986005007036e-06,
+      "loss": 0.8142,
+      "step": 8001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.478763871069111,
+      "learning_rate": 3.0007950627530197e-06,
+      "loss": 0.7055,
+      "step": 8002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5222860936784044,
+      "learning_rate": 2.999604255489894e-06,
+      "loss": 0.7067,
+      "step": 8003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4276697848398863,
+      "learning_rate": 2.9984135832980643e-06,
+      "loss": 0.714,
+      "step": 8004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7910228466134466,
+      "learning_rate": 2.9972230462579243e-06,
+      "loss": 1.0845,
+      "step": 8005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.618445048701098,
+      "learning_rate": 2.996032644449865e-06,
+      "loss": 0.7991,
+      "step": 8006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7931163516780105,
+      "learning_rate": 2.994842377954264e-06,
+      "loss": 1.0902,
+      "step": 8007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.764687241304576,
+      "learning_rate": 2.9936522468514888e-06,
+      "loss": 1.0421,
+      "step": 8008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6089082815114828,
+      "learning_rate": 2.9924622512219037e-06,
+      "loss": 0.7605,
+      "step": 8009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4993213005400576,
+      "learning_rate": 2.991272391145858e-06,
+      "loss": 0.7367,
+      "step": 8010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7604130166650378,
+      "learning_rate": 2.990082666703693e-06,
+      "loss": 1.0726,
+      "step": 8011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5490166088527242,
+      "learning_rate": 2.988893077975742e-06,
+      "loss": 0.7473,
+      "step": 8012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.555747147864406,
+      "learning_rate": 2.98770362504233e-06,
+      "loss": 0.8198,
+      "step": 8013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5837257708041734,
+      "learning_rate": 2.986514307983771e-06,
+      "loss": 0.7899,
+      "step": 8014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4624883457253859,
+      "learning_rate": 2.9853251268803674e-06,
+      "loss": 0.7189,
+      "step": 8015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7786957789800643,
+      "learning_rate": 2.984136081812421e-06,
+      "loss": 1.0359,
+      "step": 8016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4434296628097525,
+      "learning_rate": 2.9829471728602156e-06,
+      "loss": 0.7572,
+      "step": 8017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7852741949858848,
+      "learning_rate": 2.981758400104028e-06,
+      "loss": 0.7421,
+      "step": 8018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5922236010076183,
+      "learning_rate": 2.9805697636241278e-06,
+      "loss": 0.7934,
+      "step": 8019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5705563145882944,
+      "learning_rate": 2.9793812635007757e-06,
+      "loss": 0.7692,
+      "step": 8020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7982703978316215,
+      "learning_rate": 2.9781928998142217e-06,
+      "loss": 1.0826,
+      "step": 8021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4928965551209985,
+      "learning_rate": 2.9770046726447056e-06,
+      "loss": 0.7806,
+      "step": 8022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6679567521782042,
+      "learning_rate": 2.97581658207246e-06,
+      "loss": 0.7589,
+      "step": 8023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.525837198328867,
+      "learning_rate": 2.9746286281777075e-06,
+      "loss": 0.7176,
+      "step": 8024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5255339415244606,
+      "learning_rate": 2.97344081104066e-06,
+      "loss": 0.7721,
+      "step": 8025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5164366359408274,
+      "learning_rate": 2.9722531307415243e-06,
+      "loss": 0.7384,
+      "step": 8026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4379423010428263,
+      "learning_rate": 2.9710655873604943e-06,
+      "loss": 0.7035,
+      "step": 8027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4489583842644578,
+      "learning_rate": 2.969878180977755e-06,
+      "loss": 0.7461,
+      "step": 8028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.463566301609115,
+      "learning_rate": 2.968690911673482e-06,
+      "loss": 0.7532,
+      "step": 8029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3844194298239825,
+      "learning_rate": 2.967503779527845e-06,
+      "loss": 0.812,
+      "step": 8030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5572438051693611,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 0.7314,
+      "step": 8031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.556167802016975,
+      "learning_rate": 2.9651299270330945e-06,
+      "loss": 0.8426,
+      "step": 8032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6628684489413896,
+      "learning_rate": 2.9639432068442716e-06,
+      "loss": 0.7973,
+      "step": 8033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4421542175244941,
+      "learning_rate": 2.9627566241346584e-06,
+      "loss": 0.7239,
+      "step": 8034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5037599998621216,
+      "learning_rate": 2.9615701789843766e-06,
+      "loss": 0.7942,
+      "step": 8035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5272042101112238,
+      "learning_rate": 2.960383871473535e-06,
+      "loss": 0.7681,
+      "step": 8036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6937994697015775,
+      "learning_rate": 2.9591977016822406e-06,
+      "loss": 0.7941,
+      "step": 8037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7828494940349162,
+      "learning_rate": 2.9580116696905836e-06,
+      "loss": 1.0648,
+      "step": 8038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6702173253719916,
+      "learning_rate": 2.9568257755786474e-06,
+      "loss": 0.8345,
+      "step": 8039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7880414144463519,
+      "learning_rate": 2.955640019426508e-06,
+      "loss": 1.0751,
+      "step": 8040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4751484888204602,
+      "learning_rate": 2.9544544013142284e-06,
+      "loss": 0.7138,
+      "step": 8041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7655111271691917,
+      "learning_rate": 2.9532689213218657e-06,
+      "loss": 1.0534,
+      "step": 8042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6527153357583906,
+      "learning_rate": 2.9520835795294633e-06,
+      "loss": 0.7601,
+      "step": 8043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.408934616040001,
+      "learning_rate": 2.950898376017064e-06,
+      "loss": 0.7772,
+      "step": 8044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6057430189054687,
+      "learning_rate": 2.9497133108646903e-06,
+      "loss": 0.7291,
+      "step": 8045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5391777956132795,
+      "learning_rate": 2.94852838415236e-06,
+      "loss": 0.7751,
+      "step": 8046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.482991698080346,
+      "learning_rate": 2.9473435959600864e-06,
+      "loss": 0.7328,
+      "step": 8047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.424944085867061,
+      "learning_rate": 2.946158946367867e-06,
+      "loss": 0.7612,
+      "step": 8048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.471935064043554,
+      "learning_rate": 2.944974435455691e-06,
+      "loss": 0.8058,
+      "step": 8049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3962850596539222,
+      "learning_rate": 2.943790063303541e-06,
+      "loss": 0.721,
+      "step": 8050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5386898018583883,
+      "learning_rate": 2.942605829991387e-06,
+      "loss": 0.7419,
+      "step": 8051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.024697776802329,
+      "learning_rate": 2.9414217355991937e-06,
+      "loss": 0.759,
+      "step": 8052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6754737566713902,
+      "learning_rate": 2.9402377802069086e-06,
+      "loss": 0.7472,
+      "step": 8053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6139298978485828,
+      "learning_rate": 2.939053963894481e-06,
+      "loss": 0.735,
+      "step": 8054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4135510875996606,
+      "learning_rate": 2.9378702867418423e-06,
+      "loss": 0.7687,
+      "step": 8055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8389571024014597,
+      "learning_rate": 2.936686748828916e-06,
+      "loss": 1.0542,
+      "step": 8056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4783089638901947,
+      "learning_rate": 2.93550335023562e-06,
+      "loss": 0.7758,
+      "step": 8057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4453713985943246,
+      "learning_rate": 2.934320091041858e-06,
+      "loss": 0.6687,
+      "step": 8058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.650752634724723,
+      "learning_rate": 2.933136971327527e-06,
+      "loss": 0.8179,
+      "step": 8059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4788610607120913,
+      "learning_rate": 2.9319539911725136e-06,
+      "loss": 0.7814,
+      "step": 8060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8018156660344682,
+      "learning_rate": 2.930771150656696e-06,
+      "loss": 1.0852,
+      "step": 8061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7635886580720078,
+      "learning_rate": 2.9295884498599415e-06,
+      "loss": 1.098,
+      "step": 8062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5338503050890298,
+      "learning_rate": 2.9284058888621076e-06,
+      "loss": 0.8019,
+      "step": 8063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4947889838382138,
+      "learning_rate": 2.9272234677430467e-06,
+      "loss": 0.7963,
+      "step": 8064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.817480720275671,
+      "learning_rate": 2.926041186582598e-06,
+      "loss": 1.0913,
+      "step": 8065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5759143418654924,
+      "learning_rate": 2.9248590454605887e-06,
+      "loss": 0.7313,
+      "step": 8066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5047420914529652,
+      "learning_rate": 2.9236770444568428e-06,
+      "loss": 0.7023,
+      "step": 8067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4161396496439687,
+      "learning_rate": 2.922495183651171e-06,
+      "loss": 0.6961,
+      "step": 8068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6218572137866627,
+      "learning_rate": 2.921313463123375e-06,
+      "loss": 0.706,
+      "step": 8069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5159348018194834,
+      "learning_rate": 2.920131882953245e-06,
+      "loss": 0.8616,
+      "step": 8070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.573608024532517,
+      "learning_rate": 2.9189504432205685e-06,
+      "loss": 0.7217,
+      "step": 8071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4657065140139776,
+      "learning_rate": 2.9177691440051158e-06,
+      "loss": 0.8178,
+      "step": 8072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.525018251852873,
+      "learning_rate": 2.9165879853866507e-06,
+      "loss": 0.8171,
+      "step": 8073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8025038659058368,
+      "learning_rate": 2.9154069674449325e-06,
+      "loss": 1.0434,
+      "step": 8074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4337853837503074,
+      "learning_rate": 2.9142260902597003e-06,
+      "loss": 0.7809,
+      "step": 8075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3672115015814426,
+      "learning_rate": 2.9130453539106917e-06,
+      "loss": 0.7298,
+      "step": 8076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4573418958682203,
+      "learning_rate": 2.9118647584776316e-06,
+      "loss": 0.7523,
+      "step": 8077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.762183195343577,
+      "learning_rate": 2.9106843040402397e-06,
+      "loss": 1.0946,
+      "step": 8078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.600462033640718,
+      "learning_rate": 2.9095039906782207e-06,
+      "loss": 0.8476,
+      "step": 8079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5349543602300637,
+      "learning_rate": 2.9083238184712713e-06,
+      "loss": 0.7075,
+      "step": 8080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4421728261897893,
+      "learning_rate": 2.9071437874990813e-06,
+      "loss": 0.7853,
+      "step": 8081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4375242494956446,
+      "learning_rate": 2.9059638978413295e-06,
+      "loss": 0.7456,
+      "step": 8082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4579617172665835,
+      "learning_rate": 2.9047841495776812e-06,
+      "loss": 0.7253,
+      "step": 8083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.58434784649837,
+      "learning_rate": 2.9036045427878e-06,
+      "loss": 0.8164,
+      "step": 8084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4823521380097855,
+      "learning_rate": 2.902425077551334e-06,
+      "loss": 0.7978,
+      "step": 8085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7937375509882071,
+      "learning_rate": 2.901245753947923e-06,
+      "loss": 1.0667,
+      "step": 8086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7805547908756333,
+      "learning_rate": 2.9000665720571987e-06,
+      "loss": 1.0998,
+      "step": 8087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4352967288936975,
+      "learning_rate": 2.8988875319587795e-06,
+      "loss": 0.7354,
+      "step": 8088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4730829019922882,
+      "learning_rate": 2.89770863373228e-06,
+      "loss": 0.8379,
+      "step": 8089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5082691511259805,
+      "learning_rate": 2.8965298774572983e-06,
+      "loss": 0.8283,
+      "step": 8090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7936058361337938,
+      "learning_rate": 2.8953512632134305e-06,
+      "loss": 1.0916,
+      "step": 8091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.452459275453735,
+      "learning_rate": 2.8941727910802587e-06,
+      "loss": 0.7518,
+      "step": 8092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5287095106052275,
+      "learning_rate": 2.8929944611373555e-06,
+      "loss": 0.8416,
+      "step": 8093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5522357219259648,
+      "learning_rate": 2.8918162734642817e-06,
+      "loss": 0.7175,
+      "step": 8094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5609251686000156,
+      "learning_rate": 2.8906382281405965e-06,
+      "loss": 0.6848,
+      "step": 8095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5483557436464859,
+      "learning_rate": 2.8894603252458407e-06,
+      "loss": 0.7585,
+      "step": 8096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5077653530175126,
+      "learning_rate": 2.88828256485955e-06,
+      "loss": 0.7659,
+      "step": 8097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8050413001609112,
+      "learning_rate": 2.8871049470612495e-06,
+      "loss": 1.0958,
+      "step": 8098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4179232028681084,
+      "learning_rate": 2.8859274719304543e-06,
+      "loss": 0.7778,
+      "step": 8099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.522486378997395,
+      "learning_rate": 2.8847501395466704e-06,
+      "loss": 0.6731,
+      "step": 8100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5360806555064173,
+      "learning_rate": 2.8835729499893915e-06,
+      "loss": 0.8208,
+      "step": 8101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5643302481905936,
+      "learning_rate": 2.8823959033381086e-06,
+      "loss": 0.8299,
+      "step": 8102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8029753036731989,
+      "learning_rate": 2.881218999672297e-06,
+      "loss": 1.0563,
+      "step": 8103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.639315037754315,
+      "learning_rate": 2.880042239071421e-06,
+      "loss": 0.792,
+      "step": 8104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4848823908975823,
+      "learning_rate": 2.8788656216149423e-06,
+      "loss": 0.8082,
+      "step": 8105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5934031461124696,
+      "learning_rate": 2.8776891473823076e-06,
+      "loss": 0.7657,
+      "step": 8106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4339192201132445,
+      "learning_rate": 2.8765128164529545e-06,
+      "loss": 0.7581,
+      "step": 8107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4744087445007026,
+      "learning_rate": 2.875336628906312e-06,
+      "loss": 0.7651,
+      "step": 8108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5704843448597468,
+      "learning_rate": 2.874160584821798e-06,
+      "loss": 0.697,
+      "step": 8109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.561811262162164,
+      "learning_rate": 2.8729846842788223e-06,
+      "loss": 0.8051,
+      "step": 8110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4279309363445127,
+      "learning_rate": 2.871808927356783e-06,
+      "loss": 0.7676,
+      "step": 8111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3665414775725098,
+      "learning_rate": 2.870633314135073e-06,
+      "loss": 0.7168,
+      "step": 8112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5455771227612614,
+      "learning_rate": 2.869457844693071e-06,
+      "loss": 0.817,
+      "step": 8113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4362621043623574,
+      "learning_rate": 2.8682825191101447e-06,
+      "loss": 0.7459,
+      "step": 8114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3286159645409736,
+      "learning_rate": 2.86710733746566e-06,
+      "loss": 0.7069,
+      "step": 8115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5059779276284988,
+      "learning_rate": 2.865932299838964e-06,
+      "loss": 0.745,
+      "step": 8116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4986009951954842,
+      "learning_rate": 2.8647574063093995e-06,
+      "loss": 0.7604,
+      "step": 8117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.547989284907312,
+      "learning_rate": 2.8635826569562974e-06,
+      "loss": 0.7855,
+      "step": 8118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.531523906634903,
+      "learning_rate": 2.862408051858979e-06,
+      "loss": 0.745,
+      "step": 8119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5414222354532996,
+      "learning_rate": 2.861233591096758e-06,
+      "loss": 0.8215,
+      "step": 8120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4633093822555236,
+      "learning_rate": 2.860059274748933e-06,
+      "loss": 0.7695,
+      "step": 8121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5445081819621311,
+      "learning_rate": 2.8588851028948008e-06,
+      "loss": 0.7854,
+      "step": 8122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5828666202552446,
+      "learning_rate": 2.857711075613642e-06,
+      "loss": 0.8145,
+      "step": 8123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4269386679196387,
+      "learning_rate": 2.8565371929847286e-06,
+      "loss": 0.7455,
+      "step": 8124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6206372828326947,
+      "learning_rate": 2.8553634550873273e-06,
+      "loss": 0.8202,
+      "step": 8125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5694970214797648,
+      "learning_rate": 2.854189862000689e-06,
+      "loss": 0.7771,
+      "step": 8126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7997892857045377,
+      "learning_rate": 2.8530164138040585e-06,
+      "loss": 1.0341,
+      "step": 8127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5309724420575004,
+      "learning_rate": 2.851843110576667e-06,
+      "loss": 0.8447,
+      "step": 8128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7750066171906114,
+      "learning_rate": 2.850669952397743e-06,
+      "loss": 1.0629,
+      "step": 8129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.581510627470217,
+      "learning_rate": 2.849496939346498e-06,
+      "loss": 0.7296,
+      "step": 8130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7801189952109374,
+      "learning_rate": 2.848324071502137e-06,
+      "loss": 1.0627,
+      "step": 8131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5320594118233466,
+      "learning_rate": 2.8471513489438553e-06,
+      "loss": 0.742,
+      "step": 8132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6318660681902477,
+      "learning_rate": 2.845978771750837e-06,
+      "loss": 0.7914,
+      "step": 8133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3414471829840864,
+      "learning_rate": 2.8448063400022573e-06,
+      "loss": 0.6917,
+      "step": 8134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.652515575054662,
+      "learning_rate": 2.84363405377728e-06,
+      "loss": 0.7461,
+      "step": 8135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4431869706906255,
+      "learning_rate": 2.842461913155064e-06,
+      "loss": 0.716,
+      "step": 8136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4804634680711033,
+      "learning_rate": 2.8412899182147536e-06,
+      "loss": 0.7168,
+      "step": 8137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5535544693176144,
+      "learning_rate": 2.8401180690354813e-06,
+      "loss": 0.7874,
+      "step": 8138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.516954098185473,
+      "learning_rate": 2.838946365696379e-06,
+      "loss": 0.8297,
+      "step": 8139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6773037317377901,
+      "learning_rate": 2.8377748082765586e-06,
+      "loss": 0.7457,
+      "step": 8140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5923397610331045,
+      "learning_rate": 2.8366033968551277e-06,
+      "loss": 0.7579,
+      "step": 8141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4632349734107177,
+      "learning_rate": 2.835432131511182e-06,
+      "loss": 0.7724,
+      "step": 8142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5635097162607379,
+      "learning_rate": 2.834261012323809e-06,
+      "loss": 0.7651,
+      "step": 8143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4533880137645105,
+      "learning_rate": 2.8330900393720846e-06,
+      "loss": 0.6694,
+      "step": 8144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6061444499352628,
+      "learning_rate": 2.8319192127350736e-06,
+      "loss": 0.754,
+      "step": 8145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5834244901299899,
+      "learning_rate": 2.830748532491837e-06,
+      "loss": 0.7604,
+      "step": 8146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4735489759314258,
+      "learning_rate": 2.8295779987214197e-06,
+      "loss": 0.7889,
+      "step": 8147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8350176600955049,
+      "learning_rate": 2.828407611502857e-06,
+      "loss": 1.0814,
+      "step": 8148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5064134109599048,
+      "learning_rate": 2.8272373709151798e-06,
+      "loss": 0.7453,
+      "step": 8149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4548728880486381,
+      "learning_rate": 2.826067277037403e-06,
+      "loss": 0.741,
+      "step": 8150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6456121299734578,
+      "learning_rate": 2.824897329948536e-06,
+      "loss": 0.7684,
+      "step": 8151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.373144213000034,
+      "learning_rate": 2.8237275297275746e-06,
+      "loss": 0.786,
+      "step": 8152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6555856073927024,
+      "learning_rate": 2.822557876453506e-06,
+      "loss": 0.7969,
+      "step": 8153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7599400883165328,
+      "learning_rate": 2.821388370205309e-06,
+      "loss": 0.8412,
+      "step": 8154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4681089543707215,
+      "learning_rate": 2.820219011061949e-06,
+      "loss": 0.6645,
+      "step": 8155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5701251199588686,
+      "learning_rate": 2.819049799102388e-06,
+      "loss": 0.777,
+      "step": 8156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.589331090769787,
+      "learning_rate": 2.8178807344055716e-06,
+      "loss": 0.7236,
+      "step": 8157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5769274221748837,
+      "learning_rate": 2.816711817050437e-06,
+      "loss": 0.8081,
+      "step": 8158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.729006588353903,
+      "learning_rate": 2.8155430471159118e-06,
+      "loss": 0.7298,
+      "step": 8159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7797051702835283,
+      "learning_rate": 2.8143744246809167e-06,
+      "loss": 1.0931,
+      "step": 8160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3948132420441235,
+      "learning_rate": 2.813205949824358e-06,
+      "loss": 0.6949,
+      "step": 8161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.550744973865652,
+      "learning_rate": 2.8120376226251343e-06,
+      "loss": 0.7048,
+      "step": 8162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.748194246072263,
+      "learning_rate": 2.810869443162133e-06,
+      "loss": 1.0665,
+      "step": 8163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4629171201849687,
+      "learning_rate": 2.809701411514233e-06,
+      "loss": 0.7515,
+      "step": 8164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5566750323722136,
+      "learning_rate": 2.8085335277603002e-06,
+      "loss": 0.7624,
+      "step": 8165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5411964772090687,
+      "learning_rate": 2.8073657919791965e-06,
+      "loss": 0.7147,
+      "step": 8166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.517073434070002,
+      "learning_rate": 2.806198204249768e-06,
+      "loss": 0.7565,
+      "step": 8167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5043865856503862,
+      "learning_rate": 2.805030764650854e-06,
+      "loss": 0.6991,
+      "step": 8168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5912143755828456,
+      "learning_rate": 2.803863473261279e-06,
+      "loss": 0.7629,
+      "step": 8169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6921732943539733,
+      "learning_rate": 2.8026963301598668e-06,
+      "loss": 0.809,
+      "step": 8170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4296992674564826,
+      "learning_rate": 2.8015293354254223e-06,
+      "loss": 0.7495,
+      "step": 8171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4711277829924374,
+      "learning_rate": 2.8003624891367426e-06,
+      "loss": 0.7512,
+      "step": 8172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5898781228872214,
+      "learning_rate": 2.799195791372619e-06,
+      "loss": 0.6506,
+      "step": 8173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4785180225336976,
+      "learning_rate": 2.7980292422118282e-06,
+      "loss": 0.7595,
+      "step": 8174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.528239024334879,
+      "learning_rate": 2.79686284173314e-06,
+      "loss": 0.7773,
+      "step": 8175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.567854146231779,
+      "learning_rate": 2.7956965900153066e-06,
+      "loss": 0.8329,
+      "step": 8176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3674798886336972,
+      "learning_rate": 2.794530487137082e-06,
+      "loss": 0.7583,
+      "step": 8177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6200146557528319,
+      "learning_rate": 2.793364533177202e-06,
+      "loss": 0.8351,
+      "step": 8178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4492268225483043,
+      "learning_rate": 2.7921987282143927e-06,
+      "loss": 0.7118,
+      "step": 8179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5704595514648871,
+      "learning_rate": 2.791033072327375e-06,
+      "loss": 0.7619,
+      "step": 8180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6974430393766118,
+      "learning_rate": 2.789867565594856e-06,
+      "loss": 0.7995,
+      "step": 8181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5076286464717512,
+      "learning_rate": 2.78870220809553e-06,
+      "loss": 0.8573,
+      "step": 8182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4714622644339659,
+      "learning_rate": 2.7875369999080897e-06,
+      "loss": 0.6753,
+      "step": 8183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5491104866815282,
+      "learning_rate": 2.7863719411112106e-06,
+      "loss": 0.8074,
+      "step": 8184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.492394774902329,
+      "learning_rate": 2.7852070317835595e-06,
+      "loss": 0.7583,
+      "step": 8185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8194717418262882,
+      "learning_rate": 2.7840422720037943e-06,
+      "loss": 1.0918,
+      "step": 8186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8445097018164702,
+      "learning_rate": 2.7828776618505615e-06,
+      "loss": 0.7856,
+      "step": 8187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4761793697113752,
+      "learning_rate": 2.7817132014024994e-06,
+      "loss": 0.6992,
+      "step": 8188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8444929918028405,
+      "learning_rate": 2.7805488907382316e-06,
+      "loss": 1.0597,
+      "step": 8189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5522741225368732,
+      "learning_rate": 2.779384729936381e-06,
+      "loss": 0.7138,
+      "step": 8190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7891231845127261,
+      "learning_rate": 2.7782207190755496e-06,
+      "loss": 1.05,
+      "step": 8191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5305681099269213,
+      "learning_rate": 2.7770568582343364e-06,
+      "loss": 0.7289,
+      "step": 8192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6186652438868623,
+      "learning_rate": 2.7758931474913255e-06,
+      "loss": 0.8066,
+      "step": 8193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5751068155615215,
+      "learning_rate": 2.7747295869250966e-06,
+      "loss": 0.7327,
+      "step": 8194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7876555822266667,
+      "learning_rate": 2.7735661766142142e-06,
+      "loss": 1.0398,
+      "step": 8195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.515210614293708,
+      "learning_rate": 2.772402916637235e-06,
+      "loss": 0.7409,
+      "step": 8196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4587790196551986,
+      "learning_rate": 2.771239807072705e-06,
+      "loss": 0.7372,
+      "step": 8197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6144380738845585,
+      "learning_rate": 2.770076847999159e-06,
+      "loss": 0.7284,
+      "step": 8198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6076516536520802,
+      "learning_rate": 2.768914039495123e-06,
+      "loss": 0.7184,
+      "step": 8199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5285417852154481,
+      "learning_rate": 2.767751381639111e-06,
+      "loss": 0.7458,
+      "step": 8200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4840940735577408,
+      "learning_rate": 2.7665888745096326e-06,
+      "loss": 0.7436,
+      "step": 8201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7960642964042137,
+      "learning_rate": 2.7654265181851797e-06,
+      "loss": 1.0708,
+      "step": 8202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5564767371495245,
+      "learning_rate": 2.764264312744236e-06,
+      "loss": 0.6758,
+      "step": 8203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.683058950135786,
+      "learning_rate": 2.7631022582652808e-06,
+      "loss": 0.7353,
+      "step": 8204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5262474585942547,
+      "learning_rate": 2.7619403548267756e-06,
+      "loss": 0.8451,
+      "step": 8205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5811241553684552,
+      "learning_rate": 2.7607786025071754e-06,
+      "loss": 0.7695,
+      "step": 8206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4453932074438092,
+      "learning_rate": 2.7596170013849243e-06,
+      "loss": 0.7489,
+      "step": 8207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4866377333543155,
+      "learning_rate": 2.758455551538456e-06,
+      "loss": 0.7872,
+      "step": 8208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.504072959743171,
+      "learning_rate": 2.7572942530461943e-06,
+      "loss": 0.7345,
+      "step": 8209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5292456581310776,
+      "learning_rate": 2.7561331059865514e-06,
+      "loss": 0.7003,
+      "step": 8210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5616341976701913,
+      "learning_rate": 2.7549721104379335e-06,
+      "loss": 0.779,
+      "step": 8211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5180479774073772,
+      "learning_rate": 2.753811266478733e-06,
+      "loss": 0.72,
+      "step": 8212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4798054285219076,
+      "learning_rate": 2.75265057418733e-06,
+      "loss": 0.7105,
+      "step": 8213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7979725167265882,
+      "learning_rate": 2.7514900336421e-06,
+      "loss": 1.0623,
+      "step": 8214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.601102345497246,
+      "learning_rate": 2.7503296449214055e-06,
+      "loss": 0.8346,
+      "step": 8215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7980074352859905,
+      "learning_rate": 2.7491694081035975e-06,
+      "loss": 1.0591,
+      "step": 8216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.457221782327802,
+      "learning_rate": 2.748009323267016e-06,
+      "loss": 0.6734,
+      "step": 8217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7886086346592107,
+      "learning_rate": 2.7468493904899958e-06,
+      "loss": 1.0976,
+      "step": 8218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4469181252921868,
+      "learning_rate": 2.745689609850859e-06,
+      "loss": 0.8045,
+      "step": 8219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6908345128966462,
+      "learning_rate": 2.744529981427911e-06,
+      "loss": 0.8419,
+      "step": 8220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5720855433326222,
+      "learning_rate": 2.7433705052994574e-06,
+      "loss": 0.771,
+      "step": 8221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.555070697716995,
+      "learning_rate": 2.742211181543788e-06,
+      "loss": 0.7575,
+      "step": 8222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.423684606578331,
+      "learning_rate": 2.74105201023918e-06,
+      "loss": 0.6689,
+      "step": 8223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5883049216829839,
+      "learning_rate": 2.7398929914639084e-06,
+      "loss": 0.8218,
+      "step": 8224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5323578029369596,
+      "learning_rate": 2.7387341252962296e-06,
+      "loss": 0.7153,
+      "step": 8225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4825194585710566,
+      "learning_rate": 2.737575411814393e-06,
+      "loss": 0.7379,
+      "step": 8226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7805187715625739,
+      "learning_rate": 2.7364168510966367e-06,
+      "loss": 1.0679,
+      "step": 8227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7444426229544738,
+      "learning_rate": 2.735258443221192e-06,
+      "loss": 0.7455,
+      "step": 8228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5966553775158707,
+      "learning_rate": 2.734100188266276e-06,
+      "loss": 0.791,
+      "step": 8229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8784727080919141,
+      "learning_rate": 2.7329420863100963e-06,
+      "loss": 0.8027,
+      "step": 8230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.544888434203259,
+      "learning_rate": 2.731784137430852e-06,
+      "loss": 0.7925,
+      "step": 8231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5870905087155576,
+      "learning_rate": 2.730626341706728e-06,
+      "loss": 0.8558,
+      "step": 8232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7658922329321347,
+      "learning_rate": 2.729468699215903e-06,
+      "loss": 1.0666,
+      "step": 8233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5184427447612903,
+      "learning_rate": 2.728311210036542e-06,
+      "loss": 0.7758,
+      "step": 8234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.563971372833978,
+      "learning_rate": 2.727153874246804e-06,
+      "loss": 0.696,
+      "step": 8235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5323333595400572,
+      "learning_rate": 2.7259966919248336e-06,
+      "loss": 0.756,
+      "step": 8236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5050093458549125,
+      "learning_rate": 2.724839663148764e-06,
+      "loss": 0.6553,
+      "step": 8237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5745528401866362,
+      "learning_rate": 2.7236827879967255e-06,
+      "loss": 0.7865,
+      "step": 8238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4632618788397818,
+      "learning_rate": 2.72252606654683e-06,
+      "loss": 0.7203,
+      "step": 8239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6619870972870103,
+      "learning_rate": 2.7213694988771822e-06,
+      "loss": 0.8399,
+      "step": 8240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5055542259564336,
+      "learning_rate": 2.7202130850658765e-06,
+      "loss": 0.7986,
+      "step": 8241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.53580668132139,
+      "learning_rate": 2.7190568251909965e-06,
+      "loss": 0.7862,
+      "step": 8242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5916724700827931,
+      "learning_rate": 2.717900719330615e-06,
+      "loss": 0.7714,
+      "step": 8243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5399004486392494,
+      "learning_rate": 2.7167447675627933e-06,
+      "loss": 0.8145,
+      "step": 8244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5192641179967277,
+      "learning_rate": 2.715588969965588e-06,
+      "loss": 0.6827,
+      "step": 8245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5963066457127921,
+      "learning_rate": 2.7144333266170387e-06,
+      "loss": 0.7688,
+      "step": 8246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6091906856430578,
+      "learning_rate": 2.7132778375951752e-06,
+      "loss": 0.7697,
+      "step": 8247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.577973245021502,
+      "learning_rate": 2.712122502978024e-06,
+      "loss": 0.6942,
+      "step": 8248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.534652578689979,
+      "learning_rate": 2.7109673228435925e-06,
+      "loss": 0.7409,
+      "step": 8249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4787204655968955,
+      "learning_rate": 2.7098122972698815e-06,
+      "loss": 0.7318,
+      "step": 8250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5265301727902316,
+      "learning_rate": 2.7086574263348808e-06,
+      "loss": 0.7398,
+      "step": 8251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4177665504940071,
+      "learning_rate": 2.7075027101165706e-06,
+      "loss": 0.78,
+      "step": 8252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4927707994002868,
+      "learning_rate": 2.7063481486929187e-06,
+      "loss": 0.7381,
+      "step": 8253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4796579131683236,
+      "learning_rate": 2.7051937421418834e-06,
+      "loss": 0.7176,
+      "step": 8254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5919576528685537,
+      "learning_rate": 2.7040394905414156e-06,
+      "loss": 0.7548,
+      "step": 8255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6342513141230028,
+      "learning_rate": 2.7028853939694523e-06,
+      "loss": 0.7374,
+      "step": 8256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.016111651371093,
+      "learning_rate": 2.7017314525039186e-06,
+      "loss": 0.8017,
+      "step": 8257
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5204705836421444,
+      "learning_rate": 2.7005776662227312e-06,
+      "loss": 0.7902,
+      "step": 8258
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5035135609806296,
+      "learning_rate": 2.699424035203799e-06,
+      "loss": 0.77,
+      "step": 8259
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6395677048161992,
+      "learning_rate": 2.6982705595250182e-06,
+      "loss": 0.7349,
+      "step": 8260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.445221524756905,
+      "learning_rate": 2.6971172392642687e-06,
+      "loss": 0.7155,
+      "step": 8261
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.450805480445973,
+      "learning_rate": 2.695964074499432e-06,
+      "loss": 0.7328,
+      "step": 8262
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5278579527763616,
+      "learning_rate": 2.6948110653083715e-06,
+      "loss": 0.7313,
+      "step": 8263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4622310593685828,
+      "learning_rate": 2.6936582117689347e-06,
+      "loss": 0.7346,
+      "step": 8264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5642189316506692,
+      "learning_rate": 2.6925055139589705e-06,
+      "loss": 0.7728,
+      "step": 8265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5571979856865035,
+      "learning_rate": 2.6913529719563116e-06,
+      "loss": 0.7956,
+      "step": 8266
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.510178467134776,
+      "learning_rate": 2.6902005858387786e-06,
+      "loss": 0.7957,
+      "step": 8267
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5836062330763,
+      "learning_rate": 2.6890483556841817e-06,
+      "loss": 0.7863,
+      "step": 8268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6909240733681785,
+      "learning_rate": 2.6878962815703264e-06,
+      "loss": 0.8109,
+      "step": 8269
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7967985927418983,
+      "learning_rate": 2.6867443635750013e-06,
+      "loss": 1.0858,
+      "step": 8270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4660835305500082,
+      "learning_rate": 2.6855926017759837e-06,
+      "loss": 0.7128,
+      "step": 8271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4920862461130173,
+      "learning_rate": 2.6844409962510476e-06,
+      "loss": 0.6411,
+      "step": 8272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5219091682879402,
+      "learning_rate": 2.683289547077951e-06,
+      "loss": 0.7551,
+      "step": 8273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.579719795126976,
+      "learning_rate": 2.6821382543344414e-06,
+      "loss": 0.7064,
+      "step": 8274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6479014830592709,
+      "learning_rate": 2.680987118098257e-06,
+      "loss": 0.7871,
+      "step": 8275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6253991839981268,
+      "learning_rate": 2.679836138447125e-06,
+      "loss": 0.7555,
+      "step": 8276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4378880832289318,
+      "learning_rate": 2.678685315458763e-06,
+      "loss": 0.7123,
+      "step": 8277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6588005805879444,
+      "learning_rate": 2.6775346492108735e-06,
+      "loss": 0.7232,
+      "step": 8278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4767391863037627,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.7918,
+      "step": 8279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4449676568939362,
+      "learning_rate": 2.6752337872472977e-06,
+      "loss": 0.7551,
+      "step": 8280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5094335025251526,
+      "learning_rate": 2.674083591686967e-06,
+      "loss": 0.7847,
+      "step": 8281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5681406851517012,
+      "learning_rate": 2.6729335531778324e-06,
+      "loss": 0.7206,
+      "step": 8282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5116764038853105,
+      "learning_rate": 2.6717836717975448e-06,
+      "loss": 0.8246,
+      "step": 8283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7928496526314607,
+      "learning_rate": 2.670633947623748e-06,
+      "loss": 1.1011,
+      "step": 8284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4471055535164516,
+      "learning_rate": 2.669484380734073e-06,
+      "loss": 0.6978,
+      "step": 8285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5198811313458036,
+      "learning_rate": 2.6683349712061422e-06,
+      "loss": 0.7794,
+      "step": 8286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5610014070753564,
+      "learning_rate": 2.667185719117566e-06,
+      "loss": 0.764,
+      "step": 8287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5042677962157927,
+      "learning_rate": 2.6660366245459422e-06,
+      "loss": 0.7353,
+      "step": 8288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7565998165719525,
+      "learning_rate": 2.664887687568864e-06,
+      "loss": 1.0693,
+      "step": 8289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5589997059017977,
+      "learning_rate": 2.6637389082639085e-06,
+      "loss": 0.7478,
+      "step": 8290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.56216535414835,
+      "learning_rate": 2.662590286708645e-06,
+      "loss": 0.7665,
+      "step": 8291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5960946587269664,
+      "learning_rate": 2.661441822980628e-06,
+      "loss": 0.7657,
+      "step": 8292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4281754044600463,
+      "learning_rate": 2.6602935171574086e-06,
+      "loss": 0.7204,
+      "step": 8293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8872982471342359,
+      "learning_rate": 2.6591453693165204e-06,
+      "loss": 0.7338,
+      "step": 8294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5625320189673453,
+      "learning_rate": 2.6579973795354897e-06,
+      "loss": 0.7525,
+      "step": 8295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.784255830926654,
+      "learning_rate": 2.6568495478918312e-06,
+      "loss": 1.0623,
+      "step": 8296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6501774851672413,
+      "learning_rate": 2.6557018744630493e-06,
+      "loss": 0.7608,
+      "step": 8297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5589551295589876,
+      "learning_rate": 2.6545543593266376e-06,
+      "loss": 0.6947,
+      "step": 8298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.832161967403847,
+      "learning_rate": 2.6534070025600765e-06,
+      "loss": 0.7137,
+      "step": 8299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5870021199922453,
+      "learning_rate": 2.6522598042408423e-06,
+      "loss": 0.8202,
+      "step": 8300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.490317885698625,
+      "learning_rate": 2.6511127644463945e-06,
+      "loss": 0.7318,
+      "step": 8301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4310934536967337,
+      "learning_rate": 2.6499658832541824e-06,
+      "loss": 0.739,
+      "step": 8302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5615871903691583,
+      "learning_rate": 2.6488191607416493e-06,
+      "loss": 0.8113,
+      "step": 8303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7743913799613471,
+      "learning_rate": 2.6476725969862227e-06,
+      "loss": 1.0903,
+      "step": 8304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7677416170653298,
+      "learning_rate": 2.6465261920653195e-06,
+      "loss": 1.0456,
+      "step": 8305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4817362529164761,
+      "learning_rate": 2.645379946056351e-06,
+      "loss": 0.7281,
+      "step": 8306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7509586091654735,
+      "learning_rate": 2.6442338590367144e-06,
+      "loss": 1.0464,
+      "step": 8307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.668289510873252,
+      "learning_rate": 2.6430879310837933e-06,
+      "loss": 0.7898,
+      "step": 8308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5298336616719217,
+      "learning_rate": 2.641942162274962e-06,
+      "loss": 0.7664,
+      "step": 8309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4842557853779859,
+      "learning_rate": 2.64079655268759e-06,
+      "loss": 0.8014,
+      "step": 8310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5453234378926828,
+      "learning_rate": 2.639651102399029e-06,
+      "loss": 0.7824,
+      "step": 8311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5036011642971612,
+      "learning_rate": 2.6385058114866215e-06,
+      "loss": 0.8458,
+      "step": 8312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7660021211901055,
+      "learning_rate": 2.637360680027703e-06,
+      "loss": 1.0263,
+      "step": 8313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5411658763867058,
+      "learning_rate": 2.6362157080995942e-06,
+      "loss": 0.8327,
+      "step": 8314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.673054584840367,
+      "learning_rate": 2.6350708957796057e-06,
+      "loss": 0.6909,
+      "step": 8315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4596427912866035,
+      "learning_rate": 2.6339262431450365e-06,
+      "loss": 0.7661,
+      "step": 8316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5173635836734376,
+      "learning_rate": 2.632781750273179e-06,
+      "loss": 0.7384,
+      "step": 8317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5244123972029278,
+      "learning_rate": 2.6316374172413117e-06,
+      "loss": 0.7634,
+      "step": 8318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8069214500790183,
+      "learning_rate": 2.6304932441267006e-06,
+      "loss": 1.0546,
+      "step": 8319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4540523948550046,
+      "learning_rate": 2.6293492310066042e-06,
+      "loss": 0.7539,
+      "step": 8320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5434795724811048,
+      "learning_rate": 2.6282053779582683e-06,
+      "loss": 0.8189,
+      "step": 8321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5260802143096122,
+      "learning_rate": 2.627061685058927e-06,
+      "loss": 0.815,
+      "step": 8322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5315500003669564,
+      "learning_rate": 2.625918152385809e-06,
+      "loss": 0.7225,
+      "step": 8323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5458235969596263,
+      "learning_rate": 2.624774780016126e-06,
+      "loss": 0.8488,
+      "step": 8324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4135513204108237,
+      "learning_rate": 2.623631568027081e-06,
+      "loss": 0.7162,
+      "step": 8325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.768973723695802,
+      "learning_rate": 2.6224885164958642e-06,
+      "loss": 1.0803,
+      "step": 8326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5270364210756986,
+      "learning_rate": 2.6213456254996618e-06,
+      "loss": 0.8194,
+      "step": 8327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5215604022027092,
+      "learning_rate": 2.620202895115641e-06,
+      "loss": 0.7598,
+      "step": 8328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.547442459952091,
+      "learning_rate": 2.6190603254209627e-06,
+      "loss": 0.7657,
+      "step": 8329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4856411745327227,
+      "learning_rate": 2.617917916492776e-06,
+      "loss": 0.8341,
+      "step": 8330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4964497773145318,
+      "learning_rate": 2.6167756684082185e-06,
+      "loss": 0.7102,
+      "step": 8331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4885449564797126,
+      "learning_rate": 2.6156335812444174e-06,
+      "loss": 0.7541,
+      "step": 8332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5400550427616044,
+      "learning_rate": 2.614491655078487e-06,
+      "loss": 0.7547,
+      "step": 8333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5937521163858108,
+      "learning_rate": 2.6133498899875363e-06,
+      "loss": 0.7975,
+      "step": 8334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5271851735152815,
+      "learning_rate": 2.612208286048659e-06,
+      "loss": 0.7509,
+      "step": 8335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.469116636493714,
+      "learning_rate": 2.6110668433389364e-06,
+      "loss": 0.7568,
+      "step": 8336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4835749509932519,
+      "learning_rate": 2.6099255619354446e-06,
+      "loss": 0.6861,
+      "step": 8337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4571790735830914,
+      "learning_rate": 2.6087844419152455e-06,
+      "loss": 0.7816,
+      "step": 8338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.794043100723083,
+      "learning_rate": 2.607643483355387e-06,
+      "loss": 1.1015,
+      "step": 8339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.363931779289655,
+      "learning_rate": 2.6065026863329112e-06,
+      "loss": 0.8058,
+      "step": 8340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5618708578991791,
+      "learning_rate": 2.605362050924848e-06,
+      "loss": 0.6616,
+      "step": 8341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5730651032797092,
+      "learning_rate": 2.6042215772082145e-06,
+      "loss": 0.7755,
+      "step": 8342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.529371549827531,
+      "learning_rate": 2.6030812652600156e-06,
+      "loss": 0.7998,
+      "step": 8343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.512514379994665,
+      "learning_rate": 2.601941115157254e-06,
+      "loss": 0.8056,
+      "step": 8344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5051310189785445,
+      "learning_rate": 2.600801126976911e-06,
+      "loss": 0.699,
+      "step": 8345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4986973335703448,
+      "learning_rate": 2.599661300795959e-06,
+      "loss": 0.7923,
+      "step": 8346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.471068322102205,
+      "learning_rate": 2.598521636691368e-06,
+      "loss": 0.6641,
+      "step": 8347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5932328871375554,
+      "learning_rate": 2.5973821347400875e-06,
+      "loss": 0.7619,
+      "step": 8348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6251156219109892,
+      "learning_rate": 2.5962427950190584e-06,
+      "loss": 0.7123,
+      "step": 8349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6410363656149476,
+      "learning_rate": 2.5951036176052104e-06,
+      "loss": 0.7657,
+      "step": 8350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4875845499246276,
+      "learning_rate": 2.593964602575467e-06,
+      "loss": 0.7522,
+      "step": 8351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5239718470964292,
+      "learning_rate": 2.592825750006738e-06,
+      "loss": 0.7924,
+      "step": 8352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4812332997612203,
+      "learning_rate": 2.591687059975915e-06,
+      "loss": 0.8158,
+      "step": 8353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5218891112743624,
+      "learning_rate": 2.59054853255989e-06,
+      "loss": 0.7757,
+      "step": 8354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.507942353211598,
+      "learning_rate": 2.5894101678355377e-06,
+      "loss": 0.7009,
+      "step": 8355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5377710764614008,
+      "learning_rate": 2.5882719658797235e-06,
+      "loss": 0.8184,
+      "step": 8356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5660362429775845,
+      "learning_rate": 2.5871339267692984e-06,
+      "loss": 0.7999,
+      "step": 8357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4624270774768444,
+      "learning_rate": 2.58599605058111e-06,
+      "loss": 0.6185,
+      "step": 8358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5637179607164189,
+      "learning_rate": 2.584858337391988e-06,
+      "loss": 0.8311,
+      "step": 8359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7987328072619645,
+      "learning_rate": 2.5837207872787522e-06,
+      "loss": 1.0684,
+      "step": 8360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6515606773989207,
+      "learning_rate": 2.582583400318216e-06,
+      "loss": 0.7655,
+      "step": 8361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4827644290515394,
+      "learning_rate": 2.5814461765871757e-06,
+      "loss": 0.679,
+      "step": 8362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8069585849875317,
+      "learning_rate": 2.5803091161624204e-06,
+      "loss": 1.085,
+      "step": 8363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4759057158668927,
+      "learning_rate": 2.579172219120727e-06,
+      "loss": 0.7559,
+      "step": 8364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5548702280416802,
+      "learning_rate": 2.57803548553886e-06,
+      "loss": 0.7488,
+      "step": 8365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7248796059739278,
+      "learning_rate": 2.5768989154935752e-06,
+      "loss": 0.7772,
+      "step": 8366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7501609229173506,
+      "learning_rate": 2.5757625090616147e-06,
+      "loss": 1.09,
+      "step": 8367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4836467342786714,
+      "learning_rate": 2.574626266319715e-06,
+      "loss": 0.7626,
+      "step": 8368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4901006389981988,
+      "learning_rate": 2.573490187344596e-06,
+      "loss": 0.7397,
+      "step": 8369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.435554522145888,
+      "learning_rate": 2.5723542722129655e-06,
+      "loss": 0.7614,
+      "step": 8370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.763561570946773,
+      "learning_rate": 2.5712185210015283e-06,
+      "loss": 1.0629,
+      "step": 8371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.538196871892569,
+      "learning_rate": 2.57008293378697e-06,
+      "loss": 0.8044,
+      "step": 8372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7079994563724208,
+      "learning_rate": 2.5689475106459683e-06,
+      "loss": 0.7993,
+      "step": 8373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7770459303180841,
+      "learning_rate": 2.5678122516551896e-06,
+      "loss": 1.0517,
+      "step": 8374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8001471786354353,
+      "learning_rate": 2.5666771568912892e-06,
+      "loss": 1.0881,
+      "step": 8375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4127718845825796,
+      "learning_rate": 2.565542226430911e-06,
+      "loss": 0.7723,
+      "step": 8376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.742090361151367,
+      "learning_rate": 2.564407460350687e-06,
+      "loss": 1.0694,
+      "step": 8377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.14888985342812,
+      "learning_rate": 2.5632728587272427e-06,
+      "loss": 0.7065,
+      "step": 8378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5540163599327135,
+      "learning_rate": 2.562138421637186e-06,
+      "loss": 0.7356,
+      "step": 8379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4382144428626762,
+      "learning_rate": 2.561004149157116e-06,
+      "loss": 0.7309,
+      "step": 8380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.425419481907605,
+      "learning_rate": 2.559870041363625e-06,
+      "loss": 0.7611,
+      "step": 8381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.584008873824714,
+      "learning_rate": 2.558736098333289e-06,
+      "loss": 0.7838,
+      "step": 8382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6325081216459987,
+      "learning_rate": 2.5576023201426736e-06,
+      "loss": 0.79,
+      "step": 8383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4054487823566333,
+      "learning_rate": 2.5564687068683335e-06,
+      "loss": 0.6888,
+      "step": 8384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.594617861872597,
+      "learning_rate": 2.5553352585868152e-06,
+      "loss": 0.7637,
+      "step": 8385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7350851279410724,
+      "learning_rate": 2.5542019753746496e-06,
+      "loss": 0.8218,
+      "step": 8386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7866939086364111,
+      "learning_rate": 2.5530688573083574e-06,
+      "loss": 1.0623,
+      "step": 8387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.507736134345464,
+      "learning_rate": 2.551935904464453e-06,
+      "loss": 0.6908,
+      "step": 8388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7910625718056159,
+      "learning_rate": 2.550803116919435e-06,
+      "loss": 1.0803,
+      "step": 8389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4207413857699362,
+      "learning_rate": 2.5496704947497896e-06,
+      "loss": 0.726,
+      "step": 8390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.564481164980497,
+      "learning_rate": 2.5485380380319945e-06,
+      "loss": 0.6906,
+      "step": 8391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4641500491589405,
+      "learning_rate": 2.5474057468425185e-06,
+      "loss": 0.7141,
+      "step": 8392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5793549776642106,
+      "learning_rate": 2.5462736212578144e-06,
+      "loss": 0.7888,
+      "step": 8393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4445054521993803,
+      "learning_rate": 2.545141661354324e-06,
+      "loss": 0.7213,
+      "step": 8394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6413333100615628,
+      "learning_rate": 2.5440098672084845e-06,
+      "loss": 0.7825,
+      "step": 8395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7818529195443648,
+      "learning_rate": 2.542878238896716e-06,
+      "loss": 1.078,
+      "step": 8396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5310508027526801,
+      "learning_rate": 2.541746776495426e-06,
+      "loss": 0.7304,
+      "step": 8397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4288168572009228,
+      "learning_rate": 2.5406154800810125e-06,
+      "loss": 0.7502,
+      "step": 8398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6019856997365618,
+      "learning_rate": 2.539484349729868e-06,
+      "loss": 0.7558,
+      "step": 8399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4297650128586952,
+      "learning_rate": 2.5383533855183663e-06,
+      "loss": 0.7622,
+      "step": 8400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6517568364555275,
+      "learning_rate": 2.537222587522871e-06,
+      "loss": 0.7381,
+      "step": 8401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4332846174225153,
+      "learning_rate": 2.5360919558197393e-06,
+      "loss": 0.7178,
+      "step": 8402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5532658509861086,
+      "learning_rate": 2.534961490485313e-06,
+      "loss": 0.8291,
+      "step": 8403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6270131420271687,
+      "learning_rate": 2.5338311915959224e-06,
+      "loss": 0.7386,
+      "step": 8404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6435164707318857,
+      "learning_rate": 2.53270105922789e-06,
+      "loss": 0.8023,
+      "step": 8405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5189289764570817,
+      "learning_rate": 2.5315710934575245e-06,
+      "loss": 0.7635,
+      "step": 8406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5419323268118992,
+      "learning_rate": 2.5304412943611228e-06,
+      "loss": 0.8071,
+      "step": 8407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6015092877218355,
+      "learning_rate": 2.529311662014972e-06,
+      "loss": 0.7202,
+      "step": 8408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5206682794901463,
+      "learning_rate": 2.528182196495348e-06,
+      "loss": 0.8411,
+      "step": 8409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7842379052895506,
+      "learning_rate": 2.5270528978785134e-06,
+      "loss": 1.0923,
+      "step": 8410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4502227496408504,
+      "learning_rate": 2.52592376624072e-06,
+      "loss": 0.7247,
+      "step": 8411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7753518784200064,
+      "learning_rate": 2.5247948016582137e-06,
+      "loss": 1.0518,
+      "step": 8412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.466033938988233,
+      "learning_rate": 2.5236660042072215e-06,
+      "loss": 0.7485,
+      "step": 8413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5518280510084257,
+      "learning_rate": 2.5225373739639637e-06,
+      "loss": 0.6924,
+      "step": 8414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4484107891099296,
+      "learning_rate": 2.521408911004646e-06,
+      "loss": 0.7769,
+      "step": 8415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5310596627334216,
+      "learning_rate": 2.520280615405467e-06,
+      "loss": 0.7522,
+      "step": 8416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7393061497137441,
+      "learning_rate": 2.519152487242612e-06,
+      "loss": 1.0715,
+      "step": 8417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3623725067960104,
+      "learning_rate": 2.518024526592253e-06,
+      "loss": 0.6117,
+      "step": 8418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5127601497908967,
+      "learning_rate": 2.5168967335305542e-06,
+      "loss": 0.7331,
+      "step": 8419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5497545074950307,
+      "learning_rate": 2.515769108133666e-06,
+      "loss": 0.7408,
+      "step": 8420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4343647107884339,
+      "learning_rate": 2.514641650477726e-06,
+      "loss": 0.7089,
+      "step": 8421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.549562251320844,
+      "learning_rate": 2.5135143606388667e-06,
+      "loss": 0.7641,
+      "step": 8422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7720925996545489,
+      "learning_rate": 2.5123872386932037e-06,
+      "loss": 1.0545,
+      "step": 8423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4923482491361153,
+      "learning_rate": 2.511260284716842e-06,
+      "loss": 0.7875,
+      "step": 8424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5739825655294877,
+      "learning_rate": 2.510133498785875e-06,
+      "loss": 0.7532,
+      "step": 8425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5521360627895644,
+      "learning_rate": 2.50900688097639e-06,
+      "loss": 0.7916,
+      "step": 8426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7674188205702864,
+      "learning_rate": 2.5078804313644554e-06,
+      "loss": 1.0683,
+      "step": 8427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.442193330694279,
+      "learning_rate": 2.5067541500261337e-06,
+      "loss": 0.768,
+      "step": 8428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6876083903551493,
+      "learning_rate": 2.5056280370374725e-06,
+      "loss": 0.7229,
+      "step": 8429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7419561805872419,
+      "learning_rate": 2.50450209247451e-06,
+      "loss": 1.0312,
+      "step": 8430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4555961366230348,
+      "learning_rate": 2.503376316413273e-06,
+      "loss": 0.7458,
+      "step": 8431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5293103665401389,
+      "learning_rate": 2.5022507089297733e-06,
+      "loss": 0.7573,
+      "step": 8432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8012811685674973,
+      "learning_rate": 2.5011252701000194e-06,
+      "loss": 1.0735,
+      "step": 8433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5545441177260568,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.7967,
+      "step": 8434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6674759522104003,
+      "learning_rate": 2.4988748987056976e-06,
+      "loss": 0.7803,
+      "step": 8435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6078973580268119,
+      "learning_rate": 2.497749966293082e-06,
+      "loss": 0.8005,
+      "step": 8436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7703211429283107,
+      "learning_rate": 2.4966252028381113e-06,
+      "loss": 1.1068,
+      "step": 8437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6340983282383685,
+      "learning_rate": 2.495500608416728e-06,
+      "loss": 0.7466,
+      "step": 8438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7686987341424951,
+      "learning_rate": 2.494376183104873e-06,
+      "loss": 1.0497,
+      "step": 8439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.478923287697896,
+      "learning_rate": 2.4932519269784694e-06,
+      "loss": 0.7581,
+      "step": 8440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.560121133250179,
+      "learning_rate": 2.4921278401134258e-06,
+      "loss": 0.8781,
+      "step": 8441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5172347211836246,
+      "learning_rate": 2.4910039225856432e-06,
+      "loss": 0.8111,
+      "step": 8442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5686488635766143,
+      "learning_rate": 2.489880174471015e-06,
+      "loss": 0.7695,
+      "step": 8443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5871365895248375,
+      "learning_rate": 2.488756595845417e-06,
+      "loss": 0.7528,
+      "step": 8444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7711887370630394,
+      "learning_rate": 2.4876331867847138e-06,
+      "loss": 1.0517,
+      "step": 8445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2972816748496643,
+      "learning_rate": 2.4865099473647646e-06,
+      "loss": 0.7006,
+      "step": 8446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5126279923415573,
+      "learning_rate": 2.4853868776614117e-06,
+      "loss": 0.7117,
+      "step": 8447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.44772968116204,
+      "learning_rate": 2.484263977750486e-06,
+      "loss": 0.6764,
+      "step": 8448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4427981519607018,
+      "learning_rate": 2.4831412477078076e-06,
+      "loss": 0.7628,
+      "step": 8449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4914825083414722,
+      "learning_rate": 2.4820186876091893e-06,
+      "loss": 0.8245,
+      "step": 8450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4847909841052833,
+      "learning_rate": 2.4808962975304275e-06,
+      "loss": 0.8007,
+      "step": 8451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4076458706047503,
+      "learning_rate": 2.479774077547307e-06,
+      "loss": 0.6963,
+      "step": 8452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.783742448255017,
+      "learning_rate": 2.4786520277356043e-06,
+      "loss": 1.0712,
+      "step": 8453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5460566139428884,
+      "learning_rate": 2.4775301481710817e-06,
+      "loss": 0.7643,
+      "step": 8454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7464646646832671,
+      "learning_rate": 2.476408438929491e-06,
+      "loss": 1.0744,
+      "step": 8455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5901321319181445,
+      "learning_rate": 2.475286900086572e-06,
+      "loss": 0.7829,
+      "step": 8456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9767168129808872,
+      "learning_rate": 2.4741655317180565e-06,
+      "loss": 0.7345,
+      "step": 8457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7570698920037929,
+      "learning_rate": 2.4730443338996596e-06,
+      "loss": 1.082,
+      "step": 8458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6963894154363808,
+      "learning_rate": 2.471923306707086e-06,
+      "loss": 0.7574,
+      "step": 8459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5834973729346815,
+      "learning_rate": 2.4708024502160327e-06,
+      "loss": 0.854,
+      "step": 8460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6311786190052766,
+      "learning_rate": 2.469681764502182e-06,
+      "loss": 0.7323,
+      "step": 8461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7849084836681917,
+      "learning_rate": 2.4685612496412043e-06,
+      "loss": 1.1033,
+      "step": 8462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7467702167359213,
+      "learning_rate": 2.4674409057087594e-06,
+      "loss": 1.0447,
+      "step": 8463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.576550436595146,
+      "learning_rate": 2.4663207327804954e-06,
+      "loss": 0.7911,
+      "step": 8464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5310382323759892,
+      "learning_rate": 2.4652007309320497e-06,
+      "loss": 0.7868,
+      "step": 8465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5774340045688249,
+      "learning_rate": 2.464080900239045e-06,
+      "loss": 0.7546,
+      "step": 8466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4181515965360847,
+      "learning_rate": 2.4629612407770976e-06,
+      "loss": 0.7908,
+      "step": 8467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7262651524596806,
+      "learning_rate": 2.461841752621809e-06,
+      "loss": 0.7395,
+      "step": 8468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.535478838648945,
+      "learning_rate": 2.4607224358487674e-06,
+      "loss": 0.6548,
+      "step": 8469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4550847672524254,
+      "learning_rate": 2.4596032905335554e-06,
+      "loss": 0.7865,
+      "step": 8470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5443102339033274,
+      "learning_rate": 2.4584843167517376e-06,
+      "loss": 0.6896,
+      "step": 8471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5877231578804025,
+      "learning_rate": 2.4573655145788704e-06,
+      "loss": 0.8777,
+      "step": 8472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5326015925966847,
+      "learning_rate": 2.456246884090498e-06,
+      "loss": 0.7193,
+      "step": 8473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5172982366004646,
+      "learning_rate": 2.455128425362153e-06,
+      "loss": 0.7708,
+      "step": 8474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4988523396961726,
+      "learning_rate": 2.4540101384693556e-06,
+      "loss": 0.7694,
+      "step": 8475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5581546277659775,
+      "learning_rate": 2.452892023487613e-06,
+      "loss": 0.7658,
+      "step": 8476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5210484881795254,
+      "learning_rate": 2.4517740804924272e-06,
+      "loss": 0.7288,
+      "step": 8477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4332585539019278,
+      "learning_rate": 2.4506563095592826e-06,
+      "loss": 0.7656,
+      "step": 8478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4626670396198622,
+      "learning_rate": 2.449538710763652e-06,
+      "loss": 0.7722,
+      "step": 8479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5849148088265514,
+      "learning_rate": 2.448421284181001e-06,
+      "loss": 0.7434,
+      "step": 8480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5725567050276974,
+      "learning_rate": 2.4473040298867795e-06,
+      "loss": 0.7442,
+      "step": 8481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.595503875591317,
+      "learning_rate": 2.446186947956427e-06,
+      "loss": 0.7882,
+      "step": 8482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5985716994997807,
+      "learning_rate": 2.4450700384653697e-06,
+      "loss": 0.6912,
+      "step": 8483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7577981979995358,
+      "learning_rate": 2.4439533014890295e-06,
+      "loss": 1.0222,
+      "step": 8484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4589461705886189,
+      "learning_rate": 2.442836737102805e-06,
+      "loss": 0.757,
+      "step": 8485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.507856684852319,
+      "learning_rate": 2.4417203453820892e-06,
+      "loss": 0.7563,
+      "step": 8486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4954645062697478,
+      "learning_rate": 2.4406041264022677e-06,
+      "loss": 0.7851,
+      "step": 8487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8032693927492929,
+      "learning_rate": 2.4394880802387083e-06,
+      "loss": 1.076,
+      "step": 8488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.80698150158232,
+      "learning_rate": 2.4383722069667683e-06,
+      "loss": 1.0789,
+      "step": 8489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5467302227353021,
+      "learning_rate": 2.4372565066617927e-06,
+      "loss": 0.7454,
+      "step": 8490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7707819615026256,
+      "learning_rate": 2.4361409793991193e-06,
+      "loss": 1.0569,
+      "step": 8491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4413195392362863,
+      "learning_rate": 2.4350256252540697e-06,
+      "loss": 0.6664,
+      "step": 8492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5010233227479461,
+      "learning_rate": 2.4339104443019536e-06,
+      "loss": 0.7475,
+      "step": 8493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5756744878684474,
+      "learning_rate": 2.4327954366180738e-06,
+      "loss": 0.7606,
+      "step": 8494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5741863143146801,
+      "learning_rate": 2.4316806022777164e-06,
+      "loss": 0.7262,
+      "step": 8495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5548658126181871,
+      "learning_rate": 2.430565941356157e-06,
+      "loss": 0.8298,
+      "step": 8496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.807816477028093,
+      "learning_rate": 2.4294514539286613e-06,
+      "loss": 1.0734,
+      "step": 8497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8116416036386969,
+      "learning_rate": 2.428337140070481e-06,
+      "loss": 1.0402,
+      "step": 8498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5745969821687225,
+      "learning_rate": 2.4272229998568576e-06,
+      "loss": 0.7317,
+      "step": 8499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5726699494942498,
+      "learning_rate": 2.4261090333630184e-06,
+      "loss": 0.7482,
+      "step": 8500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.451988917359058,
+      "learning_rate": 2.424995240664184e-06,
+      "loss": 0.7034,
+      "step": 8501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5500404411131714,
+      "learning_rate": 2.42388162183556e-06,
+      "loss": 0.8019,
+      "step": 8502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7518438921317464,
+      "learning_rate": 2.4227681769523374e-06,
+      "loss": 1.0785,
+      "step": 8503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.333959303166974,
+      "learning_rate": 2.4216549060897026e-06,
+      "loss": 0.7951,
+      "step": 8504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5058553984262726,
+      "learning_rate": 2.420541809322824e-06,
+      "loss": 0.6886,
+      "step": 8505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.492355377308934,
+      "learning_rate": 2.419428886726861e-06,
+      "loss": 0.6907,
+      "step": 8506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4729260067494232,
+      "learning_rate": 2.4183161383769602e-06,
+      "loss": 0.8364,
+      "step": 8507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5556655955544163,
+      "learning_rate": 2.4172035643482573e-06,
+      "loss": 0.8206,
+      "step": 8508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.520560836781945,
+      "learning_rate": 2.416091164715876e-06,
+      "loss": 0.7718,
+      "step": 8509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5774677017767569,
+      "learning_rate": 2.414978939554925e-06,
+      "loss": 0.7269,
+      "step": 8510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5159135222432951,
+      "learning_rate": 2.4138668889405094e-06,
+      "loss": 0.745,
+      "step": 8511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.561068160345222,
+      "learning_rate": 2.4127550129477145e-06,
+      "loss": 0.7685,
+      "step": 8512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7624393027756743,
+      "learning_rate": 2.4116433116516182e-06,
+      "loss": 1.0559,
+      "step": 8513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5567468937949895,
+      "learning_rate": 2.4105317851272816e-06,
+      "loss": 0.7871,
+      "step": 8514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.521450805729144,
+      "learning_rate": 2.409420433449762e-06,
+      "loss": 0.7392,
+      "step": 8515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5159581843145464,
+      "learning_rate": 2.408309256694098e-06,
+      "loss": 0.8021,
+      "step": 8516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5711501502926193,
+      "learning_rate": 2.4071982549353203e-06,
+      "loss": 0.7509,
+      "step": 8517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5086312513511886,
+      "learning_rate": 2.4060874282484444e-06,
+      "loss": 0.7463,
+      "step": 8518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6272051682731492,
+      "learning_rate": 2.404976776708477e-06,
+      "loss": 0.7636,
+      "step": 8519
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4469203522673066,
+      "learning_rate": 2.4038663003904095e-06,
+      "loss": 0.8105,
+      "step": 8520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6216967759518894,
+      "learning_rate": 2.4027559993692274e-06,
+      "loss": 0.7231,
+      "step": 8521
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7651362467872992,
+      "learning_rate": 2.4016458737198995e-06,
+      "loss": 1.0808,
+      "step": 8522
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5838897908715523,
+      "learning_rate": 2.4005359235173835e-06,
+      "loss": 0.9136,
+      "step": 8523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4381843956332023,
+      "learning_rate": 2.399426148836625e-06,
+      "loss": 0.7306,
+      "step": 8524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.510335055545418,
+      "learning_rate": 2.39831654975256e-06,
+      "loss": 0.6853,
+      "step": 8525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9490788388663494,
+      "learning_rate": 2.397207126340112e-06,
+      "loss": 0.7745,
+      "step": 8526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7754369919880774,
+      "learning_rate": 2.3960978786741878e-06,
+      "loss": 1.0842,
+      "step": 8527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5257053994226537,
+      "learning_rate": 2.3949888068296927e-06,
+      "loss": 0.7792,
+      "step": 8528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4990816262870552,
+      "learning_rate": 2.3938799108815087e-06,
+      "loss": 0.7759,
+      "step": 8529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5531678195467715,
+      "learning_rate": 2.392771190904512e-06,
+      "loss": 0.7446,
+      "step": 8530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4608258718285658,
+      "learning_rate": 2.391662646973564e-06,
+      "loss": 0.816,
+      "step": 8531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7706627486168927,
+      "learning_rate": 2.3905542791635213e-06,
+      "loss": 1.0733,
+      "step": 8532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.510353106434897,
+      "learning_rate": 2.38944608754922e-06,
+      "loss": 0.7105,
+      "step": 8533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7805814581971641,
+      "learning_rate": 2.3883380722054865e-06,
+      "loss": 1.0722,
+      "step": 8534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7294263125134057,
+      "learning_rate": 2.3872302332071403e-06,
+      "loss": 1.0666,
+      "step": 8535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4851676502927478,
+      "learning_rate": 2.3861225706289824e-06,
+      "loss": 0.7353,
+      "step": 8536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.523344766286888,
+      "learning_rate": 2.385015084545805e-06,
+      "loss": 0.7297,
+      "step": 8537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.518049648251476,
+      "learning_rate": 2.383907775032389e-06,
+      "loss": 0.6912,
+      "step": 8538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.484549568057692,
+      "learning_rate": 2.382800642163503e-06,
+      "loss": 0.7738,
+      "step": 8539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5126950228592801,
+      "learning_rate": 2.381693686013902e-06,
+      "loss": 0.7309,
+      "step": 8540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4383117943177546,
+      "learning_rate": 2.3805869066583304e-06,
+      "loss": 0.8075,
+      "step": 8541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5057406824823547,
+      "learning_rate": 2.3794803041715207e-06,
+      "loss": 0.791,
+      "step": 8542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7653343647091818,
+      "learning_rate": 2.378373878628193e-06,
+      "loss": 1.064,
+      "step": 8543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.451632365771509,
+      "learning_rate": 2.377267630103054e-06,
+      "loss": 0.7748,
+      "step": 8544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7599709411696033,
+      "learning_rate": 2.376161558670803e-06,
+      "loss": 1.0372,
+      "step": 8545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7783147970953287,
+      "learning_rate": 2.375055664406124e-06,
+      "loss": 1.0484,
+      "step": 8546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7441896809449883,
+      "learning_rate": 2.3739499473836893e-06,
+      "loss": 1.0387,
+      "step": 8547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.645004072926352,
+      "learning_rate": 2.372844407678156e-06,
+      "loss": 0.8536,
+      "step": 8548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5035448161190417,
+      "learning_rate": 2.371739045364178e-06,
+      "loss": 0.7861,
+      "step": 8549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6414008265281586,
+      "learning_rate": 2.3706338605163896e-06,
+      "loss": 0.769,
+      "step": 8550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4128759869314864,
+      "learning_rate": 2.3695288532094152e-06,
+      "loss": 0.6503,
+      "step": 8551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5156163767241455,
+      "learning_rate": 2.368424023517868e-06,
+      "loss": 0.7412,
+      "step": 8552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4267419676716042,
+      "learning_rate": 2.3673193715163477e-06,
+      "loss": 0.6289,
+      "step": 8553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7858884829867728,
+      "learning_rate": 2.3662148972794434e-06,
+      "loss": 1.0688,
+      "step": 8554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4979514006745351,
+      "learning_rate": 2.3651106008817303e-06,
+      "loss": 0.7386,
+      "step": 8555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4726667690896118,
+      "learning_rate": 2.364006482397776e-06,
+      "loss": 0.7402,
+      "step": 8556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4044393374901258,
+      "learning_rate": 2.3629025419021317e-06,
+      "loss": 0.747,
+      "step": 8557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7798921104018887,
+      "learning_rate": 2.3617987794693358e-06,
+      "loss": 1.0577,
+      "step": 8558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.552262686278092,
+      "learning_rate": 2.360695195173921e-06,
+      "loss": 0.6995,
+      "step": 8559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.663446038552533,
+      "learning_rate": 2.3595917890904017e-06,
+      "loss": 0.771,
+      "step": 8560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.577542980473154,
+      "learning_rate": 2.3584885612932825e-06,
+      "loss": 0.7385,
+      "step": 8561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4474310613301555,
+      "learning_rate": 2.357385511857056e-06,
+      "loss": 0.6195,
+      "step": 8562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.585787914130582,
+      "learning_rate": 2.3562826408562016e-06,
+      "loss": 0.8609,
+      "step": 8563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7836028551372024,
+      "learning_rate": 2.3551799483651894e-06,
+      "loss": 1.0607,
+      "step": 8564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.46909051260179,
+      "learning_rate": 2.354077434458473e-06,
+      "loss": 0.8553,
+      "step": 8565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4725081174800165,
+      "learning_rate": 2.3529750992105e-06,
+      "loss": 0.7756,
+      "step": 8566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.573305977910874,
+      "learning_rate": 2.351872942695701e-06,
+      "loss": 0.7804,
+      "step": 8567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5376476203816676,
+      "learning_rate": 2.3507709649884948e-06,
+      "loss": 0.725,
+      "step": 8568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7854213151302347,
+      "learning_rate": 2.349669166163292e-06,
+      "loss": 0.773,
+      "step": 8569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4669408168907234,
+      "learning_rate": 2.348567546294488e-06,
+      "loss": 0.7704,
+      "step": 8570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7675823657644338,
+      "learning_rate": 2.347466105456466e-06,
+      "loss": 1.088,
+      "step": 8571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7616015838858193,
+      "learning_rate": 2.346364843723598e-06,
+      "loss": 1.0782,
+      "step": 8572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5180395633243218,
+      "learning_rate": 2.345263761170244e-06,
+      "loss": 0.851,
+      "step": 8573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5731372906336154,
+      "learning_rate": 2.3441628578707505e-06,
+      "loss": 0.7003,
+      "step": 8574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6382223482010612,
+      "learning_rate": 2.3430621338994527e-06,
+      "loss": 0.8144,
+      "step": 8575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5794018091609778,
+      "learning_rate": 2.3419615893306762e-06,
+      "loss": 0.7971,
+      "step": 8576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5751730128656825,
+      "learning_rate": 2.340861224238732e-06,
+      "loss": 0.6948,
+      "step": 8577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6134195032326573,
+      "learning_rate": 2.3397610386979157e-06,
+      "loss": 0.8071,
+      "step": 8578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5231983755967886,
+      "learning_rate": 2.3386610327825194e-06,
+      "loss": 0.7975,
+      "step": 8579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5725651302900405,
+      "learning_rate": 2.3375612065668158e-06,
+      "loss": 0.7585,
+      "step": 8580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4508243854844547,
+      "learning_rate": 2.3364615601250673e-06,
+      "loss": 0.7596,
+      "step": 8581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5192156165105057,
+      "learning_rate": 2.335362093531523e-06,
+      "loss": 0.7549,
+      "step": 8582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6163421518684031,
+      "learning_rate": 2.334262806860425e-06,
+      "loss": 0.7196,
+      "step": 8583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5819450021124593,
+      "learning_rate": 2.3331637001859974e-06,
+      "loss": 0.8335,
+      "step": 8584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5503334643627318,
+      "learning_rate": 2.332064773582456e-06,
+      "loss": 0.7745,
+      "step": 8585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7818150292827802,
+      "learning_rate": 2.330966027124001e-06,
+      "loss": 1.0551,
+      "step": 8586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5255440246370735,
+      "learning_rate": 2.3298674608848225e-06,
+      "loss": 0.7964,
+      "step": 8587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5875659153423067,
+      "learning_rate": 2.3287690749390994e-06,
+      "loss": 0.742,
+      "step": 8588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5950011596890226,
+      "learning_rate": 2.3276708693609947e-06,
+      "loss": 0.7255,
+      "step": 8589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4430389465991293,
+      "learning_rate": 2.326572844224665e-06,
+      "loss": 0.6812,
+      "step": 8590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7694401188895588,
+      "learning_rate": 2.32547499960425e-06,
+      "loss": 1.0649,
+      "step": 8591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5094451563851234,
+      "learning_rate": 2.3243773355738773e-06,
+      "loss": 0.7778,
+      "step": 8592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4593433639114457,
+      "learning_rate": 2.3232798522076667e-06,
+      "loss": 0.8028,
+      "step": 8593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.49924647535026,
+      "learning_rate": 2.3221825495797213e-06,
+      "loss": 0.7652,
+      "step": 8594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8780161741091508,
+      "learning_rate": 2.3210854277641333e-06,
+      "loss": 0.8135,
+      "step": 8595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7236993134868128,
+      "learning_rate": 2.3199884868349824e-06,
+      "loss": 0.7241,
+      "step": 8596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5370851161782546,
+      "learning_rate": 2.3188917268663375e-06,
+      "loss": 0.7999,
+      "step": 8597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5803946047023787,
+      "learning_rate": 2.317795147932254e-06,
+      "loss": 0.7474,
+      "step": 8598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.487836329676063,
+      "learning_rate": 2.3166987501067733e-06,
+      "loss": 0.6791,
+      "step": 8599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.535060983460703,
+      "learning_rate": 2.315602533463931e-06,
+      "loss": 0.812,
+      "step": 8600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.506178315385878,
+      "learning_rate": 2.3145064980777433e-06,
+      "loss": 0.7472,
+      "step": 8601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9120837916487803,
+      "learning_rate": 2.313410644022216e-06,
+      "loss": 0.738,
+      "step": 8602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6006623984807768,
+      "learning_rate": 2.3123149713713474e-06,
+      "loss": 0.7526,
+      "step": 8603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.62526878726048,
+      "learning_rate": 2.311219480199117e-06,
+      "loss": 0.7542,
+      "step": 8604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7879553572683522,
+      "learning_rate": 2.3101241705794962e-06,
+      "loss": 1.0662,
+      "step": 8605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.571713907593411,
+      "learning_rate": 2.309029042586442e-06,
+      "loss": 0.8138,
+      "step": 8606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5488867198100023,
+      "learning_rate": 2.3079340962939e-06,
+      "loss": 0.7938,
+      "step": 8607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.522421785501238,
+      "learning_rate": 2.3068393317758035e-06,
+      "loss": 0.6416,
+      "step": 8608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7612261063493391,
+      "learning_rate": 2.3057447491060725e-06,
+      "loss": 1.0784,
+      "step": 8609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5061028192070636,
+      "learning_rate": 2.304650348358618e-06,
+      "loss": 0.7964,
+      "step": 8610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4581792668080857,
+      "learning_rate": 2.3035561296073356e-06,
+      "loss": 0.7471,
+      "step": 8611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4980249371701648,
+      "learning_rate": 2.3024620929261087e-06,
+      "loss": 0.7967,
+      "step": 8612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.482371634552179,
+      "learning_rate": 2.301368238388808e-06,
+      "loss": 0.719,
+      "step": 8613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4995025589232869,
+      "learning_rate": 2.3002745660692967e-06,
+      "loss": 0.7229,
+      "step": 8614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.46628633950014,
+      "learning_rate": 2.2991810760414194e-06,
+      "loss": 0.7206,
+      "step": 8615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.495994099328573,
+      "learning_rate": 2.298087768379012e-06,
+      "loss": 0.816,
+      "step": 8616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4647126002887028,
+      "learning_rate": 2.2969946431558963e-06,
+      "loss": 0.6667,
+      "step": 8617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4910649822541095,
+      "learning_rate": 2.295901700445884e-06,
+      "loss": 0.7388,
+      "step": 8618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6416495754369,
+      "learning_rate": 2.29480894032277e-06,
+      "loss": 0.7862,
+      "step": 8619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5680871584305436,
+      "learning_rate": 2.2937163628603437e-06,
+      "loss": 0.7347,
+      "step": 8620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5407044245995107,
+      "learning_rate": 2.292623968132377e-06,
+      "loss": 0.6914,
+      "step": 8621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7355525773197535,
+      "learning_rate": 2.291531756212631e-06,
+      "loss": 0.7152,
+      "step": 8622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5492937108570797,
+      "learning_rate": 2.2904397271748517e-06,
+      "loss": 0.7641,
+      "step": 8623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4803060155533772,
+      "learning_rate": 2.28934788109278e-06,
+      "loss": 0.7228,
+      "step": 8624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.483808228638164,
+      "learning_rate": 2.288256218040138e-06,
+      "loss": 0.8094,
+      "step": 8625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7952761861086189,
+      "learning_rate": 2.2871647380906347e-06,
+      "loss": 1.0594,
+      "step": 8626
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7740590856570554,
+      "learning_rate": 2.286073441317973e-06,
+      "loss": 1.0522,
+      "step": 8627
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8224992494425446,
+      "learning_rate": 2.284982327795839e-06,
+      "loss": 1.0322,
+      "step": 8628
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4789466627889427,
+      "learning_rate": 2.283891397597908e-06,
+      "loss": 0.8271,
+      "step": 8629
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4021952817961376,
+      "learning_rate": 2.2828006507978373e-06,
+      "loss": 0.816,
+      "step": 8630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7678737425863551,
+      "learning_rate": 2.2817100874692816e-06,
+      "loss": 1.0701,
+      "step": 8631
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5598480149063723,
+      "learning_rate": 2.280619707685876e-06,
+      "loss": 0.8061,
+      "step": 8632
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5175424481368072,
+      "learning_rate": 2.2795295115212445e-06,
+      "loss": 0.7825,
+      "step": 8633
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7343383898610367,
+      "learning_rate": 2.2784394990490033e-06,
+      "loss": 1.0189,
+      "step": 8634
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5529729314234861,
+      "learning_rate": 2.27734967034275e-06,
+      "loss": 0.6564,
+      "step": 8635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5346131061002133,
+      "learning_rate": 2.276260025476071e-06,
+      "loss": 0.8468,
+      "step": 8636
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6775762162165697,
+      "learning_rate": 2.2751705645225446e-06,
+      "loss": 0.7595,
+      "step": 8637
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4837832078920317,
+      "learning_rate": 2.2740812875557328e-06,
+      "loss": 0.8157,
+      "step": 8638
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5451625611076296,
+      "learning_rate": 2.2729921946491855e-06,
+      "loss": 0.721,
+      "step": 8639
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.586110107963762,
+      "learning_rate": 2.2719032858764413e-06,
+      "loss": 0.7829,
+      "step": 8640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5121222727800199,
+      "learning_rate": 2.270814561311025e-06,
+      "loss": 0.8162,
+      "step": 8641
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5511901982674645,
+      "learning_rate": 2.2697260210264506e-06,
+      "loss": 0.7904,
+      "step": 8642
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8033024678419775,
+      "learning_rate": 2.268637665096216e-06,
+      "loss": 1.0598,
+      "step": 8643
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4340619245178114,
+      "learning_rate": 2.2675494935938143e-06,
+      "loss": 0.8054,
+      "step": 8644
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.371684650048675,
+      "learning_rate": 2.2664615065927182e-06,
+      "loss": 0.6732,
+      "step": 8645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.522766385738045,
+      "learning_rate": 2.2653737041663927e-06,
+      "loss": 0.7371,
+      "step": 8646
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5804006475853332,
+      "learning_rate": 2.264286086388285e-06,
+      "loss": 0.7882,
+      "step": 8647
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5839410030756629,
+      "learning_rate": 2.263198653331839e-06,
+      "loss": 0.7421,
+      "step": 8648
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4499294089114947,
+      "learning_rate": 2.2621114050704775e-06,
+      "loss": 0.7408,
+      "step": 8649
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.488134664956892,
+      "learning_rate": 2.2610243416776146e-06,
+      "loss": 0.7057,
+      "step": 8650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.426602145557958,
+      "learning_rate": 2.2599374632266514e-06,
+      "loss": 0.6487,
+      "step": 8651
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5076053867522965,
+      "learning_rate": 2.2588507697909756e-06,
+      "loss": 0.7569,
+      "step": 8652
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.409978766927716,
+      "learning_rate": 2.2577642614439643e-06,
+      "loss": 0.7761,
+      "step": 8653
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5320528891165686,
+      "learning_rate": 2.2566779382589788e-06,
+      "loss": 0.7583,
+      "step": 8654
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.385756744619177,
+      "learning_rate": 2.2555918003093737e-06,
+      "loss": 0.8209,
+      "step": 8655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5666375314583019,
+      "learning_rate": 2.2545058476684855e-06,
+      "loss": 0.8072,
+      "step": 8656
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7698046438196599,
+      "learning_rate": 2.253420080409639e-06,
+      "loss": 1.0553,
+      "step": 8657
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5774035130645043,
+      "learning_rate": 2.2523344986061508e-06,
+      "loss": 0.7389,
+      "step": 8658
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.546727535339064,
+      "learning_rate": 2.25124910233132e-06,
+      "loss": 0.7529,
+      "step": 8659
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5232079451545817,
+      "learning_rate": 2.250163891658435e-06,
+      "loss": 0.7301,
+      "step": 8660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5633243149813447,
+      "learning_rate": 2.249078866660772e-06,
+      "loss": 0.6435,
+      "step": 8661
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5481640158758458,
+      "learning_rate": 2.247994027411595e-06,
+      "loss": 0.7114,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5827406182772803,
+      "learning_rate": 2.2469093739841537e-06,
+      "loss": 0.7733,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4997896498638703,
+      "learning_rate": 2.2458249064516846e-06,
+      "loss": 0.766,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7809964236684456,
+      "learning_rate": 2.2447406248874176e-06,
+      "loss": 1.0456,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.439786395442257,
+      "learning_rate": 2.2436565293645642e-06,
+      "loss": 0.6385,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4428531723469047,
+      "learning_rate": 2.2425726199563225e-06,
+      "loss": 0.7401,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4850846910220885,
+      "learning_rate": 2.2414888967358844e-06,
+      "loss": 0.7998,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6462869377376586,
+      "learning_rate": 2.240405359776424e-06,
+      "loss": 0.7903,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7539260421140899,
+      "learning_rate": 2.2393220091511043e-06,
+      "loss": 1.0678,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5388870015743326,
+      "learning_rate": 2.2382388449330728e-06,
+      "loss": 0.7052,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4710947130064311,
+      "learning_rate": 2.237155867195472e-06,
+      "loss": 0.706,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.662251585820311,
+      "learning_rate": 2.236073076011426e-06,
+      "loss": 0.7615,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4370815809968653,
+      "learning_rate": 2.2349904714540427e-06,
+      "loss": 0.7556,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7350910090338045,
+      "learning_rate": 2.233908053596427e-06,
+      "loss": 0.7963,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6524540828366223,
+      "learning_rate": 2.2328258225116645e-06,
+      "loss": 0.8183,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4134806136348532,
+      "learning_rate": 2.231743778272828e-06,
+      "loss": 0.7979,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6477819589473648,
+      "learning_rate": 2.2306619209529832e-06,
+      "loss": 0.7172,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8223264867596763,
+      "learning_rate": 2.229580250625179e-06,
+      "loss": 1.0622,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4890441484782917,
+      "learning_rate": 2.2284987673624503e-06,
+      "loss": 0.7964,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4974516464949201,
+      "learning_rate": 2.227417471237821e-06,
+      "loss": 0.732,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.570670630902365,
+      "learning_rate": 2.2263363623243058e-06,
+      "loss": 0.7876,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.530099518427506,
+      "learning_rate": 2.225255440694901e-06,
+      "loss": 0.7752,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5858424242807503,
+      "learning_rate": 2.2241747064225942e-06,
+      "loss": 0.7705,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7012934826309962,
+      "learning_rate": 2.223094159580359e-06,
+      "loss": 0.778,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6200109715508053,
+      "learning_rate": 2.222013800241155e-06,
+      "loss": 0.8019,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5512513071049818,
+      "learning_rate": 2.220933628477932e-06,
+      "loss": 0.6898,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5762209875570257,
+      "learning_rate": 2.2198536443636233e-06,
+      "loss": 0.676,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.500662201343916,
+      "learning_rate": 2.218773847971156e-06,
+      "loss": 0.7079,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.392552909836651,
+      "learning_rate": 2.217694239373437e-06,
+      "loss": 0.7673,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4703357120974991,
+      "learning_rate": 2.2166148186433637e-06,
+      "loss": 0.7296,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5283778289806405,
+      "learning_rate": 2.2155355858538245e-06,
+      "loss": 0.7808,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8013391890255379,
+      "learning_rate": 2.21445654107769e-06,
+      "loss": 1.097,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4529341386148438,
+      "learning_rate": 2.2133776843878185e-06,
+      "loss": 0.7018,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.664215310240996,
+      "learning_rate": 2.2122990158570583e-06,
+      "loss": 0.7472,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7118800252255761,
+      "learning_rate": 2.2112205355582427e-06,
+      "loss": 0.7331,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4670626078750053,
+      "learning_rate": 2.2101422435641932e-06,
+      "loss": 0.7375,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.59443283879915,
+      "learning_rate": 2.2090641399477183e-06,
+      "loss": 0.7329,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.569408374075968,
+      "learning_rate": 2.2079862247816148e-06,
+      "loss": 0.7468,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.547386084759423,
+      "learning_rate": 2.2069084981386666e-06,
+      "loss": 0.7731,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6227734076966431,
+      "learning_rate": 2.205830960091641e-06,
+      "loss": 0.6863,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6186360633485697,
+      "learning_rate": 2.2047536107133005e-06,
+      "loss": 0.802,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4552418147882455,
+      "learning_rate": 2.203676450076388e-06,
+      "loss": 0.7622,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6257024343218032,
+      "learning_rate": 2.202599478253636e-06,
+      "loss": 0.7377,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5782620491213968,
+      "learning_rate": 2.201522695317763e-06,
+      "loss": 0.704,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7628979110573563,
+      "learning_rate": 2.2004461013414775e-06,
+      "loss": 1.0508,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.605968577314302,
+      "learning_rate": 2.1993696963974726e-06,
+      "loss": 0.8036,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5268399814401854,
+      "learning_rate": 2.198293480558428e-06,
+      "loss": 0.7783,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5687472235736113,
+      "learning_rate": 2.1972174538970155e-06,
+      "loss": 0.7484,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4264323556812994,
+      "learning_rate": 2.1961416164858897e-06,
+      "loss": 0.7229,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4414716023425886,
+      "learning_rate": 2.195065968397693e-06,
+      "loss": 0.8024,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5312019580517486,
+      "learning_rate": 2.1939905097050553e-06,
+      "loss": 0.8013,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4907439810758152,
+      "learning_rate": 2.192915240480596e-06,
+      "loss": 0.7109,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.516001962214674,
+      "learning_rate": 2.191840160796918e-06,
+      "loss": 0.6578,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6507993219831034,
+      "learning_rate": 2.190765270726612e-06,
+      "loss": 0.7416,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5015230739968666,
+      "learning_rate": 2.1896905703422605e-06,
+      "loss": 0.7275,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4984065036360317,
+      "learning_rate": 2.18861605971643e-06,
+      "loss": 0.7625,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6220732720331674,
+      "learning_rate": 2.1875417389216674e-06,
+      "loss": 0.7938,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7680073127187883,
+      "learning_rate": 2.18646760803052e-06,
+      "loss": 1.0531,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.490722018935445,
+      "learning_rate": 2.185393667115513e-06,
+      "loss": 0.7998,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4816061915696637,
+      "learning_rate": 2.1843199162491618e-06,
+      "loss": 0.8063,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5152822666832892,
+      "learning_rate": 2.1832463555039662e-06,
+      "loss": 0.7548,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7737138597227453,
+      "learning_rate": 2.18217298495242e-06,
+      "loss": 0.8539,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6382195545787797,
+      "learning_rate": 2.1810998046669964e-06,
+      "loss": 0.754,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4851896842342254,
+      "learning_rate": 2.180026814720158e-06,
+      "loss": 0.7796,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.516366262044474,
+      "learning_rate": 2.1789540151843597e-06,
+      "loss": 0.7164,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4800615733207947,
+      "learning_rate": 2.177881406132037e-06,
+      "loss": 0.7214,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.464973813729646,
+      "learning_rate": 2.1768089876356145e-06,
+      "loss": 0.76,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7866820089074741,
+      "learning_rate": 2.175736759767505e-06,
+      "loss": 1.0374,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4589495021951695,
+      "learning_rate": 2.174664722600108e-06,
+      "loss": 0.7178,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4146259021648806,
+      "learning_rate": 2.1735928762058085e-06,
+      "loss": 0.7076,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3748388644709217,
+      "learning_rate": 2.1725212206569798e-06,
+      "loss": 0.8053,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7029016608443437,
+      "learning_rate": 2.1714497560259854e-06,
+      "loss": 0.7716,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8493430153187262,
+      "learning_rate": 2.1703784823851714e-06,
+      "loss": 0.8261,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6028716591424468,
+      "learning_rate": 2.1693073998068704e-06,
+      "loss": 0.7805,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5574437686171727,
+      "learning_rate": 2.1682365083634086e-06,
+      "loss": 0.7258,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.517299510637743,
+      "learning_rate": 2.167165808127093e-06,
+      "loss": 0.7681,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7765050532695559,
+      "learning_rate": 2.16609529917022e-06,
+      "loss": 1.0553,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.45362117958091,
+      "learning_rate": 2.165024981565072e-06,
+      "loss": 0.7831,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5615396010063949,
+      "learning_rate": 2.16395485538392e-06,
+      "loss": 0.7724,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5359974451549714,
+      "learning_rate": 2.1628849206990216e-06,
+      "loss": 0.6934,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3160820506920023,
+      "learning_rate": 2.1618151775826186e-06,
+      "loss": 0.6857,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6091376974731562,
+      "learning_rate": 2.160745626106946e-06,
+      "loss": 0.7174,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4913619676924215,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.762,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7757825929527788,
+      "learning_rate": 2.1586070983666506e-06,
+      "loss": 1.048,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5338755377392468,
+      "learning_rate": 2.1575381222464236e-06,
+      "loss": 0.7809,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.421698220223875,
+      "learning_rate": 2.1564693380557246e-06,
+      "loss": 0.6771,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7808947990710995,
+      "learning_rate": 2.155400745866717e-06,
+      "loss": 1.0791,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5420577392146768,
+      "learning_rate": 2.1543323457515564e-06,
+      "loss": 0.8443,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5181520504834694,
+      "learning_rate": 2.1532641377823828e-06,
+      "loss": 0.8008,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5009966705422029,
+      "learning_rate": 2.1521961220313237e-06,
+      "loss": 0.7796,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5178397709987348,
+      "learning_rate": 2.151128298570494e-06,
+      "loss": 0.8291,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4315331585527435,
+      "learning_rate": 2.150060667471995e-06,
+      "loss": 0.7961,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.735276506523688,
+      "learning_rate": 2.148993228807918e-06,
+      "loss": 0.7672,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4012177813206146,
+      "learning_rate": 2.147925982650337e-06,
+      "loss": 0.6856,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.298923294117518,
+      "learning_rate": 2.146858929071314e-06,
+      "loss": 0.7332,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.449043455320584,
+      "learning_rate": 2.145792068142902e-06,
+      "loss": 0.7361,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5426795793525792,
+      "learning_rate": 2.1447253999371355e-06,
+      "loss": 0.7398,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5119004887439558,
+      "learning_rate": 2.1436589245260375e-06,
+      "loss": 0.7003,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5449871845477379,
+      "learning_rate": 2.142592641981622e-06,
+      "loss": 0.7512,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7790416420168379,
+      "learning_rate": 2.1415265523758877e-06,
+      "loss": 1.0614,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4972148176634292,
+      "learning_rate": 2.1404606557808142e-06,
+      "loss": 0.7455,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5248502510257462,
+      "learning_rate": 2.139394952268375e-06,
+      "loss": 0.8198,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7788280596267059,
+      "learning_rate": 2.138329441910531e-06,
+      "loss": 1.0648,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4734609676612733,
+      "learning_rate": 2.137264124779227e-06,
+      "loss": 0.7901,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4708876216899602,
+      "learning_rate": 2.1361990009463935e-06,
+      "loss": 0.8017,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7720520557917335,
+      "learning_rate": 2.1351340704839534e-06,
+      "loss": 1.0752,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4861680279470364,
+      "learning_rate": 2.134069333463812e-06,
+      "loss": 0.7073,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5345241969635814,
+      "learning_rate": 2.133004789957862e-06,
+      "loss": 0.876,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9803511984491948,
+      "learning_rate": 2.1319404400379835e-06,
+      "loss": 0.752,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5250020399165836,
+      "learning_rate": 2.130876283776046e-06,
+      "loss": 0.7938,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4597890077065672,
+      "learning_rate": 2.1298123212439028e-06,
+      "loss": 0.7364,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.503572508939869,
+      "learning_rate": 2.1287485525133943e-06,
+      "loss": 0.7306,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6003525237162908,
+      "learning_rate": 2.127684977656349e-06,
+      "loss": 0.7848,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5345098079410684,
+      "learning_rate": 2.1266215967445823e-06,
+      "loss": 0.8023,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4963824096299543,
+      "learning_rate": 2.1255584098498936e-06,
+      "loss": 0.7944,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4839727032360568,
+      "learning_rate": 2.124495417044076e-06,
+      "loss": 0.7847,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6183813996119811,
+      "learning_rate": 2.1234326183989036e-06,
+      "loss": 0.8142,
+      "step": 8778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7569152448232452,
+      "learning_rate": 2.1223700139861375e-06,
+      "loss": 0.7195,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7772811569630824,
+      "learning_rate": 2.121307603877527e-06,
+      "loss": 1.033,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5786536348925817,
+      "learning_rate": 2.120245388144811e-06,
+      "loss": 0.8001,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5547451778595185,
+      "learning_rate": 2.1191833668597117e-06,
+      "loss": 0.805,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4387759070522135,
+      "learning_rate": 2.118121540093939e-06,
+      "loss": 0.6686,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4677472818688722,
+      "learning_rate": 2.117059907919189e-06,
+      "loss": 0.6988,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.384862904762936,
+      "learning_rate": 2.1159984704071466e-06,
+      "loss": 0.7999,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.480209987129238,
+      "learning_rate": 2.114937227629482e-06,
+      "loss": 0.7289,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7506650489713488,
+      "learning_rate": 2.113876179657852e-06,
+      "loss": 1.046,
+      "step": 8787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4422946778876335,
+      "learning_rate": 2.1128153265639023e-06,
+      "loss": 0.7752,
+      "step": 8788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.535325985746534,
+      "learning_rate": 2.111754668419264e-06,
+      "loss": 0.6872,
+      "step": 8789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4779460719736055,
+      "learning_rate": 2.1106942052955535e-06,
+      "loss": 0.7725,
+      "step": 8790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.472878570430273,
+      "learning_rate": 2.109633937264379e-06,
+      "loss": 0.6803,
+      "step": 8791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.593265051125879,
+      "learning_rate": 2.10857386439733e-06,
+      "loss": 0.708,
+      "step": 8792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8504507446977863,
+      "learning_rate": 2.1075139867659853e-06,
+      "loss": 0.6844,
+      "step": 8793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5978848470688238,
+      "learning_rate": 2.1064543044419105e-06,
+      "loss": 0.7854,
+      "step": 8794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5781568863572366,
+      "learning_rate": 2.1053948174966577e-06,
+      "loss": 0.7565,
+      "step": 8795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4872738688549465,
+      "learning_rate": 2.104335526001765e-06,
+      "loss": 0.6916,
+      "step": 8796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4737916326230112,
+      "learning_rate": 2.1032764300287577e-06,
+      "loss": 0.7765,
+      "step": 8797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5909553283594644,
+      "learning_rate": 2.1022175296491516e-06,
+      "loss": 0.812,
+      "step": 8798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7552444274646167,
+      "learning_rate": 2.1011588249344434e-06,
+      "loss": 1.0449,
+      "step": 8799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.587558767961101,
+      "learning_rate": 2.1001003159561185e-06,
+      "loss": 0.7619,
+      "step": 8800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5648761358184804,
+      "learning_rate": 2.099042002785653e-06,
+      "loss": 0.7661,
+      "step": 8801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.550907305487096,
+      "learning_rate": 2.097983885494505e-06,
+      "loss": 0.7938,
+      "step": 8802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4090413484310849,
+      "learning_rate": 2.0969259641541214e-06,
+      "loss": 0.7336,
+      "step": 8803
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.45565093996669,
+      "learning_rate": 2.095868238835932e-06,
+      "loss": 0.6237,
+      "step": 8804
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4289412223810323,
+      "learning_rate": 2.0948107096113647e-06,
+      "loss": 0.8044,
+      "step": 8805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5335832421646052,
+      "learning_rate": 2.0937533765518187e-06,
+      "loss": 0.7636,
+      "step": 8806
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7776481206586978,
+      "learning_rate": 2.092696239728689e-06,
+      "loss": 1.0729,
+      "step": 8807
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5242965799380574,
+      "learning_rate": 2.0916392992133582e-06,
+      "loss": 0.6971,
+      "step": 8808
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5990687728589879,
+      "learning_rate": 2.090582555077193e-06,
+      "loss": 0.8725,
+      "step": 8809
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6627716823275616,
+      "learning_rate": 2.089526007391546e-06,
+      "loss": 0.8234,
+      "step": 8810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5435158280245527,
+      "learning_rate": 2.0884696562277566e-06,
+      "loss": 0.7634,
+      "step": 8811
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5700281647383987,
+      "learning_rate": 2.0874135016571556e-06,
+      "loss": 0.7314,
+      "step": 8812
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7575366868254246,
+      "learning_rate": 2.0863575437510547e-06,
+      "loss": 1.0786,
+      "step": 8813
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.578444607652206,
+      "learning_rate": 2.085301782580753e-06,
+      "loss": 0.8237,
+      "step": 8814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4535378818726827,
+      "learning_rate": 2.0842462182175422e-06,
+      "loss": 0.6828,
+      "step": 8815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.442577480649827,
+      "learning_rate": 2.0831908507326935e-06,
+      "loss": 0.7081,
+      "step": 8816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8041916999530423,
+      "learning_rate": 2.0821356801974686e-06,
+      "loss": 0.7357,
+      "step": 8817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6420314213733418,
+      "learning_rate": 2.0810807066831147e-06,
+      "loss": 0.7646,
+      "step": 8818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4386090029049434,
+      "learning_rate": 2.0800259302608656e-06,
+      "loss": 0.7152,
+      "step": 8819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.506198902452254,
+      "learning_rate": 2.0789713510019425e-06,
+      "loss": 0.7606,
+      "step": 8820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4896343478852094,
+      "learning_rate": 2.077916968977552e-06,
+      "loss": 0.7338,
+      "step": 8821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.570219011004925,
+      "learning_rate": 2.0768627842588906e-06,
+      "loss": 0.7891,
+      "step": 8822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7783872442248423,
+      "learning_rate": 2.0758087969171387e-06,
+      "loss": 1.0466,
+      "step": 8823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4220175868201346,
+      "learning_rate": 2.074755007023461e-06,
+      "loss": 0.7343,
+      "step": 8824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5060069237683054,
+      "learning_rate": 2.0737014146490165e-06,
+      "loss": 0.8154,
+      "step": 8825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7171139785743978,
+      "learning_rate": 2.072648019864944e-06,
+      "loss": 0.6677,
+      "step": 8826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.612845462917001,
+      "learning_rate": 2.0715948227423704e-06,
+      "loss": 0.8295,
+      "step": 8827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5573006069489843,
+      "learning_rate": 2.070541823352411e-06,
+      "loss": 0.6972,
+      "step": 8828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.570665091206447,
+      "learning_rate": 2.069489021766166e-06,
+      "loss": 0.757,
+      "step": 8829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4863189656337197,
+      "learning_rate": 2.068436418054724e-06,
+      "loss": 0.7587,
+      "step": 8830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5140775926647114,
+      "learning_rate": 2.067384012289156e-06,
+      "loss": 0.6634,
+      "step": 8831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4959774779934785,
+      "learning_rate": 2.0663318045405275e-06,
+      "loss": 0.7427,
+      "step": 8832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5253184690705592,
+      "learning_rate": 2.065279794879883e-06,
+      "loss": 0.7144,
+      "step": 8833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4819994765483997,
+      "learning_rate": 2.064227983378256e-06,
+      "loss": 0.6665,
+      "step": 8834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5579914802594237,
+      "learning_rate": 2.063176370106671e-06,
+      "loss": 0.7168,
+      "step": 8835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4919534847957625,
+      "learning_rate": 2.0621249551361323e-06,
+      "loss": 0.7087,
+      "step": 8836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.586289143791413,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.7794,
+      "step": 8837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7662943076416616,
+      "learning_rate": 2.060022720382159e-06,
+      "loss": 1.0896,
+      "step": 8838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7693620394400195,
+      "learning_rate": 2.0589719007406713e-06,
+      "loss": 1.0762,
+      "step": 8839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4271201023284485,
+      "learning_rate": 2.0579212796841257e-06,
+      "loss": 0.8374,
+      "step": 8840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4755801952957277,
+      "learning_rate": 2.0568708572834615e-06,
+      "loss": 0.6983,
+      "step": 8841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.59301061039807,
+      "learning_rate": 2.055820633609609e-06,
+      "loss": 0.7141,
+      "step": 8842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.601535374751086,
+      "learning_rate": 2.0547706087334783e-06,
+      "loss": 0.7167,
+      "step": 8843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5421393118413262,
+      "learning_rate": 2.053720782725972e-06,
+      "loss": 0.6892,
+      "step": 8844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4365636874418852,
+      "learning_rate": 2.052671155657973e-06,
+      "loss": 0.7148,
+      "step": 8845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7904456624867255,
+      "learning_rate": 2.0516217276003593e-06,
+      "loss": 1.1089,
+      "step": 8846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7746910421408888,
+      "learning_rate": 2.0505724986239883e-06,
+      "loss": 1.062,
+      "step": 8847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6070046650781,
+      "learning_rate": 2.0495234687997046e-06,
+      "loss": 0.8876,
+      "step": 8848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5002521556641006,
+      "learning_rate": 2.048474638198347e-06,
+      "loss": 0.7912,
+      "step": 8849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.517723011222634,
+      "learning_rate": 2.047426006890728e-06,
+      "loss": 0.8468,
+      "step": 8850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6622507708299004,
+      "learning_rate": 2.0463775749476556e-06,
+      "loss": 0.7554,
+      "step": 8851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5045864973317238,
+      "learning_rate": 2.0453293424399244e-06,
+      "loss": 0.8032,
+      "step": 8852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2374226391839644,
+      "learning_rate": 2.0442813094383128e-06,
+      "loss": 0.7974,
+      "step": 8853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.527037827276388,
+      "learning_rate": 2.0432334760135856e-06,
+      "loss": 0.7864,
+      "step": 8854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5266541324577658,
+      "learning_rate": 2.042185842236494e-06,
+      "loss": 0.7092,
+      "step": 8855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5402345566330267,
+      "learning_rate": 2.0411384081777785e-06,
+      "loss": 0.7295,
+      "step": 8856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5525383409897007,
+      "learning_rate": 2.040091173908164e-06,
+      "loss": 0.8301,
+      "step": 8857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6378463418474885,
+      "learning_rate": 2.0390441394983603e-06,
+      "loss": 0.7986,
+      "step": 8858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7462768752544333,
+      "learning_rate": 2.037997305019068e-06,
+      "loss": 0.775,
+      "step": 8859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5283465740048878,
+      "learning_rate": 2.0369506705409713e-06,
+      "loss": 0.6812,
+      "step": 8860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5724285233135942,
+      "learning_rate": 2.0359042361347405e-06,
+      "loss": 0.792,
+      "step": 8861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5633318386692796,
+      "learning_rate": 2.0348580018710334e-06,
+      "loss": 0.705,
+      "step": 8862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.593781633489615,
+      "learning_rate": 2.0338119678204944e-06,
+      "loss": 0.7606,
+      "step": 8863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5880504952330674,
+      "learning_rate": 2.0327661340537536e-06,
+      "loss": 0.6773,
+      "step": 8864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.510687947896017,
+      "learning_rate": 2.031720500641427e-06,
+      "loss": 0.8054,
+      "step": 8865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5845485585064583,
+      "learning_rate": 2.0306750676541214e-06,
+      "loss": 0.7387,
+      "step": 8866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4660733547426146,
+      "learning_rate": 2.029629835162425e-06,
+      "loss": 0.8358,
+      "step": 8867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3975576714120441,
+      "learning_rate": 2.028584803236914e-06,
+      "loss": 0.7317,
+      "step": 8868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3873540021136497,
+      "learning_rate": 2.02753997194815e-06,
+      "loss": 0.75,
+      "step": 8869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5099521876068092,
+      "learning_rate": 2.0264953413666856e-06,
+      "loss": 0.8006,
+      "step": 8870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.805258029689985,
+      "learning_rate": 2.0254509115630557e-06,
+      "loss": 1.0701,
+      "step": 8871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6730796369129737,
+      "learning_rate": 2.0244066826077812e-06,
+      "loss": 0.7481,
+      "step": 8872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5199666665533178,
+      "learning_rate": 2.023362654571372e-06,
+      "loss": 0.7293,
+      "step": 8873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6212072271032454,
+      "learning_rate": 2.0223188275243226e-06,
+      "loss": 0.8059,
+      "step": 8874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5240758455245407,
+      "learning_rate": 2.0212752015371135e-06,
+      "loss": 0.7005,
+      "step": 8875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4699063215170414,
+      "learning_rate": 2.0202317766802155e-06,
+      "loss": 0.6882,
+      "step": 8876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.492279589777096,
+      "learning_rate": 2.019188553024082e-06,
+      "loss": 0.7206,
+      "step": 8877
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5037508846834915,
+      "learning_rate": 2.018145530639153e-06,
+      "loss": 0.7413,
+      "step": 8878
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6654690371423002,
+      "learning_rate": 2.0171027095958543e-06,
+      "loss": 0.8282,
+      "step": 8879
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5842183915705528,
+      "learning_rate": 2.0160600899646033e-06,
+      "loss": 0.7599,
+      "step": 8880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7063844188305413,
+      "learning_rate": 2.0150176718157986e-06,
+      "loss": 0.8134,
+      "step": 8881
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.536750239168504,
+      "learning_rate": 2.013975455219826e-06,
+      "loss": 0.8133,
+      "step": 8882
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6799842872055237,
+      "learning_rate": 2.0129334402470583e-06,
+      "loss": 0.8351,
+      "step": 8883
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5693362490517175,
+      "learning_rate": 2.0118916269678555e-06,
+      "loss": 0.6895,
+      "step": 8884
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.772309244428441,
+      "learning_rate": 2.010850015452563e-06,
+      "loss": 1.0446,
+      "step": 8885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6482371646272858,
+      "learning_rate": 2.0098086057715106e-06,
+      "loss": 0.8439,
+      "step": 8886
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5693111788515117,
+      "learning_rate": 2.00876739799502e-06,
+      "loss": 0.6783,
+      "step": 8887
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4959632302397945,
+      "learning_rate": 2.0077263921933953e-06,
+      "loss": 0.7255,
+      "step": 8888
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7339190090143608,
+      "learning_rate": 2.0066855884369246e-06,
+      "loss": 0.7086,
+      "step": 8889
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4801676734125215,
+      "learning_rate": 2.0056449867958893e-06,
+      "loss": 0.7437,
+      "step": 8890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.407492087770128,
+      "learning_rate": 2.0046045873405518e-06,
+      "loss": 0.7473,
+      "step": 8891
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4877647544058141,
+      "learning_rate": 2.0035643901411593e-06,
+      "loss": 0.7895,
+      "step": 8892
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.492896532953956,
+      "learning_rate": 2.002524395267954e-06,
+      "loss": 0.7554,
+      "step": 8893
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.358161603157602,
+      "learning_rate": 2.001484602791157e-06,
+      "loss": 0.7088,
+      "step": 8894
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5313116226419123,
+      "learning_rate": 2.000445012780974e-06,
+      "loss": 0.7776,
+      "step": 8895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5851947874179244,
+      "learning_rate": 1.999405625307601e-06,
+      "loss": 0.7872,
+      "step": 8896
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6484861402420659,
+      "learning_rate": 1.9983664404412233e-06,
+      "loss": 0.8185,
+      "step": 8897
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7635757846685645,
+      "learning_rate": 1.997327458252007e-06,
+      "loss": 1.0778,
+      "step": 8898
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7740062022536703,
+      "learning_rate": 1.996288678810105e-06,
+      "loss": 1.044,
+      "step": 8899
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7653020545671417,
+      "learning_rate": 1.995250102185662e-06,
+      "loss": 1.0544,
+      "step": 8900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6138576785730476,
+      "learning_rate": 1.9942117284488026e-06,
+      "loss": 0.6421,
+      "step": 8901
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5511182672765857,
+      "learning_rate": 1.9931735576696405e-06,
+      "loss": 0.7431,
+      "step": 8902
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4893591831423465,
+      "learning_rate": 1.9921355899182733e-06,
+      "loss": 0.7616,
+      "step": 8903
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4397908554453103,
+      "learning_rate": 1.9910978252647913e-06,
+      "loss": 0.7179,
+      "step": 8904
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5567835452475074,
+      "learning_rate": 1.9900602637792637e-06,
+      "loss": 0.7321,
+      "step": 8905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5102810033285345,
+      "learning_rate": 1.9890229055317502e-06,
+      "loss": 0.7922,
+      "step": 8906
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5857857546491043,
+      "learning_rate": 1.987985750592295e-06,
+      "loss": 0.8027,
+      "step": 8907
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5722979672033213,
+      "learning_rate": 1.986948799030929e-06,
+      "loss": 0.8017,
+      "step": 8908
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5126195750399822,
+      "learning_rate": 1.9859120509176706e-06,
+      "loss": 0.7588,
+      "step": 8909
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4823179313297419,
+      "learning_rate": 1.98487550632252e-06,
+      "loss": 0.7445,
+      "step": 8910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.62994006264953,
+      "learning_rate": 1.983839165315472e-06,
+      "loss": 0.7433,
+      "step": 8911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5946409560099917,
+      "learning_rate": 1.9828030279665006e-06,
+      "loss": 0.7171,
+      "step": 8912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4600343048175033,
+      "learning_rate": 1.981767094345566e-06,
+      "loss": 0.7981,
+      "step": 8913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.509519220667268,
+      "learning_rate": 1.98073136452262e-06,
+      "loss": 0.6959,
+      "step": 8914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7723810990198323,
+      "learning_rate": 1.979695838567597e-06,
+      "loss": 1.0517,
+      "step": 8915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.452591695995745,
+      "learning_rate": 1.978660516550417e-06,
+      "loss": 0.7952,
+      "step": 8916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4557906975467643,
+      "learning_rate": 1.9776253985409876e-06,
+      "loss": 0.7206,
+      "step": 8917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4575934421030878,
+      "learning_rate": 1.9765904846092027e-06,
+      "loss": 0.7544,
+      "step": 8918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4249162511029256,
+      "learning_rate": 1.9755557748249415e-06,
+      "loss": 0.6646,
+      "step": 8919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.419907991600344,
+      "learning_rate": 1.9745212692580684e-06,
+      "loss": 0.7218,
+      "step": 8920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4319441720000685,
+      "learning_rate": 1.973486967978439e-06,
+      "loss": 0.7374,
+      "step": 8921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4621642508384134,
+      "learning_rate": 1.9724528710558906e-06,
+      "loss": 0.7877,
+      "step": 8922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5676678661571493,
+      "learning_rate": 1.971418978560245e-06,
+      "loss": 0.6781,
+      "step": 8923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4861406200085754,
+      "learning_rate": 1.970385290561317e-06,
+      "loss": 0.7484,
+      "step": 8924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4798374595798751,
+      "learning_rate": 1.969351807128902e-06,
+      "loss": 0.7226,
+      "step": 8925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.495154693849385,
+      "learning_rate": 1.968318528332783e-06,
+      "loss": 0.7367,
+      "step": 8926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7761458351933237,
+      "learning_rate": 1.9672854542427293e-06,
+      "loss": 1.0641,
+      "step": 8927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5013349752546,
+      "learning_rate": 1.9662525849284964e-06,
+      "loss": 0.7504,
+      "step": 8928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7578340808187285,
+      "learning_rate": 1.965219920459826e-06,
+      "loss": 1.06,
+      "step": 8929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4257269807062176,
+      "learning_rate": 1.9641874609064443e-06,
+      "loss": 0.6893,
+      "step": 8930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5310599744510958,
+      "learning_rate": 1.9631552063380687e-06,
+      "loss": 0.7624,
+      "step": 8931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4761888465667479,
+      "learning_rate": 1.962123156824398e-06,
+      "loss": 0.79,
+      "step": 8932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4611550376924842,
+      "learning_rate": 1.961091312435116e-06,
+      "loss": 0.6809,
+      "step": 8933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4492823579832488,
+      "learning_rate": 1.9600596732398994e-06,
+      "loss": 0.6643,
+      "step": 8934
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.58382597444594,
+      "learning_rate": 1.9590282393084054e-06,
+      "loss": 0.7809,
+      "step": 8935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4950233814421021,
+      "learning_rate": 1.957997010710278e-06,
+      "loss": 0.7793,
+      "step": 8936
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7627747854821669,
+      "learning_rate": 1.9569659875151464e-06,
+      "loss": 1.0496,
+      "step": 8937
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5817550924855095,
+      "learning_rate": 1.9559351697926337e-06,
+      "loss": 0.7309,
+      "step": 8938
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7718076636242407,
+      "learning_rate": 1.954904557612337e-06,
+      "loss": 1.0762,
+      "step": 8939
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.554068578416361,
+      "learning_rate": 1.953874151043846e-06,
+      "loss": 0.6899,
+      "step": 8940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.573678790185004,
+      "learning_rate": 1.9528439501567385e-06,
+      "loss": 0.737,
+      "step": 8941
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4809730583259506,
+      "learning_rate": 1.951813955020576e-06,
+      "loss": 0.754,
+      "step": 8942
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7749314765476829,
+      "learning_rate": 1.9507841657049056e-06,
+      "loss": 1.065,
+      "step": 8943
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5287359475853204,
+      "learning_rate": 1.9497545822792584e-06,
+      "loss": 0.7844,
+      "step": 8944
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4187913876726683,
+      "learning_rate": 1.948725204813159e-06,
+      "loss": 0.7753,
+      "step": 8945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.543293539117777,
+      "learning_rate": 1.947696033376111e-06,
+      "loss": 0.6894,
+      "step": 8946
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.512426950755788,
+      "learning_rate": 1.946667068037604e-06,
+      "loss": 0.7394,
+      "step": 8947
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4522339924106178,
+      "learning_rate": 1.9456383088671204e-06,
+      "loss": 0.7185,
+      "step": 8948
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.52756486861164,
+      "learning_rate": 1.9446097559341227e-06,
+      "loss": 0.7921,
+      "step": 8949
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5183475486028737,
+      "learning_rate": 1.94358140930806e-06,
+      "loss": 0.7196,
+      "step": 8950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6080131125802337,
+      "learning_rate": 1.9425532690583704e-06,
+      "loss": 0.7982,
+      "step": 8951
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5140270210403055,
+      "learning_rate": 1.941525335254475e-06,
+      "loss": 0.727,
+      "step": 8952
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4487319733159343,
+      "learning_rate": 1.940497607965782e-06,
+      "loss": 0.7792,
+      "step": 8953
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4280446525578032,
+      "learning_rate": 1.9394700872616856e-06,
+      "loss": 0.6951,
+      "step": 8954
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4733370358312954,
+      "learning_rate": 1.938442773211569e-06,
+      "loss": 0.6792,
+      "step": 8955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5419987252136977,
+      "learning_rate": 1.9374156658847965e-06,
+      "loss": 0.8401,
+      "step": 8956
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5644664934553532,
+      "learning_rate": 1.9363887653507195e-06,
+      "loss": 0.7855,
+      "step": 8957
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4955990936705872,
+      "learning_rate": 1.935362071678681e-06,
+      "loss": 0.7573,
+      "step": 8958
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.504117553742307,
+      "learning_rate": 1.9343355849380023e-06,
+      "loss": 0.7702,
+      "step": 8959
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4710582126617087,
+      "learning_rate": 1.933309305197995e-06,
+      "loss": 0.8685,
+      "step": 8960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5121479347556743,
+      "learning_rate": 1.9322832325279563e-06,
+      "loss": 0.7785,
+      "step": 8961
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4989900149027142,
+      "learning_rate": 1.9312573669971684e-06,
+      "loss": 0.7509,
+      "step": 8962
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6194034925716667,
+      "learning_rate": 1.9302317086749e-06,
+      "loss": 0.7763,
+      "step": 8963
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5822155097357817,
+      "learning_rate": 1.9292062576304045e-06,
+      "loss": 0.7667,
+      "step": 8964
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.118265398066812,
+      "learning_rate": 1.9281810139329255e-06,
+      "loss": 0.7717,
+      "step": 8965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5206148568119542,
+      "learning_rate": 1.927155977651689e-06,
+      "loss": 0.7362,
+      "step": 8966
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5974926437714214,
+      "learning_rate": 1.9261311488559077e-06,
+      "loss": 0.7672,
+      "step": 8967
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.551024324083652,
+      "learning_rate": 1.925106527614778e-06,
+      "loss": 0.7547,
+      "step": 8968
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4566223821588165,
+      "learning_rate": 1.924082113997488e-06,
+      "loss": 0.7671,
+      "step": 8969
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5376883582952139,
+      "learning_rate": 1.9230579080732074e-06,
+      "loss": 0.7263,
+      "step": 8970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4920376947343512,
+      "learning_rate": 1.922033909911093e-06,
+      "loss": 0.7655,
+      "step": 8971
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7986841507357801,
+      "learning_rate": 1.9210101195802873e-06,
+      "loss": 1.0425,
+      "step": 8972
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4199608676142805,
+      "learning_rate": 1.919986537149919e-06,
+      "loss": 0.7333,
+      "step": 8973
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3685757233153055,
+      "learning_rate": 1.9189631626891004e-06,
+      "loss": 0.8182,
+      "step": 8974
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7530723229379651,
+      "learning_rate": 1.9179399962669358e-06,
+      "loss": 1.0477,
+      "step": 8975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4649851605689048,
+      "learning_rate": 1.9169170379525102e-06,
+      "loss": 0.7703,
+      "step": 8976
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4903483268056776,
+      "learning_rate": 1.915894287814897e-06,
+      "loss": 0.7434,
+      "step": 8977
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.586235612978687,
+      "learning_rate": 1.9148717459231507e-06,
+      "loss": 0.8097,
+      "step": 8978
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7945707974070908,
+      "learning_rate": 1.9138494123463216e-06,
+      "loss": 1.0409,
+      "step": 8979
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.507509951507935,
+      "learning_rate": 1.9128272871534363e-06,
+      "loss": 0.7766,
+      "step": 8980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5208071403911998,
+      "learning_rate": 1.9118053704135103e-06,
+      "loss": 0.8094,
+      "step": 8981
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4189718360958496,
+      "learning_rate": 1.910783662195551e-06,
+      "loss": 0.7774,
+      "step": 8982
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6513445695919888,
+      "learning_rate": 1.909762162568541e-06,
+      "loss": 0.6669,
+      "step": 8983
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.418030958494561,
+      "learning_rate": 1.9087408716014562e-06,
+      "loss": 0.7929,
+      "step": 8984
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3738156031867634,
+      "learning_rate": 1.907719789363254e-06,
+      "loss": 0.713,
+      "step": 8985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4948792369290256,
+      "learning_rate": 1.9066989159228844e-06,
+      "loss": 0.7643,
+      "step": 8986
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3995268724255967,
+      "learning_rate": 1.9056782513492779e-06,
+      "loss": 0.7292,
+      "step": 8987
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5353213648277757,
+      "learning_rate": 1.9046577957113487e-06,
+      "loss": 0.7029,
+      "step": 8988
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6208815714457068,
+      "learning_rate": 1.9036375490780056e-06,
+      "loss": 0.7133,
+      "step": 8989
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6165035905391572,
+      "learning_rate": 1.902617511518135e-06,
+      "loss": 0.7416,
+      "step": 8990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8134995635743169,
+      "learning_rate": 1.901597683100611e-06,
+      "loss": 1.0749,
+      "step": 8991
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4725879355110554,
+      "learning_rate": 1.9005780638942982e-06,
+      "loss": 0.6841,
+      "step": 8992
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.476048028610045,
+      "learning_rate": 1.8995586539680422e-06,
+      "loss": 0.7074,
+      "step": 8993
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4953803169245197,
+      "learning_rate": 1.8985394533906749e-06,
+      "loss": 0.5939,
+      "step": 8994
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5059740685398337,
+      "learning_rate": 1.8975204622310157e-06,
+      "loss": 0.7201,
+      "step": 8995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5573322497141062,
+      "learning_rate": 1.896501680557869e-06,
+      "loss": 0.78,
+      "step": 8996
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5391471733953224,
+      "learning_rate": 1.895483108440026e-06,
+      "loss": 0.7493,
+      "step": 8997
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.774758916140538,
+      "learning_rate": 1.89446474594626e-06,
+      "loss": 1.0386,
+      "step": 8998
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7627233541379113,
+      "learning_rate": 1.8934465931453378e-06,
+      "loss": 1.076,
+      "step": 8999
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4947070734000882,
+      "learning_rate": 1.8924286501060047e-06,
+      "loss": 0.7342,
+      "step": 9000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.50427155173262,
+      "learning_rate": 1.8914109168969958e-06,
+      "loss": 0.6584,
+      "step": 9001
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4625939513214241,
+      "learning_rate": 1.8903933935870277e-06,
+      "loss": 0.6858,
+      "step": 9002
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9706659948365657,
+      "learning_rate": 1.8893760802448096e-06,
+      "loss": 0.7568,
+      "step": 9003
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5176789244714313,
+      "learning_rate": 1.888358976939032e-06,
+      "loss": 0.7798,
+      "step": 9004
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5215454887323965,
+      "learning_rate": 1.8873420837383715e-06,
+      "loss": 0.711,
+      "step": 9005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7904088371333343,
+      "learning_rate": 1.8863254007114912e-06,
+      "loss": 1.1006,
+      "step": 9006
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5006902598782765,
+      "learning_rate": 1.8853089279270393e-06,
+      "loss": 0.7748,
+      "step": 9007
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6195149948736198,
+      "learning_rate": 1.8842926654536508e-06,
+      "loss": 0.7725,
+      "step": 9008
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6218657421821607,
+      "learning_rate": 1.8832766133599445e-06,
+      "loss": 0.8206,
+      "step": 9009
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6206080808014003,
+      "learning_rate": 1.8822607717145291e-06,
+      "loss": 0.758,
+      "step": 9010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7568642495002619,
+      "learning_rate": 1.8812451405859966e-06,
+      "loss": 1.0645,
+      "step": 9011
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6008374004013357,
+      "learning_rate": 1.8802297200429215e-06,
+      "loss": 0.6304,
+      "step": 9012
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4946217142923581,
+      "learning_rate": 1.8792145101538712e-06,
+      "loss": 0.7501,
+      "step": 9013
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4814834799976933,
+      "learning_rate": 1.8781995109873929e-06,
+      "loss": 0.7281,
+      "step": 9014
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5421556514647208,
+      "learning_rate": 1.8771847226120227e-06,
+      "loss": 0.7254,
+      "step": 9015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5238179283677238,
+      "learning_rate": 1.8761701450962798e-06,
+      "loss": 0.7533,
+      "step": 9016
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5116826794758946,
+      "learning_rate": 1.8751557785086727e-06,
+      "loss": 0.7191,
+      "step": 9017
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4647682015478283,
+      "learning_rate": 1.8741416229176928e-06,
+      "loss": 0.7382,
+      "step": 9018
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5374191311853032,
+      "learning_rate": 1.8731276783918162e-06,
+      "loss": 0.8601,
+      "step": 9019
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5162653274580136,
+      "learning_rate": 1.8721139449995107e-06,
+      "loss": 0.7771,
+      "step": 9020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4756993788011128,
+      "learning_rate": 1.8711004228092233e-06,
+      "loss": 0.676,
+      "step": 9021
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5927597444191053,
+      "learning_rate": 1.8700871118893893e-06,
+      "loss": 0.6727,
+      "step": 9022
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5195555612379161,
+      "learning_rate": 1.8690740123084316e-06,
+      "loss": 0.803,
+      "step": 9023
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4816138609878466,
+      "learning_rate": 1.8680611241347557e-06,
+      "loss": 0.7798,
+      "step": 9024
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.521441307235049,
+      "learning_rate": 1.8670484474367551e-06,
+      "loss": 0.7175,
+      "step": 9025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5643611900422763,
+      "learning_rate": 1.8660359822828066e-06,
+      "loss": 0.8006,
+      "step": 9026
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4504252200206096,
+      "learning_rate": 1.8650237287412748e-06,
+      "loss": 0.7997,
+      "step": 9027
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.550270185730639,
+      "learning_rate": 1.8640116868805097e-06,
+      "loss": 0.7777,
+      "step": 9028
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7935406795138835,
+      "learning_rate": 1.8629998567688445e-06,
+      "loss": 1.0718,
+      "step": 9029
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4916607056112976,
+      "learning_rate": 1.861988238474604e-06,
+      "loss": 0.7689,
+      "step": 9030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4622805692691805,
+      "learning_rate": 1.8609768320660932e-06,
+      "loss": 0.7157,
+      "step": 9031
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8016868482922395,
+      "learning_rate": 1.8599656376116026e-06,
+      "loss": 0.8439,
+      "step": 9032
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6237913359486311,
+      "learning_rate": 1.8589546551794141e-06,
+      "loss": 0.7429,
+      "step": 9033
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.438161336741648,
+      "learning_rate": 1.8579438848377895e-06,
+      "loss": 0.7833,
+      "step": 9034
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.51947726688505,
+      "learning_rate": 1.8569333266549787e-06,
+      "loss": 0.78,
+      "step": 9035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4802229232012183,
+      "learning_rate": 1.8559229806992151e-06,
+      "loss": 0.7869,
+      "step": 9036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.451879407169695,
+      "learning_rate": 1.8549128470387229e-06,
+      "loss": 0.7266,
+      "step": 9037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4298293878553483,
+      "learning_rate": 1.8539029257417068e-06,
+      "loss": 0.7245,
+      "step": 9038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4616366727115857,
+      "learning_rate": 1.8528932168763592e-06,
+      "loss": 0.7598,
+      "step": 9039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5133308209743124,
+      "learning_rate": 1.851883720510858e-06,
+      "loss": 0.6988,
+      "step": 9040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4749691067192736,
+      "learning_rate": 1.8508744367133662e-06,
+      "loss": 0.6551,
+      "step": 9041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4783314330355415,
+      "learning_rate": 1.8498653655520337e-06,
+      "loss": 0.7165,
+      "step": 9042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7855912995164267,
+      "learning_rate": 1.8488565070949931e-06,
+      "loss": 1.0801,
+      "step": 9043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6081824859185057,
+      "learning_rate": 1.8478478614103684e-06,
+      "loss": 0.7813,
+      "step": 9044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.462659210647125,
+      "learning_rate": 1.8468394285662643e-06,
+      "loss": 0.7541,
+      "step": 9045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6442368625281671,
+      "learning_rate": 1.84583120863077e-06,
+      "loss": 0.7253,
+      "step": 9046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5951833801022253,
+      "learning_rate": 1.844823201671967e-06,
+      "loss": 0.8434,
+      "step": 9047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.54182252559575,
+      "learning_rate": 1.8438154077579157e-06,
+      "loss": 0.7709,
+      "step": 9048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7750038465905909,
+      "learning_rate": 1.8428078269566652e-06,
+      "loss": 1.0833,
+      "step": 9049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5046932926776329,
+      "learning_rate": 1.8418004593362498e-06,
+      "loss": 0.7428,
+      "step": 9050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5676843278827293,
+      "learning_rate": 1.8407933049646893e-06,
+      "loss": 0.8071,
+      "step": 9051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6406774070341008,
+      "learning_rate": 1.8397863639099884e-06,
+      "loss": 0.7393,
+      "step": 9052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3964933855449182,
+      "learning_rate": 1.8387796362401367e-06,
+      "loss": 0.6779,
+      "step": 9053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4978047015082183,
+      "learning_rate": 1.8377731220231144e-06,
+      "loss": 0.7074,
+      "step": 9054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.527750510764953,
+      "learning_rate": 1.8367668213268814e-06,
+      "loss": 0.7457,
+      "step": 9055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.506893828515584,
+      "learning_rate": 1.8357607342193844e-06,
+      "loss": 0.7193,
+      "step": 9056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5994736619599366,
+      "learning_rate": 1.83475486076856e-06,
+      "loss": 0.7697,
+      "step": 9057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4188940343381915,
+      "learning_rate": 1.8337492010423252e-06,
+      "loss": 0.7894,
+      "step": 9058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4554887404676946,
+      "learning_rate": 1.8327437551085842e-06,
+      "loss": 0.7793,
+      "step": 9059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.433485965788859,
+      "learning_rate": 1.8317385230352269e-06,
+      "loss": 0.8117,
+      "step": 9060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3707311275787393,
+      "learning_rate": 1.8307335048901299e-06,
+      "loss": 0.6985,
+      "step": 9061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4603135243118102,
+      "learning_rate": 1.829728700741153e-06,
+      "loss": 0.747,
+      "step": 9062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5288737673690223,
+      "learning_rate": 1.8287241106561422e-06,
+      "loss": 0.8003,
+      "step": 9063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.422483141260241,
+      "learning_rate": 1.8277197347029324e-06,
+      "loss": 0.6788,
+      "step": 9064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.36357307880521,
+      "learning_rate": 1.8267155729493403e-06,
+      "loss": 0.7426,
+      "step": 9065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5689701068379638,
+      "learning_rate": 1.8257116254631685e-06,
+      "loss": 0.7336,
+      "step": 9066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6465616098699154,
+      "learning_rate": 1.8247078923122046e-06,
+      "loss": 0.7397,
+      "step": 9067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3849591407513553,
+      "learning_rate": 1.8237043735642263e-06,
+      "loss": 0.7808,
+      "step": 9068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5558343400428012,
+      "learning_rate": 1.8227010692869918e-06,
+      "loss": 0.8138,
+      "step": 9069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4942701165724106,
+      "learning_rate": 1.8216979795482464e-06,
+      "loss": 0.7238,
+      "step": 9070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4897431903361553,
+      "learning_rate": 1.8206951044157212e-06,
+      "loss": 0.722,
+      "step": 9071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5522273170161136,
+      "learning_rate": 1.8196924439571323e-06,
+      "loss": 0.7658,
+      "step": 9072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5029159136555514,
+      "learning_rate": 1.8186899982401802e-06,
+      "loss": 0.7464,
+      "step": 9073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4495772517638867,
+      "learning_rate": 1.8176877673325555e-06,
+      "loss": 0.7627,
+      "step": 9074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5771838108111327,
+      "learning_rate": 1.8166857513019298e-06,
+      "loss": 0.7993,
+      "step": 9075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.582110062241659,
+      "learning_rate": 1.8156839502159606e-06,
+      "loss": 0.6791,
+      "step": 9076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5470807079626603,
+      "learning_rate": 1.814682364142291e-06,
+      "loss": 0.7233,
+      "step": 9077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6514109037357982,
+      "learning_rate": 1.8136809931485544e-06,
+      "loss": 0.7609,
+      "step": 9078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5858480028405824,
+      "learning_rate": 1.8126798373023624e-06,
+      "loss": 0.8134,
+      "step": 9079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4711547999056573,
+      "learning_rate": 1.811678896671314e-06,
+      "loss": 0.7916,
+      "step": 9080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4544737688905205,
+      "learning_rate": 1.8106781713229993e-06,
+      "loss": 0.7019,
+      "step": 9081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5628495377846643,
+      "learning_rate": 1.8096776613249872e-06,
+      "loss": 0.746,
+      "step": 9082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8108775697229388,
+      "learning_rate": 1.8086773667448359e-06,
+      "loss": 1.0602,
+      "step": 9083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7752499931276945,
+      "learning_rate": 1.8076772876500831e-06,
+      "loss": 1.076,
+      "step": 9084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4894829332743877,
+      "learning_rate": 1.8066774241082612e-06,
+      "loss": 0.733,
+      "step": 9085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.433497792699876,
+      "learning_rate": 1.8056777761868815e-06,
+      "loss": 0.7563,
+      "step": 9086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4844033200160542,
+      "learning_rate": 1.804678343953441e-06,
+      "loss": 0.7676,
+      "step": 9087
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.516014001918474,
+      "learning_rate": 1.8036791274754266e-06,
+      "loss": 0.681,
+      "step": 9088
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7933155424586475,
+      "learning_rate": 1.802680126820307e-06,
+      "loss": 1.043,
+      "step": 9089
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4707896316554034,
+      "learning_rate": 1.8016813420555346e-06,
+      "loss": 0.7876,
+      "step": 9090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.459385105568495,
+      "learning_rate": 1.8006827732485528e-06,
+      "loss": 0.7124,
+      "step": 9091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5613322008575725,
+      "learning_rate": 1.7996844204667858e-06,
+      "loss": 0.7837,
+      "step": 9092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4392652782869524,
+      "learning_rate": 1.7986862837776448e-06,
+      "loss": 0.7448,
+      "step": 9093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4542084096605008,
+      "learning_rate": 1.7976883632485258e-06,
+      "loss": 0.7714,
+      "step": 9094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7476804107469217,
+      "learning_rate": 1.7966906589468114e-06,
+      "loss": 0.7255,
+      "step": 9095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7510453886374318,
+      "learning_rate": 1.7956931709398684e-06,
+      "loss": 1.0774,
+      "step": 9096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.491524293257138,
+      "learning_rate": 1.794695899295048e-06,
+      "loss": 0.7334,
+      "step": 9097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5460250409611556,
+      "learning_rate": 1.7936988440796915e-06,
+      "loss": 0.765,
+      "step": 9098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.627236313825197,
+      "learning_rate": 1.7927020053611204e-06,
+      "loss": 0.7534,
+      "step": 9099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4942800176598443,
+      "learning_rate": 1.7917053832066444e-06,
+      "loss": 0.7092,
+      "step": 9100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4481504971410948,
+      "learning_rate": 1.790708977683555e-06,
+      "loss": 0.7142,
+      "step": 9101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4504175764977358,
+      "learning_rate": 1.789712788859136e-06,
+      "loss": 0.699,
+      "step": 9102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5420025872595933,
+      "learning_rate": 1.7887168168006498e-06,
+      "loss": 0.7575,
+      "step": 9103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.387958860348744,
+      "learning_rate": 1.7877210615753477e-06,
+      "loss": 0.6742,
+      "step": 9104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5337391832380043,
+      "learning_rate": 1.7867255232504644e-06,
+      "loss": 0.7344,
+      "step": 9105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5405551850846713,
+      "learning_rate": 1.7857302018932215e-06,
+      "loss": 0.803,
+      "step": 9106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7234841510109024,
+      "learning_rate": 1.7847350975708233e-06,
+      "loss": 1.0386,
+      "step": 9107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.502146422828761,
+      "learning_rate": 1.7837402103504653e-06,
+      "loss": 0.7489,
+      "step": 9108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5873641373470815,
+      "learning_rate": 1.7827455402993231e-06,
+      "loss": 0.819,
+      "step": 9109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5557865596874143,
+      "learning_rate": 1.7817510874845585e-06,
+      "loss": 0.7193,
+      "step": 9110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.508694521032566,
+      "learning_rate": 1.7807568519733175e-06,
+      "loss": 0.7807,
+      "step": 9111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5096515083787445,
+      "learning_rate": 1.7797628338327372e-06,
+      "loss": 0.7704,
+      "step": 9112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5344426154506512,
+      "learning_rate": 1.7787690331299334e-06,
+      "loss": 0.8115,
+      "step": 9113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6297134356488117,
+      "learning_rate": 1.7777754499320104e-06,
+      "loss": 0.7924,
+      "step": 9114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5495189129082467,
+      "learning_rate": 1.7767820843060575e-06,
+      "loss": 0.732,
+      "step": 9115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8021373905463902,
+      "learning_rate": 1.7757889363191484e-06,
+      "loss": 1.0688,
+      "step": 9116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6021482799543167,
+      "learning_rate": 1.774796006038343e-06,
+      "loss": 0.7465,
+      "step": 9117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4968258642243784,
+      "learning_rate": 1.7738032935306842e-06,
+      "loss": 0.784,
+      "step": 9118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.638683641500895,
+      "learning_rate": 1.772810798863206e-06,
+      "loss": 0.8248,
+      "step": 9119
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4415034491510006,
+      "learning_rate": 1.7718185221029217e-06,
+      "loss": 0.7458,
+      "step": 9120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.570349383337047,
+      "learning_rate": 1.770826463316831e-06,
+      "loss": 0.728,
+      "step": 9121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.495359630982366,
+      "learning_rate": 1.7698346225719232e-06,
+      "loss": 0.8612,
+      "step": 9122
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5634886432554336,
+      "learning_rate": 1.7688429999351681e-06,
+      "loss": 0.7975,
+      "step": 9123
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5376690130487365,
+      "learning_rate": 1.767851595473522e-06,
+      "loss": 0.788,
+      "step": 9124
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4947976694838772,
+      "learning_rate": 1.7668604092539255e-06,
+      "loss": 0.7445,
+      "step": 9125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4191060324094653,
+      "learning_rate": 1.7658694413433087e-06,
+      "loss": 0.7131,
+      "step": 9126
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5547640911181309,
+      "learning_rate": 1.7648786918085837e-06,
+      "loss": 0.7054,
+      "step": 9127
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4859956210165293,
+      "learning_rate": 1.763888160716644e-06,
+      "loss": 0.7457,
+      "step": 9128
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.468140694488055,
+      "learning_rate": 1.7628978481343772e-06,
+      "loss": 0.6829,
+      "step": 9129
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6418363705896792,
+      "learning_rate": 1.76190775412865e-06,
+      "loss": 0.7451,
+      "step": 9130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7305560046527197,
+      "learning_rate": 1.7609178787663135e-06,
+      "loss": 0.7623,
+      "step": 9131
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4927189414756228,
+      "learning_rate": 1.75992822211421e-06,
+      "loss": 0.8002,
+      "step": 9132
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.516872586062521,
+      "learning_rate": 1.7589387842391626e-06,
+      "loss": 0.7055,
+      "step": 9133
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.511803948932772,
+      "learning_rate": 1.7579495652079786e-06,
+      "loss": 0.7172,
+      "step": 9134
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.687899821702067,
+      "learning_rate": 1.7569605650874526e-06,
+      "loss": 0.8084,
+      "step": 9135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.557370313149239,
+      "learning_rate": 1.7559717839443664e-06,
+      "loss": 0.7309,
+      "step": 9136
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5722703795328898,
+      "learning_rate": 1.7549832218454826e-06,
+      "loss": 0.7164,
+      "step": 9137
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5441130774087743,
+      "learning_rate": 1.7539948788575524e-06,
+      "loss": 0.6253,
+      "step": 9138
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7887226525591595,
+      "learning_rate": 1.75300675504731e-06,
+      "loss": 1.0522,
+      "step": 9139
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.617572047805759,
+      "learning_rate": 1.7520188504814767e-06,
+      "loss": 0.7734,
+      "step": 9140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5936446499569914,
+      "learning_rate": 1.7510311652267576e-06,
+      "loss": 0.754,
+      "step": 9141
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4171089263601067,
+      "learning_rate": 1.7500436993498415e-06,
+      "loss": 0.7492,
+      "step": 9142
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.512967279741477,
+      "learning_rate": 1.7490564529174082e-06,
+      "loss": 0.7545,
+      "step": 9143
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4453487297923802,
+      "learning_rate": 1.7480694259961162e-06,
+      "loss": 0.7244,
+      "step": 9144
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.553368564469449,
+      "learning_rate": 1.7470826186526114e-06,
+      "loss": 0.8251,
+      "step": 9145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4786274111991848,
+      "learning_rate": 1.7460960309535286e-06,
+      "loss": 0.6713,
+      "step": 9146
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6417514608497272,
+      "learning_rate": 1.7451096629654813e-06,
+      "loss": 0.7595,
+      "step": 9147
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5466262900967662,
+      "learning_rate": 1.7441235147550728e-06,
+      "loss": 0.7004,
+      "step": 9148
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.58390902017322,
+      "learning_rate": 1.74313758638889e-06,
+      "loss": 0.7675,
+      "step": 9149
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4239810672195217,
+      "learning_rate": 1.7421518779335038e-06,
+      "loss": 0.7529,
+      "step": 9150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5490613028879163,
+      "learning_rate": 1.741166389455473e-06,
+      "loss": 0.7726,
+      "step": 9151
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.627729799866342,
+      "learning_rate": 1.7401811210213377e-06,
+      "loss": 0.7605,
+      "step": 9152
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5165486468148615,
+      "learning_rate": 1.7391960726976281e-06,
+      "loss": 0.8537,
+      "step": 9153
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5270314493592017,
+      "learning_rate": 1.7382112445508565e-06,
+      "loss": 0.6718,
+      "step": 9154
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.546583348521981,
+      "learning_rate": 1.7372266366475187e-06,
+      "loss": 0.7757,
+      "step": 9155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4414717647332593,
+      "learning_rate": 1.7362422490541003e-06,
+      "loss": 0.7204,
+      "step": 9156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5340612813870216,
+      "learning_rate": 1.7352580818370685e-06,
+      "loss": 0.7414,
+      "step": 9157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5140381274042096,
+      "learning_rate": 1.7342741350628767e-06,
+      "loss": 0.7866,
+      "step": 9158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.492093327098175,
+      "learning_rate": 1.7332904087979623e-06,
+      "loss": 0.7316,
+      "step": 9159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4981172755218095,
+      "learning_rate": 1.7323069031087498e-06,
+      "loss": 0.7695,
+      "step": 9160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.786308150197478,
+      "learning_rate": 1.7313236180616466e-06,
+      "loss": 1.068,
+      "step": 9161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.463436914739836,
+      "learning_rate": 1.7303405537230456e-06,
+      "loss": 0.6691,
+      "step": 9162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8130093397421394,
+      "learning_rate": 1.729357710159329e-06,
+      "loss": 0.7274,
+      "step": 9163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6080508084266667,
+      "learning_rate": 1.7283750874368577e-06,
+      "loss": 0.7824,
+      "step": 9164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0468505744872716,
+      "learning_rate": 1.7273926856219824e-06,
+      "loss": 0.7573,
+      "step": 9165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5189735981139543,
+      "learning_rate": 1.7264105047810341e-06,
+      "loss": 0.7795,
+      "step": 9166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5017136038688443,
+      "learning_rate": 1.725428544980336e-06,
+      "loss": 0.8509,
+      "step": 9167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4513155378685,
+      "learning_rate": 1.7244468062861897e-06,
+      "loss": 0.7278,
+      "step": 9168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5242345190456783,
+      "learning_rate": 1.7234652887648838e-06,
+      "loss": 0.7661,
+      "step": 9169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4897619809328135,
+      "learning_rate": 1.7224839924826959e-06,
+      "loss": 0.7498,
+      "step": 9170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7926154124161684,
+      "learning_rate": 1.7215029175058845e-06,
+      "loss": 1.0485,
+      "step": 9171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5438926616366768,
+      "learning_rate": 1.7205220639006893e-06,
+      "loss": 0.7366,
+      "step": 9172
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.539296307350383,
+      "learning_rate": 1.7195414317333453e-06,
+      "loss": 0.7905,
+      "step": 9173
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5138864036710689,
+      "learning_rate": 1.7185610210700654e-06,
+      "loss": 0.6634,
+      "step": 9174
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7577715114080709,
+      "learning_rate": 1.7175808319770482e-06,
+      "loss": 1.0754,
+      "step": 9175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3927023866629802,
+      "learning_rate": 1.7166008645204774e-06,
+      "loss": 0.731,
+      "step": 9176
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5990516454726418,
+      "learning_rate": 1.7156211187665262e-06,
+      "loss": 0.7743,
+      "step": 9177
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5457494203697455,
+      "learning_rate": 1.7146415947813472e-06,
+      "loss": 0.7716,
+      "step": 9178
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7159247920119958,
+      "learning_rate": 1.713662292631078e-06,
+      "loss": 0.7942,
+      "step": 9179
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5130877503498232,
+      "learning_rate": 1.7126832123818475e-06,
+      "loss": 0.8197,
+      "step": 9180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5197755172298342,
+      "learning_rate": 1.7117043540997635e-06,
+      "loss": 0.7664,
+      "step": 9181
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4396809460013287,
+      "learning_rate": 1.7107257178509202e-06,
+      "loss": 0.7901,
+      "step": 9182
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5297767596906362,
+      "learning_rate": 1.7097473037013984e-06,
+      "loss": 0.7664,
+      "step": 9183
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5133265517428123,
+      "learning_rate": 1.7087691117172617e-06,
+      "loss": 0.7531,
+      "step": 9184
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7759179337322503,
+      "learning_rate": 1.707791141964561e-06,
+      "loss": 1.0263,
+      "step": 9185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6756472773531088,
+      "learning_rate": 1.7068133945093285e-06,
+      "loss": 0.8381,
+      "step": 9186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4425332829739987,
+      "learning_rate": 1.7058358694175875e-06,
+      "loss": 0.794,
+      "step": 9187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5084276279472941,
+      "learning_rate": 1.7048585667553414e-06,
+      "loss": 0.7659,
+      "step": 9188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.715371952286538,
+      "learning_rate": 1.7038814865885779e-06,
+      "loss": 0.7565,
+      "step": 9189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7765993970604774,
+      "learning_rate": 1.7029046289832751e-06,
+      "loss": 0.7416,
+      "step": 9190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7593226878867999,
+      "learning_rate": 1.7019279940053906e-06,
+      "loss": 1.1064,
+      "step": 9191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5706020226482502,
+      "learning_rate": 1.7009515817208698e-06,
+      "loss": 0.7817,
+      "step": 9192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7599623905172376,
+      "learning_rate": 1.6999753921956425e-06,
+      "loss": 1.0828,
+      "step": 9193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7744798397643095,
+      "learning_rate": 1.6989994254956222e-06,
+      "loss": 1.0797,
+      "step": 9194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4775517024690286,
+      "learning_rate": 1.6980236816867086e-06,
+      "loss": 0.6729,
+      "step": 9195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4180842717675088,
+      "learning_rate": 1.6970481608347849e-06,
+      "loss": 0.7074,
+      "step": 9196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.509814104453549,
+      "learning_rate": 1.696072863005724e-06,
+      "loss": 0.6866,
+      "step": 9197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4640852579448582,
+      "learning_rate": 1.6950977882653779e-06,
+      "loss": 0.7856,
+      "step": 9198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4543695919843345,
+      "learning_rate": 1.6941229366795864e-06,
+      "loss": 0.7424,
+      "step": 9199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.650004186760314,
+      "learning_rate": 1.693148308314172e-06,
+      "loss": 0.7829,
+      "step": 9200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4993408708487548,
+      "learning_rate": 1.6921739032349472e-06,
+      "loss": 0.6591,
+      "step": 9201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4602831439000117,
+      "learning_rate": 1.691199721507704e-06,
+      "loss": 0.7702,
+      "step": 9202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3754567167413485,
+      "learning_rate": 1.6902257631982217e-06,
+      "loss": 0.851,
+      "step": 9203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6650390833222224,
+      "learning_rate": 1.6892520283722641e-06,
+      "loss": 0.6637,
+      "step": 9204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7721627355957261,
+      "learning_rate": 1.6882785170955801e-06,
+      "loss": 1.057,
+      "step": 9205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5172718835040933,
+      "learning_rate": 1.6873052294339015e-06,
+      "loss": 0.8031,
+      "step": 9206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.512150065910371,
+      "learning_rate": 1.6863321654529508e-06,
+      "loss": 0.7586,
+      "step": 9207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5274419661353846,
+      "learning_rate": 1.6853593252184292e-06,
+      "loss": 0.7734,
+      "step": 9208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5259049226602075,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.8279,
+      "step": 9209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5635729711579451,
+      "learning_rate": 1.6834143162514105e-06,
+      "loss": 0.7288,
+      "step": 9210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5879593116349144,
+      "learning_rate": 1.6824421476502467e-06,
+      "loss": 0.7137,
+      "step": 9211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5812325433392616,
+      "learning_rate": 1.6814702030581754e-06,
+      "loss": 0.8521,
+      "step": 9212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.548238748937519,
+      "learning_rate": 1.6804984825408227e-06,
+      "loss": 0.7431,
+      "step": 9213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4370822929663603,
+      "learning_rate": 1.6795269861638041e-06,
+      "loss": 0.6607,
+      "step": 9214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5458014722506537,
+      "learning_rate": 1.6785557139927183e-06,
+      "loss": 0.6944,
+      "step": 9215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7883556059575791,
+      "learning_rate": 1.6775846660931439e-06,
+      "loss": 1.0678,
+      "step": 9216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6140693381317457,
+      "learning_rate": 1.6766138425306483e-06,
+      "loss": 0.683,
+      "step": 9217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.478996386798182,
+      "learning_rate": 1.675643243370787e-06,
+      "loss": 0.7271,
+      "step": 9218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5605387969116387,
+      "learning_rate": 1.6746728686790952e-06,
+      "loss": 0.6783,
+      "step": 9219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4676724224422681,
+      "learning_rate": 1.6737027185210941e-06,
+      "loss": 0.7995,
+      "step": 9220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5439649777761801,
+      "learning_rate": 1.6727327929622928e-06,
+      "loss": 0.7497,
+      "step": 9221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7303049436263334,
+      "learning_rate": 1.6717630920681815e-06,
+      "loss": 0.7746,
+      "step": 9222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4467402883800913,
+      "learning_rate": 1.6707936159042364e-06,
+      "loss": 0.7097,
+      "step": 9223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.608174380629258,
+      "learning_rate": 1.669824364535918e-06,
+      "loss": 0.7583,
+      "step": 9224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.670051317342552,
+      "learning_rate": 1.6688553380286748e-06,
+      "loss": 0.7798,
+      "step": 9225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7645509018505278,
+      "learning_rate": 1.6678865364479362e-06,
+      "loss": 1.0619,
+      "step": 9226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3970654332887158,
+      "learning_rate": 1.6669179598591183e-06,
+      "loss": 0.7643,
+      "step": 9227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5606658443862658,
+      "learning_rate": 1.665949608327621e-06,
+      "loss": 0.8253,
+      "step": 9228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5686421252845744,
+      "learning_rate": 1.6649814819188304e-06,
+      "loss": 0.7244,
+      "step": 9229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4564306434564396,
+      "learning_rate": 1.664013580698114e-06,
+      "loss": 0.7325,
+      "step": 9230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5390458430165195,
+      "learning_rate": 1.6630459047308307e-06,
+      "loss": 0.7653,
+      "step": 9231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5035328713271894,
+      "learning_rate": 1.6620784540823182e-06,
+      "loss": 0.739,
+      "step": 9232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5373781548041978,
+      "learning_rate": 1.661111228817901e-06,
+      "loss": 0.6863,
+      "step": 9233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7522744764427861,
+      "learning_rate": 1.660144229002887e-06,
+      "loss": 1.0323,
+      "step": 9234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.559436927036574,
+      "learning_rate": 1.6591774547025735e-06,
+      "loss": 0.7992,
+      "step": 9235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4507404846430578,
+      "learning_rate": 1.6582109059822371e-06,
+      "loss": 0.77,
+      "step": 9236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5287730892018996,
+      "learning_rate": 1.6572445829071421e-06,
+      "loss": 0.7994,
+      "step": 9237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6472431580117108,
+      "learning_rate": 1.6562784855425362e-06,
+      "loss": 0.7361,
+      "step": 9238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5808352260586596,
+      "learning_rate": 1.6553126139536534e-06,
+      "loss": 0.7732,
+      "step": 9239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.408503030263893,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.7817,
+      "step": 9240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.500042840670344,
+      "learning_rate": 1.6533815483639094e-06,
+      "loss": 0.7767,
+      "step": 9241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7399428372035358,
+      "learning_rate": 1.6524163544934396e-06,
+      "loss": 1.0345,
+      "step": 9242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5005395534888941,
+      "learning_rate": 1.6514513866594728e-06,
+      "loss": 0.7814,
+      "step": 9243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4044075837864731,
+      "learning_rate": 1.6504866449271633e-06,
+      "loss": 0.7559,
+      "step": 9244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.534321523362935,
+      "learning_rate": 1.649522129361657e-06,
+      "loss": 0.726,
+      "step": 9245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4720599872272986,
+      "learning_rate": 1.6485578400280772e-06,
+      "loss": 0.7944,
+      "step": 9246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4471798852085327,
+      "learning_rate": 1.6475937769915357e-06,
+      "loss": 0.7075,
+      "step": 9247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.535462016350755,
+      "learning_rate": 1.646629940317128e-06,
+      "loss": 0.7691,
+      "step": 9248
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7694202822500037,
+      "learning_rate": 1.6456663300699349e-06,
+      "loss": 1.0643,
+      "step": 9249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.80988445376309,
+      "learning_rate": 1.6447029463150215e-06,
+      "loss": 0.7939,
+      "step": 9250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5110477954265107,
+      "learning_rate": 1.6437397891174357e-06,
+      "loss": 0.6929,
+      "step": 9251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.618009401875516,
+      "learning_rate": 1.6427768585422155e-06,
+      "loss": 0.8045,
+      "step": 9252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6452515639750989,
+      "learning_rate": 1.6418141546543787e-06,
+      "loss": 0.7225,
+      "step": 9253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3929055098430094,
+      "learning_rate": 1.640851677518927e-06,
+      "loss": 0.8451,
+      "step": 9254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5050459011784143,
+      "learning_rate": 1.6398894272008532e-06,
+      "loss": 0.7188,
+      "step": 9255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.520348374633757,
+      "learning_rate": 1.6389274037651288e-06,
+      "loss": 0.8369,
+      "step": 9256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7451661451877218,
+      "learning_rate": 1.6379656072767114e-06,
+      "loss": 1.0608,
+      "step": 9257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6917728561823426,
+      "learning_rate": 1.6370040378005426e-06,
+      "loss": 0.7818,
+      "step": 9258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.783513602621412,
+      "learning_rate": 1.636042695401554e-06,
+      "loss": 1.0335,
+      "step": 9259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5769046328979726,
+      "learning_rate": 1.6350815801446534e-06,
+      "loss": 0.8162,
+      "step": 9260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4722783739602536,
+      "learning_rate": 1.6341206920947373e-06,
+      "loss": 0.6947,
+      "step": 9261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5283386109159598,
+      "learning_rate": 1.6331600313166896e-06,
+      "loss": 0.847,
+      "step": 9262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.630699770796367,
+      "learning_rate": 1.6321995978753757e-06,
+      "loss": 0.7641,
+      "step": 9263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7600025328926445,
+      "learning_rate": 1.631239391835646e-06,
+      "loss": 1.0527,
+      "step": 9264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4807066417147805,
+      "learning_rate": 1.6302794132623346e-06,
+      "loss": 0.7316,
+      "step": 9265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.512905043349435,
+      "learning_rate": 1.6293196622202635e-06,
+      "loss": 0.686,
+      "step": 9266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.600044942799814,
+      "learning_rate": 1.6283601387742366e-06,
+      "loss": 0.7995,
+      "step": 9267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5083948243265983,
+      "learning_rate": 1.627400842989041e-06,
+      "loss": 0.7673,
+      "step": 9268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.340293452525885,
+      "learning_rate": 1.6264417749294543e-06,
+      "loss": 0.6478,
+      "step": 9269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7741641332582838,
+      "learning_rate": 1.6254829346602324e-06,
+      "loss": 1.0514,
+      "step": 9270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6080258312708362,
+      "learning_rate": 1.6245243222461198e-06,
+      "loss": 0.7511,
+      "step": 9271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.610195052359815,
+      "learning_rate": 1.6235659377518432e-06,
+      "loss": 0.7268,
+      "step": 9272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6105781832745332,
+      "learning_rate": 1.6226077812421154e-06,
+      "loss": 0.8147,
+      "step": 9273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.585715493929405,
+      "learning_rate": 1.6216498527816328e-06,
+      "loss": 0.8014,
+      "step": 9274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.514998315194074,
+      "learning_rate": 1.6206921524350754e-06,
+      "loss": 0.7876,
+      "step": 9275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5167648466853563,
+      "learning_rate": 1.6197346802671133e-06,
+      "loss": 0.7283,
+      "step": 9276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5544995250339124,
+      "learning_rate": 1.6187774363423946e-06,
+      "loss": 0.7421,
+      "step": 9277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4915929158597128,
+      "learning_rate": 1.6178204207255531e-06,
+      "loss": 0.7504,
+      "step": 9278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6543088200777536,
+      "learning_rate": 1.6168636334812126e-06,
+      "loss": 0.6583,
+      "step": 9279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5219753328589805,
+      "learning_rate": 1.6159070746739757e-06,
+      "loss": 0.7848,
+      "step": 9280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.521174967834369,
+      "learning_rate": 1.6149507443684314e-06,
+      "loss": 0.6967,
+      "step": 9281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.326527784619663,
+      "learning_rate": 1.613994642629153e-06,
+      "loss": 0.683,
+      "step": 9282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4723098870235911,
+      "learning_rate": 1.6130387695206989e-06,
+      "loss": 0.7515,
+      "step": 9283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6414680552081728,
+      "learning_rate": 1.6120831251076118e-06,
+      "loss": 0.6897,
+      "step": 9284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8734205261911465,
+      "learning_rate": 1.611127709454418e-06,
+      "loss": 0.7207,
+      "step": 9285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4064049558298468,
+      "learning_rate": 1.6101725226256316e-06,
+      "loss": 0.6783,
+      "step": 9286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4609954451409313,
+      "learning_rate": 1.6092175646857477e-06,
+      "loss": 0.7457,
+      "step": 9287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7764374254643186,
+      "learning_rate": 1.6082628356992453e-06,
+      "loss": 1.0757,
+      "step": 9288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6288135262362202,
+      "learning_rate": 1.607308335730594e-06,
+      "loss": 0.8444,
+      "step": 9289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5785551893804821,
+      "learning_rate": 1.6063540648442416e-06,
+      "loss": 0.7623,
+      "step": 9290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4384018372810115,
+      "learning_rate": 1.6054000231046229e-06,
+      "loss": 0.7919,
+      "step": 9291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.501478010359784,
+      "learning_rate": 1.604446210576157e-06,
+      "loss": 0.7168,
+      "step": 9292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.503561579706657,
+      "learning_rate": 1.603492627323247e-06,
+      "loss": 0.7536,
+      "step": 9293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7831922004193947,
+      "learning_rate": 1.6025392734102818e-06,
+      "loss": 1.0634,
+      "step": 9294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6359060872652826,
+      "learning_rate": 1.6015861489016316e-06,
+      "loss": 0.8094,
+      "step": 9295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5614625869077492,
+      "learning_rate": 1.6006332538616576e-06,
+      "loss": 0.727,
+      "step": 9296
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4612492630861058,
+      "learning_rate": 1.599680588354699e-06,
+      "loss": 0.7989,
+      "step": 9297
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5093069878704097,
+      "learning_rate": 1.5987281524450827e-06,
+      "loss": 0.6717,
+      "step": 9298
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.491069171032876,
+      "learning_rate": 1.597775946197117e-06,
+      "loss": 0.7329,
+      "step": 9299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5503829951855062,
+      "learning_rate": 1.5968239696751008e-06,
+      "loss": 0.7361,
+      "step": 9300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6043661239408993,
+      "learning_rate": 1.595872222943312e-06,
+      "loss": 0.8271,
+      "step": 9301
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7646790815430762,
+      "learning_rate": 1.5949207060660138e-06,
+      "loss": 1.0575,
+      "step": 9302
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4624025351710768,
+      "learning_rate": 1.5939694191074584e-06,
+      "loss": 0.7349,
+      "step": 9303
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4498767021985481,
+      "learning_rate": 1.593018362131874e-06,
+      "loss": 0.7973,
+      "step": 9304
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7568908379803088,
+      "learning_rate": 1.5920675352034792e-06,
+      "loss": 1.0259,
+      "step": 9305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5348741051838808,
+      "learning_rate": 1.5911169383864788e-06,
+      "loss": 0.7379,
+      "step": 9306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.77274836459209,
+      "learning_rate": 1.5901665717450582e-06,
+      "loss": 1.0556,
+      "step": 9307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5825214226733677,
+      "learning_rate": 1.589216435343387e-06,
+      "loss": 0.8059,
+      "step": 9308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5619764992705174,
+      "learning_rate": 1.5882665292456196e-06,
+      "loss": 0.7698,
+      "step": 9309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5812557786816608,
+      "learning_rate": 1.5873168535158995e-06,
+      "loss": 0.8173,
+      "step": 9310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4858207470537388,
+      "learning_rate": 1.586367408218349e-06,
+      "loss": 0.7379,
+      "step": 9311
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.573705484876075,
+      "learning_rate": 1.5854181934170747e-06,
+      "loss": 0.8165,
+      "step": 9312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9865636395053377,
+      "learning_rate": 1.5844692091761742e-06,
+      "loss": 0.6449,
+      "step": 9313
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6289190219703689,
+      "learning_rate": 1.5835204555597227e-06,
+      "loss": 0.7739,
+      "step": 9314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4566499135503725,
+      "learning_rate": 1.5825719326317817e-06,
+      "loss": 0.7341,
+      "step": 9315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7393616460863047,
+      "learning_rate": 1.581623640456399e-06,
+      "loss": 1.0345,
+      "step": 9316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7367761828711421,
+      "learning_rate": 1.5806755790976042e-06,
+      "loss": 1.1046,
+      "step": 9317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6437811345784097,
+      "learning_rate": 1.5797277486194136e-06,
+      "loss": 0.8238,
+      "step": 9318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7398185199898041,
+      "learning_rate": 1.578780149085824e-06,
+      "loss": 1.0275,
+      "step": 9319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.524114929252351,
+      "learning_rate": 1.577832780560824e-06,
+      "loss": 0.8089,
+      "step": 9320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.732820995424231,
+      "learning_rate": 1.5768856431083796e-06,
+      "loss": 1.0244,
+      "step": 9321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.520648622203859,
+      "learning_rate": 1.575938736792444e-06,
+      "loss": 0.7459,
+      "step": 9322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6720845091489884,
+      "learning_rate": 1.5749920616769526e-06,
+      "loss": 0.7432,
+      "step": 9323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5752076498848167,
+      "learning_rate": 1.5740456178258312e-06,
+      "loss": 0.7534,
+      "step": 9324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7656547323835673,
+      "learning_rate": 1.573099405302983e-06,
+      "loss": 1.0793,
+      "step": 9325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4598723493327956,
+      "learning_rate": 1.5721534241722996e-06,
+      "loss": 0.7454,
+      "step": 9326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.52080128296719,
+      "learning_rate": 1.5712076744976551e-06,
+      "loss": 0.774,
+      "step": 9327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5184292019755365,
+      "learning_rate": 1.5702621563429088e-06,
+      "loss": 0.7763,
+      "step": 9328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7480769700078229,
+      "learning_rate": 1.5693168697719024e-06,
+      "loss": 1.045,
+      "step": 9329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.514693176337305,
+      "learning_rate": 1.5683718148484673e-06,
+      "loss": 0.7569,
+      "step": 9330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4611740412755145,
+      "learning_rate": 1.5674269916364144e-06,
+      "loss": 0.7537,
+      "step": 9331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4797044361959082,
+      "learning_rate": 1.5664824001995398e-06,
+      "loss": 0.8024,
+      "step": 9332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4908193089495525,
+      "learning_rate": 1.5655380406016236e-06,
+      "loss": 0.686,
+      "step": 9333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.497739064921044,
+      "learning_rate": 1.5645939129064336e-06,
+      "loss": 0.7413,
+      "step": 9334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6781102045070344,
+      "learning_rate": 1.5636500171777181e-06,
+      "loss": 0.7157,
+      "step": 9335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4867146672118166,
+      "learning_rate": 1.5627063534792114e-06,
+      "loss": 0.7213,
+      "step": 9336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4921822080551297,
+      "learning_rate": 1.561762921874631e-06,
+      "loss": 0.7336,
+      "step": 9337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4695350273912682,
+      "learning_rate": 1.5608197224276806e-06,
+      "loss": 0.6651,
+      "step": 9338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5780790419708863,
+      "learning_rate": 1.5598767552020465e-06,
+      "loss": 0.7528,
+      "step": 9339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6381150503757502,
+      "learning_rate": 1.558934020261399e-06,
+      "loss": 0.7542,
+      "step": 9340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4306859683528428,
+      "learning_rate": 1.5579915176693961e-06,
+      "loss": 0.601,
+      "step": 9341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.765686831728065,
+      "learning_rate": 1.5570492474896764e-06,
+      "loss": 1.0522,
+      "step": 9342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5087266556349734,
+      "learning_rate": 1.5561072097858632e-06,
+      "loss": 0.6761,
+      "step": 9343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5140856290499989,
+      "learning_rate": 1.555165404621567e-06,
+      "loss": 0.7989,
+      "step": 9344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.549546754785703,
+      "learning_rate": 1.5542238320603802e-06,
+      "loss": 0.7371,
+      "step": 9345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.587478897101142,
+      "learning_rate": 1.5532824921658779e-06,
+      "loss": 0.7753,
+      "step": 9346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4780175105451603,
+      "learning_rate": 1.5523413850016268e-06,
+      "loss": 0.6871,
+      "step": 9347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7900091158926302,
+      "learning_rate": 1.5514005106311668e-06,
+      "loss": 1.0435,
+      "step": 9348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7518779682499173,
+      "learning_rate": 1.55045986911803e-06,
+      "loss": 1.0957,
+      "step": 9349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.542555639093399,
+      "learning_rate": 1.549519460525729e-06,
+      "loss": 0.7253,
+      "step": 9350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5429150129512195,
+      "learning_rate": 1.548579284917766e-06,
+      "loss": 0.7297,
+      "step": 9351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4813810834592238,
+      "learning_rate": 1.547639342357622e-06,
+      "loss": 0.8038,
+      "step": 9352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4109893814578336,
+      "learning_rate": 1.5466996329087618e-06,
+      "loss": 0.7827,
+      "step": 9353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4879734012714863,
+      "learning_rate": 1.5457601566346403e-06,
+      "loss": 0.7153,
+      "step": 9354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6387247773932332,
+      "learning_rate": 1.544820913598692e-06,
+      "loss": 0.8315,
+      "step": 9355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.588921834241734,
+      "learning_rate": 1.5438819038643366e-06,
+      "loss": 0.733,
+      "step": 9356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8525921625690882,
+      "learning_rate": 1.5429431274949757e-06,
+      "loss": 0.7845,
+      "step": 9357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5034040917603801,
+      "learning_rate": 1.5420045845540022e-06,
+      "loss": 0.7521,
+      "step": 9358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4760828134560084,
+      "learning_rate": 1.5410662751047855e-06,
+      "loss": 0.7029,
+      "step": 9359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5592814931972074,
+      "learning_rate": 1.5401281992106838e-06,
+      "loss": 0.7583,
+      "step": 9360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4745409887184293,
+      "learning_rate": 1.5391903569350375e-06,
+      "loss": 0.7255,
+      "step": 9361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4962860509232834,
+      "learning_rate": 1.5382527483411718e-06,
+      "loss": 0.7852,
+      "step": 9362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5551120498666697,
+      "learning_rate": 1.5373153734923945e-06,
+      "loss": 0.8046,
+      "step": 9363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.614838133850024,
+      "learning_rate": 1.5363782324520033e-06,
+      "loss": 0.676,
+      "step": 9364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5341121556399981,
+      "learning_rate": 1.5354413252832735e-06,
+      "loss": 0.753,
+      "step": 9365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4231076287033806,
+      "learning_rate": 1.5345046520494678e-06,
+      "loss": 0.7464,
+      "step": 9366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4557759749343377,
+      "learning_rate": 1.5335682128138302e-06,
+      "loss": 0.7055,
+      "step": 9367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.530119325209531,
+      "learning_rate": 1.5326320076395955e-06,
+      "loss": 0.7054,
+      "step": 9368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5163051727627412,
+      "learning_rate": 1.5316960365899757e-06,
+      "loss": 0.7615,
+      "step": 9369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.547794600921958,
+      "learning_rate": 1.5307602997281706e-06,
+      "loss": 0.7973,
+      "step": 9370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7056470821871852,
+      "learning_rate": 1.5298247971173636e-06,
+      "loss": 0.7582,
+      "step": 9371
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5559951881422502,
+      "learning_rate": 1.5288895288207205e-06,
+      "loss": 0.7856,
+      "step": 9372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5254452157091636,
+      "learning_rate": 1.5279544949013935e-06,
+      "loss": 0.8384,
+      "step": 9373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.635836028364804,
+      "learning_rate": 1.5270196954225175e-06,
+      "loss": 0.7397,
+      "step": 9374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5843857805712795,
+      "learning_rate": 1.526085130447214e-06,
+      "loss": 0.7013,
+      "step": 9375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5067935640392722,
+      "learning_rate": 1.5251508000385862e-06,
+      "loss": 0.7062,
+      "step": 9376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4183903956735697,
+      "learning_rate": 1.5242167042597206e-06,
+      "loss": 0.7839,
+      "step": 9377
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5252103139283095,
+      "learning_rate": 1.523282843173693e-06,
+      "loss": 0.7214,
+      "step": 9378
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.535968052418803,
+      "learning_rate": 1.5223492168435572e-06,
+      "loss": 0.779,
+      "step": 9379
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5535147632856845,
+      "learning_rate": 1.5214158253323546e-06,
+      "loss": 0.8615,
+      "step": 9380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.587171770504689,
+      "learning_rate": 1.5204826687031099e-06,
+      "loss": 0.7623,
+      "step": 9381
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5441700413327935,
+      "learning_rate": 1.5195497470188314e-06,
+      "loss": 0.7738,
+      "step": 9382
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.596611069874904,
+      "learning_rate": 1.5186170603425132e-06,
+      "loss": 0.778,
+      "step": 9383
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4557260709693693,
+      "learning_rate": 1.5176846087371293e-06,
+      "loss": 0.7563,
+      "step": 9384
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.414139599520713,
+      "learning_rate": 1.5167523922656458e-06,
+      "loss": 0.6932,
+      "step": 9385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4637195469447526,
+      "learning_rate": 1.5158204109910051e-06,
+      "loss": 0.7164,
+      "step": 9386
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4897474283832115,
+      "learning_rate": 1.5148886649761363e-06,
+      "loss": 0.7773,
+      "step": 9387
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7052931892247636,
+      "learning_rate": 1.513957154283955e-06,
+      "loss": 0.8368,
+      "step": 9388
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4796652064570948,
+      "learning_rate": 1.5130258789773583e-06,
+      "loss": 0.6982,
+      "step": 9389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4716998710833489,
+      "learning_rate": 1.5120948391192274e-06,
+      "loss": 0.7213,
+      "step": 9390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7495318536391611,
+      "learning_rate": 1.5111640347724293e-06,
+      "loss": 1.053,
+      "step": 9391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5074496611903578,
+      "learning_rate": 1.5102334659998124e-06,
+      "loss": 0.8159,
+      "step": 9392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4605941232842066,
+      "learning_rate": 1.509303132864212e-06,
+      "loss": 0.7737,
+      "step": 9393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.741380277165443,
+      "learning_rate": 1.5083730354284449e-06,
+      "loss": 0.7388,
+      "step": 9394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5408172244127545,
+      "learning_rate": 1.5074431737553158e-06,
+      "loss": 0.6265,
+      "step": 9395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6035364143856408,
+      "learning_rate": 1.5065135479076098e-06,
+      "loss": 0.7978,
+      "step": 9396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5354664714527295,
+      "learning_rate": 1.5055841579480974e-06,
+      "loss": 0.7002,
+      "step": 9397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.639513959373965,
+      "learning_rate": 1.5046550039395314e-06,
+      "loss": 0.7557,
+      "step": 9398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4766544727050783,
+      "learning_rate": 1.5037260859446535e-06,
+      "loss": 0.7943,
+      "step": 9399
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4615084195926922,
+      "learning_rate": 1.5027974040261855e-06,
+      "loss": 0.7291,
+      "step": 9400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.553546369932395,
+      "learning_rate": 1.5018689582468316e-06,
+      "loss": 0.7565,
+      "step": 9401
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.499828502084393,
+      "learning_rate": 1.5009407486692868e-06,
+      "loss": 0.6992,
+      "step": 9402
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.511486323622818,
+      "learning_rate": 1.5000127753562232e-06,
+      "loss": 0.693,
+      "step": 9403
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8327187105057239,
+      "learning_rate": 1.4990850383703005e-06,
+      "loss": 0.7871,
+      "step": 9404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.563588647705604,
+      "learning_rate": 1.498157537774161e-06,
+      "loss": 0.7392,
+      "step": 9405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5321043294229746,
+      "learning_rate": 1.4972302736304323e-06,
+      "loss": 0.8145,
+      "step": 9406
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5099653549498413,
+      "learning_rate": 1.4963032460017247e-06,
+      "loss": 0.7655,
+      "step": 9407
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.585409776736036,
+      "learning_rate": 1.4953764549506323e-06,
+      "loss": 0.7871,
+      "step": 9408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4959861665166552,
+      "learning_rate": 1.4944499005397372e-06,
+      "loss": 0.6751,
+      "step": 9409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4988184324233669,
+      "learning_rate": 1.4935235828316002e-06,
+      "loss": 0.7012,
+      "step": 9410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6490677551664028,
+      "learning_rate": 1.4925975018887678e-06,
+      "loss": 0.6928,
+      "step": 9411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6513745556280321,
+      "learning_rate": 1.491671657773774e-06,
+      "loss": 0.7112,
+      "step": 9412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.465946736566473,
+      "learning_rate": 1.4907460505491316e-06,
+      "loss": 0.7992,
+      "step": 9413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.635304546980997,
+      "learning_rate": 1.4898206802773408e-06,
+      "loss": 0.663,
+      "step": 9414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5284650025475603,
+      "learning_rate": 1.4888955470208837e-06,
+      "loss": 0.7224,
+      "step": 9415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6085022532294218,
+      "learning_rate": 1.4879706508422286e-06,
+      "loss": 0.843,
+      "step": 9416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4709493952870147,
+      "learning_rate": 1.4870459918038256e-06,
+      "loss": 0.7269,
+      "step": 9417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4996012349057388,
+      "learning_rate": 1.486121569968108e-06,
+      "loss": 0.7167,
+      "step": 9418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4375291186571901,
+      "learning_rate": 1.4851973853974987e-06,
+      "loss": 0.7409,
+      "step": 9419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.570236143258724,
+      "learning_rate": 1.4842734381543994e-06,
+      "loss": 0.7677,
+      "step": 9420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5411273441208346,
+      "learning_rate": 1.4833497283011967e-06,
+      "loss": 0.7622,
+      "step": 9421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5319867606181867,
+      "learning_rate": 1.4824262559002595e-06,
+      "loss": 0.8084,
+      "step": 9422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5018977059360492,
+      "learning_rate": 1.481503021013947e-06,
+      "loss": 0.7919,
+      "step": 9423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4779716643237604,
+      "learning_rate": 1.4805800237045958e-06,
+      "loss": 0.6948,
+      "step": 9424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.660853837845909,
+      "learning_rate": 1.4796572640345297e-06,
+      "loss": 0.8058,
+      "step": 9425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3929754804578618,
+      "learning_rate": 1.4787347420660541e-06,
+      "loss": 0.7202,
+      "step": 9426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7661811455217608,
+      "learning_rate": 1.4778124578614611e-06,
+      "loss": 1.0513,
+      "step": 9427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5315764418060995,
+      "learning_rate": 1.476890411483023e-06,
+      "loss": 0.7681,
+      "step": 9428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5337348892529243,
+      "learning_rate": 1.475968602993002e-06,
+      "loss": 0.7747,
+      "step": 9429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5830300581355987,
+      "learning_rate": 1.4750470324536393e-06,
+      "loss": 0.765,
+      "step": 9430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3825090451799005,
+      "learning_rate": 1.4741256999271607e-06,
+      "loss": 0.696,
+      "step": 9431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.592585856088709,
+      "learning_rate": 1.4732046054757765e-06,
+      "loss": 0.6759,
+      "step": 9432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4493961293843134,
+      "learning_rate": 1.4722837491616832e-06,
+      "loss": 0.7383,
+      "step": 9433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5901068753300758,
+      "learning_rate": 1.4713631310470571e-06,
+      "loss": 0.8079,
+      "step": 9434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3863167220821857,
+      "learning_rate": 1.4704427511940607e-06,
+      "loss": 0.7185,
+      "step": 9435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.558513537218464,
+      "learning_rate": 1.4695226096648423e-06,
+      "loss": 0.7708,
+      "step": 9436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5225431115866956,
+      "learning_rate": 1.4686027065215297e-06,
+      "loss": 0.7984,
+      "step": 9437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6689847567494265,
+      "learning_rate": 1.4676830418262372e-06,
+      "loss": 0.6857,
+      "step": 9438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5242494034987817,
+      "learning_rate": 1.466763615641061e-06,
+      "loss": 0.7716,
+      "step": 9439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5837501691709563,
+      "learning_rate": 1.4658444280280864e-06,
+      "loss": 0.7949,
+      "step": 9440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7707210931387947,
+      "learning_rate": 1.4649254790493773e-06,
+      "loss": 0.7751,
+      "step": 9441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9072790018558785,
+      "learning_rate": 1.4640067687669818e-06,
+      "loss": 0.7733,
+      "step": 9442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.443134721193111,
+      "learning_rate": 1.4630882972429367e-06,
+      "loss": 0.8092,
+      "step": 9443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5113089138249947,
+      "learning_rate": 1.4621700645392567e-06,
+      "loss": 0.763,
+      "step": 9444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4595872979348983,
+      "learning_rate": 1.4612520707179429e-06,
+      "loss": 0.7152,
+      "step": 9445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.655646916001771,
+      "learning_rate": 1.4603343158409823e-06,
+      "loss": 0.7994,
+      "step": 9446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5108828402666774,
+      "learning_rate": 1.4594167999703423e-06,
+      "loss": 0.7094,
+      "step": 9447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6977508112673279,
+      "learning_rate": 1.4584995231679778e-06,
+      "loss": 0.7123,
+      "step": 9448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5989252474850302,
+      "learning_rate": 1.457582485495821e-06,
+      "loss": 0.7355,
+      "step": 9449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7098801775461785,
+      "learning_rate": 1.4566656870157958e-06,
+      "loss": 0.7254,
+      "step": 9450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5973704030101157,
+      "learning_rate": 1.4557491277898062e-06,
+      "loss": 0.7493,
+      "step": 9451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.543448755295573,
+      "learning_rate": 1.454832807879738e-06,
+      "loss": 0.7598,
+      "step": 9452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6422934287795874,
+      "learning_rate": 1.4539167273474669e-06,
+      "loss": 0.7659,
+      "step": 9453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5071000225209674,
+      "learning_rate": 1.4530008862548472e-06,
+      "loss": 0.6921,
+      "step": 9454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5775984536071304,
+      "learning_rate": 1.4520852846637179e-06,
+      "loss": 0.7189,
+      "step": 9455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5071798338001885,
+      "learning_rate": 1.4511699226359016e-06,
+      "loss": 0.7134,
+      "step": 9456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.708951963378448,
+      "learning_rate": 1.450254800233209e-06,
+      "loss": 0.7781,
+      "step": 9457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7531615045021086,
+      "learning_rate": 1.4493399175174288e-06,
+      "loss": 1.0614,
+      "step": 9458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.375108876568127,
+      "learning_rate": 1.4484252745503363e-06,
+      "loss": 0.7687,
+      "step": 9459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5477001431370736,
+      "learning_rate": 1.4475108713936908e-06,
+      "loss": 0.7451,
+      "step": 9460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4644409094495916,
+      "learning_rate": 1.4465967081092346e-06,
+      "loss": 0.7551,
+      "step": 9461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5629186052833572,
+      "learning_rate": 1.4456827847586925e-06,
+      "loss": 0.7355,
+      "step": 9462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4175560672156606,
+      "learning_rate": 1.4447691014037774e-06,
+      "loss": 0.6657,
+      "step": 9463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5206269895611775,
+      "learning_rate": 1.4438556581061819e-06,
+      "loss": 0.6894,
+      "step": 9464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.501199603728742,
+      "learning_rate": 1.4429424549275845e-06,
+      "loss": 0.7311,
+      "step": 9465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4182629238077669,
+      "learning_rate": 1.4420294919296446e-06,
+      "loss": 0.7309,
+      "step": 9466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7592070394808534,
+      "learning_rate": 1.4411167691740109e-06,
+      "loss": 1.0692,
+      "step": 9467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5600202183335858,
+      "learning_rate": 1.4402042867223104e-06,
+      "loss": 0.7094,
+      "step": 9468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.604150752474942,
+      "learning_rate": 1.4392920446361563e-06,
+      "loss": 0.7081,
+      "step": 9469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5015229479113,
+      "learning_rate": 1.438380042977146e-06,
+      "loss": 0.7616,
+      "step": 9470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.48948228498623,
+      "learning_rate": 1.4374682818068586e-06,
+      "loss": 0.7925,
+      "step": 9471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5206480336609618,
+      "learning_rate": 1.4365567611868598e-06,
+      "loss": 0.8165,
+      "step": 9472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6591649324911522,
+      "learning_rate": 1.4356454811786947e-06,
+      "loss": 0.7714,
+      "step": 9473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7581917855484593,
+      "learning_rate": 1.434734441843899e-06,
+      "loss": 1.0318,
+      "step": 9474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4845694157894624,
+      "learning_rate": 1.4338236432439862e-06,
+      "loss": 0.7892,
+      "step": 9475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5593585834162063,
+      "learning_rate": 1.4329130854404537e-06,
+      "loss": 0.712,
+      "step": 9476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.552971160104241,
+      "learning_rate": 1.4320027684947878e-06,
+      "loss": 0.783,
+      "step": 9477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5167327666028858,
+      "learning_rate": 1.4310926924684542e-06,
+      "loss": 0.7783,
+      "step": 9478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.539114403488609,
+      "learning_rate": 1.4301828574229026e-06,
+      "loss": 0.7145,
+      "step": 9479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7850698692337855,
+      "learning_rate": 1.4292732634195677e-06,
+      "loss": 1.0832,
+      "step": 9480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5709632672979459,
+      "learning_rate": 1.4283639105198666e-06,
+      "loss": 0.6939,
+      "step": 9481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6573330335098013,
+      "learning_rate": 1.427454798785201e-06,
+      "loss": 0.7391,
+      "step": 9482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5566806261300232,
+      "learning_rate": 1.4265459282769556e-06,
+      "loss": 0.8388,
+      "step": 9483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4614264777372057,
+      "learning_rate": 1.4256372990565016e-06,
+      "loss": 0.6384,
+      "step": 9484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5199846475940837,
+      "learning_rate": 1.4247289111851902e-06,
+      "loss": 0.7961,
+      "step": 9485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7602728200499783,
+      "learning_rate": 1.423820764724357e-06,
+      "loss": 1.0664,
+      "step": 9486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3921099026872232,
+      "learning_rate": 1.4229128597353243e-06,
+      "loss": 0.7604,
+      "step": 9487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4878071251702487,
+      "learning_rate": 1.4220051962793952e-06,
+      "loss": 0.6812,
+      "step": 9488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5388958073244645,
+      "learning_rate": 1.4210977744178562e-06,
+      "loss": 0.7503,
+      "step": 9489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5015844076042044,
+      "learning_rate": 1.4201905942119782e-06,
+      "loss": 0.785,
+      "step": 9490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6119871619428312,
+      "learning_rate": 1.4192836557230182e-06,
+      "loss": 0.7523,
+      "step": 9491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703285087879522,
+      "learning_rate": 1.4183769590122138e-06,
+      "loss": 0.7554,
+      "step": 9492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5255974166207587,
+      "learning_rate": 1.4174705041407872e-06,
+      "loss": 0.7204,
+      "step": 9493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4910981011314883,
+      "learning_rate": 1.4165642911699435e-06,
+      "loss": 0.7629,
+      "step": 9494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5308285798863384,
+      "learning_rate": 1.4156583201608732e-06,
+      "loss": 0.8358,
+      "step": 9495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7769276522258671,
+      "learning_rate": 1.4147525911747495e-06,
+      "loss": 1.1205,
+      "step": 9496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6700819833988516,
+      "learning_rate": 1.413847104272727e-06,
+      "loss": 0.7884,
+      "step": 9497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4822811086769319,
+      "learning_rate": 1.41294185951595e-06,
+      "loss": 0.7822,
+      "step": 9498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4965040632647248,
+      "learning_rate": 1.4120368569655408e-06,
+      "loss": 0.7253,
+      "step": 9499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.569057144636915,
+      "learning_rate": 1.411132096682606e-06,
+      "loss": 0.8213,
+      "step": 9500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.564190279255054,
+      "learning_rate": 1.41022757872824e-06,
+      "loss": 0.7883,
+      "step": 9501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5931674786745857,
+      "learning_rate": 1.4093233031635163e-06,
+      "loss": 0.756,
+      "step": 9502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5266124900402602,
+      "learning_rate": 1.4084192700494942e-06,
+      "loss": 0.7322,
+      "step": 9503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3645980167352327,
+      "learning_rate": 1.4075154794472152e-06,
+      "loss": 0.7398,
+      "step": 9504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4992948730232998,
+      "learning_rate": 1.4066119314177056e-06,
+      "loss": 0.8278,
+      "step": 9505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5524571298765837,
+      "learning_rate": 1.4057086260219755e-06,
+      "loss": 0.7415,
+      "step": 9506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7629897494316663,
+      "learning_rate": 1.4048055633210162e-06,
+      "loss": 1.0728,
+      "step": 9507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4712892987489814,
+      "learning_rate": 1.4039027433758073e-06,
+      "loss": 0.6625,
+      "step": 9508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5533469940805962,
+      "learning_rate": 1.4030001662473086e-06,
+      "loss": 0.7277,
+      "step": 9509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.510039260252321,
+      "learning_rate": 1.4020978319964622e-06,
+      "loss": 0.8,
+      "step": 9510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7584247272877243,
+      "learning_rate": 1.4011957406841985e-06,
+      "loss": 0.7294,
+      "step": 9511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4803243765609533,
+      "learning_rate": 1.4002938923714282e-06,
+      "loss": 0.7919,
+      "step": 9512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4396424972932387,
+      "learning_rate": 1.3993922871190445e-06,
+      "loss": 0.6419,
+      "step": 9513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.53575935679358,
+      "learning_rate": 1.3984909249879275e-06,
+      "loss": 0.7236,
+      "step": 9514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.520809155464972,
+      "learning_rate": 1.3975898060389386e-06,
+      "loss": 0.7359,
+      "step": 9515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3608965450704584,
+      "learning_rate": 1.3966889303329233e-06,
+      "loss": 0.8068,
+      "step": 9516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4941496079261412,
+      "learning_rate": 1.3957882979307097e-06,
+      "loss": 0.7796,
+      "step": 9517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7711979180341535,
+      "learning_rate": 1.3948879088931128e-06,
+      "loss": 1.0452,
+      "step": 9518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.501333929489414,
+      "learning_rate": 1.3939877632809279e-06,
+      "loss": 0.729,
+      "step": 9519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.586746473000299,
+      "learning_rate": 1.3930878611549354e-06,
+      "loss": 0.7339,
+      "step": 9520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5145725388613107,
+      "learning_rate": 1.392188202575896e-06,
+      "loss": 0.7705,
+      "step": 9521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4955659026563959,
+      "learning_rate": 1.391288787604561e-06,
+      "loss": 0.6796,
+      "step": 9522
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.78207191330754,
+      "learning_rate": 1.3903896163016584e-06,
+      "loss": 1.0762,
+      "step": 9523
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.60961261475027,
+      "learning_rate": 1.389490688727903e-06,
+      "loss": 0.7435,
+      "step": 9524
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5909283086238848,
+      "learning_rate": 1.3885920049439921e-06,
+      "loss": 0.7914,
+      "step": 9525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5998790276966766,
+      "learning_rate": 1.387693565010607e-06,
+      "loss": 0.7114,
+      "step": 9526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4636930614697792,
+      "learning_rate": 1.3867953689884118e-06,
+      "loss": 0.7782,
+      "step": 9527
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6098758174786292,
+      "learning_rate": 1.3858974169380556e-06,
+      "loss": 0.8076,
+      "step": 9528
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.48981612578326,
+      "learning_rate": 1.3849997089201705e-06,
+      "loss": 0.6724,
+      "step": 9529
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7763186557175425,
+      "learning_rate": 1.3841022449953718e-06,
+      "loss": 1.0677,
+      "step": 9530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7769815046735407,
+      "learning_rate": 1.3832050252242552e-06,
+      "loss": 1.0539,
+      "step": 9531
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5549386648455308,
+      "learning_rate": 1.382308049667408e-06,
+      "loss": 0.8523,
+      "step": 9532
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5394146199977325,
+      "learning_rate": 1.3814113183853928e-06,
+      "loss": 0.7825,
+      "step": 9533
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4799590811626766,
+      "learning_rate": 1.380514831438759e-06,
+      "loss": 0.7043,
+      "step": 9534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.757302828762641,
+      "learning_rate": 1.3796185888880414e-06,
+      "loss": 1.0553,
+      "step": 9535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6250891623904644,
+      "learning_rate": 1.3787225907937552e-06,
+      "loss": 0.7852,
+      "step": 9536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.487694707527705,
+      "learning_rate": 1.3778268372164021e-06,
+      "loss": 0.7799,
+      "step": 9537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4034398294572177,
+      "learning_rate": 1.3769313282164597e-06,
+      "loss": 0.7011,
+      "step": 9538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5559268462818403,
+      "learning_rate": 1.3760360638544012e-06,
+      "loss": 0.7465,
+      "step": 9539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5438247917286623,
+      "learning_rate": 1.3751410441906737e-06,
+      "loss": 0.7612,
+      "step": 9540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5310085903196955,
+      "learning_rate": 1.37424626928571e-06,
+      "loss": 0.7319,
+      "step": 9541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.558354899647479,
+      "learning_rate": 1.3733517391999313e-06,
+      "loss": 0.7829,
+      "step": 9542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.394080410267665,
+      "learning_rate": 1.3724574539937352e-06,
+      "loss": 0.744,
+      "step": 9543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7482395514406452,
+      "learning_rate": 1.3715634137275052e-06,
+      "loss": 1.0562,
+      "step": 9544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.521987195592734,
+      "learning_rate": 1.3706696184616126e-06,
+      "loss": 0.7421,
+      "step": 9545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4225237564272504,
+      "learning_rate": 1.369776068256406e-06,
+      "loss": 0.7415,
+      "step": 9546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5887143749428199,
+      "learning_rate": 1.3688827631722202e-06,
+      "loss": 0.7465,
+      "step": 9547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.537577728954312,
+      "learning_rate": 1.3679897032693729e-06,
+      "loss": 0.8252,
+      "step": 9548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4387642407979229,
+      "learning_rate": 1.3670968886081664e-06,
+      "loss": 0.8484,
+      "step": 9549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4712697426459656,
+      "learning_rate": 1.366204319248885e-06,
+      "loss": 0.7363,
+      "step": 9550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4929074427098512,
+      "learning_rate": 1.3653119952517957e-06,
+      "loss": 0.6612,
+      "step": 9551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5131274738282197,
+      "learning_rate": 1.3644199166771531e-06,
+      "loss": 0.7756,
+      "step": 9552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6440583778847409,
+      "learning_rate": 1.363528083585191e-06,
+      "loss": 0.7472,
+      "step": 9553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8990310794231473,
+      "learning_rate": 1.3626364960361282e-06,
+      "loss": 0.8166,
+      "step": 9554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.530607098292728,
+      "learning_rate": 1.3617451540901649e-06,
+      "loss": 0.7544,
+      "step": 9555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5622471249158123,
+      "learning_rate": 1.3608540578074897e-06,
+      "loss": 0.745,
+      "step": 9556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5430420914705185,
+      "learning_rate": 1.35996320724827e-06,
+      "loss": 0.7289,
+      "step": 9557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4744979400074272,
+      "learning_rate": 1.3590726024726575e-06,
+      "loss": 0.6664,
+      "step": 9558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5299897805053908,
+      "learning_rate": 1.3581822435407889e-06,
+      "loss": 0.7714,
+      "step": 9559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5455249002153644,
+      "learning_rate": 1.3572921305127823e-06,
+      "loss": 0.7753,
+      "step": 9560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.51768656150467,
+      "learning_rate": 1.3564022634487395e-06,
+      "loss": 0.7367,
+      "step": 9561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.75018625671323,
+      "learning_rate": 1.355512642408749e-06,
+      "loss": 1.0487,
+      "step": 9562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5077494706405188,
+      "learning_rate": 1.3546232674528782e-06,
+      "loss": 0.7196,
+      "step": 9563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4253210284114108,
+      "learning_rate": 1.35373413864118e-06,
+      "loss": 0.8035,
+      "step": 9564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5095250189981502,
+      "learning_rate": 1.352845256033689e-06,
+      "loss": 0.7402,
+      "step": 9565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5414672968538685,
+      "learning_rate": 1.3519566196904278e-06,
+      "loss": 0.7586,
+      "step": 9566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5327785486704055,
+      "learning_rate": 1.3510682296713972e-06,
+      "loss": 0.7062,
+      "step": 9567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5192215335545376,
+      "learning_rate": 1.3501800860365838e-06,
+      "loss": 0.8137,
+      "step": 9568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6771151256886099,
+      "learning_rate": 1.3492921888459566e-06,
+      "loss": 0.8288,
+      "step": 9569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.496314295514803,
+      "learning_rate": 1.3484045381594684e-06,
+      "loss": 0.7965,
+      "step": 9570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5255046285968645,
+      "learning_rate": 1.3475171340370557e-06,
+      "loss": 0.691,
+      "step": 9571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7796495459882584,
+      "learning_rate": 1.346629976538637e-06,
+      "loss": 1.0578,
+      "step": 9572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6926105582105158,
+      "learning_rate": 1.3457430657241172e-06,
+      "loss": 0.7987,
+      "step": 9573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5147398155239917,
+      "learning_rate": 1.3448564016533821e-06,
+      "loss": 0.687,
+      "step": 9574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4791053719310312,
+      "learning_rate": 1.3439699843862986e-06,
+      "loss": 0.7945,
+      "step": 9575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.513324172175307,
+      "learning_rate": 1.3430838139827235e-06,
+      "loss": 0.7273,
+      "step": 9576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5623090074919714,
+      "learning_rate": 1.342197890502492e-06,
+      "loss": 0.7418,
+      "step": 9577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5601462746227026,
+      "learning_rate": 1.3413122140054219e-06,
+      "loss": 0.774,
+      "step": 9578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5420704990873941,
+      "learning_rate": 1.3404267845513165e-06,
+      "loss": 0.8706,
+      "step": 9579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5664559497710435,
+      "learning_rate": 1.3395416021999641e-06,
+      "loss": 0.7519,
+      "step": 9580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5870495164748304,
+      "learning_rate": 1.3386566670111339e-06,
+      "loss": 0.7746,
+      "step": 9581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7753911334820025,
+      "learning_rate": 1.3377719790445753e-06,
+      "loss": 1.0569,
+      "step": 9582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5529654079986897,
+      "learning_rate": 1.3368875383600277e-06,
+      "loss": 0.6815,
+      "step": 9583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7443475942759988,
+      "learning_rate": 1.3360033450172106e-06,
+      "loss": 1.0489,
+      "step": 9584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5967794829266049,
+      "learning_rate": 1.3351193990758237e-06,
+      "loss": 0.7362,
+      "step": 9585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7795442481231696,
+      "learning_rate": 1.3342357005955569e-06,
+      "loss": 0.644,
+      "step": 9586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4593669772466922,
+      "learning_rate": 1.3333522496360778e-06,
+      "loss": 0.7274,
+      "step": 9587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4896678987708647,
+      "learning_rate": 1.3324690462570395e-06,
+      "loss": 0.7804,
+      "step": 9588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.75187794423945,
+      "learning_rate": 1.3315860905180755e-06,
+      "loss": 1.0823,
+      "step": 9589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5217665948823567,
+      "learning_rate": 1.330703382478809e-06,
+      "loss": 0.7905,
+      "step": 9590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7527459970275915,
+      "learning_rate": 1.32982092219884e-06,
+      "loss": 1.0716,
+      "step": 9591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5160869901241196,
+      "learning_rate": 1.328938709737755e-06,
+      "loss": 0.6719,
+      "step": 9592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7461165581322784,
+      "learning_rate": 1.3280567451551224e-06,
+      "loss": 1.0517,
+      "step": 9593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5721973386120813,
+      "learning_rate": 1.3271750285104951e-06,
+      "loss": 0.8219,
+      "step": 9594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.55347573282422,
+      "learning_rate": 1.326293559863408e-06,
+      "loss": 0.7745,
+      "step": 9595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.747970967438058,
+      "learning_rate": 1.3254123392733793e-06,
+      "loss": 1.0202,
+      "step": 9596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5757800026573971,
+      "learning_rate": 1.3245313667999128e-06,
+      "loss": 0.7996,
+      "step": 9597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5454402984962872,
+      "learning_rate": 1.323650642502493e-06,
+      "loss": 0.7675,
+      "step": 9598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5216830322153205,
+      "learning_rate": 1.3227701664405868e-06,
+      "loss": 0.7983,
+      "step": 9599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.514376703242377,
+      "learning_rate": 1.3218899386736488e-06,
+      "loss": 0.7458,
+      "step": 9600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4694038662070055,
+      "learning_rate": 1.321009959261113e-06,
+      "loss": 0.749,
+      "step": 9601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.500593865052254,
+      "learning_rate": 1.3201302282623973e-06,
+      "loss": 0.7217,
+      "step": 9602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5297825820085655,
+      "learning_rate": 1.3192507457369025e-06,
+      "loss": 0.7406,
+      "step": 9603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.516061720849401,
+      "learning_rate": 1.3183715117440143e-06,
+      "loss": 0.7774,
+      "step": 9604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4324854677250756,
+      "learning_rate": 1.3174925263431005e-06,
+      "loss": 0.7988,
+      "step": 9605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.477445752999541,
+      "learning_rate": 1.31661378959351e-06,
+      "loss": 0.7148,
+      "step": 9606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4170814254456279,
+      "learning_rate": 1.3157353015545804e-06,
+      "loss": 0.6904,
+      "step": 9607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4226212130693878,
+      "learning_rate": 1.3148570622856282e-06,
+      "loss": 0.7634,
+      "step": 9608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.579406575996543,
+      "learning_rate": 1.3139790718459522e-06,
+      "loss": 0.712,
+      "step": 9609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.529053448061561,
+      "learning_rate": 1.3131013302948392e-06,
+      "loss": 0.7804,
+      "step": 9610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7718516249378474,
+      "learning_rate": 1.3122238376915546e-06,
+      "loss": 1.0767,
+      "step": 9611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5170924862015003,
+      "learning_rate": 1.3113465940953495e-06,
+      "loss": 0.7659,
+      "step": 9612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.45476092157081,
+      "learning_rate": 1.310469599565457e-06,
+      "loss": 0.7575,
+      "step": 9613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5173991993657474,
+      "learning_rate": 1.3095928541610936e-06,
+      "loss": 0.7719,
+      "step": 9614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.541216629205336,
+      "learning_rate": 1.3087163579414598e-06,
+      "loss": 0.751,
+      "step": 9615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5258925149485894,
+      "learning_rate": 1.3078401109657362e-06,
+      "loss": 0.7766,
+      "step": 9616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.566276447850946,
+      "learning_rate": 1.3069641132930928e-06,
+      "loss": 0.7484,
+      "step": 9617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5933224882959292,
+      "learning_rate": 1.3060883649826766e-06,
+      "loss": 0.7155,
+      "step": 9618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5523264636691527,
+      "learning_rate": 1.3052128660936193e-06,
+      "loss": 0.8087,
+      "step": 9619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.489490514051949,
+      "learning_rate": 1.3043376166850396e-06,
+      "loss": 0.7421,
+      "step": 9620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4436663030879915,
+      "learning_rate": 1.303462616816034e-06,
+      "loss": 0.751,
+      "step": 9621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5212745186773196,
+      "learning_rate": 1.302587866545686e-06,
+      "loss": 0.8891,
+      "step": 9622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.564278697001055,
+      "learning_rate": 1.3017133659330583e-06,
+      "loss": 0.8016,
+      "step": 9623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7740681373624709,
+      "learning_rate": 1.300839115037202e-06,
+      "loss": 1.0582,
+      "step": 9624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7636508265825822,
+      "learning_rate": 1.2999651139171487e-06,
+      "loss": 1.0318,
+      "step": 9625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6038596690907123,
+      "learning_rate": 1.299091362631909e-06,
+      "loss": 0.7499,
+      "step": 9626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5762151830768114,
+      "learning_rate": 1.2982178612404839e-06,
+      "loss": 0.7313,
+      "step": 9627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5122743103062353,
+      "learning_rate": 1.2973446098018543e-06,
+      "loss": 0.7747,
+      "step": 9628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6337470410589414,
+      "learning_rate": 1.2964716083749829e-06,
+      "loss": 0.7749,
+      "step": 9629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.762287517112837,
+      "learning_rate": 1.2955988570188155e-06,
+      "loss": 1.0562,
+      "step": 9630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7366152460884505,
+      "learning_rate": 1.2947263557922857e-06,
+      "loss": 1.0431,
+      "step": 9631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7462172535733403,
+      "learning_rate": 1.2938541047543046e-06,
+      "loss": 1.0501,
+      "step": 9632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5912063365960425,
+      "learning_rate": 1.2929821039637674e-06,
+      "loss": 0.7682,
+      "step": 9633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4230842832492974,
+      "learning_rate": 1.292110353479557e-06,
+      "loss": 0.7904,
+      "step": 9634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7534972158542466,
+      "learning_rate": 1.291238853360534e-06,
+      "loss": 1.036,
+      "step": 9635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4914004209373755,
+      "learning_rate": 1.2903676036655444e-06,
+      "loss": 0.7498,
+      "step": 9636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.459490155752423,
+      "learning_rate": 1.2894966044534164e-06,
+      "loss": 0.7023,
+      "step": 9637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5166228744595034,
+      "learning_rate": 1.2886258557829622e-06,
+      "loss": 0.6801,
+      "step": 9638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5130676308254578,
+      "learning_rate": 1.2877553577129776e-06,
+      "loss": 0.7627,
+      "step": 9639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4993624395259457,
+      "learning_rate": 1.2868851103022378e-06,
+      "loss": 0.7387,
+      "step": 9640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5671504758326247,
+      "learning_rate": 1.2860151136095073e-06,
+      "loss": 0.7967,
+      "step": 9641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4982710549100524,
+      "learning_rate": 1.2851453676935289e-06,
+      "loss": 0.7773,
+      "step": 9642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5099480439377788,
+      "learning_rate": 1.2842758726130283e-06,
+      "loss": 0.7138,
+      "step": 9643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6504818704952249,
+      "learning_rate": 1.2834066284267189e-06,
+      "loss": 0.8492,
+      "step": 9644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4625121203269378,
+      "learning_rate": 1.2825376351932921e-06,
+      "loss": 0.7139,
+      "step": 9645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7790379870518475,
+      "learning_rate": 1.281668892971425e-06,
+      "loss": 1.0983,
+      "step": 9646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5761273671731904,
+      "learning_rate": 1.2808004018197767e-06,
+      "loss": 0.7098,
+      "step": 9647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.486822069423648,
+      "learning_rate": 1.2799321617969895e-06,
+      "loss": 0.7429,
+      "step": 9648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5072308357759612,
+      "learning_rate": 1.2790641729616899e-06,
+      "loss": 0.8714,
+      "step": 9649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4892434754555675,
+      "learning_rate": 1.2781964353724836e-06,
+      "loss": 0.8282,
+      "step": 9650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6029313886177872,
+      "learning_rate": 1.277328949087966e-06,
+      "loss": 0.7627,
+      "step": 9651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7691575046233411,
+      "learning_rate": 1.27646171416671e-06,
+      "loss": 1.0607,
+      "step": 9652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4771049568696826,
+      "learning_rate": 1.275594730667274e-06,
+      "loss": 0.7175,
+      "step": 9653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6271113954122174,
+      "learning_rate": 1.2747279986481964e-06,
+      "loss": 0.7353,
+      "step": 9654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6580376215400123,
+      "learning_rate": 1.2738615181680043e-06,
+      "loss": 0.6362,
+      "step": 9655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5651713463623247,
+      "learning_rate": 1.272995289285202e-06,
+      "loss": 0.7288,
+      "step": 9656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4743068896551788,
+      "learning_rate": 1.2721293120582813e-06,
+      "loss": 0.7786,
+      "step": 9657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4996875300319146,
+      "learning_rate": 1.2712635865457129e-06,
+      "loss": 0.751,
+      "step": 9658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6617894287928787,
+      "learning_rate": 1.2703981128059534e-06,
+      "loss": 0.8388,
+      "step": 9659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.469818000939209,
+      "learning_rate": 1.269532890897441e-06,
+      "loss": 0.8071,
+      "step": 9660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5492949266297271,
+      "learning_rate": 1.2686679208785984e-06,
+      "loss": 0.695,
+      "step": 9661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5085355760952492,
+      "learning_rate": 1.2678032028078307e-06,
+      "loss": 0.7825,
+      "step": 9662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5490115079274849,
+      "learning_rate": 1.2669387367435243e-06,
+      "loss": 0.7017,
+      "step": 9663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6291960832211916,
+      "learning_rate": 1.2660745227440496e-06,
+      "loss": 0.6994,
+      "step": 9664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4615056493435754,
+      "learning_rate": 1.2652105608677628e-06,
+      "loss": 0.77,
+      "step": 9665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7709182794011195,
+      "learning_rate": 1.264346851172999e-06,
+      "loss": 1.084,
+      "step": 9666
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5278382425772092,
+      "learning_rate": 1.2634833937180756e-06,
+      "loss": 0.7725,
+      "step": 9667
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4560430178546355,
+      "learning_rate": 1.2626201885612999e-06,
+      "loss": 0.8331,
+      "step": 9668
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4407502640807077,
+      "learning_rate": 1.2617572357609565e-06,
+      "loss": 0.6644,
+      "step": 9669
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4560900145264788,
+      "learning_rate": 1.260894535375311e-06,
+      "loss": 0.8201,
+      "step": 9670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5638699710005548,
+      "learning_rate": 1.260032087462615e-06,
+      "loss": 0.7308,
+      "step": 9671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4466391721880791,
+      "learning_rate": 1.2591698920811057e-06,
+      "loss": 0.7436,
+      "step": 9672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5340369883760723,
+      "learning_rate": 1.2583079492889994e-06,
+      "loss": 0.7218,
+      "step": 9673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7354865172154489,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 1.0333,
+      "step": 9674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5244074894012882,
+      "learning_rate": 1.2565848217057774e-06,
+      "loss": 0.704,
+      "step": 9675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5500980992418643,
+      "learning_rate": 1.2557236370310132e-06,
+      "loss": 0.7786,
+      "step": 9676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.478064283396387,
+      "learning_rate": 1.2548627051783512e-06,
+      "loss": 0.752,
+      "step": 9677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5619859010516988,
+      "learning_rate": 1.254002026205921e-06,
+      "loss": 0.7787,
+      "step": 9678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4750369695521748,
+      "learning_rate": 1.2531416001718416e-06,
+      "loss": 0.7865,
+      "step": 9679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5518907212603452,
+      "learning_rate": 1.2522814271342093e-06,
+      "loss": 0.7942,
+      "step": 9680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4802713306744018,
+      "learning_rate": 1.2514215071511043e-06,
+      "loss": 0.7519,
+      "step": 9681
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4485021848995474,
+      "learning_rate": 1.2505618402805909e-06,
+      "loss": 0.7727,
+      "step": 9682
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.575285542803116,
+      "learning_rate": 1.2497024265807156e-06,
+      "loss": 0.8276,
+      "step": 9683
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.490654521252698,
+      "learning_rate": 1.2488432661095068e-06,
+      "loss": 0.7882,
+      "step": 9684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6594179052639808,
+      "learning_rate": 1.2479843589249796e-06,
+      "loss": 0.6682,
+      "step": 9685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5341369473561017,
+      "learning_rate": 1.2471257050851277e-06,
+      "loss": 0.7654,
+      "step": 9686
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5276446153195493,
+      "learning_rate": 1.24626730464793e-06,
+      "loss": 0.832,
+      "step": 9687
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4841018381231088,
+      "learning_rate": 1.2454091576713457e-06,
+      "loss": 0.6512,
+      "step": 9688
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5292958847370455,
+      "learning_rate": 1.2445512642133218e-06,
+      "loss": 0.7916,
+      "step": 9689
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5008857938022964,
+      "learning_rate": 1.2436936243317837e-06,
+      "loss": 0.7207,
+      "step": 9690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5411956990251847,
+      "learning_rate": 1.242836238084642e-06,
+      "loss": 0.7255,
+      "step": 9691
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4402039180309176,
+      "learning_rate": 1.2419791055297887e-06,
+      "loss": 0.8019,
+      "step": 9692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4532299315808908,
+      "learning_rate": 1.2411222267250988e-06,
+      "loss": 0.718,
+      "step": 9693
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.499403360795155,
+      "learning_rate": 1.240265601728432e-06,
+      "loss": 0.7704,
+      "step": 9694
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.660830598427563,
+      "learning_rate": 1.2394092305976274e-06,
+      "loss": 0.7415,
+      "step": 9695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6593789824022072,
+      "learning_rate": 1.238553113390512e-06,
+      "loss": 0.8008,
+      "step": 9696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.675224295136462,
+      "learning_rate": 1.2376972501648915e-06,
+      "loss": 0.8661,
+      "step": 9697
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5102014925657228,
+      "learning_rate": 1.2368416409785539e-06,
+      "loss": 0.8411,
+      "step": 9698
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5451480362890662,
+      "learning_rate": 1.2359862858892751e-06,
+      "loss": 0.8315,
+      "step": 9699
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7486421292393971,
+      "learning_rate": 1.2351311849548097e-06,
+      "loss": 1.0624,
+      "step": 9700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6105020103634604,
+      "learning_rate": 1.2342763382328954e-06,
+      "loss": 0.7851,
+      "step": 9701
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4177984611314585,
+      "learning_rate": 1.2334217457812536e-06,
+      "loss": 0.7019,
+      "step": 9702
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.460238474700881,
+      "learning_rate": 1.2325674076575884e-06,
+      "loss": 0.6801,
+      "step": 9703
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6881844776259283,
+      "learning_rate": 1.2317133239195866e-06,
+      "loss": 0.8541,
+      "step": 9704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7519646444334398,
+      "learning_rate": 1.2308594946249163e-06,
+      "loss": 1.0408,
+      "step": 9705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8080911051206523,
+      "learning_rate": 1.230005919831233e-06,
+      "loss": 0.7189,
+      "step": 9706
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4681397254596704,
+      "learning_rate": 1.2291525995961707e-06,
+      "loss": 0.759,
+      "step": 9707
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.431259141991019,
+      "learning_rate": 1.2282995339773456e-06,
+      "loss": 0.7548,
+      "step": 9708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5239988384287828,
+      "learning_rate": 1.2274467230323622e-06,
+      "loss": 0.7821,
+      "step": 9709
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.514760972254218,
+      "learning_rate": 1.226594166818803e-06,
+      "loss": 0.7801,
+      "step": 9710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5245903306172277,
+      "learning_rate": 1.2257418653942332e-06,
+      "loss": 0.7472,
+      "step": 9711
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5176787337899422,
+      "learning_rate": 1.2248898188162023e-06,
+      "loss": 0.7721,
+      "step": 9712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4211822420437397,
+      "learning_rate": 1.2240380271422459e-06,
+      "loss": 0.7768,
+      "step": 9713
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.69400932776177,
+      "learning_rate": 1.2231864904298746e-06,
+      "loss": 0.8345,
+      "step": 9714
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.441308202311701,
+      "learning_rate": 1.222335208736586e-06,
+      "loss": 0.758,
+      "step": 9715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.619096947655207,
+      "learning_rate": 1.2214841821198641e-06,
+      "loss": 0.816,
+      "step": 9716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4032898335832322,
+      "learning_rate": 1.2206334106371702e-06,
+      "loss": 0.7639,
+      "step": 9717
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5026527174104731,
+      "learning_rate": 1.219782894345949e-06,
+      "loss": 0.6903,
+      "step": 9718
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6200100515205298,
+      "learning_rate": 1.2189326333036323e-06,
+      "loss": 0.7141,
+      "step": 9719
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5216180766316332,
+      "learning_rate": 1.2180826275676294e-06,
+      "loss": 0.8207,
+      "step": 9720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4958479245592358,
+      "learning_rate": 1.2172328771953363e-06,
+      "loss": 0.6859,
+      "step": 9721
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.576534211431381,
+      "learning_rate": 1.2163833822441274e-06,
+      "loss": 0.7945,
+      "step": 9722
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4967091110773985,
+      "learning_rate": 1.2155341427713658e-06,
+      "loss": 0.8371,
+      "step": 9723
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3786310552474215,
+      "learning_rate": 1.2146851588343922e-06,
+      "loss": 0.7431,
+      "step": 9724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7666622668832891,
+      "learning_rate": 1.2138364304905326e-06,
+      "loss": 1.0507,
+      "step": 9725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5736941820800832,
+      "learning_rate": 1.212987957797095e-06,
+      "loss": 0.7474,
+      "step": 9726
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.524665198252292,
+      "learning_rate": 1.21213974081137e-06,
+      "loss": 0.796,
+      "step": 9727
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.431808437693155,
+      "learning_rate": 1.2112917795906309e-06,
+      "loss": 0.7194,
+      "step": 9728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.501053846411825,
+      "learning_rate": 1.2104440741921326e-06,
+      "loss": 0.7948,
+      "step": 9729
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.46751800037382,
+      "learning_rate": 1.2095966246731179e-06,
+      "loss": 0.7159,
+      "step": 9730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.488667175871821,
+      "learning_rate": 1.2087494310908056e-06,
+      "loss": 0.7573,
+      "step": 9731
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.587441875911726,
+      "learning_rate": 1.2079024935023998e-06,
+      "loss": 0.7805,
+      "step": 9732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7458707907196326,
+      "learning_rate": 1.2070558119650904e-06,
+      "loss": 1.0573,
+      "step": 9733
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7739026457481497,
+      "learning_rate": 1.2062093865360458e-06,
+      "loss": 1.0866,
+      "step": 9734
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5595732286686566,
+      "learning_rate": 1.2053632172724179e-06,
+      "loss": 0.7203,
+      "step": 9735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5407824636612668,
+      "learning_rate": 1.2045173042313429e-06,
+      "loss": 0.7588,
+      "step": 9736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.500108812381179,
+      "learning_rate": 1.2036716474699383e-06,
+      "loss": 0.7067,
+      "step": 9737
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5310776947466698,
+      "learning_rate": 1.202826247045305e-06,
+      "loss": 0.7994,
+      "step": 9738
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4854905065045652,
+      "learning_rate": 1.2019811030145245e-06,
+      "loss": 0.6918,
+      "step": 9739
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4509280298817377,
+      "learning_rate": 1.2011362154346668e-06,
+      "loss": 0.7545,
+      "step": 9740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9018023254750045,
+      "learning_rate": 1.2002915843627778e-06,
+      "loss": 0.7703,
+      "step": 9741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.482307323255835,
+      "learning_rate": 1.1994472098558884e-06,
+      "loss": 0.7509,
+      "step": 9742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3549629655139874,
+      "learning_rate": 1.198603091971015e-06,
+      "loss": 0.6778,
+      "step": 9743
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5192417194307386,
+      "learning_rate": 1.1977592307651536e-06,
+      "loss": 0.7921,
+      "step": 9744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5798550685357082,
+      "learning_rate": 1.196915626295283e-06,
+      "loss": 0.7758,
+      "step": 9745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4392321869405804,
+      "learning_rate": 1.196072278618366e-06,
+      "loss": 0.7229,
+      "step": 9746
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7736261874450823,
+      "learning_rate": 1.195229187791347e-06,
+      "loss": 1.0764,
+      "step": 9747
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5750348576908395,
+      "learning_rate": 1.1943863538711532e-06,
+      "loss": 0.7897,
+      "step": 9748
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.702834795877273,
+      "learning_rate": 1.193543776914693e-06,
+      "loss": 0.729,
+      "step": 9749
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6320803074220118,
+      "learning_rate": 1.1927014569788624e-06,
+      "loss": 0.7463,
+      "step": 9750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7885133474306262,
+      "learning_rate": 1.1918593941205358e-06,
+      "loss": 1.0433,
+      "step": 9751
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4500075413555893,
+      "learning_rate": 1.1910175883965708e-06,
+      "loss": 0.7061,
+      "step": 9752
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.378346423215973,
+      "learning_rate": 1.1901760398638062e-06,
+      "loss": 0.7557,
+      "step": 9753
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5610193765492888,
+      "learning_rate": 1.189334748579069e-06,
+      "loss": 0.7652,
+      "step": 9754
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4604862292763041,
+      "learning_rate": 1.1884937145991627e-06,
+      "loss": 0.7901,
+      "step": 9755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5771799074676869,
+      "learning_rate": 1.1876529379808749e-06,
+      "loss": 0.6872,
+      "step": 9756
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6100823743236223,
+      "learning_rate": 1.1868124187809815e-06,
+      "loss": 0.7079,
+      "step": 9757
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7797667891495405,
+      "learning_rate": 1.185972157056231e-06,
+      "loss": 1.066,
+      "step": 9758
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5192810219513084,
+      "learning_rate": 1.1851321528633608e-06,
+      "loss": 0.7288,
+      "step": 9759
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6745646065927855,
+      "learning_rate": 1.1842924062590922e-06,
+      "loss": 0.834,
+      "step": 9760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4885198010444902,
+      "learning_rate": 1.1834529173001253e-06,
+      "loss": 0.7685,
+      "step": 9761
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5709666958246562,
+      "learning_rate": 1.1826136860431443e-06,
+      "loss": 0.7525,
+      "step": 9762
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7136826220765258,
+      "learning_rate": 1.1817747125448148e-06,
+      "loss": 0.7255,
+      "step": 9763
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6006466238735344,
+      "learning_rate": 1.1809359968617894e-06,
+      "loss": 0.7523,
+      "step": 9764
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5182935819262782,
+      "learning_rate": 1.180097539050698e-06,
+      "loss": 0.7102,
+      "step": 9765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.597057624470285,
+      "learning_rate": 1.1792593391681545e-06,
+      "loss": 0.8017,
+      "step": 9766
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5438742813379012,
+      "learning_rate": 1.1784213972707581e-06,
+      "loss": 0.7846,
+      "step": 9767
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7502312227741539,
+      "learning_rate": 1.1775837134150875e-06,
+      "loss": 1.04,
+      "step": 9768
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5188904557511478,
+      "learning_rate": 1.1767462876577052e-06,
+      "loss": 0.8075,
+      "step": 9769
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5919566738337583,
+      "learning_rate": 1.175909120055156e-06,
+      "loss": 0.7172,
+      "step": 9770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6225264656067466,
+      "learning_rate": 1.1750722106639673e-06,
+      "loss": 0.6708,
+      "step": 9771
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5994974694142772,
+      "learning_rate": 1.1742355595406491e-06,
+      "loss": 0.7112,
+      "step": 9772
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7317714789861076,
+      "learning_rate": 1.1733991667416928e-06,
+      "loss": 1.0801,
+      "step": 9773
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2912399117457163,
+      "learning_rate": 1.1725630323235758e-06,
+      "loss": 0.7271,
+      "step": 9774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5055808834308115,
+      "learning_rate": 1.171727156342755e-06,
+      "loss": 0.6948,
+      "step": 9775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4431548789768909,
+      "learning_rate": 1.1708915388556707e-06,
+      "loss": 0.7237,
+      "step": 9776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7375877301613611,
+      "learning_rate": 1.1700561799187442e-06,
+      "loss": 1.0695,
+      "step": 9777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4304383426286993,
+      "learning_rate": 1.1692210795883835e-06,
+      "loss": 0.7183,
+      "step": 9778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7586317777390962,
+      "learning_rate": 1.1683862379209747e-06,
+      "loss": 1.0412,
+      "step": 9779
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5933403368830739,
+      "learning_rate": 1.1675516549728887e-06,
+      "loss": 0.761,
+      "step": 9780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5128870841844475,
+      "learning_rate": 1.1667173308004787e-06,
+      "loss": 0.7559,
+      "step": 9781
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7516243188097932,
+      "learning_rate": 1.1658832654600798e-06,
+      "loss": 1.0568,
+      "step": 9782
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5345618632440419,
+      "learning_rate": 1.1650494590080085e-06,
+      "loss": 0.7004,
+      "step": 9783
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.616817473468652,
+      "learning_rate": 1.164215911500568e-06,
+      "loss": 0.7813,
+      "step": 9784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6082228071908322,
+      "learning_rate": 1.1633826229940408e-06,
+      "loss": 0.8008,
+      "step": 9785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5918320137678663,
+      "learning_rate": 1.1625495935446918e-06,
+      "loss": 0.8516,
+      "step": 9786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5903029091880874,
+      "learning_rate": 1.1617168232087671e-06,
+      "loss": 0.7149,
+      "step": 9787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7626058045017418,
+      "learning_rate": 1.1608843120425012e-06,
+      "loss": 1.0636,
+      "step": 9788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4956867578873663,
+      "learning_rate": 1.1600520601021048e-06,
+      "loss": 0.7311,
+      "step": 9789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7030162017280477,
+      "learning_rate": 1.1592200674437742e-06,
+      "loss": 0.7908,
+      "step": 9790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6608069936930203,
+      "learning_rate": 1.1583883341236874e-06,
+      "loss": 0.8217,
+      "step": 9791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5405210971815246,
+      "learning_rate": 1.1575568601980043e-06,
+      "loss": 0.7321,
+      "step": 9792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7309868822689507,
+      "learning_rate": 1.1567256457228681e-06,
+      "loss": 1.0667,
+      "step": 9793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4965879483148223,
+      "learning_rate": 1.1558946907544034e-06,
+      "loss": 0.7403,
+      "step": 9794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.532478530999538,
+      "learning_rate": 1.1550639953487202e-06,
+      "loss": 0.792,
+      "step": 9795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5320896055897109,
+      "learning_rate": 1.1542335595619087e-06,
+      "loss": 0.7388,
+      "step": 9796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.549388449290148,
+      "learning_rate": 1.1534033834500391e-06,
+      "loss": 0.7209,
+      "step": 9797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4480360838824076,
+      "learning_rate": 1.1525734670691702e-06,
+      "loss": 0.7587,
+      "step": 9798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7074189503556427,
+      "learning_rate": 1.1517438104753386e-06,
+      "loss": 0.6619,
+      "step": 9799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7323665782056272,
+      "learning_rate": 1.1509144137245638e-06,
+      "loss": 1.0239,
+      "step": 9800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6005964044125225,
+      "learning_rate": 1.1500852768728515e-06,
+      "loss": 0.7681,
+      "step": 9801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4438866930342518,
+      "learning_rate": 1.1492563999761829e-06,
+      "loss": 0.6852,
+      "step": 9802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7645065682392673,
+      "learning_rate": 1.1484277830905277e-06,
+      "loss": 1.0719,
+      "step": 9803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6408394869742449,
+      "learning_rate": 1.1475994262718348e-06,
+      "loss": 0.7957,
+      "step": 9804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4873825765565303,
+      "learning_rate": 1.1467713295760386e-06,
+      "loss": 0.7247,
+      "step": 9805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4947620302894096,
+      "learning_rate": 1.1459434930590535e-06,
+      "loss": 0.7089,
+      "step": 9806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4840100849454885,
+      "learning_rate": 1.1451159167767745e-06,
+      "loss": 0.7988,
+      "step": 9807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.617265620475383,
+      "learning_rate": 1.1442886007850856e-06,
+      "loss": 0.7269,
+      "step": 9808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6467646391438429,
+      "learning_rate": 1.1434615451398467e-06,
+      "loss": 0.7153,
+      "step": 9809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6223480131670076,
+      "learning_rate": 1.142634749896903e-06,
+      "loss": 0.8032,
+      "step": 9810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4391058466220195,
+      "learning_rate": 1.1418082151120797e-06,
+      "loss": 0.6606,
+      "step": 9811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4610249275160936,
+      "learning_rate": 1.1409819408411898e-06,
+      "loss": 0.7111,
+      "step": 9812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5822543038533203,
+      "learning_rate": 1.140155927140023e-06,
+      "loss": 0.716,
+      "step": 9813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8677320064654974,
+      "learning_rate": 1.1393301740643542e-06,
+      "loss": 0.7118,
+      "step": 9814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7635182615014467,
+      "learning_rate": 1.1385046816699403e-06,
+      "loss": 1.0714,
+      "step": 9815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6602100218534417,
+      "learning_rate": 1.13767945001252e-06,
+      "loss": 0.7402,
+      "step": 9816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6010542368701006,
+      "learning_rate": 1.1368544791478132e-06,
+      "loss": 0.6883,
+      "step": 9817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7143439630482251,
+      "learning_rate": 1.136029769131527e-06,
+      "loss": 0.7529,
+      "step": 9818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4168672945516831,
+      "learning_rate": 1.1352053200193468e-06,
+      "loss": 0.7043,
+      "step": 9819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5616740727782192,
+      "learning_rate": 1.1343811318669407e-06,
+      "loss": 0.767,
+      "step": 9820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4904628913504443,
+      "learning_rate": 1.1335572047299582e-06,
+      "loss": 0.774,
+      "step": 9821
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6182970041525258,
+      "learning_rate": 1.1327335386640354e-06,
+      "loss": 0.7773,
+      "step": 9822
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4917124380420883,
+      "learning_rate": 1.1319101337247878e-06,
+      "loss": 0.6736,
+      "step": 9823
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8502454932864223,
+      "learning_rate": 1.1310869899678122e-06,
+      "loss": 0.7758,
+      "step": 9824
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7349262897712906,
+      "learning_rate": 1.1302641074486909e-06,
+      "loss": 1.0767,
+      "step": 9825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6599322203199398,
+      "learning_rate": 1.1294414862229847e-06,
+      "loss": 0.795,
+      "step": 9826
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7399811464589727,
+      "learning_rate": 1.1286191263462404e-06,
+      "loss": 1.0499,
+      "step": 9827
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7564933700779674,
+      "learning_rate": 1.1277970278739836e-06,
+      "loss": 1.0614,
+      "step": 9828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4949199448217396,
+      "learning_rate": 1.1269751908617277e-06,
+      "loss": 0.7903,
+      "step": 9829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.431780762907728,
+      "learning_rate": 1.1261536153649627e-06,
+      "loss": 0.7934,
+      "step": 9830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7638722004201676,
+      "learning_rate": 1.125332301439162e-06,
+      "loss": 1.1097,
+      "step": 9831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6883358873670051,
+      "learning_rate": 1.1245112491397859e-06,
+      "loss": 0.7276,
+      "step": 9832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.489026947929863,
+      "learning_rate": 1.1236904585222725e-06,
+      "loss": 0.7488,
+      "step": 9833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7476693707869921,
+      "learning_rate": 1.1228699296420425e-06,
+      "loss": 0.7694,
+      "step": 9834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5889766105652579,
+      "learning_rate": 1.1220496625545008e-06,
+      "loss": 0.8407,
+      "step": 9835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.591050339513009,
+      "learning_rate": 1.1212296573150332e-06,
+      "loss": 0.8001,
+      "step": 9836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5920593751824126,
+      "learning_rate": 1.1204099139790087e-06,
+      "loss": 0.795,
+      "step": 9837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4923120742993579,
+      "learning_rate": 1.119590432601776e-06,
+      "loss": 0.7192,
+      "step": 9838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5122267280194919,
+      "learning_rate": 1.1187712132386723e-06,
+      "loss": 0.7775,
+      "step": 9839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.515099090075609,
+      "learning_rate": 1.1179522559450112e-06,
+      "loss": 0.7554,
+      "step": 9840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7530193726246929,
+      "learning_rate": 1.1171335607760891e-06,
+      "loss": 1.0131,
+      "step": 9841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5490059137472882,
+      "learning_rate": 1.1163151277871892e-06,
+      "loss": 0.8482,
+      "step": 9842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.794090024004102,
+      "learning_rate": 1.1154969570335722e-06,
+      "loss": 1.067,
+      "step": 9843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6045849239379641,
+      "learning_rate": 1.1146790485704834e-06,
+      "loss": 0.7995,
+      "step": 9844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4335295952056695,
+      "learning_rate": 1.1138614024531497e-06,
+      "loss": 0.6258,
+      "step": 9845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6372444766629204,
+      "learning_rate": 1.1130440187367802e-06,
+      "loss": 0.7943,
+      "step": 9846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7043743210794442,
+      "learning_rate": 1.1122268974765665e-06,
+      "loss": 0.7695,
+      "step": 9847
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5356883791207006,
+      "learning_rate": 1.111410038727681e-06,
+      "loss": 0.7207,
+      "step": 9848
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7675642227592444,
+      "learning_rate": 1.1105934425452831e-06,
+      "loss": 1.0709,
+      "step": 9849
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4592159392136095,
+      "learning_rate": 1.1097771089845095e-06,
+      "loss": 0.7638,
+      "step": 9850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6077172196751315,
+      "learning_rate": 1.1089610381004812e-06,
+      "loss": 0.7851,
+      "step": 9851
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7714856337939658,
+      "learning_rate": 1.1081452299482999e-06,
+      "loss": 1.0556,
+      "step": 9852
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4196314992032064,
+      "learning_rate": 1.1073296845830529e-06,
+      "loss": 0.745,
+      "step": 9853
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7022937576726624,
+      "learning_rate": 1.1065144020598067e-06,
+      "loss": 0.7393,
+      "step": 9854
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5648123311305904,
+      "learning_rate": 1.1056993824336099e-06,
+      "loss": 0.7645,
+      "step": 9855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7178802731671781,
+      "learning_rate": 1.104884625759497e-06,
+      "loss": 1.0227,
+      "step": 9856
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4432176880726246,
+      "learning_rate": 1.1040701320924808e-06,
+      "loss": 0.6929,
+      "step": 9857
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4596193038806062,
+      "learning_rate": 1.1032559014875578e-06,
+      "loss": 0.7077,
+      "step": 9858
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7841268715039684,
+      "learning_rate": 1.1024419339997066e-06,
+      "loss": 1.0883,
+      "step": 9859
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3433293539264881,
+      "learning_rate": 1.1016282296838887e-06,
+      "loss": 0.6399,
+      "step": 9860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6163528774108946,
+      "learning_rate": 1.1008147885950472e-06,
+      "loss": 0.7454,
+      "step": 9861
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5280697928741949,
+      "learning_rate": 1.100001610788105e-06,
+      "loss": 0.7201,
+      "step": 9862
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6400171479130976,
+      "learning_rate": 1.0991886963179737e-06,
+      "loss": 0.7782,
+      "step": 9863
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4828515712647912,
+      "learning_rate": 1.0983760452395415e-06,
+      "loss": 0.7095,
+      "step": 9864
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5098946575176468,
+      "learning_rate": 1.0975636576076787e-06,
+      "loss": 0.7271,
+      "step": 9865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4188087783502072,
+      "learning_rate": 1.0967515334772428e-06,
+      "loss": 0.69,
+      "step": 9866
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.501738066377932,
+      "learning_rate": 1.0959396729030685e-06,
+      "loss": 0.7318,
+      "step": 9867
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.823683812317142,
+      "learning_rate": 1.0951280759399751e-06,
+      "loss": 0.7516,
+      "step": 9868
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.492994854797437,
+      "learning_rate": 1.0943167426427625e-06,
+      "loss": 0.7668,
+      "step": 9869
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7822132803101283,
+      "learning_rate": 1.0935056730662146e-06,
+      "loss": 1.0728,
+      "step": 9870
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7165255238823054,
+      "learning_rate": 1.0926948672650962e-06,
+      "loss": 0.8092,
+      "step": 9871
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5971143421669,
+      "learning_rate": 1.0918843252941537e-06,
+      "loss": 0.729,
+      "step": 9872
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5924784196165214,
+      "learning_rate": 1.0910740472081194e-06,
+      "loss": 0.7421,
+      "step": 9873
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5008518126994441,
+      "learning_rate": 1.0902640330617036e-06,
+      "loss": 0.7698,
+      "step": 9874
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7538002423896824,
+      "learning_rate": 1.0894542829095993e-06,
+      "loss": 1.0663,
+      "step": 9875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7539716316448627,
+      "learning_rate": 1.088644796806485e-06,
+      "loss": 1.108,
+      "step": 9876
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.489375556491458,
+      "learning_rate": 1.087835574807018e-06,
+      "loss": 0.854,
+      "step": 9877
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.108206439487225,
+      "learning_rate": 1.0870266169658383e-06,
+      "loss": 0.6777,
+      "step": 9878
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.488719596058092,
+      "learning_rate": 1.0862179233375697e-06,
+      "loss": 0.7429,
+      "step": 9879
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.641656566589072,
+      "learning_rate": 1.0854094939768156e-06,
+      "loss": 0.8047,
+      "step": 9880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5517601527455305,
+      "learning_rate": 1.084601328938164e-06,
+      "loss": 0.8748,
+      "step": 9881
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3909563526135977,
+      "learning_rate": 1.0837934282761825e-06,
+      "loss": 0.7519,
+      "step": 9882
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7322682018575278,
+      "learning_rate": 1.082985792045425e-06,
+      "loss": 1.0797,
+      "step": 9883
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4786547306148394,
+      "learning_rate": 1.0821784203004238e-06,
+      "loss": 0.76,
+      "step": 9884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.516290420633004,
+      "learning_rate": 1.0813713130956937e-06,
+      "loss": 0.6693,
+      "step": 9885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5157411693594403,
+      "learning_rate": 1.080564470485732e-06,
+      "loss": 0.8082,
+      "step": 9886
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.93919555750198,
+      "learning_rate": 1.0797578925250213e-06,
+      "loss": 0.8731,
+      "step": 9887
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5212995531353057,
+      "learning_rate": 1.0789515792680217e-06,
+      "loss": 0.7754,
+      "step": 9888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7705291944309108,
+      "learning_rate": 1.0781455307691767e-06,
+      "loss": 1.0471,
+      "step": 9889
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4811053407716568,
+      "learning_rate": 1.0773397470829145e-06,
+      "loss": 0.7033,
+      "step": 9890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7494576273191048,
+      "learning_rate": 1.0765342282636416e-06,
+      "loss": 1.0621,
+      "step": 9891
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6598604730943318,
+      "learning_rate": 1.0757289743657495e-06,
+      "loss": 0.739,
+      "step": 9892
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4520879018561774,
+      "learning_rate": 1.0749239854436089e-06,
+      "loss": 0.7564,
+      "step": 9893
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4444311135300092,
+      "learning_rate": 1.0741192615515772e-06,
+      "loss": 0.6963,
+      "step": 9894
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5249853746005788,
+      "learning_rate": 1.0733148027439904e-06,
+      "loss": 0.7243,
+      "step": 9895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4478012910489448,
+      "learning_rate": 1.0725106090751652e-06,
+      "loss": 0.7965,
+      "step": 9896
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7504689331268654,
+      "learning_rate": 1.0717066805994064e-06,
+      "loss": 1.0379,
+      "step": 9897
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4836974404816603,
+      "learning_rate": 1.0709030173709945e-06,
+      "loss": 0.698,
+      "step": 9898
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.489300103372873,
+      "learning_rate": 1.0700996194441944e-06,
+      "loss": 0.8164,
+      "step": 9899
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7572653679761618,
+      "learning_rate": 1.0692964868732552e-06,
+      "loss": 1.0682,
+      "step": 9900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7328233689880631,
+      "learning_rate": 1.0684936197124058e-06,
+      "loss": 1.0614,
+      "step": 9901
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6065354597497288,
+      "learning_rate": 1.067691018015858e-06,
+      "loss": 0.7501,
+      "step": 9902
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7603855919833004,
+      "learning_rate": 1.0668886818378022e-06,
+      "loss": 1.1045,
+      "step": 9903
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5174711666140337,
+      "learning_rate": 1.066086611232417e-06,
+      "loss": 0.7458,
+      "step": 9904
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7325618690199718,
+      "learning_rate": 1.0652848062538595e-06,
+      "loss": 0.664,
+      "step": 9905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4455362834855927,
+      "learning_rate": 1.0644832669562676e-06,
+      "loss": 0.7701,
+      "step": 9906
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.512443150042702,
+      "learning_rate": 1.0636819933937664e-06,
+      "loss": 0.7214,
+      "step": 9907
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4559207495817643,
+      "learning_rate": 1.062880985620458e-06,
+      "loss": 0.6966,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3672590107096925,
+      "learning_rate": 1.0620802436904275e-06,
+      "loss": 0.7051,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8436841145149356,
+      "learning_rate": 1.061279767657743e-06,
+      "loss": 0.7165,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5502076044513515,
+      "learning_rate": 1.060479557576456e-06,
+      "loss": 0.6701,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7768165685610678,
+      "learning_rate": 1.0596796135005976e-06,
+      "loss": 1.0755,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5104580199299305,
+      "learning_rate": 1.0588799354841817e-06,
+      "loss": 0.7043,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7394320967651677,
+      "learning_rate": 1.0580805235812042e-06,
+      "loss": 1.0645,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7330772966012828,
+      "learning_rate": 1.0572813778456442e-06,
+      "loss": 0.762,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7810648256856987,
+      "learning_rate": 1.0564824983314592e-06,
+      "loss": 1.0533,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4845291771217348,
+      "learning_rate": 1.0556838850925949e-06,
+      "loss": 0.7646,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.594312470393566,
+      "learning_rate": 1.0548855381829736e-06,
+      "loss": 0.6465,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7325057911174004,
+      "learning_rate": 1.0540874576565025e-06,
+      "loss": 0.9925,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6342221553449776,
+      "learning_rate": 1.053289643567067e-06,
+      "loss": 0.7191,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5023774684861384,
+      "learning_rate": 1.0524920959685414e-06,
+      "loss": 0.7377,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5492267150351635,
+      "learning_rate": 1.0516948149147755e-06,
+      "loss": 0.7442,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4662324958907165,
+      "learning_rate": 1.0508978004596043e-06,
+      "loss": 0.7311,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.462629929321534,
+      "learning_rate": 1.0501010526568439e-06,
+      "loss": 0.7265,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4578153095836521,
+      "learning_rate": 1.0493045715602924e-06,
+      "loss": 0.7703,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7530096834663219,
+      "learning_rate": 1.0485083572237297e-06,
+      "loss": 1.0742,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.763321345379345,
+      "learning_rate": 1.0477124097009172e-06,
+      "loss": 1.0566,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6643748837059584,
+      "learning_rate": 1.0469167290456016e-06,
+      "loss": 0.7817,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.554954979163268,
+      "learning_rate": 1.046121315311508e-06,
+      "loss": 0.7213,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7364978070971894,
+      "learning_rate": 1.0453261685523424e-06,
+      "loss": 1.0515,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5222219450088101,
+      "learning_rate": 1.0445312888217985e-06,
+      "loss": 0.7886,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7749117756961922,
+      "learning_rate": 1.0437366761735468e-06,
+      "loss": 1.0875,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5506141153928967,
+      "learning_rate": 1.0429423306612414e-06,
+      "loss": 0.7225,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.478725509426905,
+      "learning_rate": 1.0421482523385174e-06,
+      "loss": 0.8007,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.571929313678732,
+      "learning_rate": 1.0413544412589944e-06,
+      "loss": 0.6788,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5211343329506113,
+      "learning_rate": 1.040560897476271e-06,
+      "loss": 0.7765,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.496615796086339,
+      "learning_rate": 1.0397676210439283e-06,
+      "loss": 0.8217,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5729375399195828,
+      "learning_rate": 1.038974612015533e-06,
+      "loss": 0.816,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5276773121424616,
+      "learning_rate": 1.0381818704446296e-06,
+      "loss": 0.7977,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4034788353468814,
+      "learning_rate": 1.0373893963847436e-06,
+      "loss": 0.6919,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9235943425954065,
+      "learning_rate": 1.0365971898893884e-06,
+      "loss": 0.739,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4906536628240266,
+      "learning_rate": 1.0358052510120537e-06,
+      "loss": 0.7906,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4349694638584727,
+      "learning_rate": 1.0350135798062132e-06,
+      "loss": 0.8024,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4245457518317535,
+      "learning_rate": 1.0342221763253207e-06,
+      "loss": 0.7182,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4696083637500126,
+      "learning_rate": 1.0334310406228164e-06,
+      "loss": 0.666,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.606755677518499,
+      "learning_rate": 1.03264017275212e-06,
+      "loss": 0.7646,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.605240811464785,
+      "learning_rate": 1.0318495727666284e-06,
+      "loss": 0.8858,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7765754664569209,
+      "learning_rate": 1.0310592407197285e-06,
+      "loss": 1.0343,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5509434376885962,
+      "learning_rate": 1.0302691766647844e-06,
+      "loss": 0.8107,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.468541262536462,
+      "learning_rate": 1.029479380655143e-06,
+      "loss": 0.6667,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7754694891155596,
+      "learning_rate": 1.0286898527441308e-06,
+      "loss": 1.0359,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5940928941990904,
+      "learning_rate": 1.0279005929850626e-06,
+      "loss": 0.7508,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7483077481241737,
+      "learning_rate": 1.0271116014312293e-06,
+      "loss": 1.03,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.634862412598997,
+      "learning_rate": 1.0263228781359037e-06,
+      "loss": 0.7613,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4245357797207696,
+      "learning_rate": 1.025534423152345e-06,
+      "loss": 0.7681,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.463897032270917,
+      "learning_rate": 1.0247462365337901e-06,
+      "loss": 0.7232,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.506398343828152,
+      "learning_rate": 1.0239583183334596e-06,
+      "loss": 0.7364,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4700580875902514,
+      "learning_rate": 1.023170668604555e-06,
+      "loss": 0.7521,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.597444307123041,
+      "learning_rate": 1.0223832874002603e-06,
+      "loss": 0.6752,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6162452984148508,
+      "learning_rate": 1.021596174773742e-06,
+      "loss": 0.7665,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8112371082739316,
+      "learning_rate": 1.0208093307781452e-06,
+      "loss": 0.882,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5619348114369491,
+      "learning_rate": 1.0200227554666025e-06,
+      "loss": 0.7306,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.741890794862913,
+      "learning_rate": 1.0192364488922247e-06,
+      "loss": 1.0673,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5822581792262704,
+      "learning_rate": 1.0184504111081029e-06,
+      "loss": 0.8323,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5146102341317738,
+      "learning_rate": 1.0176646421673153e-06,
+      "loss": 0.7782,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6817894948429497,
+      "learning_rate": 1.0168791421229169e-06,
+      "loss": 0.7638,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7783819079503992,
+      "learning_rate": 1.0160939110279467e-06,
+      "loss": 1.0393,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5512464353912596,
+      "learning_rate": 1.0153089489354256e-06,
+      "loss": 0.6956,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4379954249807125,
+      "learning_rate": 1.014524255898356e-06,
+      "loss": 0.7222,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7593726451459591,
+      "learning_rate": 1.0137398319697217e-06,
+      "loss": 1.0745,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.424292545125485,
+      "learning_rate": 1.0129556772024874e-06,
+      "loss": 0.8022,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.769915689353312,
+      "learning_rate": 1.0121717916496043e-06,
+      "loss": 0.8187,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.535414714145598,
+      "learning_rate": 1.011388175364001e-06,
+      "loss": 0.7735,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.489133702307449,
+      "learning_rate": 1.010604828398587e-06,
+      "loss": 0.7515,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5283309079683616,
+      "learning_rate": 1.0098217508062587e-06,
+      "loss": 0.7379,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6369629642831451,
+      "learning_rate": 1.0090389426398894e-06,
+      "loss": 0.7574,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4667582311086997,
+      "learning_rate": 1.0082564039523368e-06,
+      "loss": 0.7276,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5152087624267274,
+      "learning_rate": 1.0074741347964395e-06,
+      "loss": 0.6962,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.540058188352714,
+      "learning_rate": 1.006692135225018e-06,
+      "loss": 0.74,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7627628662568969,
+      "learning_rate": 1.0059104052908753e-06,
+      "loss": 1.0515,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.518079240879994,
+      "learning_rate": 1.0051289450467933e-06,
+      "loss": 0.7465,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5273125539097983,
+      "learning_rate": 1.004347754545541e-06,
+      "loss": 0.7638,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7461493065854514,
+      "learning_rate": 1.0035668338398652e-06,
+      "loss": 0.8066,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7470959665143866,
+      "learning_rate": 1.0027861829824953e-06,
+      "loss": 1.0676,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5596749935107677,
+      "learning_rate": 1.002005802026141e-06,
+      "loss": 0.7336,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4822243966070272,
+      "learning_rate": 1.001225691023498e-06,
+      "loss": 0.7746,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.498488129895625,
+      "learning_rate": 1.0004458500272402e-06,
+      "loss": 0.6897,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5780239383879824,
+      "learning_rate": 9.99666279090023e-07,
+      "loss": 0.7652,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5596402968125291,
+      "learning_rate": 9.988869782644872e-07,
+      "loss": 0.7736,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4048992986528201,
+      "learning_rate": 9.98107947603253e-07,
+      "loss": 0.7363,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5054891743923948,
+      "learning_rate": 9.973291871589198e-07,
+      "loss": 0.7946,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5413199951514691,
+      "learning_rate": 9.96550696984071e-07,
+      "loss": 0.806,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5137538617524684,
+      "learning_rate": 9.957724771312754e-07,
+      "loss": 0.7591,
+      "step": 9993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4602703945897617,
+      "learning_rate": 9.949945276530782e-07,
+      "loss": 0.684,
+      "step": 9994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4244060529743618,
+      "learning_rate": 9.942168486020065e-07,
+      "loss": 0.68,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6239557235796354,
+      "learning_rate": 9.934394400305752e-07,
+      "loss": 0.6931,
+      "step": 9996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.549635646807986,
+      "learning_rate": 9.92662301991274e-07,
+      "loss": 0.7256,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5839919813283077,
+      "learning_rate": 9.918854345365758e-07,
+      "loss": 0.8486,
+      "step": 9998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.557121871986747,
+      "learning_rate": 9.911088377189405e-07,
+      "loss": 0.7209,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.73840477818183,
+      "learning_rate": 9.903325115908025e-07,
+      "loss": 1.0525,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6300699265410092,
+      "learning_rate": 9.895564562045822e-07,
+      "loss": 0.77,
+      "step": 10001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7703477922921379,
+      "learning_rate": 9.887806716126808e-07,
+      "loss": 1.0451,
+      "step": 10002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5493199621946296,
+      "learning_rate": 9.880051578674798e-07,
+      "loss": 0.7918,
+      "step": 10003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5314515802110467,
+      "learning_rate": 9.872299150213455e-07,
+      "loss": 0.7689,
+      "step": 10004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7578326334830287,
+      "learning_rate": 9.864549431266212e-07,
+      "loss": 1.0935,
+      "step": 10005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5065597472216277,
+      "learning_rate": 9.856802422356383e-07,
+      "loss": 0.6631,
+      "step": 10006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6057063613628346,
+      "learning_rate": 9.849058124007044e-07,
+      "loss": 0.7923,
+      "step": 10007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.547253788079528,
+      "learning_rate": 9.841316536741114e-07,
+      "loss": 0.8108,
+      "step": 10008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7596427705728889,
+      "learning_rate": 9.8335776610813e-07,
+      "loss": 1.0764,
+      "step": 10009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4690039205500238,
+      "learning_rate": 9.825841497550186e-07,
+      "loss": 0.6792,
+      "step": 10010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.512809560152084,
+      "learning_rate": 9.818108046670123e-07,
+      "loss": 0.7191,
+      "step": 10011
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.576048747684372,
+      "learning_rate": 9.810377308963282e-07,
+      "loss": 0.7979,
+      "step": 10012
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7526058482891663,
+      "learning_rate": 9.802649284951666e-07,
+      "loss": 1.0565,
+      "step": 10013
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5888677116655572,
+      "learning_rate": 9.794923975157083e-07,
+      "loss": 0.7316,
+      "step": 10014
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7516282159204318,
+      "learning_rate": 9.787201380101157e-07,
+      "loss": 1.0672,
+      "step": 10015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.580243063663469,
+      "learning_rate": 9.77948150030536e-07,
+      "loss": 0.7838,
+      "step": 10016
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5166439123361695,
+      "learning_rate": 9.77176433629094e-07,
+      "loss": 0.8469,
+      "step": 10017
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4448914120153866,
+      "learning_rate": 9.76404988857898e-07,
+      "loss": 0.7666,
+      "step": 10018
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5936146083393048,
+      "learning_rate": 9.75633815769036e-07,
+      "loss": 0.7674,
+      "step": 10019
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5069030230510183,
+      "learning_rate": 9.748629144145827e-07,
+      "loss": 0.8125,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6428484641500387,
+      "learning_rate": 9.740922848465894e-07,
+      "loss": 0.8094,
+      "step": 10021
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5243114974051433,
+      "learning_rate": 9.733219271170914e-07,
+      "loss": 0.6925,
+      "step": 10022
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.573857585439232,
+      "learning_rate": 9.725518412781037e-07,
+      "loss": 0.8026,
+      "step": 10023
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5717195668191197,
+      "learning_rate": 9.717820273816248e-07,
+      "loss": 0.8039,
+      "step": 10024
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.572345594584929,
+      "learning_rate": 9.71012485479635e-07,
+      "loss": 0.775,
+      "step": 10025
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4336179069881017,
+      "learning_rate": 9.702432156240937e-07,
+      "loss": 0.7668,
+      "step": 10026
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5750471469440321,
+      "learning_rate": 9.694742178669464e-07,
+      "loss": 0.7262,
+      "step": 10027
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5533922540323182,
+      "learning_rate": 9.687054922601157e-07,
+      "loss": 0.7376,
+      "step": 10028
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5524968021445842,
+      "learning_rate": 9.679370388555077e-07,
+      "loss": 0.7418,
+      "step": 10029
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.206263898031004,
+      "learning_rate": 9.671688577050114e-07,
+      "loss": 0.7759,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5711550428226237,
+      "learning_rate": 9.66400948860496e-07,
+      "loss": 0.7667,
+      "step": 10031
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.629589260593589,
+      "learning_rate": 9.656333123738116e-07,
+      "loss": 0.7673,
+      "step": 10032
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.453753722035826,
+      "learning_rate": 9.648659482967898e-07,
+      "loss": 0.7269,
+      "step": 10033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5407494070002221,
+      "learning_rate": 9.640988566812475e-07,
+      "loss": 0.7693,
+      "step": 10034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8335445615548538,
+      "learning_rate": 9.633320375789807e-07,
+      "loss": 0.8667,
+      "step": 10035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6074542537459433,
+      "learning_rate": 9.62565491041762e-07,
+      "loss": 0.7859,
+      "step": 10036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5770597358720424,
+      "learning_rate": 9.617992171213547e-07,
+      "loss": 0.7553,
+      "step": 10037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4048124219177545,
+      "learning_rate": 9.610332158694985e-07,
+      "loss": 0.7348,
+      "step": 10038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.469687655713396,
+      "learning_rate": 9.602674873379137e-07,
+      "loss": 0.7288,
+      "step": 10039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6229339728432595,
+      "learning_rate": 9.59502031578307e-07,
+      "loss": 0.7722,
+      "step": 10040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6541281315253025,
+      "learning_rate": 9.587368486423621e-07,
+      "loss": 0.7134,
+      "step": 10041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5334861149150596,
+      "learning_rate": 9.57971938581746e-07,
+      "loss": 0.7993,
+      "step": 10042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5350322184301337,
+      "learning_rate": 9.572073014481065e-07,
+      "loss": 0.7891,
+      "step": 10043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5199366027302554,
+      "learning_rate": 9.56442937293075e-07,
+      "loss": 0.7766,
+      "step": 10044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7301134784174522,
+      "learning_rate": 9.55678846168263e-07,
+      "loss": 1.0572,
+      "step": 10045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.592663081354399,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.8344,
+      "step": 10046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5822813077879134,
+      "learning_rate": 9.541514832156501e-07,
+      "loss": 0.8018,
+      "step": 10047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5946840898853993,
+      "learning_rate": 9.533882114909804e-07,
+      "loss": 0.7841,
+      "step": 10048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6320966740116614,
+      "learning_rate": 9.526252130027919e-07,
+      "loss": 0.6431,
+      "step": 10049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.551913307401585,
+      "learning_rate": 9.518624878026028e-07,
+      "loss": 0.7963,
+      "step": 10050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7544402790623255,
+      "learning_rate": 9.511000359419159e-07,
+      "loss": 1.1027,
+      "step": 10051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7548761842406448,
+      "learning_rate": 9.503378574722133e-07,
+      "loss": 1.0891,
+      "step": 10052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.291632442848924,
+      "learning_rate": 9.495759524449572e-07,
+      "loss": 0.7389,
+      "step": 10053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5290425083375367,
+      "learning_rate": 9.488143209115958e-07,
+      "loss": 0.7342,
+      "step": 10054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.46451791484958,
+      "learning_rate": 9.480529629235552e-07,
+      "loss": 0.68,
+      "step": 10055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5398079748499862,
+      "learning_rate": 9.472918785322444e-07,
+      "loss": 0.8076,
+      "step": 10056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5953283620873504,
+      "learning_rate": 9.465310677890522e-07,
+      "loss": 0.8309,
+      "step": 10057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5279790852075992,
+      "learning_rate": 9.457705307453519e-07,
+      "loss": 0.766,
+      "step": 10058
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.460450072714933,
+      "learning_rate": 9.450102674524952e-07,
+      "loss": 0.7248,
+      "step": 10059
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5441545125752085,
+      "learning_rate": 9.442502779618168e-07,
+      "loss": 0.778,
+      "step": 10060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5403706828580137,
+      "learning_rate": 9.434905623246343e-07,
+      "loss": 0.7208,
+      "step": 10061
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4957351803120378,
+      "learning_rate": 9.427311205922457e-07,
+      "loss": 0.7523,
+      "step": 10062
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6216676518096198,
+      "learning_rate": 9.419719528159271e-07,
+      "loss": 0.81,
+      "step": 10063
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5219800096421445,
+      "learning_rate": 9.412130590469438e-07,
+      "loss": 0.7479,
+      "step": 10064
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7416098807213682,
+      "learning_rate": 9.40454439336535e-07,
+      "loss": 1.0673,
+      "step": 10065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5714751296155174,
+      "learning_rate": 9.396960937359251e-07,
+      "loss": 0.8146,
+      "step": 10066
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7805042803665215,
+      "learning_rate": 9.389380222963195e-07,
+      "loss": 1.0822,
+      "step": 10067
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7504705881755415,
+      "learning_rate": 9.381802250689054e-07,
+      "loss": 1.0877,
+      "step": 10068
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5078948739238665,
+      "learning_rate": 9.374227021048499e-07,
+      "loss": 0.7925,
+      "step": 10069
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.682923275942174,
+      "learning_rate": 9.366654534553021e-07,
+      "loss": 0.7504,
+      "step": 10070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4479906332238635,
+      "learning_rate": 9.359084791713952e-07,
+      "loss": 0.7503,
+      "step": 10071
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7525012432358235,
+      "learning_rate": 9.351517793042408e-07,
+      "loss": 1.0901,
+      "step": 10072
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.73834634060162,
+      "learning_rate": 9.343953539049322e-07,
+      "loss": 1.0758,
+      "step": 10073
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.699823893955865,
+      "learning_rate": 9.336392030245473e-07,
+      "loss": 0.7704,
+      "step": 10074
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6103913498015554,
+      "learning_rate": 9.328833267141413e-07,
+      "loss": 0.7511,
+      "step": 10075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5853866310890927,
+      "learning_rate": 9.321277250247535e-07,
+      "loss": 0.8052,
+      "step": 10076
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4922542371440413,
+      "learning_rate": 9.313723980074018e-07,
+      "loss": 0.6834,
+      "step": 10077
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.593302284972048,
+      "learning_rate": 9.306173457130907e-07,
+      "loss": 0.7428,
+      "step": 10078
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7539440806364981,
+      "learning_rate": 9.298625681928031e-07,
+      "loss": 1.0396,
+      "step": 10079
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6084444942568388,
+      "learning_rate": 9.291080654974994e-07,
+      "loss": 0.7461,
+      "step": 10080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5076269613389197,
+      "learning_rate": 9.283538376781287e-07,
+      "loss": 0.761,
+      "step": 10081
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4999538985043177,
+      "learning_rate": 9.275998847856172e-07,
+      "loss": 0.7429,
+      "step": 10082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5260305013084035,
+      "learning_rate": 9.268462068708733e-07,
+      "loss": 0.7419,
+      "step": 10083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6304756674798115,
+      "learning_rate": 9.260928039847866e-07,
+      "loss": 0.8069,
+      "step": 10084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5049000628513265,
+      "learning_rate": 9.253396761782308e-07,
+      "loss": 0.7887,
+      "step": 10085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4947621776173312,
+      "learning_rate": 9.245868235020566e-07,
+      "loss": 0.7417,
+      "step": 10086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4639369683191377,
+      "learning_rate": 9.238342460070981e-07,
+      "loss": 0.7481,
+      "step": 10087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6452959363297952,
+      "learning_rate": 9.230819437441734e-07,
+      "loss": 0.8551,
+      "step": 10088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5502412084504489,
+      "learning_rate": 9.223299167640787e-07,
+      "loss": 0.8019,
+      "step": 10089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4851281093890076,
+      "learning_rate": 9.215781651175915e-07,
+      "loss": 0.7098,
+      "step": 10090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5374814833216,
+      "learning_rate": 9.208266888554729e-07,
+      "loss": 0.7741,
+      "step": 10091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7667298922962859,
+      "learning_rate": 9.200754880284635e-07,
+      "loss": 1.037,
+      "step": 10092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7556848812728668,
+      "learning_rate": 9.193245626872871e-07,
+      "loss": 1.068,
+      "step": 10093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5187962965390294,
+      "learning_rate": 9.185739128826454e-07,
+      "loss": 0.707,
+      "step": 10094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4856766099985772,
+      "learning_rate": 9.178235386652273e-07,
+      "loss": 0.7335,
+      "step": 10095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5159176186665357,
+      "learning_rate": 9.170734400856979e-07,
+      "loss": 0.7727,
+      "step": 10096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7337127933395794,
+      "learning_rate": 9.163236171947054e-07,
+      "loss": 1.0473,
+      "step": 10097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.533979402870509,
+      "learning_rate": 9.155740700428811e-07,
+      "loss": 0.8186,
+      "step": 10098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6984259719417314,
+      "learning_rate": 9.148247986808351e-07,
+      "loss": 0.7964,
+      "step": 10099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4770933395935582,
+      "learning_rate": 9.140758031591601e-07,
+      "loss": 0.7375,
+      "step": 10100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4928466786751593,
+      "learning_rate": 9.133270835284303e-07,
+      "loss": 0.7234,
+      "step": 10101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6410857605634024,
+      "learning_rate": 9.125786398392e-07,
+      "loss": 0.7761,
+      "step": 10102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5635427244039393,
+      "learning_rate": 9.118304721420068e-07,
+      "loss": 0.7941,
+      "step": 10103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5304782788452176,
+      "learning_rate": 9.110825804873668e-07,
+      "loss": 0.7546,
+      "step": 10104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5821809740114878,
+      "learning_rate": 9.10334964925782e-07,
+      "loss": 0.7638,
+      "step": 10105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4865433069373912,
+      "learning_rate": 9.095876255077318e-07,
+      "loss": 0.7078,
+      "step": 10106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4723806850996077,
+      "learning_rate": 9.088405622836788e-07,
+      "loss": 0.7581,
+      "step": 10107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5385109388010025,
+      "learning_rate": 9.080937753040647e-07,
+      "loss": 0.6749,
+      "step": 10108
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.441215162981642,
+      "learning_rate": 9.073472646193171e-07,
+      "loss": 0.7107,
+      "step": 10109
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7555390709094947,
+      "learning_rate": 9.066010302798401e-07,
+      "loss": 1.0552,
+      "step": 10110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3280024778236452,
+      "learning_rate": 9.05855072336022e-07,
+      "loss": 0.7225,
+      "step": 10111
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5359475863593215,
+      "learning_rate": 9.051093908382313e-07,
+      "loss": 0.7226,
+      "step": 10112
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4726214283860126,
+      "learning_rate": 9.04363985836818e-07,
+      "loss": 0.7456,
+      "step": 10113
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.52020436345313,
+      "learning_rate": 9.036188573821119e-07,
+      "loss": 0.7843,
+      "step": 10114
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.501936477475615,
+      "learning_rate": 9.028740055244294e-07,
+      "loss": 0.7431,
+      "step": 10115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3989649304626168,
+      "learning_rate": 9.021294303140621e-07,
+      "loss": 0.7133,
+      "step": 10116
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5871713969902226,
+      "learning_rate": 9.013851318012867e-07,
+      "loss": 0.8137,
+      "step": 10117
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6833821359491188,
+      "learning_rate": 9.006411100363577e-07,
+      "loss": 0.6829,
+      "step": 10118
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.533150142420492,
+      "learning_rate": 8.998973650695158e-07,
+      "loss": 0.7653,
+      "step": 10119
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5485299609942147,
+      "learning_rate": 8.991538969509789e-07,
+      "loss": 0.857,
+      "step": 10120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5879626897936094,
+      "learning_rate": 8.984107057309476e-07,
+      "loss": 0.6973,
+      "step": 10121
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5457025980431855,
+      "learning_rate": 8.976677914596049e-07,
+      "loss": 0.7389,
+      "step": 10122
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4821339979436228,
+      "learning_rate": 8.969251541871149e-07,
+      "loss": 0.7781,
+      "step": 10123
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7705381661315329,
+      "learning_rate": 8.961827939636198e-07,
+      "loss": 1.0655,
+      "step": 10124
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4783518080521583,
+      "learning_rate": 8.954407108392449e-07,
+      "loss": 0.725,
+      "step": 10125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6874063092655454,
+      "learning_rate": 8.946989048641003e-07,
+      "loss": 0.6991,
+      "step": 10126
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5464894536714908,
+      "learning_rate": 8.939573760882725e-07,
+      "loss": 0.8277,
+      "step": 10127
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7802660889935711,
+      "learning_rate": 8.932161245618309e-07,
+      "loss": 1.0707,
+      "step": 10128
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5134453042046034,
+      "learning_rate": 8.924751503348283e-07,
+      "loss": 0.7457,
+      "step": 10129
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.560490867114649,
+      "learning_rate": 8.917344534572958e-07,
+      "loss": 0.6965,
+      "step": 10130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5764814063442085,
+      "learning_rate": 8.909940339792461e-07,
+      "loss": 0.7546,
+      "step": 10131
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.515347282147829,
+      "learning_rate": 8.902538919506764e-07,
+      "loss": 0.7616,
+      "step": 10132
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7531315407678318,
+      "learning_rate": 8.895140274215614e-07,
+      "loss": 1.0218,
+      "step": 10133
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5008331805921915,
+      "learning_rate": 8.887744404418585e-07,
+      "loss": 0.6935,
+      "step": 10134
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.765675471255049,
+      "learning_rate": 8.88035131061506e-07,
+      "loss": 1.0759,
+      "step": 10135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7438077656652952,
+      "learning_rate": 8.872960993304242e-07,
+      "loss": 1.0404,
+      "step": 10136
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4635293951977566,
+      "learning_rate": 8.865573452985143e-07,
+      "loss": 0.7372,
+      "step": 10137
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4755754594313066,
+      "learning_rate": 8.858188690156567e-07,
+      "loss": 0.8275,
+      "step": 10138
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5126892016295028,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.7799,
+      "step": 10139
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.417690208879637,
+      "learning_rate": 8.843427498965423e-07,
+      "loss": 0.7238,
+      "step": 10140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5570456402305617,
+      "learning_rate": 8.836051071599544e-07,
+      "loss": 0.824,
+      "step": 10141
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5839204884786524,
+      "learning_rate": 8.828677423717613e-07,
+      "loss": 0.829,
+      "step": 10142
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5575821924331552,
+      "learning_rate": 8.821306555817543e-07,
+      "loss": 0.7346,
+      "step": 10143
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.501007863847518,
+      "learning_rate": 8.813938468397015e-07,
+      "loss": 0.6818,
+      "step": 10144
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5194149106942936,
+      "learning_rate": 8.806573161953536e-07,
+      "loss": 0.7458,
+      "step": 10145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5738341834858505,
+      "learning_rate": 8.799210636984428e-07,
+      "loss": 0.7722,
+      "step": 10146
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.442636467051838,
+      "learning_rate": 8.791850893986836e-07,
+      "loss": 0.7724,
+      "step": 10147
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5618563346852357,
+      "learning_rate": 8.784493933457699e-07,
+      "loss": 0.7366,
+      "step": 10148
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.528717899225521,
+      "learning_rate": 8.77713975589376e-07,
+      "loss": 0.7785,
+      "step": 10149
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5393243543183721,
+      "learning_rate": 8.76978836179162e-07,
+      "loss": 0.7188,
+      "step": 10150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7426861015636986,
+      "learning_rate": 8.76243975164765e-07,
+      "loss": 1.0519,
+      "step": 10151
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7501658562681219,
+      "learning_rate": 8.755093925958031e-07,
+      "loss": 1.0636,
+      "step": 10152
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6020802432548888,
+      "learning_rate": 8.747750885218792e-07,
+      "loss": 0.7777,
+      "step": 10153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5555029530125557,
+      "learning_rate": 8.740410629925744e-07,
+      "loss": 0.748,
+      "step": 10154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3860848196581452,
+      "learning_rate": 8.733073160574517e-07,
+      "loss": 0.7676,
+      "step": 10155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7968976988873908,
+      "learning_rate": 8.725738477660556e-07,
+      "loss": 0.7055,
+      "step": 10156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4347857152493084,
+      "learning_rate": 8.718406581679107e-07,
+      "loss": 0.7229,
+      "step": 10157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5358969048263418,
+      "learning_rate": 8.71107747312524e-07,
+      "loss": 0.7009,
+      "step": 10158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.547901260272034,
+      "learning_rate": 8.703751152493828e-07,
+      "loss": 0.7693,
+      "step": 10159
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7496180059543212,
+      "learning_rate": 8.696427620279579e-07,
+      "loss": 1.0605,
+      "step": 10160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5953470538975045,
+      "learning_rate": 8.689106876976983e-07,
+      "loss": 0.8224,
+      "step": 10161
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7774204025243533,
+      "learning_rate": 8.681788923080337e-07,
+      "loss": 1.0582,
+      "step": 10162
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4354758058779573,
+      "learning_rate": 8.674473759083801e-07,
+      "loss": 0.7371,
+      "step": 10163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.52074137938149,
+      "learning_rate": 8.667161385481288e-07,
+      "loss": 0.7244,
+      "step": 10164
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5488173634886915,
+      "learning_rate": 8.659851802766555e-07,
+      "loss": 0.6389,
+      "step": 10165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.513652808658308,
+      "learning_rate": 8.652545011433144e-07,
+      "loss": 0.7513,
+      "step": 10166
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.741752828838893,
+      "learning_rate": 8.64524101197447e-07,
+      "loss": 0.7322,
+      "step": 10167
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4414207737678317,
+      "learning_rate": 8.637939804883672e-07,
+      "loss": 0.7198,
+      "step": 10168
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.459053725213051,
+      "learning_rate": 8.630641390653743e-07,
+      "loss": 0.763,
+      "step": 10169
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.76739420214067,
+      "learning_rate": 8.623345769777514e-07,
+      "loss": 1.0642,
+      "step": 10170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5401220092356134,
+      "learning_rate": 8.616052942747599e-07,
+      "loss": 0.7616,
+      "step": 10171
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4591732511524664,
+      "learning_rate": 8.6087629100564e-07,
+      "loss": 0.735,
+      "step": 10172
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5578403503390603,
+      "learning_rate": 8.601475672196197e-07,
+      "loss": 0.7916,
+      "step": 10173
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7445508591833385,
+      "learning_rate": 8.594191229659016e-07,
+      "loss": 1.0414,
+      "step": 10174
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4300110117459108,
+      "learning_rate": 8.58690958293672e-07,
+      "loss": 0.7375,
+      "step": 10175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4912285846996498,
+      "learning_rate": 8.579630732520977e-07,
+      "loss": 0.8451,
+      "step": 10176
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4969331914012287,
+      "learning_rate": 8.572354678903289e-07,
+      "loss": 0.8301,
+      "step": 10177
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5342405687355272,
+      "learning_rate": 8.56508142257495e-07,
+      "loss": 0.7128,
+      "step": 10178
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6278346228126688,
+      "learning_rate": 8.557810964027053e-07,
+      "loss": 0.7043,
+      "step": 10179
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4520925146516004,
+      "learning_rate": 8.550543303750524e-07,
+      "loss": 0.7107,
+      "step": 10180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8977213897993432,
+      "learning_rate": 8.543278442236097e-07,
+      "loss": 0.7459,
+      "step": 10181
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4442924210143684,
+      "learning_rate": 8.536016379974299e-07,
+      "loss": 0.6952,
+      "step": 10182
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4524355796860815,
+      "learning_rate": 8.528757117455477e-07,
+      "loss": 0.6414,
+      "step": 10183
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7514701978352157,
+      "learning_rate": 8.521500655169823e-07,
+      "loss": 1.0712,
+      "step": 10184
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6348993875766722,
+      "learning_rate": 8.514246993607284e-07,
+      "loss": 0.8038,
+      "step": 10185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5118774201926732,
+      "learning_rate": 8.506996133257639e-07,
+      "loss": 0.7185,
+      "step": 10186
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4951837789466995,
+      "learning_rate": 8.499748074610508e-07,
+      "loss": 0.7413,
+      "step": 10187
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5850954121658314,
+      "learning_rate": 8.492502818155285e-07,
+      "loss": 0.7322,
+      "step": 10188
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.451877513298632,
+      "learning_rate": 8.485260364381187e-07,
+      "loss": 0.7564,
+      "step": 10189
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7565086499529938,
+      "learning_rate": 8.478020713777235e-07,
+      "loss": 1.0777,
+      "step": 10190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.704899018521599,
+      "learning_rate": 8.470783866832266e-07,
+      "loss": 0.737,
+      "step": 10191
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.542126325351262,
+      "learning_rate": 8.463549824034939e-07,
+      "loss": 0.7124,
+      "step": 10192
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5974615369250018,
+      "learning_rate": 8.456318585873691e-07,
+      "loss": 0.7208,
+      "step": 10193
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5187077342593653,
+      "learning_rate": 8.44909015283682e-07,
+      "loss": 0.6571,
+      "step": 10194
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5736099974968196,
+      "learning_rate": 8.441864525412396e-07,
+      "loss": 0.8368,
+      "step": 10195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.63263632469122,
+      "learning_rate": 8.434641704088292e-07,
+      "loss": 0.7494,
+      "step": 10196
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4477721140924287,
+      "learning_rate": 8.427421689352239e-07,
+      "loss": 0.6645,
+      "step": 10197
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.666066352949737,
+      "learning_rate": 8.420204481691734e-07,
+      "loss": 0.7191,
+      "step": 10198
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4825500970841934,
+      "learning_rate": 8.412990081594102e-07,
+      "loss": 0.7463,
+      "step": 10199
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.614495739773159,
+      "learning_rate": 8.405778489546474e-07,
+      "loss": 0.7442,
+      "step": 10200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4918497704445945,
+      "learning_rate": 8.398569706035791e-07,
+      "loss": 0.724,
+      "step": 10201
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4795152591543022,
+      "learning_rate": 8.391363731548813e-07,
+      "loss": 0.783,
+      "step": 10202
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6500498020394825,
+      "learning_rate": 8.384160566572086e-07,
+      "loss": 0.7941,
+      "step": 10203
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4742088121139285,
+      "learning_rate": 8.376960211592011e-07,
+      "loss": 0.6915,
+      "step": 10204
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4991766951381171,
+      "learning_rate": 8.369762667094755e-07,
+      "loss": 0.6646,
+      "step": 10205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.436987827435293,
+      "learning_rate": 8.362567933566318e-07,
+      "loss": 0.777,
+      "step": 10206
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4368734587817575,
+      "learning_rate": 8.355376011492494e-07,
+      "loss": 0.7457,
+      "step": 10207
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.640520049255847,
+      "learning_rate": 8.348186901358923e-07,
+      "loss": 0.6711,
+      "step": 10208
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7525985723075654,
+      "learning_rate": 8.341000603651012e-07,
+      "loss": 1.0713,
+      "step": 10209
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5059165286419782,
+      "learning_rate": 8.333817118853982e-07,
+      "loss": 0.7328,
+      "step": 10210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5650353981886584,
+      "learning_rate": 8.326636447452929e-07,
+      "loss": 0.765,
+      "step": 10211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7585011484426548,
+      "learning_rate": 8.319458589932655e-07,
+      "loss": 1.0514,
+      "step": 10212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6445875837214472,
+      "learning_rate": 8.312283546777838e-07,
+      "loss": 0.7847,
+      "step": 10213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3982329666333004,
+      "learning_rate": 8.30511131847297e-07,
+      "loss": 0.6997,
+      "step": 10214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6035742326762072,
+      "learning_rate": 8.297941905502327e-07,
+      "loss": 0.7825,
+      "step": 10215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5599981503941596,
+      "learning_rate": 8.290775308350008e-07,
+      "loss": 0.7617,
+      "step": 10216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5520083592957572,
+      "learning_rate": 8.283611527499896e-07,
+      "loss": 0.7391,
+      "step": 10217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5544708827922447,
+      "learning_rate": 8.276450563435739e-07,
+      "loss": 0.7258,
+      "step": 10218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4727551778047538,
+      "learning_rate": 8.26929241664105e-07,
+      "loss": 0.7793,
+      "step": 10219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4974960701712126,
+      "learning_rate": 8.262137087599142e-07,
+      "loss": 0.7592,
+      "step": 10220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4026477198583949,
+      "learning_rate": 8.254984576793196e-07,
+      "loss": 0.6974,
+      "step": 10221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4068663818250469,
+      "learning_rate": 8.247834884706141e-07,
+      "loss": 0.7791,
+      "step": 10222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4759468815370436,
+      "learning_rate": 8.240688011820752e-07,
+      "loss": 0.8847,
+      "step": 10223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.511019834245238,
+      "learning_rate": 8.233543958619594e-07,
+      "loss": 0.811,
+      "step": 10224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3995781304111894,
+      "learning_rate": 8.226402725585053e-07,
+      "loss": 0.7457,
+      "step": 10225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7736057707822106,
+      "learning_rate": 8.219264313199322e-07,
+      "loss": 1.0558,
+      "step": 10226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7814040401984077,
+      "learning_rate": 8.212128721944385e-07,
+      "loss": 1.0732,
+      "step": 10227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5101283017678795,
+      "learning_rate": 8.204995952302087e-07,
+      "loss": 0.7714,
+      "step": 10228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4683763197373363,
+      "learning_rate": 8.197866004754029e-07,
+      "loss": 0.7713,
+      "step": 10229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5056232147940996,
+      "learning_rate": 8.190738879781634e-07,
+      "loss": 0.6931,
+      "step": 10230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5147970630088945,
+      "learning_rate": 8.183614577866166e-07,
+      "loss": 0.8008,
+      "step": 10231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5289036058634384,
+      "learning_rate": 8.176493099488664e-07,
+      "loss": 0.778,
+      "step": 10232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6294036582417435,
+      "learning_rate": 8.169374445129979e-07,
+      "loss": 0.7987,
+      "step": 10233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5520114729667593,
+      "learning_rate": 8.162258615270779e-07,
+      "loss": 0.7522,
+      "step": 10234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5595654595584059,
+      "learning_rate": 8.155145610391552e-07,
+      "loss": 0.7779,
+      "step": 10235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5477778735512766,
+      "learning_rate": 8.148035430972573e-07,
+      "loss": 0.7606,
+      "step": 10236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5270388064118565,
+      "learning_rate": 8.140928077493937e-07,
+      "loss": 0.758,
+      "step": 10237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.64085322706877,
+      "learning_rate": 8.13382355043556e-07,
+      "loss": 0.8114,
+      "step": 10238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6146529992285388,
+      "learning_rate": 8.126721850277147e-07,
+      "loss": 0.8136,
+      "step": 10239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4926189446044338,
+      "learning_rate": 8.119622977498226e-07,
+      "loss": 0.775,
+      "step": 10240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.754038067042479,
+      "learning_rate": 8.112526932578118e-07,
+      "loss": 1.0748,
+      "step": 10241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5001416245813683,
+      "learning_rate": 8.105433715995981e-07,
+      "loss": 0.7364,
+      "step": 10242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.46228679948755,
+      "learning_rate": 8.098343328230762e-07,
+      "loss": 0.7734,
+      "step": 10243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5484875276888,
+      "learning_rate": 8.091255769761213e-07,
+      "loss": 0.8016,
+      "step": 10244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.442836367190372,
+      "learning_rate": 8.084171041065903e-07,
+      "loss": 0.7235,
+      "step": 10245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5071222364428718,
+      "learning_rate": 8.077089142623212e-07,
+      "loss": 0.7857,
+      "step": 10246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5942964847271635,
+      "learning_rate": 8.070010074911322e-07,
+      "loss": 0.8741,
+      "step": 10247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.633149424258743,
+      "learning_rate": 8.062933838408221e-07,
+      "loss": 0.7937,
+      "step": 10248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.776484005388792,
+      "learning_rate": 8.055860433591734e-07,
+      "loss": 1.0945,
+      "step": 10249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7546064764548901,
+      "learning_rate": 8.04878986093946e-07,
+      "loss": 1.062,
+      "step": 10250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4943057687077619,
+      "learning_rate": 8.041722120928814e-07,
+      "loss": 0.7331,
+      "step": 10251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4591992800753493,
+      "learning_rate": 8.034657214037044e-07,
+      "loss": 0.7466,
+      "step": 10252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4736758923055433,
+      "learning_rate": 8.027595140741179e-07,
+      "loss": 0.6754,
+      "step": 10253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7771733837920801,
+      "learning_rate": 8.02053590151805e-07,
+      "loss": 1.0858,
+      "step": 10254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.536919960923253,
+      "learning_rate": 8.013479496844356e-07,
+      "loss": 0.7553,
+      "step": 10255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.536941881898944,
+      "learning_rate": 8.00642592719652e-07,
+      "loss": 0.7422,
+      "step": 10256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4362632627505598,
+      "learning_rate": 7.999375193050828e-07,
+      "loss": 0.786,
+      "step": 10257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6465829958170026,
+      "learning_rate": 7.992327294883356e-07,
+      "loss": 0.7031,
+      "step": 10258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5027636303804737,
+      "learning_rate": 7.985282233170011e-07,
+      "loss": 0.7004,
+      "step": 10259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5610655343788522,
+      "learning_rate": 7.978240008386484e-07,
+      "loss": 0.7726,
+      "step": 10260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.551819591667791,
+      "learning_rate": 7.971200621008268e-07,
+      "loss": 0.7187,
+      "step": 10261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5493630133110987,
+      "learning_rate": 7.964164071510699e-07,
+      "loss": 0.7193,
+      "step": 10262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5785564375583097,
+      "learning_rate": 7.957130360368898e-07,
+      "loss": 0.8462,
+      "step": 10263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4630168601358768,
+      "learning_rate": 7.950099488057788e-07,
+      "loss": 0.7374,
+      "step": 10264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6070547531547,
+      "learning_rate": 7.943071455052104e-07,
+      "loss": 0.775,
+      "step": 10265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5071354970366702,
+      "learning_rate": 7.936046261826413e-07,
+      "loss": 0.7089,
+      "step": 10266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.584734252628964,
+      "learning_rate": 7.929023908855066e-07,
+      "loss": 0.7355,
+      "step": 10267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5980278197505224,
+      "learning_rate": 7.922004396612226e-07,
+      "loss": 0.741,
+      "step": 10268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5348900988624177,
+      "learning_rate": 7.914987725571866e-07,
+      "loss": 0.7261,
+      "step": 10269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7323165694928259,
+      "learning_rate": 7.907973896207765e-07,
+      "loss": 1.071,
+      "step": 10270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6220279908319377,
+      "learning_rate": 7.900962908993509e-07,
+      "loss": 0.7909,
+      "step": 10271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7649054450814116,
+      "learning_rate": 7.893954764402512e-07,
+      "loss": 1.0643,
+      "step": 10272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5417060639215268,
+      "learning_rate": 7.886949462907967e-07,
+      "loss": 0.7242,
+      "step": 10273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.612312457868986,
+      "learning_rate": 7.879947004982896e-07,
+      "loss": 0.7931,
+      "step": 10274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.597137712567434,
+      "learning_rate": 7.872947391100106e-07,
+      "loss": 0.7609,
+      "step": 10275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6347883306074404,
+      "learning_rate": 7.865950621732244e-07,
+      "loss": 0.7638,
+      "step": 10276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7470409383064237,
+      "learning_rate": 7.858956697351744e-07,
+      "loss": 1.0279,
+      "step": 10277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5180657775056046,
+      "learning_rate": 7.851965618430852e-07,
+      "loss": 0.6709,
+      "step": 10278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4825635691066743,
+      "learning_rate": 7.844977385441615e-07,
+      "loss": 0.718,
+      "step": 10279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.496532067294344,
+      "learning_rate": 7.837991998855899e-07,
+      "loss": 0.7568,
+      "step": 10280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.457051423912158,
+      "learning_rate": 7.831009459145372e-07,
+      "loss": 0.7054,
+      "step": 10281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4589054575110274,
+      "learning_rate": 7.824029766781499e-07,
+      "loss": 0.7396,
+      "step": 10282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7698748056826352,
+      "learning_rate": 7.817052922235591e-07,
+      "loss": 1.0756,
+      "step": 10283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5313419975978542,
+      "learning_rate": 7.810078925978731e-07,
+      "loss": 0.7039,
+      "step": 10284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6188830003514372,
+      "learning_rate": 7.803107778481794e-07,
+      "loss": 0.7551,
+      "step": 10285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.419430971444412,
+      "learning_rate": 7.796139480215525e-07,
+      "loss": 0.8577,
+      "step": 10286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.46528383426777,
+      "learning_rate": 7.789174031650426e-07,
+      "loss": 0.7363,
+      "step": 10287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6267693388492481,
+      "learning_rate": 7.782211433256815e-07,
+      "loss": 0.7214,
+      "step": 10288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4683304813193683,
+      "learning_rate": 7.775251685504826e-07,
+      "loss": 0.732,
+      "step": 10289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4328687078390012,
+      "learning_rate": 7.768294788864395e-07,
+      "loss": 0.7141,
+      "step": 10290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6604123596625227,
+      "learning_rate": 7.761340743805268e-07,
+      "loss": 0.7556,
+      "step": 10291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5296655377143857,
+      "learning_rate": 7.754389550796987e-07,
+      "loss": 0.7147,
+      "step": 10292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.483671714936275,
+      "learning_rate": 7.747441210308937e-07,
+      "loss": 0.7295,
+      "step": 10293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.54895338398349,
+      "learning_rate": 7.740495722810271e-07,
+      "loss": 0.7751,
+      "step": 10294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5208849836493767,
+      "learning_rate": 7.733553088769952e-07,
+      "loss": 0.7564,
+      "step": 10295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7669142752324981,
+      "learning_rate": 7.726613308656788e-07,
+      "loss": 1.0532,
+      "step": 10296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5715339787896745,
+      "learning_rate": 7.719676382939362e-07,
+      "loss": 0.8116,
+      "step": 10297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4740648374303387,
+      "learning_rate": 7.712742312086064e-07,
+      "loss": 0.721,
+      "step": 10298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4880516738219,
+      "learning_rate": 7.705811096565102e-07,
+      "loss": 0.7388,
+      "step": 10299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.501940404441746,
+      "learning_rate": 7.698882736844487e-07,
+      "loss": 0.8,
+      "step": 10300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.621992149723792,
+      "learning_rate": 7.691957233392034e-07,
+      "loss": 0.7408,
+      "step": 10301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5358297130641876,
+      "learning_rate": 7.685034586675361e-07,
+      "loss": 0.7193,
+      "step": 10302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7498570102289274,
+      "learning_rate": 7.678114797161928e-07,
+      "loss": 1.0231,
+      "step": 10303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4608801048723459,
+      "learning_rate": 7.671197865318952e-07,
+      "loss": 0.7639,
+      "step": 10304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5848901877746584,
+      "learning_rate": 7.664283791613492e-07,
+      "loss": 0.7075,
+      "step": 10305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6534338065588428,
+      "learning_rate": 7.657372576512384e-07,
+      "loss": 0.7785,
+      "step": 10306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5534680216176417,
+      "learning_rate": 7.650464220482312e-07,
+      "loss": 0.7592,
+      "step": 10307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7922126301473036,
+      "learning_rate": 7.64355872398973e-07,
+      "loss": 1.0475,
+      "step": 10308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5542972862980537,
+      "learning_rate": 7.63665608750091e-07,
+      "loss": 0.8239,
+      "step": 10309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7573813760714273,
+      "learning_rate": 7.62975631148195e-07,
+      "loss": 1.015,
+      "step": 10310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7847927430900885,
+      "learning_rate": 7.622859396398735e-07,
+      "loss": 1.0643,
+      "step": 10311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4369227988905473,
+      "learning_rate": 7.615965342716952e-07,
+      "loss": 0.7523,
+      "step": 10312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5937328612498105,
+      "learning_rate": 7.609074150902102e-07,
+      "loss": 0.7028,
+      "step": 10313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7656282234931853,
+      "learning_rate": 7.6021858214195e-07,
+      "loss": 1.0525,
+      "step": 10314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5527211629838769,
+      "learning_rate": 7.595300354734264e-07,
+      "loss": 0.7223,
+      "step": 10315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4934760246590477,
+      "learning_rate": 7.588417751311295e-07,
+      "loss": 0.7697,
+      "step": 10316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6212137554661288,
+      "learning_rate": 7.581538011615352e-07,
+      "loss": 0.7798,
+      "step": 10317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5285218912753724,
+      "learning_rate": 7.574661136110961e-07,
+      "loss": 0.7483,
+      "step": 10318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7923824465177316,
+      "learning_rate": 7.567787125262449e-07,
+      "loss": 1.0432,
+      "step": 10319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4920213332260217,
+      "learning_rate": 7.56091597953399e-07,
+      "loss": 0.7863,
+      "step": 10320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4835500345882757,
+      "learning_rate": 7.554047699389522e-07,
+      "loss": 0.8009,
+      "step": 10321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5025873877671012,
+      "learning_rate": 7.547182285292815e-07,
+      "loss": 0.6965,
+      "step": 10322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.595733219146192,
+      "learning_rate": 7.540319737707436e-07,
+      "loss": 0.7401,
+      "step": 10323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5105110648519158,
+      "learning_rate": 7.533460057096753e-07,
+      "loss": 0.7967,
+      "step": 10324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.766914754321143,
+      "learning_rate": 7.526603243923958e-07,
+      "loss": 1.0592,
+      "step": 10325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6215265805439474,
+      "learning_rate": 7.519749298652018e-07,
+      "loss": 0.7621,
+      "step": 10326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5330426107122235,
+      "learning_rate": 7.512898221743759e-07,
+      "loss": 0.7667,
+      "step": 10327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5690573473778342,
+      "learning_rate": 7.506050013661758e-07,
+      "loss": 0.8169,
+      "step": 10328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6036763239530325,
+      "learning_rate": 7.499204674868421e-07,
+      "loss": 0.7498,
+      "step": 10329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5928210510040004,
+      "learning_rate": 7.492362205825981e-07,
+      "loss": 0.7915,
+      "step": 10330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7204956188703666,
+      "learning_rate": 7.485522606996443e-07,
+      "loss": 0.7759,
+      "step": 10331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.450371505809567,
+      "learning_rate": 7.478685878841629e-07,
+      "loss": 0.7724,
+      "step": 10332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4309713569626306,
+      "learning_rate": 7.471852021823184e-07,
+      "loss": 0.7251,
+      "step": 10333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4269180575021303,
+      "learning_rate": 7.465021036402531e-07,
+      "loss": 0.745,
+      "step": 10334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6290975203159619,
+      "learning_rate": 7.458192923040919e-07,
+      "loss": 0.7194,
+      "step": 10335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4166478068792738,
+      "learning_rate": 7.451367682199389e-07,
+      "loss": 0.771,
+      "step": 10336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7440258357671208,
+      "learning_rate": 7.444545314338819e-07,
+      "loss": 1.0581,
+      "step": 10337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.537765593305599,
+      "learning_rate": 7.437725819919861e-07,
+      "loss": 0.7029,
+      "step": 10338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4606459179133346,
+      "learning_rate": 7.430909199402974e-07,
+      "loss": 0.711,
+      "step": 10339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.444786337492407,
+      "learning_rate": 7.424095453248431e-07,
+      "loss": 0.7545,
+      "step": 10340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7287186007689946,
+      "learning_rate": 7.417284581916329e-07,
+      "loss": 1.0749,
+      "step": 10341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5226347725380003,
+      "learning_rate": 7.410476585866538e-07,
+      "loss": 0.7262,
+      "step": 10342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7414041986488041,
+      "learning_rate": 7.403671465558765e-07,
+      "loss": 1.0471,
+      "step": 10343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7410818256361182,
+      "learning_rate": 7.396869221452491e-07,
+      "loss": 1.0698,
+      "step": 10344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5374568775795046,
+      "learning_rate": 7.390069854007026e-07,
+      "loss": 0.7523,
+      "step": 10345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5235908515650571,
+      "learning_rate": 7.383273363681476e-07,
+      "loss": 0.7621,
+      "step": 10346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.767277451350047,
+      "learning_rate": 7.376479750934745e-07,
+      "loss": 1.0411,
+      "step": 10347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5172849884533024,
+      "learning_rate": 7.369689016225578e-07,
+      "loss": 0.8048,
+      "step": 10348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5217305451830077,
+      "learning_rate": 7.362901160012492e-07,
+      "loss": 0.7124,
+      "step": 10349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.671789482090962,
+      "learning_rate": 7.356116182753803e-07,
+      "loss": 0.7907,
+      "step": 10350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6081906839393203,
+      "learning_rate": 7.349334084907672e-07,
+      "loss": 0.7447,
+      "step": 10351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.756013551609216,
+      "learning_rate": 7.342554866932028e-07,
+      "loss": 1.0433,
+      "step": 10352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5682380616308471,
+      "learning_rate": 7.335778529284615e-07,
+      "loss": 0.7626,
+      "step": 10353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4565073083883056,
+      "learning_rate": 7.329005072423001e-07,
+      "loss": 0.7094,
+      "step": 10354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.543418417449908,
+      "learning_rate": 7.322234496804536e-07,
+      "loss": 0.7607,
+      "step": 10355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4818711511818212,
+      "learning_rate": 7.315466802886401e-07,
+      "loss": 0.7397,
+      "step": 10356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.55167643444967,
+      "learning_rate": 7.308701991125527e-07,
+      "loss": 0.7767,
+      "step": 10357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8701745948581088,
+      "learning_rate": 7.301940061978724e-07,
+      "loss": 0.7821,
+      "step": 10358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4877157008475492,
+      "learning_rate": 7.295181015902569e-07,
+      "loss": 0.7717,
+      "step": 10359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.492867331982862,
+      "learning_rate": 7.288424853353426e-07,
+      "loss": 0.7301,
+      "step": 10360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5286979661644513,
+      "learning_rate": 7.281671574787513e-07,
+      "loss": 0.7952,
+      "step": 10361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4723094658909126,
+      "learning_rate": 7.274921180660821e-07,
+      "loss": 0.8476,
+      "step": 10362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4939130870980386,
+      "learning_rate": 7.268173671429147e-07,
+      "loss": 0.7871,
+      "step": 10363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.556158302633824,
+      "learning_rate": 7.261429047548085e-07,
+      "loss": 0.7343,
+      "step": 10364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3939570313579885,
+      "learning_rate": 7.254687309473074e-07,
+      "loss": 0.7868,
+      "step": 10365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6452253118852742,
+      "learning_rate": 7.247948457659315e-07,
+      "loss": 0.7938,
+      "step": 10366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6410398683002072,
+      "learning_rate": 7.241212492561839e-07,
+      "loss": 0.7998,
+      "step": 10367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5047545742220603,
+      "learning_rate": 7.23447941463547e-07,
+      "loss": 0.8085,
+      "step": 10368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3598967667450848,
+      "learning_rate": 7.22774922433484e-07,
+      "loss": 0.7128,
+      "step": 10369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4368467622978538,
+      "learning_rate": 7.221021922114374e-07,
+      "loss": 0.7689,
+      "step": 10370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7321282115736699,
+      "learning_rate": 7.214297508428336e-07,
+      "loss": 1.0398,
+      "step": 10371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5961704318913943,
+      "learning_rate": 7.207575983730774e-07,
+      "loss": 0.7505,
+      "step": 10372
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5874872168562826,
+      "learning_rate": 7.200857348475526e-07,
+      "loss": 0.8058,
+      "step": 10373
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5624671074458882,
+      "learning_rate": 7.194141603116244e-07,
+      "loss": 0.792,
+      "step": 10374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4817360058821694,
+      "learning_rate": 7.187428748106418e-07,
+      "loss": 0.7506,
+      "step": 10375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5225122412042733,
+      "learning_rate": 7.180718783899298e-07,
+      "loss": 0.7696,
+      "step": 10376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4556160859518892,
+      "learning_rate": 7.174011710947959e-07,
+      "loss": 0.77,
+      "step": 10377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4754991747118285,
+      "learning_rate": 7.167307529705275e-07,
+      "loss": 0.737,
+      "step": 10378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6087676355742386,
+      "learning_rate": 7.16060624062393e-07,
+      "loss": 0.7624,
+      "step": 10379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4733683845101624,
+      "learning_rate": 7.153907844156411e-07,
+      "loss": 0.7222,
+      "step": 10380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5751049190670687,
+      "learning_rate": 7.147212340754994e-07,
+      "loss": 0.784,
+      "step": 10381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5822974695301222,
+      "learning_rate": 7.140519730871804e-07,
+      "loss": 0.803,
+      "step": 10382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5736565683593153,
+      "learning_rate": 7.13383001495872e-07,
+      "loss": 0.7541,
+      "step": 10383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4972999396497033,
+      "learning_rate": 7.127143193467445e-07,
+      "loss": 0.6669,
+      "step": 10384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5076944113034005,
+      "learning_rate": 7.120459266849511e-07,
+      "loss": 0.713,
+      "step": 10385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4111353403347884,
+      "learning_rate": 7.113778235556212e-07,
+      "loss": 0.7454,
+      "step": 10386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4975962764849649,
+      "learning_rate": 7.107100100038672e-07,
+      "loss": 0.7375,
+      "step": 10387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.431892913448652,
+      "learning_rate": 7.100424860747817e-07,
+      "loss": 0.7112,
+      "step": 10388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.552786234205501,
+      "learning_rate": 7.093752518134367e-07,
+      "loss": 0.8015,
+      "step": 10389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4809457687093304,
+      "learning_rate": 7.087083072648865e-07,
+      "loss": 0.7583,
+      "step": 10390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.422177878433341,
+      "learning_rate": 7.080416524741623e-07,
+      "loss": 0.7519,
+      "step": 10391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.352242991668909,
+      "learning_rate": 7.07375287486281e-07,
+      "loss": 0.6241,
+      "step": 10392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.609716033823568,
+      "learning_rate": 7.067092123462361e-07,
+      "loss": 0.7603,
+      "step": 10393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4493573413328664,
+      "learning_rate": 7.060434270990013e-07,
+      "loss": 0.7849,
+      "step": 10394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5505538315015905,
+      "learning_rate": 7.053779317895343e-07,
+      "loss": 0.6872,
+      "step": 10395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4844584330618307,
+      "learning_rate": 7.047127264627696e-07,
+      "loss": 0.7813,
+      "step": 10396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6785934385288757,
+      "learning_rate": 7.040478111636229e-07,
+      "loss": 0.7712,
+      "step": 10397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.551043294529431,
+      "learning_rate": 7.033831859369905e-07,
+      "loss": 0.7561,
+      "step": 10398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6275683778412977,
+      "learning_rate": 7.027188508277516e-07,
+      "loss": 0.7181,
+      "step": 10399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6460406975846045,
+      "learning_rate": 7.02054805880763e-07,
+      "loss": 0.7061,
+      "step": 10400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.643622283812845,
+      "learning_rate": 7.013910511408595e-07,
+      "loss": 0.7441,
+      "step": 10401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.006361350256462,
+      "learning_rate": 7.007275866528623e-07,
+      "loss": 0.7239,
+      "step": 10402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5755313017684354,
+      "learning_rate": 7.000644124615702e-07,
+      "loss": 0.7964,
+      "step": 10403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7612019366299966,
+      "learning_rate": 6.994015286117606e-07,
+      "loss": 1.0402,
+      "step": 10404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6293684743276031,
+      "learning_rate": 6.987389351481933e-07,
+      "loss": 0.8086,
+      "step": 10405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.40190700398306,
+      "learning_rate": 6.980766321156091e-07,
+      "loss": 0.7017,
+      "step": 10406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7745999216855157,
+      "learning_rate": 6.974146195587278e-07,
+      "loss": 1.0444,
+      "step": 10407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.558690876148342,
+      "learning_rate": 6.967528975222487e-07,
+      "loss": 0.7021,
+      "step": 10408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.484925340531523,
+      "learning_rate": 6.96091466050855e-07,
+      "loss": 0.7164,
+      "step": 10409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.626538985161877,
+      "learning_rate": 6.95430325189207e-07,
+      "loss": 0.7675,
+      "step": 10410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7345292324539173,
+      "learning_rate": 6.947694749819467e-07,
+      "loss": 1.0465,
+      "step": 10411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4969143328667123,
+      "learning_rate": 6.941089154736958e-07,
+      "loss": 0.7919,
+      "step": 10412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4362062591228537,
+      "learning_rate": 6.934486467090568e-07,
+      "loss": 0.6731,
+      "step": 10413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4731033817943517,
+      "learning_rate": 6.927886687326129e-07,
+      "loss": 0.8045,
+      "step": 10414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5893601955211494,
+      "learning_rate": 6.921289815889259e-07,
+      "loss": 0.7433,
+      "step": 10415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4235624541398297,
+      "learning_rate": 6.914695853225417e-07,
+      "loss": 0.6569,
+      "step": 10416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5348053873130933,
+      "learning_rate": 6.90810479977983e-07,
+      "loss": 0.7742,
+      "step": 10417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.743148788674857,
+      "learning_rate": 6.901516655997536e-07,
+      "loss": 1.0821,
+      "step": 10418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5451662833929896,
+      "learning_rate": 6.894931422323398e-07,
+      "loss": 0.7192,
+      "step": 10419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4718227314500927,
+      "learning_rate": 6.888349099202051e-07,
+      "loss": 0.7832,
+      "step": 10420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.746766551768053,
+      "learning_rate": 6.881769687077955e-07,
+      "loss": 1.0748,
+      "step": 10421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.547023583495951,
+      "learning_rate": 6.875193186395368e-07,
+      "loss": 0.7808,
+      "step": 10422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.799574900743195,
+      "learning_rate": 6.868619597598347e-07,
+      "loss": 0.7452,
+      "step": 10423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4666801513776357,
+      "learning_rate": 6.86204892113076e-07,
+      "loss": 0.8121,
+      "step": 10424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4096099335048984,
+      "learning_rate": 6.855481157436256e-07,
+      "loss": 0.6789,
+      "step": 10425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7531968088940726,
+      "learning_rate": 6.84891630695833e-07,
+      "loss": 1.0379,
+      "step": 10426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5299006193082145,
+      "learning_rate": 6.842354370140247e-07,
+      "loss": 0.7406,
+      "step": 10427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4416928193368106,
+      "learning_rate": 6.835795347425073e-07,
+      "loss": 0.7035,
+      "step": 10428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4436125806461273,
+      "learning_rate": 6.829239239255708e-07,
+      "loss": 0.7448,
+      "step": 10429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5249002123904924,
+      "learning_rate": 6.822686046074828e-07,
+      "loss": 0.805,
+      "step": 10430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5790945422586637,
+      "learning_rate": 6.816135768324916e-07,
+      "loss": 0.7873,
+      "step": 10431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4538199438046915,
+      "learning_rate": 6.809588406448264e-07,
+      "loss": 0.7548,
+      "step": 10432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4990017582592883,
+      "learning_rate": 6.803043960886957e-07,
+      "loss": 0.8747,
+      "step": 10433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.474141685193094,
+      "learning_rate": 6.796502432082902e-07,
+      "loss": 0.799,
+      "step": 10434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4861493096194343,
+      "learning_rate": 6.789963820477785e-07,
+      "loss": 0.7811,
+      "step": 10435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4687373557534227,
+      "learning_rate": 6.783428126513125e-07,
+      "loss": 0.6686,
+      "step": 10436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7486296172418683,
+      "learning_rate": 6.776895350630219e-07,
+      "loss": 1.0375,
+      "step": 10437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7469736403092688,
+      "learning_rate": 6.770365493270176e-07,
+      "loss": 1.0219,
+      "step": 10438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7608148781235129,
+      "learning_rate": 6.763838554873892e-07,
+      "loss": 1.0241,
+      "step": 10439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4099427752002593,
+      "learning_rate": 6.757314535882104e-07,
+      "loss": 0.773,
+      "step": 10440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4737713668865997,
+      "learning_rate": 6.75079343673532e-07,
+      "loss": 0.7092,
+      "step": 10441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4966978458705134,
+      "learning_rate": 6.74427525787385e-07,
+      "loss": 0.7206,
+      "step": 10442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.559992155865875,
+      "learning_rate": 6.737759999737836e-07,
+      "loss": 0.7553,
+      "step": 10443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.531756538675424,
+      "learning_rate": 6.731247662767199e-07,
+      "loss": 0.7447,
+      "step": 10444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.888783732287368,
+      "learning_rate": 6.724738247401652e-07,
+      "loss": 0.7465,
+      "step": 10445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4584868350945932,
+      "learning_rate": 6.718231754080723e-07,
+      "loss": 0.7999,
+      "step": 10446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7354782209157104,
+      "learning_rate": 6.711728183243766e-07,
+      "loss": 1.0423,
+      "step": 10447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.453862820522831,
+      "learning_rate": 6.7052275353299e-07,
+      "loss": 0.7663,
+      "step": 10448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4542831935814535,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.7195,
+      "step": 10449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4129335768813776,
+      "learning_rate": 6.69223501002702e-07,
+      "loss": 0.7789,
+      "step": 10450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.789275292532673,
+      "learning_rate": 6.685743133515293e-07,
+      "loss": 0.8127,
+      "step": 10451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4125882547420212,
+      "learning_rate": 6.679254181681228e-07,
+      "loss": 0.6727,
+      "step": 10452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7032565382561737,
+      "learning_rate": 6.672768154962983e-07,
+      "loss": 0.7173,
+      "step": 10453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4668332389472905,
+      "learning_rate": 6.66628505379851e-07,
+      "loss": 0.7233,
+      "step": 10454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.53065052875446,
+      "learning_rate": 6.659804878625559e-07,
+      "loss": 0.7841,
+      "step": 10455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7849040164090195,
+      "learning_rate": 6.653327629881689e-07,
+      "loss": 1.0516,
+      "step": 10456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4882027396296327,
+      "learning_rate": 6.646853308004253e-07,
+      "loss": 0.6437,
+      "step": 10457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6882672720891754,
+      "learning_rate": 6.64038191343041e-07,
+      "loss": 0.8217,
+      "step": 10458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4216467836302558,
+      "learning_rate": 6.633913446597124e-07,
+      "loss": 0.7152,
+      "step": 10459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4392467760459209,
+      "learning_rate": 6.62744790794117e-07,
+      "loss": 0.7012,
+      "step": 10460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5119470045627927,
+      "learning_rate": 6.620985297899113e-07,
+      "loss": 0.6963,
+      "step": 10461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4712938680726466,
+      "learning_rate": 6.614525616907319e-07,
+      "loss": 0.7257,
+      "step": 10462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6267643994987406,
+      "learning_rate": 6.608068865401957e-07,
+      "loss": 0.7624,
+      "step": 10463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4800174923931173,
+      "learning_rate": 6.60161504381901e-07,
+      "loss": 0.7676,
+      "step": 10464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4798795178230548,
+      "learning_rate": 6.595164152594258e-07,
+      "loss": 0.7476,
+      "step": 10465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6827016802595964,
+      "learning_rate": 6.588716192163269e-07,
+      "loss": 0.7759,
+      "step": 10466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3798405874164192,
+      "learning_rate": 6.582271162961428e-07,
+      "loss": 0.65,
+      "step": 10467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8684535040393007,
+      "learning_rate": 6.575829065423922e-07,
+      "loss": 0.6797,
+      "step": 10468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7827010100263714,
+      "learning_rate": 6.569389899985723e-07,
+      "loss": 1.0717,
+      "step": 10469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.400417183628508,
+      "learning_rate": 6.562953667081634e-07,
+      "loss": 0.704,
+      "step": 10470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.798061893404342,
+      "learning_rate": 6.556520367146246e-07,
+      "loss": 0.6846,
+      "step": 10471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5111573280707653,
+      "learning_rate": 6.550090000613935e-07,
+      "loss": 0.759,
+      "step": 10472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7389934228175498,
+      "learning_rate": 6.543662567918895e-07,
+      "loss": 1.0652,
+      "step": 10473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.489247324426463,
+      "learning_rate": 6.537238069495133e-07,
+      "loss": 0.7711,
+      "step": 10474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4523888456299305,
+      "learning_rate": 6.530816505776444e-07,
+      "loss": 0.8114,
+      "step": 10475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7402449252069958,
+      "learning_rate": 6.524397877196426e-07,
+      "loss": 1.0583,
+      "step": 10476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7606411275246279,
+      "learning_rate": 6.51798218418847e-07,
+      "loss": 0.7941,
+      "step": 10477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5600404326924076,
+      "learning_rate": 6.511569427185788e-07,
+      "loss": 0.7618,
+      "step": 10478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7803944325378279,
+      "learning_rate": 6.505159606621381e-07,
+      "loss": 1.03,
+      "step": 10479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7600361654022643,
+      "learning_rate": 6.498752722928042e-07,
+      "loss": 1.0262,
+      "step": 10480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5294669689718934,
+      "learning_rate": 6.492348776538398e-07,
+      "loss": 0.7593,
+      "step": 10481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.536786675000522,
+      "learning_rate": 6.48594776788486e-07,
+      "loss": 0.7547,
+      "step": 10482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7613802910804228,
+      "learning_rate": 6.479549697399612e-07,
+      "loss": 1.0398,
+      "step": 10483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4587158909304605,
+      "learning_rate": 6.473154565514695e-07,
+      "loss": 0.7985,
+      "step": 10484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.436131729704655,
+      "learning_rate": 6.466762372661911e-07,
+      "loss": 0.7075,
+      "step": 10485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3842884281038133,
+      "learning_rate": 6.460373119272867e-07,
+      "loss": 0.6712,
+      "step": 10486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6293583953349016,
+      "learning_rate": 6.453986805779006e-07,
+      "loss": 0.7617,
+      "step": 10487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.532988824443648,
+      "learning_rate": 6.447603432611533e-07,
+      "loss": 0.7726,
+      "step": 10488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4569271563263309,
+      "learning_rate": 6.441223000201457e-07,
+      "loss": 0.7952,
+      "step": 10489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.519212098672324,
+      "learning_rate": 6.434845508979598e-07,
+      "loss": 0.779,
+      "step": 10490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4604464914824096,
+      "learning_rate": 6.428470959376593e-07,
+      "loss": 0.7345,
+      "step": 10491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.557633962388726,
+      "learning_rate": 6.422099351822864e-07,
+      "loss": 1.0641,
+      "step": 10492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5456850324523328,
+      "learning_rate": 6.415730686748628e-07,
+      "loss": 0.6857,
+      "step": 10493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5852065714181305,
+      "learning_rate": 6.409364964583919e-07,
+      "loss": 0.7969,
+      "step": 10494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5970580993416825,
+      "learning_rate": 6.403002185758572e-07,
+      "loss": 0.8334,
+      "step": 10495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4631418824599978,
+      "learning_rate": 6.396642350702204e-07,
+      "loss": 0.6722,
+      "step": 10496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7403562797043832,
+      "learning_rate": 6.390285459844236e-07,
+      "loss": 1.0476,
+      "step": 10497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4839784079230407,
+      "learning_rate": 6.383931513613928e-07,
+      "loss": 0.7027,
+      "step": 10498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5650671328739711,
+      "learning_rate": 6.377580512440301e-07,
+      "loss": 0.7683,
+      "step": 10499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5469041121509528,
+      "learning_rate": 6.37123245675218e-07,
+      "loss": 0.6991,
+      "step": 10500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5479319778099485,
+      "learning_rate": 6.364887346978211e-07,
+      "loss": 0.7404,
+      "step": 10501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.566916579044323,
+      "learning_rate": 6.358545183546827e-07,
+      "loss": 0.6948,
+      "step": 10502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5878718975369863,
+      "learning_rate": 6.35220596688626e-07,
+      "loss": 0.7527,
+      "step": 10503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5189381160540856,
+      "learning_rate": 6.345869697424544e-07,
+      "loss": 0.6923,
+      "step": 10504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4767972658213235,
+      "learning_rate": 6.339536375589539e-07,
+      "loss": 0.8199,
+      "step": 10505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4415498090505827,
+      "learning_rate": 6.333206001808878e-07,
+      "loss": 0.7508,
+      "step": 10506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.534528153153127,
+      "learning_rate": 6.326878576509982e-07,
+      "loss": 0.8246,
+      "step": 10507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5364062504817957,
+      "learning_rate": 6.320554100120119e-07,
+      "loss": 0.7566,
+      "step": 10508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.461410407609989,
+      "learning_rate": 6.314232573066326e-07,
+      "loss": 0.7617,
+      "step": 10509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7569940145523752,
+      "learning_rate": 6.307913995775439e-07,
+      "loss": 0.8022,
+      "step": 10510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7246711628430602,
+      "learning_rate": 6.301598368674106e-07,
+      "loss": 1.0583,
+      "step": 10511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5319012186580074,
+      "learning_rate": 6.295285692188779e-07,
+      "loss": 0.7682,
+      "step": 10512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5217823381412638,
+      "learning_rate": 6.288975966745697e-07,
+      "loss": 0.7861,
+      "step": 10513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7571552220228096,
+      "learning_rate": 6.282669192770896e-07,
+      "loss": 1.0637,
+      "step": 10514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.546997368330712,
+      "learning_rate": 6.276365370690246e-07,
+      "loss": 0.7407,
+      "step": 10515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.518796557351805,
+      "learning_rate": 6.27006450092939e-07,
+      "loss": 0.8571,
+      "step": 10516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4646669307594735,
+      "learning_rate": 6.263766583913766e-07,
+      "loss": 0.7764,
+      "step": 10517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4832945058331344,
+      "learning_rate": 6.257471620068634e-07,
+      "loss": 0.7563,
+      "step": 10518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4436715239143716,
+      "learning_rate": 6.251179609819047e-07,
+      "loss": 0.8176,
+      "step": 10519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6196485537825582,
+      "learning_rate": 6.24489055358985e-07,
+      "loss": 0.8144,
+      "step": 10520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7433239592448581,
+      "learning_rate": 6.238604451805691e-07,
+      "loss": 1.0759,
+      "step": 10521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5846108999451023,
+      "learning_rate": 6.232321304891032e-07,
+      "loss": 0.8317,
+      "step": 10522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5950412344049762,
+      "learning_rate": 6.226041113270115e-07,
+      "loss": 0.7044,
+      "step": 10523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5757830643935897,
+      "learning_rate": 6.219763877366986e-07,
+      "loss": 0.7374,
+      "step": 10524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4973662899147278,
+      "learning_rate": 6.213489597605526e-07,
+      "loss": 0.7549,
+      "step": 10525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7715024760722176,
+      "learning_rate": 6.207218274409366e-07,
+      "loss": 0.7436,
+      "step": 10526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.560094627528198,
+      "learning_rate": 6.200949908201959e-07,
+      "loss": 0.7365,
+      "step": 10527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7509627516517307,
+      "learning_rate": 6.194684499406578e-07,
+      "loss": 1.0429,
+      "step": 10528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.579229615721901,
+      "learning_rate": 6.188422048446263e-07,
+      "loss": 0.8039,
+      "step": 10529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4175984073126289,
+      "learning_rate": 6.182162555743876e-07,
+      "loss": 0.703,
+      "step": 10530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5432174668425107,
+      "learning_rate": 6.175906021722055e-07,
+      "loss": 0.7652,
+      "step": 10531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.466592482822433,
+      "learning_rate": 6.169652446803292e-07,
+      "loss": 0.7603,
+      "step": 10532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4882773340321385,
+      "learning_rate": 6.16340183140981e-07,
+      "loss": 0.6865,
+      "step": 10533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7546393740179701,
+      "learning_rate": 6.157154175963665e-07,
+      "loss": 1.072,
+      "step": 10534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7588800931127216,
+      "learning_rate": 6.15090948088673e-07,
+      "loss": 1.0725,
+      "step": 10535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7590148260440126,
+      "learning_rate": 6.144667746600652e-07,
+      "loss": 1.0482,
+      "step": 10536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6369816884381203,
+      "learning_rate": 6.138428973526894e-07,
+      "loss": 0.8314,
+      "step": 10537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4528960419440884,
+      "learning_rate": 6.132193162086697e-07,
+      "loss": 0.6958,
+      "step": 10538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.510794793469759,
+      "learning_rate": 6.125960312701135e-07,
+      "loss": 0.8457,
+      "step": 10539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7655761919027867,
+      "learning_rate": 6.119730425791059e-07,
+      "loss": 1.0858,
+      "step": 10540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7416002711462988,
+      "learning_rate": 6.113503501777113e-07,
+      "loss": 0.7998,
+      "step": 10541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5648510335105579,
+      "learning_rate": 6.107279541079769e-07,
+      "loss": 0.7753,
+      "step": 10542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7467911789172526,
+      "learning_rate": 6.101058544119282e-07,
+      "loss": 1.057,
+      "step": 10543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7465331206212319,
+      "learning_rate": 6.094840511315703e-07,
+      "loss": 1.0873,
+      "step": 10544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5918462737205172,
+      "learning_rate": 6.088625443088885e-07,
+      "loss": 0.842,
+      "step": 10545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.453487384422699,
+      "learning_rate": 6.082413339858489e-07,
+      "loss": 0.7267,
+      "step": 10546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8232227349125614,
+      "learning_rate": 6.076204202043968e-07,
+      "loss": 0.6874,
+      "step": 10547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.49449028243409,
+      "learning_rate": 6.069998030064561e-07,
+      "loss": 0.7403,
+      "step": 10548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5052313433227271,
+      "learning_rate": 6.063794824339359e-07,
+      "loss": 0.7989,
+      "step": 10549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7350464854517975,
+      "learning_rate": 6.057594585287191e-07,
+      "loss": 0.7927,
+      "step": 10550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7028366267284434,
+      "learning_rate": 6.051397313326707e-07,
+      "loss": 0.7049,
+      "step": 10551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5287994030572938,
+      "learning_rate": 6.045203008876383e-07,
+      "loss": 0.7653,
+      "step": 10552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.76898414335785,
+      "learning_rate": 6.039011672354456e-07,
+      "loss": 1.0963,
+      "step": 10553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7478283684865068,
+      "learning_rate": 6.032823304178986e-07,
+      "loss": 1.0555,
+      "step": 10554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6455016538588079,
+      "learning_rate": 6.026637904767824e-07,
+      "loss": 0.6986,
+      "step": 10555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5621064560096705,
+      "learning_rate": 6.020455474538622e-07,
+      "loss": 0.7768,
+      "step": 10556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4870109306205488,
+      "learning_rate": 6.014276013908832e-07,
+      "loss": 0.7511,
+      "step": 10557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4465316600824325,
+      "learning_rate": 6.008099523295696e-07,
+      "loss": 0.8117,
+      "step": 10558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.560575544240619,
+      "learning_rate": 6.001926003116282e-07,
+      "loss": 0.7536,
+      "step": 10559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4566470195293502,
+      "learning_rate": 5.995755453787433e-07,
+      "loss": 0.6899,
+      "step": 10560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4862953275351058,
+      "learning_rate": 5.9895878757258e-07,
+      "loss": 0.7832,
+      "step": 10561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5247669532911683,
+      "learning_rate": 5.983423269347816e-07,
+      "loss": 0.6803,
+      "step": 10562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6797097384646174,
+      "learning_rate": 5.977261635069753e-07,
+      "loss": 0.6405,
+      "step": 10563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7546939017864326,
+      "learning_rate": 5.971102973307646e-07,
+      "loss": 1.0781,
+      "step": 10564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.594278704695481,
+      "learning_rate": 5.964947284477346e-07,
+      "loss": 0.7218,
+      "step": 10565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.560490576755668,
+      "learning_rate": 5.958794568994503e-07,
+      "loss": 0.7716,
+      "step": 10566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4795719876748201,
+      "learning_rate": 5.95264482727455e-07,
+      "loss": 0.7818,
+      "step": 10567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4794055424884966,
+      "learning_rate": 5.946498059732731e-07,
+      "loss": 0.7354,
+      "step": 10568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.565433059041811,
+      "learning_rate": 5.940354266784109e-07,
+      "loss": 0.762,
+      "step": 10569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5304103532499334,
+      "learning_rate": 5.934213448843512e-07,
+      "loss": 0.7328,
+      "step": 10570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5413959449152705,
+      "learning_rate": 5.92807560632559e-07,
+      "loss": 0.8255,
+      "step": 10571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5585873185341812,
+      "learning_rate": 5.921940739644766e-07,
+      "loss": 0.7549,
+      "step": 10572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7538045039970029,
+      "learning_rate": 5.915808849215304e-07,
+      "loss": 1.03,
+      "step": 10573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4711792948008318,
+      "learning_rate": 5.909679935451235e-07,
+      "loss": 0.8114,
+      "step": 10574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6215686485155403,
+      "learning_rate": 5.903553998766387e-07,
+      "loss": 0.7932,
+      "step": 10575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3989742993734906,
+      "learning_rate": 5.897431039574414e-07,
+      "loss": 0.655,
+      "step": 10576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7540439000225984,
+      "learning_rate": 5.891311058288751e-07,
+      "loss": 1.0383,
+      "step": 10577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8160916585887958,
+      "learning_rate": 5.885194055322618e-07,
+      "loss": 1.0623,
+      "step": 10578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4122437461049024,
+      "learning_rate": 5.879080031089047e-07,
+      "loss": 0.7229,
+      "step": 10579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.560511260000694,
+      "learning_rate": 5.872968986000893e-07,
+      "loss": 0.7371,
+      "step": 10580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4425415410589826,
+      "learning_rate": 5.866860920470773e-07,
+      "loss": 0.7791,
+      "step": 10581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5056419993815482,
+      "learning_rate": 5.860755834911108e-07,
+      "loss": 0.7832,
+      "step": 10582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4566053990461625,
+      "learning_rate": 5.854653729734156e-07,
+      "loss": 0.7244,
+      "step": 10583
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4881113783124142,
+      "learning_rate": 5.848554605351925e-07,
+      "loss": 0.7012,
+      "step": 10584
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6786858323430598,
+      "learning_rate": 5.842458462176231e-07,
+      "loss": 0.7227,
+      "step": 10585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.553182825504059,
+      "learning_rate": 5.83636530061873e-07,
+      "loss": 0.7909,
+      "step": 10586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6468691723998505,
+      "learning_rate": 5.830275121090828e-07,
+      "loss": 0.8033,
+      "step": 10587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.602047408575517,
+      "learning_rate": 5.824187924003749e-07,
+      "loss": 0.8198,
+      "step": 10588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4679298566231678,
+      "learning_rate": 5.818103709768519e-07,
+      "loss": 0.7484,
+      "step": 10589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4254932942182705,
+      "learning_rate": 5.812022478795954e-07,
+      "loss": 0.7359,
+      "step": 10590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6418651659946617,
+      "learning_rate": 5.805944231496669e-07,
+      "loss": 0.7693,
+      "step": 10591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.517608652066018,
+      "learning_rate": 5.799868968281075e-07,
+      "loss": 0.7509,
+      "step": 10592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.43812808123009,
+      "learning_rate": 5.793796689559411e-07,
+      "loss": 0.707,
+      "step": 10593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7133115875627756,
+      "learning_rate": 5.787727395741682e-07,
+      "loss": 0.7097,
+      "step": 10594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5570657839993853,
+      "learning_rate": 5.781661087237689e-07,
+      "loss": 0.7463,
+      "step": 10595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.679370956468701,
+      "learning_rate": 5.775597764457047e-07,
+      "loss": 0.8101,
+      "step": 10596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5270219386733994,
+      "learning_rate": 5.769537427809174e-07,
+      "loss": 0.69,
+      "step": 10597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4543781993299443,
+      "learning_rate": 5.763480077703276e-07,
+      "loss": 0.7765,
+      "step": 10598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.524973318580101,
+      "learning_rate": 5.757425714548354e-07,
+      "loss": 0.8068,
+      "step": 10599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.51206405713927,
+      "learning_rate": 5.751374338753218e-07,
+      "loss": 0.6935,
+      "step": 10600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5687466764301068,
+      "learning_rate": 5.745325950726466e-07,
+      "loss": 0.7385,
+      "step": 10601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9809507542362046,
+      "learning_rate": 5.739280550876497e-07,
+      "loss": 0.8308,
+      "step": 10602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7515309932223118,
+      "learning_rate": 5.733238139611508e-07,
+      "loss": 1.0716,
+      "step": 10603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5373506235425038,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.7572,
+      "step": 10604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.901289004258174,
+      "learning_rate": 5.72116228446829e-07,
+      "loss": 0.7689,
+      "step": 10605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.555798067149641,
+      "learning_rate": 5.715128841405432e-07,
+      "loss": 0.8148,
+      "step": 10606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6860447046847298,
+      "learning_rate": 5.709098388558348e-07,
+      "loss": 0.7504,
+      "step": 10607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4323710878262288,
+      "learning_rate": 5.70307092633422e-07,
+      "loss": 0.6863,
+      "step": 10608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.53317670527803,
+      "learning_rate": 5.697046455140031e-07,
+      "loss": 0.7997,
+      "step": 10609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7608030377371746,
+      "learning_rate": 5.69102497538257e-07,
+      "loss": 1.0451,
+      "step": 10610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6228869480140762,
+      "learning_rate": 5.685006487468426e-07,
+      "loss": 0.7675,
+      "step": 10611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5163830932052216,
+      "learning_rate": 5.678990991803973e-07,
+      "loss": 0.7108,
+      "step": 10612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4974097507295567,
+      "learning_rate": 5.672978488795383e-07,
+      "loss": 0.7499,
+      "step": 10613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5732594439684298,
+      "learning_rate": 5.666968978848659e-07,
+      "loss": 0.7551,
+      "step": 10614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4230785523989802,
+      "learning_rate": 5.660962462369562e-07,
+      "loss": 0.796,
+      "step": 10615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5072422378450723,
+      "learning_rate": 5.654958939763655e-07,
+      "loss": 0.7186,
+      "step": 10616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4838842175019953,
+      "learning_rate": 5.648958411436334e-07,
+      "loss": 0.798,
+      "step": 10617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7589315873358868,
+      "learning_rate": 5.642960877792752e-07,
+      "loss": 1.0258,
+      "step": 10618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8951134175728026,
+      "learning_rate": 5.636966339237882e-07,
+      "loss": 0.7227,
+      "step": 10619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5077349614705646,
+      "learning_rate": 5.630974796176481e-07,
+      "loss": 0.8248,
+      "step": 10620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.478941503368043,
+      "learning_rate": 5.624986249013131e-07,
+      "loss": 0.7602,
+      "step": 10621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.49673180997114,
+      "learning_rate": 5.619000698152171e-07,
+      "loss": 0.7704,
+      "step": 10622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7298171328792727,
+      "learning_rate": 5.613018143997762e-07,
+      "loss": 1.082,
+      "step": 10623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6805563470773388,
+      "learning_rate": 5.607038586953873e-07,
+      "loss": 0.7124,
+      "step": 10624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7777007368980453,
+      "learning_rate": 5.601062027424243e-07,
+      "loss": 1.0249,
+      "step": 10625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5706155518184461,
+      "learning_rate": 5.595088465812426e-07,
+      "loss": 0.7515,
+      "step": 10626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5097634878890067,
+      "learning_rate": 5.589117902521779e-07,
+      "loss": 0.6507,
+      "step": 10627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4633041537008902,
+      "learning_rate": 5.583150337955445e-07,
+      "loss": 0.7434,
+      "step": 10628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.404378663824516,
+      "learning_rate": 5.577185772516369e-07,
+      "loss": 0.6991,
+      "step": 10629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4176453633116144,
+      "learning_rate": 5.571224206607274e-07,
+      "loss": 0.6669,
+      "step": 10630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4784067185545693,
+      "learning_rate": 5.565265640630724e-07,
+      "loss": 0.678,
+      "step": 10631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5577111898788731,
+      "learning_rate": 5.559310074989044e-07,
+      "loss": 0.7399,
+      "step": 10632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7459198882130866,
+      "learning_rate": 5.553357510084368e-07,
+      "loss": 1.0284,
+      "step": 10633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4857550489252411,
+      "learning_rate": 5.547407946318628e-07,
+      "loss": 0.6937,
+      "step": 10634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7361966300294183,
+      "learning_rate": 5.541461384093549e-07,
+      "loss": 1.0421,
+      "step": 10635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6208521134451581,
+      "learning_rate": 5.535517823810654e-07,
+      "loss": 0.6652,
+      "step": 10636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.762287939537756,
+      "learning_rate": 5.529577265871266e-07,
+      "loss": 1.0584,
+      "step": 10637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5966156835091907,
+      "learning_rate": 5.523639710676515e-07,
+      "loss": 0.7307,
+      "step": 10638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8513473404729726,
+      "learning_rate": 5.517705158627313e-07,
+      "loss": 0.8013,
+      "step": 10639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3546384272239413,
+      "learning_rate": 5.511773610124366e-07,
+      "loss": 0.6617,
+      "step": 10640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5471172118421526,
+      "learning_rate": 5.5058450655682e-07,
+      "loss": 0.7299,
+      "step": 10641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.554386273443815,
+      "learning_rate": 5.499919525359121e-07,
+      "loss": 0.6868,
+      "step": 10642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4492036587014332,
+      "learning_rate": 5.493996989897227e-07,
+      "loss": 0.7384,
+      "step": 10643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5925727195266548,
+      "learning_rate": 5.488077459582425e-07,
+      "loss": 0.78,
+      "step": 10644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5962278724681638,
+      "learning_rate": 5.482160934814418e-07,
+      "loss": 0.7888,
+      "step": 10645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.501442019924188,
+      "learning_rate": 5.476247415992702e-07,
+      "loss": 0.7524,
+      "step": 10646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6827083466978405,
+      "learning_rate": 5.47033690351656e-07,
+      "loss": 0.7151,
+      "step": 10647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6102823960522417,
+      "learning_rate": 5.464429397785099e-07,
+      "loss": 0.7018,
+      "step": 10648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.490299365644473,
+      "learning_rate": 5.458524899197204e-07,
+      "loss": 0.686,
+      "step": 10649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4864402783449124,
+      "learning_rate": 5.452623408151553e-07,
+      "loss": 0.853,
+      "step": 10650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7611929007203355,
+      "learning_rate": 5.446724925046637e-07,
+      "loss": 1.033,
+      "step": 10651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7541218496393737,
+      "learning_rate": 5.440829450280732e-07,
+      "loss": 1.0902,
+      "step": 10652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6199859173166222,
+      "learning_rate": 5.434936984251916e-07,
+      "loss": 0.7417,
+      "step": 10653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4269645401209525,
+      "learning_rate": 5.429047527358056e-07,
+      "loss": 0.7247,
+      "step": 10654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6074244734168412,
+      "learning_rate": 5.423161079996824e-07,
+      "loss": 0.6565,
+      "step": 10655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5036907991710482,
+      "learning_rate": 5.417277642565694e-07,
+      "loss": 0.7488,
+      "step": 10656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4611061702109365,
+      "learning_rate": 5.411397215461905e-07,
+      "loss": 0.879,
+      "step": 10657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.496054497251319,
+      "learning_rate": 5.405519799082548e-07,
+      "loss": 0.7658,
+      "step": 10658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.489377056496907,
+      "learning_rate": 5.399645393824465e-07,
+      "loss": 0.7393,
+      "step": 10659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5232933804575872,
+      "learning_rate": 5.393774000084307e-07,
+      "loss": 0.6974,
+      "step": 10660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7608169839726203,
+      "learning_rate": 5.387905618258521e-07,
+      "loss": 1.0407,
+      "step": 10661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4993444599169157,
+      "learning_rate": 5.382040248743364e-07,
+      "loss": 0.745,
+      "step": 10662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6753462121695666,
+      "learning_rate": 5.37617789193488e-07,
+      "loss": 0.7328,
+      "step": 10663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.570468865217088,
+      "learning_rate": 5.370318548228886e-07,
+      "loss": 0.8535,
+      "step": 10664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6027263463616042,
+      "learning_rate": 5.364462218021066e-07,
+      "loss": 0.8198,
+      "step": 10665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5809352373172658,
+      "learning_rate": 5.358608901706802e-07,
+      "loss": 0.7555,
+      "step": 10666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5646452410986769,
+      "learning_rate": 5.352758599681341e-07,
+      "loss": 0.7864,
+      "step": 10667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5513899139727896,
+      "learning_rate": 5.346911312339719e-07,
+      "loss": 0.7528,
+      "step": 10668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6679122564609645,
+      "learning_rate": 5.341067040076752e-07,
+      "loss": 0.736,
+      "step": 10669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6261253776910327,
+      "learning_rate": 5.335225783287051e-07,
+      "loss": 0.7118,
+      "step": 10670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6282146964290545,
+      "learning_rate": 5.329387542365033e-07,
+      "loss": 0.7479,
+      "step": 10671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4673581294017208,
+      "learning_rate": 5.323552317704922e-07,
+      "loss": 0.7149,
+      "step": 10672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5941734477844849,
+      "learning_rate": 5.317720109700719e-07,
+      "loss": 0.7499,
+      "step": 10673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4419007244734823,
+      "learning_rate": 5.311890918746216e-07,
+      "loss": 0.7151,
+      "step": 10674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7627666863689653,
+      "learning_rate": 5.306064745235035e-07,
+      "loss": 1.0591,
+      "step": 10675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6717186124925845,
+      "learning_rate": 5.300241589560556e-07,
+      "loss": 0.7445,
+      "step": 10676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7739806745865697,
+      "learning_rate": 5.294421452115983e-07,
+      "loss": 1.035,
+      "step": 10677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5334945218157354,
+      "learning_rate": 5.288604333294295e-07,
+      "loss": 0.7564,
+      "step": 10678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4603335532675281,
+      "learning_rate": 5.282790233488283e-07,
+      "loss": 0.7062,
+      "step": 10679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5477972123553387,
+      "learning_rate": 5.276979153090528e-07,
+      "loss": 0.7939,
+      "step": 10680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7677028954458824,
+      "learning_rate": 5.271171092493393e-07,
+      "loss": 1.1006,
+      "step": 10681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4964386917454109,
+      "learning_rate": 5.265366052089077e-07,
+      "loss": 0.6849,
+      "step": 10682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7622269651061627,
+      "learning_rate": 5.259564032269538e-07,
+      "loss": 1.079,
+      "step": 10683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5876986363839836,
+      "learning_rate": 5.253765033426528e-07,
+      "loss": 0.7443,
+      "step": 10684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5082872924646236,
+      "learning_rate": 5.24796905595163e-07,
+      "loss": 0.7194,
+      "step": 10685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5832711948298928,
+      "learning_rate": 5.242176100236195e-07,
+      "loss": 0.7424,
+      "step": 10686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6574993504247448,
+      "learning_rate": 5.236386166671376e-07,
+      "loss": 0.7026,
+      "step": 10687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8040597391468379,
+      "learning_rate": 5.230599255648116e-07,
+      "loss": 1.05,
+      "step": 10688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5334024951417946,
+      "learning_rate": 5.224815367557173e-07,
+      "loss": 0.7088,
+      "step": 10689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4733085546152755,
+      "learning_rate": 5.219034502789078e-07,
+      "loss": 0.7965,
+      "step": 10690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5127337881537215,
+      "learning_rate": 5.213256661734162e-07,
+      "loss": 0.7084,
+      "step": 10691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5528205234778825,
+      "learning_rate": 5.207481844782575e-07,
+      "loss": 0.8372,
+      "step": 10692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7483865997239115,
+      "learning_rate": 5.201710052324238e-07,
+      "loss": 1.0572,
+      "step": 10693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7582654611289715,
+      "learning_rate": 5.195941284748879e-07,
+      "loss": 1.0846,
+      "step": 10694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7349772349306587,
+      "learning_rate": 5.190175542446002e-07,
+      "loss": 1.0276,
+      "step": 10695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7400493499720503,
+      "learning_rate": 5.184412825804947e-07,
+      "loss": 1.0207,
+      "step": 10696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4404320840150981,
+      "learning_rate": 5.178653135214811e-07,
+      "loss": 0.6797,
+      "step": 10697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7545429608427342,
+      "learning_rate": 5.172896471064514e-07,
+      "loss": 1.0821,
+      "step": 10698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6037273647412593,
+      "learning_rate": 5.167142833742744e-07,
+      "loss": 0.7371,
+      "step": 10699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6070630630046192,
+      "learning_rate": 5.161392223638012e-07,
+      "loss": 0.7721,
+      "step": 10700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5528749814279657,
+      "learning_rate": 5.155644641138602e-07,
+      "loss": 0.6195,
+      "step": 10701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.53077067565577,
+      "learning_rate": 5.149900086632597e-07,
+      "loss": 0.7224,
+      "step": 10702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5688498916884153,
+      "learning_rate": 5.144158560507912e-07,
+      "loss": 0.7769,
+      "step": 10703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.50992264124819,
+      "learning_rate": 5.138420063152205e-07,
+      "loss": 0.7747,
+      "step": 10704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7522032466641091,
+      "learning_rate": 5.132684594952946e-07,
+      "loss": 1.0448,
+      "step": 10705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4384324068280459,
+      "learning_rate": 5.126952156297433e-07,
+      "loss": 0.7015,
+      "step": 10706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.463693899665976,
+      "learning_rate": 5.121222747572712e-07,
+      "loss": 0.7026,
+      "step": 10707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.467132058479348,
+      "learning_rate": 5.115496369165651e-07,
+      "loss": 0.7636,
+      "step": 10708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5576195332032134,
+      "learning_rate": 5.109773021462921e-07,
+      "loss": 0.79,
+      "step": 10709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5140269972669043,
+      "learning_rate": 5.10405270485096e-07,
+      "loss": 0.6938,
+      "step": 10710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.49063356643001,
+      "learning_rate": 5.098335419716022e-07,
+      "loss": 0.722,
+      "step": 10711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7815024520261052,
+      "learning_rate": 5.092621166444139e-07,
+      "loss": 1.0772,
+      "step": 10712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7424635467214203,
+      "learning_rate": 5.08690994542117e-07,
+      "loss": 1.0491,
+      "step": 10713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6162737516251804,
+      "learning_rate": 5.081201757032744e-07,
+      "loss": 0.743,
+      "step": 10714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.501818171953974,
+      "learning_rate": 5.075496601664276e-07,
+      "loss": 0.7671,
+      "step": 10715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5465526333440849,
+      "learning_rate": 5.069794479701013e-07,
+      "loss": 0.7221,
+      "step": 10716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.507311805627084,
+      "learning_rate": 5.064095391527968e-07,
+      "loss": 0.6899,
+      "step": 10717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5830778815080264,
+      "learning_rate": 5.058399337529957e-07,
+      "loss": 0.8064,
+      "step": 10718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5309835635536504,
+      "learning_rate": 5.052706318091572e-07,
+      "loss": 0.7175,
+      "step": 10719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6451625038781048,
+      "learning_rate": 5.047016333597248e-07,
+      "loss": 0.7597,
+      "step": 10720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7650018947353433,
+      "learning_rate": 5.04132938443117e-07,
+      "loss": 1.046,
+      "step": 10721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.654864128033921,
+      "learning_rate": 5.03564547097734e-07,
+      "loss": 0.8197,
+      "step": 10722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5502362618439645,
+      "learning_rate": 5.029964593619541e-07,
+      "loss": 0.7866,
+      "step": 10723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.753674645585248,
+      "learning_rate": 5.024286752741364e-07,
+      "loss": 1.0435,
+      "step": 10724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5817776832524295,
+      "learning_rate": 5.018611948726182e-07,
+      "loss": 0.7713,
+      "step": 10725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5760535695866238,
+      "learning_rate": 5.012940181957182e-07,
+      "loss": 0.761,
+      "step": 10726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5417930896154035,
+      "learning_rate": 5.00727145281733e-07,
+      "loss": 0.7898,
+      "step": 10727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6038448645944507,
+      "learning_rate": 5.001605761689399e-07,
+      "loss": 0.7431,
+      "step": 10728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6170973341904578,
+      "learning_rate": 4.995943108955926e-07,
+      "loss": 0.8111,
+      "step": 10729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.734378877963947,
+      "learning_rate": 4.990283494999293e-07,
+      "loss": 0.6849,
+      "step": 10730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4127169500503385,
+      "learning_rate": 4.984626920201641e-07,
+      "loss": 0.7272,
+      "step": 10731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7293267566120726,
+      "learning_rate": 4.978973384944913e-07,
+      "loss": 1.0434,
+      "step": 10732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5911336987141511,
+      "learning_rate": 4.973322889610849e-07,
+      "loss": 0.7567,
+      "step": 10733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.578865247044147,
+      "learning_rate": 4.967675434580982e-07,
+      "loss": 0.7785,
+      "step": 10734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.519050588420745,
+      "learning_rate": 4.96203102023664e-07,
+      "loss": 0.7441,
+      "step": 10735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4829550624644494,
+      "learning_rate": 4.956389646958943e-07,
+      "loss": 0.7438,
+      "step": 10736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4374848474141362,
+      "learning_rate": 4.950751315128821e-07,
+      "loss": 0.6736,
+      "step": 10737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5602851056143778,
+      "learning_rate": 4.945116025126984e-07,
+      "loss": 0.7864,
+      "step": 10738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5512055640208706,
+      "learning_rate": 4.939483777333931e-07,
+      "loss": 0.7395,
+      "step": 10739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6742643399237433,
+      "learning_rate": 4.933854572129975e-07,
+      "loss": 0.7614,
+      "step": 10740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.605704445786202,
+      "learning_rate": 4.928228409895214e-07,
+      "loss": 0.7742,
+      "step": 10741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5005978452659055,
+      "learning_rate": 4.922605291009525e-07,
+      "loss": 0.7678,
+      "step": 10742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5275244881984402,
+      "learning_rate": 4.91698521585261e-07,
+      "loss": 0.7169,
+      "step": 10743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.602446129312948,
+      "learning_rate": 4.911368184803939e-07,
+      "loss": 0.7834,
+      "step": 10744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4295698713918725,
+      "learning_rate": 4.90575419824279e-07,
+      "loss": 0.7101,
+      "step": 10745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4767143973314751,
+      "learning_rate": 4.900143256548223e-07,
+      "loss": 0.7396,
+      "step": 10746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6422440857329006,
+      "learning_rate": 4.894535360099117e-07,
+      "loss": 0.8586,
+      "step": 10747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7583225780138904,
+      "learning_rate": 4.888930509274125e-07,
+      "loss": 1.0765,
+      "step": 10748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5774375851330238,
+      "learning_rate": 4.883328704451689e-07,
+      "loss": 0.6701,
+      "step": 10749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4796662597927759,
+      "learning_rate": 4.877729946010073e-07,
+      "loss": 0.7659,
+      "step": 10750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5681507967343238,
+      "learning_rate": 4.872134234327308e-07,
+      "loss": 0.8154,
+      "step": 10751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.478846645597066,
+      "learning_rate": 4.866541569781235e-07,
+      "loss": 0.7795,
+      "step": 10752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7452529042061491,
+      "learning_rate": 4.860951952749477e-07,
+      "loss": 1.075,
+      "step": 10753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.535543027183284,
+      "learning_rate": 4.855365383609457e-07,
+      "loss": 0.8176,
+      "step": 10754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4926332519109893,
+      "learning_rate": 4.849781862738401e-07,
+      "loss": 0.7389,
+      "step": 10755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.587253276206484,
+      "learning_rate": 4.844201390513297e-07,
+      "loss": 0.7457,
+      "step": 10756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4699854656303906,
+      "learning_rate": 4.83862396731099e-07,
+      "loss": 0.6814,
+      "step": 10757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5288191605767416,
+      "learning_rate": 4.833049593508055e-07,
+      "loss": 0.6933,
+      "step": 10758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6214061477740267,
+      "learning_rate": 4.827478269480895e-07,
+      "loss": 0.7883,
+      "step": 10759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5931764849037424,
+      "learning_rate": 4.821909995605684e-07,
+      "loss": 0.7166,
+      "step": 10760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7592697801337348,
+      "learning_rate": 4.816344772258425e-07,
+      "loss": 1.0793,
+      "step": 10761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5219655014312945,
+      "learning_rate": 4.810782599814884e-07,
+      "loss": 0.7452,
+      "step": 10762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5361478950549132,
+      "learning_rate": 4.805223478650628e-07,
+      "loss": 0.7826,
+      "step": 10763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6620316855875783,
+      "learning_rate": 4.799667409141035e-07,
+      "loss": 0.7281,
+      "step": 10764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4915248735035098,
+      "learning_rate": 4.794114391661253e-07,
+      "loss": 0.8015,
+      "step": 10765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3795497675119444,
+      "learning_rate": 4.78856442658624e-07,
+      "loss": 0.7223,
+      "step": 10766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5006046817147776,
+      "learning_rate": 4.783017514290739e-07,
+      "loss": 0.7259,
+      "step": 10767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6752794783781924,
+      "learning_rate": 4.77747365514929e-07,
+      "loss": 0.7531,
+      "step": 10768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4916175605608066,
+      "learning_rate": 4.77193284953622e-07,
+      "loss": 0.7963,
+      "step": 10769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7439317600220771,
+      "learning_rate": 4.7663950978256657e-07,
+      "loss": 1.0725,
+      "step": 10770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5048080432530966,
+      "learning_rate": 4.760860400391548e-07,
+      "loss": 0.7688,
+      "step": 10771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5303043080785739,
+      "learning_rate": 4.755328757607586e-07,
+      "loss": 0.6779,
+      "step": 10772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7614386974255439,
+      "learning_rate": 4.7498001698472793e-07,
+      "loss": 1.0662,
+      "step": 10773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3839866251104858,
+      "learning_rate": 4.7442746374839363e-07,
+      "loss": 0.7342,
+      "step": 10774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4184984169412276,
+      "learning_rate": 4.7387521608906585e-07,
+      "loss": 0.796,
+      "step": 10775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4507836053134406,
+      "learning_rate": 4.73323274044033e-07,
+      "loss": 0.7284,
+      "step": 10776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.425163291562211,
+      "learning_rate": 4.727716376505637e-07,
+      "loss": 0.7033,
+      "step": 10777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.558501208118554,
+      "learning_rate": 4.722203069459053e-07,
+      "loss": 0.8186,
+      "step": 10778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7717244550680494,
+      "learning_rate": 4.7166928196728524e-07,
+      "loss": 1.0596,
+      "step": 10779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4471640301629336,
+      "learning_rate": 4.7111856275190937e-07,
+      "loss": 0.6609,
+      "step": 10780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.49447412732685,
+      "learning_rate": 4.705681493369646e-07,
+      "loss": 0.71,
+      "step": 10781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.756293533275371,
+      "learning_rate": 4.700180417596156e-07,
+      "loss": 1.0765,
+      "step": 10782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.582992954933307,
+      "learning_rate": 4.6946824005700606e-07,
+      "loss": 0.6741,
+      "step": 10783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7463881622146418,
+      "learning_rate": 4.6891874426626125e-07,
+      "loss": 1.0418,
+      "step": 10784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5246374399681542,
+      "learning_rate": 4.683695544244843e-07,
+      "loss": 0.7455,
+      "step": 10785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6077003125687528,
+      "learning_rate": 4.678206705687566e-07,
+      "loss": 0.7668,
+      "step": 10786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5461386255268992,
+      "learning_rate": 4.672720927361413e-07,
+      "loss": 0.7456,
+      "step": 10787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6126972276297873,
+      "learning_rate": 4.667238209636782e-07,
+      "loss": 0.7584,
+      "step": 10788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6057232003618762,
+      "learning_rate": 4.6617585528838937e-07,
+      "loss": 0.7825,
+      "step": 10789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5496909320708245,
+      "learning_rate": 4.6562819574727304e-07,
+      "loss": 0.716,
+      "step": 10790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.58685017039491,
+      "learning_rate": 4.650808423773101e-07,
+      "loss": 0.7661,
+      "step": 10791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5882866282604633,
+      "learning_rate": 4.645337952154583e-07,
+      "loss": 0.7195,
+      "step": 10792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.502770437280089,
+      "learning_rate": 4.6398705429865574e-07,
+      "loss": 0.7107,
+      "step": 10793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4941326371463215,
+      "learning_rate": 4.634406196638186e-07,
+      "loss": 0.754,
+      "step": 10794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.509496447284733,
+      "learning_rate": 4.62894491347845e-07,
+      "loss": 0.6891,
+      "step": 10795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5645790937728679,
+      "learning_rate": 4.623486693876106e-07,
+      "loss": 0.7143,
+      "step": 10796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4794202390874187,
+      "learning_rate": 4.6180315381996963e-07,
+      "loss": 0.726,
+      "step": 10797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.589974092047516,
+      "learning_rate": 4.6125794468175723e-07,
+      "loss": 0.7225,
+      "step": 10798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.454638255372068,
+      "learning_rate": 4.607130420097866e-07,
+      "loss": 0.7733,
+      "step": 10799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5838284963981806,
+      "learning_rate": 4.601684458408506e-07,
+      "loss": 0.7507,
+      "step": 10800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4978858351776485,
+      "learning_rate": 4.5962415621172205e-07,
+      "loss": 0.7576,
+      "step": 10801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5015797210270119,
+      "learning_rate": 4.590801731591532e-07,
+      "loss": 0.8012,
+      "step": 10802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.544530860167229,
+      "learning_rate": 4.5853649671987464e-07,
+      "loss": 0.781,
+      "step": 10803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5519572991136248,
+      "learning_rate": 4.579931269305954e-07,
+      "loss": 0.7494,
+      "step": 10804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.504341838994756,
+      "learning_rate": 4.574500638280072e-07,
+      "loss": 0.7815,
+      "step": 10805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5032595960959758,
+      "learning_rate": 4.5690730744877733e-07,
+      "loss": 0.7027,
+      "step": 10806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4196085629133373,
+      "learning_rate": 4.563648578295543e-07,
+      "loss": 0.7103,
+      "step": 10807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.362049089544699,
+      "learning_rate": 4.5582271500696607e-07,
+      "loss": 0.694,
+      "step": 10808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.535717556051172,
+      "learning_rate": 4.552808790176194e-07,
+      "loss": 0.8219,
+      "step": 10809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4118023088878167,
+      "learning_rate": 4.5473934989810064e-07,
+      "loss": 0.7463,
+      "step": 10810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7594779698868047,
+      "learning_rate": 4.5419812768497274e-07,
+      "loss": 1.0131,
+      "step": 10811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5970778011653948,
+      "learning_rate": 4.5365721241478256e-07,
+      "loss": 0.7528,
+      "step": 10812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.748541265263908,
+      "learning_rate": 4.531166041240531e-07,
+      "loss": 1.0409,
+      "step": 10813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7352203158506543,
+      "learning_rate": 4.525763028492869e-07,
+      "loss": 1.0644,
+      "step": 10814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6181497760065366,
+      "learning_rate": 4.5203630862696803e-07,
+      "loss": 0.7538,
+      "step": 10815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.401876461513214,
+      "learning_rate": 4.514966214935573e-07,
+      "loss": 0.7475,
+      "step": 10816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.522176580198074,
+      "learning_rate": 4.5095724148549515e-07,
+      "loss": 0.7864,
+      "step": 10817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.436349681437484,
+      "learning_rate": 4.504181686392012e-07,
+      "loss": 0.7563,
+      "step": 10818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4925702902415443,
+      "learning_rate": 4.498794029910769e-07,
+      "loss": 0.7118,
+      "step": 10819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.612995103467293,
+      "learning_rate": 4.4934094457749934e-07,
+      "loss": 0.7689,
+      "step": 10820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7477687298888301,
+      "learning_rate": 4.4880279343482713e-07,
+      "loss": 1.0679,
+      "step": 10821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5697601672896522,
+      "learning_rate": 4.482649495993974e-07,
+      "loss": 0.6962,
+      "step": 10822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5751908979681308,
+      "learning_rate": 4.4772741310752653e-07,
+      "loss": 0.7833,
+      "step": 10823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4855339785603408,
+      "learning_rate": 4.4719018399550893e-07,
+      "loss": 0.6843,
+      "step": 10824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.431638528527591,
+      "learning_rate": 4.4665326229962167e-07,
+      "loss": 0.7307,
+      "step": 10825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6236283987898097,
+      "learning_rate": 4.46116648056118e-07,
+      "loss": 0.7217,
+      "step": 10826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5773995367136033,
+      "learning_rate": 4.455803413012316e-07,
+      "loss": 0.7706,
+      "step": 10827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.591105818424367,
+      "learning_rate": 4.4504434207117363e-07,
+      "loss": 0.824,
+      "step": 10828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7419227931692505,
+      "learning_rate": 4.445086504021384e-07,
+      "loss": 1.093,
+      "step": 10829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6440755997786751,
+      "learning_rate": 4.439732663302954e-07,
+      "loss": 0.7256,
+      "step": 10830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7665875502893416,
+      "learning_rate": 4.434381898917961e-07,
+      "loss": 0.7668,
+      "step": 10831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7618737919593795,
+      "learning_rate": 4.4290342112276895e-07,
+      "loss": 1.0827,
+      "step": 10832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7819819426092904,
+      "learning_rate": 4.423689600593234e-07,
+      "loss": 1.0409,
+      "step": 10833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4977293824654383,
+      "learning_rate": 4.418348067375472e-07,
+      "loss": 0.8294,
+      "step": 10834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7449262311134747,
+      "learning_rate": 4.4130096119350707e-07,
+      "loss": 1.0324,
+      "step": 10835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5676177864437064,
+      "learning_rate": 4.4076742346325086e-07,
+      "loss": 0.7231,
+      "step": 10836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4885152709754226,
+      "learning_rate": 4.4023419358280307e-07,
+      "loss": 0.7561,
+      "step": 10837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.585636121606808,
+      "learning_rate": 4.397012715881688e-07,
+      "loss": 0.9106,
+      "step": 10838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7250152348376198,
+      "learning_rate": 4.3916865751533313e-07,
+      "loss": 0.7634,
+      "step": 10839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5883764889785326,
+      "learning_rate": 4.38636351400259e-07,
+      "loss": 0.7862,
+      "step": 10840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5437528587024798,
+      "learning_rate": 4.3810435327888814e-07,
+      "loss": 0.7308,
+      "step": 10841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5543574001658695,
+      "learning_rate": 4.37572663187143e-07,
+      "loss": 0.7234,
+      "step": 10842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5635287132782059,
+      "learning_rate": 4.3704128116092423e-07,
+      "loss": 0.7016,
+      "step": 10843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4471386716701087,
+      "learning_rate": 4.365102072361116e-07,
+      "loss": 0.6631,
+      "step": 10844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4305168448675005,
+      "learning_rate": 4.359794414485646e-07,
+      "loss": 0.7912,
+      "step": 10845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6476042030217537,
+      "learning_rate": 4.354489838341225e-07,
+      "loss": 0.8229,
+      "step": 10846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6195577112356045,
+      "learning_rate": 4.3491883442860263e-07,
+      "loss": 0.8073,
+      "step": 10847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7432980531406017,
+      "learning_rate": 4.343889932678008e-07,
+      "loss": 1.0165,
+      "step": 10848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.615904267883659,
+      "learning_rate": 4.338594603874946e-07,
+      "loss": 0.6997,
+      "step": 10849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7759843240832383,
+      "learning_rate": 4.3333023582343925e-07,
+      "loss": 1.0559,
+      "step": 10850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5663035775603884,
+      "learning_rate": 4.328013196113684e-07,
+      "loss": 0.8689,
+      "step": 10851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4449564823603271,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.6754,
+      "step": 10852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6791153537287589,
+      "learning_rate": 4.317444123860143e-07,
+      "loss": 0.8228,
+      "step": 10853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.521176159129267,
+      "learning_rate": 4.3121642144409726e-07,
+      "loss": 0.7174,
+      "step": 10854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.619480829500294,
+      "learning_rate": 4.306887389968928e-07,
+      "loss": 0.7431,
+      "step": 10855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7226325045998638,
+      "learning_rate": 4.3016136508003404e-07,
+      "loss": 0.9968,
+      "step": 10856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5059336600920292,
+      "learning_rate": 4.296342997291292e-07,
+      "loss": 0.7245,
+      "step": 10857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4866489154432223,
+      "learning_rate": 4.2910754297976755e-07,
+      "loss": 0.7428,
+      "step": 10858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.750979320381967,
+      "learning_rate": 4.285810948675156e-07,
+      "loss": 1.054,
+      "step": 10859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5955691763209754,
+      "learning_rate": 4.280549554279223e-07,
+      "loss": 0.8723,
+      "step": 10860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5728529398262352,
+      "learning_rate": 4.27529124696513e-07,
+      "loss": 0.7514,
+      "step": 10861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.473266208648008,
+      "learning_rate": 4.270036027087915e-07,
+      "loss": 0.6887,
+      "step": 10862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4987134534024584,
+      "learning_rate": 4.2647838950024445e-07,
+      "loss": 0.8458,
+      "step": 10863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6637440350760873,
+      "learning_rate": 4.259534851063346e-07,
+      "loss": 0.7062,
+      "step": 10864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6371718619071305,
+      "learning_rate": 4.2542888956250475e-07,
+      "loss": 0.7437,
+      "step": 10865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.505095516169909,
+      "learning_rate": 4.2490460290417645e-07,
+      "loss": 0.7983,
+      "step": 10866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4874924709777637,
+      "learning_rate": 4.243806251667509e-07,
+      "loss": 0.7013,
+      "step": 10867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.414192099784101,
+      "learning_rate": 4.2385695638560874e-07,
+      "loss": 0.6876,
+      "step": 10868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5367233329125694,
+      "learning_rate": 4.2333359659610715e-07,
+      "loss": 0.8357,
+      "step": 10869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5259162089539835,
+      "learning_rate": 4.228105458335879e-07,
+      "loss": 0.7668,
+      "step": 10870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7509049937396648,
+      "learning_rate": 4.222878041333672e-07,
+      "loss": 1.0728,
+      "step": 10871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4850529327171975,
+      "learning_rate": 4.2176537153074014e-07,
+      "loss": 0.7803,
+      "step": 10872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5081266063100325,
+      "learning_rate": 4.212432480609846e-07,
+      "loss": 0.729,
+      "step": 10873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.800014874269248,
+      "learning_rate": 4.207214337593557e-07,
+      "loss": 0.7534,
+      "step": 10874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5796778189469272,
+      "learning_rate": 4.2019992866108637e-07,
+      "loss": 0.6964,
+      "step": 10875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4199688347483326,
+      "learning_rate": 4.1967873280139017e-07,
+      "loss": 0.7706,
+      "step": 10876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4414603757546947,
+      "learning_rate": 4.191578462154594e-07,
+      "loss": 0.6974,
+      "step": 10877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.531750112399289,
+      "learning_rate": 4.186372689384655e-07,
+      "loss": 0.832,
+      "step": 10878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6257890063176823,
+      "learning_rate": 4.181170010055585e-07,
+      "loss": 0.7619,
+      "step": 10879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5189046642418236,
+      "learning_rate": 4.1759704245186936e-07,
+      "loss": 0.6893,
+      "step": 10880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5925069284966133,
+      "learning_rate": 4.170773933125061e-07,
+      "loss": 0.8033,
+      "step": 10881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5804564878128753,
+      "learning_rate": 4.165580536225561e-07,
+      "loss": 0.7841,
+      "step": 10882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5997023754792323,
+      "learning_rate": 4.1603902341708804e-07,
+      "loss": 0.7178,
+      "step": 10883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4802467174291711,
+      "learning_rate": 4.1552030273114665e-07,
+      "loss": 0.7551,
+      "step": 10884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7621645153175403,
+      "learning_rate": 4.150018915997578e-07,
+      "loss": 0.8266,
+      "step": 10885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4672942145649917,
+      "learning_rate": 4.1448379005792517e-07,
+      "loss": 0.671,
+      "step": 10886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4908575742250816,
+      "learning_rate": 4.1396599814063244e-07,
+      "loss": 0.6795,
+      "step": 10887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7591012438197998,
+      "learning_rate": 4.1344851588284216e-07,
+      "loss": 1.1026,
+      "step": 10888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4849777900195205,
+      "learning_rate": 4.129313433194948e-07,
+      "loss": 0.6946,
+      "step": 10889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5730539321805446,
+      "learning_rate": 4.124144804855135e-07,
+      "loss": 0.7879,
+      "step": 10890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6567694862738451,
+      "learning_rate": 4.118979274157964e-07,
+      "loss": 0.7686,
+      "step": 10891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5752157648285017,
+      "learning_rate": 4.113816841452223e-07,
+      "loss": 0.7535,
+      "step": 10892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8718092463510594,
+      "learning_rate": 4.1086575070864885e-07,
+      "loss": 0.7495,
+      "step": 10893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4919671070474896,
+      "learning_rate": 4.1035012714091436e-07,
+      "loss": 0.7357,
+      "step": 10894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3918360649688855,
+      "learning_rate": 4.098348134768343e-07,
+      "loss": 0.6677,
+      "step": 10895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7337175479942323,
+      "learning_rate": 4.0931980975120246e-07,
+      "loss": 1.0458,
+      "step": 10896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5823529764943791,
+      "learning_rate": 4.0880511599879545e-07,
+      "loss": 0.8521,
+      "step": 10897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.53303764219004,
+      "learning_rate": 4.0829073225436613e-07,
+      "loss": 0.7706,
+      "step": 10898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4716144815591148,
+      "learning_rate": 4.077766585526444e-07,
+      "loss": 0.6956,
+      "step": 10899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5956499949241874,
+      "learning_rate": 4.072628949283447e-07,
+      "loss": 0.7485,
+      "step": 10900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5238950269877727,
+      "learning_rate": 4.067494414161555e-07,
+      "loss": 0.7841,
+      "step": 10901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4876642918472216,
+      "learning_rate": 4.0623629805074784e-07,
+      "loss": 0.6775,
+      "step": 10902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4149890774524612,
+      "learning_rate": 4.057234648667685e-07,
+      "loss": 0.6657,
+      "step": 10903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.452399398702419,
+      "learning_rate": 4.05210941898847e-07,
+      "loss": 0.7044,
+      "step": 10904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3609363650430728,
+      "learning_rate": 4.046987291815896e-07,
+      "loss": 0.7823,
+      "step": 10905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4273855151067214,
+      "learning_rate": 4.0418682674958074e-07,
+      "loss": 0.7584,
+      "step": 10906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.528420336150421,
+      "learning_rate": 4.036752346373868e-07,
+      "loss": 0.7856,
+      "step": 10907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4910433458487766,
+      "learning_rate": 4.0316395287955166e-07,
+      "loss": 0.7972,
+      "step": 10908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6430989164408867,
+      "learning_rate": 4.0265298151059785e-07,
+      "loss": 0.7386,
+      "step": 10909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7353311932513732,
+      "learning_rate": 4.0214232056502653e-07,
+      "loss": 1.0638,
+      "step": 10910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4727645104989497,
+      "learning_rate": 4.016319700773197e-07,
+      "loss": 0.7766,
+      "step": 10911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.440976912736856,
+      "learning_rate": 4.0112193008193746e-07,
+      "loss": 0.7371,
+      "step": 10912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7377174199134139,
+      "learning_rate": 4.006122006133173e-07,
+      "loss": 1.0355,
+      "step": 10913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.410285765460856,
+      "learning_rate": 4.001027817058789e-07,
+      "loss": 0.77,
+      "step": 10914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5743927414276078,
+      "learning_rate": 3.995936733940198e-07,
+      "loss": 0.7924,
+      "step": 10915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4891792665549142,
+      "learning_rate": 3.9908487571211463e-07,
+      "loss": 0.742,
+      "step": 10916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5335841003067048,
+      "learning_rate": 3.985763886945188e-07,
+      "loss": 0.7066,
+      "step": 10917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5586759305478077,
+      "learning_rate": 3.9806821237556805e-07,
+      "loss": 0.6873,
+      "step": 10918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.561460467590906,
+      "learning_rate": 3.975603467895739e-07,
+      "loss": 0.7813,
+      "step": 10919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7658741678068239,
+      "learning_rate": 3.9705279197083003e-07,
+      "loss": 1.0692,
+      "step": 10920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.498333711543909,
+      "learning_rate": 3.9654554795360624e-07,
+      "loss": 0.7132,
+      "step": 10921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4639315271740825,
+      "learning_rate": 3.96038614772154e-07,
+      "loss": 0.7003,
+      "step": 10922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6184995695559106,
+      "learning_rate": 3.95531992460701e-07,
+      "loss": 0.8107,
+      "step": 10923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6549879459964296,
+      "learning_rate": 3.9502568105345753e-07,
+      "loss": 0.8111,
+      "step": 10924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5087869999785404,
+      "learning_rate": 3.9451968058460967e-07,
+      "loss": 0.7518,
+      "step": 10925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4931849029850557,
+      "learning_rate": 3.940139910883245e-07,
+      "loss": 0.8085,
+      "step": 10926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.439156383846267,
+      "learning_rate": 3.9350861259874586e-07,
+      "loss": 0.7111,
+      "step": 10927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5367679471867846,
+      "learning_rate": 3.930035451499997e-07,
+      "loss": 0.8175,
+      "step": 10928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4523201130013537,
+      "learning_rate": 3.9249878877618886e-07,
+      "loss": 0.6872,
+      "step": 10929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4068640721745465,
+      "learning_rate": 3.9199434351139544e-07,
+      "loss": 0.6547,
+      "step": 10930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5219191792677558,
+      "learning_rate": 3.914902093896811e-07,
+      "loss": 0.6802,
+      "step": 10931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5915826367163501,
+      "learning_rate": 3.909863864450852e-07,
+      "loss": 0.7386,
+      "step": 10932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7537331106421039,
+      "learning_rate": 3.9048287471162847e-07,
+      "loss": 1.0912,
+      "step": 10933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7019205329750964,
+      "learning_rate": 3.8997967422330693e-07,
+      "loss": 0.763,
+      "step": 10934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5220772044557014,
+      "learning_rate": 3.8947678501410014e-07,
+      "loss": 0.8192,
+      "step": 10935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5628237485440022,
+      "learning_rate": 3.889742071179636e-07,
+      "loss": 0.7386,
+      "step": 10936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7524469453717851,
+      "learning_rate": 3.884719405688314e-07,
+      "loss": 1.082,
+      "step": 10937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4110842848276386,
+      "learning_rate": 3.8796998540061916e-07,
+      "loss": 0.8143,
+      "step": 10938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.485213912445654,
+      "learning_rate": 3.8746834164722024e-07,
+      "loss": 0.7806,
+      "step": 10939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.879718417042932,
+      "learning_rate": 3.8696700934250485e-07,
+      "loss": 0.7611,
+      "step": 10940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7606604894187893,
+      "learning_rate": 3.8646598852032593e-07,
+      "loss": 1.0472,
+      "step": 10941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5618685182129253,
+      "learning_rate": 3.859652792145141e-07,
+      "loss": 0.782,
+      "step": 10942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4350105166458642,
+      "learning_rate": 3.8546488145887627e-07,
+      "loss": 0.714,
+      "step": 10943
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6468784960142941,
+      "learning_rate": 3.84964795287201e-07,
+      "loss": 0.7932,
+      "step": 10944
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5339274857138525,
+      "learning_rate": 3.844650207332562e-07,
+      "loss": 0.7755,
+      "step": 10945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5113205656119701,
+      "learning_rate": 3.8396555783078717e-07,
+      "loss": 0.8319,
+      "step": 10946
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5860154363324779,
+      "learning_rate": 3.8346640661351795e-07,
+      "loss": 0.6949,
+      "step": 10947
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.696046099297465,
+      "learning_rate": 3.8296756711515446e-07,
+      "loss": 0.7767,
+      "step": 10948
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5666204426998542,
+      "learning_rate": 3.8246903936937806e-07,
+      "loss": 0.6784,
+      "step": 10949
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7625153174115986,
+      "learning_rate": 3.819708234098507e-07,
+      "loss": 1.0759,
+      "step": 10950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5965289860890333,
+      "learning_rate": 3.8147291927021213e-07,
+      "loss": 0.7501,
+      "step": 10951
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6654556425391605,
+      "learning_rate": 3.8097532698408436e-07,
+      "loss": 0.7991,
+      "step": 10952
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.650604801938321,
+      "learning_rate": 3.804780465850644e-07,
+      "loss": 0.7445,
+      "step": 10953
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5602761237847655,
+      "learning_rate": 3.799810781067298e-07,
+      "loss": 0.7677,
+      "step": 10954
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3925359297347504,
+      "learning_rate": 3.794844215826371e-07,
+      "loss": 0.7513,
+      "step": 10955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.576378133715264,
+      "learning_rate": 3.789880770463217e-07,
+      "loss": 0.7435,
+      "step": 10956
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5223247566696465,
+      "learning_rate": 3.784920445312978e-07,
+      "loss": 0.764,
+      "step": 10957
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4307736332204142,
+      "learning_rate": 3.779963240710577e-07,
+      "loss": 0.6782,
+      "step": 10958
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.334372188484022,
+      "learning_rate": 3.775009156990761e-07,
+      "loss": 0.7624,
+      "step": 10959
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4987633409644332,
+      "learning_rate": 3.7700581944880246e-07,
+      "loss": 0.7582,
+      "step": 10960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6890232753226628,
+      "learning_rate": 3.765110353536661e-07,
+      "loss": 0.8296,
+      "step": 10961
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5624747119192703,
+      "learning_rate": 3.7601656344707746e-07,
+      "loss": 0.6624,
+      "step": 10962
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4953135106282294,
+      "learning_rate": 3.755224037624239e-07,
+      "loss": 0.769,
+      "step": 10963
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4351060883818185,
+      "learning_rate": 3.7502855633307246e-07,
+      "loss": 0.7233,
+      "step": 10964
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4485235587462515,
+      "learning_rate": 3.745350211923682e-07,
+      "loss": 0.7421,
+      "step": 10965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4426987004814353,
+      "learning_rate": 3.7404179837363665e-07,
+      "loss": 0.7554,
+      "step": 10966
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.479405918577684,
+      "learning_rate": 3.735488879101801e-07,
+      "loss": 0.7301,
+      "step": 10967
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6216574133211057,
+      "learning_rate": 3.730562898352813e-07,
+      "loss": 0.7961,
+      "step": 10968
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.467189544336084,
+      "learning_rate": 3.725640041822026e-07,
+      "loss": 0.7265,
+      "step": 10969
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4876668454848354,
+      "learning_rate": 3.7207203098418354e-07,
+      "loss": 0.8603,
+      "step": 10970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5783907627603446,
+      "learning_rate": 3.715803702744425e-07,
+      "loss": 0.8183,
+      "step": 10971
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4770757276812514,
+      "learning_rate": 3.71089022086179e-07,
+      "loss": 0.7995,
+      "step": 10972
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6008629648848236,
+      "learning_rate": 3.7059798645256996e-07,
+      "loss": 0.7436,
+      "step": 10973
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5665279358286361,
+      "learning_rate": 3.701072634067704e-07,
+      "loss": 0.7398,
+      "step": 10974
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7503801235726272,
+      "learning_rate": 3.6961685298191496e-07,
+      "loss": 1.0382,
+      "step": 10975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5306359479740053,
+      "learning_rate": 3.691267552111183e-07,
+      "loss": 0.7382,
+      "step": 10976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5814265836627432,
+      "learning_rate": 3.686369701274717e-07,
+      "loss": 0.7531,
+      "step": 10977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6662224173174685,
+      "learning_rate": 3.681474977640465e-07,
+      "loss": 0.7672,
+      "step": 10978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6167283556678482,
+      "learning_rate": 3.676583381538945e-07,
+      "loss": 0.7509,
+      "step": 10979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5642568650651039,
+      "learning_rate": 3.671694913300439e-07,
+      "loss": 0.696,
+      "step": 10980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4840671086771284,
+      "learning_rate": 3.6668095732550203e-07,
+      "loss": 0.7116,
+      "step": 10981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7147326429796217,
+      "learning_rate": 3.66192736173257e-07,
+      "loss": 0.7208,
+      "step": 10982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.379189352853151,
+      "learning_rate": 3.6570482790627526e-07,
+      "loss": 0.7257,
+      "step": 10983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4131510046607434,
+      "learning_rate": 3.652172325574999e-07,
+      "loss": 0.747,
+      "step": 10984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5013898256092313,
+      "learning_rate": 3.647299501598539e-07,
+      "loss": 0.7528,
+      "step": 10985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6270174736009082,
+      "learning_rate": 3.6424298074624333e-07,
+      "loss": 0.7802,
+      "step": 10986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.502568192627465,
+      "learning_rate": 3.6375632434954564e-07,
+      "loss": 0.7705,
+      "step": 10987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.707811120973351,
+      "learning_rate": 3.6326998100262134e-07,
+      "loss": 0.8572,
+      "step": 10988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7374993424738078,
+      "learning_rate": 3.6278395073831183e-07,
+      "loss": 1.054,
+      "step": 10989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5533169856310978,
+      "learning_rate": 3.622982335894332e-07,
+      "loss": 0.698,
+      "step": 10990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7595724411641711,
+      "learning_rate": 3.6181282958878295e-07,
+      "loss": 1.0451,
+      "step": 10991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.444656803839999,
+      "learning_rate": 3.6132773876913495e-07,
+      "loss": 0.787,
+      "step": 10992
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6168321970799713,
+      "learning_rate": 3.6084296116324637e-07,
+      "loss": 0.7611,
+      "step": 10993
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5031100369159351,
+      "learning_rate": 3.603584968038487e-07,
+      "loss": 0.7079,
+      "step": 10994
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3672294757163161,
+      "learning_rate": 3.598743457236542e-07,
+      "loss": 0.6527,
+      "step": 10995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6075434588299722,
+      "learning_rate": 3.5939050795535446e-07,
+      "loss": 0.6865,
+      "step": 10996
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.588607134089166,
+      "learning_rate": 3.5890698353161947e-07,
+      "loss": 0.7268,
+      "step": 10997
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.471690486971903,
+      "learning_rate": 3.5842377248509696e-07,
+      "loss": 0.7552,
+      "step": 10998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9660411802613984,
+      "learning_rate": 3.5794087484841534e-07,
+      "loss": 0.7971,
+      "step": 10999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4468538421342834,
+      "learning_rate": 3.5745829065418013e-07,
+      "loss": 0.6409,
+      "step": 11000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6797742275037628,
+      "learning_rate": 3.56976019934977e-07,
+      "loss": 0.7181,
+      "step": 11001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.469969120821769,
+      "learning_rate": 3.5649406272336926e-07,
+      "loss": 0.7576,
+      "step": 11002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5435041712455972,
+      "learning_rate": 3.5601241905190153e-07,
+      "loss": 0.7762,
+      "step": 11003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.495842426979491,
+      "learning_rate": 3.5553108895309443e-07,
+      "loss": 0.7379,
+      "step": 11004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.569247196517556,
+      "learning_rate": 3.550500724594469e-07,
+      "loss": 0.8037,
+      "step": 11005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4236960188204681,
+      "learning_rate": 3.545693696034413e-07,
+      "loss": 0.6986,
+      "step": 11006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5359390912035282,
+      "learning_rate": 3.54088980417534e-07,
+      "loss": 0.7034,
+      "step": 11007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7647765252402849,
+      "learning_rate": 3.5360890493416235e-07,
+      "loss": 1.0737,
+      "step": 11008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4664208388494542,
+      "learning_rate": 3.53129143185742e-07,
+      "loss": 0.7341,
+      "step": 11009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6337413694314546,
+      "learning_rate": 3.5264969520466765e-07,
+      "loss": 0.7932,
+      "step": 11010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4586087315406375,
+      "learning_rate": 3.521705610233123e-07,
+      "loss": 0.7153,
+      "step": 11011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7353728430896286,
+      "learning_rate": 3.5169174067402833e-07,
+      "loss": 1.0817,
+      "step": 11012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7537308377300403,
+      "learning_rate": 3.5121323418914723e-07,
+      "loss": 1.065,
+      "step": 11013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4368479478626945,
+      "learning_rate": 3.5073504160097913e-07,
+      "loss": 0.5962,
+      "step": 11014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5013169349365483,
+      "learning_rate": 3.502571629418122e-07,
+      "loss": 0.7932,
+      "step": 11015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.479773614881406,
+      "learning_rate": 3.4977959824391285e-07,
+      "loss": 0.7826,
+      "step": 11016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7576735520904445,
+      "learning_rate": 3.4930234753952963e-07,
+      "loss": 1.0703,
+      "step": 11017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5335334592636083,
+      "learning_rate": 3.488254108608857e-07,
+      "loss": 0.7709,
+      "step": 11018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4557138408128887,
+      "learning_rate": 3.483487882401859e-07,
+      "loss": 0.6993,
+      "step": 11019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.488197374703026,
+      "learning_rate": 3.478724797096128e-07,
+      "loss": 0.8006,
+      "step": 11020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7559731160109061,
+      "learning_rate": 3.473964853013273e-07,
+      "loss": 1.08,
+      "step": 11021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6358992732156503,
+      "learning_rate": 3.4692080504746926e-07,
+      "loss": 0.816,
+      "step": 11022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7409542379487125,
+      "learning_rate": 3.464454389801597e-07,
+      "loss": 1.0722,
+      "step": 11023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.609956343512661,
+      "learning_rate": 3.4597038713149455e-07,
+      "loss": 0.7169,
+      "step": 11024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6205486501016622,
+      "learning_rate": 3.45495649533551e-07,
+      "loss": 0.731,
+      "step": 11025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5250345800515839,
+      "learning_rate": 3.450212262183833e-07,
+      "loss": 0.6936,
+      "step": 11026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4978815365001965,
+      "learning_rate": 3.4454711721802757e-07,
+      "loss": 0.748,
+      "step": 11027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4348511161941986,
+      "learning_rate": 3.44073322564496e-07,
+      "loss": 0.749,
+      "step": 11028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5315407208297407,
+      "learning_rate": 3.4359984228977907e-07,
+      "loss": 0.7341,
+      "step": 11029
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5563067478184318,
+      "learning_rate": 3.431266764258501e-07,
+      "loss": 0.7309,
+      "step": 11030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5835311062703263,
+      "learning_rate": 3.4265382500465574e-07,
+      "loss": 0.7172,
+      "step": 11031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.493435371298899,
+      "learning_rate": 3.421812880581249e-07,
+      "loss": 0.8348,
+      "step": 11032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5460081782634507,
+      "learning_rate": 3.417090656181632e-07,
+      "loss": 0.7702,
+      "step": 11033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4394257856035668,
+      "learning_rate": 3.4123715771665786e-07,
+      "loss": 0.7922,
+      "step": 11034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5400473262501295,
+      "learning_rate": 3.4076556438547294e-07,
+      "loss": 0.7747,
+      "step": 11035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4704212777838803,
+      "learning_rate": 3.402942856564501e-07,
+      "loss": 0.7721,
+      "step": 11036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5288603626255577,
+      "learning_rate": 3.398233215614127e-07,
+      "loss": 0.7424,
+      "step": 11037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5455326262385387,
+      "learning_rate": 3.3935267213216163e-07,
+      "loss": 0.7528,
+      "step": 11038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6149503135211043,
+      "learning_rate": 3.388823374004735e-07,
+      "loss": 0.8393,
+      "step": 11039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6007372849225834,
+      "learning_rate": 3.384123173981096e-07,
+      "loss": 0.7704,
+      "step": 11040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5543947543635532,
+      "learning_rate": 3.3794261215680525e-07,
+      "loss": 0.7146,
+      "step": 11041
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5458531108575948,
+      "learning_rate": 3.3747322170827656e-07,
+      "loss": 0.7698,
+      "step": 11042
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5959077644225101,
+      "learning_rate": 3.370041460842166e-07,
+      "loss": 0.7278,
+      "step": 11043
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5741018908522728,
+      "learning_rate": 3.3653538531630006e-07,
+      "loss": 0.8237,
+      "step": 11044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4708693042316467,
+      "learning_rate": 3.3606693943617764e-07,
+      "loss": 0.7494,
+      "step": 11045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5567242926608835,
+      "learning_rate": 3.3559880847547965e-07,
+      "loss": 0.6624,
+      "step": 11046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5654550355716215,
+      "learning_rate": 3.3513099246581636e-07,
+      "loss": 0.7637,
+      "step": 11047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.600159050000962,
+      "learning_rate": 3.346634914387753e-07,
+      "loss": 0.8179,
+      "step": 11048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3134972523034971,
+      "learning_rate": 3.3419630542592343e-07,
+      "loss": 0.7866,
+      "step": 11049
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5256731107356623,
+      "learning_rate": 3.337294344588055e-07,
+      "loss": 0.7443,
+      "step": 11050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5369712477818505,
+      "learning_rate": 3.332628785689462e-07,
+      "loss": 0.7656,
+      "step": 11051
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.573666141803481,
+      "learning_rate": 3.327966377878494e-07,
+      "loss": 0.7967,
+      "step": 11052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.731535928574638,
+      "learning_rate": 3.323307121469954e-07,
+      "loss": 0.8713,
+      "step": 11053
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.63106374085155,
+      "learning_rate": 3.3186510167784456e-07,
+      "loss": 0.6757,
+      "step": 11054
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5974079207498195,
+      "learning_rate": 3.313998064118373e-07,
+      "loss": 0.7401,
+      "step": 11055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6680928299266247,
+      "learning_rate": 3.3093482638038963e-07,
+      "loss": 0.7336,
+      "step": 11056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5588846108935976,
+      "learning_rate": 3.3047016161489867e-07,
+      "loss": 0.7476,
+      "step": 11057
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6818346922757996,
+      "learning_rate": 3.3000581214674086e-07,
+      "loss": 0.7561,
+      "step": 11058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.526085003276602,
+      "learning_rate": 3.295417780072685e-07,
+      "loss": 0.7641,
+      "step": 11059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4415828415768006,
+      "learning_rate": 3.290780592278148e-07,
+      "loss": 0.7694,
+      "step": 11060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5435710536209635,
+      "learning_rate": 3.286146558396919e-07,
+      "loss": 0.7361,
+      "step": 11061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6377427950484176,
+      "learning_rate": 3.2815156787418925e-07,
+      "loss": 0.811,
+      "step": 11062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.473475459016356,
+      "learning_rate": 3.276887953625751e-07,
+      "loss": 0.7684,
+      "step": 11063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6634186712148324,
+      "learning_rate": 3.2722633833609797e-07,
+      "loss": 0.8335,
+      "step": 11064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5733499458681153,
+      "learning_rate": 3.2676419682598325e-07,
+      "loss": 0.7203,
+      "step": 11065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5540484411116942,
+      "learning_rate": 3.263023708634355e-07,
+      "loss": 0.6922,
+      "step": 11066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5180307903587187,
+      "learning_rate": 3.258408604796387e-07,
+      "loss": 0.7523,
+      "step": 11067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5702897044992534,
+      "learning_rate": 3.2537966570575566e-07,
+      "loss": 0.7209,
+      "step": 11068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4017798212283716,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.6724,
+      "step": 11069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5216823019588588,
+      "learning_rate": 3.2445822311227003e-07,
+      "loss": 0.8391,
+      "step": 11070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7630209103467722,
+      "learning_rate": 3.239979753548872e-07,
+      "loss": 1.0913,
+      "step": 11071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7363059981320395,
+      "learning_rate": 3.23538043331853e-07,
+      "loss": 1.0508,
+      "step": 11072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4410785088342541,
+      "learning_rate": 3.2307842707422324e-07,
+      "loss": 0.7282,
+      "step": 11073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5710786186475005,
+      "learning_rate": 3.2261912661303297e-07,
+      "loss": 0.727,
+      "step": 11074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4838948764394324,
+      "learning_rate": 3.2216014197929413e-07,
+      "loss": 0.761,
+      "step": 11075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5004303920154511,
+      "learning_rate": 3.217014732039997e-07,
+      "loss": 0.6929,
+      "step": 11076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7486107578822753,
+      "learning_rate": 3.2124312031811823e-07,
+      "loss": 1.0482,
+      "step": 11077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5033042459130572,
+      "learning_rate": 3.207850833526005e-07,
+      "loss": 0.7734,
+      "step": 11078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7652063197074335,
+      "learning_rate": 3.20327362338374e-07,
+      "loss": 1.0665,
+      "step": 11079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.49440311736683,
+      "learning_rate": 3.1986995730634404e-07,
+      "loss": 0.6626,
+      "step": 11080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4766376726770714,
+      "learning_rate": 3.194128682873965e-07,
+      "loss": 0.7188,
+      "step": 11081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5837729503858349,
+      "learning_rate": 3.1895609531239545e-07,
+      "loss": 0.7425,
+      "step": 11082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5691954533046482,
+      "learning_rate": 3.1849963841218244e-07,
+      "loss": 0.7766,
+      "step": 11083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.488387640694689,
+      "learning_rate": 3.180434976175784e-07,
+      "loss": 0.7393,
+      "step": 11084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.45229830219342,
+      "learning_rate": 3.1758767295938356e-07,
+      "loss": 0.7402,
+      "step": 11085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.502770014179707,
+      "learning_rate": 3.1713216446837613e-07,
+      "loss": 0.7468,
+      "step": 11086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.607685596364651,
+      "learning_rate": 3.1667697217531324e-07,
+      "loss": 0.8521,
+      "step": 11087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5305685294045883,
+      "learning_rate": 3.1622209611093023e-07,
+      "loss": 0.7437,
+      "step": 11088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7631086789271833,
+      "learning_rate": 3.157675363059409e-07,
+      "loss": 1.0755,
+      "step": 11089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4330018539225748,
+      "learning_rate": 3.1531329279103905e-07,
+      "loss": 0.6925,
+      "step": 11090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4965084725542455,
+      "learning_rate": 3.148593655968951e-07,
+      "loss": 0.8227,
+      "step": 11091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7495933806925165,
+      "learning_rate": 3.144057547541607e-07,
+      "loss": 1.0675,
+      "step": 11092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7535336703476626,
+      "learning_rate": 3.1395246029346405e-07,
+      "loss": 1.0574,
+      "step": 11093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7542244074805167,
+      "learning_rate": 3.1349948224541183e-07,
+      "loss": 1.0621,
+      "step": 11094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7613576971654857,
+      "learning_rate": 3.1304682064059177e-07,
+      "loss": 1.097,
+      "step": 11095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6207985871307777,
+      "learning_rate": 3.1259447550956777e-07,
+      "loss": 0.7086,
+      "step": 11096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4210154238674653,
+      "learning_rate": 3.1214244688288263e-07,
+      "loss": 0.7794,
+      "step": 11097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.459571018567844,
+      "learning_rate": 3.116907347910597e-07,
+      "loss": 0.7719,
+      "step": 11098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.572077166043891,
+      "learning_rate": 3.112393392645985e-07,
+      "loss": 0.7646,
+      "step": 11099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5066274261242514,
+      "learning_rate": 3.1078826033397845e-07,
+      "loss": 0.6904,
+      "step": 11100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7226837222371575,
+      "learning_rate": 3.1033749802965694e-07,
+      "loss": 1.0536,
+      "step": 11101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5414470772811537,
+      "learning_rate": 3.098870523820718e-07,
+      "loss": 0.7056,
+      "step": 11102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5043675942812087,
+      "learning_rate": 3.094369234216371e-07,
+      "loss": 0.7682,
+      "step": 11103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4754611139138323,
+      "learning_rate": 3.089871111787468e-07,
+      "loss": 0.788,
+      "step": 11104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5153370976574505,
+      "learning_rate": 3.085376156837738e-07,
+      "loss": 0.7751,
+      "step": 11105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5970161107419645,
+      "learning_rate": 3.080884369670689e-07,
+      "loss": 0.8021,
+      "step": 11106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8556855385147228,
+      "learning_rate": 3.076395750589617e-07,
+      "loss": 0.761,
+      "step": 11107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7702770675156423,
+      "learning_rate": 3.071910299897596e-07,
+      "loss": 1.0587,
+      "step": 11108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4804488829888518,
+      "learning_rate": 3.0674280178975e-07,
+      "loss": 0.7301,
+      "step": 11109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4379106700339863,
+      "learning_rate": 3.062948904891988e-07,
+      "loss": 0.749,
+      "step": 11110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4674472648119166,
+      "learning_rate": 3.0584729611834785e-07,
+      "loss": 0.7531,
+      "step": 11111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6081638999178967,
+      "learning_rate": 3.054000187074224e-07,
+      "loss": 0.8096,
+      "step": 11112
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.656772717771923,
+      "learning_rate": 3.049530582866228e-07,
+      "loss": 0.7817,
+      "step": 11113
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4965200978167352,
+      "learning_rate": 3.045064148861282e-07,
+      "loss": 0.6786,
+      "step": 11114
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4442389330689,
+      "learning_rate": 3.040600885360967e-07,
+      "loss": 0.7003,
+      "step": 11115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5909026499593781,
+      "learning_rate": 3.0361407926666644e-07,
+      "loss": 0.7411,
+      "step": 11116
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6257001686865984,
+      "learning_rate": 3.031683871079527e-07,
+      "loss": 0.6991,
+      "step": 11117
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5235540981042892,
+      "learning_rate": 3.0272301209004873e-07,
+      "loss": 0.8388,
+      "step": 11118
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5234892219202067,
+      "learning_rate": 3.022779542430293e-07,
+      "loss": 0.7576,
+      "step": 11119
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5618575996523274,
+      "learning_rate": 3.018332135969443e-07,
+      "loss": 0.7849,
+      "step": 11120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5011259729150346,
+      "learning_rate": 3.0138879018182243e-07,
+      "loss": 0.7075,
+      "step": 11121
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5634672964473333,
+      "learning_rate": 3.009446840276742e-07,
+      "loss": 0.8107,
+      "step": 11122
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5315534292063606,
+      "learning_rate": 3.0050089516448553e-07,
+      "loss": 0.7655,
+      "step": 11123
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5288700280827774,
+      "learning_rate": 3.000574236222231e-07,
+      "loss": 0.7594,
+      "step": 11124
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9538907399958063,
+      "learning_rate": 2.996142694308296e-07,
+      "loss": 0.7729,
+      "step": 11125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4225168814962479,
+      "learning_rate": 2.991714326202294e-07,
+      "loss": 0.6651,
+      "step": 11126
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.518797373285312,
+      "learning_rate": 2.9872891322032307e-07,
+      "loss": 0.7757,
+      "step": 11127
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4442787857491621,
+      "learning_rate": 2.9828671126098995e-07,
+      "loss": 0.7213,
+      "step": 11128
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5138077300199504,
+      "learning_rate": 2.9784482677209013e-07,
+      "loss": 0.7866,
+      "step": 11129
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7488686022148093,
+      "learning_rate": 2.9740325978345976e-07,
+      "loss": 1.0785,
+      "step": 11130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.52477985990948,
+      "learning_rate": 2.9696201032491434e-07,
+      "loss": 0.6429,
+      "step": 11131
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.529601856509054,
+      "learning_rate": 2.965210784262479e-07,
+      "loss": 0.7249,
+      "step": 11132
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5025061182541974,
+      "learning_rate": 2.9608046411723325e-07,
+      "loss": 0.7722,
+      "step": 11133
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6472983172781575,
+      "learning_rate": 2.9564016742762214e-07,
+      "loss": 0.7476,
+      "step": 11134
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5268752727115371,
+      "learning_rate": 2.952001883871436e-07,
+      "loss": 0.7602,
+      "step": 11135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7540060255321654,
+      "learning_rate": 2.9476052702550716e-07,
+      "loss": 1.0449,
+      "step": 11136
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4449056837729604,
+      "learning_rate": 2.9432118337239857e-07,
+      "loss": 0.7427,
+      "step": 11137
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5268550208876228,
+      "learning_rate": 2.9388215745748347e-07,
+      "loss": 0.7237,
+      "step": 11138
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3448534174589557,
+      "learning_rate": 2.934434493104071e-07,
+      "loss": 0.7442,
+      "step": 11139
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6334527995966484,
+      "learning_rate": 2.9300505896079135e-07,
+      "loss": 0.8298,
+      "step": 11140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6124570887749639,
+      "learning_rate": 2.925669864382369e-07,
+      "loss": 0.7608,
+      "step": 11141
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6007905334160644,
+      "learning_rate": 2.92129231772324e-07,
+      "loss": 0.7522,
+      "step": 11142
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.538164758555258,
+      "learning_rate": 2.916917949926107e-07,
+      "loss": 0.7655,
+      "step": 11143
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6751777019348677,
+      "learning_rate": 2.912546761286333e-07,
+      "loss": 0.7735,
+      "step": 11144
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7538311398031248,
+      "learning_rate": 2.9081787520990665e-07,
+      "loss": 1.0829,
+      "step": 11145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4697033471218546,
+      "learning_rate": 2.90381392265926e-07,
+      "loss": 0.7049,
+      "step": 11146
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5704763312678693,
+      "learning_rate": 2.899452273261627e-07,
+      "loss": 0.7396,
+      "step": 11147
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4429982874019476,
+      "learning_rate": 2.895093804200683e-07,
+      "loss": 0.6887,
+      "step": 11148
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4562277318305354,
+      "learning_rate": 2.890738515770708e-07,
+      "loss": 0.7517,
+      "step": 11149
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7608110630876941,
+      "learning_rate": 2.8863864082657955e-07,
+      "loss": 1.0354,
+      "step": 11150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7670376629116403,
+      "learning_rate": 2.882037481979805e-07,
+      "loss": 1.1139,
+      "step": 11151
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6430535815584841,
+      "learning_rate": 2.8776917372063896e-07,
+      "loss": 0.7894,
+      "step": 11152
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7414853330520232,
+      "learning_rate": 2.8733491742389765e-07,
+      "loss": 0.749,
+      "step": 11153
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.559874876878309,
+      "learning_rate": 2.8690097933707863e-07,
+      "loss": 0.761,
+      "step": 11154
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5622344693841999,
+      "learning_rate": 2.8646735948948234e-07,
+      "loss": 0.7514,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.584200461198389,
+      "learning_rate": 2.8603405791038876e-07,
+      "loss": 0.8195,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7484765503962355,
+      "learning_rate": 2.856010746290544e-07,
+      "loss": 1.0632,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.736996704500388,
+      "learning_rate": 2.851684096747159e-07,
+      "loss": 1.0417,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5741538709505476,
+      "learning_rate": 2.847360630765866e-07,
+      "loss": 0.6575,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4600336146316644,
+      "learning_rate": 2.843040348638615e-07,
+      "loss": 0.7327,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5553370524861467,
+      "learning_rate": 2.8387232506571105e-07,
+      "loss": 0.6892,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5227563776069415,
+      "learning_rate": 2.834409337112842e-07,
+      "loss": 0.7674,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4364587937439275,
+      "learning_rate": 2.8300986082971214e-07,
+      "loss": 0.7126,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5044351871457913,
+      "learning_rate": 2.8257910645009935e-07,
+      "loss": 0.7957,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6691165460395936,
+      "learning_rate": 2.821486706015325e-07,
+      "loss": 0.7365,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4833382921410052,
+      "learning_rate": 2.817185533130751e-07,
+      "loss": 0.746,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7440988017640782,
+      "learning_rate": 2.812887546137705e-07,
+      "loss": 1.0457,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5031318516521872,
+      "learning_rate": 2.8085927453263883e-07,
+      "loss": 0.728,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6655269631777834,
+      "learning_rate": 2.804301130986797e-07,
+      "loss": 0.7151,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5812249830621519,
+      "learning_rate": 2.8000127034087165e-07,
+      "loss": 0.819,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7457674211927122,
+      "learning_rate": 2.795727462881709e-07,
+      "loss": 1.0107,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6202076514880337,
+      "learning_rate": 2.7914454096951206e-07,
+      "loss": 0.7564,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5114659472752774,
+      "learning_rate": 2.7871665441380814e-07,
+      "loss": 0.69,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6009334540339515,
+      "learning_rate": 2.7828908664995216e-07,
+      "loss": 0.7792,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8783175508850718,
+      "learning_rate": 2.7786183770681376e-07,
+      "loss": 0.7629,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5527537809739251,
+      "learning_rate": 2.774349076132421e-07,
+      "loss": 0.8582,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.538307224641372,
+      "learning_rate": 2.770082963980647e-07,
+      "loss": 0.7147,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6369990047400909,
+      "learning_rate": 2.7658200409008626e-07,
+      "loss": 0.828,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4729337445463648,
+      "learning_rate": 2.7615603071809103e-07,
+      "loss": 0.6735,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7659668188755265,
+      "learning_rate": 2.757303763108432e-07,
+      "loss": 1.061,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4427532812894908,
+      "learning_rate": 2.753050408970831e-07,
+      "loss": 0.7518,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.734512275662547,
+      "learning_rate": 2.748800245055305e-07,
+      "loss": 1.0607,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5379605408438441,
+      "learning_rate": 2.74455327164882e-07,
+      "loss": 0.6932,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.497355428426903,
+      "learning_rate": 2.7403094890381674e-07,
+      "loss": 0.7118,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4332489589790323,
+      "learning_rate": 2.7360688975098806e-07,
+      "loss": 0.7701,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7630941521841698,
+      "learning_rate": 2.7318314973502957e-07,
+      "loss": 1.032,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6217914698184144,
+      "learning_rate": 2.7275972888455347e-07,
+      "loss": 0.7633,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7442088758093761,
+      "learning_rate": 2.7233662722815024e-07,
+      "loss": 1.0731,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8329429703670623,
+      "learning_rate": 2.719138447943881e-07,
+      "loss": 0.7793,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.757490377558337,
+      "learning_rate": 2.714913816118142e-07,
+      "loss": 1.058,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5174357912285095,
+      "learning_rate": 2.7106923770895466e-07,
+      "loss": 0.8166,
+      "step": 11191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5485125485370086,
+      "learning_rate": 2.706474131143144e-07,
+      "loss": 0.7787,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4765659884508924,
+      "learning_rate": 2.7022590785637406e-07,
+      "loss": 0.7836,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6444051691627424,
+      "learning_rate": 2.698047219635963e-07,
+      "loss": 0.803,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4481311027530321,
+      "learning_rate": 2.693838554644196e-07,
+      "loss": 0.6756,
+      "step": 11195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5586522906935116,
+      "learning_rate": 2.689633083872628e-07,
+      "loss": 0.8328,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4078306578321635,
+      "learning_rate": 2.68543080760521e-07,
+      "loss": 0.7105,
+      "step": 11197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.577800511719139,
+      "learning_rate": 2.6812317261256995e-07,
+      "loss": 0.8326,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5548554409192066,
+      "learning_rate": 2.6770358397176233e-07,
+      "loss": 0.8341,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7661293905237565,
+      "learning_rate": 2.672843148664289e-07,
+      "loss": 1.0635,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7255393645150351,
+      "learning_rate": 2.668653653248815e-07,
+      "loss": 1.0493,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4968942774606921,
+      "learning_rate": 2.6644673537540746e-07,
+      "loss": 0.7704,
+      "step": 11202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.582346429737381,
+      "learning_rate": 2.66028425046273e-07,
+      "loss": 0.79,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4851103457608008,
+      "learning_rate": 2.656104343657251e-07,
+      "loss": 0.787,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6537778387379753,
+      "learning_rate": 2.6519276336198665e-07,
+      "loss": 0.827,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5327496712017237,
+      "learning_rate": 2.6477541206325896e-07,
+      "loss": 0.756,
+      "step": 11206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4514045570242222,
+      "learning_rate": 2.6435838049772346e-07,
+      "loss": 0.7441,
+      "step": 11207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.486791672045777,
+      "learning_rate": 2.6394166869353923e-07,
+      "loss": 0.7589,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6014507005606395,
+      "learning_rate": 2.6352527667884264e-07,
+      "loss": 0.7383,
+      "step": 11209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7855351634420427,
+      "learning_rate": 2.6310920448174957e-07,
+      "loss": 0.7637,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4557309067323967,
+      "learning_rate": 2.6269345213035536e-07,
+      "loss": 0.7544,
+      "step": 11211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.692361618884787,
+      "learning_rate": 2.622780196527314e-07,
+      "loss": 0.7621,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5314135517072656,
+      "learning_rate": 2.6186290707692907e-07,
+      "loss": 0.8127,
+      "step": 11213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5264408982331834,
+      "learning_rate": 2.614481144309772e-07,
+      "loss": 0.7276,
+      "step": 11214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5382460004775107,
+      "learning_rate": 2.6103364174288435e-07,
+      "loss": 0.6276,
+      "step": 11215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5183983600786577,
+      "learning_rate": 2.6061948904063663e-07,
+      "loss": 0.6778,
+      "step": 11216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7403674630537943,
+      "learning_rate": 2.602056563521976e-07,
+      "loss": 1.0477,
+      "step": 11217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6134141454191993,
+      "learning_rate": 2.597921437055112e-07,
+      "loss": 0.8577,
+      "step": 11218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5231508426191889,
+      "learning_rate": 2.5937895112849886e-07,
+      "loss": 0.7553,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5263139922591729,
+      "learning_rate": 2.5896607864905944e-07,
+      "loss": 0.7042,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7460763140995438,
+      "learning_rate": 2.585535262950717e-07,
+      "loss": 1.047,
+      "step": 11221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4540421339177418,
+      "learning_rate": 2.581412940943917e-07,
+      "loss": 0.7006,
+      "step": 11222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.496683974960104,
+      "learning_rate": 2.577293820748544e-07,
+      "loss": 0.7387,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.53687619552774,
+      "learning_rate": 2.573177902642726e-07,
+      "loss": 0.744,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7506978887917596,
+      "learning_rate": 2.569065186904385e-07,
+      "loss": 1.0293,
+      "step": 11225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5451994418085164,
+      "learning_rate": 2.564955673811226e-07,
+      "loss": 0.727,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4206439122441608,
+      "learning_rate": 2.560849363640716e-07,
+      "loss": 0.648,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3878093579943576,
+      "learning_rate": 2.556746256670145e-07,
+      "loss": 0.7749,
+      "step": 11228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4985749226422698,
+      "learning_rate": 2.5526463531765467e-07,
+      "loss": 0.6971,
+      "step": 11229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.483609937464334,
+      "learning_rate": 2.5485496534367657e-07,
+      "loss": 0.7335,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4554354859203376,
+      "learning_rate": 2.544456157727415e-07,
+      "loss": 0.7992,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4282278999025104,
+      "learning_rate": 2.5403658663248953e-07,
+      "loss": 0.7034,
+      "step": 11232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4945602828862585,
+      "learning_rate": 2.536278779505402e-07,
+      "loss": 0.7517,
+      "step": 11233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4317518934981917,
+      "learning_rate": 2.532194897544882e-07,
+      "loss": 0.7324,
+      "step": 11234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5545037796736385,
+      "learning_rate": 2.528114220719119e-07,
+      "loss": 0.7616,
+      "step": 11235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5307505978044706,
+      "learning_rate": 2.524036749303632e-07,
+      "loss": 0.7475,
+      "step": 11236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6197481615382159,
+      "learning_rate": 2.5199624835737345e-07,
+      "loss": 0.7125,
+      "step": 11237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5287344884858842,
+      "learning_rate": 2.5158914238045507e-07,
+      "loss": 0.7605,
+      "step": 11238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6447468489407668,
+      "learning_rate": 2.511823570270955e-07,
+      "loss": 0.7608,
+      "step": 11239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5135873557017792,
+      "learning_rate": 2.5077589232476217e-07,
+      "loss": 0.7674,
+      "step": 11240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5472361340390777,
+      "learning_rate": 2.503697483008999e-07,
+      "loss": 0.7194,
+      "step": 11241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5351681018963006,
+      "learning_rate": 2.4996392498293334e-07,
+      "loss": 0.7385,
+      "step": 11242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5142152721575817,
+      "learning_rate": 2.495584223982644e-07,
+      "loss": 0.778,
+      "step": 11243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5751234467732027,
+      "learning_rate": 2.491532405742719e-07,
+      "loss": 0.7701,
+      "step": 11244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5424047807534318,
+      "learning_rate": 2.4874837953831723e-07,
+      "loss": 0.8033,
+      "step": 11245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6466924936224756,
+      "learning_rate": 2.483438393177362e-07,
+      "loss": 0.7513,
+      "step": 11246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4497958496034145,
+      "learning_rate": 2.479396199398448e-07,
+      "loss": 0.7117,
+      "step": 11247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.609320427328318,
+      "learning_rate": 2.475357214319357e-07,
+      "loss": 0.7709,
+      "step": 11248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.51061343446991,
+      "learning_rate": 2.4713214382128204e-07,
+      "loss": 0.7699,
+      "step": 11249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5620083902440538,
+      "learning_rate": 2.4672888713513476e-07,
+      "loss": 0.756,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7337232870886723,
+      "learning_rate": 2.463259514007216e-07,
+      "loss": 1.0703,
+      "step": 11251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4620361613376,
+      "learning_rate": 2.459233366452507e-07,
+      "loss": 0.7589,
+      "step": 11252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6237781717899087,
+      "learning_rate": 2.455210428959065e-07,
+      "loss": 0.7129,
+      "step": 11253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5225875120319627,
+      "learning_rate": 2.451190701798523e-07,
+      "loss": 0.7526,
+      "step": 11254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4723323202050016,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.8148,
+      "step": 11255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.504256939558105,
+      "learning_rate": 2.443160879561657e-07,
+      "loss": 0.743,
+      "step": 11256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6387728527545875,
+      "learning_rate": 2.4391507850275166e-07,
+      "loss": 0.7532,
+      "step": 11257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5247521225676715,
+      "learning_rate": 2.4351439019106584e-07,
+      "loss": 0.7555,
+      "step": 11258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5244893211370647,
+      "learning_rate": 2.4311402304816546e-07,
+      "loss": 0.8505,
+      "step": 11259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7558001718713542,
+      "learning_rate": 2.427139771010839e-07,
+      "loss": 1.0383,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5075921059561654,
+      "learning_rate": 2.423142523768318e-07,
+      "loss": 0.7504,
+      "step": 11261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.412525831422138,
+      "learning_rate": 2.4191484890240093e-07,
+      "loss": 0.7564,
+      "step": 11262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7437306126323908,
+      "learning_rate": 2.4151576670476016e-07,
+      "loss": 1.0573,
+      "step": 11263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5577401122491874,
+      "learning_rate": 2.411170058108558e-07,
+      "loss": 0.7351,
+      "step": 11264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4010333689266676,
+      "learning_rate": 2.407185662476119e-07,
+      "loss": 0.6593,
+      "step": 11265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4899102286359993,
+      "learning_rate": 2.403204480419341e-07,
+      "loss": 0.7949,
+      "step": 11266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7572016066556934,
+      "learning_rate": 2.3992265122070314e-07,
+      "loss": 1.0539,
+      "step": 11267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.242723894728857,
+      "learning_rate": 2.395251758107786e-07,
+      "loss": 0.8296,
+      "step": 11268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6297353910050398,
+      "learning_rate": 2.391280218390002e-07,
+      "loss": 0.7051,
+      "step": 11269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6118987226141208,
+      "learning_rate": 2.387311893321842e-07,
+      "loss": 0.6917,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4844342512957134,
+      "learning_rate": 2.3833467831712587e-07,
+      "loss": 0.7423,
+      "step": 11271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6988900950856047,
+      "learning_rate": 2.3793848882059768e-07,
+      "loss": 0.8332,
+      "step": 11272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5450056836410146,
+      "learning_rate": 2.375426208693521e-07,
+      "loss": 0.8017,
+      "step": 11273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7406808298954016,
+      "learning_rate": 2.3714707449011886e-07,
+      "loss": 1.0831,
+      "step": 11274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5965278225440347,
+      "learning_rate": 2.3675184970960607e-07,
+      "loss": 0.7352,
+      "step": 11275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.597125082974124,
+      "learning_rate": 2.363569465545007e-07,
+      "loss": 0.6943,
+      "step": 11276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4499118102016848,
+      "learning_rate": 2.3596236505146642e-07,
+      "loss": 0.7126,
+      "step": 11277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7630486501097397,
+      "learning_rate": 2.3556810522714636e-07,
+      "loss": 1.0552,
+      "step": 11278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5250295568857868,
+      "learning_rate": 2.351741671081631e-07,
+      "loss": 0.7611,
+      "step": 11279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4796313149393943,
+      "learning_rate": 2.3478055072111538e-07,
+      "loss": 0.7482,
+      "step": 11280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.741368244625926,
+      "learning_rate": 2.3438725609258138e-07,
+      "loss": 1.0596,
+      "step": 11281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4945696793466325,
+      "learning_rate": 2.3399428324911654e-07,
+      "loss": 0.7183,
+      "step": 11282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4549555749028342,
+      "learning_rate": 2.336016322172563e-07,
+      "loss": 0.724,
+      "step": 11283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6178294908945854,
+      "learning_rate": 2.332093030235133e-07,
+      "loss": 0.7177,
+      "step": 11284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5218919229653736,
+      "learning_rate": 2.328172956943775e-07,
+      "loss": 0.7362,
+      "step": 11285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.465339177273194,
+      "learning_rate": 2.3242561025631882e-07,
+      "loss": 0.7268,
+      "step": 11286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5309234880549285,
+      "learning_rate": 2.32034246735785e-07,
+      "loss": 0.8257,
+      "step": 11287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.455840489393032,
+      "learning_rate": 2.3164320515920101e-07,
+      "loss": 0.7587,
+      "step": 11288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6622149837574556,
+      "learning_rate": 2.3125248555297074e-07,
+      "loss": 0.831,
+      "step": 11289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5272210281184677,
+      "learning_rate": 2.308620879434781e-07,
+      "loss": 0.7808,
+      "step": 11290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7286223045301974,
+      "learning_rate": 2.3047201235708195e-07,
+      "loss": 1.065,
+      "step": 11291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8661284347836218,
+      "learning_rate": 2.3008225882012125e-07,
+      "loss": 0.7071,
+      "step": 11292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6061117268080947,
+      "learning_rate": 2.296928273589144e-07,
+      "loss": 0.7859,
+      "step": 11293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4078803341145592,
+      "learning_rate": 2.2930371799975593e-07,
+      "loss": 0.7951,
+      "step": 11294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0831678717499917,
+      "learning_rate": 2.2891493076891924e-07,
+      "loss": 0.6557,
+      "step": 11295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5118646676983432,
+      "learning_rate": 2.2852646569265556e-07,
+      "loss": 0.6535,
+      "step": 11296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7497925286566581,
+      "learning_rate": 2.2813832279719615e-07,
+      "loss": 1.0748,
+      "step": 11297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5103118472819625,
+      "learning_rate": 2.277505021087484e-07,
+      "loss": 0.7435,
+      "step": 11298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7518153876725854,
+      "learning_rate": 2.2736300365349905e-07,
+      "loss": 1.0412,
+      "step": 11299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5363083465712088,
+      "learning_rate": 2.2697582745761282e-07,
+      "loss": 0.6852,
+      "step": 11300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4770625130362167,
+      "learning_rate": 2.2658897354723373e-07,
+      "loss": 0.7149,
+      "step": 11301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5868537327168775,
+      "learning_rate": 2.2620244194848096e-07,
+      "loss": 0.756,
+      "step": 11302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7789083676839663,
+      "learning_rate": 2.258162326874558e-07,
+      "loss": 1.0448,
+      "step": 11303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5333064982832316,
+      "learning_rate": 2.2543034579023572e-07,
+      "loss": 0.8183,
+      "step": 11304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3206393528267721,
+      "learning_rate": 2.2504478128287654e-07,
+      "loss": 0.7324,
+      "step": 11305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7549097864176597,
+      "learning_rate": 2.2465953919141136e-07,
+      "loss": 1.0371,
+      "step": 11306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4460452026287987,
+      "learning_rate": 2.2427461954185493e-07,
+      "loss": 0.7151,
+      "step": 11307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.446209712062776,
+      "learning_rate": 2.2389002236019642e-07,
+      "loss": 0.6694,
+      "step": 11308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8449757154346536,
+      "learning_rate": 2.2350574767240395e-07,
+      "loss": 0.7062,
+      "step": 11309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3944072738625088,
+      "learning_rate": 2.231217955044257e-07,
+      "loss": 0.7548,
+      "step": 11310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.742147293239935,
+      "learning_rate": 2.22738165882187e-07,
+      "loss": 1.0695,
+      "step": 11311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4312518944682713,
+      "learning_rate": 2.2235485883159159e-07,
+      "loss": 0.7088,
+      "step": 11312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.544638094897233,
+      "learning_rate": 2.2197187437851985e-07,
+      "loss": 0.7522,
+      "step": 11313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4267707711497692,
+      "learning_rate": 2.2158921254883337e-07,
+      "loss": 0.6941,
+      "step": 11314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6184482967929608,
+      "learning_rate": 2.2120687336837033e-07,
+      "loss": 0.7202,
+      "step": 11315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7575072789247222,
+      "learning_rate": 2.2082485686294507e-07,
+      "loss": 1.0774,
+      "step": 11316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7333684617798224,
+      "learning_rate": 2.204431630583548e-07,
+      "loss": 1.0486,
+      "step": 11317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7364018148926658,
+      "learning_rate": 2.200617919803716e-07,
+      "loss": 1.0358,
+      "step": 11318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5036891181160223,
+      "learning_rate": 2.1968074365474544e-07,
+      "loss": 0.7823,
+      "step": 11319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5531406433195596,
+      "learning_rate": 2.1930001810720692e-07,
+      "loss": 0.6739,
+      "step": 11320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8497853453226605,
+      "learning_rate": 2.1891961536346262e-07,
+      "loss": 0.7879,
+      "step": 11321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6778807227948678,
+      "learning_rate": 2.185395354491987e-07,
+      "loss": 0.8188,
+      "step": 11322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.514165802985255,
+      "learning_rate": 2.1815977839007795e-07,
+      "loss": 0.735,
+      "step": 11323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5664206309518336,
+      "learning_rate": 2.1778034421174433e-07,
+      "loss": 0.8323,
+      "step": 11324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7606371933012575,
+      "learning_rate": 2.1740123293981675e-07,
+      "loss": 1.0451,
+      "step": 11325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4856853089915207,
+      "learning_rate": 2.1702244459989308e-07,
+      "loss": 0.7862,
+      "step": 11326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5329465777044016,
+      "learning_rate": 2.166439792175523e-07,
+      "loss": 0.795,
+      "step": 11327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.513790992225619,
+      "learning_rate": 2.1626583681834733e-07,
+      "loss": 0.6951,
+      "step": 11328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6075026525574054,
+      "learning_rate": 2.1588801742781163e-07,
+      "loss": 0.7723,
+      "step": 11329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7890698689826547,
+      "learning_rate": 2.1551052107145698e-07,
+      "loss": 0.7465,
+      "step": 11330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7722807042340272,
+      "learning_rate": 2.1513334777477192e-07,
+      "loss": 1.0566,
+      "step": 11331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5260360661567334,
+      "learning_rate": 2.1475649756322436e-07,
+      "loss": 0.7765,
+      "step": 11332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7636934219662053,
+      "learning_rate": 2.1437997046226012e-07,
+      "loss": 1.064,
+      "step": 11333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.594049569361944,
+      "learning_rate": 2.140037664973038e-07,
+      "loss": 0.7953,
+      "step": 11334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5307857207867879,
+      "learning_rate": 2.1362788569375682e-07,
+      "loss": 0.809,
+      "step": 11335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6095733688955813,
+      "learning_rate": 2.132523280769988e-07,
+      "loss": 0.8303,
+      "step": 11336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5315309849958798,
+      "learning_rate": 2.1287709367239008e-07,
+      "loss": 0.7319,
+      "step": 11337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7404872477536104,
+      "learning_rate": 2.1250218250526643e-07,
+      "loss": 1.0412,
+      "step": 11338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5452586403658264,
+      "learning_rate": 2.1212759460094268e-07,
+      "loss": 0.6767,
+      "step": 11339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.632780722184232,
+      "learning_rate": 2.1175332998471189e-07,
+      "loss": 0.7774,
+      "step": 11340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7470833439375405,
+      "learning_rate": 2.1137938868184493e-07,
+      "loss": 1.0894,
+      "step": 11341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4998593138579897,
+      "learning_rate": 2.1100577071759164e-07,
+      "loss": 0.7136,
+      "step": 11342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.602828556297149,
+      "learning_rate": 2.1063247611717908e-07,
+      "loss": 0.7868,
+      "step": 11343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.527764059481734,
+      "learning_rate": 2.102595049058137e-07,
+      "loss": 0.639,
+      "step": 11344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4891679578059003,
+      "learning_rate": 2.0988685710867874e-07,
+      "loss": 0.7705,
+      "step": 11345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7288237016125884,
+      "learning_rate": 2.095145327509368e-07,
+      "loss": 1.0536,
+      "step": 11346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4776278588501102,
+      "learning_rate": 2.0914253185772727e-07,
+      "loss": 0.7539,
+      "step": 11347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7550206097661549,
+      "learning_rate": 2.0877085445416889e-07,
+      "loss": 1.0514,
+      "step": 11348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6751800895574427,
+      "learning_rate": 2.0839950056535884e-07,
+      "loss": 0.7304,
+      "step": 11349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4781372518518667,
+      "learning_rate": 2.080284702163704e-07,
+      "loss": 0.73,
+      "step": 11350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4821034265889417,
+      "learning_rate": 2.076577634322574e-07,
+      "loss": 0.7404,
+      "step": 11351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.570714681835568,
+      "learning_rate": 2.072873802380515e-07,
+      "loss": 0.7397,
+      "step": 11352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.409165223593943,
+      "learning_rate": 2.069173206587599e-07,
+      "loss": 0.7792,
+      "step": 11353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4938012946881862,
+      "learning_rate": 2.0654758471937098e-07,
+      "loss": 0.7804,
+      "step": 11354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9595966796617497,
+      "learning_rate": 2.0617817244485027e-07,
+      "loss": 0.6948,
+      "step": 11355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4792613334112894,
+      "learning_rate": 2.058090838601412e-07,
+      "loss": 0.6942,
+      "step": 11356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3962570996578123,
+      "learning_rate": 2.0544031899016437e-07,
+      "loss": 0.7095,
+      "step": 11357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5848901259174668,
+      "learning_rate": 2.0507187785982153e-07,
+      "loss": 0.7265,
+      "step": 11358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7531521104197739,
+      "learning_rate": 2.0470376049398944e-07,
+      "loss": 1.0795,
+      "step": 11359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4841083065028131,
+      "learning_rate": 2.0433596691752432e-07,
+      "loss": 0.7602,
+      "step": 11360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4704723805143989,
+      "learning_rate": 2.0396849715526134e-07,
+      "loss": 0.7574,
+      "step": 11361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7379352724767565,
+      "learning_rate": 2.0360135123201175e-07,
+      "loss": 1.0383,
+      "step": 11362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.513674713153626,
+      "learning_rate": 2.0323452917256736e-07,
+      "loss": 0.7532,
+      "step": 11363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5558452918709833,
+      "learning_rate": 2.0286803100169507e-07,
+      "loss": 0.7105,
+      "step": 11364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.515950691166428,
+      "learning_rate": 2.0250185674414336e-07,
+      "loss": 0.6443,
+      "step": 11365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5660753671523235,
+      "learning_rate": 2.0213600642463583e-07,
+      "loss": 0.7562,
+      "step": 11366
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6262792025691646,
+      "learning_rate": 2.0177048006787604e-07,
+      "loss": 0.8486,
+      "step": 11367
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.567495656211438,
+      "learning_rate": 2.014052776985459e-07,
+      "loss": 0.7563,
+      "step": 11368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.548256727096305,
+      "learning_rate": 2.0104039934130348e-07,
+      "loss": 0.8065,
+      "step": 11369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6502457144534302,
+      "learning_rate": 2.0067584502078742e-07,
+      "loss": 0.7915,
+      "step": 11370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.572075210524325,
+      "learning_rate": 2.0031161476161132e-07,
+      "loss": 0.7781,
+      "step": 11371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5542465922257869,
+      "learning_rate": 1.999477085883711e-07,
+      "loss": 0.718,
+      "step": 11372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6215007243798356,
+      "learning_rate": 1.9958412652563763e-07,
+      "loss": 0.8079,
+      "step": 11373
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5850927859204047,
+      "learning_rate": 1.9922086859796074e-07,
+      "loss": 0.781,
+      "step": 11374
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7578599866840234,
+      "learning_rate": 1.9885793482986858e-07,
+      "loss": 1.0551,
+      "step": 11375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.554565299097713,
+      "learning_rate": 1.984953252458671e-07,
+      "loss": 0.7055,
+      "step": 11376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7057016109504752,
+      "learning_rate": 1.981330398704395e-07,
+      "loss": 0.8442,
+      "step": 11377
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7687786945677014,
+      "learning_rate": 1.9777107872805012e-07,
+      "loss": 1.0357,
+      "step": 11378
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6309992009151821,
+      "learning_rate": 1.9740944184313882e-07,
+      "loss": 0.7719,
+      "step": 11379
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4823164888700506,
+      "learning_rate": 1.9704812924012328e-07,
+      "loss": 0.7197,
+      "step": 11380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6254977957477459,
+      "learning_rate": 1.9668714094340012e-07,
+      "loss": 0.7759,
+      "step": 11381
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6053344074178564,
+      "learning_rate": 1.963264769773454e-07,
+      "loss": 0.7735,
+      "step": 11382
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.544267985245219,
+      "learning_rate": 1.9596613736631133e-07,
+      "loss": 0.7226,
+      "step": 11383
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5744098460798852,
+      "learning_rate": 1.9560612213462837e-07,
+      "loss": 0.7621,
+      "step": 11384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6187999976972132,
+      "learning_rate": 1.9524643130660658e-07,
+      "loss": 0.7536,
+      "step": 11385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6159556627650153,
+      "learning_rate": 1.948870649065321e-07,
+      "loss": 0.811,
+      "step": 11386
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7418941784714713,
+      "learning_rate": 1.9452802295867047e-07,
+      "loss": 1.0383,
+      "step": 11387
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5325691043304228,
+      "learning_rate": 1.9416930548726453e-07,
+      "loss": 0.7722,
+      "step": 11388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6224231126453805,
+      "learning_rate": 1.9381091251653717e-07,
+      "loss": 0.7764,
+      "step": 11389
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4410075295578681,
+      "learning_rate": 1.9345284407068677e-07,
+      "loss": 0.7782,
+      "step": 11390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.474499161180752,
+      "learning_rate": 1.930951001738901e-07,
+      "loss": 0.7641,
+      "step": 11391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5108285936493677,
+      "learning_rate": 1.9273768085030508e-07,
+      "loss": 0.6794,
+      "step": 11392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5307231143686597,
+      "learning_rate": 1.9238058612406408e-07,
+      "loss": 0.7439,
+      "step": 11393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4680332193654813,
+      "learning_rate": 1.920238160192789e-07,
+      "loss": 0.6644,
+      "step": 11394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7656016827979132,
+      "learning_rate": 1.9166737056004025e-07,
+      "loss": 1.0204,
+      "step": 11395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3834336412804957,
+      "learning_rate": 1.9131124977041616e-07,
+      "loss": 0.7062,
+      "step": 11396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.590309511905686,
+      "learning_rate": 1.909554536744518e-07,
+      "loss": 0.7402,
+      "step": 11397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4822310325950672,
+      "learning_rate": 1.9059998229617072e-07,
+      "loss": 0.7234,
+      "step": 11398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4564739918891403,
+      "learning_rate": 1.9024483565957707e-07,
+      "loss": 0.6794,
+      "step": 11399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8429645304930484,
+      "learning_rate": 1.8989001378865058e-07,
+      "loss": 0.7239,
+      "step": 11400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5391680154779082,
+      "learning_rate": 1.8953551670734816e-07,
+      "loss": 0.7659,
+      "step": 11401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7575024262552371,
+      "learning_rate": 1.8918134443960844e-07,
+      "loss": 1.1141,
+      "step": 11402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.584139506191855,
+      "learning_rate": 1.888274970093451e-07,
+      "loss": 0.8057,
+      "step": 11403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6322459930469535,
+      "learning_rate": 1.8847397444045013e-07,
+      "loss": 0.799,
+      "step": 11404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5485833049923485,
+      "learning_rate": 1.881207767567944e-07,
+      "loss": 0.7613,
+      "step": 11405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5269753627444194,
+      "learning_rate": 1.8776790398222722e-07,
+      "loss": 0.7603,
+      "step": 11406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5910138953860988,
+      "learning_rate": 1.8741535614057505e-07,
+      "loss": 0.7769,
+      "step": 11407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.590937775856611,
+      "learning_rate": 1.8706313325564274e-07,
+      "loss": 0.7356,
+      "step": 11408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7633210468185405,
+      "learning_rate": 1.8671123535121294e-07,
+      "loss": 1.0654,
+      "step": 11409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7067924136317525,
+      "learning_rate": 1.8635966245104663e-07,
+      "loss": 0.8412,
+      "step": 11410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6043094717052977,
+      "learning_rate": 1.8600841457888264e-07,
+      "loss": 0.8123,
+      "step": 11411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4873829583076292,
+      "learning_rate": 1.8565749175843916e-07,
+      "loss": 0.7361,
+      "step": 11412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6174578101058774,
+      "learning_rate": 1.8530689401341006e-07,
+      "loss": 0.862,
+      "step": 11413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5642273546638945,
+      "learning_rate": 1.8495662136746916e-07,
+      "loss": 0.7116,
+      "step": 11414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5601130418088414,
+      "learning_rate": 1.84606673844267e-07,
+      "loss": 0.8246,
+      "step": 11415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7310380754671386,
+      "learning_rate": 1.8425705146743355e-07,
+      "loss": 1.0731,
+      "step": 11416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.577411879267811,
+      "learning_rate": 1.8390775426057604e-07,
+      "loss": 0.7092,
+      "step": 11417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5208825002750896,
+      "learning_rate": 1.8355878224728008e-07,
+      "loss": 0.7129,
+      "step": 11418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5348885302444468,
+      "learning_rate": 1.8321013545110788e-07,
+      "loss": 0.7849,
+      "step": 11419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7471209462423316,
+      "learning_rate": 1.8286181389560176e-07,
+      "loss": 1.0848,
+      "step": 11420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7289920642512187,
+      "learning_rate": 1.825138176042812e-07,
+      "loss": 1.0694,
+      "step": 11421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5342736798233756,
+      "learning_rate": 1.8216614660064246e-07,
+      "loss": 0.783,
+      "step": 11422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7287690144027845,
+      "learning_rate": 1.818188009081634e-07,
+      "loss": 1.0852,
+      "step": 11423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7280327765079978,
+      "learning_rate": 1.814717805502958e-07,
+      "loss": 1.0823,
+      "step": 11424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5351236312759993,
+      "learning_rate": 1.8112508555047149e-07,
+      "loss": 0.6963,
+      "step": 11425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5109366756364304,
+      "learning_rate": 1.8077871593210116e-07,
+      "loss": 0.7441,
+      "step": 11426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5252065508654027,
+      "learning_rate": 1.804326717185717e-07,
+      "loss": 0.7719,
+      "step": 11427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.535205607011373,
+      "learning_rate": 1.800869529332483e-07,
+      "loss": 0.666,
+      "step": 11428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4664660078046563,
+      "learning_rate": 1.7974155959947614e-07,
+      "loss": 0.6822,
+      "step": 11429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5804398644730142,
+      "learning_rate": 1.793964917405755e-07,
+      "loss": 0.8742,
+      "step": 11430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7745488719338536,
+      "learning_rate": 1.790517493798466e-07,
+      "loss": 0.7553,
+      "step": 11431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5029140423058138,
+      "learning_rate": 1.7870733254056692e-07,
+      "loss": 0.7652,
+      "step": 11432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.727022547733256,
+      "learning_rate": 1.7836324124599348e-07,
+      "loss": 0.8645,
+      "step": 11433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5004434146490058,
+      "learning_rate": 1.780194755193593e-07,
+      "loss": 0.7657,
+      "step": 11434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.419702684106565,
+      "learning_rate": 1.7767603538387523e-07,
+      "loss": 0.7247,
+      "step": 11435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.381237635860276,
+      "learning_rate": 1.7733292086273336e-07,
+      "loss": 0.7743,
+      "step": 11436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4854012378178016,
+      "learning_rate": 1.7699013197909954e-07,
+      "loss": 0.705,
+      "step": 11437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5441834987042333,
+      "learning_rate": 1.7664766875612137e-07,
+      "loss": 0.8339,
+      "step": 11438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.533571948357021,
+      "learning_rate": 1.7630553121692097e-07,
+      "loss": 0.701,
+      "step": 11439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5596171382716337,
+      "learning_rate": 1.75963719384602e-07,
+      "loss": 0.7969,
+      "step": 11440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.478467838908533,
+      "learning_rate": 1.7562223328224327e-07,
+      "loss": 0.6741,
+      "step": 11441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.531512562772538,
+      "learning_rate": 1.752810729329024e-07,
+      "loss": 0.7,
+      "step": 11442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5911478878946343,
+      "learning_rate": 1.7494023835961604e-07,
+      "loss": 0.793,
+      "step": 11443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7508469808723031,
+      "learning_rate": 1.745997295853985e-07,
+      "loss": 1.0568,
+      "step": 11444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5729371124251879,
+      "learning_rate": 1.7425954663324085e-07,
+      "loss": 0.8133,
+      "step": 11445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4916463496518486,
+      "learning_rate": 1.7391968952611304e-07,
+      "loss": 0.7611,
+      "step": 11446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7475668508850616,
+      "learning_rate": 1.735801582869634e-07,
+      "loss": 1.0877,
+      "step": 11447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.642997472506341,
+      "learning_rate": 1.73240952938718e-07,
+      "loss": 0.8914,
+      "step": 11448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.520179586908612,
+      "learning_rate": 1.7290207350428024e-07,
+      "loss": 0.7442,
+      "step": 11449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4933845113980004,
+      "learning_rate": 1.725635200065323e-07,
+      "loss": 0.7784,
+      "step": 11450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5030162891800423,
+      "learning_rate": 1.722252924683343e-07,
+      "loss": 0.6874,
+      "step": 11451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.504863120299521,
+      "learning_rate": 1.7188739091252405e-07,
+      "loss": 0.692,
+      "step": 11452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5257174758740994,
+      "learning_rate": 1.7154981536191718e-07,
+      "loss": 0.7823,
+      "step": 11453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4887136748046441,
+      "learning_rate": 1.7121256583930824e-07,
+      "loss": 0.7152,
+      "step": 11454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5918879692774812,
+      "learning_rate": 1.708756423674679e-07,
+      "loss": 0.6986,
+      "step": 11455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6225856745691278,
+      "learning_rate": 1.7053904496914632e-07,
+      "loss": 0.7048,
+      "step": 11456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4147256629482556,
+      "learning_rate": 1.7020277366707193e-07,
+      "loss": 0.7272,
+      "step": 11457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5811459998931663,
+      "learning_rate": 1.6986682848395053e-07,
+      "loss": 0.7987,
+      "step": 11458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.48517917347595,
+      "learning_rate": 1.6953120944246503e-07,
+      "loss": 0.739,
+      "step": 11459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.493051272487321,
+      "learning_rate": 1.6919591656527846e-07,
+      "loss": 0.7188,
+      "step": 11460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4409274952811308,
+      "learning_rate": 1.6886094987502987e-07,
+      "loss": 0.6987,
+      "step": 11461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5051182837032335,
+      "learning_rate": 1.685263093943368e-07,
+      "loss": 0.6591,
+      "step": 11462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5481860752983105,
+      "learning_rate": 1.6819199514579553e-07,
+      "loss": 0.7234,
+      "step": 11463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7621709945666707,
+      "learning_rate": 1.678580071519792e-07,
+      "loss": 1.0852,
+      "step": 11464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4285512874484294,
+      "learning_rate": 1.6752434543543917e-07,
+      "loss": 0.741,
+      "step": 11465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7571139393318409,
+      "learning_rate": 1.671910100187052e-07,
+      "loss": 1.0345,
+      "step": 11466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5117682035717561,
+      "learning_rate": 1.6685800092428595e-07,
+      "loss": 0.7861,
+      "step": 11467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7401157623599486,
+      "learning_rate": 1.6652531817466566e-07,
+      "loss": 1.0679,
+      "step": 11468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5346386629059132,
+      "learning_rate": 1.6619296179230859e-07,
+      "loss": 0.7865,
+      "step": 11469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6011090696668449,
+      "learning_rate": 1.6586093179965513e-07,
+      "loss": 0.7928,
+      "step": 11470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7512501314124169,
+      "learning_rate": 1.655292282191262e-07,
+      "loss": 1.0424,
+      "step": 11471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4579883700896463,
+      "learning_rate": 1.651978510731189e-07,
+      "loss": 0.7104,
+      "step": 11472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4351079828639386,
+      "learning_rate": 1.648668003840076e-07,
+      "loss": 0.6896,
+      "step": 11473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1237431096594652,
+      "learning_rate": 1.6453607617414603e-07,
+      "loss": 0.7255,
+      "step": 11474
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5396295854504067,
+      "learning_rate": 1.6420567846586577e-07,
+      "loss": 0.734,
+      "step": 11475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3840271584890722,
+      "learning_rate": 1.6387560728147512e-07,
+      "loss": 0.8165,
+      "step": 11476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4694661420840485,
+      "learning_rate": 1.635458626432629e-07,
+      "loss": 0.7257,
+      "step": 11477
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6232443236922922,
+      "learning_rate": 1.6321644457349294e-07,
+      "loss": 0.7529,
+      "step": 11478
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.620190345586592,
+      "learning_rate": 1.6288735309440863e-07,
+      "loss": 0.8184,
+      "step": 11479
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4371915734488587,
+      "learning_rate": 1.6255858822823044e-07,
+      "loss": 0.6873,
+      "step": 11480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6035948310596557,
+      "learning_rate": 1.6223014999715847e-07,
+      "loss": 0.7394,
+      "step": 11481
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5069007876758964,
+      "learning_rate": 1.6190203842336882e-07,
+      "loss": 0.7166,
+      "step": 11482
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5006330195248903,
+      "learning_rate": 1.6157425352901602e-07,
+      "loss": 0.6951,
+      "step": 11483
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4940483553666957,
+      "learning_rate": 1.6124679533623456e-07,
+      "loss": 0.6348,
+      "step": 11484
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5270409123424928,
+      "learning_rate": 1.6091966386713342e-07,
+      "loss": 0.7084,
+      "step": 11485
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5582611788860636,
+      "learning_rate": 1.6059285914380164e-07,
+      "loss": 0.754,
+      "step": 11486
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7752840448714791,
+      "learning_rate": 1.602663811883054e-07,
+      "loss": 1.0279,
+      "step": 11487
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3511309722943619,
+      "learning_rate": 1.5994023002269043e-07,
+      "loss": 0.7303,
+      "step": 11488
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5511290017045107,
+      "learning_rate": 1.5961440566897913e-07,
+      "loss": 0.7698,
+      "step": 11489
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5806244139689942,
+      "learning_rate": 1.5928890814916997e-07,
+      "loss": 0.7871,
+      "step": 11490
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4312077983675053,
+      "learning_rate": 1.5896373748524375e-07,
+      "loss": 0.7092,
+      "step": 11491
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4761960861286887,
+      "learning_rate": 1.5863889369915564e-07,
+      "loss": 0.7281,
+      "step": 11492
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5174314626407077,
+      "learning_rate": 1.5831437681283924e-07,
+      "loss": 0.7342,
+      "step": 11493
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.570956077400536,
+      "learning_rate": 1.5799018684820756e-07,
+      "loss": 0.7139,
+      "step": 11494
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4660625720569322,
+      "learning_rate": 1.5766632382715084e-07,
+      "loss": 0.6747,
+      "step": 11495
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.551357353272507,
+      "learning_rate": 1.573427877715361e-07,
+      "loss": 0.7003,
+      "step": 11496
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6342115826928392,
+      "learning_rate": 1.5701957870321026e-07,
+      "loss": 0.7643,
+      "step": 11497
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5214767706236938,
+      "learning_rate": 1.5669669664399645e-07,
+      "loss": 0.7309,
+      "step": 11498
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4477134552366848,
+      "learning_rate": 1.5637414161569663e-07,
+      "loss": 0.6941,
+      "step": 11499
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4450729044845816,
+      "learning_rate": 1.5605191364008954e-07,
+      "loss": 0.7229,
+      "step": 11500
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5317795901649498,
+      "learning_rate": 1.557300127389344e-07,
+      "loss": 0.6943,
+      "step": 11501
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4936487279546737,
+      "learning_rate": 1.554084389339655e-07,
+      "loss": 0.76,
+      "step": 11502
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7766714721370767,
+      "learning_rate": 1.5508719224689716e-07,
+      "loss": 0.7396,
+      "step": 11503
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.470656583729276,
+      "learning_rate": 1.5476627269941925e-07,
+      "loss": 0.8036,
+      "step": 11504
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5772042853455661,
+      "learning_rate": 1.5444568031320272e-07,
+      "loss": 0.7142,
+      "step": 11505
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8829054092821704,
+      "learning_rate": 1.5412541510989364e-07,
+      "loss": 0.7398,
+      "step": 11506
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5853339349726727,
+      "learning_rate": 1.538054771111175e-07,
+      "loss": 0.7857,
+      "step": 11507
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5359688014008768,
+      "learning_rate": 1.5348586633847695e-07,
+      "loss": 0.7898,
+      "step": 11508
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4844990476034075,
+      "learning_rate": 1.5316658281355313e-07,
+      "loss": 0.74,
+      "step": 11509
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5395766919084806,
+      "learning_rate": 1.5284762655790374e-07,
+      "loss": 0.8287,
+      "step": 11510
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6214939415226979,
+      "learning_rate": 1.5252899759306716e-07,
+      "loss": 0.7061,
+      "step": 11511
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7619562953047476,
+      "learning_rate": 1.522106959405567e-07,
+      "loss": 1.0944,
+      "step": 11512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5728129142286058,
+      "learning_rate": 1.518927216218652e-07,
+      "loss": 0.7533,
+      "step": 11513
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.502451237987401,
+      "learning_rate": 1.5157507465846267e-07,
+      "loss": 0.6511,
+      "step": 11514
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.488325766313183,
+      "learning_rate": 1.5125775507179806e-07,
+      "loss": 0.7312,
+      "step": 11515
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7712915684782717,
+      "learning_rate": 1.5094076288329762e-07,
+      "loss": 0.729,
+      "step": 11516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4699292089255036,
+      "learning_rate": 1.5062409811436474e-07,
+      "loss": 0.7306,
+      "step": 11517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4309383359486156,
+      "learning_rate": 1.503077607863812e-07,
+      "loss": 0.7297,
+      "step": 11518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.423943505346532,
+      "learning_rate": 1.4999175092070716e-07,
+      "loss": 0.763,
+      "step": 11519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7360472015572849,
+      "learning_rate": 1.4967606853868056e-07,
+      "loss": 1.0295,
+      "step": 11520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5082824794770961,
+      "learning_rate": 1.4936071366161598e-07,
+      "loss": 0.7838,
+      "step": 11521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7686269211682747,
+      "learning_rate": 1.4904568631080807e-07,
+      "loss": 1.0536,
+      "step": 11522
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5122478384894131,
+      "learning_rate": 1.4873098650752815e-07,
+      "loss": 0.7676,
+      "step": 11523
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7489220062439181,
+      "learning_rate": 1.484166142730248e-07,
+      "loss": 1.0754,
+      "step": 11524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.88312768089819,
+      "learning_rate": 1.4810256962852543e-07,
+      "loss": 0.7102,
+      "step": 11525
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6164030281615904,
+      "learning_rate": 1.4778885259523535e-07,
+      "loss": 0.7318,
+      "step": 11526
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4930876423843507,
+      "learning_rate": 1.4747546319433702e-07,
+      "loss": 0.7382,
+      "step": 11527
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5738756583285554,
+      "learning_rate": 1.471624014469919e-07,
+      "loss": 0.6884,
+      "step": 11528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5184991791901412,
+      "learning_rate": 1.4684966737433748e-07,
+      "loss": 0.7916,
+      "step": 11529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6068277119774252,
+      "learning_rate": 1.4653726099749133e-07,
+      "loss": 0.861,
+      "step": 11530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.612661183723356,
+      "learning_rate": 1.4622518233754713e-07,
+      "loss": 0.7134,
+      "step": 11531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4472523931138574,
+      "learning_rate": 1.4591343141557746e-07,
+      "loss": 0.7751,
+      "step": 11532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5380144683059969,
+      "learning_rate": 1.456020082526327e-07,
+      "loss": 0.8304,
+      "step": 11533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5151271295596551,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.7453,
+      "step": 11534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7450856670748359,
+      "learning_rate": 1.4498014528790628e-07,
+      "loss": 1.0703,
+      "step": 11535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.460660110049066,
+      "learning_rate": 1.4466970552811488e-07,
+      "loss": 0.7464,
+      "step": 11536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7236001373892466,
+      "learning_rate": 1.4435959361132735e-07,
+      "loss": 1.0751,
+      "step": 11537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4578068009337162,
+      "learning_rate": 1.4404980955848246e-07,
+      "loss": 0.707,
+      "step": 11538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6139453857157324,
+      "learning_rate": 1.4374035339049908e-07,
+      "loss": 0.7078,
+      "step": 11539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5165889812499682,
+      "learning_rate": 1.4343122512827102e-07,
+      "loss": 0.7671,
+      "step": 11540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.429989095557889,
+      "learning_rate": 1.4312242479267213e-07,
+      "loss": 0.7091,
+      "step": 11541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5183989451795163,
+      "learning_rate": 1.4281395240455353e-07,
+      "loss": 0.8018,
+      "step": 11542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4837379586155433,
+      "learning_rate": 1.4250580798474301e-07,
+      "loss": 0.688,
+      "step": 11543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4264785878017912,
+      "learning_rate": 1.4219799155404778e-07,
+      "loss": 0.755,
+      "step": 11544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7540764389059333,
+      "learning_rate": 1.4189050313325126e-07,
+      "loss": 0.7595,
+      "step": 11545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6029274298480944,
+      "learning_rate": 1.4158334274311791e-07,
+      "loss": 0.8616,
+      "step": 11546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7457878076437932,
+      "learning_rate": 1.4127651040438618e-07,
+      "loss": 1.0556,
+      "step": 11547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.530962641062409,
+      "learning_rate": 1.4097000613777445e-07,
+      "loss": 0.775,
+      "step": 11548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7647828382705414,
+      "learning_rate": 1.4066382996397898e-07,
+      "loss": 1.0516,
+      "step": 11549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5597439567025544,
+      "learning_rate": 1.4035798190367322e-07,
+      "loss": 0.7381,
+      "step": 11550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4642254267128032,
+      "learning_rate": 1.40052461977509e-07,
+      "loss": 0.763,
+      "step": 11551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7313468943551944,
+      "learning_rate": 1.3974727020611534e-07,
+      "loss": 1.0644,
+      "step": 11552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7382761449142697,
+      "learning_rate": 1.3944240661009968e-07,
+      "loss": 1.0681,
+      "step": 11553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.464398057092571,
+      "learning_rate": 1.3913787121004717e-07,
+      "loss": 0.7665,
+      "step": 11554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4887305471778307,
+      "learning_rate": 1.3883366402652032e-07,
+      "loss": 0.7714,
+      "step": 11555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5986057351770586,
+      "learning_rate": 1.3852978508006044e-07,
+      "loss": 0.7464,
+      "step": 11556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5698639474155487,
+      "learning_rate": 1.3822623439118556e-07,
+      "loss": 0.7659,
+      "step": 11557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.501359788032596,
+      "learning_rate": 1.379230119803926e-07,
+      "loss": 0.7362,
+      "step": 11558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4830455050215656,
+      "learning_rate": 1.3762011786815576e-07,
+      "loss": 0.7035,
+      "step": 11559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5145768472936716,
+      "learning_rate": 1.3731755207492703e-07,
+      "loss": 0.6618,
+      "step": 11560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5104530912108303,
+      "learning_rate": 1.370153146211367e-07,
+      "loss": 0.7293,
+      "step": 11561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7482113044089997,
+      "learning_rate": 1.367134055271918e-07,
+      "loss": 1.0175,
+      "step": 11562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.498538637472767,
+      "learning_rate": 1.364118248134788e-07,
+      "loss": 0.729,
+      "step": 11563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6119710991041687,
+      "learning_rate": 1.361105725003603e-07,
+      "loss": 0.7223,
+      "step": 11564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5062787926744414,
+      "learning_rate": 1.358096486081778e-07,
+      "loss": 0.6824,
+      "step": 11565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5693797918692383,
+      "learning_rate": 1.3550905315725061e-07,
+      "loss": 0.6722,
+      "step": 11566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.735448322485531,
+      "learning_rate": 1.3520878616787525e-07,
+      "loss": 1.0911,
+      "step": 11567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4064885760342318,
+      "learning_rate": 1.349088476603272e-07,
+      "loss": 0.6831,
+      "step": 11568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.475807224312743,
+      "learning_rate": 1.3460923765485745e-07,
+      "loss": 0.7511,
+      "step": 11569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4950323557007452,
+      "learning_rate": 1.3430995617169817e-07,
+      "loss": 0.6697,
+      "step": 11570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7628161873871482,
+      "learning_rate": 1.340110032310571e-07,
+      "loss": 1.0825,
+      "step": 11571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4528555469083737,
+      "learning_rate": 1.3371237885311916e-07,
+      "loss": 0.684,
+      "step": 11572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.703514470748917,
+      "learning_rate": 1.3341408305804938e-07,
+      "loss": 0.7711,
+      "step": 11573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.762305753271549,
+      "learning_rate": 1.3311611586598828e-07,
+      "loss": 1.048,
+      "step": 11574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6370509409793772,
+      "learning_rate": 1.3281847729705589e-07,
+      "loss": 0.816,
+      "step": 11575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6320169244200504,
+      "learning_rate": 1.3252116737134946e-07,
+      "loss": 0.7536,
+      "step": 11576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5704851102391233,
+      "learning_rate": 1.3222418610894405e-07,
+      "loss": 0.7761,
+      "step": 11577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4843012711189025,
+      "learning_rate": 1.3192753352989252e-07,
+      "loss": 0.7631,
+      "step": 11578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.766732575021258,
+      "learning_rate": 1.3163120965422494e-07,
+      "loss": 0.6944,
+      "step": 11579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5028483984168686,
+      "learning_rate": 1.3133521450195086e-07,
+      "loss": 0.7177,
+      "step": 11580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7506150602632006,
+      "learning_rate": 1.3103954809305596e-07,
+      "loss": 1.0868,
+      "step": 11581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.453024776622909,
+      "learning_rate": 1.3074421044750375e-07,
+      "loss": 0.6995,
+      "step": 11582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7279100129985856,
+      "learning_rate": 1.3044920158523766e-07,
+      "loss": 0.7605,
+      "step": 11583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5481841179736573,
+      "learning_rate": 1.301545215261757e-07,
+      "loss": 0.7632,
+      "step": 11584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7375625349562257,
+      "learning_rate": 1.2986017029021692e-07,
+      "loss": 1.0305,
+      "step": 11585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.491996027660803,
+      "learning_rate": 1.2956614789723433e-07,
+      "loss": 0.7341,
+      "step": 11586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.818206201928402,
+      "learning_rate": 1.2927245436708314e-07,
+      "loss": 0.7471,
+      "step": 11587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.256319607334584,
+      "learning_rate": 1.2897908971959307e-07,
+      "loss": 0.8297,
+      "step": 11588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3751614442131928,
+      "learning_rate": 1.2868605397457324e-07,
+      "loss": 0.6566,
+      "step": 11589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5240284079646402,
+      "learning_rate": 1.2839334715181005e-07,
+      "loss": 0.8412,
+      "step": 11590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4800894079046318,
+      "learning_rate": 1.2810096927106773e-07,
+      "loss": 0.6998,
+      "step": 11591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5495902247994005,
+      "learning_rate": 1.278089203520877e-07,
+      "loss": 0.7096,
+      "step": 11592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4875233499089704,
+      "learning_rate": 1.275172004145908e-07,
+      "loss": 0.6832,
+      "step": 11593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3464019484190954,
+      "learning_rate": 1.272258094782741e-07,
+      "loss": 0.7559,
+      "step": 11594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7619689046036213,
+      "learning_rate": 1.26934747562813e-07,
+      "loss": 1.048,
+      "step": 11595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.515918322230553,
+      "learning_rate": 1.2664401468786114e-07,
+      "loss": 0.7986,
+      "step": 11596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4228711042862285,
+      "learning_rate": 1.2635361087304844e-07,
+      "loss": 0.7471,
+      "step": 11597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5307554784079018,
+      "learning_rate": 1.2606353613798418e-07,
+      "loss": 0.6432,
+      "step": 11598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.71454343275002,
+      "learning_rate": 1.2577379050225491e-07,
+      "loss": 1.0501,
+      "step": 11599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5409218078554658,
+      "learning_rate": 1.2548437398542502e-07,
+      "loss": 0.7137,
+      "step": 11600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.678614326946253,
+      "learning_rate": 1.2519528660703663e-07,
+      "loss": 0.6968,
+      "step": 11601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0663517715771924,
+      "learning_rate": 1.249065283866091e-07,
+      "loss": 0.8285,
+      "step": 11602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7523775099071293,
+      "learning_rate": 1.246180993436402e-07,
+      "loss": 1.0747,
+      "step": 11603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7576258466765845,
+      "learning_rate": 1.243299994976055e-07,
+      "loss": 1.0566,
+      "step": 11604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8168605808587195,
+      "learning_rate": 1.240422288679588e-07,
+      "loss": 0.7224,
+      "step": 11605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5584679345055115,
+      "learning_rate": 1.2375478747413017e-07,
+      "loss": 0.706,
+      "step": 11606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4582320735744128,
+      "learning_rate": 1.2346767533552795e-07,
+      "loss": 0.7576,
+      "step": 11607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6141420662043873,
+      "learning_rate": 1.231808924715394e-07,
+      "loss": 0.7484,
+      "step": 11608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4935690414434502,
+      "learning_rate": 1.2289443890152852e-07,
+      "loss": 0.7857,
+      "step": 11609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5721604342563642,
+      "learning_rate": 1.226083146448376e-07,
+      "loss": 0.7694,
+      "step": 11610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7151994843761079,
+      "learning_rate": 1.2232251972078558e-07,
+      "loss": 1.0367,
+      "step": 11611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7651664023690908,
+      "learning_rate": 1.2203705414867097e-07,
+      "loss": 1.0586,
+      "step": 11612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5360545154623024,
+      "learning_rate": 1.2175191794776775e-07,
+      "loss": 0.7562,
+      "step": 11613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.564692319699779,
+      "learning_rate": 1.2146711113733055e-07,
+      "loss": 0.7384,
+      "step": 11614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.449899796210899,
+      "learning_rate": 1.2118263373658956e-07,
+      "loss": 0.7087,
+      "step": 11615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4829106428918795,
+      "learning_rate": 1.2089848576475327e-07,
+      "loss": 0.82,
+      "step": 11616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4995414997408316,
+      "learning_rate": 1.20614667241008e-07,
+      "loss": 0.7471,
+      "step": 11617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.474125184970781,
+      "learning_rate": 1.2033117818451734e-07,
+      "loss": 0.7759,
+      "step": 11618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5862116446521561,
+      "learning_rate": 1.2004801861442373e-07,
+      "loss": 0.8041,
+      "step": 11619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5568671946989048,
+      "learning_rate": 1.197651885498463e-07,
+      "loss": 0.6689,
+      "step": 11620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6924081969267455,
+      "learning_rate": 1.1948268800988317e-07,
+      "loss": 0.75,
+      "step": 11621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7575881843739094,
+      "learning_rate": 1.1920051701360902e-07,
+      "loss": 1.0565,
+      "step": 11622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5289170285323785,
+      "learning_rate": 1.1891867558007586e-07,
+      "loss": 0.7796,
+      "step": 11623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5852809974989661,
+      "learning_rate": 1.1863716372831513e-07,
+      "loss": 0.7703,
+      "step": 11624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4779618454892063,
+      "learning_rate": 1.1835598147733552e-07,
+      "loss": 0.7801,
+      "step": 11625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5679329794102597,
+      "learning_rate": 1.1807512884612238e-07,
+      "loss": 0.7333,
+      "step": 11626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5041057881219142,
+      "learning_rate": 1.1779460585363945e-07,
+      "loss": 0.8837,
+      "step": 11627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6718821377989863,
+      "learning_rate": 1.1751441251882878e-07,
+      "loss": 0.737,
+      "step": 11628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5036640463753483,
+      "learning_rate": 1.1723454886061026e-07,
+      "loss": 0.7223,
+      "step": 11629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5509103320002373,
+      "learning_rate": 1.1695501489787875e-07,
+      "loss": 0.7911,
+      "step": 11630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5525405654021054,
+      "learning_rate": 1.1667581064951139e-07,
+      "loss": 0.7675,
+      "step": 11631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5766100412179762,
+      "learning_rate": 1.1639693613435921e-07,
+      "loss": 0.7956,
+      "step": 11632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7983363624813645,
+      "learning_rate": 1.161183913712527e-07,
+      "loss": 1.0588,
+      "step": 11633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5884298783633715,
+      "learning_rate": 1.158401763790007e-07,
+      "loss": 0.7725,
+      "step": 11634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7581949641237834,
+      "learning_rate": 1.1556229117638818e-07,
+      "loss": 1.0706,
+      "step": 11635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6017286776264739,
+      "learning_rate": 1.1528473578217847e-07,
+      "loss": 0.8001,
+      "step": 11636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.626937196023273,
+      "learning_rate": 1.1500751021511269e-07,
+      "loss": 0.7518,
+      "step": 11637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.562771297310525,
+      "learning_rate": 1.1473061449391032e-07,
+      "loss": 0.853,
+      "step": 11638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6042007113071641,
+      "learning_rate": 1.1445404863726806e-07,
+      "loss": 0.8139,
+      "step": 11639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7356215844187232,
+      "learning_rate": 1.1417781266385986e-07,
+      "loss": 1.0623,
+      "step": 11640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6939681652147394,
+      "learning_rate": 1.1390190659233746e-07,
+      "loss": 0.6997,
+      "step": 11641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7326529089180442,
+      "learning_rate": 1.1362633044133153e-07,
+      "loss": 1.0348,
+      "step": 11642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5102708604273685,
+      "learning_rate": 1.1335108422944884e-07,
+      "loss": 0.7754,
+      "step": 11643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6296167812020075,
+      "learning_rate": 1.1307616797527454e-07,
+      "loss": 0.817,
+      "step": 11644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5427396611259516,
+      "learning_rate": 1.1280158169737265e-07,
+      "loss": 0.783,
+      "step": 11645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4716251201785033,
+      "learning_rate": 1.125273254142828e-07,
+      "loss": 0.7673,
+      "step": 11646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4133705990483192,
+      "learning_rate": 1.1225339914452349e-07,
+      "loss": 0.6856,
+      "step": 11647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7767174657172488,
+      "learning_rate": 1.1197980290659216e-07,
+      "loss": 1.0576,
+      "step": 11648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7560863160825426,
+      "learning_rate": 1.1170653671896125e-07,
+      "loss": 1.0631,
+      "step": 11649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6036443781579655,
+      "learning_rate": 1.1143360060008268e-07,
+      "loss": 0.7592,
+      "step": 11650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5682676971547131,
+      "learning_rate": 1.1116099456838558e-07,
+      "loss": 0.7581,
+      "step": 11651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4463084254141356,
+      "learning_rate": 1.1088871864227745e-07,
+      "loss": 0.7311,
+      "step": 11652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5830524820119454,
+      "learning_rate": 1.1061677284014305e-07,
+      "loss": 0.7903,
+      "step": 11653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3679279255993526,
+      "learning_rate": 1.1034515718034321e-07,
+      "loss": 0.7699,
+      "step": 11654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4763963006796172,
+      "learning_rate": 1.1007387168121997e-07,
+      "loss": 0.7237,
+      "step": 11655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5955436492884354,
+      "learning_rate": 1.0980291636109031e-07,
+      "loss": 0.7934,
+      "step": 11656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.647589204024626,
+      "learning_rate": 1.095322912382496e-07,
+      "loss": 0.7575,
+      "step": 11657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5130994375861957,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.7561,
+      "step": 11658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4894364273882432,
+      "learning_rate": 1.0899203165750716e-07,
+      "loss": 0.7607,
+      "step": 11659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6005623521984795,
+      "learning_rate": 1.0872239723608457e-07,
+      "loss": 0.7551,
+      "step": 11660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7202470614442928,
+      "learning_rate": 1.0845309308491036e-07,
+      "loss": 0.8608,
+      "step": 11661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5052670296196762,
+      "learning_rate": 1.0818411922216832e-07,
+      "loss": 0.8036,
+      "step": 11662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.596561042550473,
+      "learning_rate": 1.0791547566602056e-07,
+      "loss": 0.775,
+      "step": 11663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4948386271658631,
+      "learning_rate": 1.0764716243460594e-07,
+      "loss": 0.7221,
+      "step": 11664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4812744827755475,
+      "learning_rate": 1.0737917954604216e-07,
+      "loss": 0.8114,
+      "step": 11665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5219337879150776,
+      "learning_rate": 1.071115270184242e-07,
+      "loss": 0.6932,
+      "step": 11666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.571391280105402,
+      "learning_rate": 1.0684420486982316e-07,
+      "loss": 0.7971,
+      "step": 11667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4406504200464427,
+      "learning_rate": 1.0657721311829128e-07,
+      "loss": 0.7413,
+      "step": 11668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5071843342454418,
+      "learning_rate": 1.0631055178185523e-07,
+      "loss": 0.722,
+      "step": 11669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4840383478285324,
+      "learning_rate": 1.0604422087852063e-07,
+      "loss": 0.7412,
+      "step": 11670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.517206568814955,
+      "learning_rate": 1.0577822042627084e-07,
+      "loss": 0.6418,
+      "step": 11671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.437331734191744,
+      "learning_rate": 1.0551255044306707e-07,
+      "loss": 0.7035,
+      "step": 11672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.477440188740493,
+      "learning_rate": 1.0524721094684831e-07,
+      "loss": 0.7532,
+      "step": 11673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5720278056210752,
+      "learning_rate": 1.0498220195552966e-07,
+      "loss": 0.7326,
+      "step": 11674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7383120522287375,
+      "learning_rate": 1.0471752348700625e-07,
+      "loss": 1.0968,
+      "step": 11675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7283716498415547,
+      "learning_rate": 1.0445317555914992e-07,
+      "loss": 1.0843,
+      "step": 11676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.754663957061084,
+      "learning_rate": 1.0418915818980857e-07,
+      "loss": 1.0235,
+      "step": 11677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4402222363095767,
+      "learning_rate": 1.039254713968102e-07,
+      "loss": 0.7515,
+      "step": 11678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5575922438225422,
+      "learning_rate": 1.0366211519795999e-07,
+      "loss": 0.7632,
+      "step": 11679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.565138290457019,
+      "learning_rate": 1.0339908961103984e-07,
+      "loss": 0.7136,
+      "step": 11680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6135672842189217,
+      "learning_rate": 1.0313639465380998e-07,
+      "loss": 0.751,
+      "step": 11681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5675855260024139,
+      "learning_rate": 1.0287403034400789e-07,
+      "loss": 0.8167,
+      "step": 11682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5635533743025227,
+      "learning_rate": 1.0261199669934941e-07,
+      "loss": 0.7103,
+      "step": 11683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5562356871555558,
+      "learning_rate": 1.0235029373752758e-07,
+      "loss": 0.7826,
+      "step": 11684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.491671569523729,
+      "learning_rate": 1.020889214762133e-07,
+      "loss": 0.7948,
+      "step": 11685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6927158954051493,
+      "learning_rate": 1.018278799330541e-07,
+      "loss": 0.8548,
+      "step": 11686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4646273274852137,
+      "learning_rate": 1.0156716912567755e-07,
+      "loss": 0.7548,
+      "step": 11687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.640934436190595,
+      "learning_rate": 1.0130678907168568e-07,
+      "loss": 0.8405,
+      "step": 11688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7008136356663095,
+      "learning_rate": 1.0104673978866164e-07,
+      "loss": 0.7203,
+      "step": 11689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5631924363582475,
+      "learning_rate": 1.0078702129416362e-07,
+      "loss": 0.7474,
+      "step": 11690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4361952670313323,
+      "learning_rate": 1.005276336057287e-07,
+      "loss": 0.8235,
+      "step": 11691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.759478753749927,
+      "learning_rate": 1.002685767408712e-07,
+      "loss": 0.7807,
+      "step": 11692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4679536496463905,
+      "learning_rate": 1.000098507170838e-07,
+      "loss": 0.8096,
+      "step": 11693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4268991337632444,
+      "learning_rate": 9.97514555518353e-08,
+      "loss": 0.7725,
+      "step": 11694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6825621810625369,
+      "learning_rate": 9.949339126257396e-08,
+      "loss": 0.7562,
+      "step": 11695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7696151699666551,
+      "learning_rate": 9.923565786672418e-08,
+      "loss": 1.084,
+      "step": 11696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.65942114914688,
+      "learning_rate": 9.897825538168926e-08,
+      "loss": 0.79,
+      "step": 11697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4115584379576358,
+      "learning_rate": 9.872118382484918e-08,
+      "loss": 0.7058,
+      "step": 11698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6377153500107164,
+      "learning_rate": 9.84644432135623e-08,
+      "loss": 0.7047,
+      "step": 11699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7526202265906664,
+      "learning_rate": 9.820803356516472e-08,
+      "loss": 1.0592,
+      "step": 11700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4973918869544123,
+      "learning_rate": 9.795195489696874e-08,
+      "loss": 0.7236,
+      "step": 11701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4071118366236521,
+      "learning_rate": 9.769620722626605e-08,
+      "loss": 0.6988,
+      "step": 11702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5142132814618308,
+      "learning_rate": 9.744079057032563e-08,
+      "loss": 0.7094,
+      "step": 11703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.457555360873698,
+      "learning_rate": 9.718570494639312e-08,
+      "loss": 0.7352,
+      "step": 11704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7536889764557025,
+      "learning_rate": 9.693095037169254e-08,
+      "loss": 1.0516,
+      "step": 11705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5300935630667072,
+      "learning_rate": 9.667652686342621e-08,
+      "loss": 0.7693,
+      "step": 11706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5335702734497474,
+      "learning_rate": 9.642243443877264e-08,
+      "loss": 0.7715,
+      "step": 11707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7402598132782229,
+      "learning_rate": 9.61686731148892e-08,
+      "loss": 1.0319,
+      "step": 11708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.44513957444645,
+      "learning_rate": 9.591524290890996e-08,
+      "loss": 0.7283,
+      "step": 11709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7497276782012292,
+      "learning_rate": 9.566214383794736e-08,
+      "loss": 1.0732,
+      "step": 11710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6702294475564803,
+      "learning_rate": 9.54093759190916e-08,
+      "loss": 0.8138,
+      "step": 11711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4267981502015816,
+      "learning_rate": 9.515693916940961e-08,
+      "loss": 0.7001,
+      "step": 11712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7620638907709514,
+      "learning_rate": 9.49048336059466e-08,
+      "loss": 0.8312,
+      "step": 11713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5493834322913107,
+      "learning_rate": 9.465305924572566e-08,
+      "loss": 0.7315,
+      "step": 11714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.421446190792139,
+      "learning_rate": 9.44016161057465e-08,
+      "loss": 0.6651,
+      "step": 11715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5454016833831585,
+      "learning_rate": 9.415050420298777e-08,
+      "loss": 0.7884,
+      "step": 11716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6827834355275006,
+      "learning_rate": 9.389972355440535e-08,
+      "loss": 0.7933,
+      "step": 11717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5605163666960093,
+      "learning_rate": 9.364927417693181e-08,
+      "loss": 0.7742,
+      "step": 11718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7468549194798393,
+      "learning_rate": 9.339915608747807e-08,
+      "loss": 1.0954,
+      "step": 11719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5365899628746467,
+      "learning_rate": 9.314936930293283e-08,
+      "loss": 0.8506,
+      "step": 11720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5644233425427656,
+      "learning_rate": 9.289991384016262e-08,
+      "loss": 0.7252,
+      "step": 11721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6111768938064022,
+      "learning_rate": 9.26507897160106e-08,
+      "loss": 0.8175,
+      "step": 11722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4939074605879605,
+      "learning_rate": 9.240199694729946e-08,
+      "loss": 0.758,
+      "step": 11723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5073270860292758,
+      "learning_rate": 9.215353555082685e-08,
+      "loss": 0.7041,
+      "step": 11724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5646367091803335,
+      "learning_rate": 9.19054055433699e-08,
+      "loss": 0.7921,
+      "step": 11725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7594709582291391,
+      "learning_rate": 9.165760694168302e-08,
+      "loss": 1.0629,
+      "step": 11726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5213469211719661,
+      "learning_rate": 9.141013976249835e-08,
+      "loss": 0.7832,
+      "step": 11727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5029727825529715,
+      "learning_rate": 9.116300402252476e-08,
+      "loss": 0.7058,
+      "step": 11728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.459059259712718,
+      "learning_rate": 9.091619973845056e-08,
+      "loss": 0.8228,
+      "step": 11729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5044771140903666,
+      "learning_rate": 9.066972692693966e-08,
+      "loss": 0.7632,
+      "step": 11730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.354584204528621,
+      "learning_rate": 9.042358560463427e-08,
+      "loss": 0.6751,
+      "step": 11731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4213141956802156,
+      "learning_rate": 9.0177775788155e-08,
+      "loss": 0.7574,
+      "step": 11732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4762576893190678,
+      "learning_rate": 8.993229749409915e-08,
+      "loss": 0.7532,
+      "step": 11733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5659915854633282,
+      "learning_rate": 8.968715073904232e-08,
+      "loss": 0.7847,
+      "step": 11734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.521450188874236,
+      "learning_rate": 8.944233553953741e-08,
+      "loss": 0.7387,
+      "step": 11735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6296541462549299,
+      "learning_rate": 8.919785191211395e-08,
+      "loss": 0.753,
+      "step": 11736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.405896126873497,
+      "learning_rate": 8.89536998732815e-08,
+      "loss": 0.7359,
+      "step": 11737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4416633891187987,
+      "learning_rate": 8.87098794395247e-08,
+      "loss": 0.6953,
+      "step": 11738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4364479772065941,
+      "learning_rate": 8.8466390627307e-08,
+      "loss": 0.7225,
+      "step": 11739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.542870236938211,
+      "learning_rate": 8.822323345306971e-08,
+      "loss": 0.77,
+      "step": 11740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.405822696575828,
+      "learning_rate": 8.798040793323137e-08,
+      "loss": 0.694,
+      "step": 11741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4648410254790962,
+      "learning_rate": 8.773791408418775e-08,
+      "loss": 0.729,
+      "step": 11742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5016611829850137,
+      "learning_rate": 8.74957519223124e-08,
+      "loss": 0.8453,
+      "step": 11743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7536762483976661,
+      "learning_rate": 8.725392146395728e-08,
+      "loss": 1.0755,
+      "step": 11744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5275084427509664,
+      "learning_rate": 8.701242272545151e-08,
+      "loss": 0.7887,
+      "step": 11745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5887238417834932,
+      "learning_rate": 8.677125572310041e-08,
+      "loss": 0.7894,
+      "step": 11746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6900224027099913,
+      "learning_rate": 8.653042047318927e-08,
+      "loss": 0.7691,
+      "step": 11747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7309195717792222,
+      "learning_rate": 8.628991699197953e-08,
+      "loss": 1.039,
+      "step": 11748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5591826654440557,
+      "learning_rate": 8.604974529571042e-08,
+      "loss": 0.7408,
+      "step": 11749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4437597580533539,
+      "learning_rate": 8.580990540059896e-08,
+      "loss": 0.7517,
+      "step": 11750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5659668206839303,
+      "learning_rate": 8.557039732283945e-08,
+      "loss": 0.7728,
+      "step": 11751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.637517525048418,
+      "learning_rate": 8.533122107860448e-08,
+      "loss": 0.8054,
+      "step": 11752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7650628139697091,
+      "learning_rate": 8.50923766840428e-08,
+      "loss": 1.0245,
+      "step": 11753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5802260661457388,
+      "learning_rate": 8.485386415528318e-08,
+      "loss": 0.7142,
+      "step": 11754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6172581696048949,
+      "learning_rate": 8.461568350842997e-08,
+      "loss": 0.6795,
+      "step": 11755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.433323643050015,
+      "learning_rate": 8.437783475956473e-08,
+      "loss": 0.7556,
+      "step": 11756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4810190263797607,
+      "learning_rate": 8.414031792474908e-08,
+      "loss": 0.7069,
+      "step": 11757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7883920303387122,
+      "learning_rate": 8.390313302001962e-08,
+      "loss": 1.0397,
+      "step": 11758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7647436816102453,
+      "learning_rate": 8.366628006139243e-08,
+      "loss": 1.0707,
+      "step": 11759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3451805958147889,
+      "learning_rate": 8.342975906485918e-08,
+      "loss": 0.7395,
+      "step": 11760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7429381294556304,
+      "learning_rate": 8.31935700463915e-08,
+      "loss": 1.0591,
+      "step": 11761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7493206279580875,
+      "learning_rate": 8.295771302193723e-08,
+      "loss": 1.0598,
+      "step": 11762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5036452488215097,
+      "learning_rate": 8.272218800742083e-08,
+      "loss": 0.7285,
+      "step": 11763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5667500417700027,
+      "learning_rate": 8.24869950187468e-08,
+      "loss": 0.6764,
+      "step": 11764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5100358619090153,
+      "learning_rate": 8.225213407179522e-08,
+      "loss": 0.7183,
+      "step": 11765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4856117843247338,
+      "learning_rate": 8.201760518242452e-08,
+      "loss": 0.7752,
+      "step": 11766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.763072008276388,
+      "learning_rate": 8.178340836647147e-08,
+      "loss": 1.0634,
+      "step": 11767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5330096045685622,
+      "learning_rate": 8.154954363974843e-08,
+      "loss": 0.7716,
+      "step": 11768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7297756673796976,
+      "learning_rate": 8.13160110180472e-08,
+      "loss": 1.0294,
+      "step": 11769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5234862438734786,
+      "learning_rate": 8.108281051713519e-08,
+      "loss": 0.8424,
+      "step": 11770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3784341346016933,
+      "learning_rate": 8.084994215276031e-08,
+      "loss": 0.6756,
+      "step": 11771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4456075443285483,
+      "learning_rate": 8.061740594064615e-08,
+      "loss": 0.7322,
+      "step": 11772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7633586193682407,
+      "learning_rate": 8.03852018964929e-08,
+      "loss": 1.0205,
+      "step": 11773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6930761463925912,
+      "learning_rate": 8.015333003598025e-08,
+      "loss": 0.81,
+      "step": 11774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4759540470342736,
+      "learning_rate": 7.992179037476511e-08,
+      "loss": 0.7099,
+      "step": 11775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.488881180583393,
+      "learning_rate": 7.969058292848108e-08,
+      "loss": 0.6745,
+      "step": 11776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.529987281848464,
+      "learning_rate": 7.945970771273903e-08,
+      "loss": 0.7346,
+      "step": 11777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7631228291735633,
+      "learning_rate": 7.922916474312981e-08,
+      "loss": 1.0691,
+      "step": 11778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7597206052154587,
+      "learning_rate": 7.899895403521928e-08,
+      "loss": 1.0515,
+      "step": 11779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.48563065674476,
+      "learning_rate": 7.876907560455116e-08,
+      "loss": 0.7739,
+      "step": 11780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5482618933332608,
+      "learning_rate": 7.853952946664911e-08,
+      "loss": 0.7047,
+      "step": 11781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4722832610193968,
+      "learning_rate": 7.831031563701131e-08,
+      "loss": 0.819,
+      "step": 11782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.471805380998007,
+      "learning_rate": 7.808143413111535e-08,
+      "loss": 0.774,
+      "step": 11783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5841009706294713,
+      "learning_rate": 7.7852884964415e-08,
+      "loss": 0.7523,
+      "step": 11784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7869525292017979,
+      "learning_rate": 7.762466815234349e-08,
+      "loss": 1.078,
+      "step": 11785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7448045129001342,
+      "learning_rate": 7.739678371031013e-08,
+      "loss": 1.0402,
+      "step": 11786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4946439987848343,
+      "learning_rate": 7.716923165370204e-08,
+      "loss": 0.6985,
+      "step": 11787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5327709730214734,
+      "learning_rate": 7.69420119978842e-08,
+      "loss": 0.794,
+      "step": 11788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4973269304359622,
+      "learning_rate": 7.671512475819876e-08,
+      "loss": 0.6748,
+      "step": 11789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6621814856259682,
+      "learning_rate": 7.648856994996568e-08,
+      "loss": 0.7956,
+      "step": 11790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6820457244043674,
+      "learning_rate": 7.626234758848272e-08,
+      "loss": 0.7497,
+      "step": 11791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6471086436213207,
+      "learning_rate": 7.603645768902546e-08,
+      "loss": 0.7746,
+      "step": 11792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.670613391628085,
+      "learning_rate": 7.581090026684556e-08,
+      "loss": 0.7717,
+      "step": 11793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7322210122207348,
+      "learning_rate": 7.558567533717365e-08,
+      "loss": 1.1143,
+      "step": 11794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.503369080556754,
+      "learning_rate": 7.536078291521698e-08,
+      "loss": 0.7096,
+      "step": 11795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7570594101176908,
+      "learning_rate": 7.51362230161612e-08,
+      "loss": 1.0491,
+      "step": 11796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4491774827244222,
+      "learning_rate": 7.491199565516916e-08,
+      "loss": 0.7693,
+      "step": 11797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4031622879402925,
+      "learning_rate": 7.468810084738099e-08,
+      "loss": 0.7359,
+      "step": 11798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4828331438888935,
+      "learning_rate": 7.446453860791458e-08,
+      "loss": 0.7295,
+      "step": 11799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7605032363006203,
+      "learning_rate": 7.424130895186565e-08,
+      "loss": 1.05,
+      "step": 11800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3871194432115348,
+      "learning_rate": 7.401841189430659e-08,
+      "loss": 0.7131,
+      "step": 11801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.613260509150101,
+      "learning_rate": 7.379584745028867e-08,
+      "loss": 0.7826,
+      "step": 11802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9685827129665163,
+      "learning_rate": 7.357361563483934e-08,
+      "loss": 0.7757,
+      "step": 11803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.660722804320975,
+      "learning_rate": 7.335171646296436e-08,
+      "loss": 0.6767,
+      "step": 11804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.435894053211387,
+      "learning_rate": 7.313014994964729e-08,
+      "loss": 0.6453,
+      "step": 11805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.778537895212343,
+      "learning_rate": 7.29089161098484e-08,
+      "loss": 0.7194,
+      "step": 11806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4889314270757301,
+      "learning_rate": 7.268801495850575e-08,
+      "loss": 0.6992,
+      "step": 11807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7263062675888676,
+      "learning_rate": 7.246744651053573e-08,
+      "loss": 1.0713,
+      "step": 11808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6231280791850418,
+      "learning_rate": 7.224721078083086e-08,
+      "loss": 0.7586,
+      "step": 11809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6767833076080592,
+      "learning_rate": 7.202730778426204e-08,
+      "loss": 0.842,
+      "step": 11810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5541955677301218,
+      "learning_rate": 7.180773753567793e-08,
+      "loss": 0.7208,
+      "step": 11811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7636402699805047,
+      "learning_rate": 7.158850004990503e-08,
+      "loss": 1.059,
+      "step": 11812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5619522565460409,
+      "learning_rate": 7.136959534174592e-08,
+      "loss": 0.6761,
+      "step": 11813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5351168819064709,
+      "learning_rate": 7.115102342598101e-08,
+      "loss": 0.7013,
+      "step": 11814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6007756214991877,
+      "learning_rate": 7.093278431737016e-08,
+      "loss": 0.7931,
+      "step": 11815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5443329083230197,
+      "learning_rate": 7.07148780306488e-08,
+      "loss": 0.7653,
+      "step": 11816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6489810493605999,
+      "learning_rate": 7.049730458053017e-08,
+      "loss": 0.8171,
+      "step": 11817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.605251702096956,
+      "learning_rate": 7.028006398170528e-08,
+      "loss": 0.7482,
+      "step": 11818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7316970400325731,
+      "learning_rate": 7.006315624884296e-08,
+      "loss": 1.0807,
+      "step": 11819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6823894590673796,
+      "learning_rate": 6.984658139658984e-08,
+      "loss": 0.7445,
+      "step": 11820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6008308001096743,
+      "learning_rate": 6.963033943956866e-08,
+      "loss": 0.7123,
+      "step": 11821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4605018128225222,
+      "learning_rate": 6.94144303923805e-08,
+      "loss": 0.7786,
+      "step": 11822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5153186239937928,
+      "learning_rate": 6.919885426960538e-08,
+      "loss": 0.7537,
+      "step": 11823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7578658994123622,
+      "learning_rate": 6.898361108579832e-08,
+      "loss": 1.0276,
+      "step": 11824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6558304940929784,
+      "learning_rate": 6.87687008554927e-08,
+      "loss": 0.7044,
+      "step": 11825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4244957382143164,
+      "learning_rate": 6.85541235932008e-08,
+      "loss": 0.7569,
+      "step": 11826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.571055713405727,
+      "learning_rate": 6.833987931341046e-08,
+      "loss": 0.7225,
+      "step": 11827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.55827626802485,
+      "learning_rate": 6.812596803058902e-08,
+      "loss": 0.7051,
+      "step": 11828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5682387513698197,
+      "learning_rate": 6.791238975917935e-08,
+      "loss": 0.6608,
+      "step": 11829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7503477238160616,
+      "learning_rate": 6.769914451360327e-08,
+      "loss": 1.0621,
+      "step": 11830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4760381737277677,
+      "learning_rate": 6.748623230825868e-08,
+      "loss": 0.7645,
+      "step": 11831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.738219894819387,
+      "learning_rate": 6.727365315752299e-08,
+      "loss": 1.0536,
+      "step": 11832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7562067209889934,
+      "learning_rate": 6.706140707574971e-08,
+      "loss": 1.078,
+      "step": 11833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5788411561381988,
+      "learning_rate": 6.684949407727015e-08,
+      "loss": 0.7997,
+      "step": 11834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4785188416419548,
+      "learning_rate": 6.663791417639287e-08,
+      "loss": 0.7535,
+      "step": 11835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5261526960380205,
+      "learning_rate": 6.642666738740422e-08,
+      "loss": 0.7497,
+      "step": 11836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6229806474519988,
+      "learning_rate": 6.62157537245689e-08,
+      "loss": 0.7638,
+      "step": 11837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4880118524546733,
+      "learning_rate": 6.600517320212719e-08,
+      "loss": 0.7491,
+      "step": 11838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7295032853882354,
+      "learning_rate": 6.57949258342988e-08,
+      "loss": 1.0625,
+      "step": 11839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5428007236647525,
+      "learning_rate": 6.558501163527964e-08,
+      "loss": 0.7688,
+      "step": 11840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7253240834836505,
+      "learning_rate": 6.537543061924334e-08,
+      "loss": 1.0715,
+      "step": 11841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5211454557679596,
+      "learning_rate": 6.516618280034192e-08,
+      "loss": 0.8785,
+      "step": 11842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6932149778271643,
+      "learning_rate": 6.49572681927041e-08,
+      "loss": 0.761,
+      "step": 11843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.463600828220428,
+      "learning_rate": 6.474868681043578e-08,
+      "loss": 0.7455,
+      "step": 11844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.429585529211455,
+      "learning_rate": 6.454043866762128e-08,
+      "loss": 0.6949,
+      "step": 11845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.44658496207876,
+      "learning_rate": 6.433252377832267e-08,
+      "loss": 0.6835,
+      "step": 11846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5407830540850413,
+      "learning_rate": 6.41249421565776e-08,
+      "loss": 0.7493,
+      "step": 11847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7480594566474317,
+      "learning_rate": 6.391769381640268e-08,
+      "loss": 1.0832,
+      "step": 11848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4741967470020076,
+      "learning_rate": 6.371077877179277e-08,
+      "loss": 0.6312,
+      "step": 11849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4219572412207049,
+      "learning_rate": 6.350419703671896e-08,
+      "loss": 0.7435,
+      "step": 11850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6989914135557433,
+      "learning_rate": 6.329794862512895e-08,
+      "loss": 0.782,
+      "step": 11851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5478382487931646,
+      "learning_rate": 6.309203355094995e-08,
+      "loss": 0.7662,
+      "step": 11852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5865679651948137,
+      "learning_rate": 6.288645182808583e-08,
+      "loss": 0.7619,
+      "step": 11853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4610707411945616,
+      "learning_rate": 6.268120347041829e-08,
+      "loss": 0.6829,
+      "step": 11854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7724639742333462,
+      "learning_rate": 6.24762884918051e-08,
+      "loss": 1.0535,
+      "step": 11855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7353164079383536,
+      "learning_rate": 6.227170690608353e-08,
+      "loss": 1.066,
+      "step": 11856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4653647413841036,
+      "learning_rate": 6.206745872706754e-08,
+      "loss": 0.762,
+      "step": 11857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5894842309074142,
+      "learning_rate": 6.186354396854776e-08,
+      "loss": 0.8143,
+      "step": 11858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4327530580489907,
+      "learning_rate": 6.165996264429264e-08,
+      "loss": 0.7615,
+      "step": 11859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7592551944699043,
+      "learning_rate": 6.145671476804948e-08,
+      "loss": 1.0752,
+      "step": 11860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.428390695972985,
+      "learning_rate": 6.125380035354179e-08,
+      "loss": 0.7861,
+      "step": 11861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5184525491224714,
+      "learning_rate": 6.105121941447024e-08,
+      "loss": 0.7518,
+      "step": 11862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7307610400806844,
+      "learning_rate": 6.084897196451445e-08,
+      "loss": 1.083,
+      "step": 11863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.489314247619539,
+      "learning_rate": 6.064705801732962e-08,
+      "loss": 0.7702,
+      "step": 11864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5085468312501191,
+      "learning_rate": 6.044547758654983e-08,
+      "loss": 0.7568,
+      "step": 11865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7302487957318256,
+      "learning_rate": 6.0244230685787e-08,
+      "loss": 1.0335,
+      "step": 11866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4364162069238844,
+      "learning_rate": 6.004331732862856e-08,
+      "loss": 0.7216,
+      "step": 11867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5065270305792933,
+      "learning_rate": 5.9842737528642e-08,
+      "loss": 0.6913,
+      "step": 11868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7652646780345472,
+      "learning_rate": 5.964249129936927e-08,
+      "loss": 1.0619,
+      "step": 11869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4935031705324187,
+      "learning_rate": 5.9442578654332895e-08,
+      "loss": 0.754,
+      "step": 11870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5362796411753907,
+      "learning_rate": 5.9242999607030396e-08,
+      "loss": 0.8099,
+      "step": 11871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.556900369970313,
+      "learning_rate": 5.904375417093877e-08,
+      "loss": 0.6864,
+      "step": 11872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4441505435825608,
+      "learning_rate": 5.8844842359511155e-08,
+      "loss": 0.7893,
+      "step": 11873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5689316608983594,
+      "learning_rate": 5.8646264186177914e-08,
+      "loss": 0.7376,
+      "step": 11874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6377251779978368,
+      "learning_rate": 5.844801966434832e-08,
+      "loss": 0.7669,
+      "step": 11875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4442666505215322,
+      "learning_rate": 5.8250108807407777e-08,
+      "loss": 0.7123,
+      "step": 11876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5905658606002402,
+      "learning_rate": 5.805253162872004e-08,
+      "loss": 0.7203,
+      "step": 11877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5143451941449155,
+      "learning_rate": 5.785528814162555e-08,
+      "loss": 0.7638,
+      "step": 11878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7165477432974746,
+      "learning_rate": 5.7658378359443104e-08,
+      "loss": 1.0629,
+      "step": 11879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5146432856326892,
+      "learning_rate": 5.746180229546816e-08,
+      "loss": 0.7553,
+      "step": 11880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6041348213634052,
+      "learning_rate": 5.726555996297456e-08,
+      "loss": 0.8156,
+      "step": 11881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.42502399154685,
+      "learning_rate": 5.7069651375212255e-08,
+      "loss": 0.7099,
+      "step": 11882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6198292179986065,
+      "learning_rate": 5.687407654540955e-08,
+      "loss": 0.7805,
+      "step": 11883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5593460687913763,
+      "learning_rate": 5.667883548677311e-08,
+      "loss": 0.744,
+      "step": 11884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7458700879376744,
+      "learning_rate": 5.648392821248461e-08,
+      "loss": 1.0394,
+      "step": 11885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5247179507537945,
+      "learning_rate": 5.628935473570518e-08,
+      "loss": 0.6685,
+      "step": 11886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4687403438170146,
+      "learning_rate": 5.6095115069573216e-08,
+      "loss": 0.7159,
+      "step": 11887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4950160674669732,
+      "learning_rate": 5.590120922720432e-08,
+      "loss": 0.7446,
+      "step": 11888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4818913538848177,
+      "learning_rate": 5.5707637221690815e-08,
+      "loss": 0.8281,
+      "step": 11889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2057876418645366,
+      "learning_rate": 5.551439906610334e-08,
+      "loss": 0.6878,
+      "step": 11890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5490490633282796,
+      "learning_rate": 5.532149477349036e-08,
+      "loss": 0.6937,
+      "step": 11891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.490267414347693,
+      "learning_rate": 5.512892435687645e-08,
+      "loss": 0.7975,
+      "step": 11892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6173505439428486,
+      "learning_rate": 5.493668782926453e-08,
+      "loss": 0.7398,
+      "step": 11893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7588969226169844,
+      "learning_rate": 5.4744785203635355e-08,
+      "loss": 1.0928,
+      "step": 11894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7583972150800963,
+      "learning_rate": 5.4553216492946317e-08,
+      "loss": 1.0491,
+      "step": 11895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5299857730422597,
+      "learning_rate": 5.436198171013207e-08,
+      "loss": 0.7267,
+      "step": 11896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6182229180564647,
+      "learning_rate": 5.417108086810618e-08,
+      "loss": 0.832,
+      "step": 11897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7658880206484194,
+      "learning_rate": 5.3980513979758344e-08,
+      "loss": 1.0478,
+      "step": 11898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4170249597618734,
+      "learning_rate": 5.379028105795547e-08,
+      "loss": 0.7398,
+      "step": 11899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.493441539105011,
+      "learning_rate": 5.360038211554286e-08,
+      "loss": 0.7791,
+      "step": 11900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5595171473853402,
+      "learning_rate": 5.3410817165343576e-08,
+      "loss": 0.8202,
+      "step": 11901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7438996169881661,
+      "learning_rate": 5.322158622015683e-08,
+      "loss": 1.0532,
+      "step": 11902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5188510827706196,
+      "learning_rate": 5.3032689292760176e-08,
+      "loss": 0.8028,
+      "step": 11903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7411912264654605,
+      "learning_rate": 5.284412639590786e-08,
+      "loss": 0.7786,
+      "step": 11904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.549813520505425,
+      "learning_rate": 5.265589754233302e-08,
+      "loss": 0.8436,
+      "step": 11905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4828792420680912,
+      "learning_rate": 5.246800274474439e-08,
+      "loss": 0.6895,
+      "step": 11906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7681539552428225,
+      "learning_rate": 5.22804420158296e-08,
+      "loss": 1.0521,
+      "step": 11907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5616601959942615,
+      "learning_rate": 5.209321536825351e-08,
+      "loss": 0.7376,
+      "step": 11908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4866465643175135,
+      "learning_rate": 5.190632281465713e-08,
+      "loss": 0.6735,
+      "step": 11909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5063916260380588,
+      "learning_rate": 5.17197643676598e-08,
+      "loss": 0.7231,
+      "step": 11910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.656488460358884,
+      "learning_rate": 5.153354003985922e-08,
+      "loss": 0.848,
+      "step": 11911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7358928921710034,
+      "learning_rate": 5.134764984382978e-08,
+      "loss": 1.0635,
+      "step": 11912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.562811942513881,
+      "learning_rate": 5.116209379212256e-08,
+      "loss": 0.7965,
+      "step": 11913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5415302469121195,
+      "learning_rate": 5.0976871897266966e-08,
+      "loss": 0.8139,
+      "step": 11914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5080493925235028,
+      "learning_rate": 5.079198417176967e-08,
+      "loss": 0.7226,
+      "step": 11915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5259115828397163,
+      "learning_rate": 5.060743062811457e-08,
+      "loss": 0.7683,
+      "step": 11916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.741953001199755,
+      "learning_rate": 5.042321127876337e-08,
+      "loss": 1.044,
+      "step": 11917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5949082969816555,
+      "learning_rate": 5.0239326136154454e-08,
+      "loss": 0.7461,
+      "step": 11918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4631116214277187,
+      "learning_rate": 5.0055775212705107e-08,
+      "loss": 0.6662,
+      "step": 11919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.52818776203499,
+      "learning_rate": 4.9872558520807635e-08,
+      "loss": 0.7775,
+      "step": 11920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5479289744028941,
+      "learning_rate": 4.968967607283493e-08,
+      "loss": 0.8117,
+      "step": 11921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5863330323715512,
+      "learning_rate": 4.9507127881134876e-08,
+      "loss": 0.7693,
+      "step": 11922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5258386282299399,
+      "learning_rate": 4.932491395803318e-08,
+      "loss": 0.7536,
+      "step": 11923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.672003749123705,
+      "learning_rate": 4.914303431583389e-08,
+      "loss": 0.8013,
+      "step": 11924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7836663401862163,
+      "learning_rate": 4.896148896681774e-08,
+      "loss": 1.0473,
+      "step": 11925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4106046622418669,
+      "learning_rate": 4.8780277923243244e-08,
+      "loss": 0.7594,
+      "step": 11926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.68338462395273,
+      "learning_rate": 4.859940119734563e-08,
+      "loss": 0.7371,
+      "step": 11927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4645087854832732,
+      "learning_rate": 4.8418858801339007e-08,
+      "loss": 0.7456,
+      "step": 11928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6195679909651803,
+      "learning_rate": 4.823865074741307e-08,
+      "loss": 0.8092,
+      "step": 11929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3812763119736708,
+      "learning_rate": 4.8058777047736427e-08,
+      "loss": 0.7511,
+      "step": 11930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7567138053457433,
+      "learning_rate": 4.787923771445435e-08,
+      "loss": 1.0452,
+      "step": 11931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6407055102804997,
+      "learning_rate": 4.7700032759690484e-08,
+      "loss": 0.688,
+      "step": 11932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5418341145876793,
+      "learning_rate": 4.752116219554403e-08,
+      "loss": 0.6797,
+      "step": 11933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4994473307634746,
+      "learning_rate": 4.734262603409312e-08,
+      "loss": 0.791,
+      "step": 11934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6392252038486956,
+      "learning_rate": 4.71644242873931e-08,
+      "loss": 0.8438,
+      "step": 11935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4149812640305448,
+      "learning_rate": 4.698655696747656e-08,
+      "loss": 0.7777,
+      "step": 11936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5725697353290597,
+      "learning_rate": 4.680902408635335e-08,
+      "loss": 0.7694,
+      "step": 11937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5184439599193844,
+      "learning_rate": 4.66318256560111e-08,
+      "loss": 0.7753,
+      "step": 11938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7333901239806647,
+      "learning_rate": 4.645496168841468e-08,
+      "loss": 1.0329,
+      "step": 11939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6960488922416148,
+      "learning_rate": 4.627843219550621e-08,
+      "loss": 0.8026,
+      "step": 11940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7358954413653698,
+      "learning_rate": 4.6102237189205036e-08,
+      "loss": 1.0411,
+      "step": 11941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7709938868977536,
+      "learning_rate": 4.592637668140887e-08,
+      "loss": 1.0877,
+      "step": 11942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7571276015230615,
+      "learning_rate": 4.575085068399154e-08,
+      "loss": 1.0449,
+      "step": 11943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4326696685120044,
+      "learning_rate": 4.55756592088058e-08,
+      "loss": 0.7484,
+      "step": 11944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.530256857722689,
+      "learning_rate": 4.5400802267679955e-08,
+      "loss": 0.7647,
+      "step": 11945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5259967522541673,
+      "learning_rate": 4.5226279872421784e-08,
+      "loss": 0.7557,
+      "step": 11946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7238977806797886,
+      "learning_rate": 4.50520920348152e-08,
+      "loss": 0.7986,
+      "step": 11947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4636137136076974,
+      "learning_rate": 4.4878238766620784e-08,
+      "loss": 0.7481,
+      "step": 11948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6645420255325523,
+      "learning_rate": 4.4704720079579155e-08,
+      "loss": 0.7524,
+      "step": 11949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7606768903387768,
+      "learning_rate": 4.453153598540538e-08,
+      "loss": 1.0753,
+      "step": 11950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5547758855050897,
+      "learning_rate": 4.4358686495793444e-08,
+      "loss": 0.7759,
+      "step": 11951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7611789257486212,
+      "learning_rate": 4.418617162241512e-08,
+      "loss": 1.0382,
+      "step": 11952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.525104812367159,
+      "learning_rate": 4.4013991376918306e-08,
+      "loss": 0.7631,
+      "step": 11953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6968682697981758,
+      "learning_rate": 4.384214577092927e-08,
+      "loss": 0.8147,
+      "step": 11954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5701560272377886,
+      "learning_rate": 4.3670634816052046e-08,
+      "loss": 0.7426,
+      "step": 11955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.738403380793251,
+      "learning_rate": 4.349945852386628e-08,
+      "loss": 1.0553,
+      "step": 11956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.490444927169988,
+      "learning_rate": 4.3328616905931595e-08,
+      "loss": 0.6965,
+      "step": 11957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5948105701882,
+      "learning_rate": 4.315810997378212e-08,
+      "loss": 0.7251,
+      "step": 11958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5583364533491446,
+      "learning_rate": 4.29879377389314e-08,
+      "loss": 0.7971,
+      "step": 11959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7287752849192682,
+      "learning_rate": 4.281810021287081e-08,
+      "loss": 1.0379,
+      "step": 11960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5911090162255506,
+      "learning_rate": 4.264859740706673e-08,
+      "loss": 0.7685,
+      "step": 11961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4605787117398492,
+      "learning_rate": 4.2479429332965006e-08,
+      "loss": 0.7763,
+      "step": 11962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5430515702741172,
+      "learning_rate": 4.2310596001988725e-08,
+      "loss": 0.7806,
+      "step": 11963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.58108609579862,
+      "learning_rate": 4.21420974255371e-08,
+      "loss": 0.7083,
+      "step": 11964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.739197160727827,
+      "learning_rate": 4.1973933614987693e-08,
+      "loss": 1.0845,
+      "step": 11965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4802605234572235,
+      "learning_rate": 4.180610458169587e-08,
+      "loss": 0.7017,
+      "step": 11966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.631052052292385,
+      "learning_rate": 4.163861033699368e-08,
+      "loss": 0.7641,
+      "step": 11967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.669491476714782,
+      "learning_rate": 4.147145089218985e-08,
+      "loss": 0.7324,
+      "step": 11968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4257264678704684,
+      "learning_rate": 4.130462625857257e-08,
+      "loss": 0.6853,
+      "step": 11969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4832146218789752,
+      "learning_rate": 4.1138136447405606e-08,
+      "loss": 0.7328,
+      "step": 11970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6500276753073948,
+      "learning_rate": 4.097198146993053e-08,
+      "loss": 0.7906,
+      "step": 11971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4924071764496711,
+      "learning_rate": 4.08061613373667e-08,
+      "loss": 0.7789,
+      "step": 11972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.488221216646638,
+      "learning_rate": 4.0640676060911267e-08,
+      "loss": 0.7905,
+      "step": 11973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7242425339719398,
+      "learning_rate": 4.047552565173751e-08,
+      "loss": 0.7885,
+      "step": 11974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.656934514029331,
+      "learning_rate": 4.0310710120995966e-08,
+      "loss": 0.8449,
+      "step": 11975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.487160068651432,
+      "learning_rate": 4.014622947981717e-08,
+      "loss": 0.7569,
+      "step": 11976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7039889842019553,
+      "learning_rate": 3.998208373930612e-08,
+      "loss": 0.8199,
+      "step": 11977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5409773225870376,
+      "learning_rate": 3.981827291054674e-08,
+      "loss": 0.7206,
+      "step": 11978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.481048774691378,
+      "learning_rate": 3.965479700459962e-08,
+      "loss": 0.803,
+      "step": 11979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.567369732938257,
+      "learning_rate": 3.949165603250316e-08,
+      "loss": 0.7681,
+      "step": 11980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5258270200374175,
+      "learning_rate": 3.932885000527298e-08,
+      "loss": 0.8035,
+      "step": 11981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5182144855270965,
+      "learning_rate": 3.916637893390196e-08,
+      "loss": 0.7164,
+      "step": 11982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5313258106643597,
+      "learning_rate": 3.900424282936077e-08,
+      "loss": 0.7806,
+      "step": 11983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4531249492882428,
+      "learning_rate": 3.884244170259732e-08,
+      "loss": 0.7364,
+      "step": 11984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5484451732470077,
+      "learning_rate": 3.8680975564536206e-08,
+      "loss": 0.7606,
+      "step": 11985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.74215290038469,
+      "learning_rate": 3.851984442608036e-08,
+      "loss": 1.0606,
+      "step": 11986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6473463094015097,
+      "learning_rate": 3.835904829810999e-08,
+      "loss": 0.7448,
+      "step": 11987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5249144636943242,
+      "learning_rate": 3.8198587191481394e-08,
+      "loss": 0.7682,
+      "step": 11988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.564264827407454,
+      "learning_rate": 3.8038461117030914e-08,
+      "loss": 0.7725,
+      "step": 11989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5875614394893987,
+      "learning_rate": 3.787867008556934e-08,
+      "loss": 0.8201,
+      "step": 11990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4470535230395765,
+      "learning_rate": 3.771921410788693e-08,
+      "loss": 0.7402,
+      "step": 11991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6477725145761186,
+      "learning_rate": 3.756009319474951e-08,
+      "loss": 0.7028,
+      "step": 11992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.419199782037499,
+      "learning_rate": 3.7401307356902395e-08,
+      "loss": 0.7247,
+      "step": 11993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.676819214103522,
+      "learning_rate": 3.724285660506699e-08,
+      "loss": 0.7165,
+      "step": 11994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.538935410488219,
+      "learning_rate": 3.708474094994141e-08,
+      "loss": 0.6726,
+      "step": 11995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5708417726003436,
+      "learning_rate": 3.6926960402202674e-08,
+      "loss": 0.814,
+      "step": 11996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.755790789578418,
+      "learning_rate": 3.676951497250447e-08,
+      "loss": 1.0644,
+      "step": 11997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5481452784040748,
+      "learning_rate": 3.6612404671477197e-08,
+      "loss": 0.7383,
+      "step": 11998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5774879174507113,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.7276,
+      "step": 11999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4950879757144344,
+      "learning_rate": 3.629918949784872e-08,
+      "loss": 0.7425,
+      "step": 12000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4720948102130558,
+      "learning_rate": 3.6143084646396156e-08,
+      "loss": 0.7873,
+      "step": 12001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5572405382743781,
+      "learning_rate": 3.5987314965913456e-08,
+      "loss": 0.7572,
+      "step": 12002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6108717533855452,
+      "learning_rate": 3.583188046691777e-08,
+      "loss": 0.7542,
+      "step": 12003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4656890129282163,
+      "learning_rate": 3.567678115990514e-08,
+      "loss": 0.6733,
+      "step": 12004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5758244132348622,
+      "learning_rate": 3.552201705534775e-08,
+      "loss": 0.7518,
+      "step": 12005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5796032396367683,
+      "learning_rate": 3.5367588163695566e-08,
+      "loss": 0.7415,
+      "step": 12006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.459730184716668,
+      "learning_rate": 3.5213494495376364e-08,
+      "loss": 0.7294,
+      "step": 12007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7585135262931583,
+      "learning_rate": 3.505973606079405e-08,
+      "loss": 1.0586,
+      "step": 12008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5394306903499744,
+      "learning_rate": 3.4906312870331973e-08,
+      "loss": 0.7943,
+      "step": 12009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5911551930270873,
+      "learning_rate": 3.475322493434907e-08,
+      "loss": 0.7899,
+      "step": 12010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6514995022182055,
+      "learning_rate": 3.460047226318208e-08,
+      "loss": 0.822,
+      "step": 12011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4984882179802255,
+      "learning_rate": 3.4448054867144424e-08,
+      "loss": 0.7139,
+      "step": 12012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5631710317787282,
+      "learning_rate": 3.429597275652952e-08,
+      "loss": 0.7907,
+      "step": 12013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4462195018317614,
+      "learning_rate": 3.414422594160527e-08,
+      "loss": 0.7945,
+      "step": 12014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5420659129850651,
+      "learning_rate": 3.399281443261793e-08,
+      "loss": 0.752,
+      "step": 12015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.485810319178031,
+      "learning_rate": 3.384173823979098e-08,
+      "loss": 0.7221,
+      "step": 12016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7425265740157266,
+      "learning_rate": 3.3690997373325705e-08,
+      "loss": 1.0642,
+      "step": 12017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5021346667771052,
+      "learning_rate": 3.354059184340064e-08,
+      "loss": 0.7374,
+      "step": 12018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4789291484268914,
+      "learning_rate": 3.339052166017098e-08,
+      "loss": 0.8289,
+      "step": 12019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6791615753343079,
+      "learning_rate": 3.32407868337703e-08,
+      "loss": 0.7754,
+      "step": 12020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6482472975226365,
+      "learning_rate": 3.309138737430884e-08,
+      "loss": 0.7562,
+      "step": 12021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7575972704545427,
+      "learning_rate": 3.294232329187408e-08,
+      "loss": 1.0659,
+      "step": 12022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7483060896071535,
+      "learning_rate": 3.279359459653242e-08,
+      "loss": 1.0567,
+      "step": 12023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6227054438209503,
+      "learning_rate": 3.264520129832471e-08,
+      "loss": 0.7142,
+      "step": 12024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5648380980115593,
+      "learning_rate": 3.2497143407271837e-08,
+      "loss": 0.7769,
+      "step": 12025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.770718914632636,
+      "learning_rate": 3.234942093337079e-08,
+      "loss": 1.07,
+      "step": 12026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.511612786798848,
+      "learning_rate": 3.220203388659582e-08,
+      "loss": 0.7516,
+      "step": 12027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6309551109201752,
+      "learning_rate": 3.2054982276899516e-08,
+      "loss": 0.8533,
+      "step": 12028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5443701878247627,
+      "learning_rate": 3.1908266114210054e-08,
+      "loss": 0.7216,
+      "step": 12029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7726090847214648,
+      "learning_rate": 3.1761885408435055e-08,
+      "loss": 0.7975,
+      "step": 12030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5886782185557369,
+      "learning_rate": 3.161584016945829e-08,
+      "loss": 0.75,
+      "step": 12031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5973790357198578,
+      "learning_rate": 3.147013040714075e-08,
+      "loss": 0.7522,
+      "step": 12032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5721669701526224,
+      "learning_rate": 3.1324756131320685e-08,
+      "loss": 0.8757,
+      "step": 12033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4695380893574135,
+      "learning_rate": 3.1179717351815245e-08,
+      "loss": 0.6312,
+      "step": 12034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6217474470555753,
+      "learning_rate": 3.1035014078417136e-08,
+      "loss": 0.7235,
+      "step": 12035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7516883031973185,
+      "learning_rate": 3.089064632089689e-08,
+      "loss": 0.7184,
+      "step": 12036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6957071983562524,
+      "learning_rate": 3.0746614089002814e-08,
+      "loss": 0.6764,
+      "step": 12037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4528770045594799,
+      "learning_rate": 3.0602917392460463e-08,
+      "loss": 0.7174,
+      "step": 12038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5121904655596383,
+      "learning_rate": 3.0459556240972635e-08,
+      "loss": 0.7968,
+      "step": 12039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5496453186608181,
+      "learning_rate": 3.03165306442188e-08,
+      "loss": 0.7928,
+      "step": 12040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.512845943940522,
+      "learning_rate": 3.017384061185624e-08,
+      "loss": 0.8261,
+      "step": 12041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.538267013249084,
+      "learning_rate": 3.003148615352058e-08,
+      "loss": 0.7503,
+      "step": 12042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5608459940011419,
+      "learning_rate": 2.988946727882303e-08,
+      "loss": 0.7148,
+      "step": 12043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4595637563485164,
+      "learning_rate": 2.974778399735423e-08,
+      "loss": 0.7502,
+      "step": 12044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5753283212260905,
+      "learning_rate": 2.9606436318679878e-08,
+      "loss": 0.718,
+      "step": 12045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5020035950143162,
+      "learning_rate": 2.9465424252343998e-08,
+      "loss": 0.7015,
+      "step": 12046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4379269748535652,
+      "learning_rate": 2.932474780786898e-08,
+      "loss": 0.7363,
+      "step": 12047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5917438528305938,
+      "learning_rate": 2.9184406994753335e-08,
+      "loss": 0.6887,
+      "step": 12048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5500904830223958,
+      "learning_rate": 2.904440182247281e-08,
+      "loss": 0.7487,
+      "step": 12049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7342263536085256,
+      "learning_rate": 2.8904732300480965e-08,
+      "loss": 1.0791,
+      "step": 12050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.779008490156847,
+      "learning_rate": 2.8765398438208582e-08,
+      "loss": 0.8237,
+      "step": 12051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5132382132552953,
+      "learning_rate": 2.8626400245064247e-08,
+      "loss": 0.7223,
+      "step": 12052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.667518437491313,
+      "learning_rate": 2.8487737730432674e-08,
+      "loss": 0.7882,
+      "step": 12053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8040965769087123,
+      "learning_rate": 2.8349410903677488e-08,
+      "loss": 0.7262,
+      "step": 12054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.495068151923647,
+      "learning_rate": 2.8211419774137882e-08,
+      "loss": 0.7767,
+      "step": 12055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2951634006422768,
+      "learning_rate": 2.8073764351132517e-08,
+      "loss": 0.5863,
+      "step": 12056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.759360951724486,
+      "learning_rate": 2.7936444643955063e-08,
+      "loss": 1.0661,
+      "step": 12057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5143809292066703,
+      "learning_rate": 2.7799460661878663e-08,
+      "loss": 0.7292,
+      "step": 12058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6715454818568098,
+      "learning_rate": 2.766281241415203e-08,
+      "loss": 0.7805,
+      "step": 12059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7298366674264836,
+      "learning_rate": 2.7526499910002225e-08,
+      "loss": 0.7579,
+      "step": 12060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5370114786502842,
+      "learning_rate": 2.7390523158633552e-08,
+      "loss": 0.7856,
+      "step": 12061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.529890359233118,
+      "learning_rate": 2.7254882169227002e-08,
+      "loss": 0.6484,
+      "step": 12062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4655259074787756,
+      "learning_rate": 2.7119576950941918e-08,
+      "loss": 0.8372,
+      "step": 12063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4905858429465033,
+      "learning_rate": 2.698460751291432e-08,
+      "loss": 0.7676,
+      "step": 12064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7281693143567856,
+      "learning_rate": 2.684997386425692e-08,
+      "loss": 1.0332,
+      "step": 12065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5131673673512678,
+      "learning_rate": 2.6715676014061887e-08,
+      "loss": 0.8068,
+      "step": 12066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.562162478433678,
+      "learning_rate": 2.658171397139586e-08,
+      "loss": 0.7659,
+      "step": 12067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4347873939509137,
+      "learning_rate": 2.644808774530494e-08,
+      "loss": 0.7743,
+      "step": 12068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5621475830483658,
+      "learning_rate": 2.631479734481246e-08,
+      "loss": 0.807,
+      "step": 12069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5821912201303974,
+      "learning_rate": 2.6181842778917332e-08,
+      "loss": 0.7561,
+      "step": 12070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4940135049611913,
+      "learning_rate": 2.6049224056597933e-08,
+      "loss": 0.754,
+      "step": 12071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7337815418747571,
+      "learning_rate": 2.591694118680821e-08,
+      "loss": 1.063,
+      "step": 12072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5872876443814612,
+      "learning_rate": 2.578499417848046e-08,
+      "loss": 0.685,
+      "step": 12073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5030384380613944,
+      "learning_rate": 2.5653383040524228e-08,
+      "loss": 0.8115,
+      "step": 12074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6400037322774594,
+      "learning_rate": 2.552210778182629e-08,
+      "loss": 0.7785,
+      "step": 12075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6309772703087977,
+      "learning_rate": 2.5391168411250665e-08,
+      "loss": 0.7908,
+      "step": 12076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7310115041380982,
+      "learning_rate": 2.526056493763751e-08,
+      "loss": 1.061,
+      "step": 12077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5721089136938338,
+      "learning_rate": 2.5130297369807543e-08,
+      "loss": 0.751,
+      "step": 12078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6283089586074007,
+      "learning_rate": 2.5000365716554843e-08,
+      "loss": 0.6954,
+      "step": 12079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6143667127654806,
+      "learning_rate": 2.4870769986654054e-08,
+      "loss": 0.8307,
+      "step": 12080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.544125946752776,
+      "learning_rate": 2.4741510188854843e-08,
+      "loss": 0.7587,
+      "step": 12081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7368203431129738,
+      "learning_rate": 2.461258633188579e-08,
+      "loss": 1.0404,
+      "step": 12082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4762284314298035,
+      "learning_rate": 2.4483998424451593e-08,
+      "loss": 0.8304,
+      "step": 12083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7566756581564131,
+      "learning_rate": 2.4355746475234755e-08,
+      "loss": 1.0532,
+      "step": 12084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7545129706382248,
+      "learning_rate": 2.422783049289612e-08,
+      "loss": 1.0522,
+      "step": 12085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6314778991217525,
+      "learning_rate": 2.4100250486071565e-08,
+      "loss": 0.7771,
+      "step": 12086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7445415248824638,
+      "learning_rate": 2.3973006463376412e-08,
+      "loss": 1.0722,
+      "step": 12087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7472520441798517,
+      "learning_rate": 2.384609843340213e-08,
+      "loss": 1.0784,
+      "step": 12088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4437707071227226,
+      "learning_rate": 2.3719526404717975e-08,
+      "loss": 0.7269,
+      "step": 12089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6031409134867034,
+      "learning_rate": 2.3593290385870436e-08,
+      "loss": 0.6971,
+      "step": 12090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5957027219055668,
+      "learning_rate": 2.3467390385382706e-08,
+      "loss": 0.6955,
+      "step": 12091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4825727413774954,
+      "learning_rate": 2.3341826411756863e-08,
+      "loss": 0.757,
+      "step": 12092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5795420642215066,
+      "learning_rate": 2.3216598473470575e-08,
+      "loss": 0.8322,
+      "step": 12093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6821250601939293,
+      "learning_rate": 2.3091706578979857e-08,
+      "loss": 0.7579,
+      "step": 12094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4766261286680555,
+      "learning_rate": 2.2967150736717402e-08,
+      "loss": 0.7179,
+      "step": 12095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5280060479286353,
+      "learning_rate": 2.2842930955093158e-08,
+      "loss": 0.7735,
+      "step": 12096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5691790528506049,
+      "learning_rate": 2.2719047242495406e-08,
+      "loss": 0.7902,
+      "step": 12097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5957527229822552,
+      "learning_rate": 2.2595499607289127e-08,
+      "loss": 0.7689,
+      "step": 12098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5255819363272125,
+      "learning_rate": 2.2472288057815984e-08,
+      "loss": 0.6977,
+      "step": 12099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6159646744970761,
+      "learning_rate": 2.234941260239598e-08,
+      "loss": 0.7433,
+      "step": 12100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3920847198574284,
+      "learning_rate": 2.2226873249325263e-08,
+      "loss": 0.6379,
+      "step": 12101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5910744886051125,
+      "learning_rate": 2.2104670006878883e-08,
+      "loss": 0.7379,
+      "step": 12102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4659846137387946,
+      "learning_rate": 2.1982802883307453e-08,
+      "loss": 0.7469,
+      "step": 12103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.500338275173877,
+      "learning_rate": 2.1861271886840508e-08,
+      "loss": 0.7952,
+      "step": 12104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4366635451948462,
+      "learning_rate": 2.1740077025683703e-08,
+      "loss": 0.7385,
+      "step": 12105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.586521187657662,
+      "learning_rate": 2.161921830801994e-08,
+      "loss": 0.75,
+      "step": 12106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.564629514885037,
+      "learning_rate": 2.149869574201047e-08,
+      "loss": 0.7362,
+      "step": 12107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7391356847167303,
+      "learning_rate": 2.1378509335793217e-08,
+      "loss": 1.1054,
+      "step": 12108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.588425342510197,
+      "learning_rate": 2.125865909748337e-08,
+      "loss": 0.7634,
+      "step": 12109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4743592309817515,
+      "learning_rate": 2.1139145035173336e-08,
+      "loss": 0.7177,
+      "step": 12110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7408690567085984,
+      "learning_rate": 2.1019967156932774e-08,
+      "loss": 1.0512,
+      "step": 12111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4660455431411379,
+      "learning_rate": 2.090112547080969e-08,
+      "loss": 0.7893,
+      "step": 12112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.49794075140457,
+      "learning_rate": 2.0782619984827667e-08,
+      "loss": 0.7573,
+      "step": 12113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5478142221315458,
+      "learning_rate": 2.0664450706988636e-08,
+      "loss": 0.7745,
+      "step": 12114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5592176632428518,
+      "learning_rate": 2.0546617645272325e-08,
+      "loss": 0.7687,
+      "step": 12115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.520778902736634,
+      "learning_rate": 2.0429120807634595e-08,
+      "loss": 0.7783,
+      "step": 12116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4993376085191434,
+      "learning_rate": 2.0311960202009097e-08,
+      "loss": 0.7079,
+      "step": 12117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5637398314716326,
+      "learning_rate": 2.0195135836306168e-08,
+      "loss": 0.7597,
+      "step": 12118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4371945917201197,
+      "learning_rate": 2.0078647718415058e-08,
+      "loss": 0.7039,
+      "step": 12119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5112828324028913,
+      "learning_rate": 1.9962495856201135e-08,
+      "loss": 0.81,
+      "step": 12120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.478274913631058,
+      "learning_rate": 1.984668025750647e-08,
+      "loss": 0.7395,
+      "step": 12121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4730426256989468,
+      "learning_rate": 1.9731200930152572e-08,
+      "loss": 0.769,
+      "step": 12122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7249290090168846,
+      "learning_rate": 1.9616057881935436e-08,
+      "loss": 1.0065,
+      "step": 12123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.934300895769558,
+      "learning_rate": 1.950125112063106e-08,
+      "loss": 0.6933,
+      "step": 12124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5452861105453484,
+      "learning_rate": 1.9386780653989913e-08,
+      "loss": 0.8033,
+      "step": 12125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5322404185331482,
+      "learning_rate": 1.927264648974303e-08,
+      "loss": 0.7366,
+      "step": 12126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5599745376133634,
+      "learning_rate": 1.9158848635595915e-08,
+      "loss": 0.7927,
+      "step": 12127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.732465566611681,
+      "learning_rate": 1.9045387099232425e-08,
+      "loss": 1.048,
+      "step": 12128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7429380353699753,
+      "learning_rate": 1.8932261888314207e-08,
+      "loss": 1.0279,
+      "step": 12129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7733017992885716,
+      "learning_rate": 1.8819473010479594e-08,
+      "loss": 1.0631,
+      "step": 12130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4892529781677015,
+      "learning_rate": 1.8707020473344163e-08,
+      "loss": 0.7897,
+      "step": 12131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.562627988393497,
+      "learning_rate": 1.8594904284501282e-08,
+      "loss": 0.8471,
+      "step": 12132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4155780910907019,
+      "learning_rate": 1.8483124451521005e-08,
+      "loss": 0.7685,
+      "step": 12133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.730907653826536,
+      "learning_rate": 1.837168098195119e-08,
+      "loss": 0.8345,
+      "step": 12134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5351954783836654,
+      "learning_rate": 1.8260573883316924e-08,
+      "loss": 0.793,
+      "step": 12135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7594083629174347,
+      "learning_rate": 1.8149803163119984e-08,
+      "loss": 1.0336,
+      "step": 12136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4819924222313472,
+      "learning_rate": 1.8039368828839942e-08,
+      "loss": 0.7892,
+      "step": 12137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4968492206310025,
+      "learning_rate": 1.7929270887933615e-08,
+      "loss": 0.7435,
+      "step": 12138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5138212975706289,
+      "learning_rate": 1.781950934783505e-08,
+      "loss": 0.7165,
+      "step": 12139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9518202903729691,
+      "learning_rate": 1.7710084215956104e-08,
+      "loss": 0.784,
+      "step": 12140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4884694456253735,
+      "learning_rate": 1.7600995499684193e-08,
+      "loss": 0.7749,
+      "step": 12141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.47020519026117,
+      "learning_rate": 1.749224320638676e-08,
+      "loss": 0.7358,
+      "step": 12142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6770860230917108,
+      "learning_rate": 1.7383827343405712e-08,
+      "loss": 0.7223,
+      "step": 12143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6592490153695485,
+      "learning_rate": 1.7275747918062414e-08,
+      "loss": 0.8116,
+      "step": 12144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4850221318698864,
+      "learning_rate": 1.7168004937653803e-08,
+      "loss": 0.7656,
+      "step": 12145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5254932174204305,
+      "learning_rate": 1.7060598409456286e-08,
+      "loss": 0.7162,
+      "step": 12146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6630279755455382,
+      "learning_rate": 1.6953528340720726e-08,
+      "loss": 0.7856,
+      "step": 12147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4404833276850257,
+      "learning_rate": 1.684679473867745e-08,
+      "loss": 0.7861,
+      "step": 12148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5288606432333816,
+      "learning_rate": 1.6740397610533477e-08,
+      "loss": 0.7445,
+      "step": 12149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5270304084888775,
+      "learning_rate": 1.66343369634725e-08,
+      "loss": 0.7462,
+      "step": 12150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4014274947959138,
+      "learning_rate": 1.6528612804656565e-08,
+      "loss": 0.6737,
+      "step": 12151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.75149451056459,
+      "learning_rate": 1.6423225141223854e-08,
+      "loss": 1.0709,
+      "step": 12152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5714300105812147,
+      "learning_rate": 1.631817398029034e-08,
+      "loss": 0.7818,
+      "step": 12153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.761393438369206,
+      "learning_rate": 1.6213459328950355e-08,
+      "loss": 1.0507,
+      "step": 12154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4792679649349916,
+      "learning_rate": 1.6109081194273235e-08,
+      "loss": 0.7382,
+      "step": 12155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.615248952856287,
+      "learning_rate": 1.6005039583307237e-08,
+      "loss": 0.8374,
+      "step": 12156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4979722703922982,
+      "learning_rate": 1.5901334503077294e-08,
+      "loss": 0.6995,
+      "step": 12157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4820566431298845,
+      "learning_rate": 1.5797965960586693e-08,
+      "loss": 0.8018,
+      "step": 12158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5412998835312395,
+      "learning_rate": 1.5694933962814295e-08,
+      "loss": 0.8163,
+      "step": 12159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.473489735798859,
+      "learning_rate": 1.5592238516717317e-08,
+      "loss": 0.81,
+      "step": 12160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5761408308232585,
+      "learning_rate": 1.5489879629229654e-08,
+      "loss": 0.741,
+      "step": 12161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7337510317941677,
+      "learning_rate": 1.5387857307262998e-08,
+      "loss": 1.0419,
+      "step": 12162
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6268546975060227,
+      "learning_rate": 1.528617155770684e-08,
+      "loss": 0.716,
+      "step": 12163
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.586119697029699,
+      "learning_rate": 1.5184822387426246e-08,
+      "loss": 0.8348,
+      "step": 12164
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5635134674236129,
+      "learning_rate": 1.508380980326518e-08,
+      "loss": 0.8243,
+      "step": 12165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.629279359497004,
+      "learning_rate": 1.4983133812043193e-08,
+      "loss": 0.7169,
+      "step": 12166
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5289622337579658,
+      "learning_rate": 1.4882794420559843e-08,
+      "loss": 0.8071,
+      "step": 12167
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5628548278801926,
+      "learning_rate": 1.4782791635588601e-08,
+      "loss": 0.7987,
+      "step": 12168
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.399331753968418,
+      "learning_rate": 1.4683125463882952e-08,
+      "loss": 0.7418,
+      "step": 12169
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5046578779024695,
+      "learning_rate": 1.4583795912172516e-08,
+      "loss": 0.7832,
+      "step": 12170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4167722045976292,
+      "learning_rate": 1.4484802987164147e-08,
+      "loss": 0.7759,
+      "step": 12171
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4340380497918501,
+      "learning_rate": 1.4386146695541947e-08,
+      "loss": 0.7545,
+      "step": 12172
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.538427771093233,
+      "learning_rate": 1.4287827043966696e-08,
+      "loss": 0.7348,
+      "step": 12173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4646971790107552,
+      "learning_rate": 1.4189844039078638e-08,
+      "loss": 0.783,
+      "step": 12174
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7407695081387405,
+      "learning_rate": 1.4092197687492481e-08,
+      "loss": 1.0788,
+      "step": 12175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.557378403086291,
+      "learning_rate": 1.3994887995802397e-08,
+      "loss": 0.7435,
+      "step": 12176
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7582511984515554,
+      "learning_rate": 1.3897914970578685e-08,
+      "loss": 1.0737,
+      "step": 12177
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5429132344068073,
+      "learning_rate": 1.380127861836944e-08,
+      "loss": 0.7072,
+      "step": 12178
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7238813197452245,
+      "learning_rate": 1.3704978945698888e-08,
+      "loss": 1.0689,
+      "step": 12179
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7547551122783296,
+      "learning_rate": 1.360901595907016e-08,
+      "loss": 1.0833,
+      "step": 12180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5796905791865619,
+      "learning_rate": 1.3513389664963073e-08,
+      "loss": 0.7853,
+      "step": 12181
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5687554019722012,
+      "learning_rate": 1.3418100069834128e-08,
+      "loss": 0.7726,
+      "step": 12182
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7966044576258473,
+      "learning_rate": 1.3323147180117624e-08,
+      "loss": 0.8358,
+      "step": 12183
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4957682004706594,
+      "learning_rate": 1.3228531002224543e-08,
+      "loss": 0.7773,
+      "step": 12184
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5078880755380488,
+      "learning_rate": 1.3134251542544774e-08,
+      "loss": 0.7344,
+      "step": 12185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6518981545978477,
+      "learning_rate": 1.304030880744267e-08,
+      "loss": 0.7499,
+      "step": 12186
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4220032050240292,
+      "learning_rate": 1.29467028032626e-08,
+      "loss": 0.7305,
+      "step": 12187
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5724802340640325,
+      "learning_rate": 1.2853433536324512e-08,
+      "loss": 0.7265,
+      "step": 12188
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7606724698007911,
+      "learning_rate": 1.2760501012926696e-08,
+      "loss": 1.035,
+      "step": 12189
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8088481903441134,
+      "learning_rate": 1.2667905239343581e-08,
+      "loss": 0.7663,
+      "step": 12190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.704325475671645,
+      "learning_rate": 1.2575646221828497e-08,
+      "loss": 0.7927,
+      "step": 12191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7422223512058327,
+      "learning_rate": 1.248372396660924e-08,
+      "loss": 1.012,
+      "step": 12192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5346367435299728,
+      "learning_rate": 1.2392138479894178e-08,
+      "loss": 0.6972,
+      "step": 12193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.604897177068113,
+      "learning_rate": 1.2300889767866697e-08,
+      "loss": 0.772,
+      "step": 12194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5386245259233673,
+      "learning_rate": 1.220997783668798e-08,
+      "loss": 0.71,
+      "step": 12195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5163094757362094,
+      "learning_rate": 1.2119402692497007e-08,
+      "loss": 0.7425,
+      "step": 12196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5600418730945607,
+      "learning_rate": 1.2029164341409438e-08,
+      "loss": 0.755,
+      "step": 12197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5438999322038198,
+      "learning_rate": 1.193926278951818e-08,
+      "loss": 0.7851,
+      "step": 12198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9039441628649647,
+      "learning_rate": 1.1849698042893932e-08,
+      "loss": 0.7971,
+      "step": 12199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5244553125306592,
+      "learning_rate": 1.1760470107584077e-08,
+      "loss": 0.669,
+      "step": 12200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6571796498573543,
+      "learning_rate": 1.1671578989613796e-08,
+      "loss": 0.8851,
+      "step": 12201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5792420161315612,
+      "learning_rate": 1.1583024694984956e-08,
+      "loss": 0.8281,
+      "step": 12202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4954007442504227,
+      "learning_rate": 1.1494807229677218e-08,
+      "loss": 0.7824,
+      "step": 12203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5326699272644617,
+      "learning_rate": 1.1406926599646373e-08,
+      "loss": 0.7484,
+      "step": 12204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4790226908789161,
+      "learning_rate": 1.1319382810827673e-08,
+      "loss": 0.7237,
+      "step": 12205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6044290431390875,
+      "learning_rate": 1.1232175869130835e-08,
+      "loss": 0.7247,
+      "step": 12206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4967648529623887,
+      "learning_rate": 1.1145305780445036e-08,
+      "loss": 0.735,
+      "step": 12207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.399727849332947,
+      "learning_rate": 1.1058772550636699e-08,
+      "loss": 0.6726,
+      "step": 12208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.469642876023869,
+      "learning_rate": 1.0972576185547256e-08,
+      "loss": 0.682,
+      "step": 12209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5457970431485608,
+      "learning_rate": 1.0886716690997612e-08,
+      "loss": 0.7142,
+      "step": 12210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7337792759909446,
+      "learning_rate": 1.0801194072785348e-08,
+      "loss": 1.0753,
+      "step": 12211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5445360428739587,
+      "learning_rate": 1.0716008336684736e-08,
+      "loss": 0.8391,
+      "step": 12212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4321594692713804,
+      "learning_rate": 1.0631159488448395e-08,
+      "loss": 0.6916,
+      "step": 12213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5586477823684421,
+      "learning_rate": 1.0546647533804521e-08,
+      "loss": 0.7221,
+      "step": 12214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6090382372569705,
+      "learning_rate": 1.0462472478460217e-08,
+      "loss": 0.7399,
+      "step": 12215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.764550963972967,
+      "learning_rate": 1.0378634328099268e-08,
+      "loss": 1.0562,
+      "step": 12216
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7500419281453206,
+      "learning_rate": 1.0295133088382147e-08,
+      "loss": 1.0978,
+      "step": 12217
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.53997796571693,
+      "learning_rate": 1.0211968764947122e-08,
+      "loss": 0.7174,
+      "step": 12218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.467213267723149,
+      "learning_rate": 1.0129141363410256e-08,
+      "loss": 0.757,
+      "step": 12219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.561631613941726,
+      "learning_rate": 1.0046650889363741e-08,
+      "loss": 0.7382,
+      "step": 12220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5240943044044848,
+      "learning_rate": 9.964497348377012e-09,
+      "loss": 0.7563,
+      "step": 12221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.752855470745556,
+      "learning_rate": 9.882680745998408e-09,
+      "loss": 1.0589,
+      "step": 12222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.625856349735403,
+      "learning_rate": 9.801201087751843e-09,
+      "loss": 0.65,
+      "step": 12223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4978713995477744,
+      "learning_rate": 9.720058379138474e-09,
+      "loss": 0.7269,
+      "step": 12224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6235485307082411,
+      "learning_rate": 9.639252625638363e-09,
+      "loss": 0.7029,
+      "step": 12225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.456299330964164,
+      "learning_rate": 9.558783832706586e-09,
+      "loss": 0.6858,
+      "step": 12226
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4815290204210438,
+      "learning_rate": 9.478652005777134e-09,
+      "loss": 0.7101,
+      "step": 12227
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5480540597804195,
+      "learning_rate": 9.398857150260676e-09,
+      "loss": 0.7437,
+      "step": 12228
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.504194315596324,
+      "learning_rate": 9.319399271545126e-09,
+      "loss": 0.7913,
+      "step": 12229
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7205311415841864,
+      "learning_rate": 9.240278374995637e-09,
+      "loss": 1.0272,
+      "step": 12230
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7565575528862646,
+      "learning_rate": 9.161494465954046e-09,
+      "loss": 1.0857,
+      "step": 12231
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7223992819351541,
+      "learning_rate": 9.083047549741098e-09,
+      "loss": 1.0589,
+      "step": 12232
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7398082890804571,
+      "learning_rate": 9.004937631653664e-09,
+      "loss": 1.0458,
+      "step": 12233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.473484726937989,
+      "learning_rate": 8.927164716964754e-09,
+      "loss": 0.7676,
+      "step": 12234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4150576374942574,
+      "learning_rate": 8.849728810926273e-09,
+      "loss": 0.7829,
+      "step": 12235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7437032756234213,
+      "learning_rate": 8.772629918767927e-09,
+      "loss": 1.0891,
+      "step": 12236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.55966802403685,
+      "learning_rate": 8.695868045693889e-09,
+      "loss": 0.7449,
+      "step": 12237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7276919885787497,
+      "learning_rate": 8.6194431968889e-09,
+      "loss": 1.0696,
+      "step": 12238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6172237678374088,
+      "learning_rate": 8.543355377512164e-09,
+      "loss": 0.7746,
+      "step": 12239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.827159983902288,
+      "learning_rate": 8.467604592701795e-09,
+      "loss": 0.6425,
+      "step": 12240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6346821276296246,
+      "learning_rate": 8.392190847572035e-09,
+      "loss": 0.7592,
+      "step": 12241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.505159639603996,
+      "learning_rate": 8.317114147216587e-09,
+      "loss": 0.7164,
+      "step": 12242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5455081144281047,
+      "learning_rate": 8.242374496703065e-09,
+      "loss": 0.717,
+      "step": 12243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4721416534375171,
+      "learning_rate": 8.167971901079097e-09,
+      "loss": 0.707,
+      "step": 12244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4624194382802544,
+      "learning_rate": 8.093906365367888e-09,
+      "loss": 0.7247,
+      "step": 12245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.733955807661556,
+      "learning_rate": 8.02017789457099e-09,
+      "loss": 1.0643,
+      "step": 12246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5876674612042119,
+      "learning_rate": 7.946786493666647e-09,
+      "loss": 0.7661,
+      "step": 12247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4459884898297926,
+      "learning_rate": 7.873732167609782e-09,
+      "loss": 0.6675,
+      "step": 12248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4128138476298862,
+      "learning_rate": 7.801014921334227e-09,
+      "loss": 0.7762,
+      "step": 12249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5298272637513937,
+      "learning_rate": 7.728634759749387e-09,
+      "loss": 0.6915,
+      "step": 12250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5177313017399514,
+      "learning_rate": 7.656591687742465e-09,
+      "loss": 0.7556,
+      "step": 12251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5280165687731564,
+      "learning_rate": 7.584885710178457e-09,
+      "loss": 0.721,
+      "step": 12252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6622074214088207,
+      "learning_rate": 7.513516831898493e-09,
+      "loss": 0.8115,
+      "step": 12253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4622559665760364,
+      "learning_rate": 7.442485057722048e-09,
+      "loss": 0.7728,
+      "step": 12254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4417249800598522,
+      "learning_rate": 7.371790392445288e-09,
+      "loss": 0.7786,
+      "step": 12255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5158115866071773,
+      "learning_rate": 7.301432840841061e-09,
+      "loss": 0.7685,
+      "step": 12256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7558309309576888,
+      "learning_rate": 7.2314124076611205e-09,
+      "loss": 1.0598,
+      "step": 12257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.606740303559623,
+      "learning_rate": 7.161729097632797e-09,
+      "loss": 0.8409,
+      "step": 12258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.512776199757214,
+      "learning_rate": 7.092382915461215e-09,
+      "loss": 0.7463,
+      "step": 12259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.519617019758592,
+      "learning_rate": 7.0233738658292974e-09,
+      "loss": 0.7287,
+      "step": 12260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4857528512043399,
+      "learning_rate": 6.954701953395538e-09,
+      "loss": 0.7144,
+      "step": 12261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4078448056642565,
+      "learning_rate": 6.886367182798448e-09,
+      "loss": 0.7416,
+      "step": 12262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5805433124971253,
+      "learning_rate": 6.81836955865045e-09,
+      "loss": 0.7657,
+      "step": 12263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4551513256776019,
+      "learning_rate": 6.750709085544538e-09,
+      "loss": 0.6867,
+      "step": 12264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7461464435757642,
+      "learning_rate": 6.683385768047612e-09,
+      "loss": 1.0363,
+      "step": 12265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.505883029816718,
+      "learning_rate": 6.616399610707147e-09,
+      "loss": 0.693,
+      "step": 12266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4967490856542023,
+      "learning_rate": 6.5497506180450806e-09,
+      "loss": 0.8166,
+      "step": 12267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7696148995996944,
+      "learning_rate": 6.483438794562258e-09,
+      "loss": 1.0609,
+      "step": 12268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6188971110285095,
+      "learning_rate": 6.417464144736208e-09,
+      "loss": 0.8396,
+      "step": 12269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4369398514815348,
+      "learning_rate": 6.351826673021144e-09,
+      "loss": 0.8023,
+      "step": 12270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.744131829892106,
+      "learning_rate": 6.286526383849078e-09,
+      "loss": 0.7583,
+      "step": 12271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4962403839258664,
+      "learning_rate": 6.221563281630372e-09,
+      "loss": 0.7545,
+      "step": 12272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5675149444466239,
+      "learning_rate": 6.156937370750405e-09,
+      "loss": 0.7732,
+      "step": 12273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6276991159329492,
+      "learning_rate": 6.092648655572908e-09,
+      "loss": 0.7693,
+      "step": 12274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.420451375620458,
+      "learning_rate": 6.028697140438855e-09,
+      "loss": 0.7683,
+      "step": 12275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.449157585532989,
+      "learning_rate": 5.965082829667013e-09,
+      "loss": 0.8051,
+      "step": 12276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5036171976180839,
+      "learning_rate": 5.901805727552279e-09,
+      "loss": 0.7015,
+      "step": 12277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5565950931177601,
+      "learning_rate": 5.838865838366792e-09,
+      "loss": 0.6848,
+      "step": 12278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5244985592092153,
+      "learning_rate": 5.7762631663615955e-09,
+      "loss": 0.7202,
+      "step": 12279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4919569470831324,
+      "learning_rate": 5.713997715762754e-09,
+      "loss": 0.8281,
+      "step": 12280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.723380816698547,
+      "learning_rate": 5.652069490775236e-09,
+      "loss": 1.0816,
+      "step": 12281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3695862600501325,
+      "learning_rate": 5.590478495580143e-09,
+      "loss": 0.7026,
+      "step": 12282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7490532668940209,
+      "learning_rate": 5.529224734335814e-09,
+      "loss": 1.036,
+      "step": 12283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4354644292201042,
+      "learning_rate": 5.468308211179496e-09,
+      "loss": 0.6877,
+      "step": 12284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.627537603888248,
+      "learning_rate": 5.407728930223455e-09,
+      "loss": 0.7639,
+      "step": 12285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5356280054695592,
+      "learning_rate": 5.347486895558307e-09,
+      "loss": 0.7435,
+      "step": 12286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.417226413121078,
+      "learning_rate": 5.2875821112513544e-09,
+      "loss": 0.6819,
+      "step": 12287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7496895079320679,
+      "learning_rate": 5.228014581348806e-09,
+      "loss": 1.0641,
+      "step": 12288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7009314785075278,
+      "learning_rate": 5.168784309871333e-09,
+      "loss": 0.7172,
+      "step": 12289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6564218583516788,
+      "learning_rate": 5.10989130081907e-09,
+      "loss": 0.7294,
+      "step": 12290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7624380478789122,
+      "learning_rate": 5.051335558168835e-09,
+      "loss": 1.0592,
+      "step": 12291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4539798483309492,
+      "learning_rate": 4.993117085873578e-09,
+      "loss": 0.6984,
+      "step": 12292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5996949254168764,
+      "learning_rate": 4.935235887865153e-09,
+      "loss": 0.7422,
+      "step": 12293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6242460162536,
+      "learning_rate": 4.877691968051545e-09,
+      "loss": 0.7289,
+      "step": 12294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.627077630921616,
+      "learning_rate": 4.820485330317981e-09,
+      "loss": 0.7909,
+      "step": 12295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6660590276334288,
+      "learning_rate": 4.763615978526925e-09,
+      "loss": 0.7124,
+      "step": 12296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4695998113360933,
+      "learning_rate": 4.70708391651975e-09,
+      "loss": 0.7634,
+      "step": 12297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7557761399981202,
+      "learning_rate": 4.650889148112292e-09,
+      "loss": 0.819,
+      "step": 12298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5096338109168732,
+      "learning_rate": 4.595031677099293e-09,
+      "loss": 0.7463,
+      "step": 12299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4398946937485608,
+      "learning_rate": 4.539511507252181e-09,
+      "loss": 0.8178,
+      "step": 12300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7334287062633463,
+      "learning_rate": 4.484328642320734e-09,
+      "loss": 1.0539,
+      "step": 12301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.614745375485301,
+      "learning_rate": 4.429483086029751e-09,
+      "loss": 0.8122,
+      "step": 12302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5355098288615516,
+      "learning_rate": 4.37497484208349e-09,
+      "loss": 0.7743,
+      "step": 12303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7420222044383968,
+      "learning_rate": 4.320803914162341e-09,
+      "loss": 1.0718,
+      "step": 12304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5042719620185068,
+      "learning_rate": 4.266970305923379e-09,
+      "loss": 0.7242,
+      "step": 12305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.488958591933096,
+      "learning_rate": 4.213474021002029e-09,
+      "loss": 0.739,
+      "step": 12306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9413089085789392,
+      "learning_rate": 4.1603150630104005e-09,
+      "loss": 0.7143,
+      "step": 12307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5291371954056623,
+      "learning_rate": 4.1074934355384015e-09,
+      "loss": 0.664,
+      "step": 12308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5226469832861347,
+      "learning_rate": 4.055009142152066e-09,
+      "loss": 0.7468,
+      "step": 12309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.472435045984113,
+      "learning_rate": 4.002862186395229e-09,
+      "loss": 0.7149,
+      "step": 12310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.551233199829396,
+      "learning_rate": 3.95105257178896e-09,
+      "loss": 0.6887,
+      "step": 12311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4505864695931416,
+      "learning_rate": 3.8995803018321285e-09,
+      "loss": 0.7775,
+      "step": 12312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9558651676799548,
+      "learning_rate": 3.848445380000288e-09,
+      "loss": 0.757,
+      "step": 12313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4798073500410793,
+      "learning_rate": 3.7976478097451196e-09,
+      "loss": 0.8093,
+      "step": 12314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7611462589757173,
+      "learning_rate": 3.74718759449777e-09,
+      "loss": 1.0348,
+      "step": 12315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5488027863679448,
+      "learning_rate": 3.697064737664402e-09,
+      "loss": 0.7572,
+      "step": 12316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5590785671528982,
+      "learning_rate": 3.6472792426306413e-09,
+      "loss": 0.7432,
+      "step": 12317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4510692780082382,
+      "learning_rate": 3.597831112757133e-09,
+      "loss": 0.6941,
+      "step": 12318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4500311150031706,
+      "learning_rate": 3.548720351382873e-09,
+      "loss": 0.8135,
+      "step": 12319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7558188513076751,
+      "learning_rate": 3.499946961824097e-09,
+      "loss": 1.0481,
+      "step": 12320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4804567419309809,
+      "learning_rate": 3.4515109473742815e-09,
+      "loss": 0.7222,
+      "step": 12321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5024753755205242,
+      "learning_rate": 3.4034123113035888e-09,
+      "loss": 0.7078,
+      "step": 12322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5108404869825345,
+      "learning_rate": 3.3556510568599763e-09,
+      "loss": 0.7315,
+      "step": 12323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7361429246047747,
+      "learning_rate": 3.308227187268642e-09,
+      "loss": 1.056,
+      "step": 12324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.635077365001119,
+      "learning_rate": 3.261140705730914e-09,
+      "loss": 0.7758,
+      "step": 12325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7740550873080957,
+      "learning_rate": 3.214391615427026e-09,
+      "loss": 1.0498,
+      "step": 12326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.731086193580534,
+      "learning_rate": 3.167979919512787e-09,
+      "loss": 1.0604,
+      "step": 12327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4093174091429432,
+      "learning_rate": 3.1219056211229117e-09,
+      "loss": 0.7245,
+      "step": 12328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4871815680536813,
+      "learning_rate": 3.0761687233682445e-09,
+      "loss": 0.7547,
+      "step": 12329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7943795299152325,
+      "learning_rate": 3.0307692293363165e-09,
+      "loss": 0.7,
+      "step": 12330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4706743874043986,
+      "learning_rate": 2.9857071420935636e-09,
+      "loss": 0.7739,
+      "step": 12331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4529138111504463,
+      "learning_rate": 2.9409824646819962e-09,
+      "loss": 0.7916,
+      "step": 12332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4086396144610462,
+      "learning_rate": 2.8965952001214217e-09,
+      "loss": 0.7088,
+      "step": 12333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.590369034122501,
+      "learning_rate": 2.8525453514099966e-09,
+      "loss": 0.7223,
+      "step": 12334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.503029607013171,
+      "learning_rate": 2.808832921520899e-09,
+      "loss": 0.7441,
+      "step": 12335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7512640571878971,
+      "learning_rate": 2.765457913406211e-09,
+      "loss": 1.0645,
+      "step": 12336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3814650380745912,
+      "learning_rate": 2.7224203299947017e-09,
+      "loss": 0.8012,
+      "step": 12337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5183323609979942,
+      "learning_rate": 2.6797201741923797e-09,
+      "loss": 0.7867,
+      "step": 12338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4902387076345842,
+      "learning_rate": 2.637357448882494e-09,
+      "loss": 0.7897,
+      "step": 12339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7513237719229644,
+      "learning_rate": 2.595332156925534e-09,
+      "loss": 0.8393,
+      "step": 12340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5716354831052264,
+      "learning_rate": 2.5536443011586753e-09,
+      "loss": 0.7885,
+      "step": 12341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7228462327458982,
+      "learning_rate": 2.512293884396888e-09,
+      "loss": 1.0603,
+      "step": 12342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4718607566052602,
+      "learning_rate": 2.4712809094329382e-09,
+      "loss": 0.8277,
+      "step": 12343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5376839158606592,
+      "learning_rate": 2.430605379035722e-09,
+      "loss": 0.703,
+      "step": 12344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5045276046640932,
+      "learning_rate": 2.3902672959513763e-09,
+      "loss": 0.7565,
+      "step": 12345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.51695777515481,
+      "learning_rate": 2.350266662903833e-09,
+      "loss": 0.7423,
+      "step": 12346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5991948080269145,
+      "learning_rate": 2.3106034825942647e-09,
+      "loss": 0.6999,
+      "step": 12347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5413705772405415,
+      "learning_rate": 2.2712777577005297e-09,
+      "loss": 0.8805,
+      "step": 12348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5083754354977248,
+      "learning_rate": 2.232289490878281e-09,
+      "loss": 0.7604,
+      "step": 12349
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5105730557216628,
+      "learning_rate": 2.193638684759858e-09,
+      "loss": 0.6205,
+      "step": 12350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5449658289786963,
+      "learning_rate": 2.1553253419553943e-09,
+      "loss": 0.7027,
+      "step": 12351
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4684063082632595,
+      "learning_rate": 2.117349465051155e-09,
+      "loss": 0.7107,
+      "step": 12352
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6665053839933863,
+      "learning_rate": 2.079711056611755e-09,
+      "loss": 0.777,
+      "step": 12353
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4839248520617672,
+      "learning_rate": 2.04241011917905e-09,
+      "loss": 0.7693,
+      "step": 12354
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.486740195732885,
+      "learning_rate": 2.005446655271581e-09,
+      "loss": 0.7065,
+      "step": 12355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.493732473635201,
+      "learning_rate": 1.9688206673845746e-09,
+      "loss": 0.749,
+      "step": 12356
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.690767987170318,
+      "learning_rate": 1.9325321579916066e-09,
+      "loss": 0.7226,
+      "step": 12357
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4783019805803208,
+      "learning_rate": 1.8965811295423854e-09,
+      "loss": 0.6912,
+      "step": 12358
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7579287411647264,
+      "learning_rate": 1.8609675844655228e-09,
+      "loss": 1.0781,
+      "step": 12359
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.406472877462492,
+      "learning_rate": 1.8256915251646524e-09,
+      "loss": 0.7407,
+      "step": 12360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5482737380476663,
+      "learning_rate": 1.7907529540223123e-09,
+      "loss": 0.7179,
+      "step": 12361
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.383327947631836,
+      "learning_rate": 1.7561518733977267e-09,
+      "loss": 0.7153,
+      "step": 12362
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5136647498613403,
+      "learning_rate": 1.7218882856262498e-09,
+      "loss": 0.7669,
+      "step": 12363
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6695650755174116,
+      "learning_rate": 1.6879621930226963e-09,
+      "loss": 0.6661,
+      "step": 12364
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3339222890611686,
+      "learning_rate": 1.6543735978769016e-09,
+      "loss": 0.7798,
+      "step": 12365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4379200470380373,
+      "learning_rate": 1.621122502457606e-09,
+      "loss": 0.7277,
+      "step": 12366
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5596792388106255,
+      "learning_rate": 1.5882089090091257e-09,
+      "loss": 0.807,
+      "step": 12367
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5379239496515134,
+      "learning_rate": 1.555632819754682e-09,
+      "loss": 0.7234,
+      "step": 12368
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.472812001735286,
+      "learning_rate": 1.523394236893072e-09,
+      "loss": 0.7131,
+      "step": 12369
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5090647393358532,
+      "learning_rate": 1.491493162601998e-09,
+      "loss": 0.7962,
+      "step": 12370
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.554072329024694,
+      "learning_rate": 1.4599295990352924e-09,
+      "loss": 0.7458,
+      "step": 12371
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5105888924677444,
+      "learning_rate": 1.4287035483234734e-09,
+      "loss": 0.7906,
+      "step": 12372
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.755455342623131,
+      "learning_rate": 1.3978150125759649e-09,
+      "loss": 1.0282,
+      "step": 12373
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.615004117411443,
+      "learning_rate": 1.3672639938777655e-09,
+      "loss": 0.705,
+      "step": 12374
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4652872621022046,
+      "learning_rate": 1.3370504942922246e-09,
+      "loss": 0.7823,
+      "step": 12375
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6091127744552964,
+      "learning_rate": 1.3071745158588222e-09,
+      "loss": 0.7278,
+      "step": 12376
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5352274209639358,
+      "learning_rate": 1.2776360605953887e-09,
+      "loss": 0.7839,
+      "step": 12377
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4072945170315727,
+      "learning_rate": 1.2484351304958841e-09,
+      "loss": 0.6929,
+      "step": 12378
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5433956173817382,
+      "learning_rate": 1.2195717275326201e-09,
+      "loss": 0.8055,
+      "step": 12379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4568118489246025,
+      "learning_rate": 1.1910458536545932e-09,
+      "loss": 0.7475,
+      "step": 12380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.554751420041771,
+      "learning_rate": 1.16285751078693e-09,
+      "loss": 0.7412,
+      "step": 12381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.385779361034528,
+      "learning_rate": 1.135006700834218e-09,
+      "loss": 0.7669,
+      "step": 12382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5270886371657446,
+      "learning_rate": 1.107493425676065e-09,
+      "loss": 0.7365,
+      "step": 12383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.419052047834366,
+      "learning_rate": 1.080317687170984e-09,
+      "loss": 0.6503,
+      "step": 12384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5640293123782238,
+      "learning_rate": 1.0534794871536192e-09,
+      "loss": 0.748,
+      "step": 12385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3438032628627603,
+      "learning_rate": 1.026978827435854e-09,
+      "loss": 0.7754,
+      "step": 12386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5077917461770538,
+      "learning_rate": 1.0008157098073678e-09,
+      "loss": 0.8437,
+      "step": 12387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3589242801510977,
+      "learning_rate": 9.749901360345259e-10,
+      "loss": 0.7538,
+      "step": 12388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5695586354462252,
+      "learning_rate": 9.495021078614885e-10,
+      "loss": 0.7758,
+      "step": 12389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4540572797076856,
+      "learning_rate": 9.243516270091013e-10,
+      "loss": 0.7178,
+      "step": 12390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5701355305410813,
+      "learning_rate": 8.995386951754503e-10,
+      "loss": 0.7607,
+      "step": 12391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5546246274554316,
+      "learning_rate": 8.750633140364173e-10,
+      "loss": 0.6916,
+      "step": 12392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.508841982687282,
+      "learning_rate": 8.509254852440141e-10,
+      "loss": 0.8504,
+      "step": 12393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5253749977451414,
+      "learning_rate": 8.271252104286032e-10,
+      "loss": 0.7436,
+      "step": 12394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4940266725044131,
+      "learning_rate": 8.036624911966773e-10,
+      "loss": 0.6499,
+      "step": 12395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5403224290178037,
+      "learning_rate": 7.805373291330797e-10,
+      "loss": 0.7586,
+      "step": 12396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7716576136903933,
+      "learning_rate": 7.577497257987842e-10,
+      "loss": 1.0623,
+      "step": 12397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5126627112817044,
+      "learning_rate": 7.352996827325598e-10,
+      "loss": 0.7789,
+      "step": 12398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.46529754912899,
+      "learning_rate": 7.131872014509711e-10,
+      "loss": 0.713,
+      "step": 12399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5364920105125914,
+      "learning_rate": 6.914122834461578e-10,
+      "loss": 0.7531,
+      "step": 12400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5392092102316044,
+      "learning_rate": 6.699749301886105e-10,
+      "loss": 0.7368,
+      "step": 12401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.563051643212222,
+      "learning_rate": 6.488751431266149e-10,
+      "loss": 0.7665,
+      "step": 12402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5778020215977786,
+      "learning_rate": 6.281129236834771e-10,
+      "loss": 0.7361,
+      "step": 12403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5131734586385042,
+      "learning_rate": 6.076882732625189e-10,
+      "loss": 0.7586,
+      "step": 12404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5682417403069193,
+      "learning_rate": 5.876011932420822e-10,
+      "loss": 0.7201,
+      "step": 12405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4273772435684127,
+      "learning_rate": 5.678516849788596e-10,
+      "loss": 0.7603,
+      "step": 12406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6289370658945002,
+      "learning_rate": 5.484397498056737e-10,
+      "loss": 0.7714,
+      "step": 12407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.581494306729302,
+      "learning_rate": 5.293653890342531e-10,
+      "loss": 0.6706,
+      "step": 12408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5645787385282115,
+      "learning_rate": 5.106286039519015e-10,
+      "loss": 0.7714,
+      "step": 12409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4123927399625205,
+      "learning_rate": 4.922293958237179e-10,
+      "loss": 0.7601,
+      "step": 12410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7379264995326732,
+      "learning_rate": 4.74167765892597e-10,
+      "loss": 1.0322,
+      "step": 12411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7369683261807992,
+      "learning_rate": 4.5644371537756363e-10,
+      "loss": 1.0672,
+      "step": 12412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.649078715418298,
+      "learning_rate": 4.390572454759934e-10,
+      "loss": 0.8638,
+      "step": 12413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7412885669364163,
+      "learning_rate": 4.220083573608369e-10,
+      "loss": 1.032,
+      "step": 12414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4385200448089004,
+      "learning_rate": 4.0529705218450565e-10,
+      "loss": 0.6634,
+      "step": 12415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7443068490867211,
+      "learning_rate": 3.8892333107443114e-10,
+      "loss": 1.0158,
+      "step": 12416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6079441955540794,
+      "learning_rate": 3.7288719513639547e-10,
+      "loss": 0.7567,
+      "step": 12417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4130708091305078,
+      "learning_rate": 3.5718864545397636e-10,
+      "loss": 0.6372,
+      "step": 12418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5859746330420321,
+      "learning_rate": 3.4182768308577143e-10,
+      "loss": 0.7583,
+      "step": 12419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4701033062344357,
+      "learning_rate": 3.268043090703943e-10,
+      "loss": 0.703,
+      "step": 12420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4179732202659425,
+      "learning_rate": 3.121185244214786e-10,
+      "loss": 0.7286,
+      "step": 12421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7472972937964611,
+      "learning_rate": 2.9777033013045354e-10,
+      "loss": 1.025,
+      "step": 12422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5326423140331222,
+      "learning_rate": 2.8375972716709887e-10,
+      "loss": 0.7196,
+      "step": 12423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5818943315646454,
+      "learning_rate": 2.7008671647621443e-10,
+      "loss": 0.818,
+      "step": 12424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5629073842611907,
+      "learning_rate": 2.5675129898206086e-10,
+      "loss": 0.706,
+      "step": 12425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5064179116304723,
+      "learning_rate": 2.437534755844739e-10,
+      "loss": 0.7921,
+      "step": 12426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.435174155150378,
+      "learning_rate": 2.3109324716108494e-10,
+      "loss": 0.7652,
+      "step": 12427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7633479043862755,
+      "learning_rate": 2.1877061456732073e-10,
+      "loss": 1.0864,
+      "step": 12428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.607472226370612,
+      "learning_rate": 2.0678557863473837e-10,
+      "loss": 0.901,
+      "step": 12429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.040320903586556,
+      "learning_rate": 1.951381401726904e-10,
+      "loss": 0.7566,
+      "step": 12430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4686642639889502,
+      "learning_rate": 1.8382829996776986e-10,
+      "loss": 0.7557,
+      "step": 12431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.46272487402183,
+      "learning_rate": 1.7285605878325507e-10,
+      "loss": 0.789,
+      "step": 12432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.591254863078844,
+      "learning_rate": 1.622214173602199e-10,
+      "loss": 0.6589,
+      "step": 12433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.604645232425831,
+      "learning_rate": 1.5192437641753376e-10,
+      "loss": 0.6246,
+      "step": 12434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3995279883125502,
+      "learning_rate": 1.4196493664908607e-10,
+      "loss": 0.6972,
+      "step": 12435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.453823331467375,
+      "learning_rate": 1.3234309872822703e-10,
+      "loss": 0.7527,
+      "step": 12436
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8064464675817895,
+      "learning_rate": 1.2305886330443716e-10,
+      "loss": 0.8389,
+      "step": 12437
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4556165441019453,
+      "learning_rate": 1.1411223100443735e-10,
+      "loss": 0.6374,
+      "step": 12438
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7506123674036117,
+      "learning_rate": 1.0550320243274403e-10,
+      "loss": 1.0391,
+      "step": 12439
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.428876842869932,
+      "learning_rate": 9.723177817055896e-11,
+      "loss": 0.6849,
+      "step": 12440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.450156566962703,
+      "learning_rate": 8.92979587757692e-11,
+      "loss": 0.6504,
+      "step": 12441
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5118516833768456,
+      "learning_rate": 8.170174478516756e-11,
+      "loss": 0.845,
+      "step": 12442
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6105020279930118,
+      "learning_rate": 7.44431367105669e-11,
+      "loss": 0.7797,
+      "step": 12443
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6155013061660637,
+      "learning_rate": 6.752213504268579e-11,
+      "loss": 0.8018,
+      "step": 12444
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7001580338741185,
+      "learning_rate": 6.09387402489281e-11,
+      "loss": 0.7511,
+      "step": 12445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3942888866988503,
+      "learning_rate": 5.4692952773938155e-11,
+      "loss": 0.704,
+      "step": 12446
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5279643983982922,
+      "learning_rate": 4.87847730384905e-11,
+      "loss": 0.7237,
+      "step": 12447
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.605312948576224,
+      "learning_rate": 4.32142014422654e-11,
+      "loss": 0.77,
+      "step": 12448
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6121600648041308,
+      "learning_rate": 3.798123836162848e-11,
+      "loss": 0.7497,
+      "step": 12449
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6002711686780942,
+      "learning_rate": 3.3085884149630654e-11,
+      "loss": 0.803,
+      "step": 12450
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5475687904048794,
+      "learning_rate": 2.8528139136563272e-11,
+      "loss": 0.7028,
+      "step": 12451
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4948529313309675,
+      "learning_rate": 2.4308003630513222e-11,
+      "loss": 0.7782,
+      "step": 12452
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.519321225948849,
+      "learning_rate": 2.042547791625271e-11,
+      "loss": 0.7944,
+      "step": 12453
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6002422473651947,
+      "learning_rate": 1.6880562255794375e-11,
+      "loss": 0.7763,
+      "step": 12454
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.580476020350656,
+      "learning_rate": 1.3673256889501496e-11,
+      "loss": 0.7203,
+      "step": 12455
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4724385027106994,
+      "learning_rate": 1.080356203220223e-11,
+      "loss": 0.6918,
+      "step": 12456
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4346772718229068,
+      "learning_rate": 8.27147787929583e-12,
+      "loss": 0.7416,
+      "step": 12457
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6017589775414862,
+      "learning_rate": 6.077004600646419e-12,
+      "loss": 0.7453,
+      "step": 12458
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5739089051251347,
+      "learning_rate": 4.220142345023881e-12,
+      "loss": 0.714,
+      "step": 12459
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6094207268184746,
+      "learning_rate": 2.700891237328307e-12,
+      "loss": 0.725,
+      "step": 12460
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5980993674607418,
+      "learning_rate": 1.5192513808104381e-12,
+      "loss": 0.8227,
+      "step": 12461
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5140374223815996,
+      "learning_rate": 6.752228548512208e-13,
+      "loss": 0.7531,
+      "step": 12462
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7687265906311198,
+      "learning_rate": 1.6880571662714064e-13,
+      "loss": 1.0721,
+      "step": 12463
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4548610715303427,
+      "learning_rate": 0.0,
+      "loss": 0.6391,
+      "step": 12464
+    },
+    {
+      "epoch": 1.0,
+      "step": 12464,
+      "total_flos": 7038884969086976.0,
+      "train_loss": 0.8383206710774112,
+      "train_runtime": 88593.6851,
+      "train_samples_per_second": 36.017,
+      "train_steps_per_second": 0.141
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12464,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 7038884969086976.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}