diff --git "a/final_qwen_f.json" "b/final_qwen_f.json"
new file mode 100644--- /dev/null
+++ "b/final_qwen_f.json"
@@ -0,0 +1,42646 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999589372972529,
+  "eval_steps": 500,
+  "global_step": 6088,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 124.91422296568267,
+      "learning_rate": 1.0928961748633881e-07,
+      "loss": 2.8678,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 129.26119700373192,
+      "learning_rate": 2.1857923497267762e-07,
+      "loss": 3.0253,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 111.1662959350857,
+      "learning_rate": 3.278688524590164e-07,
+      "loss": 2.8913,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 93.48384555462891,
+      "learning_rate": 4.3715846994535524e-07,
+      "loss": 2.8807,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 105.36479378390489,
+      "learning_rate": 5.46448087431694e-07,
+      "loss": 3.0221,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 85.38432037633636,
+      "learning_rate": 6.557377049180328e-07,
+      "loss": 2.8835,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 101.98717428722452,
+      "learning_rate": 7.650273224043716e-07,
+      "loss": 2.7392,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 107.71690758035587,
+      "learning_rate": 8.743169398907105e-07,
+      "loss": 2.6117,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 89.70063914160424,
+      "learning_rate": 9.836065573770493e-07,
+      "loss": 2.6908,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.2476827440964,
+      "learning_rate": 1.092896174863388e-06,
+      "loss": 2.0703,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 67.65403685548536,
+      "learning_rate": 1.2021857923497268e-06,
+      "loss": 2.1587,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.367796751977151,
+      "learning_rate": 1.3114754098360657e-06,
+      "loss": 1.0399,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 55.46848050878616,
+      "learning_rate": 1.4207650273224043e-06,
+      "loss": 1.9881,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.693112867620982,
+      "learning_rate": 1.5300546448087432e-06,
+      "loss": 0.9603,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.0570263156503,
+      "learning_rate": 1.6393442622950819e-06,
+      "loss": 1.4842,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 27.772872845486685,
+      "learning_rate": 1.748633879781421e-06,
+      "loss": 1.5673,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.18948653712698,
+      "learning_rate": 1.8579234972677599e-06,
+      "loss": 1.5187,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.849181815277223,
+      "learning_rate": 1.9672131147540985e-06,
+      "loss": 1.2983,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.374247020732028,
+      "learning_rate": 2.0765027322404376e-06,
+      "loss": 1.0801,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.709977543601987,
+      "learning_rate": 2.185792349726776e-06,
+      "loss": 0.7882,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.872012420378546,
+      "learning_rate": 2.295081967213115e-06,
+      "loss": 0.9737,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.693594763668879,
+      "learning_rate": 2.4043715846994536e-06,
+      "loss": 1.1161,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.707194251973675,
+      "learning_rate": 2.5136612021857927e-06,
+      "loss": 0.7324,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4570853045424337,
+      "learning_rate": 2.6229508196721314e-06,
+      "loss": 0.9268,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.334244582193526,
+      "learning_rate": 2.7322404371584705e-06,
+      "loss": 0.7205,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5459021379973916,
+      "learning_rate": 2.8415300546448087e-06,
+      "loss": 1.026,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.138585625547749,
+      "learning_rate": 2.9508196721311478e-06,
+      "loss": 0.9478,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3505240476053455,
+      "learning_rate": 3.0601092896174864e-06,
+      "loss": 0.9829,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.489463154905634,
+      "learning_rate": 3.1693989071038255e-06,
+      "loss": 0.8952,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8855915210413334,
+      "learning_rate": 3.2786885245901638e-06,
+      "loss": 0.8309,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.956197928717932,
+      "learning_rate": 3.387978142076503e-06,
+      "loss": 0.8779,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4250446550064835,
+      "learning_rate": 3.497267759562842e-06,
+      "loss": 0.8184,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4668320128753973,
+      "learning_rate": 3.6065573770491806e-06,
+      "loss": 0.8518,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.468915545975964,
+      "learning_rate": 3.7158469945355197e-06,
+      "loss": 0.8357,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2377789013130953,
+      "learning_rate": 3.825136612021858e-06,
+      "loss": 0.7452,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0231637236966162,
+      "learning_rate": 3.934426229508197e-06,
+      "loss": 0.8431,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.443372995560802,
+      "learning_rate": 4.043715846994536e-06,
+      "loss": 0.8716,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.209691816826,
+      "learning_rate": 4.153005464480875e-06,
+      "loss": 0.7917,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9854963436243693,
+      "learning_rate": 4.2622950819672135e-06,
+      "loss": 0.8147,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9932479458157422,
+      "learning_rate": 4.371584699453552e-06,
+      "loss": 0.6884,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.664913776836722,
+      "learning_rate": 4.480874316939891e-06,
+      "loss": 0.6828,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1499479821555436,
+      "learning_rate": 4.59016393442623e-06,
+      "loss": 0.7388,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.48649766188542,
+      "learning_rate": 4.699453551912569e-06,
+      "loss": 0.7236,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8418004288239636,
+      "learning_rate": 4.808743169398907e-06,
+      "loss": 0.7094,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1763880409861414,
+      "learning_rate": 4.918032786885246e-06,
+      "loss": 0.8259,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.177132654795525,
+      "learning_rate": 5.027322404371585e-06,
+      "loss": 0.7197,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2860227946208238,
+      "learning_rate": 5.1366120218579245e-06,
+      "loss": 0.6733,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.95072222572888,
+      "learning_rate": 5.245901639344263e-06,
+      "loss": 0.7829,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1375216057471786,
+      "learning_rate": 5.355191256830602e-06,
+      "loss": 0.7692,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.245087646583787,
+      "learning_rate": 5.464480874316941e-06,
+      "loss": 0.7275,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.231867850038757,
+      "learning_rate": 5.573770491803278e-06,
+      "loss": 0.7303,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9493216782867173,
+      "learning_rate": 5.683060109289617e-06,
+      "loss": 0.8101,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.739113317234144,
+      "learning_rate": 5.7923497267759565e-06,
+      "loss": 0.7329,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.125767126353231,
+      "learning_rate": 5.9016393442622956e-06,
+      "loss": 0.7475,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6938928984821926,
+      "learning_rate": 6.010928961748635e-06,
+      "loss": 0.7655,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.154858842607843,
+      "learning_rate": 6.120218579234973e-06,
+      "loss": 0.7603,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1576586180292523,
+      "learning_rate": 6.229508196721312e-06,
+      "loss": 0.6622,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.559429521579539,
+      "learning_rate": 6.338797814207651e-06,
+      "loss": 0.8148,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5346928976239487,
+      "learning_rate": 6.44808743169399e-06,
+      "loss": 0.6621,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.806663859740348,
+      "learning_rate": 6.5573770491803276e-06,
+      "loss": 0.7498,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.052555195795816,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.7273,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.422529381681842,
+      "learning_rate": 6.775956284153006e-06,
+      "loss": 0.7473,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3147664721821157,
+      "learning_rate": 6.885245901639345e-06,
+      "loss": 0.7306,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.304040870544587,
+      "learning_rate": 6.994535519125684e-06,
+      "loss": 0.7212,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3122870044021284,
+      "learning_rate": 7.103825136612022e-06,
+      "loss": 0.7867,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2420217829623206,
+      "learning_rate": 7.213114754098361e-06,
+      "loss": 0.7175,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.471957258687583,
+      "learning_rate": 7.3224043715847e-06,
+      "loss": 0.7548,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.500280200600227,
+      "learning_rate": 7.4316939890710394e-06,
+      "loss": 0.7336,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.293773968782794,
+      "learning_rate": 7.540983606557377e-06,
+      "loss": 0.6471,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.667098489000959,
+      "learning_rate": 7.650273224043716e-06,
+      "loss": 0.7274,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9406122866057136,
+      "learning_rate": 7.759562841530056e-06,
+      "loss": 0.6606,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6741871659872425,
+      "learning_rate": 7.868852459016394e-06,
+      "loss": 0.6769,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.644570616717336,
+      "learning_rate": 7.978142076502732e-06,
+      "loss": 0.7779,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.576234854691406,
+      "learning_rate": 8.087431693989072e-06,
+      "loss": 0.6949,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6588160570189476,
+      "learning_rate": 8.19672131147541e-06,
+      "loss": 0.6045,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2950534681407913,
+      "learning_rate": 8.30601092896175e-06,
+      "loss": 0.7367,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1039093569794862,
+      "learning_rate": 8.415300546448089e-06,
+      "loss": 0.7307,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2119779539688444,
+      "learning_rate": 8.524590163934427e-06,
+      "loss": 0.6565,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1359809464419426,
+      "learning_rate": 8.633879781420765e-06,
+      "loss": 0.688,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7873896787885015,
+      "learning_rate": 8.743169398907103e-06,
+      "loss": 0.7126,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6512522202046815,
+      "learning_rate": 8.852459016393443e-06,
+      "loss": 0.7768,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.072299245544699,
+      "learning_rate": 8.961748633879782e-06,
+      "loss": 0.6434,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.494963838140089,
+      "learning_rate": 9.071038251366122e-06,
+      "loss": 0.6524,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3806119002537938,
+      "learning_rate": 9.18032786885246e-06,
+      "loss": 0.7801,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8068463194562834,
+      "learning_rate": 9.2896174863388e-06,
+      "loss": 0.6928,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2110464813693658,
+      "learning_rate": 9.398907103825138e-06,
+      "loss": 0.7059,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8921687835556087,
+      "learning_rate": 9.508196721311476e-06,
+      "loss": 0.6653,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.539373886489346,
+      "learning_rate": 9.617486338797814e-06,
+      "loss": 0.8046,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6011786201729907,
+      "learning_rate": 9.726775956284153e-06,
+      "loss": 0.6815,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4289960831068154,
+      "learning_rate": 9.836065573770493e-06,
+      "loss": 0.7711,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.367116514803818,
+      "learning_rate": 9.945355191256831e-06,
+      "loss": 0.7676,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.058991282012093,
+      "learning_rate": 1.005464480874317e-05,
+      "loss": 0.6858,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.042490782057154,
+      "learning_rate": 1.0163934426229509e-05,
+      "loss": 0.7288,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4727393084192837,
+      "learning_rate": 1.0273224043715849e-05,
+      "loss": 0.7375,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4361077512225693,
+      "learning_rate": 1.0382513661202187e-05,
+      "loss": 0.6828,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.199248438494068,
+      "learning_rate": 1.0491803278688525e-05,
+      "loss": 0.6903,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0976306322965823,
+      "learning_rate": 1.0601092896174865e-05,
+      "loss": 0.6563,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3922381926228584,
+      "learning_rate": 1.0710382513661204e-05,
+      "loss": 0.7122,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.114880049495975,
+      "learning_rate": 1.0819672131147544e-05,
+      "loss": 0.6399,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2403220400457415,
+      "learning_rate": 1.0928961748633882e-05,
+      "loss": 0.6602,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8458458971043683,
+      "learning_rate": 1.1038251366120218e-05,
+      "loss": 0.6843,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.38588192830122,
+      "learning_rate": 1.1147540983606557e-05,
+      "loss": 0.6793,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5859287053691977,
+      "learning_rate": 1.1256830601092897e-05,
+      "loss": 0.6382,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6332128120429155,
+      "learning_rate": 1.1366120218579235e-05,
+      "loss": 0.7626,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8502590207430667,
+      "learning_rate": 1.1475409836065575e-05,
+      "loss": 0.6591,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.225894169878997,
+      "learning_rate": 1.1584699453551913e-05,
+      "loss": 0.7788,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.765382087216269,
+      "learning_rate": 1.1693989071038251e-05,
+      "loss": 0.6506,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5682905807373775,
+      "learning_rate": 1.1803278688524591e-05,
+      "loss": 0.6703,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.397795607871973,
+      "learning_rate": 1.191256830601093e-05,
+      "loss": 0.6714,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8262209711534458,
+      "learning_rate": 1.202185792349727e-05,
+      "loss": 0.6007,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.623578944019624,
+      "learning_rate": 1.2131147540983608e-05,
+      "loss": 0.6739,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4066461751264643,
+      "learning_rate": 1.2240437158469946e-05,
+      "loss": 0.623,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.343873517454208,
+      "learning_rate": 1.2349726775956286e-05,
+      "loss": 0.6829,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.783385202973862,
+      "learning_rate": 1.2459016393442624e-05,
+      "loss": 0.6218,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3577833772192323,
+      "learning_rate": 1.2568306010928964e-05,
+      "loss": 0.645,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.388475726941176,
+      "learning_rate": 1.2677595628415302e-05,
+      "loss": 0.7175,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.794823889469251,
+      "learning_rate": 1.2786885245901642e-05,
+      "loss": 0.6714,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.817073927028771,
+      "learning_rate": 1.289617486338798e-05,
+      "loss": 0.6624,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6514517612367245,
+      "learning_rate": 1.3005464480874317e-05,
+      "loss": 0.7726,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4674554382672618,
+      "learning_rate": 1.3114754098360655e-05,
+      "loss": 0.7085,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.172835720963696,
+      "learning_rate": 1.3224043715846995e-05,
+      "loss": 0.6675,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.737823969170508,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8285,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1612673132409816,
+      "learning_rate": 1.3442622950819673e-05,
+      "loss": 0.6393,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1713544298269536,
+      "learning_rate": 1.3551912568306011e-05,
+      "loss": 0.7049,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0160650149980213,
+      "learning_rate": 1.366120218579235e-05,
+      "loss": 0.6854,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5224464399453637,
+      "learning_rate": 1.377049180327869e-05,
+      "loss": 0.6712,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.208313769874623,
+      "learning_rate": 1.3879781420765028e-05,
+      "loss": 0.6758,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4846590168227203,
+      "learning_rate": 1.3989071038251368e-05,
+      "loss": 0.6042,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9307952355985067,
+      "learning_rate": 1.4098360655737706e-05,
+      "loss": 0.7783,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3934776523856622,
+      "learning_rate": 1.4207650273224044e-05,
+      "loss": 0.7223,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5612849608094033,
+      "learning_rate": 1.4316939890710384e-05,
+      "loss": 0.6912,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3641729942484733,
+      "learning_rate": 1.4426229508196722e-05,
+      "loss": 0.7694,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6183644033757356,
+      "learning_rate": 1.4535519125683062e-05,
+      "loss": 0.6912,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6466504676073974,
+      "learning_rate": 1.46448087431694e-05,
+      "loss": 0.6799,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.891267722925287,
+      "learning_rate": 1.4754098360655739e-05,
+      "loss": 0.6845,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.076879287953429,
+      "learning_rate": 1.4863387978142079e-05,
+      "loss": 0.6616,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1553465336686473,
+      "learning_rate": 1.4972677595628417e-05,
+      "loss": 0.6897,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5855992143512494,
+      "learning_rate": 1.5081967213114754e-05,
+      "loss": 0.6436,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.501845183205708,
+      "learning_rate": 1.5191256830601094e-05,
+      "loss": 0.7119,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9560827919992483,
+      "learning_rate": 1.5300546448087432e-05,
+      "loss": 0.5899,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.084191350309575,
+      "learning_rate": 1.5409836065573772e-05,
+      "loss": 0.7404,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2248025800437263,
+      "learning_rate": 1.551912568306011e-05,
+      "loss": 0.6738,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7919991732187364,
+      "learning_rate": 1.5628415300546448e-05,
+      "loss": 0.6997,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.078068890326367,
+      "learning_rate": 1.5737704918032788e-05,
+      "loss": 0.8073,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9891495101089394,
+      "learning_rate": 1.5846994535519128e-05,
+      "loss": 0.6831,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5013755768773835,
+      "learning_rate": 1.5956284153005465e-05,
+      "loss": 0.6648,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7095739741536264,
+      "learning_rate": 1.6065573770491805e-05,
+      "loss": 0.7117,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5574974823527414,
+      "learning_rate": 1.6174863387978145e-05,
+      "loss": 0.6776,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.226905034196009,
+      "learning_rate": 1.628415300546448e-05,
+      "loss": 0.6139,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4559138834132783,
+      "learning_rate": 1.639344262295082e-05,
+      "loss": 0.6972,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.011539543809936,
+      "learning_rate": 1.650273224043716e-05,
+      "loss": 0.6201,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.844795112209871,
+      "learning_rate": 1.66120218579235e-05,
+      "loss": 0.7332,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2276049480857023,
+      "learning_rate": 1.6721311475409837e-05,
+      "loss": 0.6749,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.302829022684234,
+      "learning_rate": 1.6830601092896177e-05,
+      "loss": 0.6549,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.804539229363819,
+      "learning_rate": 1.6939890710382517e-05,
+      "loss": 0.7303,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.142031564776917,
+      "learning_rate": 1.7049180327868854e-05,
+      "loss": 0.5977,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0705365725161475,
+      "learning_rate": 1.715846994535519e-05,
+      "loss": 0.6483,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.650896318369193,
+      "learning_rate": 1.726775956284153e-05,
+      "loss": 0.5602,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.369310425143655,
+      "learning_rate": 1.737704918032787e-05,
+      "loss": 0.6349,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.74274606338942,
+      "learning_rate": 1.7486338797814207e-05,
+      "loss": 0.7051,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0593212022116556,
+      "learning_rate": 1.7595628415300547e-05,
+      "loss": 0.7693,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.767283862549674,
+      "learning_rate": 1.7704918032786887e-05,
+      "loss": 0.6915,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3920863352001875,
+      "learning_rate": 1.7814207650273227e-05,
+      "loss": 0.7924,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.681227765938785,
+      "learning_rate": 1.7923497267759563e-05,
+      "loss": 0.563,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.227485027373034,
+      "learning_rate": 1.8032786885245903e-05,
+      "loss": 0.6726,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.837118766178178,
+      "learning_rate": 1.8142076502732243e-05,
+      "loss": 0.6637,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3439661054048644,
+      "learning_rate": 1.825136612021858e-05,
+      "loss": 0.5313,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.452750533292834,
+      "learning_rate": 1.836065573770492e-05,
+      "loss": 0.7498,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.151043732160486,
+      "learning_rate": 1.846994535519126e-05,
+      "loss": 0.673,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.001234276634474,
+      "learning_rate": 1.85792349726776e-05,
+      "loss": 0.6234,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4617378053393004,
+      "learning_rate": 1.8688524590163936e-05,
+      "loss": 0.7483,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5306289032485365,
+      "learning_rate": 1.8797814207650276e-05,
+      "loss": 0.7391,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.561707135969862,
+      "learning_rate": 1.8907103825136616e-05,
+      "loss": 0.5882,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5616303426729408,
+      "learning_rate": 1.9016393442622952e-05,
+      "loss": 0.7233,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.244970716824333,
+      "learning_rate": 1.912568306010929e-05,
+      "loss": 0.588,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0795333322295027,
+      "learning_rate": 1.923497267759563e-05,
+      "loss": 0.6257,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6058741866192743,
+      "learning_rate": 1.934426229508197e-05,
+      "loss": 0.6245,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3784566974105483,
+      "learning_rate": 1.9453551912568305e-05,
+      "loss": 0.7043,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8906394374263216,
+      "learning_rate": 1.9562841530054645e-05,
+      "loss": 0.6696,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7674792664925154,
+      "learning_rate": 1.9672131147540985e-05,
+      "loss": 0.6962,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.159088905005779,
+      "learning_rate": 1.9781420765027325e-05,
+      "loss": 0.7846,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2361074026288494,
+      "learning_rate": 1.9890710382513662e-05,
+      "loss": 0.6908,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5554839716230022,
+      "learning_rate": 2e-05,
+      "loss": 0.6634,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2452964182399913,
+      "learning_rate": 1.999999858476052e-05,
+      "loss": 0.6385,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0677303993061225,
+      "learning_rate": 1.999999433904248e-05,
+      "loss": 0.6954,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4439739451745814,
+      "learning_rate": 1.999998726284708e-05,
+      "loss": 0.716,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.155494046370752,
+      "learning_rate": 1.999997735617632e-05,
+      "loss": 0.6652,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.42464794157875,
+      "learning_rate": 1.999996461903301e-05,
+      "loss": 0.7727,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.655071312851942,
+      "learning_rate": 1.999994905142075e-05,
+      "loss": 0.6687,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2752575680092884,
+      "learning_rate": 1.999993065334395e-05,
+      "loss": 0.7296,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.176423425061127,
+      "learning_rate": 1.999990942480782e-05,
+      "loss": 0.6563,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2421547347632202,
+      "learning_rate": 1.999988536581836e-05,
+      "loss": 0.6451,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.297022332276414,
+      "learning_rate": 1.9999858476382388e-05,
+      "loss": 0.6181,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4483627458507735,
+      "learning_rate": 1.9999828756507512e-05,
+      "loss": 0.7215,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7261972485319617,
+      "learning_rate": 1.999979620620214e-05,
+      "loss": 0.6698,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.422479564303686,
+      "learning_rate": 1.9999760825475496e-05,
+      "loss": 0.7276,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.043276637935657,
+      "learning_rate": 1.9999722614337585e-05,
+      "loss": 0.6441,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8949882244810776,
+      "learning_rate": 1.9999681572799226e-05,
+      "loss": 0.6764,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0652960249353027,
+      "learning_rate": 1.9999637700872037e-05,
+      "loss": 0.6945,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4316387808694033,
+      "learning_rate": 1.9999590998568432e-05,
+      "loss": 0.6839,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0917440478061398,
+      "learning_rate": 1.9999541465901636e-05,
+      "loss": 0.7427,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.24381527499914,
+      "learning_rate": 1.9999489102885657e-05,
+      "loss": 0.6075,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.639028055305177,
+      "learning_rate": 1.9999433909535333e-05,
+      "loss": 0.7147,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.010905166913826,
+      "learning_rate": 1.9999375885866272e-05,
+      "loss": 0.6178,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1059456586457834,
+      "learning_rate": 1.9999315031894908e-05,
+      "loss": 0.7078,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2684507792078143,
+      "learning_rate": 1.999925134763846e-05,
+      "loss": 0.7205,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.417153826272357,
+      "learning_rate": 1.9999184833114952e-05,
+      "loss": 0.7505,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.793971683080793,
+      "learning_rate": 1.9999115488343213e-05,
+      "loss": 0.7138,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9445393460267777,
+      "learning_rate": 1.9999043313342875e-05,
+      "loss": 0.5583,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 47.91550933486743,
+      "learning_rate": 1.9998968308134362e-05,
+      "loss": 0.6135,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6944053887230472,
+      "learning_rate": 1.9998890472738902e-05,
+      "loss": 0.7755,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7404534892959527,
+      "learning_rate": 1.9998809807178533e-05,
+      "loss": 0.7033,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8983404644326924,
+      "learning_rate": 1.999872631147608e-05,
+      "loss": 0.6165,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.334566102867691,
+      "learning_rate": 1.9998639985655183e-05,
+      "loss": 0.6213,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8835734551828829,
+      "learning_rate": 1.9998550829740277e-05,
+      "loss": 0.5916,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.257914470108109,
+      "learning_rate": 1.9998458843756587e-05,
+      "loss": 0.6981,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.375144093098851,
+      "learning_rate": 1.999836402773016e-05,
+      "loss": 0.7223,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.233213027098855,
+      "learning_rate": 1.999826638168783e-05,
+      "loss": 0.5931,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8364951137626127,
+      "learning_rate": 1.999816590565723e-05,
+      "loss": 0.6393,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.442137280130426,
+      "learning_rate": 1.999806259966681e-05,
+      "loss": 0.6688,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9524760211664511,
+      "learning_rate": 1.9997956463745806e-05,
+      "loss": 0.6201,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3579452459993617,
+      "learning_rate": 1.9997847497924258e-05,
+      "loss": 0.7269,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0506728529668004,
+      "learning_rate": 1.9997735702233006e-05,
+      "loss": 0.637,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.803958489441051,
+      "learning_rate": 1.99976210767037e-05,
+      "loss": 0.6442,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.796866868918954,
+      "learning_rate": 1.999750362136878e-05,
+      "loss": 0.6237,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3798872350328195,
+      "learning_rate": 1.999738333626149e-05,
+      "loss": 0.6093,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0558291684927603,
+      "learning_rate": 1.9997260221415883e-05,
+      "loss": 0.6645,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6784315233245515,
+      "learning_rate": 1.99971342768668e-05,
+      "loss": 0.722,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0921064181738918,
+      "learning_rate": 1.99970055026499e-05,
+      "loss": 0.757,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2717248966707055,
+      "learning_rate": 1.9996873898801617e-05,
+      "loss": 0.5808,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0107090061394137,
+      "learning_rate": 1.9996739465359208e-05,
+      "loss": 0.6493,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9697989936784703,
+      "learning_rate": 1.9996602202360728e-05,
+      "loss": 0.6948,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.091836679751457,
+      "learning_rate": 1.999646210984502e-05,
+      "loss": 0.6635,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.110102296167636,
+      "learning_rate": 1.999631918785175e-05,
+      "loss": 0.5514,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1501767352236874,
+      "learning_rate": 1.999617343642136e-05,
+      "loss": 0.6258,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8126193898586698,
+      "learning_rate": 1.999602485559511e-05,
+      "loss": 0.6516,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7341896858141206,
+      "learning_rate": 1.9995873445415056e-05,
+      "loss": 0.6578,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3863395174830684,
+      "learning_rate": 1.999571920592405e-05,
+      "loss": 0.6915,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7064279469335784,
+      "learning_rate": 1.9995562137165752e-05,
+      "loss": 0.6669,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9566939008635937,
+      "learning_rate": 1.9995402239184622e-05,
+      "loss": 0.682,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0137421223356107,
+      "learning_rate": 1.9995239512025916e-05,
+      "loss": 0.6471,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9131184062523627,
+      "learning_rate": 1.9995073955735695e-05,
+      "loss": 0.6283,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9204560236172767,
+      "learning_rate": 1.9994905570360817e-05,
+      "loss": 0.5305,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0926496770370133,
+      "learning_rate": 1.9994734355948943e-05,
+      "loss": 0.7489,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.906585130393534,
+      "learning_rate": 1.9994560312548538e-05,
+      "loss": 0.6316,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.146173577941802,
+      "learning_rate": 1.9994383440208867e-05,
+      "loss": 0.6326,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8793278577422377,
+      "learning_rate": 1.9994203738979984e-05,
+      "loss": 0.6652,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.237039446115609,
+      "learning_rate": 1.999402120891276e-05,
+      "loss": 0.6821,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.961652841644488,
+      "learning_rate": 1.999383585005886e-05,
+      "loss": 0.6663,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9222643222364602,
+      "learning_rate": 1.9993647662470746e-05,
+      "loss": 0.7155,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.441135335145979,
+      "learning_rate": 1.999345664620169e-05,
+      "loss": 0.6491,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9129707942206144,
+      "learning_rate": 1.999326280130575e-05,
+      "loss": 0.6294,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3579709951350334,
+      "learning_rate": 1.99930661278378e-05,
+      "loss": 0.686,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.778930568167307,
+      "learning_rate": 1.999286662585351e-05,
+      "loss": 0.5974,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.795657684162946,
+      "learning_rate": 1.9992664295409338e-05,
+      "loss": 0.6392,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0599999508329447,
+      "learning_rate": 1.999245913656256e-05,
+      "loss": 0.6683,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.366848708360587,
+      "learning_rate": 1.9992251149371253e-05,
+      "loss": 0.5873,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4144383736620503,
+      "learning_rate": 1.9992040333894273e-05,
+      "loss": 0.6523,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.093559499928844,
+      "learning_rate": 1.9991826690191304e-05,
+      "loss": 0.6669,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9483591881334004,
+      "learning_rate": 1.9991610218322804e-05,
+      "loss": 0.6804,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.30596393677097,
+      "learning_rate": 1.9991390918350054e-05,
+      "loss": 0.6818,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9175164700755054,
+      "learning_rate": 1.999116879033513e-05,
+      "loss": 0.662,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.215682288987195,
+      "learning_rate": 1.9990943834340893e-05,
+      "loss": 0.6099,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.380955451262495,
+      "learning_rate": 1.9990716050431026e-05,
+      "loss": 0.7238,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.973839858730111,
+      "learning_rate": 1.9990485438669998e-05,
+      "loss": 0.6782,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.910351731391163,
+      "learning_rate": 1.9990251999123086e-05,
+      "loss": 0.6257,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3270474086707047,
+      "learning_rate": 1.9990015731856366e-05,
+      "loss": 0.6933,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9423194272144007,
+      "learning_rate": 1.9989776636936705e-05,
+      "loss": 0.6361,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2067226985049886,
+      "learning_rate": 1.9989534714431788e-05,
+      "loss": 0.6686,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0544774041174088,
+      "learning_rate": 1.9989289964410082e-05,
+      "loss": 0.5703,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9725585264244263,
+      "learning_rate": 1.9989042386940872e-05,
+      "loss": 0.5732,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1762743219360046,
+      "learning_rate": 1.9988791982094224e-05,
+      "loss": 0.6762,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.107637076473886,
+      "learning_rate": 1.9988538749941024e-05,
+      "loss": 0.6955,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1514814534094175,
+      "learning_rate": 1.998828269055294e-05,
+      "loss": 0.5536,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3035204041063118,
+      "learning_rate": 1.9988023804002463e-05,
+      "loss": 0.6481,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2821755440062885,
+      "learning_rate": 1.9987762090362855e-05,
+      "loss": 0.6554,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5504201153659927,
+      "learning_rate": 1.9987497549708197e-05,
+      "loss": 0.7015,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.580261231184916,
+      "learning_rate": 1.9987230182113374e-05,
+      "loss": 0.6936,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9687663516579637,
+      "learning_rate": 1.998695998765406e-05,
+      "loss": 0.6972,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3613742285681285,
+      "learning_rate": 1.9986686966406728e-05,
+      "loss": 0.642,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.678585470200449,
+      "learning_rate": 1.9986411118448666e-05,
+      "loss": 0.647,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1392588122709646,
+      "learning_rate": 1.9986132443857942e-05,
+      "loss": 0.5715,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8016617447356555,
+      "learning_rate": 1.998585094271344e-05,
+      "loss": 0.6029,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7486987731696313,
+      "learning_rate": 1.9985566615094836e-05,
+      "loss": 0.6984,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0000971184812135,
+      "learning_rate": 1.998527946108261e-05,
+      "loss": 0.5244,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.450642729510117,
+      "learning_rate": 1.9984989480758038e-05,
+      "loss": 0.6619,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0037122485075596,
+      "learning_rate": 1.9984696674203204e-05,
+      "loss": 0.7625,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0481794233252666,
+      "learning_rate": 1.998440104150098e-05,
+      "loss": 0.7304,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.233235925940712,
+      "learning_rate": 1.998410258273505e-05,
+      "loss": 0.6873,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.029208621713401,
+      "learning_rate": 1.9983801297989883e-05,
+      "loss": 0.5942,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7991518978534315,
+      "learning_rate": 1.9983497187350767e-05,
+      "loss": 0.6214,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.273244083457427,
+      "learning_rate": 1.998319025090377e-05,
+      "loss": 0.6766,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.107822005813553,
+      "learning_rate": 1.998288048873578e-05,
+      "loss": 0.6324,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8466070233707572,
+      "learning_rate": 1.998256790093447e-05,
+      "loss": 0.5806,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8032119340425912,
+      "learning_rate": 1.9982252487588315e-05,
+      "loss": 0.5856,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.049604082600251,
+      "learning_rate": 1.9981934248786596e-05,
+      "loss": 0.6158,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0207844174574316,
+      "learning_rate": 1.9981613184619387e-05,
+      "loss": 0.6794,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8138232510836205,
+      "learning_rate": 1.9981289295177566e-05,
+      "loss": 0.7283,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5076381664189165,
+      "learning_rate": 1.9980962580552808e-05,
+      "loss": 0.6619,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.190885614926032,
+      "learning_rate": 1.998063304083759e-05,
+      "loss": 0.586,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9350487220442327,
+      "learning_rate": 1.9980300676125188e-05,
+      "loss": 0.649,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8938225049656199,
+      "learning_rate": 1.9979965486509676e-05,
+      "loss": 0.7037,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9731403313978366,
+      "learning_rate": 1.9979627472085927e-05,
+      "loss": 0.6749,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7902149997281007,
+      "learning_rate": 1.997928663294962e-05,
+      "loss": 0.6534,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0546680158686144,
+      "learning_rate": 1.9978942969197224e-05,
+      "loss": 0.6722,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.978504936224877,
+      "learning_rate": 1.9978596480926017e-05,
+      "loss": 0.6897,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9183914118778955,
+      "learning_rate": 1.9978247168234065e-05,
+      "loss": 0.6651,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8474590417447287,
+      "learning_rate": 1.997789503122025e-05,
+      "loss": 0.643,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1395602661720687,
+      "learning_rate": 1.9977540069984233e-05,
+      "loss": 0.6961,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1297072216008788,
+      "learning_rate": 1.9977182284626492e-05,
+      "loss": 0.6318,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.234709579087438,
+      "learning_rate": 1.9976821675248298e-05,
+      "loss": 0.684,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1938237602125072,
+      "learning_rate": 1.9976458241951715e-05,
+      "loss": 0.6417,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9343820845026296,
+      "learning_rate": 1.9976091984839616e-05,
+      "loss": 0.5885,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2012392568972694,
+      "learning_rate": 1.997572290401567e-05,
+      "loss": 0.6564,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.887153066972596,
+      "learning_rate": 1.9975350999584342e-05,
+      "loss": 0.6345,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6664035221875986,
+      "learning_rate": 1.99749762716509e-05,
+      "loss": 0.5371,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0198026750488487,
+      "learning_rate": 1.9974598720321407e-05,
+      "loss": 0.6613,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3643031045633203,
+      "learning_rate": 1.9974218345702733e-05,
+      "loss": 0.6758,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.253772413376972,
+      "learning_rate": 1.997383514790254e-05,
+      "loss": 0.7029,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.152130621968117,
+      "learning_rate": 1.9973449127029296e-05,
+      "loss": 0.6124,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6489660237865287,
+      "learning_rate": 1.9973060283192253e-05,
+      "loss": 0.6315,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.132978973363815,
+      "learning_rate": 1.997266861650148e-05,
+      "loss": 0.6359,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8781058194538227,
+      "learning_rate": 1.9972274127067838e-05,
+      "loss": 0.6498,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4885921649021103,
+      "learning_rate": 1.997187681500298e-05,
+      "loss": 0.6884,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8715203775436586,
+      "learning_rate": 1.9971476680419372e-05,
+      "loss": 0.6542,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.996059821077678,
+      "learning_rate": 1.9971073723430266e-05,
+      "loss": 0.7282,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.53457708562039,
+      "learning_rate": 1.997066794414972e-05,
+      "loss": 0.6138,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.722271845762915,
+      "learning_rate": 1.997025934269259e-05,
+      "loss": 0.6992,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.401534516490874,
+      "learning_rate": 1.9969847919174525e-05,
+      "loss": 0.6331,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9377053511580682,
+      "learning_rate": 1.9969433673711984e-05,
+      "loss": 0.7064,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9258133798579458,
+      "learning_rate": 1.9969016606422215e-05,
+      "loss": 0.6333,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7644835593691077,
+      "learning_rate": 1.996859671742327e-05,
+      "loss": 0.6427,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.098588937764492,
+      "learning_rate": 1.9968174006833996e-05,
+      "loss": 0.7746,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0273803698301482,
+      "learning_rate": 1.996774847477404e-05,
+      "loss": 0.5809,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9874946431105853,
+      "learning_rate": 1.9967320121363845e-05,
+      "loss": 0.6899,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.87714853043082,
+      "learning_rate": 1.9966888946724663e-05,
+      "loss": 0.6057,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.313758650576544,
+      "learning_rate": 1.9966454950978534e-05,
+      "loss": 0.7344,
+      "step": 337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.812543846184006,
+      "learning_rate": 1.9966018134248296e-05,
+      "loss": 0.6898,
+      "step": 338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.295146216148921,
+      "learning_rate": 1.9965578496657593e-05,
+      "loss": 0.7272,
+      "step": 339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.150715763295496,
+      "learning_rate": 1.996513603833086e-05,
+      "loss": 0.755,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5244697818040596,
+      "learning_rate": 1.996469075939334e-05,
+      "loss": 0.682,
+      "step": 341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0640478330853043,
+      "learning_rate": 1.9964242659971063e-05,
+      "loss": 0.5912,
+      "step": 342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9092895334550861,
+      "learning_rate": 1.9963791740190863e-05,
+      "loss": 0.607,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.148912326721727,
+      "learning_rate": 1.9963338000180374e-05,
+      "loss": 0.6428,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4082504540040888,
+      "learning_rate": 1.9962881440068026e-05,
+      "loss": 0.5607,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.438065663728744,
+      "learning_rate": 1.9962422059983045e-05,
+      "loss": 0.7121,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3393310863785723,
+      "learning_rate": 1.9961959860055456e-05,
+      "loss": 0.7211,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2444534285593316,
+      "learning_rate": 1.996149484041609e-05,
+      "loss": 0.6959,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1078173588924556,
+      "learning_rate": 1.9961027001196568e-05,
+      "loss": 0.7591,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9555521382159569,
+      "learning_rate": 1.996055634252931e-05,
+      "loss": 0.6523,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.139134857388536,
+      "learning_rate": 1.9960082864547528e-05,
+      "loss": 0.6131,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9735451045598689,
+      "learning_rate": 1.995960656738525e-05,
+      "loss": 0.6152,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8305026826093493,
+      "learning_rate": 1.9959127451177287e-05,
+      "loss": 0.6376,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.977966271836292,
+      "learning_rate": 1.9958645516059247e-05,
+      "loss": 0.6986,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.080168742289415,
+      "learning_rate": 1.995816076216755e-05,
+      "loss": 0.7578,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7465662173066798,
+      "learning_rate": 1.9957673189639398e-05,
+      "loss": 0.6328,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.284784919176215,
+      "learning_rate": 1.9957182798612797e-05,
+      "loss": 0.6837,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9758202684568542,
+      "learning_rate": 1.9956689589226555e-05,
+      "loss": 0.5828,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7254492472804877,
+      "learning_rate": 1.9956193561620267e-05,
+      "loss": 0.6951,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6796922348502967,
+      "learning_rate": 1.9955694715934344e-05,
+      "loss": 0.6293,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9682555458760838,
+      "learning_rate": 1.9955193052309972e-05,
+      "loss": 0.6009,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7580639527168036,
+      "learning_rate": 1.9954688570889152e-05,
+      "loss": 0.6176,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0673079706358366,
+      "learning_rate": 1.9954181271814673e-05,
+      "loss": 0.67,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.390260950420611,
+      "learning_rate": 1.995367115523013e-05,
+      "loss": 0.7041,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.099083294964416,
+      "learning_rate": 1.9953158221279906e-05,
+      "loss": 0.7193,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2132458593216224,
+      "learning_rate": 1.9952642470109185e-05,
+      "loss": 0.612,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9283353158213443,
+      "learning_rate": 1.9952123901863953e-05,
+      "loss": 0.6213,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9306724973848788,
+      "learning_rate": 1.9951602516690988e-05,
+      "loss": 0.7029,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0923495597544046,
+      "learning_rate": 1.995107831473787e-05,
+      "loss": 0.631,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4079158181176954,
+      "learning_rate": 1.9950551296152965e-05,
+      "loss": 0.6424,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.091440347201453,
+      "learning_rate": 1.9950021461085452e-05,
+      "loss": 0.5988,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.120586479674253,
+      "learning_rate": 1.99494888096853e-05,
+      "loss": 0.6827,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1470523255001606,
+      "learning_rate": 1.9948953342103268e-05,
+      "loss": 0.6423,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3075684028008787,
+      "learning_rate": 1.9948415058490926e-05,
+      "loss": 0.6787,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.55241210678369,
+      "learning_rate": 1.994787395900063e-05,
+      "loss": 0.6542,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.879136301233314,
+      "learning_rate": 1.994733004378554e-05,
+      "loss": 0.7053,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.409215834458604,
+      "learning_rate": 1.9946783312999606e-05,
+      "loss": 0.6659,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0064477602182147,
+      "learning_rate": 1.994623376679758e-05,
+      "loss": 0.6088,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3780993208256407,
+      "learning_rate": 1.9945681405335016e-05,
+      "loss": 0.7499,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8494119675565783,
+      "learning_rate": 1.9945126228768252e-05,
+      "loss": 0.6128,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0196297478825644,
+      "learning_rate": 1.994456823725443e-05,
+      "loss": 0.6672,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.897008360443399,
+      "learning_rate": 1.994400743095149e-05,
+      "loss": 0.6967,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0758027108020585,
+      "learning_rate": 1.9943443810018174e-05,
+      "loss": 0.7052,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2644996552861905,
+      "learning_rate": 1.9942877374614e-05,
+      "loss": 0.665,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.143234586192905,
+      "learning_rate": 1.9942308124899307e-05,
+      "loss": 0.6515,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.955943049692675,
+      "learning_rate": 1.9941736061035213e-05,
+      "loss": 0.6249,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.668581878622091,
+      "learning_rate": 1.9941161183183643e-05,
+      "loss": 0.6474,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2617183906754526,
+      "learning_rate": 1.9940583491507314e-05,
+      "loss": 0.5208,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8435252248602325,
+      "learning_rate": 1.994000298616974e-05,
+      "loss": 0.6415,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.004989452919752,
+      "learning_rate": 1.9939419667335236e-05,
+      "loss": 0.5346,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.470218058521019,
+      "learning_rate": 1.99388335351689e-05,
+      "loss": 0.6173,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1179713732034133,
+      "learning_rate": 1.9938244589836646e-05,
+      "loss": 0.578,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.182357850039801,
+      "learning_rate": 1.993765283150517e-05,
+      "loss": 0.5192,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.564175793960651,
+      "learning_rate": 1.9937058260341967e-05,
+      "loss": 0.6329,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9347374728904405,
+      "learning_rate": 1.9936460876515323e-05,
+      "loss": 0.6581,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4533410925900934,
+      "learning_rate": 1.993586068019434e-05,
+      "loss": 0.5819,
+      "step": 396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.056330038552092,
+      "learning_rate": 1.993525767154889e-05,
+      "loss": 0.5539,
+      "step": 397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7166549662862143,
+      "learning_rate": 1.9934651850749663e-05,
+      "loss": 0.7157,
+      "step": 398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2081556322190217,
+      "learning_rate": 1.9934043217968124e-05,
+      "loss": 0.7119,
+      "step": 399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7853340024959392,
+      "learning_rate": 1.9933431773376554e-05,
+      "loss": 0.6124,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.339680896295054,
+      "learning_rate": 1.993281751714802e-05,
+      "loss": 0.6193,
+      "step": 401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6322822470805969,
+      "learning_rate": 1.9932200449456385e-05,
+      "loss": 0.572,
+      "step": 402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.673356693662832,
+      "learning_rate": 1.9931580570476306e-05,
+      "loss": 0.5571,
+      "step": 403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5808193307703693,
+      "learning_rate": 1.993095788038324e-05,
+      "loss": 0.5901,
+      "step": 404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7040058170585108,
+      "learning_rate": 1.993033237935344e-05,
+      "loss": 0.5975,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.527040449774983,
+      "learning_rate": 1.9929704067563957e-05,
+      "loss": 0.6418,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.099934016611324,
+      "learning_rate": 1.992907294519262e-05,
+      "loss": 0.625,
+      "step": 407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.89598961330044,
+      "learning_rate": 1.9928439012418076e-05,
+      "loss": 0.733,
+      "step": 408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9621027652313192,
+      "learning_rate": 1.992780226941976e-05,
+      "loss": 0.6371,
+      "step": 409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.481631527678462,
+      "learning_rate": 1.9927162716377896e-05,
+      "loss": 0.6641,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8343981583497864,
+      "learning_rate": 1.992652035347351e-05,
+      "loss": 0.6608,
+      "step": 411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3717523059107313,
+      "learning_rate": 1.9925875180888426e-05,
+      "loss": 0.7596,
+      "step": 412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.077942449595362,
+      "learning_rate": 1.9925227198805247e-05,
+      "loss": 0.5288,
+      "step": 413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5039263784142554,
+      "learning_rate": 1.9924576407407398e-05,
+      "loss": 0.6883,
+      "step": 414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1468651604563997,
+      "learning_rate": 1.992392280687907e-05,
+      "loss": 0.6976,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.162299299122582,
+      "learning_rate": 1.9923266397405273e-05,
+      "loss": 0.6154,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4437305465754595,
+      "learning_rate": 1.9922607179171796e-05,
+      "loss": 0.7657,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.895826879632121,
+      "learning_rate": 1.9921945152365235e-05,
+      "loss": 0.6429,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8471527035913358,
+      "learning_rate": 1.992128031717297e-05,
+      "loss": 0.6168,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.012126025957949,
+      "learning_rate": 1.9920612673783186e-05,
+      "loss": 0.7497,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9669434182940015,
+      "learning_rate": 1.9919942222384855e-05,
+      "loss": 0.5515,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.079693415215127,
+      "learning_rate": 1.9919268963167747e-05,
+      "loss": 0.5367,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7384838148933404,
+      "learning_rate": 1.9918592896322432e-05,
+      "loss": 0.604,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.280840052907835,
+      "learning_rate": 1.9917914022040258e-05,
+      "loss": 0.5942,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.644689349945528,
+      "learning_rate": 1.9917232340513388e-05,
+      "loss": 0.579,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.813799480875315,
+      "learning_rate": 1.9916547851934768e-05,
+      "loss": 0.5626,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2881197782499663,
+      "learning_rate": 1.991586055649814e-05,
+      "loss": 0.6155,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.053536215203151,
+      "learning_rate": 1.9915170454398045e-05,
+      "loss": 0.6073,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0363897014288397,
+      "learning_rate": 1.991447754582981e-05,
+      "loss": 0.7098,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1098782547635855,
+      "learning_rate": 1.9913781830989568e-05,
+      "loss": 0.5724,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0346975338186395,
+      "learning_rate": 1.991308331007423e-05,
+      "loss": 0.648,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.229552569951582,
+      "learning_rate": 1.9912381983281518e-05,
+      "loss": 0.6509,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1380949386154224,
+      "learning_rate": 1.9911677850809943e-05,
+      "loss": 0.5927,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6970900436798129,
+      "learning_rate": 1.9910970912858802e-05,
+      "loss": 0.5335,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6148059141833992,
+      "learning_rate": 1.9910261169628195e-05,
+      "loss": 0.5965,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.149233091870205,
+      "learning_rate": 1.9909548621319014e-05,
+      "loss": 0.5378,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9114551835690154,
+      "learning_rate": 1.9908833268132943e-05,
+      "loss": 0.6758,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7565957192195754,
+      "learning_rate": 1.9908115110272463e-05,
+      "loss": 0.6263,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8206876481965093,
+      "learning_rate": 1.9907394147940845e-05,
+      "loss": 0.7416,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7937587379385174,
+      "learning_rate": 1.9906670381342156e-05,
+      "loss": 0.6301,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.389432874634321,
+      "learning_rate": 1.9905943810681257e-05,
+      "loss": 0.7042,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2055480838440036,
+      "learning_rate": 1.9905214436163806e-05,
+      "loss": 0.6453,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0926227336521603,
+      "learning_rate": 1.9904482257996244e-05,
+      "loss": 0.5924,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1344211392996892,
+      "learning_rate": 1.9903747276385816e-05,
+      "loss": 0.5894,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.906680767734538,
+      "learning_rate": 1.9903009491540558e-05,
+      "loss": 0.6165,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.546876080333997,
+      "learning_rate": 1.99022689036693e-05,
+      "loss": 0.5177,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3733353729037305,
+      "learning_rate": 1.9901525512981656e-05,
+      "loss": 0.6842,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0157959967158554,
+      "learning_rate": 1.990077931968805e-05,
+      "loss": 0.6019,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.243288604865424,
+      "learning_rate": 1.990003032399969e-05,
+      "loss": 0.7061,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9409756008584556,
+      "learning_rate": 1.989927852612857e-05,
+      "loss": 0.5336,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.13473736470724,
+      "learning_rate": 1.9898523926287488e-05,
+      "loss": 0.6166,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.254516241262999,
+      "learning_rate": 1.9897766524690038e-05,
+      "loss": 0.6536,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0076921990242544,
+      "learning_rate": 1.9897006321550592e-05,
+      "loss": 0.651,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.320624289310723,
+      "learning_rate": 1.9896243317084333e-05,
+      "loss": 0.5711,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2512764688308096,
+      "learning_rate": 1.989547751150722e-05,
+      "loss": 0.5615,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3556068183759478,
+      "learning_rate": 1.9894708905036015e-05,
+      "loss": 0.6268,
+      "step": 456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.442809031247582,
+      "learning_rate": 1.989393749788827e-05,
+      "loss": 0.6824,
+      "step": 457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4328276882184645,
+      "learning_rate": 1.9893163290282335e-05,
+      "loss": 0.6289,
+      "step": 458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.133451140625721,
+      "learning_rate": 1.9892386282437344e-05,
+      "loss": 0.6489,
+      "step": 459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.184973578661761,
+      "learning_rate": 1.9891606474573225e-05,
+      "loss": 0.7446,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0494942578476962,
+      "learning_rate": 1.9890823866910702e-05,
+      "loss": 0.7027,
+      "step": 461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9771302109234297,
+      "learning_rate": 1.9890038459671297e-05,
+      "loss": 0.6883,
+      "step": 462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.893179741102865,
+      "learning_rate": 1.9889250253077306e-05,
+      "loss": 0.6854,
+      "step": 463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.063305340970488,
+      "learning_rate": 1.9888459247351844e-05,
+      "loss": 0.6879,
+      "step": 464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7924857295342795,
+      "learning_rate": 1.988766544271879e-05,
+      "loss": 0.5722,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1255845851863935,
+      "learning_rate": 1.9886868839402837e-05,
+      "loss": 0.5782,
+      "step": 466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7662397087877515,
+      "learning_rate": 1.9886069437629456e-05,
+      "loss": 0.5952,
+      "step": 467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.551864684621323,
+      "learning_rate": 1.9885267237624923e-05,
+      "loss": 0.7176,
+      "step": 468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.758088995750622,
+      "learning_rate": 1.9884462239616292e-05,
+      "loss": 0.6269,
+      "step": 469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9688385871431207,
+      "learning_rate": 1.988365444383142e-05,
+      "loss": 0.645,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6015353660958997,
+      "learning_rate": 1.988284385049895e-05,
+      "loss": 0.6259,
+      "step": 471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3441345972241963,
+      "learning_rate": 1.9882030459848325e-05,
+      "loss": 0.6608,
+      "step": 472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.186504423120599,
+      "learning_rate": 1.988121427210976e-05,
+      "loss": 0.578,
+      "step": 473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1953333764349154,
+      "learning_rate": 1.9880395287514292e-05,
+      "loss": 0.6178,
+      "step": 474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.134962498200742,
+      "learning_rate": 1.987957350629372e-05,
+      "loss": 0.5502,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1470429203928627,
+      "learning_rate": 1.987874892868065e-05,
+      "loss": 0.6384,
+      "step": 476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2885327524060455,
+      "learning_rate": 1.987792155490848e-05,
+      "loss": 0.6119,
+      "step": 477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1611510414098065,
+      "learning_rate": 1.98770913852114e-05,
+      "loss": 0.5592,
+      "step": 478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.222314946254348,
+      "learning_rate": 1.9876258419824375e-05,
+      "loss": 0.6551,
+      "step": 479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1200313775936754,
+      "learning_rate": 1.987542265898319e-05,
+      "loss": 0.6327,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2314307156406734,
+      "learning_rate": 1.9874584102924394e-05,
+      "loss": 0.5931,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.020378809994487,
+      "learning_rate": 1.987374275188534e-05,
+      "loss": 0.6985,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.287148762323182,
+      "learning_rate": 1.9872898606104175e-05,
+      "loss": 0.6148,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0268699008129945,
+      "learning_rate": 1.9872051665819828e-05,
+      "loss": 0.681,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9158555186526622,
+      "learning_rate": 1.9871201931272027e-05,
+      "loss": 0.6653,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5609866512211448,
+      "learning_rate": 1.987034940270129e-05,
+      "loss": 0.6772,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9424978924559555,
+      "learning_rate": 1.9869494080348916e-05,
+      "loss": 0.7011,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.148548907552869,
+      "learning_rate": 1.9868635964457007e-05,
+      "loss": 0.6651,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3982347762921714,
+      "learning_rate": 1.986777505526845e-05,
+      "loss": 0.6975,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8322575443241664,
+      "learning_rate": 1.986691135302692e-05,
+      "loss": 0.6456,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9215399572651608,
+      "learning_rate": 1.9866044857976897e-05,
+      "loss": 0.627,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8848914349709949,
+      "learning_rate": 1.986517557036363e-05,
+      "loss": 0.6404,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3207444027444026,
+      "learning_rate": 1.986430349043317e-05,
+      "loss": 0.6083,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.886880803977155,
+      "learning_rate": 1.9863428618432365e-05,
+      "loss": 0.6114,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.975817998168313,
+      "learning_rate": 1.9862550954608837e-05,
+      "loss": 0.6321,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9320411660665786,
+      "learning_rate": 1.986167049921101e-05,
+      "loss": 0.5972,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.214206855721271,
+      "learning_rate": 1.9860787252488096e-05,
+      "loss": 0.6252,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.192683780486275,
+      "learning_rate": 1.9859901214690094e-05,
+      "loss": 0.6984,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.005999317068309,
+      "learning_rate": 1.98590123860678e-05,
+      "loss": 0.7346,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8081035153371934,
+      "learning_rate": 1.9858120766872785e-05,
+      "loss": 0.5147,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7184350887663717,
+      "learning_rate": 1.9857226357357432e-05,
+      "loss": 0.5725,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9834629180180308,
+      "learning_rate": 1.9856329157774897e-05,
+      "loss": 0.7221,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9383888546021777,
+      "learning_rate": 1.9855429168379127e-05,
+      "loss": 0.7216,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7493648379968807,
+      "learning_rate": 1.9854526389424867e-05,
+      "loss": 0.6213,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.873334772626765,
+      "learning_rate": 1.985362082116764e-05,
+      "loss": 0.4932,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.091354598838092,
+      "learning_rate": 1.9852712463863777e-05,
+      "loss": 0.5471,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.984534824380136,
+      "learning_rate": 1.9851801317770375e-05,
+      "loss": 0.5883,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0217359394315,
+      "learning_rate": 1.985088738314534e-05,
+      "loss": 0.6241,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.28725487499651,
+      "learning_rate": 1.9849970660247352e-05,
+      "loss": 0.6603,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.071424516203741,
+      "learning_rate": 1.984905114933589e-05,
+      "loss": 0.5976,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2635586701264034,
+      "learning_rate": 1.9848128850671222e-05,
+      "loss": 0.7012,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.788343204047133,
+      "learning_rate": 1.98472037645144e-05,
+      "loss": 0.5841,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7751332679316114,
+      "learning_rate": 1.9846275891127275e-05,
+      "loss": 0.5706,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.506020723937374,
+      "learning_rate": 1.9845345230772467e-05,
+      "loss": 0.6809,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0420640169457824,
+      "learning_rate": 1.984441178371341e-05,
+      "loss": 0.6501,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9717873827405425,
+      "learning_rate": 1.9843475550214306e-05,
+      "loss": 0.6234,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2617080873217996,
+      "learning_rate": 1.9842536530540154e-05,
+      "loss": 0.6278,
+      "step": 517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8963845716953496,
+      "learning_rate": 1.9841594724956746e-05,
+      "loss": 0.5571,
+      "step": 518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6118290232079515,
+      "learning_rate": 1.9840650133730657e-05,
+      "loss": 0.6197,
+      "step": 519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.700031470648018,
+      "learning_rate": 1.9839702757129248e-05,
+      "loss": 0.5983,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8702603394150221,
+      "learning_rate": 1.9838752595420674e-05,
+      "loss": 0.6177,
+      "step": 521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7214583947073763,
+      "learning_rate": 1.983779964887388e-05,
+      "loss": 0.6776,
+      "step": 522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0754758649404383,
+      "learning_rate": 1.9836843917758593e-05,
+      "loss": 0.5405,
+      "step": 523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0747927215169613,
+      "learning_rate": 1.9835885402345325e-05,
+      "loss": 0.6864,
+      "step": 524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6238706898229323,
+      "learning_rate": 1.983492410290539e-05,
+      "loss": 0.6695,
+      "step": 525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7323919394458793,
+      "learning_rate": 1.9833960019710878e-05,
+      "loss": 0.5692,
+      "step": 526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5734462475620388,
+      "learning_rate": 1.983299315303467e-05,
+      "loss": 0.752,
+      "step": 527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1360470743259685,
+      "learning_rate": 1.983202350315044e-05,
+      "loss": 0.6923,
+      "step": 528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.745035923122398,
+      "learning_rate": 1.9831051070332642e-05,
+      "loss": 0.6211,
+      "step": 529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0638919774481297,
+      "learning_rate": 1.983007585485652e-05,
+      "loss": 0.7421,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.06554807632873,
+      "learning_rate": 1.9829097856998105e-05,
+      "loss": 0.5444,
+      "step": 531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.130272878967899,
+      "learning_rate": 1.9828117077034225e-05,
+      "loss": 0.5383,
+      "step": 532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6477787156164538,
+      "learning_rate": 1.982713351524248e-05,
+      "loss": 0.6296,
+      "step": 533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.905956414952208,
+      "learning_rate": 1.982614717190127e-05,
+      "loss": 0.6357,
+      "step": 534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.327752565820475,
+      "learning_rate": 1.982515804728977e-05,
+      "loss": 0.6254,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8630302688053053,
+      "learning_rate": 1.982416614168796e-05,
+      "loss": 0.6315,
+      "step": 536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9880441023888256,
+      "learning_rate": 1.9823171455376592e-05,
+      "loss": 0.5993,
+      "step": 537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.031442009613123,
+      "learning_rate": 1.982217398863721e-05,
+      "loss": 0.6082,
+      "step": 538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8286622429425803,
+      "learning_rate": 1.982117374175214e-05,
+      "loss": 0.5587,
+      "step": 539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1779302291114098,
+      "learning_rate": 1.982017071500451e-05,
+      "loss": 0.6722,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.767831867415724,
+      "learning_rate": 1.9819164908678216e-05,
+      "loss": 0.5636,
+      "step": 541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6449521900136204,
+      "learning_rate": 1.9818156323057952e-05,
+      "loss": 0.681,
+      "step": 542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0485274717599307,
+      "learning_rate": 1.98171449584292e-05,
+      "loss": 0.6771,
+      "step": 543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9074717302139765,
+      "learning_rate": 1.9816130815078216e-05,
+      "loss": 0.6681,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.925118720710693,
+      "learning_rate": 1.9815113893292058e-05,
+      "loss": 0.6233,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0194588913773104,
+      "learning_rate": 1.981409419335856e-05,
+      "loss": 0.5585,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7494746493635942,
+      "learning_rate": 1.981307171556635e-05,
+      "loss": 0.6767,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.143467224863585,
+      "learning_rate": 1.9812046460204837e-05,
+      "loss": 0.6051,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8160103524676259,
+      "learning_rate": 1.9811018427564213e-05,
+      "loss": 0.5514,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.836243389500368,
+      "learning_rate": 1.9809987617935468e-05,
+      "loss": 0.7098,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.078032713316881,
+      "learning_rate": 1.9808954031610362e-05,
+      "loss": 0.6446,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.236348689436439,
+      "learning_rate": 1.9807917668881455e-05,
+      "loss": 0.6347,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4669140725283567,
+      "learning_rate": 1.9806878530042083e-05,
+      "loss": 0.6174,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.773772753233092,
+      "learning_rate": 1.9805836615386378e-05,
+      "loss": 0.6413,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9669104811336906,
+      "learning_rate": 1.980479192520925e-05,
+      "loss": 0.6387,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4792048619028946,
+      "learning_rate": 1.9803744459806393e-05,
+      "loss": 0.6339,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9392623696774893,
+      "learning_rate": 1.980269421947429e-05,
+      "loss": 0.6421,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.067503542292842,
+      "learning_rate": 1.9801641204510216e-05,
+      "loss": 0.6511,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7221558068381637,
+      "learning_rate": 1.9800585415212214e-05,
+      "loss": 0.6842,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.204573593448233,
+      "learning_rate": 1.979952685187913e-05,
+      "loss": 0.6147,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3299342817207864,
+      "learning_rate": 1.979846551481059e-05,
+      "loss": 0.6062,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.490731449899381,
+      "learning_rate": 1.9797401404307e-05,
+      "loss": 0.6559,
+      "step": 562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6828010373644284,
+      "learning_rate": 1.9796334520669555e-05,
+      "loss": 0.6434,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2659442314243097,
+      "learning_rate": 1.9795264864200233e-05,
+      "loss": 0.5843,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2135820313974914,
+      "learning_rate": 1.9794192435201797e-05,
+      "loss": 0.6362,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3996758195780512,
+      "learning_rate": 1.97931172339778e-05,
+      "loss": 0.7801,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1290762909253798,
+      "learning_rate": 1.979203926083257e-05,
+      "loss": 0.684,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.603933953083028,
+      "learning_rate": 1.9790958516071228e-05,
+      "loss": 0.695,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2887990690312767,
+      "learning_rate": 1.9789874999999678e-05,
+      "loss": 0.6263,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.100229760369699,
+      "learning_rate": 1.9788788712924606e-05,
+      "loss": 0.6592,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.032859247701893,
+      "learning_rate": 1.978769965515348e-05,
+      "loss": 0.5438,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7847771549911897,
+      "learning_rate": 1.9786607826994557e-05,
+      "loss": 0.5453,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2570009908639634,
+      "learning_rate": 1.978551322875688e-05,
+      "loss": 0.595,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5783991648182165,
+      "learning_rate": 1.978441586075027e-05,
+      "loss": 0.6722,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.121752735804931,
+      "learning_rate": 1.978331572328534e-05,
+      "loss": 0.5781,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.767091219126598,
+      "learning_rate": 1.9782212816673468e-05,
+      "loss": 0.5965,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.28811625758068,
+      "learning_rate": 1.9781107141226845e-05,
+      "loss": 0.6695,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.545520022578118,
+      "learning_rate": 1.977999869725842e-05,
+      "loss": 0.6045,
+      "step": 578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.073703473105074,
+      "learning_rate": 1.977888748508194e-05,
+      "loss": 0.5977,
+      "step": 579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.327100044030786,
+      "learning_rate": 1.977777350501193e-05,
+      "loss": 0.7425,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.000777728411005,
+      "learning_rate": 1.97766567573637e-05,
+      "loss": 0.6006,
+      "step": 581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9697898032986811,
+      "learning_rate": 1.9775537242453347e-05,
+      "loss": 0.6806,
+      "step": 582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0172780270691058,
+      "learning_rate": 1.977441496059774e-05,
+      "loss": 0.6075,
+      "step": 583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5754986102027035,
+      "learning_rate": 1.9773289912114543e-05,
+      "loss": 0.5212,
+      "step": 584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3516025235882947,
+      "learning_rate": 1.9772162097322195e-05,
+      "loss": 0.6123,
+      "step": 585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.916912178881169,
+      "learning_rate": 1.9771031516539928e-05,
+      "loss": 0.5473,
+      "step": 586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9116611793931944,
+      "learning_rate": 1.9769898170087743e-05,
+      "loss": 0.6521,
+      "step": 587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2040573509148706,
+      "learning_rate": 1.9768762058286433e-05,
+      "loss": 0.7345,
+      "step": 588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.44234632537945,
+      "learning_rate": 1.976762318145758e-05,
+      "loss": 0.607,
+      "step": 589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3310018149649996,
+      "learning_rate": 1.9766481539923533e-05,
+      "loss": 0.5854,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7472590924827147,
+      "learning_rate": 1.9765337134007432e-05,
+      "loss": 0.5687,
+      "step": 591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.762456894945371,
+      "learning_rate": 1.9764189964033198e-05,
+      "loss": 0.5702,
+      "step": 592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0314602947757017,
+      "learning_rate": 1.976304003032554e-05,
+      "loss": 0.606,
+      "step": 593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.231768818645365,
+      "learning_rate": 1.976188733320994e-05,
+      "loss": 0.5779,
+      "step": 594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.268601675655289,
+      "learning_rate": 1.9760731873012668e-05,
+      "loss": 0.5558,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0933619702427517,
+      "learning_rate": 1.9759573650060774e-05,
+      "loss": 0.7383,
+      "step": 596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9134963777948708,
+      "learning_rate": 1.9758412664682088e-05,
+      "loss": 0.5136,
+      "step": 597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.470559243751316,
+      "learning_rate": 1.9757248917205228e-05,
+      "loss": 0.5894,
+      "step": 598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9806114796522238,
+      "learning_rate": 1.975608240795959e-05,
+      "loss": 0.5985,
+      "step": 599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.600636486154216,
+      "learning_rate": 1.9754913137275355e-05,
+      "loss": 0.5222,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8550737532406276,
+      "learning_rate": 1.9753741105483475e-05,
+      "loss": 0.6106,
+      "step": 601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7055688109135068,
+      "learning_rate": 1.9752566312915697e-05,
+      "loss": 0.6461,
+      "step": 602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2010359750533453,
+      "learning_rate": 1.975138875990454e-05,
+      "loss": 0.6187,
+      "step": 603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3862477597639633,
+      "learning_rate": 1.975020844678331e-05,
+      "loss": 0.6538,
+      "step": 604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.139772294478368,
+      "learning_rate": 1.974902537388609e-05,
+      "loss": 0.6073,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.907729501815567,
+      "learning_rate": 1.9747839541547754e-05,
+      "loss": 0.7607,
+      "step": 606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.28350475307537,
+      "learning_rate": 1.974665095010394e-05,
+      "loss": 0.6198,
+      "step": 607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1257109840995456,
+      "learning_rate": 1.974545959989108e-05,
+      "loss": 0.6321,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.749808349619169,
+      "learning_rate": 1.974426549124638e-05,
+      "loss": 0.6578,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.128763468032316,
+      "learning_rate": 1.9743068624507837e-05,
+      "loss": 0.6916,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9743242049721,
+      "learning_rate": 1.9741869000014217e-05,
+      "loss": 0.6659,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6610823633090928,
+      "learning_rate": 1.9740666618105067e-05,
+      "loss": 0.6319,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8077050140987267,
+      "learning_rate": 1.9739461479120727e-05,
+      "loss": 0.6291,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 40.53266062387016,
+      "learning_rate": 1.9738253583402306e-05,
+      "loss": 0.6477,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2453207005653835,
+      "learning_rate": 1.973704293129169e-05,
+      "loss": 0.6068,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.611760734948503,
+      "learning_rate": 1.9735829523131564e-05,
+      "loss": 0.6384,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0322959973347854,
+      "learning_rate": 1.9734613359265373e-05,
+      "loss": 0.6346,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0430468269494058,
+      "learning_rate": 1.973339444003735e-05,
+      "loss": 0.6456,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.297901647034911,
+      "learning_rate": 1.9732172765792507e-05,
+      "loss": 0.6035,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.252787896043794,
+      "learning_rate": 1.9730948336876637e-05,
+      "loss": 0.6493,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8835913046909107,
+      "learning_rate": 1.9729721153636312e-05,
+      "loss": 0.6734,
+      "step": 621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.073658061701117,
+      "learning_rate": 1.9728491216418884e-05,
+      "loss": 0.6411,
+      "step": 622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.803789478782545,
+      "learning_rate": 1.9727258525572487e-05,
+      "loss": 0.5975,
+      "step": 623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9378075901592977,
+      "learning_rate": 1.9726023081446026e-05,
+      "loss": 0.6203,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7969844270129638,
+      "learning_rate": 1.9724784884389195e-05,
+      "loss": 0.6851,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8425008930692306,
+      "learning_rate": 1.972354393475246e-05,
+      "loss": 0.6618,
+      "step": 626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6955420051092227,
+      "learning_rate": 1.9722300232887073e-05,
+      "loss": 0.6764,
+      "step": 627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.89669386695366,
+      "learning_rate": 1.9721053779145057e-05,
+      "loss": 0.6989,
+      "step": 628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.502403796085501,
+      "learning_rate": 1.9719804573879223e-05,
+      "loss": 0.6069,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.117604688929045,
+      "learning_rate": 1.971855261744315e-05,
+      "loss": 0.6738,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8246383078681538,
+      "learning_rate": 1.9717297910191204e-05,
+      "loss": 0.5412,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5197325014090572,
+      "learning_rate": 1.9716040452478527e-05,
+      "loss": 0.5972,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3928697103746934,
+      "learning_rate": 1.9714780244661044e-05,
+      "loss": 0.6727,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5456374044656085,
+      "learning_rate": 1.971351728709545e-05,
+      "loss": 0.4633,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.200349106229375,
+      "learning_rate": 1.9712251580139225e-05,
+      "loss": 0.7282,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.333669617003356,
+      "learning_rate": 1.971098312415062e-05,
+      "loss": 0.6459,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.813134563036079,
+      "learning_rate": 1.9709711919488673e-05,
+      "loss": 0.6762,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5401041111688825,
+      "learning_rate": 1.9708437966513196e-05,
+      "loss": 0.643,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9410985370726759,
+      "learning_rate": 1.9707161265584775e-05,
+      "loss": 0.6248,
+      "step": 639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2506015211429946,
+      "learning_rate": 1.970588181706478e-05,
+      "loss": 0.723,
+      "step": 640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7754919861717886,
+      "learning_rate": 1.970459962131536e-05,
+      "loss": 0.5629,
+      "step": 641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4143364360153345,
+      "learning_rate": 1.9703314678699426e-05,
+      "loss": 0.5899,
+      "step": 642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0337025879115194,
+      "learning_rate": 1.9702026989580694e-05,
+      "loss": 0.71,
+      "step": 643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0077871189254797,
+      "learning_rate": 1.970073655432363e-05,
+      "loss": 0.6649,
+      "step": 644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.881799472306852,
+      "learning_rate": 1.9699443373293496e-05,
+      "loss": 0.5391,
+      "step": 645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5711998750138056,
+      "learning_rate": 1.9698147446856316e-05,
+      "loss": 0.6067,
+      "step": 646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6915253880314076,
+      "learning_rate": 1.969684877537891e-05,
+      "loss": 0.6382,
+      "step": 647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.742652189591396,
+      "learning_rate": 1.969554735922885e-05,
+      "loss": 0.625,
+      "step": 648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0994294428627396,
+      "learning_rate": 1.9694243198774516e-05,
+      "loss": 0.5357,
+      "step": 649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6306018902272432,
+      "learning_rate": 1.9692936294385038e-05,
+      "loss": 0.6983,
+      "step": 650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8307323544204717,
+      "learning_rate": 1.969162664643033e-05,
+      "loss": 0.6166,
+      "step": 651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8641522341812466,
+      "learning_rate": 1.9690314255281092e-05,
+      "loss": 0.62,
+      "step": 652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5759706841755026,
+      "learning_rate": 1.968899912130879e-05,
+      "loss": 0.6175,
+      "step": 653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6162384881325895,
+      "learning_rate": 1.968768124488567e-05,
+      "loss": 0.5785,
+      "step": 654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7564903336227404,
+      "learning_rate": 1.9686360626384756e-05,
+      "loss": 0.6114,
+      "step": 655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6549849087613613,
+      "learning_rate": 1.9685037266179846e-05,
+      "loss": 0.5876,
+      "step": 656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9692228088685277,
+      "learning_rate": 1.9683711164645508e-05,
+      "loss": 0.6164,
+      "step": 657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.871591461730637,
+      "learning_rate": 1.9682382322157103e-05,
+      "loss": 0.4991,
+      "step": 658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.077244299987256,
+      "learning_rate": 1.968105073909075e-05,
+      "loss": 0.4933,
+      "step": 659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9155774015207567,
+      "learning_rate": 1.9679716415823352e-05,
+      "loss": 0.7273,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.604980055681089,
+      "learning_rate": 1.9678379352732587e-05,
+      "loss": 0.6118,
+      "step": 661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5399813698615934,
+      "learning_rate": 1.967703955019691e-05,
+      "loss": 0.5843,
+      "step": 662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6121720954834873,
+      "learning_rate": 1.9675697008595545e-05,
+      "loss": 0.5785,
+      "step": 663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.797897179869437,
+      "learning_rate": 1.9674351728308498e-05,
+      "loss": 0.594,
+      "step": 664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7824402242204016,
+      "learning_rate": 1.9673003709716548e-05,
+      "loss": 0.6333,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.806593669207902,
+      "learning_rate": 1.9671652953201245e-05,
+      "loss": 0.6345,
+      "step": 666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.791743208147295,
+      "learning_rate": 1.9670299459144923e-05,
+      "loss": 0.6444,
+      "step": 667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.221125628610726,
+      "learning_rate": 1.9668943227930686e-05,
+      "loss": 0.7795,
+      "step": 668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6950834828868273,
+      "learning_rate": 1.9667584259942408e-05,
+      "loss": 0.5726,
+      "step": 669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.544046033366278,
+      "learning_rate": 1.9666222555564744e-05,
+      "loss": 0.5661,
+      "step": 670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.160440154783599,
+      "learning_rate": 1.9664858115183122e-05,
+      "loss": 0.7553,
+      "step": 671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.934462623303267,
+      "learning_rate": 1.9663490939183744e-05,
+      "loss": 0.6078,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7188506059481328,
+      "learning_rate": 1.966212102795358e-05,
+      "loss": 0.5886,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.844713402351757,
+      "learning_rate": 1.9660748381880394e-05,
+      "loss": 0.6788,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5608710404437298,
+      "learning_rate": 1.96593730013527e-05,
+      "loss": 0.5873,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5371014999321082,
+      "learning_rate": 1.96579948867598e-05,
+      "loss": 0.6536,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7247429844434359,
+      "learning_rate": 1.9656614038491765e-05,
+      "loss": 0.5759,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.605022785167086,
+      "learning_rate": 1.965523045693944e-05,
+      "loss": 0.6484,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.31874960027235,
+      "learning_rate": 1.965384414249445e-05,
+      "loss": 0.5849,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7807265787860524,
+      "learning_rate": 1.9652455095549188e-05,
+      "loss": 0.6231,
+      "step": 680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1040870914871326,
+      "learning_rate": 1.9651063316496813e-05,
+      "loss": 0.7359,
+      "step": 681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8409899699765577,
+      "learning_rate": 1.9649668805731274e-05,
+      "loss": 0.6076,
+      "step": 682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.534732589386806,
+      "learning_rate": 1.964827156364728e-05,
+      "loss": 0.6201,
+      "step": 683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8513036684869464,
+      "learning_rate": 1.9646871590640317e-05,
+      "loss": 0.6739,
+      "step": 684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.609029200642991,
+      "learning_rate": 1.9645468887106645e-05,
+      "loss": 0.6656,
+      "step": 685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2624281971976905,
+      "learning_rate": 1.9644063453443296e-05,
+      "loss": 0.6714,
+      "step": 686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8060980917809355,
+      "learning_rate": 1.9642655290048077e-05,
+      "loss": 0.6479,
+      "step": 687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8699635224886388,
+      "learning_rate": 1.964124439731957e-05,
+      "loss": 0.5458,
+      "step": 688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.018037357333053,
+      "learning_rate": 1.9639830775657113e-05,
+      "loss": 0.5983,
+      "step": 689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6810940145326865,
+      "learning_rate": 1.9638414425460834e-05,
+      "loss": 0.6269,
+      "step": 690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0815141095583505,
+      "learning_rate": 1.9636995347131634e-05,
+      "loss": 0.7389,
+      "step": 691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2648320275142217,
+      "learning_rate": 1.9635573541071174e-05,
+      "loss": 0.6063,
+      "step": 692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8636806596473825,
+      "learning_rate": 1.9634149007681894e-05,
+      "loss": 0.7218,
+      "step": 693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8951189954056595,
+      "learning_rate": 1.9632721747367005e-05,
+      "loss": 0.5902,
+      "step": 694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9083114660422067,
+      "learning_rate": 1.9631291760530492e-05,
+      "loss": 0.6251,
+      "step": 695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6844738023559984,
+      "learning_rate": 1.962985904757711e-05,
+      "loss": 0.6537,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6665136627222314,
+      "learning_rate": 1.962842360891238e-05,
+      "loss": 0.5135,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8334362981132104,
+      "learning_rate": 1.962698544494261e-05,
+      "loss": 0.6857,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9528435013737366,
+      "learning_rate": 1.9625544556074857e-05,
+      "loss": 0.6002,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7777573658045336,
+      "learning_rate": 1.962410094271697e-05,
+      "loss": 0.5981,
+      "step": 700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.013245356552458,
+      "learning_rate": 1.962265460527756e-05,
+      "loss": 0.439,
+      "step": 701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0433544003471744,
+      "learning_rate": 1.962120554416601e-05,
+      "loss": 0.6178,
+      "step": 702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7430159680733397,
+      "learning_rate": 1.9619753759792466e-05,
+      "loss": 0.6157,
+      "step": 703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5784918812971733,
+      "learning_rate": 1.9618299252567863e-05,
+      "loss": 0.5869,
+      "step": 704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5078469152200153,
+      "learning_rate": 1.961684202290389e-05,
+      "loss": 0.572,
+      "step": 705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5721776627567945,
+      "learning_rate": 1.9615382071213017e-05,
+      "loss": 0.6326,
+      "step": 706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.83007200275257,
+      "learning_rate": 1.9613919397908473e-05,
+      "loss": 0.6603,
+      "step": 707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6650978293043017,
+      "learning_rate": 1.961245400340427e-05,
+      "loss": 0.6015,
+      "step": 708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2311782651553997,
+      "learning_rate": 1.961098588811519e-05,
+      "loss": 0.6822,
+      "step": 709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.809672533752102,
+      "learning_rate": 1.9609515052456768e-05,
+      "loss": 0.6459,
+      "step": 710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7701386027756905,
+      "learning_rate": 1.960804149684533e-05,
+      "loss": 0.6207,
+      "step": 711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.048237809511126,
+      "learning_rate": 1.960656522169796e-05,
+      "loss": 0.6308,
+      "step": 712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9851920855110308,
+      "learning_rate": 1.9605086227432512e-05,
+      "loss": 0.6232,
+      "step": 713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7780424403852941,
+      "learning_rate": 1.9603604514467616e-05,
+      "loss": 0.6204,
+      "step": 714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6671929476439753,
+      "learning_rate": 1.9602120083222665e-05,
+      "loss": 0.6602,
+      "step": 715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6050480803661777,
+      "learning_rate": 1.960063293411783e-05,
+      "loss": 0.5656,
+      "step": 716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9233733101686274,
+      "learning_rate": 1.9599143067574037e-05,
+      "loss": 0.5747,
+      "step": 717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7947814560086943,
+      "learning_rate": 1.9597650484012997e-05,
+      "loss": 0.6024,
+      "step": 718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7293523702496045,
+      "learning_rate": 1.9596155183857176e-05,
+      "loss": 0.5855,
+      "step": 719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8894637021501985,
+      "learning_rate": 1.9594657167529818e-05,
+      "loss": 0.6264,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.646411639295727,
+      "learning_rate": 1.9593156435454936e-05,
+      "loss": 0.6323,
+      "step": 721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.624136714819338,
+      "learning_rate": 1.9591652988057305e-05,
+      "loss": 0.6529,
+      "step": 722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.446551631195427,
+      "learning_rate": 1.9590146825762476e-05,
+      "loss": 0.6497,
+      "step": 723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1827970835469896,
+      "learning_rate": 1.9588637948996766e-05,
+      "loss": 0.7506,
+      "step": 724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.698113901709904,
+      "learning_rate": 1.9587126358187257e-05,
+      "loss": 0.6229,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0622635477920612,
+      "learning_rate": 1.9585612053761798e-05,
+      "loss": 0.6556,
+      "step": 726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6381942044855715,
+      "learning_rate": 1.9584095036149016e-05,
+      "loss": 0.6508,
+      "step": 727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5721333086650842,
+      "learning_rate": 1.9582575305778297e-05,
+      "loss": 0.5629,
+      "step": 728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7703811650821941,
+      "learning_rate": 1.95810528630798e-05,
+      "loss": 0.643,
+      "step": 729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.918572930015132,
+      "learning_rate": 1.9579527708484444e-05,
+      "loss": 0.6722,
+      "step": 730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6365349579787993,
+      "learning_rate": 1.957799984242392e-05,
+      "loss": 0.669,
+      "step": 731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6414957401086223,
+      "learning_rate": 1.95764692653307e-05,
+      "loss": 0.6218,
+      "step": 732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.135919642983536,
+      "learning_rate": 1.9574935977637994e-05,
+      "loss": 0.6847,
+      "step": 733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5916691050584089,
+      "learning_rate": 1.9573399979779807e-05,
+      "loss": 0.5474,
+      "step": 734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.174638863781073,
+      "learning_rate": 1.9571861272190898e-05,
+      "loss": 0.6504,
+      "step": 735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8266366236570375,
+      "learning_rate": 1.957031985530679e-05,
+      "loss": 0.6118,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7743026874637837,
+      "learning_rate": 1.9568775729563782e-05,
+      "loss": 0.733,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.069246017744117,
+      "learning_rate": 1.9567228895398936e-05,
+      "loss": 0.6091,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.85030898840822,
+      "learning_rate": 1.9565679353250077e-05,
+      "loss": 0.5875,
+      "step": 739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6868160084312573,
+      "learning_rate": 1.95641271035558e-05,
+      "loss": 0.6281,
+      "step": 740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8821726727239105,
+      "learning_rate": 1.9562572146755473e-05,
+      "loss": 0.5993,
+      "step": 741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9543107867308462,
+      "learning_rate": 1.956101448328921e-05,
+      "loss": 0.575,
+      "step": 742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8961696811686588,
+      "learning_rate": 1.955945411359792e-05,
+      "loss": 0.6068,
+      "step": 743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.929678575630021,
+      "learning_rate": 1.955789103812325e-05,
+      "loss": 0.6503,
+      "step": 744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8538228687289433,
+      "learning_rate": 1.955632525730763e-05,
+      "loss": 0.664,
+      "step": 745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9131946628070533,
+      "learning_rate": 1.955475677159425e-05,
+      "loss": 0.5659,
+      "step": 746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5711195279144858,
+      "learning_rate": 1.955318558142707e-05,
+      "loss": 0.5667,
+      "step": 747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8190214999841572,
+      "learning_rate": 1.9551611687250808e-05,
+      "loss": 0.6342,
+      "step": 748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0693442621959535,
+      "learning_rate": 1.9550035089510952e-05,
+      "loss": 0.5915,
+      "step": 749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.61863478892151,
+      "learning_rate": 1.9548455788653754e-05,
+      "loss": 0.6493,
+      "step": 750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.130821643908958,
+      "learning_rate": 1.9546873785126237e-05,
+      "loss": 0.6441,
+      "step": 751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.663046036814242,
+      "learning_rate": 1.954528907937618e-05,
+      "loss": 0.5208,
+      "step": 752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6596277194682536,
+      "learning_rate": 1.9543701671852127e-05,
+      "loss": 0.5353,
+      "step": 753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8749611572997356,
+      "learning_rate": 1.9542111563003393e-05,
+      "loss": 0.5696,
+      "step": 754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.460289346385702,
+      "learning_rate": 1.954051875328006e-05,
+      "loss": 0.6806,
+      "step": 755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8750334696210917,
+      "learning_rate": 1.9538923243132967e-05,
+      "loss": 0.5486,
+      "step": 756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.283312146104184,
+      "learning_rate": 1.9537325033013714e-05,
+      "loss": 0.5677,
+      "step": 757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8871214976324757,
+      "learning_rate": 1.9535724123374674e-05,
+      "loss": 0.5801,
+      "step": 758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7574703810155845,
+      "learning_rate": 1.9534120514668987e-05,
+      "loss": 0.563,
+      "step": 759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7656371634941834,
+      "learning_rate": 1.9532514207350543e-05,
+      "loss": 0.5549,
+      "step": 760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6480776723709407,
+      "learning_rate": 1.953090520187401e-05,
+      "loss": 0.6401,
+      "step": 761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9699093686457665,
+      "learning_rate": 1.952929349869481e-05,
+      "loss": 0.6706,
+      "step": 762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1013578255496745,
+      "learning_rate": 1.952767909826913e-05,
+      "loss": 0.6338,
+      "step": 763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0057519966778097,
+      "learning_rate": 1.9526062001053928e-05,
+      "loss": 0.663,
+      "step": 764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7107240695323769,
+      "learning_rate": 1.9524442207506915e-05,
+      "loss": 0.7149,
+      "step": 765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.281213430137285,
+      "learning_rate": 1.9522819718086578e-05,
+      "loss": 0.7329,
+      "step": 766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8892538093243092,
+      "learning_rate": 1.952119453325215e-05,
+      "loss": 0.5786,
+      "step": 767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.963398363243217,
+      "learning_rate": 1.951956665346364e-05,
+      "loss": 0.5701,
+      "step": 768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.175190468764164,
+      "learning_rate": 1.951793607918182e-05,
+      "loss": 0.7584,
+      "step": 769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0257251787691515,
+      "learning_rate": 1.9516302810868212e-05,
+      "loss": 0.7099,
+      "step": 770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0116489308149443,
+      "learning_rate": 1.9514666848985116e-05,
+      "loss": 0.6464,
+      "step": 771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6783302133166016,
+      "learning_rate": 1.9513028193995588e-05,
+      "loss": 0.6579,
+      "step": 772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7828637461608936,
+      "learning_rate": 1.951138684636344e-05,
+      "loss": 0.6407,
+      "step": 773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8672349567569246,
+      "learning_rate": 1.950974280655326e-05,
+      "loss": 0.528,
+      "step": 774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7314843065627585,
+      "learning_rate": 1.9508096075030378e-05,
+      "loss": 0.6941,
+      "step": 775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3314174938386965,
+      "learning_rate": 1.950644665226091e-05,
+      "loss": 0.7236,
+      "step": 776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6466335431514536,
+      "learning_rate": 1.9504794538711715e-05,
+      "loss": 0.5753,
+      "step": 777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.143286238107289,
+      "learning_rate": 1.9503139734850426e-05,
+      "loss": 0.6126,
+      "step": 778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4392970587289247,
+      "learning_rate": 1.9501482241145422e-05,
+      "loss": 0.606,
+      "step": 779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8333701847696537,
+      "learning_rate": 1.9499822058065863e-05,
+      "loss": 0.5667,
+      "step": 780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6603117485352794,
+      "learning_rate": 1.9498159186081656e-05,
+      "loss": 0.6442,
+      "step": 781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.868236751189393,
+      "learning_rate": 1.949649362566347e-05,
+      "loss": 0.5925,
+      "step": 782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9524354994097128,
+      "learning_rate": 1.9494825377282746e-05,
+      "loss": 0.6949,
+      "step": 783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2737290773528347,
+      "learning_rate": 1.9493154441411673e-05,
+      "loss": 0.5907,
+      "step": 784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3845708135081845,
+      "learning_rate": 1.949148081852321e-05,
+      "loss": 0.6018,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1738918995352807,
+      "learning_rate": 1.9489804509091066e-05,
+      "loss": 0.6137,
+      "step": 786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.867753153035319,
+      "learning_rate": 1.9488125513589722e-05,
+      "loss": 0.7503,
+      "step": 787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9935064962703957,
+      "learning_rate": 1.9486443832494414e-05,
+      "loss": 0.6591,
+      "step": 788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7981612752510605,
+      "learning_rate": 1.9484759466281132e-05,
+      "loss": 0.6417,
+      "step": 789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1934315334439924,
+      "learning_rate": 1.948307241542664e-05,
+      "loss": 0.5312,
+      "step": 790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1160063174038184,
+      "learning_rate": 1.9481382680408455e-05,
+      "loss": 0.5702,
+      "step": 791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7650033694806282,
+      "learning_rate": 1.9479690261704848e-05,
+      "loss": 0.5728,
+      "step": 792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.890128406891613,
+      "learning_rate": 1.9477995159794854e-05,
+      "loss": 0.5456,
+      "step": 793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.020059878365379,
+      "learning_rate": 1.947629737515827e-05,
+      "loss": 0.5551,
+      "step": 794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9993923704629624,
+      "learning_rate": 1.947459690827565e-05,
+      "loss": 0.7318,
+      "step": 795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.724341878195876,
+      "learning_rate": 1.9472893759628307e-05,
+      "loss": 0.6534,
+      "step": 796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.274455958252066,
+      "learning_rate": 1.9471187929698317e-05,
+      "loss": 0.6215,
+      "step": 797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3132373820486802,
+      "learning_rate": 1.9469479418968506e-05,
+      "loss": 0.7022,
+      "step": 798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8426393149693552,
+      "learning_rate": 1.946776822792247e-05,
+      "loss": 0.5579,
+      "step": 799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7173647401664893,
+      "learning_rate": 1.9466054357044558e-05,
+      "loss": 0.6112,
+      "step": 800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8202492106991148,
+      "learning_rate": 1.9464337806819872e-05,
+      "loss": 0.6945,
+      "step": 801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.654506939575743,
+      "learning_rate": 1.946261857773428e-05,
+      "loss": 0.5835,
+      "step": 802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4419700973540355,
+      "learning_rate": 1.9460896670274408e-05,
+      "loss": 0.617,
+      "step": 803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7966285197370981,
+      "learning_rate": 1.9459172084927638e-05,
+      "loss": 0.5804,
+      "step": 804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.410204508273117,
+      "learning_rate": 1.945744482218211e-05,
+      "loss": 0.6707,
+      "step": 805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4759531970274942,
+      "learning_rate": 1.945571488252672e-05,
+      "loss": 0.5366,
+      "step": 806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8190886743904062,
+      "learning_rate": 1.945398226645113e-05,
+      "loss": 0.5765,
+      "step": 807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.018322929879264,
+      "learning_rate": 1.9452246974445743e-05,
+      "loss": 0.5879,
+      "step": 808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7726209371608117,
+      "learning_rate": 1.9450509007001738e-05,
+      "loss": 0.6455,
+      "step": 809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.527932331058197,
+      "learning_rate": 1.9448768364611043e-05,
+      "loss": 0.6396,
+      "step": 810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8837155398931429,
+      "learning_rate": 1.944702504776634e-05,
+      "loss": 0.505,
+      "step": 811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.957492136051977,
+      "learning_rate": 1.944527905696107e-05,
+      "loss": 0.7274,
+      "step": 812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8115814342408663,
+      "learning_rate": 1.9443530392689434e-05,
+      "loss": 0.5771,
+      "step": 813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7489449564095538,
+      "learning_rate": 1.9441779055446387e-05,
+      "loss": 0.6122,
+      "step": 814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6761432996869425,
+      "learning_rate": 1.9440025045727645e-05,
+      "loss": 0.5186,
+      "step": 815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7082774106670753,
+      "learning_rate": 1.9438268364029674e-05,
+      "loss": 0.554,
+      "step": 816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5179060375494233,
+      "learning_rate": 1.9436509010849696e-05,
+      "loss": 0.576,
+      "step": 817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2450790887972563,
+      "learning_rate": 1.94347469866857e-05,
+      "loss": 0.564,
+      "step": 818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.341998987711259,
+      "learning_rate": 1.9432982292036414e-05,
+      "loss": 0.5813,
+      "step": 819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4369212692293283,
+      "learning_rate": 1.9431214927401337e-05,
+      "loss": 0.5868,
+      "step": 820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.464191473731214,
+      "learning_rate": 1.9429444893280717e-05,
+      "loss": 0.6144,
+      "step": 821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6562257789024781,
+      "learning_rate": 1.9427672190175557e-05,
+      "loss": 0.6376,
+      "step": 822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8137843292093998,
+      "learning_rate": 1.9425896818587615e-05,
+      "loss": 0.6396,
+      "step": 823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5823969416832617,
+      "learning_rate": 1.9424118779019415e-05,
+      "loss": 0.5378,
+      "step": 824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6258443123082578,
+      "learning_rate": 1.9422338071974215e-05,
+      "loss": 0.6823,
+      "step": 825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7857008793355629,
+      "learning_rate": 1.9420554697956052e-05,
+      "loss": 0.6359,
+      "step": 826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3451747302990604,
+      "learning_rate": 1.9418768657469695e-05,
+      "loss": 0.5455,
+      "step": 827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9786510381605344,
+      "learning_rate": 1.941697995102069e-05,
+      "loss": 0.6386,
+      "step": 828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1513926002656865,
+      "learning_rate": 1.9415188579115315e-05,
+      "loss": 0.6162,
+      "step": 829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.889927539658958,
+      "learning_rate": 1.941339454226063e-05,
+      "loss": 0.7133,
+      "step": 830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6767719082528572,
+      "learning_rate": 1.9411597840964414e-05,
+      "loss": 0.5057,
+      "step": 831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9505128422012947,
+      "learning_rate": 1.9409798475735234e-05,
+      "loss": 0.5676,
+      "step": 832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9737167160717939,
+      "learning_rate": 1.9407996447082394e-05,
+      "loss": 0.559,
+      "step": 833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7412617723959494,
+      "learning_rate": 1.940619175551595e-05,
+      "loss": 0.6103,
+      "step": 834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.644092860835786,
+      "learning_rate": 1.940438440154672e-05,
+      "loss": 0.6102,
+      "step": 835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0883471948813042,
+      "learning_rate": 1.940257438568627e-05,
+      "loss": 0.7191,
+      "step": 836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7065102895520572,
+      "learning_rate": 1.9400761708446918e-05,
+      "loss": 0.6719,
+      "step": 837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8843157767685141,
+      "learning_rate": 1.939894637034174e-05,
+      "loss": 0.6561,
+      "step": 838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.033857868090308,
+      "learning_rate": 1.9397128371884575e-05,
+      "loss": 0.711,
+      "step": 839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4321004225351524,
+      "learning_rate": 1.9395307713589987e-05,
+      "loss": 0.5119,
+      "step": 840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9739255127988922,
+      "learning_rate": 1.939348439597332e-05,
+      "loss": 0.6417,
+      "step": 841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.230205146115749,
+      "learning_rate": 1.9391658419550653e-05,
+      "loss": 0.6797,
+      "step": 842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.051285184781089,
+      "learning_rate": 1.9389829784838833e-05,
+      "loss": 0.5411,
+      "step": 843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0888053514870313,
+      "learning_rate": 1.9387998492355444e-05,
+      "loss": 0.5463,
+      "step": 844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.371392420550488,
+      "learning_rate": 1.9386164542618836e-05,
+      "loss": 0.6696,
+      "step": 845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7433685324608175,
+      "learning_rate": 1.9384327936148095e-05,
+      "loss": 0.6668,
+      "step": 846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.067802743258445,
+      "learning_rate": 1.938248867346308e-05,
+      "loss": 0.5763,
+      "step": 847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8754844039345144,
+      "learning_rate": 1.938064675508438e-05,
+      "loss": 0.7222,
+      "step": 848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.326387174136724,
+      "learning_rate": 1.9378802181533354e-05,
+      "loss": 0.7704,
+      "step": 849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.055158304887679,
+      "learning_rate": 1.9376954953332104e-05,
+      "loss": 0.6332,
+      "step": 850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8925517701256989,
+      "learning_rate": 1.9375105071003476e-05,
+      "loss": 0.6326,
+      "step": 851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7406906324029456,
+      "learning_rate": 1.9373252535071087e-05,
+      "loss": 0.6404,
+      "step": 852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.62457230824466,
+      "learning_rate": 1.9371397346059286e-05,
+      "loss": 0.5957,
+      "step": 853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.524877249830909,
+      "learning_rate": 1.936953950449318e-05,
+      "loss": 0.6508,
+      "step": 854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.634273213597407,
+      "learning_rate": 1.936767901089863e-05,
+      "loss": 0.5135,
+      "step": 855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1481590708428753,
+      "learning_rate": 1.9365815865802243e-05,
+      "loss": 0.6711,
+      "step": 856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7295810330071002,
+      "learning_rate": 1.936395006973138e-05,
+      "loss": 0.6556,
+      "step": 857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.263507774294202,
+      "learning_rate": 1.936208162321415e-05,
+      "loss": 0.6882,
+      "step": 858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.308361777905462,
+      "learning_rate": 1.9360210526779414e-05,
+      "loss": 0.6774,
+      "step": 859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8695459872051503,
+      "learning_rate": 1.9358336780956777e-05,
+      "loss": 0.5512,
+      "step": 860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1674452924796626,
+      "learning_rate": 1.9356460386276606e-05,
+      "loss": 0.7304,
+      "step": 861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.594294831771164,
+      "learning_rate": 1.9354581343270006e-05,
+      "loss": 0.6333,
+      "step": 862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1237695891665798,
+      "learning_rate": 1.9352699652468835e-05,
+      "loss": 0.6188,
+      "step": 863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4636190766799357,
+      "learning_rate": 1.9350815314405703e-05,
+      "loss": 0.6162,
+      "step": 864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1820884343348452,
+      "learning_rate": 1.934892832961397e-05,
+      "loss": 0.5834,
+      "step": 865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.950861410695652,
+      "learning_rate": 1.9347038698627744e-05,
+      "loss": 0.6776,
+      "step": 866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.760536078777778,
+      "learning_rate": 1.9345146421981878e-05,
+      "loss": 0.5306,
+      "step": 867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.871309726923071,
+      "learning_rate": 1.9343251500211977e-05,
+      "loss": 0.5266,
+      "step": 868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9454068206309962,
+      "learning_rate": 1.9341353933854396e-05,
+      "loss": 0.5782,
+      "step": 869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7037013789981632,
+      "learning_rate": 1.9339453723446234e-05,
+      "loss": 0.6294,
+      "step": 870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7761667372454581,
+      "learning_rate": 1.9337550869525344e-05,
+      "loss": 0.6005,
+      "step": 871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.806680978952498,
+      "learning_rate": 1.933564537263033e-05,
+      "loss": 0.5734,
+      "step": 872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.856475456901691,
+      "learning_rate": 1.933373723330053e-05,
+      "loss": 0.6665,
+      "step": 873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6596203720135592,
+      "learning_rate": 1.9331826452076044e-05,
+      "loss": 0.5046,
+      "step": 874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.903988827890489,
+      "learning_rate": 1.932991302949771e-05,
+      "loss": 0.6865,
+      "step": 875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.033747813653558,
+      "learning_rate": 1.9327996966107122e-05,
+      "loss": 0.6919,
+      "step": 876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.502972494894146,
+      "learning_rate": 1.932607826244662e-05,
+      "loss": 0.5783,
+      "step": 877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7098300139546272,
+      "learning_rate": 1.9324156919059286e-05,
+      "loss": 0.5513,
+      "step": 878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.493243515203421,
+      "learning_rate": 1.932223293648895e-05,
+      "loss": 0.5697,
+      "step": 879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5538568024946955,
+      "learning_rate": 1.9320306315280196e-05,
+      "loss": 0.6524,
+      "step": 880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6216042300263807,
+      "learning_rate": 1.9318377055978342e-05,
+      "loss": 0.6758,
+      "step": 881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.958317816645062,
+      "learning_rate": 1.9316445159129474e-05,
+      "loss": 0.597,
+      "step": 882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8390127355213475,
+      "learning_rate": 1.93145106252804e-05,
+      "loss": 0.6103,
+      "step": 883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.684044338821304,
+      "learning_rate": 1.931257345497869e-05,
+      "loss": 0.5628,
+      "step": 884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.852928676015327,
+      "learning_rate": 1.9310633648772656e-05,
+      "loss": 0.5306,
+      "step": 885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.925316273996792,
+      "learning_rate": 1.9308691207211353e-05,
+      "loss": 0.679,
+      "step": 886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6764196464055603,
+      "learning_rate": 1.9306746130844593e-05,
+      "loss": 0.7077,
+      "step": 887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5849855042523577,
+      "learning_rate": 1.9304798420222918e-05,
+      "loss": 0.6172,
+      "step": 888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8215253276766983,
+      "learning_rate": 1.9302848075897624e-05,
+      "loss": 0.6622,
+      "step": 889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1513833511539695,
+      "learning_rate": 1.9300895098420753e-05,
+      "loss": 0.6196,
+      "step": 890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3917063566093786,
+      "learning_rate": 1.929893948834509e-05,
+      "loss": 0.644,
+      "step": 891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.804803883440053,
+      "learning_rate": 1.9296981246224173e-05,
+      "loss": 0.6317,
+      "step": 892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.574737877991451,
+      "learning_rate": 1.9295020372612276e-05,
+      "loss": 0.5476,
+      "step": 893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.727255966989649,
+      "learning_rate": 1.929305686806441e-05,
+      "loss": 0.6538,
+      "step": 894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8652296646292938,
+      "learning_rate": 1.9291090733136352e-05,
+      "loss": 0.5683,
+      "step": 895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5723720962163956,
+      "learning_rate": 1.928912196838461e-05,
+      "loss": 0.5772,
+      "step": 896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4933644539724202,
+      "learning_rate": 1.9287150574366432e-05,
+      "loss": 0.6271,
+      "step": 897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.76744104843635,
+      "learning_rate": 1.9285176551639826e-05,
+      "loss": 0.6739,
+      "step": 898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9069638819269022,
+      "learning_rate": 1.9283199900763533e-05,
+      "loss": 0.7033,
+      "step": 899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5763530261868464,
+      "learning_rate": 1.9281220622297033e-05,
+      "loss": 0.5796,
+      "step": 900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.532769232352761,
+      "learning_rate": 1.927923871680057e-05,
+      "loss": 0.4702,
+      "step": 901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8926763070736328,
+      "learning_rate": 1.9277254184835105e-05,
+      "loss": 0.7309,
+      "step": 902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8024439861207613,
+      "learning_rate": 1.9275267026962358e-05,
+      "loss": 0.6266,
+      "step": 903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.871072757139364,
+      "learning_rate": 1.9273277243744797e-05,
+      "loss": 0.6438,
+      "step": 904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6722465936477147,
+      "learning_rate": 1.927128483574562e-05,
+      "loss": 0.5734,
+      "step": 905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.302977532562798,
+      "learning_rate": 1.9269289803528775e-05,
+      "loss": 0.6365,
+      "step": 906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7508614736263166,
+      "learning_rate": 1.926729214765895e-05,
+      "loss": 0.5395,
+      "step": 907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5811146764013144,
+      "learning_rate": 1.9265291868701584e-05,
+      "loss": 0.6165,
+      "step": 908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7146459134105925,
+      "learning_rate": 1.9263288967222843e-05,
+      "loss": 0.7272,
+      "step": 909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4775895113844413,
+      "learning_rate": 1.926128344378965e-05,
+      "loss": 0.5312,
+      "step": 910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7304818227986019,
+      "learning_rate": 1.9259275298969663e-05,
+      "loss": 0.5354,
+      "step": 911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3895155172971463,
+      "learning_rate": 1.925726453333128e-05,
+      "loss": 0.6241,
+      "step": 912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.400025506287465,
+      "learning_rate": 1.9255251147443646e-05,
+      "loss": 0.5766,
+      "step": 913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6235968216293228,
+      "learning_rate": 1.9253235141876646e-05,
+      "loss": 0.6478,
+      "step": 914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6869370394953147,
+      "learning_rate": 1.925121651720091e-05,
+      "loss": 0.5317,
+      "step": 915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.466887971109699,
+      "learning_rate": 1.92491952739878e-05,
+      "loss": 0.6985,
+      "step": 916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5445468583051873,
+      "learning_rate": 1.9247171412809423e-05,
+      "loss": 0.5861,
+      "step": 917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.310505180905813,
+      "learning_rate": 1.924514493423864e-05,
+      "loss": 0.5751,
+      "step": 918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0135617676361195,
+      "learning_rate": 1.9243115838849023e-05,
+      "loss": 0.7795,
+      "step": 919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7569521793067802,
+      "learning_rate": 1.924108412721492e-05,
+      "loss": 0.6166,
+      "step": 920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.133576031212046,
+      "learning_rate": 1.9239049799911397e-05,
+      "loss": 0.61,
+      "step": 921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.967531294057922,
+      "learning_rate": 1.923701285751426e-05,
+      "loss": 0.652,
+      "step": 922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0982868061151168,
+      "learning_rate": 1.9234973300600074e-05,
+      "loss": 0.648,
+      "step": 923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6334514428095237,
+      "learning_rate": 1.9232931129746116e-05,
+      "loss": 0.5993,
+      "step": 924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.481049014947498,
+      "learning_rate": 1.9230886345530432e-05,
+      "loss": 0.6364,
+      "step": 925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5959629110706173,
+      "learning_rate": 1.9228838948531786e-05,
+      "loss": 0.6089,
+      "step": 926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5736776525413994,
+      "learning_rate": 1.9226788939329693e-05,
+      "loss": 0.5689,
+      "step": 927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7969359333753454,
+      "learning_rate": 1.92247363185044e-05,
+      "loss": 0.5489,
+      "step": 928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8847621144944555,
+      "learning_rate": 1.92226810866369e-05,
+      "loss": 0.5565,
+      "step": 929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.022126364808558,
+      "learning_rate": 1.922062324430892e-05,
+      "loss": 0.6372,
+      "step": 930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7334611811746106,
+      "learning_rate": 1.921856279210293e-05,
+      "loss": 0.5962,
+      "step": 931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8414568576268548,
+      "learning_rate": 1.9216499730602135e-05,
+      "loss": 0.6497,
+      "step": 932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4735638331530065,
+      "learning_rate": 1.9214434060390484e-05,
+      "loss": 0.6104,
+      "step": 933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1122149191966293,
+      "learning_rate": 1.9212365782052656e-05,
+      "loss": 0.6189,
+      "step": 934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.919111222374491,
+      "learning_rate": 1.9210294896174074e-05,
+      "loss": 0.5558,
+      "step": 935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0613408729189673,
+      "learning_rate": 1.9208221403340895e-05,
+      "loss": 0.5154,
+      "step": 936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.357046408761208,
+      "learning_rate": 1.9206145304140026e-05,
+      "loss": 0.5786,
+      "step": 937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0747828249402556,
+      "learning_rate": 1.9204066599159094e-05,
+      "loss": 0.5629,
+      "step": 938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.762561695239933,
+      "learning_rate": 1.920198528898648e-05,
+      "loss": 0.5253,
+      "step": 939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.59771973393716,
+      "learning_rate": 1.919990137421128e-05,
+      "loss": 0.5693,
+      "step": 940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8460648110502127,
+      "learning_rate": 1.9197814855423357e-05,
+      "loss": 0.6479,
+      "step": 941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9826749935450603,
+      "learning_rate": 1.919572573321329e-05,
+      "loss": 0.7355,
+      "step": 942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3059526262294825,
+      "learning_rate": 1.9193634008172396e-05,
+      "loss": 0.5576,
+      "step": 943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.995697900311234,
+      "learning_rate": 1.9191539680892738e-05,
+      "loss": 0.5841,
+      "step": 944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7493648317886854,
+      "learning_rate": 1.9189442751967117e-05,
+      "loss": 0.5663,
+      "step": 945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.828732857968477,
+      "learning_rate": 1.9187343221989052e-05,
+      "loss": 0.6075,
+      "step": 946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0624268188705215,
+      "learning_rate": 1.918524109155282e-05,
+      "loss": 0.6854,
+      "step": 947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.006592996496279,
+      "learning_rate": 1.9183136361253417e-05,
+      "loss": 0.5862,
+      "step": 948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1769276012056173,
+      "learning_rate": 1.918102903168659e-05,
+      "loss": 0.6471,
+      "step": 949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.591786461448892,
+      "learning_rate": 1.9178919103448807e-05,
+      "loss": 0.5989,
+      "step": 950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.213985841788408,
+      "learning_rate": 1.9176806577137285e-05,
+      "loss": 0.6627,
+      "step": 951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.619588650891793,
+      "learning_rate": 1.9174691453349967e-05,
+      "loss": 0.6143,
+      "step": 952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9935827858499202,
+      "learning_rate": 1.917257373268554e-05,
+      "loss": 0.5933,
+      "step": 953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7761517628294243,
+      "learning_rate": 1.917045341574341e-05,
+      "loss": 0.6321,
+      "step": 954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.150647575650903,
+      "learning_rate": 1.916833050312373e-05,
+      "loss": 0.5852,
+      "step": 955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8709253984566043,
+      "learning_rate": 1.9166204995427398e-05,
+      "loss": 0.6518,
+      "step": 956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.817905961582922,
+      "learning_rate": 1.916407689325602e-05,
+      "loss": 0.6139,
+      "step": 957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5685734734643924,
+      "learning_rate": 1.916194619721196e-05,
+      "loss": 0.5175,
+      "step": 958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8154768146646345,
+      "learning_rate": 1.9159812907898304e-05,
+      "loss": 0.5631,
+      "step": 959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.188801583208716,
+      "learning_rate": 1.915767702591887e-05,
+      "loss": 0.623,
+      "step": 960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6458796158928133,
+      "learning_rate": 1.9155538551878225e-05,
+      "loss": 0.5275,
+      "step": 961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4137374632093245,
+      "learning_rate": 1.9153397486381657e-05,
+      "loss": 0.667,
+      "step": 962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.405283737717209,
+      "learning_rate": 1.915125383003518e-05,
+      "loss": 0.6425,
+      "step": 963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6224219578928296,
+      "learning_rate": 1.9149107583445566e-05,
+      "loss": 0.6541,
+      "step": 964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6958896242732435,
+      "learning_rate": 1.9146958747220292e-05,
+      "loss": 0.5308,
+      "step": 965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9237983536449437,
+      "learning_rate": 1.9144807321967594e-05,
+      "loss": 0.6712,
+      "step": 966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7454164519991082,
+      "learning_rate": 1.914265330829642e-05,
+      "loss": 0.5297,
+      "step": 967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.750652047776378,
+      "learning_rate": 1.914049670681646e-05,
+      "loss": 0.6562,
+      "step": 968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8288309067745445,
+      "learning_rate": 1.913833751813814e-05,
+      "loss": 0.5858,
+      "step": 969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.149113177891929,
+      "learning_rate": 1.9136175742872608e-05,
+      "loss": 0.6797,
+      "step": 970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.625613018492278,
+      "learning_rate": 1.9134011381631755e-05,
+      "loss": 0.5522,
+      "step": 971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.3482847552340465,
+      "learning_rate": 1.9131844435028196e-05,
+      "loss": 0.6362,
+      "step": 972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6647569702362335,
+      "learning_rate": 1.912967490367528e-05,
+      "loss": 0.6251,
+      "step": 973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1980855953452583,
+      "learning_rate": 1.912750278818709e-05,
+      "loss": 0.6321,
+      "step": 974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0541754783918993,
+      "learning_rate": 1.9125328089178442e-05,
+      "loss": 0.5737,
+      "step": 975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8128869012893993,
+      "learning_rate": 1.9123150807264872e-05,
+      "loss": 0.5895,
+      "step": 976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1404405055971918,
+      "learning_rate": 1.912097094306266e-05,
+      "loss": 0.6908,
+      "step": 977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9764602757727396,
+      "learning_rate": 1.9118788497188815e-05,
+      "loss": 0.7093,
+      "step": 978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.810254489375759,
+      "learning_rate": 1.9116603470261065e-05,
+      "loss": 0.5562,
+      "step": 979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3420387458056187,
+      "learning_rate": 1.9114415862897883e-05,
+      "loss": 0.6595,
+      "step": 980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.716507888126097,
+      "learning_rate": 1.911222567571847e-05,
+      "loss": 0.6683,
+      "step": 981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2725477625914587,
+      "learning_rate": 1.911003290934275e-05,
+      "loss": 0.7095,
+      "step": 982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7360631444911283,
+      "learning_rate": 1.9107837564391376e-05,
+      "loss": 0.5783,
+      "step": 983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.069686956575179,
+      "learning_rate": 1.910563964148574e-05,
+      "loss": 0.6701,
+      "step": 984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0124787433034994,
+      "learning_rate": 1.9103439141247966e-05,
+      "loss": 0.6133,
+      "step": 985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9172845300162518,
+      "learning_rate": 1.9101236064300895e-05,
+      "loss": 0.5934,
+      "step": 986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.041311393371292,
+      "learning_rate": 1.90990304112681e-05,
+      "loss": 0.7016,
+      "step": 987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9107782210610409,
+      "learning_rate": 1.9096822182773887e-05,
+      "loss": 0.6863,
+      "step": 988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.101349128110604,
+      "learning_rate": 1.9094611379443298e-05,
+      "loss": 0.617,
+      "step": 989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.585677591801241,
+      "learning_rate": 1.9092398001902092e-05,
+      "loss": 0.5872,
+      "step": 990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.191837696058044,
+      "learning_rate": 1.909018205077676e-05,
+      "loss": 0.6337,
+      "step": 991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.382913216432954,
+      "learning_rate": 1.908796352669452e-05,
+      "loss": 0.6445,
+      "step": 992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.239523252356875,
+      "learning_rate": 1.9085742430283322e-05,
+      "loss": 0.6686,
+      "step": 993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7864260930461942,
+      "learning_rate": 1.9083518762171847e-05,
+      "loss": 0.5991,
+      "step": 994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.791996672464703,
+      "learning_rate": 1.9081292522989493e-05,
+      "loss": 0.5615,
+      "step": 995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9882291176877311,
+      "learning_rate": 1.90790637133664e-05,
+      "loss": 0.6559,
+      "step": 996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2195809827207182,
+      "learning_rate": 1.9076832333933423e-05,
+      "loss": 0.548,
+      "step": 997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.373436152443858,
+      "learning_rate": 1.907459838532215e-05,
+      "loss": 0.547,
+      "step": 998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9205535256488149,
+      "learning_rate": 1.9072361868164892e-05,
+      "loss": 0.6161,
+      "step": 999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.087939148697894,
+      "learning_rate": 1.9070122783094695e-05,
+      "loss": 0.6039,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9025978707194457,
+      "learning_rate": 1.9067881130745325e-05,
+      "loss": 0.526,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9533995111510756,
+      "learning_rate": 1.906563691175128e-05,
+      "loss": 0.6573,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.39124882960071,
+      "learning_rate": 1.9063390126747778e-05,
+      "loss": 0.6522,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1242336347937498,
+      "learning_rate": 1.906114077637077e-05,
+      "loss": 0.718,
+      "step": 1004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4574801825457686,
+      "learning_rate": 1.905888886125693e-05,
+      "loss": 0.739,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4636242324608242,
+      "learning_rate": 1.9056634382043653e-05,
+      "loss": 0.6072,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8375401996369404,
+      "learning_rate": 1.905437733936907e-05,
+      "loss": 0.67,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.521212870775441,
+      "learning_rate": 1.9052117733872025e-05,
+      "loss": 0.5593,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4640378159280134,
+      "learning_rate": 1.9049855566192105e-05,
+      "loss": 0.5896,
+      "step": 1009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9922037443740852,
+      "learning_rate": 1.9047590836969603e-05,
+      "loss": 0.5406,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.74398057466398,
+      "learning_rate": 1.904532354684555e-05,
+      "loss": 0.6205,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.79596958019087,
+      "learning_rate": 1.9043053696461696e-05,
+      "loss": 0.6105,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1570935667531947,
+      "learning_rate": 1.904078128646052e-05,
+      "loss": 0.5494,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4498979112962413,
+      "learning_rate": 1.903850631748522e-05,
+      "loss": 0.583,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7678727578348574,
+      "learning_rate": 1.9036228790179722e-05,
+      "loss": 0.5113,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5355244690520675,
+      "learning_rate": 1.9033948705188673e-05,
+      "loss": 0.6408,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8901723928811676,
+      "learning_rate": 1.9031666063157453e-05,
+      "loss": 0.4856,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.046527025548339,
+      "learning_rate": 1.902938086473215e-05,
+      "loss": 0.6022,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.809416079648886,
+      "learning_rate": 1.9027093110559596e-05,
+      "loss": 0.5753,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.254403891644121,
+      "learning_rate": 1.9024802801287327e-05,
+      "loss": 0.59,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9746273248324238,
+      "learning_rate": 1.9022509937563606e-05,
+      "loss": 0.5898,
+      "step": 1021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6366200245193796,
+      "learning_rate": 1.9020214520037433e-05,
+      "loss": 0.5829,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.457710488384124,
+      "learning_rate": 1.901791654935852e-05,
+      "loss": 0.5984,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9926375080663197,
+      "learning_rate": 1.9015616026177294e-05,
+      "loss": 0.5569,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6746502615061614,
+      "learning_rate": 1.901331295114492e-05,
+      "loss": 0.5012,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.831630747740797,
+      "learning_rate": 1.9011007324913277e-05,
+      "loss": 0.6484,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1997799118158574,
+      "learning_rate": 1.9008699148134967e-05,
+      "loss": 0.5215,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6428992405588858,
+      "learning_rate": 1.9006388421463322e-05,
+      "loss": 0.5547,
+      "step": 1028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.891504510071167,
+      "learning_rate": 1.9004075145552378e-05,
+      "loss": 0.6556,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1724526069874113,
+      "learning_rate": 1.9001759321056905e-05,
+      "loss": 0.6074,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.977753375117036,
+      "learning_rate": 1.8999440948632397e-05,
+      "loss": 0.5811,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9658168009400012,
+      "learning_rate": 1.899712002893506e-05,
+      "loss": 0.615,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7701225634745041,
+      "learning_rate": 1.899479656262183e-05,
+      "loss": 0.6234,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7767318990917924,
+      "learning_rate": 1.8992470550350356e-05,
+      "loss": 0.5903,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.466032522561397,
+      "learning_rate": 1.8990141992779008e-05,
+      "loss": 0.5373,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.702004883274978,
+      "learning_rate": 1.8987810890566885e-05,
+      "loss": 0.6174,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8835695493474378,
+      "learning_rate": 1.8985477244373796e-05,
+      "loss": 0.6725,
+      "step": 1037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9318841758753866,
+      "learning_rate": 1.898314105486028e-05,
+      "loss": 0.4789,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2374247051744196,
+      "learning_rate": 1.8980802322687584e-05,
+      "loss": 0.474,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6885223177499413,
+      "learning_rate": 1.8978461048517686e-05,
+      "loss": 0.5381,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2379396990222884,
+      "learning_rate": 1.8976117233013278e-05,
+      "loss": 0.6344,
+      "step": 1041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5767332209189808,
+      "learning_rate": 1.8973770876837772e-05,
+      "loss": 0.5456,
+      "step": 1042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.09438517431333,
+      "learning_rate": 1.8971421980655295e-05,
+      "loss": 0.6676,
+      "step": 1043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53232937930043,
+      "learning_rate": 1.8969070545130702e-05,
+      "loss": 0.6695,
+      "step": 1044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6315381597309822,
+      "learning_rate": 1.896671657092956e-05,
+      "loss": 0.6313,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3725470832921574,
+      "learning_rate": 1.8964360058718162e-05,
+      "loss": 0.5394,
+      "step": 1046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4510902640686456,
+      "learning_rate": 1.8962001009163506e-05,
+      "loss": 0.4225,
+      "step": 1047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.688433766085792,
+      "learning_rate": 1.8959639422933316e-05,
+      "loss": 0.5366,
+      "step": 1048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8041894069243771,
+      "learning_rate": 1.8957275300696036e-05,
+      "loss": 0.5436,
+      "step": 1049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2899517083677385,
+      "learning_rate": 1.895490864312083e-05,
+      "loss": 0.7155,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.532930914788766,
+      "learning_rate": 1.895253945087757e-05,
+      "loss": 0.6417,
+      "step": 1051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9951963872554477,
+      "learning_rate": 1.8950167724636856e-05,
+      "loss": 0.6174,
+      "step": 1052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4802120122989038,
+      "learning_rate": 1.894779346506999e-05,
+      "loss": 0.5807,
+      "step": 1053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.73396589248917,
+      "learning_rate": 1.8945416672849014e-05,
+      "loss": 0.516,
+      "step": 1054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2850523288460347,
+      "learning_rate": 1.8943037348646668e-05,
+      "loss": 0.5705,
+      "step": 1055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9493168160057928,
+      "learning_rate": 1.8940655493136415e-05,
+      "loss": 0.5483,
+      "step": 1056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6548939980627275,
+      "learning_rate": 1.8938271106992433e-05,
+      "loss": 0.6304,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6096367233004916,
+      "learning_rate": 1.893588419088962e-05,
+      "loss": 0.6511,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6391471102425252,
+      "learning_rate": 1.8933494745503585e-05,
+      "loss": 0.4976,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.896683183424374,
+      "learning_rate": 1.8931102771510657e-05,
+      "loss": 0.6074,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.920051239259327,
+      "learning_rate": 1.8928708269587876e-05,
+      "loss": 0.5702,
+      "step": 1061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9961909912605602,
+      "learning_rate": 1.8926311240413008e-05,
+      "loss": 0.5523,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.224743053352204,
+      "learning_rate": 1.892391168466452e-05,
+      "loss": 0.5424,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5320042183687028,
+      "learning_rate": 1.8921509603021606e-05,
+      "loss": 0.6239,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9592121331861676,
+      "learning_rate": 1.8919104996164167e-05,
+      "loss": 0.5878,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2571130147055323,
+      "learning_rate": 1.8916697864772822e-05,
+      "loss": 0.6344,
+      "step": 1066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.758327572488884,
+      "learning_rate": 1.8914288209528907e-05,
+      "loss": 0.6576,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.123499160256295,
+      "learning_rate": 1.891187603111447e-05,
+      "loss": 0.5785,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7048072525922366,
+      "learning_rate": 1.8909461330212267e-05,
+      "loss": 0.6189,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.899966059678097,
+      "learning_rate": 1.890704410750578e-05,
+      "loss": 0.544,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5323636878536173,
+      "learning_rate": 1.89046243636792e-05,
+      "loss": 0.5343,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.667750035532417,
+      "learning_rate": 1.8902202099417425e-05,
+      "loss": 0.6207,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7558831170789135,
+      "learning_rate": 1.8899777315406073e-05,
+      "loss": 0.5915,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8218793989964532,
+      "learning_rate": 1.8897350012331478e-05,
+      "loss": 0.6305,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.129753051396958,
+      "learning_rate": 1.889492019088068e-05,
+      "loss": 0.6505,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7155847038199314,
+      "learning_rate": 1.889248785174143e-05,
+      "loss": 0.6303,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.804039802327124,
+      "learning_rate": 1.8890052995602207e-05,
+      "loss": 0.6016,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9334062473752995,
+      "learning_rate": 1.8887615623152188e-05,
+      "loss": 0.6456,
+      "step": 1078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7641348712751566,
+      "learning_rate": 1.888517573508126e-05,
+      "loss": 0.589,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9952305133691843,
+      "learning_rate": 1.8882733332080038e-05,
+      "loss": 0.5919,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.682505492573714,
+      "learning_rate": 1.888028841483983e-05,
+      "loss": 0.5052,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5895316158861066,
+      "learning_rate": 1.887784098405267e-05,
+      "loss": 0.5386,
+      "step": 1082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8538730970240576,
+      "learning_rate": 1.88753910404113e-05,
+      "loss": 0.6721,
+      "step": 1083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9152304571062087,
+      "learning_rate": 1.8872938584609164e-05,
+      "loss": 0.5754,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.239711805326106,
+      "learning_rate": 1.887048361734043e-05,
+      "loss": 0.5197,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5732426912597839,
+      "learning_rate": 1.886802613929997e-05,
+      "loss": 0.5301,
+      "step": 1086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8962821563981125,
+      "learning_rate": 1.8865566151183365e-05,
+      "loss": 0.6018,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.254402207240487,
+      "learning_rate": 1.8863103653686917e-05,
+      "loss": 0.7679,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8733447164697048,
+      "learning_rate": 1.8860638647507622e-05,
+      "loss": 0.6739,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5799484966091293,
+      "learning_rate": 1.8858171133343202e-05,
+      "loss": 0.6204,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5883668483065347,
+      "learning_rate": 1.885570111189208e-05,
+      "loss": 0.6011,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2279169836983295,
+      "learning_rate": 1.8853228583853384e-05,
+      "loss": 0.641,
+      "step": 1092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8161785445590806,
+      "learning_rate": 1.8850753549926967e-05,
+      "loss": 0.6818,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9072856063202688,
+      "learning_rate": 1.8848276010813377e-05,
+      "loss": 0.6783,
+      "step": 1094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7315524026448734,
+      "learning_rate": 1.8845795967213876e-05,
+      "loss": 0.5363,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.633384226079634,
+      "learning_rate": 1.884331341983044e-05,
+      "loss": 0.5168,
+      "step": 1096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.825950197202617,
+      "learning_rate": 1.8840828369365743e-05,
+      "loss": 0.6335,
+      "step": 1097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8097762802568034,
+      "learning_rate": 1.8838340816523175e-05,
+      "loss": 0.7129,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.732785162190959,
+      "learning_rate": 1.883585076200683e-05,
+      "loss": 0.6131,
+      "step": 1099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7041382834824075,
+      "learning_rate": 1.883335820652152e-05,
+      "loss": 0.5915,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4490540031498975,
+      "learning_rate": 1.8830863150772754e-05,
+      "loss": 0.6075,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4620699363292047,
+      "learning_rate": 1.882836559546675e-05,
+      "loss": 0.5014,
+      "step": 1102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0092642708040869,
+      "learning_rate": 1.8825865541310438e-05,
+      "loss": 0.6948,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9730692551316555,
+      "learning_rate": 1.882336298901145e-05,
+      "loss": 0.7389,
+      "step": 1104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5003853079654117,
+      "learning_rate": 1.8820857939278136e-05,
+      "loss": 0.5018,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1960139092808704,
+      "learning_rate": 1.8818350392819535e-05,
+      "loss": 0.6568,
+      "step": 1106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9950116797664248,
+      "learning_rate": 1.881584035034541e-05,
+      "loss": 0.6263,
+      "step": 1107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8612691057436996,
+      "learning_rate": 1.8813327812566217e-05,
+      "loss": 0.5387,
+      "step": 1108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.481900008459278,
+      "learning_rate": 1.881081278019313e-05,
+      "loss": 0.5865,
+      "step": 1109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5020146140586885,
+      "learning_rate": 1.8808295253938025e-05,
+      "loss": 0.5789,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6499261780515717,
+      "learning_rate": 1.8805775234513476e-05,
+      "loss": 0.6057,
+      "step": 1111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3040075170111236,
+      "learning_rate": 1.8803252722632775e-05,
+      "loss": 0.7035,
+      "step": 1112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5461510560684477,
+      "learning_rate": 1.880072771900991e-05,
+      "loss": 0.6097,
+      "step": 1113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.339627597126912,
+      "learning_rate": 1.879820022435958e-05,
+      "loss": 0.5055,
+      "step": 1114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0110202237476558,
+      "learning_rate": 1.8795670239397184e-05,
+      "loss": 0.5283,
+      "step": 1115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.113711577796458,
+      "learning_rate": 1.8793137764838834e-05,
+      "loss": 0.5891,
+      "step": 1116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5761852982338955,
+      "learning_rate": 1.879060280140134e-05,
+      "loss": 0.6154,
+      "step": 1117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.710997046313092,
+      "learning_rate": 1.878806534980221e-05,
+      "loss": 0.6282,
+      "step": 1118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9261231962566886,
+      "learning_rate": 1.8785525410759676e-05,
+      "loss": 0.6976,
+      "step": 1119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9916162446356909,
+      "learning_rate": 1.878298298499266e-05,
+      "loss": 0.6148,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8765494078027776,
+      "learning_rate": 1.8780438073220785e-05,
+      "loss": 0.665,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.122273560802957,
+      "learning_rate": 1.8777890676164387e-05,
+      "loss": 0.5608,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7213086834326279,
+      "learning_rate": 1.8775340794544497e-05,
+      "loss": 0.5276,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3360314965375784,
+      "learning_rate": 1.877278842908286e-05,
+      "loss": 0.6887,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.944967586350892,
+      "learning_rate": 1.8770233580501913e-05,
+      "loss": 0.5885,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.173925875358129,
+      "learning_rate": 1.87676762495248e-05,
+      "loss": 0.5118,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6924342818766434,
+      "learning_rate": 1.8765116436875374e-05,
+      "loss": 0.5255,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4232110272495797,
+      "learning_rate": 1.876255414327818e-05,
+      "loss": 0.6539,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8924147482220333,
+      "learning_rate": 1.8759989369458468e-05,
+      "loss": 0.6361,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8302770351607525,
+      "learning_rate": 1.8757422116142198e-05,
+      "loss": 0.5763,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7463591297110725,
+      "learning_rate": 1.875485238405602e-05,
+      "loss": 0.651,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.10843835805321,
+      "learning_rate": 1.875228017392729e-05,
+      "loss": 0.5326,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7412255865575683,
+      "learning_rate": 1.8749705486484074e-05,
+      "loss": 0.585,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2185292936904895,
+      "learning_rate": 1.8747128322455128e-05,
+      "loss": 0.6332,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7287191923188843,
+      "learning_rate": 1.8744548682569914e-05,
+      "loss": 0.5855,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1290449080091722,
+      "learning_rate": 1.874196656755859e-05,
+      "loss": 0.6022,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.697787268686174,
+      "learning_rate": 1.873938197815202e-05,
+      "loss": 0.6051,
+      "step": 1137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5254113882849853,
+      "learning_rate": 1.8736794915081765e-05,
+      "loss": 0.6541,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8883533700092177,
+      "learning_rate": 1.8734205379080093e-05,
+      "loss": 0.55,
+      "step": 1139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9422693243309248,
+      "learning_rate": 1.8731613370879963e-05,
+      "loss": 0.6302,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5233645424287734,
+      "learning_rate": 1.8729018891215042e-05,
+      "loss": 0.5929,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.783214520963635,
+      "learning_rate": 1.8726421940819683e-05,
+      "loss": 0.5403,
+      "step": 1142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3807944419197455,
+      "learning_rate": 1.8723822520428954e-05,
+      "loss": 0.4982,
+      "step": 1143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6888607665056987,
+      "learning_rate": 1.8721220630778613e-05,
+      "loss": 0.6101,
+      "step": 1144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9322692003359307,
+      "learning_rate": 1.871861627260512e-05,
+      "loss": 0.638,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7355398861731453,
+      "learning_rate": 1.8716009446645636e-05,
+      "loss": 0.5384,
+      "step": 1146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8379530314640322,
+      "learning_rate": 1.8713400153638013e-05,
+      "loss": 0.6364,
+      "step": 1147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8618847473922935,
+      "learning_rate": 1.8710788394320807e-05,
+      "loss": 0.5983,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6333091212537691,
+      "learning_rate": 1.870817416943327e-05,
+      "loss": 0.5929,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6618370463001289,
+      "learning_rate": 1.8705557479715363e-05,
+      "loss": 0.6303,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9527753420125138,
+      "learning_rate": 1.870293832590772e-05,
+      "loss": 0.5977,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6660612944595454,
+      "learning_rate": 1.8700316708751693e-05,
+      "loss": 0.593,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7066355182692223,
+      "learning_rate": 1.8697692628989327e-05,
+      "loss": 0.5616,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4504635275207403,
+      "learning_rate": 1.869506608736336e-05,
+      "loss": 0.5071,
+      "step": 1154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8846540427556524,
+      "learning_rate": 1.869243708461723e-05,
+      "loss": 0.5546,
+      "step": 1155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.887688463476996,
+      "learning_rate": 1.8689805621495072e-05,
+      "loss": 0.5504,
+      "step": 1156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1579446653340506,
+      "learning_rate": 1.8687171698741714e-05,
+      "loss": 0.6042,
+      "step": 1157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.835356282843814,
+      "learning_rate": 1.868453531710268e-05,
+      "loss": 0.6965,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6583327283042895,
+      "learning_rate": 1.8681896477324198e-05,
+      "loss": 0.6122,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8161586887567571,
+      "learning_rate": 1.8679255180153184e-05,
+      "loss": 0.6715,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9726597479026775,
+      "learning_rate": 1.867661142633725e-05,
+      "loss": 0.6598,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7991087254361058,
+      "learning_rate": 1.8673965216624704e-05,
+      "loss": 0.6085,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.804261428797347,
+      "learning_rate": 1.8671316551764552e-05,
+      "loss": 0.6246,
+      "step": 1163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9207852747346121,
+      "learning_rate": 1.8668665432506496e-05,
+      "loss": 0.6117,
+      "step": 1164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.215985183862306,
+      "learning_rate": 1.8666011859600925e-05,
+      "loss": 0.6396,
+      "step": 1165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.637712392820533,
+      "learning_rate": 1.8663355833798927e-05,
+      "loss": 0.6163,
+      "step": 1166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6888730468310489,
+      "learning_rate": 1.8660697355852288e-05,
+      "loss": 0.6195,
+      "step": 1167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4457770065520212,
+      "learning_rate": 1.865803642651348e-05,
+      "loss": 0.5702,
+      "step": 1168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6850807173065991,
+      "learning_rate": 1.8655373046535682e-05,
+      "loss": 0.598,
+      "step": 1169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8330919878165486,
+      "learning_rate": 1.8652707216672747e-05,
+      "loss": 0.5772,
+      "step": 1170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9034453091420747,
+      "learning_rate": 1.865003893767924e-05,
+      "loss": 0.5693,
+      "step": 1171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1065384862348813,
+      "learning_rate": 1.864736821031041e-05,
+      "loss": 0.69,
+      "step": 1172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.323865767953787,
+      "learning_rate": 1.8644695035322203e-05,
+      "loss": 0.5667,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0861398697204647,
+      "learning_rate": 1.864201941347125e-05,
+      "loss": 0.5549,
+      "step": 1174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.756114109672989,
+      "learning_rate": 1.863934134551488e-05,
+      "loss": 0.6242,
+      "step": 1175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7073186403874572,
+      "learning_rate": 1.8636660832211126e-05,
+      "loss": 0.6022,
+      "step": 1176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6482743937156552,
+      "learning_rate": 1.8633977874318686e-05,
+      "loss": 0.5884,
+      "step": 1177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.605758807596704,
+      "learning_rate": 1.8631292472596978e-05,
+      "loss": 0.6283,
+      "step": 1178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.156635211651928,
+      "learning_rate": 1.862860462780609e-05,
+      "loss": 0.626,
+      "step": 1179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.106674649180831,
+      "learning_rate": 1.8625914340706818e-05,
+      "loss": 0.6712,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5540086683180194,
+      "learning_rate": 1.862322161206064e-05,
+      "loss": 0.5668,
+      "step": 1181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5883023505967413,
+      "learning_rate": 1.862052644262972e-05,
+      "loss": 0.5729,
+      "step": 1182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8047681131635638,
+      "learning_rate": 1.8617828833176935e-05,
+      "loss": 0.513,
+      "step": 1183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5417358935059224,
+      "learning_rate": 1.8615128784465826e-05,
+      "loss": 0.5975,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7104766791845873,
+      "learning_rate": 1.861242629726064e-05,
+      "loss": 0.5711,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7271991024480775,
+      "learning_rate": 1.8609721372326305e-05,
+      "loss": 0.5425,
+      "step": 1186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9553781738296483,
+      "learning_rate": 1.8607014010428454e-05,
+      "loss": 0.6409,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3712979983009042,
+      "learning_rate": 1.860430421233339e-05,
+      "loss": 0.6238,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6784715376195491,
+      "learning_rate": 1.8601591978808126e-05,
+      "loss": 0.5694,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1314302465698547,
+      "learning_rate": 1.8598877310620347e-05,
+      "loss": 0.5447,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.170103476078789,
+      "learning_rate": 1.859616020853843e-05,
+      "loss": 0.5717,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8725094968887104,
+      "learning_rate": 1.859344067333146e-05,
+      "loss": 0.6319,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5891810609725092,
+      "learning_rate": 1.859071870576918e-05,
+      "loss": 0.6631,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.261091025040417,
+      "learning_rate": 1.8587994306622047e-05,
+      "loss": 0.6529,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6162489218074974,
+      "learning_rate": 1.8585267476661187e-05,
+      "loss": 0.6705,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.468864119684275,
+      "learning_rate": 1.8582538216658435e-05,
+      "loss": 0.6371,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1587129101321763,
+      "learning_rate": 1.85798065273863e-05,
+      "loss": 0.6299,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9084719535108412,
+      "learning_rate": 1.857707240961797e-05,
+      "loss": 0.6272,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.655247211088756,
+      "learning_rate": 1.8574335864127344e-05,
+      "loss": 0.6288,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0419484236634324,
+      "learning_rate": 1.8571596891688988e-05,
+      "loss": 0.6396,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9670638061627015,
+      "learning_rate": 1.8568855493078165e-05,
+      "loss": 0.7279,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.813613343612414,
+      "learning_rate": 1.8566111669070822e-05,
+      "loss": 0.6209,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8996201966407016,
+      "learning_rate": 1.8563365420443594e-05,
+      "loss": 0.5873,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1658208080410124,
+      "learning_rate": 1.85606167479738e-05,
+      "loss": 0.5597,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8679680595829264,
+      "learning_rate": 1.8557865652439445e-05,
+      "loss": 0.5539,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7978368850714,
+      "learning_rate": 1.8555112134619218e-05,
+      "loss": 0.5613,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0982402746445947,
+      "learning_rate": 1.85523561952925e-05,
+      "loss": 0.6154,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6951394984145132,
+      "learning_rate": 1.854959783523936e-05,
+      "loss": 0.5276,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5966286982008586,
+      "learning_rate": 1.8546837055240536e-05,
+      "loss": 0.5403,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.603948232025248,
+      "learning_rate": 1.854407385607746e-05,
+      "loss": 0.6132,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2634505172980326,
+      "learning_rate": 1.8541308238532257e-05,
+      "loss": 0.6504,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7688397866457595,
+      "learning_rate": 1.853854020338773e-05,
+      "loss": 0.6251,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8034603744384519,
+      "learning_rate": 1.853576975142736e-05,
+      "loss": 0.6175,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6392409102848131,
+      "learning_rate": 1.853299688343532e-05,
+      "loss": 0.5751,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7897696381516346,
+      "learning_rate": 1.8530221600196462e-05,
+      "loss": 0.5544,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5794587447785386,
+      "learning_rate": 1.8527443902496325e-05,
+      "loss": 0.6947,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8204420988309062,
+      "learning_rate": 1.8524663791121134e-05,
+      "loss": 0.6611,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8237674132476802,
+      "learning_rate": 1.852188126685779e-05,
+      "loss": 0.5002,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8538965517047314,
+      "learning_rate": 1.851909633049388e-05,
+      "loss": 0.4553,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4084532951589739,
+      "learning_rate": 1.8516308982817685e-05,
+      "loss": 0.45,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7879971164594732,
+      "learning_rate": 1.851351922461814e-05,
+      "loss": 0.593,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8219884534041593,
+      "learning_rate": 1.851072705668489e-05,
+      "loss": 0.6031,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9522516145852415,
+      "learning_rate": 1.8507932479808254e-05,
+      "loss": 0.4863,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.151814211197761,
+      "learning_rate": 1.8505135494779228e-05,
+      "loss": 0.5029,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5639374455329988,
+      "learning_rate": 1.8502336102389494e-05,
+      "loss": 0.6208,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9635132016238852,
+      "learning_rate": 1.8499534303431414e-05,
+      "loss": 0.76,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.521895907794719,
+      "learning_rate": 1.849673009869803e-05,
+      "loss": 0.6254,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.349848083600444,
+      "learning_rate": 1.8493923488983066e-05,
+      "loss": 0.6209,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.807335806291125,
+      "learning_rate": 1.849111447508093e-05,
+      "loss": 0.5969,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.784411257227522,
+      "learning_rate": 1.8488303057786707e-05,
+      "loss": 0.5708,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7827601766495464,
+      "learning_rate": 1.848548923789616e-05,
+      "loss": 0.484,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9544999003482917,
+      "learning_rate": 1.8482673016205734e-05,
+      "loss": 0.601,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8414612470243565,
+      "learning_rate": 1.847985439351256e-05,
+      "loss": 0.6411,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9199524256630016,
+      "learning_rate": 1.8477033370614438e-05,
+      "loss": 0.7465,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7655323226404491,
+      "learning_rate": 1.8474209948309852e-05,
+      "loss": 0.6078,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6324113424890832,
+      "learning_rate": 1.8471384127397974e-05,
+      "loss": 0.562,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.796857522650803,
+      "learning_rate": 1.8468555908678638e-05,
+      "loss": 0.6879,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8269532309198572,
+      "learning_rate": 1.846572529295237e-05,
+      "loss": 0.6978,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9147766742026686,
+      "learning_rate": 1.8462892281020365e-05,
+      "loss": 0.5578,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7167603780206757,
+      "learning_rate": 1.8460056873684503e-05,
+      "loss": 0.4788,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4170784075636333,
+      "learning_rate": 1.8457219071747345e-05,
+      "loss": 0.5239,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8112853100500073,
+      "learning_rate": 1.845437887601212e-05,
+      "loss": 0.5838,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6101532821530384,
+      "learning_rate": 1.845153628728274e-05,
+      "loss": 0.6505,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5422081037599227,
+      "learning_rate": 1.8448691306363798e-05,
+      "loss": 0.6402,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.984449546796776,
+      "learning_rate": 1.844584393406055e-05,
+      "loss": 0.4963,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8871130175391342,
+      "learning_rate": 1.8442994171178948e-05,
+      "loss": 0.6479,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7988580629737183,
+      "learning_rate": 1.844014201852561e-05,
+      "loss": 0.6007,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5839819157839576,
+      "learning_rate": 1.8437287476907828e-05,
+      "loss": 0.6633,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.729988144343722,
+      "learning_rate": 1.8434430547133576e-05,
+      "loss": 0.6056,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3771190524596706,
+      "learning_rate": 1.8431571230011504e-05,
+      "loss": 0.6305,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.48493100466631,
+      "learning_rate": 1.8428709526350932e-05,
+      "loss": 0.5867,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8314966246038369,
+      "learning_rate": 1.8425845436961863e-05,
+      "loss": 0.6697,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6270067054286494,
+      "learning_rate": 1.842297896265497e-05,
+      "loss": 0.657,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3992817019536579,
+      "learning_rate": 1.8420110104241598e-05,
+      "loss": 0.4389,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5989230359283355,
+      "learning_rate": 1.841723886253378e-05,
+      "loss": 0.5221,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7468142325920302,
+      "learning_rate": 1.8414365238344208e-05,
+      "loss": 0.6739,
+      "step": 1256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0201291122976475,
+      "learning_rate": 1.8411489232486256e-05,
+      "loss": 0.6245,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7307849650414062,
+      "learning_rate": 1.8408610845773974e-05,
+      "loss": 0.5748,
+      "step": 1258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7072196458941935,
+      "learning_rate": 1.8405730079022083e-05,
+      "loss": 0.5515,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.463891956250344,
+      "learning_rate": 1.8402846933045974e-05,
+      "loss": 0.6339,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6380946110155314,
+      "learning_rate": 1.8399961408661725e-05,
+      "loss": 0.575,
+      "step": 1261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8989063120894096,
+      "learning_rate": 1.8397073506686066e-05,
+      "loss": 0.5585,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.735675271436225,
+      "learning_rate": 1.8394183227936418e-05,
+      "loss": 0.547,
+      "step": 1263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.124340892143138,
+      "learning_rate": 1.839129057323087e-05,
+      "loss": 0.5854,
+      "step": 1264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5911632600900032,
+      "learning_rate": 1.8388395543388174e-05,
+      "loss": 0.6033,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.733830144325735,
+      "learning_rate": 1.838549813922777e-05,
+      "loss": 0.687,
+      "step": 1266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6814223823792713,
+      "learning_rate": 1.838259836156976e-05,
+      "loss": 0.6733,
+      "step": 1267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8202133353045973,
+      "learning_rate": 1.8379696211234918e-05,
+      "loss": 0.6258,
+      "step": 1268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.05478889532283,
+      "learning_rate": 1.8376791689044693e-05,
+      "loss": 0.5309,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6002201815642687,
+      "learning_rate": 1.8373884795821203e-05,
+      "loss": 0.5989,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6795166342313073,
+      "learning_rate": 1.8370975532387237e-05,
+      "loss": 0.5634,
+      "step": 1271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.715722304244586,
+      "learning_rate": 1.8368063899566263e-05,
+      "loss": 0.5515,
+      "step": 1272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.39094143514639,
+      "learning_rate": 1.8365149898182403e-05,
+      "loss": 0.5904,
+      "step": 1273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8088101363662044,
+      "learning_rate": 1.8362233529060464e-05,
+      "loss": 0.6037,
+      "step": 1274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.960902241109917,
+      "learning_rate": 1.8359314793025914e-05,
+      "loss": 0.6655,
+      "step": 1275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7498513463456609,
+      "learning_rate": 1.8356393690904904e-05,
+      "loss": 0.6266,
+      "step": 1276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6304165154261505,
+      "learning_rate": 1.8353470223524237e-05,
+      "loss": 0.5708,
+      "step": 1277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5499202771956044,
+      "learning_rate": 1.8350544391711396e-05,
+      "loss": 0.5578,
+      "step": 1278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7566882133526913,
+      "learning_rate": 1.8347616196294536e-05,
+      "loss": 0.6116,
+      "step": 1279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8559295967130796,
+      "learning_rate": 1.8344685638102472e-05,
+      "loss": 0.5003,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3387043239677814,
+      "learning_rate": 1.8341752717964696e-05,
+      "loss": 0.5783,
+      "step": 1281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0056985497524569,
+      "learning_rate": 1.8338817436711358e-05,
+      "loss": 0.7243,
+      "step": 1282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.47998146951202,
+      "learning_rate": 1.833587979517329e-05,
+      "loss": 0.576,
+      "step": 1283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3650254295517112,
+      "learning_rate": 1.8332939794181986e-05,
+      "loss": 0.652,
+      "step": 1284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7792313094473025,
+      "learning_rate": 1.83299974345696e-05,
+      "loss": 0.5499,
+      "step": 1285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6406680001601277,
+      "learning_rate": 1.832705271716897e-05,
+      "loss": 0.5796,
+      "step": 1286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4986339632861518,
+      "learning_rate": 1.832410564281358e-05,
+      "loss": 0.4794,
+      "step": 1287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8621527384074734,
+      "learning_rate": 1.8321156212337604e-05,
+      "loss": 0.6644,
+      "step": 1288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6457817231662772,
+      "learning_rate": 1.8318204426575873e-05,
+      "loss": 0.569,
+      "step": 1289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8507444114187144,
+      "learning_rate": 1.831525028636387e-05,
+      "loss": 0.6372,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1978072210157658,
+      "learning_rate": 1.8312293792537773e-05,
+      "loss": 0.6132,
+      "step": 1291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6390729951606373,
+      "learning_rate": 1.8309334945934402e-05,
+      "loss": 0.5993,
+      "step": 1292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.54151024246081,
+      "learning_rate": 1.830637374739126e-05,
+      "loss": 0.4934,
+      "step": 1293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9508744145995702,
+      "learning_rate": 1.83034101977465e-05,
+      "loss": 0.5921,
+      "step": 1294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7983180266414853,
+      "learning_rate": 1.8300444297838955e-05,
+      "loss": 0.6212,
+      "step": 1295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5362298090075197,
+      "learning_rate": 1.8297476048508113e-05,
+      "loss": 0.5209,
+      "step": 1296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0603967835189163,
+      "learning_rate": 1.829450545059413e-05,
+      "loss": 0.6417,
+      "step": 1297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5294795402537695,
+      "learning_rate": 1.829153250493783e-05,
+      "loss": 0.5569,
+      "step": 1298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6345793962504318,
+      "learning_rate": 1.8288557212380703e-05,
+      "loss": 0.5572,
+      "step": 1299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5110228073103502,
+      "learning_rate": 1.828557957376489e-05,
+      "loss": 0.6095,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6967088909233559,
+      "learning_rate": 1.8282599589933214e-05,
+      "loss": 0.5465,
+      "step": 1301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.599893545063506,
+      "learning_rate": 1.8279617261729142e-05,
+      "loss": 0.6212,
+      "step": 1302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0546020953060045,
+      "learning_rate": 1.827663258999683e-05,
+      "loss": 0.7443,
+      "step": 1303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7776874769784723,
+      "learning_rate": 1.827364557558107e-05,
+      "loss": 0.5279,
+      "step": 1304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5150715827318568,
+      "learning_rate": 1.827065621932734e-05,
+      "loss": 0.5307,
+      "step": 1305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7992637488152552,
+      "learning_rate": 1.8267664522081767e-05,
+      "loss": 0.5708,
+      "step": 1306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8564705098626024,
+      "learning_rate": 1.8264670484691144e-05,
+      "loss": 0.62,
+      "step": 1307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.144346764065768,
+      "learning_rate": 1.8261674108002925e-05,
+      "loss": 0.5938,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5841707691070395,
+      "learning_rate": 1.8258675392865235e-05,
+      "loss": 0.5358,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9520868421952041,
+      "learning_rate": 1.8255674340126847e-05,
+      "loss": 0.7091,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9114641947224174,
+      "learning_rate": 1.8252670950637206e-05,
+      "loss": 0.6593,
+      "step": 1311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9212054700238175,
+      "learning_rate": 1.8249665225246417e-05,
+      "loss": 0.6121,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.089772106528644,
+      "learning_rate": 1.824665716480524e-05,
+      "loss": 0.6866,
+      "step": 1313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3945097963628266,
+      "learning_rate": 1.82436467701651e-05,
+      "loss": 0.5817,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0547276626207855,
+      "learning_rate": 1.8240634042178086e-05,
+      "loss": 0.6147,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7254264744027337,
+      "learning_rate": 1.8237618981696944e-05,
+      "loss": 0.5817,
+      "step": 1316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.429476733387876,
+      "learning_rate": 1.823460158957508e-05,
+      "loss": 0.591,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7885333709490934,
+      "learning_rate": 1.823158186666656e-05,
+      "loss": 0.5915,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0652200551169986,
+      "learning_rate": 1.8228559813826106e-05,
+      "loss": 0.6047,
+      "step": 1319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.34964932553087,
+      "learning_rate": 1.8225535431909113e-05,
+      "loss": 0.6351,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6325247479918772,
+      "learning_rate": 1.822250872177162e-05,
+      "loss": 0.5765,
+      "step": 1321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.616373441526426,
+      "learning_rate": 1.821947968427033e-05,
+      "loss": 0.6975,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.888514559288,
+      "learning_rate": 1.821644832026261e-05,
+      "loss": 0.6428,
+      "step": 1323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8107686321109064,
+      "learning_rate": 1.821341463060648e-05,
+      "loss": 0.6216,
+      "step": 1324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7375821624267775,
+      "learning_rate": 1.8210378616160617e-05,
+      "loss": 0.5389,
+      "step": 1325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.334798477144555,
+      "learning_rate": 1.8207340277784357e-05,
+      "loss": 0.6086,
+      "step": 1326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5234801166901815,
+      "learning_rate": 1.82042996163377e-05,
+      "loss": 0.598,
+      "step": 1327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9434772970136951,
+      "learning_rate": 1.82012566326813e-05,
+      "loss": 0.7402,
+      "step": 1328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.887694191398026,
+      "learning_rate": 1.819821132767646e-05,
+      "loss": 0.5368,
+      "step": 1329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.384908605813789,
+      "learning_rate": 1.8195163702185153e-05,
+      "loss": 0.6673,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.489002304993147,
+      "learning_rate": 1.819211375707e-05,
+      "loss": 0.6456,
+      "step": 1331
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.89034694154408,
+      "learning_rate": 1.8189061493194283e-05,
+      "loss": 0.6979,
+      "step": 1332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4855757922897614,
+      "learning_rate": 1.8186006911421937e-05,
+      "loss": 0.5732,
+      "step": 1333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6255787264485764,
+      "learning_rate": 1.818295001261756e-05,
+      "loss": 0.6045,
+      "step": 1334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.664134360233818,
+      "learning_rate": 1.8179890797646398e-05,
+      "loss": 0.5207,
+      "step": 1335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7174624303397763,
+      "learning_rate": 1.817682926737435e-05,
+      "loss": 0.5029,
+      "step": 1336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9426462729387601,
+      "learning_rate": 1.8173765422667987e-05,
+      "loss": 0.5398,
+      "step": 1337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7337475940857994,
+      "learning_rate": 1.817069926439451e-05,
+      "loss": 0.5166,
+      "step": 1338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8284325682523377,
+      "learning_rate": 1.81676307934218e-05,
+      "loss": 0.5989,
+      "step": 1339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.090064534381596,
+      "learning_rate": 1.8164560010618377e-05,
+      "loss": 0.573,
+      "step": 1340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.473481839369123,
+      "learning_rate": 1.816148691685342e-05,
+      "loss": 0.6048,
+      "step": 1341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5851250288228336,
+      "learning_rate": 1.815841151299676e-05,
+      "loss": 0.5309,
+      "step": 1342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.019406505529589,
+      "learning_rate": 1.8155333799918883e-05,
+      "loss": 0.7317,
+      "step": 1343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.568657897694507,
+      "learning_rate": 1.8152253778490933e-05,
+      "loss": 0.6552,
+      "step": 1344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1838778038779223,
+      "learning_rate": 1.8149171449584705e-05,
+      "loss": 0.6169,
+      "step": 1345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5790659423794096,
+      "learning_rate": 1.814608681407264e-05,
+      "loss": 0.6117,
+      "step": 1346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7610401641825324,
+      "learning_rate": 1.814299987282784e-05,
+      "loss": 0.5366,
+      "step": 1347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.497098482162508,
+      "learning_rate": 1.8139910626724058e-05,
+      "loss": 0.5532,
+      "step": 1348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6857512396187493,
+      "learning_rate": 1.8136819076635696e-05,
+      "loss": 0.6104,
+      "step": 1349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6646167589047016,
+      "learning_rate": 1.8133725223437815e-05,
+      "loss": 0.5206,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6693314475035927,
+      "learning_rate": 1.813062906800612e-05,
+      "loss": 0.6594,
+      "step": 1351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3152811465189664,
+      "learning_rate": 1.8127530611216973e-05,
+      "loss": 0.478,
+      "step": 1352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5725423506537217,
+      "learning_rate": 1.8124429853947387e-05,
+      "loss": 0.5264,
+      "step": 1353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7983651325337942,
+      "learning_rate": 1.8121326797075022e-05,
+      "loss": 0.5489,
+      "step": 1354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6810782160270077,
+      "learning_rate": 1.811822144147819e-05,
+      "loss": 0.6156,
+      "step": 1355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.805456547548261,
+      "learning_rate": 1.8115113788035863e-05,
+      "loss": 0.6581,
+      "step": 1356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.040937898422809,
+      "learning_rate": 1.8112003837627646e-05,
+      "loss": 0.5286,
+      "step": 1357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7487459540896526,
+      "learning_rate": 1.8108891591133812e-05,
+      "loss": 0.6092,
+      "step": 1358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.999590752694541,
+      "learning_rate": 1.810577704943527e-05,
+      "loss": 0.6877,
+      "step": 1359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4341381580915313,
+      "learning_rate": 1.8102660213413593e-05,
+      "loss": 0.5649,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.129670250015723,
+      "learning_rate": 1.8099541083950988e-05,
+      "loss": 0.6771,
+      "step": 1361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.17749681288786,
+      "learning_rate": 1.8096419661930313e-05,
+      "loss": 0.5766,
+      "step": 1362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.597431665336124,
+      "learning_rate": 1.809329594823509e-05,
+      "loss": 0.6634,
+      "step": 1363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.543785551863803,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.656,
+      "step": 1364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9994739313030316,
+      "learning_rate": 1.808704164935828e-05,
+      "loss": 0.6547,
+      "step": 1365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0807861008693456,
+      "learning_rate": 1.8083911065946958e-05,
+      "loss": 0.5884,
+      "step": 1366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5729983279196316,
+      "learning_rate": 1.8080778194401615e-05,
+      "loss": 0.7203,
+      "step": 1367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.172983125571174,
+      "learning_rate": 1.8077643035609006e-05,
+      "loss": 0.6256,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4811602163098794,
+      "learning_rate": 1.807450559045653e-05,
+      "loss": 0.5328,
+      "step": 1369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.588898355152529,
+      "learning_rate": 1.807136585983223e-05,
+      "loss": 0.5186,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8019301718996508,
+      "learning_rate": 1.8068223844624806e-05,
+      "loss": 0.692,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6245846178044714,
+      "learning_rate": 1.80650795457236e-05,
+      "loss": 0.7051,
+      "step": 1372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6006581999106753,
+      "learning_rate": 1.806193296401859e-05,
+      "loss": 0.6358,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5274613766596532,
+      "learning_rate": 1.8058784100400418e-05,
+      "loss": 0.6364,
+      "step": 1374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1122924038729671,
+      "learning_rate": 1.8055632955760364e-05,
+      "loss": 0.7048,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7634010287853437,
+      "learning_rate": 1.8052479530990348e-05,
+      "loss": 0.6292,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0015475078698364,
+      "learning_rate": 1.8049323826982942e-05,
+      "loss": 0.6197,
+      "step": 1377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6627154387349075,
+      "learning_rate": 1.804616584463136e-05,
+      "loss": 0.6409,
+      "step": 1378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5627020572008659,
+      "learning_rate": 1.804300558482946e-05,
+      "loss": 0.4585,
+      "step": 1379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7012212259965924,
+      "learning_rate": 1.8039843048471756e-05,
+      "loss": 0.5956,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3852960340075495,
+      "learning_rate": 1.8036678236453387e-05,
+      "loss": 0.5648,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9493532789964745,
+      "learning_rate": 1.8033511149670152e-05,
+      "loss": 0.5107,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9578744395285947,
+      "learning_rate": 1.803034178901849e-05,
+      "loss": 0.6063,
+      "step": 1383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0308549675212975,
+      "learning_rate": 1.8027170155395476e-05,
+      "loss": 0.5878,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4638113242527075,
+      "learning_rate": 1.8023996249698836e-05,
+      "loss": 0.5867,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6958022989124528,
+      "learning_rate": 1.802082007282694e-05,
+      "loss": 0.6342,
+      "step": 1386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6130645266905699,
+      "learning_rate": 1.8017641625678797e-05,
+      "loss": 0.6463,
+      "step": 1387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0744065750754905,
+      "learning_rate": 1.8014460909154058e-05,
+      "loss": 0.5068,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5517994510986008,
+      "learning_rate": 1.8011277924153022e-05,
+      "loss": 0.5338,
+      "step": 1389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.743740959230237,
+      "learning_rate": 1.8008092671576624e-05,
+      "loss": 0.6284,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.394179051742786,
+      "learning_rate": 1.8004905152326443e-05,
+      "loss": 0.6298,
+      "step": 1391
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.664303670015105,
+      "learning_rate": 1.8001715367304696e-05,
+      "loss": 0.5526,
+      "step": 1392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.580691123113156,
+      "learning_rate": 1.799852331741425e-05,
+      "loss": 0.5628,
+      "step": 1393
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4729978149633338,
+      "learning_rate": 1.799532900355861e-05,
+      "loss": 0.5746,
+      "step": 1394
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6296241968922562,
+      "learning_rate": 1.799213242664191e-05,
+      "loss": 0.5335,
+      "step": 1395
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9396555963673867,
+      "learning_rate": 1.7988933587568948e-05,
+      "loss": 0.5665,
+      "step": 1396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.512253024111374,
+      "learning_rate": 1.7985732487245132e-05,
+      "loss": 0.5865,
+      "step": 1397
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9360094889433204,
+      "learning_rate": 1.7982529126576543e-05,
+      "loss": 0.5966,
+      "step": 1398
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8228509551304328,
+      "learning_rate": 1.7979323506469878e-05,
+      "loss": 0.6189,
+      "step": 1399
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.92579396367746,
+      "learning_rate": 1.797611562783248e-05,
+      "loss": 0.5141,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4296567077551736,
+      "learning_rate": 1.7972905491572334e-05,
+      "loss": 0.5947,
+      "step": 1401
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3108532906484722,
+      "learning_rate": 1.7969693098598063e-05,
+      "loss": 0.6237,
+      "step": 1402
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1078597814473907,
+      "learning_rate": 1.7966478449818925e-05,
+      "loss": 0.7615,
+      "step": 1403
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8516749242771606,
+      "learning_rate": 1.7963261546144823e-05,
+      "loss": 0.6039,
+      "step": 1404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9738821239588953,
+      "learning_rate": 1.7960042388486293e-05,
+      "loss": 0.6434,
+      "step": 1405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1540057127834347,
+      "learning_rate": 1.795682097775451e-05,
+      "loss": 0.5164,
+      "step": 1406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7355703167737988,
+      "learning_rate": 1.795359731486129e-05,
+      "loss": 0.6495,
+      "step": 1407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6063768327140515,
+      "learning_rate": 1.7950371400719087e-05,
+      "loss": 0.5859,
+      "step": 1408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.704568003400319,
+      "learning_rate": 1.7947143236240983e-05,
+      "loss": 0.6021,
+      "step": 1409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5814186696877963,
+      "learning_rate": 1.7943912822340702e-05,
+      "loss": 0.5593,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5984575619644508,
+      "learning_rate": 1.7940680159932612e-05,
+      "loss": 0.634,
+      "step": 1411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6204782684006602,
+      "learning_rate": 1.7937445249931706e-05,
+      "loss": 0.6306,
+      "step": 1412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0662357578540136,
+      "learning_rate": 1.7934208093253625e-05,
+      "loss": 0.5598,
+      "step": 1413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5853798183647962,
+      "learning_rate": 1.7930968690814634e-05,
+      "loss": 0.5203,
+      "step": 1414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.736752973492914,
+      "learning_rate": 1.7927727043531637e-05,
+      "loss": 0.6523,
+      "step": 1415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7443758062041033,
+      "learning_rate": 1.7924483152322183e-05,
+      "loss": 0.6616,
+      "step": 1416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.882873723917877,
+      "learning_rate": 1.7921237018104443e-05,
+      "loss": 0.4866,
+      "step": 1417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5255790290189373,
+      "learning_rate": 1.7917988641797227e-05,
+      "loss": 0.6305,
+      "step": 1418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6609793571281894,
+      "learning_rate": 1.791473802431999e-05,
+      "loss": 0.4987,
+      "step": 1419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7221714210840162,
+      "learning_rate": 1.7911485166592802e-05,
+      "loss": 0.5538,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.862182131572846,
+      "learning_rate": 1.790823006953638e-05,
+      "loss": 0.6078,
+      "step": 1421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6471139040217986,
+      "learning_rate": 1.790497273407208e-05,
+      "loss": 0.6488,
+      "step": 1422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8714278565565399,
+      "learning_rate": 1.7901713161121873e-05,
+      "loss": 0.6215,
+      "step": 1423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.452342543158586,
+      "learning_rate": 1.7898451351608385e-05,
+      "loss": 0.679,
+      "step": 1424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.791297439859586,
+      "learning_rate": 1.7895187306454852e-05,
+      "loss": 0.6458,
+      "step": 1425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.635495544114762,
+      "learning_rate": 1.7891921026585167e-05,
+      "loss": 0.6206,
+      "step": 1426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7747427733918049,
+      "learning_rate": 1.7888652512923836e-05,
+      "loss": 0.5684,
+      "step": 1427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.66340250822401,
+      "learning_rate": 1.7885381766396008e-05,
+      "loss": 0.5159,
+      "step": 1428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6761753512035333,
+      "learning_rate": 1.788210878792746e-05,
+      "loss": 0.5626,
+      "step": 1429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5513340844386265,
+      "learning_rate": 1.7878833578444603e-05,
+      "loss": 0.476,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6246129562654357,
+      "learning_rate": 1.787555613887448e-05,
+      "loss": 0.6357,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.004328916422176,
+      "learning_rate": 1.7872276470144756e-05,
+      "loss": 0.6002,
+      "step": 1432
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9740593632840997,
+      "learning_rate": 1.786899457318374e-05,
+      "loss": 0.5731,
+      "step": 1433
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8076186864474895,
+      "learning_rate": 1.7865710448920365e-05,
+      "loss": 0.6544,
+      "step": 1434
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4789500079644418,
+      "learning_rate": 1.7862424098284197e-05,
+      "loss": 0.6167,
+      "step": 1435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5949357986673456,
+      "learning_rate": 1.7859135522205426e-05,
+      "loss": 0.5852,
+      "step": 1436
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6008658003430347,
+      "learning_rate": 1.7855844721614883e-05,
+      "loss": 0.604,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8110911256845879,
+      "learning_rate": 1.7852551697444017e-05,
+      "loss": 0.6025,
+      "step": 1438
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3348489111396145,
+      "learning_rate": 1.7849256450624914e-05,
+      "loss": 0.5571,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9929597730575458,
+      "learning_rate": 1.7845958982090287e-05,
+      "loss": 0.7483,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8978961942965176,
+      "learning_rate": 1.7842659292773474e-05,
+      "loss": 0.681,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6869800249828661,
+      "learning_rate": 1.783935738360845e-05,
+      "loss": 0.5409,
+      "step": 1442
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.812180468042523,
+      "learning_rate": 1.783605325552981e-05,
+      "loss": 0.6505,
+      "step": 1443
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3443016870364946,
+      "learning_rate": 1.7832746909472783e-05,
+      "loss": 0.5265,
+      "step": 1444
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6990694969656897,
+      "learning_rate": 1.782943834637322e-05,
+      "loss": 0.678,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5134777399880284,
+      "learning_rate": 1.7826127567167602e-05,
+      "loss": 0.4589,
+      "step": 1446
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7650516051807565,
+      "learning_rate": 1.7822814572793047e-05,
+      "loss": 0.6646,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0072665946629833,
+      "learning_rate": 1.7819499364187282e-05,
+      "loss": 0.5739,
+      "step": 1448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8147864904111761,
+      "learning_rate": 1.7816181942288672e-05,
+      "loss": 0.5496,
+      "step": 1449
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.66769397412382,
+      "learning_rate": 1.781286230803621e-05,
+      "loss": 0.6141,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5129944935457258,
+      "learning_rate": 1.7809540462369505e-05,
+      "loss": 0.5405,
+      "step": 1451
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8499325677201452,
+      "learning_rate": 1.7806216406228805e-05,
+      "loss": 0.5622,
+      "step": 1452
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7868175899983274,
+      "learning_rate": 1.780289014055497e-05,
+      "loss": 0.625,
+      "step": 1453
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.987190640624128,
+      "learning_rate": 1.77995616662895e-05,
+      "loss": 0.6678,
+      "step": 1454
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6493329252929299,
+      "learning_rate": 1.779623098437451e-05,
+      "loss": 0.5907,
+      "step": 1455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.695197766665084,
+      "learning_rate": 1.779289809575274e-05,
+      "loss": 0.5668,
+      "step": 1456
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0025024683432937,
+      "learning_rate": 1.7789563001367557e-05,
+      "loss": 0.6453,
+      "step": 1457
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4505146976229584,
+      "learning_rate": 1.7786225702162955e-05,
+      "loss": 0.5592,
+      "step": 1458
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.103314849083466,
+      "learning_rate": 1.778288619908355e-05,
+      "loss": 0.5595,
+      "step": 1459
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7005887972762133,
+      "learning_rate": 1.777954449307458e-05,
+      "loss": 0.4735,
+      "step": 1460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.106454144365622,
+      "learning_rate": 1.777620058508191e-05,
+      "loss": 0.6419,
+      "step": 1461
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6982631526545497,
+      "learning_rate": 1.7772854476052023e-05,
+      "loss": 0.6659,
+      "step": 1462
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.828978967651756,
+      "learning_rate": 1.7769506166932026e-05,
+      "loss": 0.6381,
+      "step": 1463
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7716564467836597,
+      "learning_rate": 1.7766155658669655e-05,
+      "loss": 0.6638,
+      "step": 1464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6476139691122402,
+      "learning_rate": 1.7762802952213264e-05,
+      "loss": 0.5837,
+      "step": 1465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5112756557711935,
+      "learning_rate": 1.7759448048511833e-05,
+      "loss": 0.5996,
+      "step": 1466
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.06521049549914,
+      "learning_rate": 1.775609094851495e-05,
+      "loss": 0.5455,
+      "step": 1467
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0479157696371804,
+      "learning_rate": 1.7752731653172847e-05,
+      "loss": 0.6251,
+      "step": 1468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6429613978595778,
+      "learning_rate": 1.7749370163436353e-05,
+      "loss": 0.5844,
+      "step": 1469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.075429705647753,
+      "learning_rate": 1.7746006480256943e-05,
+      "loss": 0.6258,
+      "step": 1470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5402929887443453,
+      "learning_rate": 1.7742640604586694e-05,
+      "loss": 0.4802,
+      "step": 1471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.093484003740606,
+      "learning_rate": 1.773927253737831e-05,
+      "loss": 0.6399,
+      "step": 1472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8325431084357657,
+      "learning_rate": 1.7735902279585118e-05,
+      "loss": 0.6571,
+      "step": 1473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.077222542029501,
+      "learning_rate": 1.7732529832161057e-05,
+      "loss": 0.7022,
+      "step": 1474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7917847934988276,
+      "learning_rate": 1.7729155196060697e-05,
+      "loss": 0.571,
+      "step": 1475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7605502129418067,
+      "learning_rate": 1.772577837223922e-05,
+      "loss": 0.5272,
+      "step": 1476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7586736840481019,
+      "learning_rate": 1.772239936165243e-05,
+      "loss": 0.5783,
+      "step": 1477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6692808499022862,
+      "learning_rate": 1.7719018165256745e-05,
+      "loss": 0.6266,
+      "step": 1478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8591609581242317,
+      "learning_rate": 1.7715634784009207e-05,
+      "loss": 0.6283,
+      "step": 1479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.752610388613009,
+      "learning_rate": 1.7712249218867476e-05,
+      "loss": 0.5537,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8138801820464667,
+      "learning_rate": 1.770886147078983e-05,
+      "loss": 0.6155,
+      "step": 1481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.95881430021073,
+      "learning_rate": 1.770547154073516e-05,
+      "loss": 0.5889,
+      "step": 1482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.603599177415643,
+      "learning_rate": 1.7702079429662986e-05,
+      "loss": 0.5429,
+      "step": 1483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.788501841321269,
+      "learning_rate": 1.769868513853343e-05,
+      "loss": 0.6775,
+      "step": 1484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.948761021336707,
+      "learning_rate": 1.769528866830724e-05,
+      "loss": 0.6571,
+      "step": 1485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3851562365879193,
+      "learning_rate": 1.7691890019945785e-05,
+      "loss": 0.4999,
+      "step": 1486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8928584719351742,
+      "learning_rate": 1.768848919441104e-05,
+      "loss": 0.6825,
+      "step": 1487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6507283372434334,
+      "learning_rate": 1.7685086192665605e-05,
+      "loss": 0.5629,
+      "step": 1488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2863720002775305,
+      "learning_rate": 1.7681681015672693e-05,
+      "loss": 0.4673,
+      "step": 1489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0055844243832555,
+      "learning_rate": 1.767827366439613e-05,
+      "loss": 0.5759,
+      "step": 1490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.659954894598363,
+      "learning_rate": 1.7674864139800356e-05,
+      "loss": 0.5535,
+      "step": 1491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9473428426875588,
+      "learning_rate": 1.7671452442850438e-05,
+      "loss": 0.7004,
+      "step": 1492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7256543622545366,
+      "learning_rate": 1.7668038574512045e-05,
+      "loss": 0.6322,
+      "step": 1493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5446255225949728,
+      "learning_rate": 1.7664622535751463e-05,
+      "loss": 0.4859,
+      "step": 1494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4914854142370502,
+      "learning_rate": 1.76612043275356e-05,
+      "loss": 0.5435,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.694123256245639,
+      "learning_rate": 1.7657783950831965e-05,
+      "loss": 0.6264,
+      "step": 1496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5903715276887511,
+      "learning_rate": 1.76543614066087e-05,
+      "loss": 0.5616,
+      "step": 1497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1489443265449832,
+      "learning_rate": 1.7650936695834536e-05,
+      "loss": 0.6984,
+      "step": 1498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.975990242576158,
+      "learning_rate": 1.764750981947884e-05,
+      "loss": 0.7701,
+      "step": 1499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7928090613408367,
+      "learning_rate": 1.764408077851158e-05,
+      "loss": 0.6232,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2831546918892978,
+      "learning_rate": 1.7640649573903334e-05,
+      "loss": 0.5575,
+      "step": 1501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0356430512405805,
+      "learning_rate": 1.7637216206625303e-05,
+      "loss": 0.5928,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.623775438824219,
+      "learning_rate": 1.763378067764929e-05,
+      "loss": 0.6336,
+      "step": 1503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.112263064436643,
+      "learning_rate": 1.7630342987947718e-05,
+      "loss": 0.5265,
+      "step": 1504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4687533262874206,
+      "learning_rate": 1.7626903138493614e-05,
+      "loss": 0.6013,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9074006114363584,
+      "learning_rate": 1.7623461130260625e-05,
+      "loss": 0.5421,
+      "step": 1506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6855344908813241,
+      "learning_rate": 1.7620016964223e-05,
+      "loss": 0.6285,
+      "step": 1507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.612606045979136,
+      "learning_rate": 1.7616570641355602e-05,
+      "loss": 0.6103,
+      "step": 1508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.811183374955782,
+      "learning_rate": 1.761312216263391e-05,
+      "loss": 0.5801,
+      "step": 1509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6380489752174836,
+      "learning_rate": 1.7609671529034006e-05,
+      "loss": 0.5996,
+      "step": 1510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6802709610546347,
+      "learning_rate": 1.7606218741532588e-05,
+      "loss": 0.6333,
+      "step": 1511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5492316877313264,
+      "learning_rate": 1.7602763801106952e-05,
+      "loss": 0.6341,
+      "step": 1512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8751968064731694,
+      "learning_rate": 1.759930670873502e-05,
+      "loss": 0.6345,
+      "step": 1513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0659517859541934,
+      "learning_rate": 1.759584746539531e-05,
+      "loss": 0.5994,
+      "step": 1514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6831538625027578,
+      "learning_rate": 1.7592386072066962e-05,
+      "loss": 0.5366,
+      "step": 1515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.545300099924969,
+      "learning_rate": 1.7588922529729703e-05,
+      "loss": 0.5468,
+      "step": 1516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.385488645213335,
+      "learning_rate": 1.7585456839363886e-05,
+      "loss": 0.5978,
+      "step": 1517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7185482353242412,
+      "learning_rate": 1.758198900195047e-05,
+      "loss": 0.5435,
+      "step": 1518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.535828132642818,
+      "learning_rate": 1.757851901847102e-05,
+      "loss": 0.4104,
+      "step": 1519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.315356712613748,
+      "learning_rate": 1.7575046889907708e-05,
+      "loss": 0.5876,
+      "step": 1520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.199217841139352,
+      "learning_rate": 1.7571572617243307e-05,
+      "loss": 0.5263,
+      "step": 1521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4782175775993134,
+      "learning_rate": 1.7568096201461204e-05,
+      "loss": 0.6267,
+      "step": 1522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5152280003488472,
+      "learning_rate": 1.7564617643545395e-05,
+      "loss": 0.6135,
+      "step": 1523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7429933174946635,
+      "learning_rate": 1.7561136944480478e-05,
+      "loss": 0.5437,
+      "step": 1524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.665175551970124,
+      "learning_rate": 1.7557654105251657e-05,
+      "loss": 0.5467,
+      "step": 1525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.682153916320299,
+      "learning_rate": 1.7554169126844736e-05,
+      "loss": 0.5951,
+      "step": 1526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9022218791317282,
+      "learning_rate": 1.755068201024614e-05,
+      "loss": 0.6782,
+      "step": 1527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.492631873514721,
+      "learning_rate": 1.7547192756442887e-05,
+      "loss": 0.5602,
+      "step": 1528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.814261777678974,
+      "learning_rate": 1.75437013664226e-05,
+      "loss": 0.5171,
+      "step": 1529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.760746054682778,
+      "learning_rate": 1.7540207841173513e-05,
+      "loss": 0.5757,
+      "step": 1530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6385046943292012,
+      "learning_rate": 1.753671218168446e-05,
+      "loss": 0.578,
+      "step": 1531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8416666223542895,
+      "learning_rate": 1.753321438894488e-05,
+      "loss": 0.6983,
+      "step": 1532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8231027491175296,
+      "learning_rate": 1.7529714463944815e-05,
+      "loss": 0.553,
+      "step": 1533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.281860427222861,
+      "learning_rate": 1.7526212407674916e-05,
+      "loss": 0.4853,
+      "step": 1534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1319823442264116,
+      "learning_rate": 1.7522708221126424e-05,
+      "loss": 0.6121,
+      "step": 1535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.500811232203332,
+      "learning_rate": 1.7519201905291195e-05,
+      "loss": 0.5193,
+      "step": 1536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7115857427545145,
+      "learning_rate": 1.7515693461161687e-05,
+      "loss": 0.6095,
+      "step": 1537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4865101882228546,
+      "learning_rate": 1.751218288973096e-05,
+      "loss": 0.5686,
+      "step": 1538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7108548347020047,
+      "learning_rate": 1.7508670191992667e-05,
+      "loss": 0.5398,
+      "step": 1539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.888853688565269,
+      "learning_rate": 1.7505155368941074e-05,
+      "loss": 0.5869,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9453656518013238,
+      "learning_rate": 1.7501638421571045e-05,
+      "loss": 0.5912,
+      "step": 1541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6351684457849576,
+      "learning_rate": 1.749811935087804e-05,
+      "loss": 0.6267,
+      "step": 1542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7375741904963116,
+      "learning_rate": 1.7494598157858127e-05,
+      "loss": 0.5486,
+      "step": 1543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.95278155866444,
+      "learning_rate": 1.7491074843507974e-05,
+      "loss": 0.6048,
+      "step": 1544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.497405045498587,
+      "learning_rate": 1.7487549408824845e-05,
+      "loss": 0.5284,
+      "step": 1545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5122410430249236,
+      "learning_rate": 1.748402185480661e-05,
+      "loss": 0.5208,
+      "step": 1546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0407037625440574,
+      "learning_rate": 1.7480492182451735e-05,
+      "loss": 0.5769,
+      "step": 1547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.744272023722274,
+      "learning_rate": 1.7476960392759284e-05,
+      "loss": 0.6227,
+      "step": 1548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.089311786187679,
+      "learning_rate": 1.7473426486728925e-05,
+      "loss": 0.4933,
+      "step": 1549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6227643182603484,
+      "learning_rate": 1.746989046536092e-05,
+      "loss": 0.5972,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6334630029934145,
+      "learning_rate": 1.7466352329656134e-05,
+      "loss": 0.6102,
+      "step": 1551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.401447564346569,
+      "learning_rate": 1.746281208061603e-05,
+      "loss": 0.5874,
+      "step": 1552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5428827679856107,
+      "learning_rate": 1.7459269719242665e-05,
+      "loss": 0.6863,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6936991729759492,
+      "learning_rate": 1.74557252465387e-05,
+      "loss": 0.5383,
+      "step": 1554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9174467843886849,
+      "learning_rate": 1.745217866350739e-05,
+      "loss": 0.4908,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5109800574458176,
+      "learning_rate": 1.744862997115259e-05,
+      "loss": 0.4951,
+      "step": 1556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5314231774000808,
+      "learning_rate": 1.7445079170478743e-05,
+      "loss": 0.5331,
+      "step": 1557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8267164654173922,
+      "learning_rate": 1.74415262624909e-05,
+      "loss": 0.6396,
+      "step": 1558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.41055929100685,
+      "learning_rate": 1.7437971248194706e-05,
+      "loss": 0.5143,
+      "step": 1559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.509050029154316,
+      "learning_rate": 1.74344141285964e-05,
+      "loss": 0.5881,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.520592393249489,
+      "learning_rate": 1.7430854904702806e-05,
+      "loss": 0.6655,
+      "step": 1561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5797868653413925,
+      "learning_rate": 1.7427293577521377e-05,
+      "loss": 0.4891,
+      "step": 1562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6713496671772459,
+      "learning_rate": 1.742373014806012e-05,
+      "loss": 0.4641,
+      "step": 1563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8154248006590745,
+      "learning_rate": 1.7420164617327662e-05,
+      "loss": 0.5672,
+      "step": 1564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5411591677707044,
+      "learning_rate": 1.7416596986333222e-05,
+      "loss": 0.5113,
+      "step": 1565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7048377307810259,
+      "learning_rate": 1.7413027256086606e-05,
+      "loss": 0.5911,
+      "step": 1566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5819962573253636,
+      "learning_rate": 1.7409455427598224e-05,
+      "loss": 0.5742,
+      "step": 1567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8508144291749267,
+      "learning_rate": 1.740588150187907e-05,
+      "loss": 0.6406,
+      "step": 1568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8482264352106017,
+      "learning_rate": 1.7402305479940735e-05,
+      "loss": 0.6438,
+      "step": 1569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8252419500925907,
+      "learning_rate": 1.739872736279541e-05,
+      "loss": 0.6638,
+      "step": 1570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9042871038933304,
+      "learning_rate": 1.7395147151455868e-05,
+      "loss": 0.6101,
+      "step": 1571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.454995196496883,
+      "learning_rate": 1.7391564846935484e-05,
+      "loss": 0.5562,
+      "step": 1572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.349670388673249,
+      "learning_rate": 1.7387980450248222e-05,
+      "loss": 0.6793,
+      "step": 1573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.171494238172583,
+      "learning_rate": 1.7384393962408633e-05,
+      "loss": 0.5935,
+      "step": 1574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6128531609679186,
+      "learning_rate": 1.7380805384431875e-05,
+      "loss": 0.5722,
+      "step": 1575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.520168924920876,
+      "learning_rate": 1.7377214717333675e-05,
+      "loss": 0.5429,
+      "step": 1576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7131220863185805,
+      "learning_rate": 1.7373621962130373e-05,
+      "loss": 0.6932,
+      "step": 1577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.609778529112429,
+      "learning_rate": 1.7370027119838884e-05,
+      "loss": 0.5884,
+      "step": 1578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4282729592320522,
+      "learning_rate": 1.7366430191476723e-05,
+      "loss": 0.551,
+      "step": 1579
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.397588037646445,
+      "learning_rate": 1.7362831178062e-05,
+      "loss": 0.6621,
+      "step": 1580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5881456268790852,
+      "learning_rate": 1.73592300806134e-05,
+      "loss": 0.5448,
+      "step": 1581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9589341536239254,
+      "learning_rate": 1.7355626900150202e-05,
+      "loss": 0.7689,
+      "step": 1582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7284488168600827,
+      "learning_rate": 1.735202163769229e-05,
+      "loss": 0.6314,
+      "step": 1583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1021162915592453,
+      "learning_rate": 1.734841429426012e-05,
+      "loss": 0.6516,
+      "step": 1584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8851285345817537,
+      "learning_rate": 1.7344804870874744e-05,
+      "loss": 0.6718,
+      "step": 1585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7437997489295767,
+      "learning_rate": 1.73411933685578e-05,
+      "loss": 0.6082,
+      "step": 1586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5097763311987626,
+      "learning_rate": 1.733757978833152e-05,
+      "loss": 0.5886,
+      "step": 1587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4340487459270097,
+      "learning_rate": 1.7333964131218714e-05,
+      "loss": 0.5628,
+      "step": 1588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1218923251654496,
+      "learning_rate": 1.7330346398242794e-05,
+      "loss": 0.5356,
+      "step": 1589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.551967562459985,
+      "learning_rate": 1.7326726590427747e-05,
+      "loss": 0.5669,
+      "step": 1590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.721768136378051,
+      "learning_rate": 1.732310470879815e-05,
+      "loss": 0.5017,
+      "step": 1591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.500991464905295,
+      "learning_rate": 1.7319480754379175e-05,
+      "loss": 0.6746,
+      "step": 1592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4065742231990443,
+      "learning_rate": 1.7315854728196568e-05,
+      "loss": 0.5395,
+      "step": 1593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.971169895577976,
+      "learning_rate": 1.7312226631276675e-05,
+      "loss": 0.6117,
+      "step": 1594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7972545146275887,
+      "learning_rate": 1.7308596464646413e-05,
+      "loss": 0.6253,
+      "step": 1595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5644968661312688,
+      "learning_rate": 1.7304964229333302e-05,
+      "loss": 0.4722,
+      "step": 1596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7059397770488656,
+      "learning_rate": 1.7301329926365432e-05,
+      "loss": 0.5374,
+      "step": 1597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8863588221130354,
+      "learning_rate": 1.729769355677149e-05,
+      "loss": 0.5676,
+      "step": 1598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7947481177539129,
+      "learning_rate": 1.7294055121580735e-05,
+      "loss": 0.633,
+      "step": 1599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8855244209152113,
+      "learning_rate": 1.7290414621823026e-05,
+      "loss": 0.6163,
+      "step": 1600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4723897868153786,
+      "learning_rate": 1.7286772058528796e-05,
+      "loss": 0.6125,
+      "step": 1601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7808067080890366,
+      "learning_rate": 1.7283127432729063e-05,
+      "loss": 0.5928,
+      "step": 1602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5794625639688085,
+      "learning_rate": 1.7279480745455433e-05,
+      "loss": 0.5965,
+      "step": 1603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.656626662305367,
+      "learning_rate": 1.7275831997740095e-05,
+      "loss": 0.4963,
+      "step": 1604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7431956939474584,
+      "learning_rate": 1.7272181190615812e-05,
+      "loss": 0.5864,
+      "step": 1605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7178820047380559,
+      "learning_rate": 1.7268528325115947e-05,
+      "loss": 0.5962,
+      "step": 1606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.785212942689208,
+      "learning_rate": 1.726487340227443e-05,
+      "loss": 0.5557,
+      "step": 1607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5303574900031987,
+      "learning_rate": 1.7261216423125782e-05,
+      "loss": 0.5813,
+      "step": 1608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8781397715090062,
+      "learning_rate": 1.7257557388705098e-05,
+      "loss": 0.5969,
+      "step": 1609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5421102140085994,
+      "learning_rate": 1.725389630004807e-05,
+      "loss": 0.5821,
+      "step": 1610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5778504047247934,
+      "learning_rate": 1.7250233158190948e-05,
+      "loss": 0.5549,
+      "step": 1611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.52575872432632,
+      "learning_rate": 1.7246567964170585e-05,
+      "loss": 0.5779,
+      "step": 1612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6392552577036152,
+      "learning_rate": 1.724290071902441e-05,
+      "loss": 0.6314,
+      "step": 1613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8084585985189834,
+      "learning_rate": 1.723923142379042e-05,
+      "loss": 0.6143,
+      "step": 1614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.846456574791803,
+      "learning_rate": 1.7235560079507204e-05,
+      "loss": 0.6387,
+      "step": 1615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.366423657203043,
+      "learning_rate": 1.7231886687213936e-05,
+      "loss": 0.4575,
+      "step": 1616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.320727738774158,
+      "learning_rate": 1.722821124795035e-05,
+      "loss": 0.5398,
+      "step": 1617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7374407910225576,
+      "learning_rate": 1.7224533762756775e-05,
+      "loss": 0.569,
+      "step": 1618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.750022891866523,
+      "learning_rate": 1.7220854232674127e-05,
+      "loss": 0.6604,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.785903799435068,
+      "learning_rate": 1.721717265874387e-05,
+      "loss": 0.5575,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.567061073652123,
+      "learning_rate": 1.721348904200808e-05,
+      "loss": 0.5989,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3681979570166645,
+      "learning_rate": 1.7209803383509394e-05,
+      "loss": 0.5442,
+      "step": 1622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5503966144000811,
+      "learning_rate": 1.720611568429103e-05,
+      "loss": 0.6813,
+      "step": 1623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4076302037796646,
+      "learning_rate": 1.7202425945396774e-05,
+      "loss": 0.5346,
+      "step": 1624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.881861278285824,
+      "learning_rate": 1.7198734167871015e-05,
+      "loss": 0.5678,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7280595787050586,
+      "learning_rate": 1.719504035275869e-05,
+      "loss": 0.5985,
+      "step": 1626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9536386294151566,
+      "learning_rate": 1.7191344501105328e-05,
+      "loss": 0.6139,
+      "step": 1627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6320884133110671,
+      "learning_rate": 1.718764661395704e-05,
+      "loss": 0.6023,
+      "step": 1628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5305795652668062,
+      "learning_rate": 1.7183946692360495e-05,
+      "loss": 0.5418,
+      "step": 1629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7620783615619744,
+      "learning_rate": 1.7180244737362956e-05,
+      "loss": 0.5552,
+      "step": 1630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.693477444936704,
+      "learning_rate": 1.717654075001225e-05,
+      "loss": 0.6068,
+      "step": 1631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.641959191325776,
+      "learning_rate": 1.717283473135678e-05,
+      "loss": 0.5142,
+      "step": 1632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4171401225814295,
+      "learning_rate": 1.716912668244553e-05,
+      "loss": 0.6212,
+      "step": 1633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2890074958641435,
+      "learning_rate": 1.7165416604328054e-05,
+      "loss": 0.7494,
+      "step": 1634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9662978549279084,
+      "learning_rate": 1.7161704498054485e-05,
+      "loss": 0.5725,
+      "step": 1635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.930761878641957,
+      "learning_rate": 1.7157990364675524e-05,
+      "loss": 0.5668,
+      "step": 1636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.931782132921812,
+      "learning_rate": 1.7154274205242448e-05,
+      "loss": 0.6259,
+      "step": 1637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6853847784299971,
+      "learning_rate": 1.715055602080711e-05,
+      "loss": 0.5777,
+      "step": 1638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9099806159211272,
+      "learning_rate": 1.7146835812421937e-05,
+      "loss": 0.6667,
+      "step": 1639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.188027563767982,
+      "learning_rate": 1.714311358113992e-05,
+      "loss": 0.4764,
+      "step": 1640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.807691390792332,
+      "learning_rate": 1.7139389328014634e-05,
+      "loss": 0.6189,
+      "step": 1641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6059902029137525,
+      "learning_rate": 1.7135663054100216e-05,
+      "loss": 0.6087,
+      "step": 1642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6641427666015625,
+      "learning_rate": 1.7131934760451385e-05,
+      "loss": 0.5647,
+      "step": 1643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5456880134568864,
+      "learning_rate": 1.7128204448123422e-05,
+      "loss": 0.6161,
+      "step": 1644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5893356506846827,
+      "learning_rate": 1.7124472118172187e-05,
+      "loss": 0.5785,
+      "step": 1645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.616274761200397,
+      "learning_rate": 1.712073777165411e-05,
+      "loss": 0.58,
+      "step": 1646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6979885893984399,
+      "learning_rate": 1.7117001409626185e-05,
+      "loss": 0.61,
+      "step": 1647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4482423930150583,
+      "learning_rate": 1.7113263033145985e-05,
+      "loss": 0.5744,
+      "step": 1648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8152577157038425,
+      "learning_rate": 1.7109522643271646e-05,
+      "loss": 0.5588,
+      "step": 1649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.934034837913026,
+      "learning_rate": 1.7105780241061884e-05,
+      "loss": 0.6293,
+      "step": 1650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.593698907841797,
+      "learning_rate": 1.710203582757597e-05,
+      "loss": 0.5992,
+      "step": 1651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7603689590296485,
+      "learning_rate": 1.7098289403873754e-05,
+      "loss": 0.69,
+      "step": 1652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8616008749301756,
+      "learning_rate": 1.7094540971015663e-05,
+      "loss": 0.5677,
+      "step": 1653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9746553335231278,
+      "learning_rate": 1.709079053006267e-05,
+      "loss": 0.6578,
+      "step": 1654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9580217454155562,
+      "learning_rate": 1.708703808207634e-05,
+      "loss": 0.7596,
+      "step": 1655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8202875746498255,
+      "learning_rate": 1.7083283628118786e-05,
+      "loss": 0.7275,
+      "step": 1656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1466235908293427,
+      "learning_rate": 1.7079527169252706e-05,
+      "loss": 0.5226,
+      "step": 1657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7627165563557425,
+      "learning_rate": 1.7075768706541355e-05,
+      "loss": 0.5432,
+      "step": 1658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.220821648198032,
+      "learning_rate": 1.7072008241048555e-05,
+      "loss": 0.587,
+      "step": 1659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8565421260748418,
+      "learning_rate": 1.7068245773838703e-05,
+      "loss": 0.6168,
+      "step": 1660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5866691048631851,
+      "learning_rate": 1.7064481305976754e-05,
+      "loss": 0.6333,
+      "step": 1661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.647748149669201,
+      "learning_rate": 1.7060714838528234e-05,
+      "loss": 0.5893,
+      "step": 1662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4743622191433263,
+      "learning_rate": 1.7056946372559234e-05,
+      "loss": 0.6105,
+      "step": 1663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6183698242674305,
+      "learning_rate": 1.7053175909136406e-05,
+      "loss": 0.5662,
+      "step": 1664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9512502280717507,
+      "learning_rate": 1.7049403449326982e-05,
+      "loss": 0.5954,
+      "step": 1665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.540462047929474,
+      "learning_rate": 1.704562899419874e-05,
+      "loss": 0.5468,
+      "step": 1666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.145491376915714,
+      "learning_rate": 1.704185254482003e-05,
+      "loss": 0.6385,
+      "step": 1667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.191139039607457,
+      "learning_rate": 1.7038074102259775e-05,
+      "loss": 0.6034,
+      "step": 1668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5540657384257601,
+      "learning_rate": 1.703429366758745e-05,
+      "loss": 0.5735,
+      "step": 1669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7841752420172432,
+      "learning_rate": 1.7030511241873107e-05,
+      "loss": 0.5812,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7385534285260278,
+      "learning_rate": 1.7026726826187343e-05,
+      "loss": 0.6774,
+      "step": 1671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5353182068356614,
+      "learning_rate": 1.7022940421601334e-05,
+      "loss": 0.5439,
+      "step": 1672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7219233489235108,
+      "learning_rate": 1.7019152029186817e-05,
+      "loss": 0.5211,
+      "step": 1673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.399818160413672,
+      "learning_rate": 1.701536165001608e-05,
+      "loss": 0.5377,
+      "step": 1674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.566019922087614,
+      "learning_rate": 1.701156928516199e-05,
+      "loss": 0.4758,
+      "step": 1675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8360261872255683,
+      "learning_rate": 1.7007774935697966e-05,
+      "loss": 0.6309,
+      "step": 1676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5736446439578913,
+      "learning_rate": 1.7003978602697988e-05,
+      "loss": 0.5453,
+      "step": 1677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5162857075663483,
+      "learning_rate": 1.70001802872366e-05,
+      "loss": 0.6104,
+      "step": 1678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.824396389580153,
+      "learning_rate": 1.6996379990388908e-05,
+      "loss": 0.5632,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8009438640238418,
+      "learning_rate": 1.6992577713230582e-05,
+      "loss": 0.665,
+      "step": 1680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.787746950301064,
+      "learning_rate": 1.6988773456837847e-05,
+      "loss": 0.5985,
+      "step": 1681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7895865344883215,
+      "learning_rate": 1.6984967222287484e-05,
+      "loss": 0.6674,
+      "step": 1682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6986134858891297,
+      "learning_rate": 1.6981159010656847e-05,
+      "loss": 0.6192,
+      "step": 1683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6214397562835465,
+      "learning_rate": 1.6977348823023838e-05,
+      "loss": 0.4762,
+      "step": 1684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.800617946852894,
+      "learning_rate": 1.697353666046692e-05,
+      "loss": 0.5565,
+      "step": 1685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8587497946736469,
+      "learning_rate": 1.6969722524065124e-05,
+      "loss": 0.5382,
+      "step": 1686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6327882800340008,
+      "learning_rate": 1.696590641489803e-05,
+      "loss": 0.5367,
+      "step": 1687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.199364758439497,
+      "learning_rate": 1.6962088334045785e-05,
+      "loss": 0.5823,
+      "step": 1688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4223515010941632,
+      "learning_rate": 1.695826828258908e-05,
+      "loss": 0.4664,
+      "step": 1689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.628503423600149,
+      "learning_rate": 1.6954446261609176e-05,
+      "loss": 0.563,
+      "step": 1690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.078574941242636,
+      "learning_rate": 1.6950622272187888e-05,
+      "loss": 0.6604,
+      "step": 1691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4005428222171183,
+      "learning_rate": 1.6946796315407593e-05,
+      "loss": 0.5106,
+      "step": 1692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4170845541778903,
+      "learning_rate": 1.694296839235121e-05,
+      "loss": 0.5837,
+      "step": 1693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5748667900489977,
+      "learning_rate": 1.6939138504102235e-05,
+      "loss": 0.6041,
+      "step": 1694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.866412318564003,
+      "learning_rate": 1.6935306651744704e-05,
+      "loss": 0.5062,
+      "step": 1695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7084444131151377,
+      "learning_rate": 1.6931472836363214e-05,
+      "loss": 0.5674,
+      "step": 1696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9499698789842288,
+      "learning_rate": 1.6927637059042918e-05,
+      "loss": 0.597,
+      "step": 1697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3575283910085636,
+      "learning_rate": 1.692379932086953e-05,
+      "loss": 0.7863,
+      "step": 1698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.422075832635088,
+      "learning_rate": 1.6919959622929312e-05,
+      "loss": 0.5932,
+      "step": 1699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.885140588175973,
+      "learning_rate": 1.691611796630908e-05,
+      "loss": 0.5883,
+      "step": 1700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4816612310046724,
+      "learning_rate": 1.6912274352096207e-05,
+      "loss": 0.4875,
+      "step": 1701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8891011854698574,
+      "learning_rate": 1.690842878137862e-05,
+      "loss": 0.4347,
+      "step": 1702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.8048701191151215,
+      "learning_rate": 1.6904581255244802e-05,
+      "loss": 0.661,
+      "step": 1703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9934126657246285,
+      "learning_rate": 1.690073177478379e-05,
+      "loss": 0.5615,
+      "step": 1704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.952658063898271,
+      "learning_rate": 1.6896880341085158e-05,
+      "loss": 0.5281,
+      "step": 1705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8809201838770074,
+      "learning_rate": 1.6893026955239062e-05,
+      "loss": 0.7095,
+      "step": 1706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.02439580100413,
+      "learning_rate": 1.688917161833618e-05,
+      "loss": 0.5772,
+      "step": 1707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6711392222187644,
+      "learning_rate": 1.688531433146777e-05,
+      "loss": 0.6178,
+      "step": 1708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5494811747235786,
+      "learning_rate": 1.6881455095725627e-05,
+      "loss": 0.6392,
+      "step": 1709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7574610168960088,
+      "learning_rate": 1.6877593912202094e-05,
+      "loss": 0.6751,
+      "step": 1710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3071661237440355,
+      "learning_rate": 1.6873730781990073e-05,
+      "loss": 0.5706,
+      "step": 1711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7847046945019054,
+      "learning_rate": 1.6869865706183017e-05,
+      "loss": 0.5238,
+      "step": 1712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5722430006583543,
+      "learning_rate": 1.6865998685874923e-05,
+      "loss": 0.5561,
+      "step": 1713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0146101896677755,
+      "learning_rate": 1.6862129722160347e-05,
+      "loss": 0.5599,
+      "step": 1714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8917256703008103,
+      "learning_rate": 1.685825881613439e-05,
+      "loss": 0.6146,
+      "step": 1715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0381194617893659,
+      "learning_rate": 1.6854385968892702e-05,
+      "loss": 0.7808,
+      "step": 1716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2126468979528435,
+      "learning_rate": 1.6850511181531487e-05,
+      "loss": 0.6033,
+      "step": 1717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7736058033040891,
+      "learning_rate": 1.6846634455147498e-05,
+      "loss": 0.6597,
+      "step": 1718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0599327617315772,
+      "learning_rate": 1.6842755790838025e-05,
+      "loss": 0.6025,
+      "step": 1719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7374055266438744,
+      "learning_rate": 1.6838875189700924e-05,
+      "loss": 0.5379,
+      "step": 1720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2408787222432984,
+      "learning_rate": 1.6834992652834586e-05,
+      "loss": 0.59,
+      "step": 1721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.690844297214501,
+      "learning_rate": 1.6831108181337957e-05,
+      "loss": 0.5576,
+      "step": 1722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.209250775690139,
+      "learning_rate": 1.6827221776310532e-05,
+      "loss": 0.556,
+      "step": 1723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.974745038766743,
+      "learning_rate": 1.6823333438852346e-05,
+      "loss": 0.5957,
+      "step": 1724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6277068391496945,
+      "learning_rate": 1.6819443170063983e-05,
+      "loss": 0.5823,
+      "step": 1725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9103519327869254,
+      "learning_rate": 1.681555097104658e-05,
+      "loss": 0.6345,
+      "step": 1726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5145732837109966,
+      "learning_rate": 1.681165684290181e-05,
+      "loss": 0.558,
+      "step": 1727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6352998709311264,
+      "learning_rate": 1.6807760786731905e-05,
+      "loss": 0.6349,
+      "step": 1728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6278812158169913,
+      "learning_rate": 1.680386280363963e-05,
+      "loss": 0.6098,
+      "step": 1729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3049682311304478,
+      "learning_rate": 1.67999628947283e-05,
+      "loss": 0.5001,
+      "step": 1730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4882575995965315,
+      "learning_rate": 1.6796061061101782e-05,
+      "loss": 0.5539,
+      "step": 1731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9463043497494144,
+      "learning_rate": 1.6792157303864475e-05,
+      "loss": 0.5886,
+      "step": 1732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9974010838819132,
+      "learning_rate": 1.6788251624121335e-05,
+      "loss": 0.5819,
+      "step": 1733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3651081903659357,
+      "learning_rate": 1.678434402297785e-05,
+      "loss": 0.6476,
+      "step": 1734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0204512970872264,
+      "learning_rate": 1.678043450154007e-05,
+      "loss": 0.7507,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8364709137060462,
+      "learning_rate": 1.6776523060914565e-05,
+      "loss": 0.7089,
+      "step": 1736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4756568135478763,
+      "learning_rate": 1.677260970220846e-05,
+      "loss": 0.4476,
+      "step": 1737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.808647599989326,
+      "learning_rate": 1.6768694426529432e-05,
+      "loss": 0.5429,
+      "step": 1738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2931478190167285,
+      "learning_rate": 1.676477723498569e-05,
+      "loss": 0.5247,
+      "step": 1739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.222791477489505,
+      "learning_rate": 1.6760858128685974e-05,
+      "loss": 0.5229,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.718310597173031,
+      "learning_rate": 1.6756937108739596e-05,
+      "loss": 0.5293,
+      "step": 1741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.971058237757801,
+      "learning_rate": 1.675301417625638e-05,
+      "loss": 0.4998,
+      "step": 1742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.699470489653138,
+      "learning_rate": 1.6749089332346714e-05,
+      "loss": 0.5831,
+      "step": 1743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.407425084811213,
+      "learning_rate": 1.6745162578121504e-05,
+      "loss": 0.5637,
+      "step": 1744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.268205817539003,
+      "learning_rate": 1.6741233914692223e-05,
+      "loss": 0.6268,
+      "step": 1745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8742642959528228,
+      "learning_rate": 1.6737303343170863e-05,
+      "loss": 0.5726,
+      "step": 1746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4331753140438748,
+      "learning_rate": 1.6733370864669965e-05,
+      "loss": 0.5608,
+      "step": 1747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.760357780524026,
+      "learning_rate": 1.672943648030261e-05,
+      "loss": 0.6154,
+      "step": 1748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8894264495613733,
+      "learning_rate": 1.6725500191182415e-05,
+      "loss": 0.5785,
+      "step": 1749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6807077472873437,
+      "learning_rate": 1.672156199842354e-05,
+      "loss": 0.6138,
+      "step": 1750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8251512870721747,
+      "learning_rate": 1.6717621903140686e-05,
+      "loss": 0.5921,
+      "step": 1751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.956008279840701,
+      "learning_rate": 1.6713679906449084e-05,
+      "loss": 0.7154,
+      "step": 1752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.958840378816245,
+      "learning_rate": 1.6709736009464504e-05,
+      "loss": 0.5674,
+      "step": 1753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7224492095681432,
+      "learning_rate": 1.670579021330327e-05,
+      "loss": 0.521,
+      "step": 1754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9056294749221332,
+      "learning_rate": 1.670184251908222e-05,
+      "loss": 0.6208,
+      "step": 1755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2280441906903294,
+      "learning_rate": 1.6697892927918742e-05,
+      "loss": 0.6078,
+      "step": 1756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.986980765876602,
+      "learning_rate": 1.6693941440930768e-05,
+      "loss": 0.5488,
+      "step": 1757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9104141831911015,
+      "learning_rate": 1.668998805923675e-05,
+      "loss": 0.5765,
+      "step": 1758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0451957482060408,
+      "learning_rate": 1.668603278395568e-05,
+      "loss": 0.5837,
+      "step": 1759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.883796564567973,
+      "learning_rate": 1.6682075616207103e-05,
+      "loss": 0.4488,
+      "step": 1760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9560877905003975,
+      "learning_rate": 1.667811655711108e-05,
+      "loss": 0.4987,
+      "step": 1761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5805245213963952,
+      "learning_rate": 1.6674155607788214e-05,
+      "loss": 0.5436,
+      "step": 1762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7873515969389406,
+      "learning_rate": 1.6670192769359643e-05,
+      "loss": 0.5713,
+      "step": 1763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.523686297622767,
+      "learning_rate": 1.666622804294704e-05,
+      "loss": 0.5288,
+      "step": 1764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9030195774052296,
+      "learning_rate": 1.666226142967262e-05,
+      "loss": 0.6689,
+      "step": 1765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5601634730030498,
+      "learning_rate": 1.6658292930659115e-05,
+      "loss": 0.613,
+      "step": 1766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3250430773836284,
+      "learning_rate": 1.6654322547029803e-05,
+      "loss": 0.5555,
+      "step": 1767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1827701017328374,
+      "learning_rate": 1.6650350279908497e-05,
+      "loss": 0.5858,
+      "step": 1768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0437640584852677,
+      "learning_rate": 1.664637613041953e-05,
+      "loss": 0.5795,
+      "step": 1769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5162455708412292,
+      "learning_rate": 1.6642400099687787e-05,
+      "loss": 0.5382,
+      "step": 1770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.875776404331716,
+      "learning_rate": 1.6638422188838667e-05,
+      "loss": 0.6152,
+      "step": 1771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.576811803971752,
+      "learning_rate": 1.6634442398998115e-05,
+      "loss": 0.6296,
+      "step": 1772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0601649302030154,
+      "learning_rate": 1.6630460731292597e-05,
+      "loss": 0.6647,
+      "step": 1773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8693262199803287,
+      "learning_rate": 1.662647718684912e-05,
+      "loss": 0.7194,
+      "step": 1774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5079443150338812,
+      "learning_rate": 1.6622491766795215e-05,
+      "loss": 0.6084,
+      "step": 1775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6060537077802777,
+      "learning_rate": 1.661850447225895e-05,
+      "loss": 0.6934,
+      "step": 1776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3895134078985008,
+      "learning_rate": 1.6614515304368915e-05,
+      "loss": 0.6111,
+      "step": 1777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6381180522713956,
+      "learning_rate": 1.661052426425424e-05,
+      "loss": 0.5692,
+      "step": 1778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8180923594741786,
+      "learning_rate": 1.6606531353044585e-05,
+      "loss": 0.7504,
+      "step": 1779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8451154841164523,
+      "learning_rate": 1.660253657187012e-05,
+      "loss": 0.7125,
+      "step": 1780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3134459225375588,
+      "learning_rate": 1.6598539921861573e-05,
+      "loss": 0.5105,
+      "step": 1781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.775024105969445,
+      "learning_rate": 1.6594541404150187e-05,
+      "loss": 0.6133,
+      "step": 1782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4296248133688252,
+      "learning_rate": 1.6590541019867722e-05,
+      "loss": 0.4955,
+      "step": 1783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4339270553512375,
+      "learning_rate": 1.6586538770146495e-05,
+      "loss": 0.518,
+      "step": 1784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5924180520747544,
+      "learning_rate": 1.658253465611932e-05,
+      "loss": 0.5476,
+      "step": 1785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8168139340365097,
+      "learning_rate": 1.6578528678919564e-05,
+      "loss": 0.5549,
+      "step": 1786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8743753767634869,
+      "learning_rate": 1.65745208396811e-05,
+      "loss": 0.6888,
+      "step": 1787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8097404033651732,
+      "learning_rate": 1.6570511139538348e-05,
+      "loss": 0.5541,
+      "step": 1788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6014871644424995,
+      "learning_rate": 1.6566499579626237e-05,
+      "loss": 0.621,
+      "step": 1789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0977908251518826,
+      "learning_rate": 1.656248616108024e-05,
+      "loss": 0.574,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7261565734754525,
+      "learning_rate": 1.655847088503634e-05,
+      "loss": 0.7267,
+      "step": 1791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4748206084989661,
+      "learning_rate": 1.655445375263105e-05,
+      "loss": 0.619,
+      "step": 1792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6244719641940673,
+      "learning_rate": 1.655043476500142e-05,
+      "loss": 0.5976,
+      "step": 1793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7394104631823142,
+      "learning_rate": 1.6546413923285008e-05,
+      "loss": 0.6256,
+      "step": 1794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.891080986125921,
+      "learning_rate": 1.6542391228619906e-05,
+      "loss": 0.5527,
+      "step": 1795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8346144366547559,
+      "learning_rate": 1.6538366682144734e-05,
+      "loss": 0.7264,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9433074614893262,
+      "learning_rate": 1.6534340284998626e-05,
+      "loss": 0.6524,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4761919779293204,
+      "learning_rate": 1.6530312038321247e-05,
+      "loss": 0.5628,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5375602433982254,
+      "learning_rate": 1.6526281943252782e-05,
+      "loss": 0.5735,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.824233707231009,
+      "learning_rate": 1.6522250000933948e-05,
+      "loss": 0.6149,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.75971326776135,
+      "learning_rate": 1.6518216212505968e-05,
+      "loss": 0.5902,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.485887642708432,
+      "learning_rate": 1.6514180579110606e-05,
+      "loss": 0.511,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4223432005719914,
+      "learning_rate": 1.6510143101890136e-05,
+      "loss": 0.5872,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.796179275603924,
+      "learning_rate": 1.6506103781987355e-05,
+      "loss": 0.5813,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6778964554454854,
+      "learning_rate": 1.650206262054559e-05,
+      "loss": 0.5327,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1775995655871005,
+      "learning_rate": 1.6498019618708673e-05,
+      "loss": 0.6159,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9842781849450826,
+      "learning_rate": 1.6493974777620976e-05,
+      "loss": 0.6454,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8189789390952502,
+      "learning_rate": 1.6489928098427383e-05,
+      "loss": 0.6634,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7838863999224612,
+      "learning_rate": 1.648587958227329e-05,
+      "loss": 0.5539,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4881815997966652,
+      "learning_rate": 1.648182923030463e-05,
+      "loss": 0.5502,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9982050596128362,
+      "learning_rate": 1.6477777043667846e-05,
+      "loss": 0.5784,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.714635705711105,
+      "learning_rate": 1.647372302350989e-05,
+      "loss": 0.5353,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.480895481122811,
+      "learning_rate": 1.6469667170978258e-05,
+      "loss": 0.5626,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8917810478553163,
+      "learning_rate": 1.6465609487220942e-05,
+      "loss": 0.6579,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4861650782460307,
+      "learning_rate": 1.6461549973386464e-05,
+      "loss": 0.5824,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6168122222130117,
+      "learning_rate": 1.645748863062386e-05,
+      "loss": 0.4995,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0433832558657183,
+      "learning_rate": 1.6453425460082685e-05,
+      "loss": 0.6004,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2934480952457443,
+      "learning_rate": 1.6449360462913005e-05,
+      "loss": 0.538,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1033785506464318,
+      "learning_rate": 1.644529364026542e-05,
+      "loss": 0.6711,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4492932193537034,
+      "learning_rate": 1.644122499329103e-05,
+      "loss": 0.5474,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9857766923401432,
+      "learning_rate": 1.6437154523141453e-05,
+      "loss": 0.7679,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.622898622549106,
+      "learning_rate": 1.6433082230968833e-05,
+      "loss": 0.6391,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7904123055996795,
+      "learning_rate": 1.642900811792582e-05,
+      "loss": 0.6378,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6753702253293856,
+      "learning_rate": 1.6424932185165587e-05,
+      "loss": 0.6016,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8518005537799287,
+      "learning_rate": 1.6420854433841817e-05,
+      "loss": 0.508,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8028173828410232,
+      "learning_rate": 1.6416774865108706e-05,
+      "loss": 0.7245,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8934034190008409,
+      "learning_rate": 1.641269348012097e-05,
+      "loss": 0.583,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7569739113116807,
+      "learning_rate": 1.640861028003383e-05,
+      "loss": 0.5574,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4680357907361623,
+      "learning_rate": 1.6404525266003037e-05,
+      "loss": 0.4514,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2274139065072096,
+      "learning_rate": 1.6400438439184842e-05,
+      "loss": 0.5817,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9759003740499013,
+      "learning_rate": 1.6396349800736012e-05,
+      "loss": 0.6518,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5881934023893434,
+      "learning_rate": 1.6392259351813827e-05,
+      "loss": 0.5874,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.071273365773276,
+      "learning_rate": 1.6388167093576083e-05,
+      "loss": 0.6032,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.085038022577111,
+      "learning_rate": 1.638407302718108e-05,
+      "loss": 0.4957,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1443137130224152,
+      "learning_rate": 1.6379977153787637e-05,
+      "loss": 0.5286,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8254222606476893,
+      "learning_rate": 1.6375879474555084e-05,
+      "loss": 0.6502,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.080325056272222,
+      "learning_rate": 1.637177999064326e-05,
+      "loss": 0.7485,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.656038975348432,
+      "learning_rate": 1.6367678703212515e-05,
+      "loss": 0.6727,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4903419885003844,
+      "learning_rate": 1.636357561342371e-05,
+      "loss": 0.5394,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.116910323375124,
+      "learning_rate": 1.6359470722438212e-05,
+      "loss": 0.606,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5351775836310262,
+      "learning_rate": 1.6355364031417903e-05,
+      "loss": 0.5948,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.593284562495536,
+      "learning_rate": 1.6351255541525182e-05,
+      "loss": 0.6185,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5585107943765084,
+      "learning_rate": 1.6347145253922942e-05,
+      "loss": 0.6769,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4904462827534126,
+      "learning_rate": 1.6343033169774587e-05,
+      "loss": 0.612,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4768445900202125,
+      "learning_rate": 1.633891929024404e-05,
+      "loss": 0.5885,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8619114445447766,
+      "learning_rate": 1.6334803616495722e-05,
+      "loss": 0.6122,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4757523757211555,
+      "learning_rate": 1.633068614969457e-05,
+      "loss": 0.5602,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6574110842590142,
+      "learning_rate": 1.632656689100602e-05,
+      "loss": 0.6057,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.630231043079719,
+      "learning_rate": 1.632244584159602e-05,
+      "loss": 0.6197,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.640238670297492,
+      "learning_rate": 1.631832300263103e-05,
+      "loss": 0.6146,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.482546150050166,
+      "learning_rate": 1.6314198375278003e-05,
+      "loss": 0.5079,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5875493501119153,
+      "learning_rate": 1.6310071960704412e-05,
+      "loss": 0.5745,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8822963368004704,
+      "learning_rate": 1.6305943760078226e-05,
+      "loss": 0.6323,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7316834210017404,
+      "learning_rate": 1.630181377456793e-05,
+      "loss": 0.6401,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7061933543216643,
+      "learning_rate": 1.6297682005342497e-05,
+      "loss": 0.7042,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2594059248246467,
+      "learning_rate": 1.6293548453571422e-05,
+      "loss": 0.563,
+      "step": 1856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.991621141391356,
+      "learning_rate": 1.62894131204247e-05,
+      "loss": 0.5962,
+      "step": 1857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6286114748646603,
+      "learning_rate": 1.628527600707283e-05,
+      "loss": 0.5178,
+      "step": 1858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7407152932755687,
+      "learning_rate": 1.62811371146868e-05,
+      "loss": 0.6412,
+      "step": 1859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6651784443107005,
+      "learning_rate": 1.627699644443813e-05,
+      "loss": 0.4339,
+      "step": 1860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.272867484274822,
+      "learning_rate": 1.6272853997498822e-05,
+      "loss": 0.623,
+      "step": 1861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8072398372842282,
+      "learning_rate": 1.6268709775041385e-05,
+      "loss": 0.5863,
+      "step": 1862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8869152150600886,
+      "learning_rate": 1.6264563778238834e-05,
+      "loss": 0.7346,
+      "step": 1863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8533031537164993,
+      "learning_rate": 1.6260416008264685e-05,
+      "loss": 0.7025,
+      "step": 1864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.884523807794649,
+      "learning_rate": 1.625626646629296e-05,
+      "loss": 0.7649,
+      "step": 1865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.654255253104252,
+      "learning_rate": 1.625211515349817e-05,
+      "loss": 0.5493,
+      "step": 1866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4604307579125644,
+      "learning_rate": 1.624796207105534e-05,
+      "loss": 0.5464,
+      "step": 1867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.060603329591662,
+      "learning_rate": 1.6243807220139988e-05,
+      "loss": 0.671,
+      "step": 1868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.730234619713422,
+      "learning_rate": 1.623965060192814e-05,
+      "loss": 0.6622,
+      "step": 1869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9779354717189478,
+      "learning_rate": 1.623549221759632e-05,
+      "loss": 0.7487,
+      "step": 1870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.767369975664117,
+      "learning_rate": 1.6231332068321538e-05,
+      "loss": 0.6727,
+      "step": 1871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1610866353660554,
+      "learning_rate": 1.622717015528133e-05,
+      "loss": 0.5956,
+      "step": 1872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4602846970721437,
+      "learning_rate": 1.6223006479653708e-05,
+      "loss": 0.6139,
+      "step": 1873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6259694836099683,
+      "learning_rate": 1.6218841042617196e-05,
+      "loss": 0.5275,
+      "step": 1874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.740390642003812,
+      "learning_rate": 1.621467384535081e-05,
+      "loss": 0.5565,
+      "step": 1875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.245813655399566,
+      "learning_rate": 1.6210504889034063e-05,
+      "loss": 0.5717,
+      "step": 1876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8091563400453152,
+      "learning_rate": 1.6206334174846974e-05,
+      "loss": 0.6165,
+      "step": 1877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9438207718540448,
+      "learning_rate": 1.6202161703970057e-05,
+      "loss": 0.5393,
+      "step": 1878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1657048056938994,
+      "learning_rate": 1.6197987477584315e-05,
+      "loss": 0.6276,
+      "step": 1879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2527721490440813,
+      "learning_rate": 1.6193811496871256e-05,
+      "loss": 0.6143,
+      "step": 1880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7998053335769388,
+      "learning_rate": 1.6189633763012885e-05,
+      "loss": 0.4983,
+      "step": 1881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8361504873061876,
+      "learning_rate": 1.61854542771917e-05,
+      "loss": 0.5355,
+      "step": 1882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.606907635695084,
+      "learning_rate": 1.6181273040590696e-05,
+      "loss": 0.5435,
+      "step": 1883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1200815904203276,
+      "learning_rate": 1.617709005439336e-05,
+      "loss": 0.5132,
+      "step": 1884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1221514959897014,
+      "learning_rate": 1.617290531978368e-05,
+      "loss": 0.6332,
+      "step": 1885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5840044533190376,
+      "learning_rate": 1.6168718837946133e-05,
+      "loss": 0.556,
+      "step": 1886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.538574649094453,
+      "learning_rate": 1.61645306100657e-05,
+      "loss": 0.5684,
+      "step": 1887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.490259736361592,
+      "learning_rate": 1.616034063732785e-05,
+      "loss": 0.5791,
+      "step": 1888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5974068974817892,
+      "learning_rate": 1.6156148920918538e-05,
+      "loss": 0.6097,
+      "step": 1889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5609893262238592,
+      "learning_rate": 1.6151955462024225e-05,
+      "loss": 0.5548,
+      "step": 1890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.802694542980552,
+      "learning_rate": 1.6147760261831866e-05,
+      "loss": 0.5894,
+      "step": 1891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8808737405875706,
+      "learning_rate": 1.6143563321528893e-05,
+      "loss": 0.6977,
+      "step": 1892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4788946651233557,
+      "learning_rate": 1.613936464230325e-05,
+      "loss": 0.6503,
+      "step": 1893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5521851591877955,
+      "learning_rate": 1.6135164225343357e-05,
+      "loss": 0.538,
+      "step": 1894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.377292898972991,
+      "learning_rate": 1.613096207183814e-05,
+      "loss": 0.715,
+      "step": 1895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9146658835204358,
+      "learning_rate": 1.6126758182977007e-05,
+      "loss": 0.4738,
+      "step": 1896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4908262893523576,
+      "learning_rate": 1.612255255994986e-05,
+      "loss": 0.5873,
+      "step": 1897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3419185678475003,
+      "learning_rate": 1.6118345203947093e-05,
+      "loss": 0.6095,
+      "step": 1898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8896581892738005,
+      "learning_rate": 1.6114136116159585e-05,
+      "loss": 0.5524,
+      "step": 1899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.973964528533699,
+      "learning_rate": 1.6109925297778717e-05,
+      "loss": 0.642,
+      "step": 1900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5529696246960794,
+      "learning_rate": 1.6105712749996345e-05,
+      "loss": 0.554,
+      "step": 1901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3185573246392104,
+      "learning_rate": 1.610149847400482e-05,
+      "loss": 0.5437,
+      "step": 1902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8181514961221077,
+      "learning_rate": 1.6097282470996997e-05,
+      "loss": 0.6192,
+      "step": 1903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.948272654750491,
+      "learning_rate": 1.609306474216619e-05,
+      "loss": 0.7453,
+      "step": 1904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8893753217410142,
+      "learning_rate": 1.608884528870623e-05,
+      "loss": 0.6781,
+      "step": 1905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.468136948724301,
+      "learning_rate": 1.608462411181142e-05,
+      "loss": 0.5691,
+      "step": 1906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.200105954934806,
+      "learning_rate": 1.6080401212676558e-05,
+      "loss": 0.5461,
+      "step": 1907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6494198788959125,
+      "learning_rate": 1.6076176592496926e-05,
+      "loss": 0.5961,
+      "step": 1908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.531448698006727,
+      "learning_rate": 1.6071950252468288e-05,
+      "loss": 0.4976,
+      "step": 1909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7756991128262507,
+      "learning_rate": 1.6067722193786907e-05,
+      "loss": 0.6213,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.96373630271348,
+      "learning_rate": 1.6063492417649528e-05,
+      "loss": 0.6173,
+      "step": 1911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5258069895005248,
+      "learning_rate": 1.605926092525337e-05,
+      "loss": 0.5037,
+      "step": 1912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.714484987530834,
+      "learning_rate": 1.605502771779616e-05,
+      "loss": 0.4878,
+      "step": 1913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6526386274517564,
+      "learning_rate": 1.6050792796476092e-05,
+      "loss": 0.6117,
+      "step": 1914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6637710630630411,
+      "learning_rate": 1.6046556162491852e-05,
+      "loss": 0.5731,
+      "step": 1915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7068054609444658,
+      "learning_rate": 1.604231781704261e-05,
+      "loss": 0.573,
+      "step": 1916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.172228558466989,
+      "learning_rate": 1.6038077761328024e-05,
+      "loss": 0.7552,
+      "step": 1917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2675941975155365,
+      "learning_rate": 1.603383599654823e-05,
+      "loss": 0.5668,
+      "step": 1918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9228309229151231,
+      "learning_rate": 1.602959252390385e-05,
+      "loss": 0.7632,
+      "step": 1919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5425077436819772,
+      "learning_rate": 1.602534734459599e-05,
+      "loss": 0.5188,
+      "step": 1920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.683196260419575,
+      "learning_rate": 1.6021100459826243e-05,
+      "loss": 0.5875,
+      "step": 1921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.17620981395371,
+      "learning_rate": 1.601685187079668e-05,
+      "loss": 0.6056,
+      "step": 1922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4695935550673918,
+      "learning_rate": 1.601260157870985e-05,
+      "loss": 0.5597,
+      "step": 1923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1601677770475143,
+      "learning_rate": 1.6008349584768793e-05,
+      "loss": 0.6204,
+      "step": 1924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5462101268748254,
+      "learning_rate": 1.6004095890177026e-05,
+      "loss": 0.5728,
+      "step": 1925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7277309899816644,
+      "learning_rate": 1.599984049613855e-05,
+      "loss": 0.6354,
+      "step": 1926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8855262985337515,
+      "learning_rate": 1.5995583403857845e-05,
+      "loss": 0.4907,
+      "step": 1927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7125837289729702,
+      "learning_rate": 1.599132461453987e-05,
+      "loss": 0.5332,
+      "step": 1928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7494773796927037,
+      "learning_rate": 1.5987064129390066e-05,
+      "loss": 0.5644,
+      "step": 1929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.478237463342407,
+      "learning_rate": 1.5982801949614358e-05,
+      "loss": 0.5355,
+      "step": 1930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.323714244721281,
+      "learning_rate": 1.5978538076419143e-05,
+      "loss": 0.5309,
+      "step": 1931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.450740727593207,
+      "learning_rate": 1.5974272511011305e-05,
+      "loss": 0.528,
+      "step": 1932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4283327977053386,
+      "learning_rate": 1.5970005254598204e-05,
+      "loss": 0.4582,
+      "step": 1933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5761563952227469,
+      "learning_rate": 1.5965736308387668e-05,
+      "loss": 0.5032,
+      "step": 1934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6348596921300047,
+      "learning_rate": 1.5961465673588027e-05,
+      "loss": 0.5995,
+      "step": 1935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8971904376028015,
+      "learning_rate": 1.5957193351408065e-05,
+      "loss": 0.5918,
+      "step": 1936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9982117459479718,
+      "learning_rate": 1.595291934305706e-05,
+      "loss": 0.5545,
+      "step": 1937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9611629383551747,
+      "learning_rate": 1.594864364974476e-05,
+      "loss": 0.7066,
+      "step": 1938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0302631207727577,
+      "learning_rate": 1.5944366272681386e-05,
+      "loss": 0.5327,
+      "step": 1939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7434086174249472,
+      "learning_rate": 1.5940087213077648e-05,
+      "loss": 0.5439,
+      "step": 1940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7426500511460916,
+      "learning_rate": 1.593580647214472e-05,
+      "loss": 0.5456,
+      "step": 1941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.8448640849034454,
+      "learning_rate": 1.5931524051094254e-05,
+      "loss": 0.5333,
+      "step": 1942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.509687991364699,
+      "learning_rate": 1.592723995113839e-05,
+      "loss": 0.5757,
+      "step": 1943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0313841363830223,
+      "learning_rate": 1.5922954173489726e-05,
+      "loss": 0.5441,
+      "step": 1944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2598599042952685,
+      "learning_rate": 1.5918666719361346e-05,
+      "loss": 0.4633,
+      "step": 1945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8749270554037687,
+      "learning_rate": 1.5914377589966798e-05,
+      "loss": 0.6052,
+      "step": 1946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2259524156987314,
+      "learning_rate": 1.5910086786520118e-05,
+      "loss": 0.5283,
+      "step": 1947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.729541712593477,
+      "learning_rate": 1.5905794310235808e-05,
+      "loss": 0.5756,
+      "step": 1948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.792925121929122,
+      "learning_rate": 1.590150016232884e-05,
+      "loss": 0.5296,
+      "step": 1949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6466414038402881,
+      "learning_rate": 1.589720434401467e-05,
+      "loss": 0.571,
+      "step": 1950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.977947324878164,
+      "learning_rate": 1.5892906856509214e-05,
+      "loss": 0.5489,
+      "step": 1951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2453556672940858,
+      "learning_rate": 1.5888607701028877e-05,
+      "loss": 0.546,
+      "step": 1952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.714338065581333,
+      "learning_rate": 1.5884306878790512e-05,
+      "loss": 0.5598,
+      "step": 1953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.597840002975269,
+      "learning_rate": 1.5880004391011464e-05,
+      "loss": 0.6492,
+      "step": 1954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.236056663374785,
+      "learning_rate": 1.5875700238909547e-05,
+      "loss": 0.6379,
+      "step": 1955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4001542018324729,
+      "learning_rate": 1.5871394423703036e-05,
+      "loss": 0.5944,
+      "step": 1956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.232064816353976,
+      "learning_rate": 1.5867086946610687e-05,
+      "loss": 0.7658,
+      "step": 1957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0596977324991186,
+      "learning_rate": 1.586277780885172e-05,
+      "loss": 0.5813,
+      "step": 1958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.066289829836428,
+      "learning_rate": 1.585846701164583e-05,
+      "loss": 0.5977,
+      "step": 1959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.097338404854039,
+      "learning_rate": 1.585415455621318e-05,
+      "loss": 0.6335,
+      "step": 1960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.934262455689717,
+      "learning_rate": 1.5849840443774393e-05,
+      "loss": 0.546,
+      "step": 1961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.661719271620289,
+      "learning_rate": 1.584552467555058e-05,
+      "loss": 0.6601,
+      "step": 1962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0967724190552226,
+      "learning_rate": 1.58412072527633e-05,
+      "loss": 0.6205,
+      "step": 1963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8514470148847162,
+      "learning_rate": 1.58368881766346e-05,
+      "loss": 0.5603,
+      "step": 1964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6968448599330619,
+      "learning_rate": 1.5832567448386985e-05,
+      "loss": 0.5253,
+      "step": 1965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3421697993003097,
+      "learning_rate": 1.5828245069243417e-05,
+      "loss": 0.5761,
+      "step": 1966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.479928974593355,
+      "learning_rate": 1.5823921040427348e-05,
+      "loss": 0.4433,
+      "step": 1967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4710983456672908,
+      "learning_rate": 1.5819595363162682e-05,
+      "loss": 0.4823,
+      "step": 1968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.867414858402519,
+      "learning_rate": 1.5815268038673786e-05,
+      "loss": 0.5523,
+      "step": 1969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7805306747458167,
+      "learning_rate": 1.581093906818551e-05,
+      "loss": 0.56,
+      "step": 1970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7637801659465324,
+      "learning_rate": 1.5806608452923158e-05,
+      "loss": 0.6328,
+      "step": 1971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2101810405113804,
+      "learning_rate": 1.5802276194112498e-05,
+      "loss": 0.4957,
+      "step": 1972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.01036544685945,
+      "learning_rate": 1.5797942292979767e-05,
+      "loss": 0.5002,
+      "step": 1973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.310412233705687,
+      "learning_rate": 1.5793606750751668e-05,
+      "loss": 0.5277,
+      "step": 1974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8183536125174877,
+      "learning_rate": 1.578926956865537e-05,
+      "loss": 0.5344,
+      "step": 1975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.466407477168745,
+      "learning_rate": 1.5784930747918492e-05,
+      "loss": 0.5957,
+      "step": 1976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8355299152601063,
+      "learning_rate": 1.578059028976914e-05,
+      "loss": 0.6878,
+      "step": 1977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8040565869843077,
+      "learning_rate": 1.577624819543587e-05,
+      "loss": 0.6676,
+      "step": 1978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5259490869488355,
+      "learning_rate": 1.57719044661477e-05,
+      "loss": 0.5366,
+      "step": 1979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.225492281895553,
+      "learning_rate": 1.5767559103134114e-05,
+      "loss": 0.5517,
+      "step": 1980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2660837939073373,
+      "learning_rate": 1.5763212107625055e-05,
+      "loss": 0.624,
+      "step": 1981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5277160758983246,
+      "learning_rate": 1.5758863480850936e-05,
+      "loss": 0.559,
+      "step": 1982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6720845828185178,
+      "learning_rate": 1.5754513224042625e-05,
+      "loss": 0.5816,
+      "step": 1983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4867980385586128,
+      "learning_rate": 1.5750161338431452e-05,
+      "loss": 0.5492,
+      "step": 1984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0699066715506773,
+      "learning_rate": 1.5745807825249208e-05,
+      "loss": 0.5381,
+      "step": 1985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6973736456668629,
+      "learning_rate": 1.574145268572815e-05,
+      "loss": 0.6305,
+      "step": 1986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7585127871713673,
+      "learning_rate": 1.5737095921100983e-05,
+      "loss": 0.4819,
+      "step": 1987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0653285981732505,
+      "learning_rate": 1.573273753260089e-05,
+      "loss": 0.569,
+      "step": 1988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.102468219986763,
+      "learning_rate": 1.5728377521461496e-05,
+      "loss": 0.4997,
+      "step": 1989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.478621759192977,
+      "learning_rate": 1.57240158889169e-05,
+      "loss": 0.5816,
+      "step": 1990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8105439019481875,
+      "learning_rate": 1.5719652636201646e-05,
+      "loss": 0.5991,
+      "step": 1991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.680389953550436,
+      "learning_rate": 1.5715287764550745e-05,
+      "loss": 0.5744,
+      "step": 1992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1321981381812556,
+      "learning_rate": 1.571092127519967e-05,
+      "loss": 0.5717,
+      "step": 1993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.992287590778129,
+      "learning_rate": 1.570655316938434e-05,
+      "loss": 0.6019,
+      "step": 1994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0095911803842506,
+      "learning_rate": 1.5702183448341143e-05,
+      "loss": 0.5514,
+      "step": 1995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8422761517093529,
+      "learning_rate": 1.5697812113306917e-05,
+      "loss": 0.5741,
+      "step": 1996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6532520340390149,
+      "learning_rate": 1.5693439165518957e-05,
+      "loss": 0.6059,
+      "step": 1997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5528371460659918,
+      "learning_rate": 1.568906460621502e-05,
+      "loss": 0.5334,
+      "step": 1998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.345975707067101,
+      "learning_rate": 1.5684688436633314e-05,
+      "loss": 0.5221,
+      "step": 1999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9973674112340918,
+      "learning_rate": 1.5680310658012507e-05,
+      "loss": 0.5423,
+      "step": 2000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6432731596385055,
+      "learning_rate": 1.5675931271591717e-05,
+      "loss": 0.5964,
+      "step": 2001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8967680270326189,
+      "learning_rate": 1.5671550278610526e-05,
+      "loss": 0.7259,
+      "step": 2002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8098526423772028,
+      "learning_rate": 1.566716768030896e-05,
+      "loss": 0.7271,
+      "step": 2003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.582375518893886,
+      "learning_rate": 1.56627834779275e-05,
+      "loss": 0.5702,
+      "step": 2004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9438441584949042,
+      "learning_rate": 1.5658397672707093e-05,
+      "loss": 0.5168,
+      "step": 2005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5811696939180158,
+      "learning_rate": 1.565401026588913e-05,
+      "loss": 0.4945,
+      "step": 2006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.573858829384964,
+      "learning_rate": 1.5649621258715454e-05,
+      "loss": 0.5055,
+      "step": 2007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.427548096963402,
+      "learning_rate": 1.5645230652428367e-05,
+      "loss": 0.6556,
+      "step": 2008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.603412713064446,
+      "learning_rate": 1.564083844827062e-05,
+      "loss": 0.5117,
+      "step": 2009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.642782060311387,
+      "learning_rate": 1.563644464748542e-05,
+      "loss": 0.5348,
+      "step": 2010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6324823315852275,
+      "learning_rate": 1.563204925131642e-05,
+      "loss": 0.5993,
+      "step": 2011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9328018794751243,
+      "learning_rate": 1.5627652261007726e-05,
+      "loss": 0.6599,
+      "step": 2012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5748294851321112,
+      "learning_rate": 1.5623253677803897e-05,
+      "loss": 0.5778,
+      "step": 2013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.712573056118339,
+      "learning_rate": 1.5618853502949948e-05,
+      "loss": 0.681,
+      "step": 2014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.615056106239427,
+      "learning_rate": 1.5614451737691335e-05,
+      "loss": 0.5286,
+      "step": 2015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.380482602798708,
+      "learning_rate": 1.561004838327397e-05,
+      "loss": 0.5468,
+      "step": 2016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5546302808463708,
+      "learning_rate": 1.5605643440944213e-05,
+      "loss": 0.5771,
+      "step": 2017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.919091369966952,
+      "learning_rate": 1.5601236911948876e-05,
+      "loss": 0.5703,
+      "step": 2018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0723459838567724,
+      "learning_rate": 1.559682879753521e-05,
+      "loss": 0.5599,
+      "step": 2019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7655152587908076,
+      "learning_rate": 1.559241909895093e-05,
+      "loss": 0.5946,
+      "step": 2020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6107729805177995,
+      "learning_rate": 1.558800781744419e-05,
+      "loss": 0.6338,
+      "step": 2021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7217293141608987,
+      "learning_rate": 1.5583594954263593e-05,
+      "loss": 0.523,
+      "step": 2022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4470021586330943,
+      "learning_rate": 1.5579180510658187e-05,
+      "loss": 0.5829,
+      "step": 2023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.474964323577696,
+      "learning_rate": 1.557476448787748e-05,
+      "loss": 0.5032,
+      "step": 2024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8008702294071333,
+      "learning_rate": 1.557034688717141e-05,
+      "loss": 0.6964,
+      "step": 2025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.772785377654187,
+      "learning_rate": 1.5565927709790377e-05,
+      "loss": 0.5256,
+      "step": 2026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4313551270200757,
+      "learning_rate": 1.5561506956985213e-05,
+      "loss": 0.6281,
+      "step": 2027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8431531810883373,
+      "learning_rate": 1.5557084630007206e-05,
+      "loss": 0.6016,
+      "step": 2028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5962748023778202,
+      "learning_rate": 1.5552660730108084e-05,
+      "loss": 0.628,
+      "step": 2029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6435601843694523,
+      "learning_rate": 1.5548235258540023e-05,
+      "loss": 0.5314,
+      "step": 2030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4568229106386668,
+      "learning_rate": 1.5543808216555645e-05,
+      "loss": 0.4903,
+      "step": 2031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8473591206401785,
+      "learning_rate": 1.5539379605408015e-05,
+      "loss": 0.489,
+      "step": 2032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.644337342989557,
+      "learning_rate": 1.5534949426350642e-05,
+      "loss": 0.5723,
+      "step": 2033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.561204792498149,
+      "learning_rate": 1.5530517680637478e-05,
+      "loss": 0.579,
+      "step": 2034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7157725078433526,
+      "learning_rate": 1.552608436952292e-05,
+      "loss": 0.5116,
+      "step": 2035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.279276488268159,
+      "learning_rate": 1.552164949426181e-05,
+      "loss": 0.5516,
+      "step": 2036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9448567376187627,
+      "learning_rate": 1.551721305610942e-05,
+      "loss": 0.6199,
+      "step": 2037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4350515695727877,
+      "learning_rate": 1.551277505632149e-05,
+      "loss": 0.4542,
+      "step": 2038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9058705829362568,
+      "learning_rate": 1.550833549615417e-05,
+      "loss": 0.5072,
+      "step": 2039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3698388117490694,
+      "learning_rate": 1.550389437686408e-05,
+      "loss": 0.5292,
+      "step": 2040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6890138783453028,
+      "learning_rate": 1.549945169970827e-05,
+      "loss": 0.5687,
+      "step": 2041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6495092662985626,
+      "learning_rate": 1.549500746594422e-05,
+      "loss": 0.6296,
+      "step": 2042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6633738133038807,
+      "learning_rate": 1.549056167682987e-05,
+      "loss": 0.5332,
+      "step": 2043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5596818682374485,
+      "learning_rate": 1.5486114333623587e-05,
+      "loss": 0.5927,
+      "step": 2044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5789987780447226,
+      "learning_rate": 1.5481665437584185e-05,
+      "loss": 0.6822,
+      "step": 2045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5951347273665486,
+      "learning_rate": 1.5477214989970916e-05,
+      "loss": 0.5415,
+      "step": 2046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8685249454254265,
+      "learning_rate": 1.547276299204346e-05,
+      "loss": 0.6914,
+      "step": 2047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5270701699511866,
+      "learning_rate": 1.546830944506196e-05,
+      "loss": 0.5511,
+      "step": 2048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7208753650843431,
+      "learning_rate": 1.5463854350286972e-05,
+      "loss": 0.5415,
+      "step": 2049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6773670344031832,
+      "learning_rate": 1.5459397708979508e-05,
+      "loss": 0.5218,
+      "step": 2050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7598693951528879,
+      "learning_rate": 1.5454939522401e-05,
+      "loss": 0.5264,
+      "step": 2051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6576439836004746,
+      "learning_rate": 1.5450479791813348e-05,
+      "loss": 0.5343,
+      "step": 2052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3690026557724977,
+      "learning_rate": 1.544601851847885e-05,
+      "loss": 0.4861,
+      "step": 2053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.718229986081342,
+      "learning_rate": 1.544155570366027e-05,
+      "loss": 0.5754,
+      "step": 2054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.366214671230858,
+      "learning_rate": 1.5437091348620798e-05,
+      "loss": 0.4493,
+      "step": 2055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7610959911469175,
+      "learning_rate": 1.5432625454624054e-05,
+      "loss": 0.5801,
+      "step": 2056
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6358955248475493,
+      "learning_rate": 1.5428158022934106e-05,
+      "loss": 0.6243,
+      "step": 2057
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5342081183737413,
+      "learning_rate": 1.542368905481545e-05,
+      "loss": 0.5252,
+      "step": 2058
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1967855999662427,
+      "learning_rate": 1.5419218551533017e-05,
+      "loss": 0.5049,
+      "step": 2059
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.081346520287825,
+      "learning_rate": 1.5414746514352178e-05,
+      "loss": 0.5695,
+      "step": 2060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7810486151893183,
+      "learning_rate": 1.5410272944538725e-05,
+      "loss": 0.5759,
+      "step": 2061
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6518660846888655,
+      "learning_rate": 1.5405797843358897e-05,
+      "loss": 0.5616,
+      "step": 2062
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8194679396673614,
+      "learning_rate": 1.5401321212079366e-05,
+      "loss": 0.5435,
+      "step": 2063
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6802904881702074,
+      "learning_rate": 1.5396843051967225e-05,
+      "loss": 0.6595,
+      "step": 2064
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.682306924438514,
+      "learning_rate": 1.5392363364290016e-05,
+      "loss": 0.6858,
+      "step": 2065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7867494229098102,
+      "learning_rate": 1.53878821503157e-05,
+      "loss": 0.637,
+      "step": 2066
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.722913895035403,
+      "learning_rate": 1.5383399411312673e-05,
+      "loss": 0.632,
+      "step": 2067
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.318954645143978,
+      "learning_rate": 1.5378915148549772e-05,
+      "loss": 0.4918,
+      "step": 2068
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.430643939105609,
+      "learning_rate": 1.5374429363296252e-05,
+      "loss": 0.5916,
+      "step": 2069
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.110428214069378,
+      "learning_rate": 1.536994205682181e-05,
+      "loss": 0.5875,
+      "step": 2070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5576469864918727,
+      "learning_rate": 1.536545323039657e-05,
+      "loss": 0.5294,
+      "step": 2071
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6348666527924305,
+      "learning_rate": 1.5360962885291074e-05,
+      "loss": 0.5762,
+      "step": 2072
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9946661144498552,
+      "learning_rate": 1.5356471022776315e-05,
+      "loss": 0.6181,
+      "step": 2073
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5786834104758016,
+      "learning_rate": 1.5351977644123703e-05,
+      "loss": 0.6186,
+      "step": 2074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8306852923539912,
+      "learning_rate": 1.5347482750605072e-05,
+      "loss": 0.6778,
+      "step": 2075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5362104243991417,
+      "learning_rate": 1.5342986343492704e-05,
+      "loss": 0.5961,
+      "step": 2076
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.663356048397953,
+      "learning_rate": 1.533848842405929e-05,
+      "loss": 0.6163,
+      "step": 2077
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.732764278274162,
+      "learning_rate": 1.5333988993577958e-05,
+      "loss": 0.4974,
+      "step": 2078
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.607700876131261,
+      "learning_rate": 1.5329488053322266e-05,
+      "loss": 0.6124,
+      "step": 2079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4334479628610641,
+      "learning_rate": 1.5324985604566194e-05,
+      "loss": 0.5351,
+      "step": 2080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5940836406109138,
+      "learning_rate": 1.5320481648584147e-05,
+      "loss": 0.5465,
+      "step": 2081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5412140178410823,
+      "learning_rate": 1.5315976186650962e-05,
+      "loss": 0.5923,
+      "step": 2082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9203482014595423,
+      "learning_rate": 1.5311469220041903e-05,
+      "loss": 0.5952,
+      "step": 2083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.625966584504904,
+      "learning_rate": 1.5306960750032657e-05,
+      "loss": 0.5159,
+      "step": 2084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2528375861327605,
+      "learning_rate": 1.5302450777899332e-05,
+      "loss": 0.5428,
+      "step": 2085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5042836787061684,
+      "learning_rate": 1.529793930491847e-05,
+      "loss": 0.54,
+      "step": 2086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8577992800394587,
+      "learning_rate": 1.5293426332367034e-05,
+      "loss": 0.471,
+      "step": 2087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0669579919711127,
+      "learning_rate": 1.5288911861522413e-05,
+      "loss": 0.5598,
+      "step": 2088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8578459531309022,
+      "learning_rate": 1.5284395893662414e-05,
+      "loss": 0.71,
+      "step": 2089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.946007133190683,
+      "learning_rate": 1.5279878430065277e-05,
+      "loss": 0.4929,
+      "step": 2090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7932467139689369,
+      "learning_rate": 1.5275359472009656e-05,
+      "loss": 0.741,
+      "step": 2091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5628712529631494,
+      "learning_rate": 1.5270839020774638e-05,
+      "loss": 0.6275,
+      "step": 2092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9774392203335451,
+      "learning_rate": 1.526631707763972e-05,
+      "loss": 0.6253,
+      "step": 2093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5423203954118023,
+      "learning_rate": 1.5261793643884835e-05,
+      "loss": 0.5669,
+      "step": 2094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.455667604739606,
+      "learning_rate": 1.5257268720790328e-05,
+      "loss": 0.4929,
+      "step": 2095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.664045905894347,
+      "learning_rate": 1.5252742309636972e-05,
+      "loss": 0.6247,
+      "step": 2096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7002573508633356,
+      "learning_rate": 1.5248214411705955e-05,
+      "loss": 0.554,
+      "step": 2097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4841519614177003,
+      "learning_rate": 1.5243685028278888e-05,
+      "loss": 0.5896,
+      "step": 2098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2624128173095697,
+      "learning_rate": 1.5239154160637805e-05,
+      "loss": 0.4357,
+      "step": 2099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.400400471634663,
+      "learning_rate": 1.5234621810065164e-05,
+      "loss": 0.5241,
+      "step": 2100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5451744880220692,
+      "learning_rate": 1.5230087977843826e-05,
+      "loss": 0.5445,
+      "step": 2101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.561023760484766,
+      "learning_rate": 1.5225552665257092e-05,
+      "loss": 0.5945,
+      "step": 2102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4558688865684664,
+      "learning_rate": 1.5221015873588672e-05,
+      "loss": 0.4958,
+      "step": 2103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.418163641138415,
+      "learning_rate": 1.521647760412269e-05,
+      "loss": 0.4949,
+      "step": 2104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3254503101823982,
+      "learning_rate": 1.5211937858143699e-05,
+      "loss": 0.5304,
+      "step": 2105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8916918552496282,
+      "learning_rate": 1.5207396636936662e-05,
+      "loss": 0.6346,
+      "step": 2106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5193443433255707,
+      "learning_rate": 1.520285394178696e-05,
+      "loss": 0.506,
+      "step": 2107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1295195157505056,
+      "learning_rate": 1.5198309773980397e-05,
+      "loss": 0.6147,
+      "step": 2108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0204743289746525,
+      "learning_rate": 1.5193764134803188e-05,
+      "loss": 0.5547,
+      "step": 2109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7401546839520257,
+      "learning_rate": 1.5189217025541969e-05,
+      "loss": 0.6288,
+      "step": 2110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5033884318106514,
+      "learning_rate": 1.5184668447483785e-05,
+      "loss": 0.565,
+      "step": 2111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4053112603151483,
+      "learning_rate": 1.5180118401916107e-05,
+      "loss": 0.6429,
+      "step": 2112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6119326243375487,
+      "learning_rate": 1.5175566890126812e-05,
+      "loss": 0.5404,
+      "step": 2113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3788375925637864,
+      "learning_rate": 1.51710139134042e-05,
+      "loss": 0.6002,
+      "step": 2114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1704361987466663,
+      "learning_rate": 1.5166459473036977e-05,
+      "loss": 0.5269,
+      "step": 2115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.190912312027071,
+      "learning_rate": 1.5161903570314268e-05,
+      "loss": 0.5527,
+      "step": 2116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6115949037842308,
+      "learning_rate": 1.5157346206525613e-05,
+      "loss": 0.5343,
+      "step": 2117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.408699071801244,
+      "learning_rate": 1.5152787382960968e-05,
+      "loss": 0.5873,
+      "step": 2118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.320583100995214,
+      "learning_rate": 1.5148227100910691e-05,
+      "loss": 0.4813,
+      "step": 2119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.842768385422223,
+      "learning_rate": 1.5143665361665565e-05,
+      "loss": 0.543,
+      "step": 2120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6061795818979692,
+      "learning_rate": 1.5139102166516782e-05,
+      "loss": 0.5416,
+      "step": 2121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.332260775466258,
+      "learning_rate": 1.5134537516755938e-05,
+      "loss": 0.5785,
+      "step": 2122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4692831151276253,
+      "learning_rate": 1.5129971413675055e-05,
+      "loss": 0.5409,
+      "step": 2123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7648568017461943,
+      "learning_rate": 1.5125403858566552e-05,
+      "loss": 0.5426,
+      "step": 2124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4225573025532134,
+      "learning_rate": 1.5120834852723276e-05,
+      "loss": 0.4201,
+      "step": 2125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.841850449960772,
+      "learning_rate": 1.5116264397438465e-05,
+      "loss": 0.5975,
+      "step": 2126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5596968750328493,
+      "learning_rate": 1.511169249400578e-05,
+      "loss": 0.5796,
+      "step": 2127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6929386049716404,
+      "learning_rate": 1.510711914371929e-05,
+      "loss": 0.6151,
+      "step": 2128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.687116512605874,
+      "learning_rate": 1.5102544347873469e-05,
+      "loss": 0.679,
+      "step": 2129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5604224533103477,
+      "learning_rate": 1.5097968107763205e-05,
+      "loss": 0.5651,
+      "step": 2130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7933787394079834,
+      "learning_rate": 1.5093390424683796e-05,
+      "loss": 0.5239,
+      "step": 2131
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3844844650664756,
+      "learning_rate": 1.5088811299930942e-05,
+      "loss": 0.5979,
+      "step": 2132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4880287299161743,
+      "learning_rate": 1.5084230734800754e-05,
+      "loss": 0.5236,
+      "step": 2133
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.789326482523847,
+      "learning_rate": 1.5079648730589753e-05,
+      "loss": 0.5169,
+      "step": 2134
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5985334826725477,
+      "learning_rate": 1.5075065288594864e-05,
+      "loss": 0.5445,
+      "step": 2135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9557997906866255,
+      "learning_rate": 1.5070480410113427e-05,
+      "loss": 0.698,
+      "step": 2136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9297821580750467,
+      "learning_rate": 1.5065894096443173e-05,
+      "loss": 0.7128,
+      "step": 2137
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7956811639715125,
+      "learning_rate": 1.5061306348882252e-05,
+      "loss": 0.589,
+      "step": 2138
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9120449145217548,
+      "learning_rate": 1.505671716872922e-05,
+      "loss": 0.557,
+      "step": 2139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8449436869949487,
+      "learning_rate": 1.5052126557283031e-05,
+      "loss": 0.6235,
+      "step": 2140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.902989474837001,
+      "learning_rate": 1.5047534515843047e-05,
+      "loss": 0.6436,
+      "step": 2141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.999740186106436,
+      "learning_rate": 1.5042941045709039e-05,
+      "loss": 0.6335,
+      "step": 2142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.915822085002048,
+      "learning_rate": 1.5038346148181178e-05,
+      "loss": 0.5638,
+      "step": 2143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.347406025924967,
+      "learning_rate": 1.5033749824560037e-05,
+      "loss": 0.5702,
+      "step": 2144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9830108025555098,
+      "learning_rate": 1.50291520761466e-05,
+      "loss": 0.6076,
+      "step": 2145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6983848692680266,
+      "learning_rate": 1.5024552904242246e-05,
+      "loss": 0.4854,
+      "step": 2146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7806916085166188,
+      "learning_rate": 1.5019952310148766e-05,
+      "loss": 0.6415,
+      "step": 2147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6424072732675183,
+      "learning_rate": 1.5015350295168344e-05,
+      "loss": 0.6094,
+      "step": 2148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.087527042553089,
+      "learning_rate": 1.5010746860603575e-05,
+      "loss": 0.5647,
+      "step": 2149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.569087033751852,
+      "learning_rate": 1.5006142007757446e-05,
+      "loss": 0.5755,
+      "step": 2150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.671219466074002,
+      "learning_rate": 1.5001535737933355e-05,
+      "loss": 0.469,
+      "step": 2151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9170238876017622,
+      "learning_rate": 1.4996928052435095e-05,
+      "loss": 0.5369,
+      "step": 2152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.902176633425763,
+      "learning_rate": 1.4992318952566862e-05,
+      "loss": 0.5587,
+      "step": 2153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6710848121316388,
+      "learning_rate": 1.4987708439633255e-05,
+      "loss": 0.6879,
+      "step": 2154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5272504474079964,
+      "learning_rate": 1.4983096514939263e-05,
+      "loss": 0.579,
+      "step": 2155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0531547554168204,
+      "learning_rate": 1.497848317979029e-05,
+      "loss": 0.59,
+      "step": 2156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6553443695522392,
+      "learning_rate": 1.4973868435492125e-05,
+      "loss": 0.6595,
+      "step": 2157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6496670575352685,
+      "learning_rate": 1.4969252283350964e-05,
+      "loss": 0.5426,
+      "step": 2158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.248094187071528,
+      "learning_rate": 1.4964634724673397e-05,
+      "loss": 0.5837,
+      "step": 2159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5629225624820076,
+      "learning_rate": 1.4960015760766418e-05,
+      "loss": 0.5048,
+      "step": 2160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8501122236187155,
+      "learning_rate": 1.495539539293741e-05,
+      "loss": 0.7006,
+      "step": 2161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.389085051274074,
+      "learning_rate": 1.4950773622494166e-05,
+      "loss": 0.5867,
+      "step": 2162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.562426909915579,
+      "learning_rate": 1.4946150450744859e-05,
+      "loss": 0.5869,
+      "step": 2163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7297662458843923,
+      "learning_rate": 1.4941525878998073e-05,
+      "loss": 0.6587,
+      "step": 2164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.094283667135443,
+      "learning_rate": 1.4936899908562788e-05,
+      "loss": 0.7019,
+      "step": 2165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.580470406440832,
+      "learning_rate": 1.4932272540748366e-05,
+      "loss": 0.6172,
+      "step": 2166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6967568608445955,
+      "learning_rate": 1.4927643776864577e-05,
+      "loss": 0.5766,
+      "step": 2167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.643675359212008,
+      "learning_rate": 1.4923013618221584e-05,
+      "loss": 0.5571,
+      "step": 2168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6235340987143598,
+      "learning_rate": 1.4918382066129946e-05,
+      "loss": 0.628,
+      "step": 2169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4659031123249098,
+      "learning_rate": 1.4913749121900611e-05,
+      "loss": 0.5606,
+      "step": 2170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7480132180294723,
+      "learning_rate": 1.4909114786844925e-05,
+      "loss": 0.6732,
+      "step": 2171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4985750660778308,
+      "learning_rate": 1.4904479062274627e-05,
+      "loss": 0.5692,
+      "step": 2172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5972827368637292,
+      "learning_rate": 1.4899841949501845e-05,
+      "loss": 0.6267,
+      "step": 2173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.680229856181742,
+      "learning_rate": 1.4895203449839111e-05,
+      "loss": 0.6928,
+      "step": 2174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7827067333987952,
+      "learning_rate": 1.4890563564599337e-05,
+      "loss": 0.5505,
+      "step": 2175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8724783528593185,
+      "learning_rate": 1.4885922295095836e-05,
+      "loss": 0.6805,
+      "step": 2176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7608620513851632,
+      "learning_rate": 1.4881279642642308e-05,
+      "loss": 0.6285,
+      "step": 2177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5252044771678739,
+      "learning_rate": 1.4876635608552845e-05,
+      "loss": 0.5595,
+      "step": 2178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.033686240704862,
+      "learning_rate": 1.4871990194141934e-05,
+      "loss": 0.5476,
+      "step": 2179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8655201668022767,
+      "learning_rate": 1.486734340072445e-05,
+      "loss": 0.5257,
+      "step": 2180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5587661724130502,
+      "learning_rate": 1.4862695229615654e-05,
+      "loss": 0.5791,
+      "step": 2181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.626791990196542,
+      "learning_rate": 1.4858045682131203e-05,
+      "loss": 0.5462,
+      "step": 2182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7714745613147924,
+      "learning_rate": 1.4853394759587146e-05,
+      "loss": 0.5863,
+      "step": 2183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5366786919317976,
+      "learning_rate": 1.4848742463299907e-05,
+      "loss": 0.5122,
+      "step": 2184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1855025663584184,
+      "learning_rate": 1.484408879458632e-05,
+      "loss": 0.5404,
+      "step": 2185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5718702436284169,
+      "learning_rate": 1.4839433754763588e-05,
+      "loss": 0.4984,
+      "step": 2186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5903927848167614,
+      "learning_rate": 1.4834777345149313e-05,
+      "loss": 0.6487,
+      "step": 2187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.137074381208324,
+      "learning_rate": 1.4830119567061484e-05,
+      "loss": 0.5641,
+      "step": 2188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.387082592620697,
+      "learning_rate": 1.4825460421818472e-05,
+      "loss": 0.5315,
+      "step": 2189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7995251358406565,
+      "learning_rate": 1.4820799910739042e-05,
+      "loss": 0.6237,
+      "step": 2190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5248472338967933,
+      "learning_rate": 1.4816138035142334e-05,
+      "loss": 0.3661,
+      "step": 2191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6332968985824572,
+      "learning_rate": 1.481147479634789e-05,
+      "loss": 0.5759,
+      "step": 2192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.32327210116816,
+      "learning_rate": 1.4806810195675627e-05,
+      "loss": 0.674,
+      "step": 2193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5992753349181528,
+      "learning_rate": 1.480214423444585e-05,
+      "loss": 0.6623,
+      "step": 2194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9318255349602784,
+      "learning_rate": 1.4797476913979251e-05,
+      "loss": 0.6462,
+      "step": 2195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.324151094164312,
+      "learning_rate": 1.4792808235596907e-05,
+      "loss": 0.5467,
+      "step": 2196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5696865847665358,
+      "learning_rate": 1.4788138200620272e-05,
+      "loss": 0.6047,
+      "step": 2197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.580919269868831,
+      "learning_rate": 1.4783466810371195e-05,
+      "loss": 0.511,
+      "step": 2198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6406575873027627,
+      "learning_rate": 1.47787940661719e-05,
+      "loss": 0.5965,
+      "step": 2199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.26162754450886,
+      "learning_rate": 1.4774119969344996e-05,
+      "loss": 0.4876,
+      "step": 2200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5763011795357045,
+      "learning_rate": 1.4769444521213482e-05,
+      "loss": 0.5625,
+      "step": 2201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5841008181624854,
+      "learning_rate": 1.4764767723100729e-05,
+      "loss": 0.5521,
+      "step": 2202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9870852656958622,
+      "learning_rate": 1.4760089576330493e-05,
+      "loss": 0.5279,
+      "step": 2203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5165748059603108,
+      "learning_rate": 1.475541008222692e-05,
+      "loss": 0.5827,
+      "step": 2204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7910354154351844,
+      "learning_rate": 1.4750729242114527e-05,
+      "loss": 0.544,
+      "step": 2205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.627190025113575,
+      "learning_rate": 1.4746047057318217e-05,
+      "loss": 0.6188,
+      "step": 2206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6301766658376895,
+      "learning_rate": 1.4741363529163273e-05,
+      "loss": 0.5042,
+      "step": 2207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.191685798796917,
+      "learning_rate": 1.4736678658975357e-05,
+      "loss": 0.5282,
+      "step": 2208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6466860395525273,
+      "learning_rate": 1.4731992448080509e-05,
+      "loss": 0.5797,
+      "step": 2209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.75250496627278,
+      "learning_rate": 1.4727304897805157e-05,
+      "loss": 0.5689,
+      "step": 2210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5965166480348436,
+      "learning_rate": 1.47226160094761e-05,
+      "loss": 0.5438,
+      "step": 2211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9303328670738478,
+      "learning_rate": 1.4717925784420514e-05,
+      "loss": 0.5632,
+      "step": 2212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4412157186088925,
+      "learning_rate": 1.471323422396596e-05,
+      "loss": 0.5927,
+      "step": 2213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5178346754787353,
+      "learning_rate": 1.4708541329440375e-05,
+      "loss": 0.6047,
+      "step": 2214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.480474397255128,
+      "learning_rate": 1.4703847102172074e-05,
+      "loss": 0.5136,
+      "step": 2215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5825143724076056,
+      "learning_rate": 1.4699151543489745e-05,
+      "loss": 0.566,
+      "step": 2216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6143640574819123,
+      "learning_rate": 1.4694454654722459e-05,
+      "loss": 0.4851,
+      "step": 2217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6226938967337858,
+      "learning_rate": 1.4689756437199658e-05,
+      "loss": 0.5553,
+      "step": 2218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6096059448496076,
+      "learning_rate": 1.4685056892251167e-05,
+      "loss": 0.6164,
+      "step": 2219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4942813012601461,
+      "learning_rate": 1.4680356021207176e-05,
+      "loss": 0.5636,
+      "step": 2220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6195173895603796,
+      "learning_rate": 1.4675653825398261e-05,
+      "loss": 0.6079,
+      "step": 2221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8638607721576768,
+      "learning_rate": 1.4670950306155368e-05,
+      "loss": 0.5103,
+      "step": 2222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9461037952659221,
+      "learning_rate": 1.4666245464809818e-05,
+      "loss": 0.5318,
+      "step": 2223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4879140120419523,
+      "learning_rate": 1.4661539302693306e-05,
+      "loss": 0.4639,
+      "step": 2224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.580086685648302,
+      "learning_rate": 1.46568318211379e-05,
+      "loss": 0.6336,
+      "step": 2225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8685993358064115,
+      "learning_rate": 1.4652123021476044e-05,
+      "loss": 0.5976,
+      "step": 2226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.514935059118785,
+      "learning_rate": 1.4647412905040553e-05,
+      "loss": 0.5214,
+      "step": 2227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9793550721813808,
+      "learning_rate": 1.4642701473164618e-05,
+      "loss": 0.4568,
+      "step": 2228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1196414669089516,
+      "learning_rate": 1.4637988727181798e-05,
+      "loss": 0.6096,
+      "step": 2229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6485708161073005,
+      "learning_rate": 1.4633274668426028e-05,
+      "loss": 0.6043,
+      "step": 2230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4888971290722646,
+      "learning_rate": 1.462855929823161e-05,
+      "loss": 0.6633,
+      "step": 2231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.948532974537591,
+      "learning_rate": 1.4623842617933219e-05,
+      "loss": 0.5707,
+      "step": 2232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5679755342598753,
+      "learning_rate": 1.4619124628865904e-05,
+      "loss": 0.6424,
+      "step": 2233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6453069870021026,
+      "learning_rate": 1.461440533236508e-05,
+      "loss": 0.5561,
+      "step": 2234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.252614506518848,
+      "learning_rate": 1.4609684729766536e-05,
+      "loss": 0.5966,
+      "step": 2235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.382029748897145,
+      "learning_rate": 1.4604962822406426e-05,
+      "loss": 0.5467,
+      "step": 2236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3580942726959955,
+      "learning_rate": 1.4600239611621274e-05,
+      "loss": 0.5751,
+      "step": 2237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1373069813304157,
+      "learning_rate": 1.459551509874798e-05,
+      "loss": 0.5066,
+      "step": 2238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5989412025582488,
+      "learning_rate": 1.4590789285123808e-05,
+      "loss": 0.5648,
+      "step": 2239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9157777612932592,
+      "learning_rate": 1.4586062172086383e-05,
+      "loss": 0.703,
+      "step": 2240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4147239943718124,
+      "learning_rate": 1.4581333760973713e-05,
+      "loss": 0.6353,
+      "step": 2241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5499004733071127,
+      "learning_rate": 1.4576604053124154e-05,
+      "loss": 0.4556,
+      "step": 2242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5341935592848361,
+      "learning_rate": 1.4571873049876452e-05,
+      "loss": 0.5862,
+      "step": 2243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.829721236245094,
+      "learning_rate": 1.4567140752569701e-05,
+      "loss": 0.635,
+      "step": 2244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4887403656158904,
+      "learning_rate": 1.4562407162543367e-05,
+      "loss": 0.5128,
+      "step": 2245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8626225946040234,
+      "learning_rate": 1.4557672281137286e-05,
+      "loss": 0.4994,
+      "step": 2246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3971570360625947,
+      "learning_rate": 1.455293610969165e-05,
+      "loss": 0.5993,
+      "step": 2247
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3953673708676866,
+      "learning_rate": 1.454819864954703e-05,
+      "loss": 0.501,
+      "step": 2248
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5200583565629504,
+      "learning_rate": 1.4543459902044347e-05,
+      "loss": 0.4785,
+      "step": 2249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5351657159968681,
+      "learning_rate": 1.45387198685249e-05,
+      "loss": 0.6234,
+      "step": 2250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8724311418953191,
+      "learning_rate": 1.4533978550330343e-05,
+      "loss": 0.6855,
+      "step": 2251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4513773960082763,
+      "learning_rate": 1.4529235948802696e-05,
+      "loss": 0.5581,
+      "step": 2252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7410388433980446,
+      "learning_rate": 1.4524492065284344e-05,
+      "loss": 0.5505,
+      "step": 2253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.125170519101055,
+      "learning_rate": 1.4519746901118029e-05,
+      "loss": 0.6381,
+      "step": 2254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7651821426829648,
+      "learning_rate": 1.4515000457646866e-05,
+      "loss": 0.569,
+      "step": 2255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4810471938826315,
+      "learning_rate": 1.4510252736214318e-05,
+      "loss": 0.6172,
+      "step": 2256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4123541858042408,
+      "learning_rate": 1.4505503738164225e-05,
+      "loss": 0.5469,
+      "step": 2257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6224487074012481,
+      "learning_rate": 1.4500753464840775e-05,
+      "loss": 0.5139,
+      "step": 2258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4956040173578873,
+      "learning_rate": 1.4496001917588528e-05,
+      "loss": 0.5125,
+      "step": 2259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4957736730512243,
+      "learning_rate": 1.4491249097752393e-05,
+      "loss": 0.4693,
+      "step": 2260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7888906642228968,
+      "learning_rate": 1.448649500667765e-05,
+      "loss": 0.6148,
+      "step": 2261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3323488618021981,
+      "learning_rate": 1.4481739645709935e-05,
+      "loss": 0.4756,
+      "step": 2262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5305888435501909,
+      "learning_rate": 1.4476983016195245e-05,
+      "loss": 0.5667,
+      "step": 2263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.830755077128761,
+      "learning_rate": 1.4472225119479928e-05,
+      "loss": 0.6244,
+      "step": 2264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8688054943956072,
+      "learning_rate": 1.4467465956910704e-05,
+      "loss": 0.6295,
+      "step": 2265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8437090642547683,
+      "learning_rate": 1.4462705529834635e-05,
+      "loss": 0.7032,
+      "step": 2266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.00564972466476,
+      "learning_rate": 1.4457943839599158e-05,
+      "loss": 0.5985,
+      "step": 2267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.891836251697731,
+      "learning_rate": 1.4453180887552052e-05,
+      "loss": 0.49,
+      "step": 2268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8952799771755648,
+      "learning_rate": 1.4448416675041465e-05,
+      "loss": 0.5927,
+      "step": 2269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6516924388248209,
+      "learning_rate": 1.44436512034159e-05,
+      "loss": 0.5432,
+      "step": 2270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4184347194771691,
+      "learning_rate": 1.443888447402421e-05,
+      "loss": 0.5039,
+      "step": 2271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7143680789021207,
+      "learning_rate": 1.4434116488215603e-05,
+      "loss": 0.6643,
+      "step": 2272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.996344802742972,
+      "learning_rate": 1.4429347247339656e-05,
+      "loss": 0.5566,
+      "step": 2273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.644588811724676,
+      "learning_rate": 1.4424576752746288e-05,
+      "loss": 0.5496,
+      "step": 2274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4923719662890853,
+      "learning_rate": 1.4419805005785783e-05,
+      "loss": 0.6215,
+      "step": 2275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5420877894521292,
+      "learning_rate": 1.4415032007808767e-05,
+      "loss": 0.5187,
+      "step": 2276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8894110087514407,
+      "learning_rate": 1.441025776016623e-05,
+      "loss": 0.7416,
+      "step": 2277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6146626501430352,
+      "learning_rate": 1.4405482264209512e-05,
+      "loss": 0.6449,
+      "step": 2278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4582988562585144,
+      "learning_rate": 1.4400705521290306e-05,
+      "loss": 0.5708,
+      "step": 2279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7679503967464893,
+      "learning_rate": 1.4395927532760664e-05,
+      "loss": 0.6057,
+      "step": 2280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.539816219248465,
+      "learning_rate": 1.4391148299972978e-05,
+      "loss": 0.5496,
+      "step": 2281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.370459832604894,
+      "learning_rate": 1.4386367824280006e-05,
+      "loss": 0.4825,
+      "step": 2282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4687200879070392,
+      "learning_rate": 1.4381586107034849e-05,
+      "loss": 0.5178,
+      "step": 2283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6523540272711936,
+      "learning_rate": 1.437680314959096e-05,
+      "loss": 0.6263,
+      "step": 2284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6556767699566548,
+      "learning_rate": 1.437201895330215e-05,
+      "loss": 0.578,
+      "step": 2285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4356555502608916,
+      "learning_rate": 1.4367233519522571e-05,
+      "loss": 0.5143,
+      "step": 2286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.40792799523231,
+      "learning_rate": 1.4362446849606737e-05,
+      "loss": 0.5529,
+      "step": 2287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.827860097923529,
+      "learning_rate": 1.4357658944909496e-05,
+      "loss": 0.6199,
+      "step": 2288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5963306997940863,
+      "learning_rate": 1.4352869806786061e-05,
+      "loss": 0.5784,
+      "step": 2289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8411205274630092,
+      "learning_rate": 1.4348079436591982e-05,
+      "loss": 0.733,
+      "step": 2290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5446470975965718,
+      "learning_rate": 1.4343287835683168e-05,
+      "loss": 0.4142,
+      "step": 2291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5163468568809857,
+      "learning_rate": 1.4338495005415869e-05,
+      "loss": 0.5531,
+      "step": 2292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6043896337140406,
+      "learning_rate": 1.4333700947146686e-05,
+      "loss": 0.5343,
+      "step": 2293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6945799138502526,
+      "learning_rate": 1.4328905662232567e-05,
+      "loss": 0.5922,
+      "step": 2294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7188146326949165,
+      "learning_rate": 1.4324109152030807e-05,
+      "loss": 0.6442,
+      "step": 2295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1213596296445023,
+      "learning_rate": 1.4319311417899048e-05,
+      "loss": 0.5746,
+      "step": 2296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6955187614940954,
+      "learning_rate": 1.431451246119528e-05,
+      "loss": 0.6009,
+      "step": 2297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5977323249889828,
+      "learning_rate": 1.4309712283277839e-05,
+      "loss": 0.4416,
+      "step": 2298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.647925428751749,
+      "learning_rate": 1.4304910885505404e-05,
+      "loss": 0.5229,
+      "step": 2299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7875087304803217,
+      "learning_rate": 1.4300108269236997e-05,
+      "loss": 0.6442,
+      "step": 2300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6587971633975462,
+      "learning_rate": 1.429530443583199e-05,
+      "loss": 0.5678,
+      "step": 2301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8953205882794357,
+      "learning_rate": 1.4290499386650099e-05,
+      "loss": 0.4854,
+      "step": 2302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.863300669997199,
+      "learning_rate": 1.4285693123051385e-05,
+      "loss": 0.6849,
+      "step": 2303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5681366342080192,
+      "learning_rate": 1.4280885646396248e-05,
+      "loss": 0.5717,
+      "step": 2304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4786436573664536,
+      "learning_rate": 1.4276076958045436e-05,
+      "loss": 0.5385,
+      "step": 2305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6258164036530085,
+      "learning_rate": 1.4271267059360035e-05,
+      "loss": 0.5416,
+      "step": 2306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3141472704874095,
+      "learning_rate": 1.4266455951701476e-05,
+      "loss": 0.6476,
+      "step": 2307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3536651521722134,
+      "learning_rate": 1.4261643636431539e-05,
+      "loss": 0.5316,
+      "step": 2308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8430962760338988,
+      "learning_rate": 1.4256830114912341e-05,
+      "loss": 0.725,
+      "step": 2309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9282282291720794,
+      "learning_rate": 1.4252015388506328e-05,
+      "loss": 0.5719,
+      "step": 2310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0054863108074517,
+      "learning_rate": 1.4247199458576308e-05,
+      "loss": 0.6295,
+      "step": 2311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8945905997166146,
+      "learning_rate": 1.4242382326485416e-05,
+      "loss": 0.5162,
+      "step": 2312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.580006080598253,
+      "learning_rate": 1.4237563993597133e-05,
+      "loss": 0.6098,
+      "step": 2313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4492563418191526,
+      "learning_rate": 1.4232744461275273e-05,
+      "loss": 0.5386,
+      "step": 2314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.684353156892466,
+      "learning_rate": 1.4227923730884001e-05,
+      "loss": 0.62,
+      "step": 2315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9854025357841643,
+      "learning_rate": 1.4223101803787811e-05,
+      "loss": 0.6218,
+      "step": 2316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.679348043256461,
+      "learning_rate": 1.421827868135154e-05,
+      "loss": 0.5929,
+      "step": 2317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.953542819936109,
+      "learning_rate": 1.4213454364940362e-05,
+      "loss": 0.5822,
+      "step": 2318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.473109282838677,
+      "learning_rate": 1.420862885591979e-05,
+      "loss": 0.6219,
+      "step": 2319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5267231781185215,
+      "learning_rate": 1.4203802155655677e-05,
+      "loss": 0.5029,
+      "step": 2320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5723499355542458,
+      "learning_rate": 1.4198974265514207e-05,
+      "loss": 0.5877,
+      "step": 2321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8173258692377363,
+      "learning_rate": 1.4194145186861902e-05,
+      "loss": 0.4992,
+      "step": 2322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5747063852943795,
+      "learning_rate": 1.4189314921065629e-05,
+      "loss": 0.5836,
+      "step": 2323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.474996772480832,
+      "learning_rate": 1.418448346949258e-05,
+      "loss": 0.5339,
+      "step": 2324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6462857379870819,
+      "learning_rate": 1.417965083351029e-05,
+      "loss": 0.5919,
+      "step": 2325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5128227880710363,
+      "learning_rate": 1.4174817014486622e-05,
+      "loss": 0.4981,
+      "step": 2326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9874105881776587,
+      "learning_rate": 1.4169982013789782e-05,
+      "loss": 0.6085,
+      "step": 2327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4838923542408515,
+      "learning_rate": 1.4165145832788305e-05,
+      "loss": 0.5765,
+      "step": 2328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4937438033282933,
+      "learning_rate": 1.4160308472851065e-05,
+      "loss": 0.5328,
+      "step": 2329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.655962685936461,
+      "learning_rate": 1.4155469935347264e-05,
+      "loss": 0.4935,
+      "step": 2330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.567831251007788,
+      "learning_rate": 1.415063022164644e-05,
+      "loss": 0.4697,
+      "step": 2331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3410922098185676,
+      "learning_rate": 1.4145789333118462e-05,
+      "loss": 0.5466,
+      "step": 2332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8865585314747169,
+      "learning_rate": 1.4140947271133536e-05,
+      "loss": 0.5255,
+      "step": 2333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5165467204233563,
+      "learning_rate": 1.4136104037062197e-05,
+      "loss": 0.5761,
+      "step": 2334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.493552111405945,
+      "learning_rate": 1.4131259632275312e-05,
+      "loss": 0.4416,
+      "step": 2335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6214476939133677,
+      "learning_rate": 1.412641405814408e-05,
+      "loss": 0.5986,
+      "step": 2336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7414073600792945,
+      "learning_rate": 1.412156731604003e-05,
+      "loss": 0.6023,
+      "step": 2337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.722883913725722,
+      "learning_rate": 1.4116719407335022e-05,
+      "loss": 0.6439,
+      "step": 2338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7801715666069706,
+      "learning_rate": 1.4111870333401246e-05,
+      "loss": 0.5393,
+      "step": 2339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8473317428954235,
+      "learning_rate": 1.4107020095611223e-05,
+      "loss": 0.536,
+      "step": 2340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0686968526580887,
+      "learning_rate": 1.4102168695337804e-05,
+      "loss": 0.6472,
+      "step": 2341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.107678179598744,
+      "learning_rate": 1.4097316133954163e-05,
+      "loss": 0.5709,
+      "step": 2342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5595916306124793,
+      "learning_rate": 1.4092462412833811e-05,
+      "loss": 0.6157,
+      "step": 2343
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5233266424936323,
+      "learning_rate": 1.4087607533350585e-05,
+      "loss": 0.5577,
+      "step": 2344
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7845214984589308,
+      "learning_rate": 1.4082751496878644e-05,
+      "loss": 0.4926,
+      "step": 2345
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0058011546073393,
+      "learning_rate": 1.4077894304792481e-05,
+      "loss": 0.5661,
+      "step": 2346
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4510358575475875,
+      "learning_rate": 1.4073035958466916e-05,
+      "loss": 0.6222,
+      "step": 2347
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3499344831525308,
+      "learning_rate": 1.406817645927709e-05,
+      "loss": 0.5235,
+      "step": 2348
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.270580624172594,
+      "learning_rate": 1.4063315808598477e-05,
+      "loss": 0.557,
+      "step": 2349
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6460132593657593,
+      "learning_rate": 1.4058454007806874e-05,
+      "loss": 0.5998,
+      "step": 2350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8273661083453222,
+      "learning_rate": 1.4053591058278402e-05,
+      "loss": 0.6833,
+      "step": 2351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.91977929660956,
+      "learning_rate": 1.4048726961389508e-05,
+      "loss": 0.5562,
+      "step": 2352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7460645154621746,
+      "learning_rate": 1.4043861718516964e-05,
+      "loss": 0.6542,
+      "step": 2353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.543341658684004,
+      "learning_rate": 1.403899533103787e-05,
+      "loss": 0.5344,
+      "step": 2354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.609425695228496,
+      "learning_rate": 1.4034127800329645e-05,
+      "loss": 0.5051,
+      "step": 2355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7044568288552995,
+      "learning_rate": 1.4029259127770032e-05,
+      "loss": 0.5343,
+      "step": 2356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.902699248574413,
+      "learning_rate": 1.40243893147371e-05,
+      "loss": 0.4956,
+      "step": 2357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4221856127521553,
+      "learning_rate": 1.4019518362609239e-05,
+      "loss": 0.5455,
+      "step": 2358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3797306745628037,
+      "learning_rate": 1.4014646272765162e-05,
+      "loss": 0.5633,
+      "step": 2359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6106690197265419,
+      "learning_rate": 1.4009773046583904e-05,
+      "loss": 0.5362,
+      "step": 2360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.802967922016013,
+      "learning_rate": 1.4004898685444819e-05,
+      "loss": 0.7313,
+      "step": 2361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9774950460248586,
+      "learning_rate": 1.4000023190727587e-05,
+      "loss": 0.5533,
+      "step": 2362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9549694536840034,
+      "learning_rate": 1.399514656381221e-05,
+      "loss": 0.5729,
+      "step": 2363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.43572025116169,
+      "learning_rate": 1.3990268806078999e-05,
+      "loss": 0.5873,
+      "step": 2364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6671114703715733,
+      "learning_rate": 1.39853899189086e-05,
+      "loss": 0.5895,
+      "step": 2365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.583209970208007,
+      "learning_rate": 1.3980509903681968e-05,
+      "loss": 0.5499,
+      "step": 2366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5407299508197116,
+      "learning_rate": 1.397562876178038e-05,
+      "loss": 0.5473,
+      "step": 2367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9050696316890225,
+      "learning_rate": 1.3970746494585439e-05,
+      "loss": 0.4771,
+      "step": 2368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.885785081233274,
+      "learning_rate": 1.3965863103479054e-05,
+      "loss": 0.622,
+      "step": 2369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.012303548332027,
+      "learning_rate": 1.3960978589843458e-05,
+      "loss": 0.5491,
+      "step": 2370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9312964008819045,
+      "learning_rate": 1.3956092955061208e-05,
+      "loss": 0.5406,
+      "step": 2371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9940812052397636,
+      "learning_rate": 1.3951206200515173e-05,
+      "loss": 0.5834,
+      "step": 2372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4109357074229683,
+      "learning_rate": 1.3946318327588534e-05,
+      "loss": 0.5095,
+      "step": 2373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8305470900772817,
+      "learning_rate": 1.3941429337664791e-05,
+      "loss": 0.6849,
+      "step": 2374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8363469062290076,
+      "learning_rate": 1.3936539232127771e-05,
+      "loss": 0.71,
+      "step": 2375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.020706391280267,
+      "learning_rate": 1.3931648012361599e-05,
+      "loss": 0.5855,
+      "step": 2376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8962565072740944,
+      "learning_rate": 1.392675567975073e-05,
+      "loss": 0.4523,
+      "step": 2377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.863028581789806,
+      "learning_rate": 1.3921862235679929e-05,
+      "loss": 0.5374,
+      "step": 2378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5483698999857622,
+      "learning_rate": 1.391696768153427e-05,
+      "loss": 0.5237,
+      "step": 2379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.554392388912573,
+      "learning_rate": 1.3912072018699152e-05,
+      "loss": 0.4764,
+      "step": 2380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4900042816419363,
+      "learning_rate": 1.3907175248560276e-05,
+      "loss": 0.5596,
+      "step": 2381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9362325778623993,
+      "learning_rate": 1.390227737250367e-05,
+      "loss": 0.7216,
+      "step": 2382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9599746119825794,
+      "learning_rate": 1.389737839191566e-05,
+      "loss": 0.5404,
+      "step": 2383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7358948176711475,
+      "learning_rate": 1.38924783081829e-05,
+      "loss": 0.607,
+      "step": 2384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8946819381506759,
+      "learning_rate": 1.3887577122692337e-05,
+      "loss": 0.5673,
+      "step": 2385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7010413814192178,
+      "learning_rate": 1.3882674836831251e-05,
+      "loss": 0.5812,
+      "step": 2386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6579676040077507,
+      "learning_rate": 1.3877771451987223e-05,
+      "loss": 0.5448,
+      "step": 2387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.626669200444195,
+      "learning_rate": 1.3872866969548143e-05,
+      "loss": 0.568,
+      "step": 2388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.605551644449508,
+      "learning_rate": 1.3867961390902214e-05,
+      "loss": 0.5553,
+      "step": 2389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.485014382870232,
+      "learning_rate": 1.3863054717437952e-05,
+      "loss": 0.476,
+      "step": 2390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.765612740608783,
+      "learning_rate": 1.3858146950544178e-05,
+      "loss": 0.5468,
+      "step": 2391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6782285414106906,
+      "learning_rate": 1.3853238091610029e-05,
+      "loss": 0.5882,
+      "step": 2392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.75416577190845,
+      "learning_rate": 1.384832814202494e-05,
+      "loss": 0.5113,
+      "step": 2393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7187260830773547,
+      "learning_rate": 1.3843417103178669e-05,
+      "loss": 0.5295,
+      "step": 2394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0049110936153043,
+      "learning_rate": 1.3838504976461278e-05,
+      "loss": 0.6555,
+      "step": 2395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.972611485191297,
+      "learning_rate": 1.3833591763263123e-05,
+      "loss": 0.6404,
+      "step": 2396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8042083182245087,
+      "learning_rate": 1.3828677464974885e-05,
+      "loss": 0.6592,
+      "step": 2397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.597912465501303,
+      "learning_rate": 1.3823762082987544e-05,
+      "loss": 0.4748,
+      "step": 2398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7536121164517375,
+      "learning_rate": 1.381884561869239e-05,
+      "loss": 0.5967,
+      "step": 2399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9817667653065087,
+      "learning_rate": 1.3813928073481023e-05,
+      "loss": 0.555,
+      "step": 2400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0011274802173094,
+      "learning_rate": 1.3809009448745334e-05,
+      "loss": 0.6486,
+      "step": 2401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7876621510051626,
+      "learning_rate": 1.3804089745877536e-05,
+      "loss": 0.7013,
+      "step": 2402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8740464438472715,
+      "learning_rate": 1.3799168966270139e-05,
+      "loss": 0.7402,
+      "step": 2403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8179034554307272,
+      "learning_rate": 1.3794247111315955e-05,
+      "loss": 0.6958,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4164907238318103,
+      "learning_rate": 1.3789324182408112e-05,
+      "loss": 0.5806,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6749230448466432,
+      "learning_rate": 1.3784400180940032e-05,
+      "loss": 0.5642,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5205632658848995,
+      "learning_rate": 1.377947510830544e-05,
+      "loss": 0.5634,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4116619180114243,
+      "learning_rate": 1.3774548965898371e-05,
+      "loss": 0.5485,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4140500536840441,
+      "learning_rate": 1.3769621755113156e-05,
+      "loss": 0.5025,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9383204902845078,
+      "learning_rate": 1.3764693477344435e-05,
+      "loss": 0.7412,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.249270461904804,
+      "learning_rate": 1.3759764133987146e-05,
+      "loss": 0.5712,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4205877671091083,
+      "learning_rate": 1.375483372643653e-05,
+      "loss": 0.5226,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6550108213322396,
+      "learning_rate": 1.3749902256088125e-05,
+      "loss": 0.5714,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0598565943733167,
+      "learning_rate": 1.3744969724337779e-05,
+      "loss": 0.6791,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8537270397186887,
+      "learning_rate": 1.3740036132581626e-05,
+      "loss": 0.56,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6748994965128352,
+      "learning_rate": 1.3735101482216117e-05,
+      "loss": 0.5354,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5382257670511725,
+      "learning_rate": 1.3730165774637994e-05,
+      "loss": 0.5085,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9908320548047724,
+      "learning_rate": 1.3725229011244294e-05,
+      "loss": 0.5173,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5441933182103056,
+      "learning_rate": 1.3720291193432357e-05,
+      "loss": 0.5062,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5953849044566082,
+      "learning_rate": 1.3715352322599826e-05,
+      "loss": 0.5816,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4979289697338602,
+      "learning_rate": 1.3710412400144637e-05,
+      "loss": 0.5471,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8240536133168803,
+      "learning_rate": 1.3705471427465025e-05,
+      "loss": 0.7329,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.805383232397561,
+      "learning_rate": 1.3700529405959517e-05,
+      "loss": 0.5531,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8293195392217272,
+      "learning_rate": 1.3695586337026949e-05,
+      "loss": 0.6585,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9130031098906937,
+      "learning_rate": 1.3690642222066445e-05,
+      "loss": 0.4726,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6451887401228258,
+      "learning_rate": 1.3685697062477421e-05,
+      "loss": 0.5628,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7765781096733486,
+      "learning_rate": 1.3680750859659599e-05,
+      "loss": 0.4936,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5469996582073577,
+      "learning_rate": 1.3675803615012993e-05,
+      "loss": 0.5134,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.820455639031437,
+      "learning_rate": 1.3670855329937905e-05,
+      "loss": 0.7006,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0907593878411537,
+      "learning_rate": 1.3665906005834938e-05,
+      "loss": 0.5648,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2853267568007682,
+      "learning_rate": 1.3660955644104985e-05,
+      "loss": 0.542,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3531764790960554,
+      "learning_rate": 1.3656004246149242e-05,
+      "loss": 0.4059,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6515378132651108,
+      "learning_rate": 1.3651051813369188e-05,
+      "loss": 0.5479,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.027715060503047,
+      "learning_rate": 1.3646098347166598e-05,
+      "loss": 0.5089,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7528300753648328,
+      "learning_rate": 1.3641143848943545e-05,
+      "loss": 0.6849,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8978436196735533,
+      "learning_rate": 1.3636188320102384e-05,
+      "loss": 0.638,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.667659883430721,
+      "learning_rate": 1.3631231762045768e-05,
+      "loss": 0.5711,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8610179271764153,
+      "learning_rate": 1.3626274176176645e-05,
+      "loss": 0.5235,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.288625876299832,
+      "learning_rate": 1.3621315563898243e-05,
+      "loss": 0.5148,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.296933057542676,
+      "learning_rate": 1.3616355926614089e-05,
+      "loss": 0.5278,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4328748660418182,
+      "learning_rate": 1.3611395265727998e-05,
+      "loss": 0.5439,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4368606823730234,
+      "learning_rate": 1.3606433582644077e-05,
+      "loss": 0.5359,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5887728915578228,
+      "learning_rate": 1.3601470878766714e-05,
+      "loss": 0.4663,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2994782564708514,
+      "learning_rate": 1.3596507155500596e-05,
+      "loss": 0.5901,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.790209767393109,
+      "learning_rate": 1.3591542414250694e-05,
+      "loss": 0.6819,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3877946140380186,
+      "learning_rate": 1.3586576656422268e-05,
+      "loss": 0.5012,
+      "step": 2446
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2281541728382668,
+      "learning_rate": 1.3581609883420866e-05,
+      "loss": 0.5172,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8226134562061059,
+      "learning_rate": 1.3576642096652322e-05,
+      "loss": 0.668,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1373420281477724,
+      "learning_rate": 1.3571673297522759e-05,
+      "loss": 0.5973,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.62915984660785,
+      "learning_rate": 1.3566703487438579e-05,
+      "loss": 0.4896,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9274871546098227,
+      "learning_rate": 1.3561732667806481e-05,
+      "loss": 0.5193,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7925397580260025,
+      "learning_rate": 1.3556760840033447e-05,
+      "loss": 0.54,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9190547001787102,
+      "learning_rate": 1.3551788005526738e-05,
+      "loss": 0.6489,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5945690331307258,
+      "learning_rate": 1.3546814165693909e-05,
+      "loss": 0.5267,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7840678380960804,
+      "learning_rate": 1.3541839321942786e-05,
+      "loss": 0.5565,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7788735637329602,
+      "learning_rate": 1.35368634756815e-05,
+      "loss": 0.5561,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.760494251069017,
+      "learning_rate": 1.3531886628318447e-05,
+      "loss": 0.51,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6001462049560418,
+      "learning_rate": 1.3526908781262314e-05,
+      "loss": 0.5703,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7764429172353595,
+      "learning_rate": 1.352192993592207e-05,
+      "loss": 0.5296,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5626437980070171,
+      "learning_rate": 1.3516950093706968e-05,
+      "loss": 0.5022,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3569951633234214,
+      "learning_rate": 1.3511969256026542e-05,
+      "loss": 0.5024,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8109575843609884,
+      "learning_rate": 1.3506987424290605e-05,
+      "loss": 0.5732,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8186130948025077,
+      "learning_rate": 1.3502004599909255e-05,
+      "loss": 0.5572,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.919087171978144,
+      "learning_rate": 1.349702078429287e-05,
+      "loss": 0.4902,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0221905823774535,
+      "learning_rate": 1.3492035978852114e-05,
+      "loss": 0.5317,
+      "step": 2465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8822756648177752,
+      "learning_rate": 1.3487050184997916e-05,
+      "loss": 0.5654,
+      "step": 2466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6867315273507297,
+      "learning_rate": 1.3482063404141496e-05,
+      "loss": 0.5518,
+      "step": 2467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6256346297655586,
+      "learning_rate": 1.3477075637694362e-05,
+      "loss": 0.5984,
+      "step": 2468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.028275341113145,
+      "learning_rate": 1.347208688706828e-05,
+      "loss": 0.623,
+      "step": 2469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7302174449135732,
+      "learning_rate": 1.3467097153675313e-05,
+      "loss": 0.5689,
+      "step": 2470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.62354805534783,
+      "learning_rate": 1.3462106438927788e-05,
+      "loss": 0.5979,
+      "step": 2471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4458584636872285,
+      "learning_rate": 1.345711474423832e-05,
+      "loss": 0.4532,
+      "step": 2472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.985520721498813,
+      "learning_rate": 1.3452122071019797e-05,
+      "loss": 0.5494,
+      "step": 2473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8408824700453112,
+      "learning_rate": 1.3447128420685385e-05,
+      "loss": 0.6967,
+      "step": 2474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.730461478252986,
+      "learning_rate": 1.3442133794648521e-05,
+      "loss": 0.6323,
+      "step": 2475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7554538936902726,
+      "learning_rate": 1.3437138194322934e-05,
+      "loss": 0.6024,
+      "step": 2476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5720989692447034,
+      "learning_rate": 1.3432141621122608e-05,
+      "loss": 0.473,
+      "step": 2477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6797556966996965,
+      "learning_rate": 1.3427144076461818e-05,
+      "loss": 0.6086,
+      "step": 2478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9143840097221156,
+      "learning_rate": 1.3422145561755106e-05,
+      "loss": 0.5635,
+      "step": 2479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4560308620777083,
+      "learning_rate": 1.3417146078417294e-05,
+      "loss": 0.5423,
+      "step": 2480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1882489757236248,
+      "learning_rate": 1.3412145627863473e-05,
+      "loss": 0.6079,
+      "step": 2481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6181663342596653,
+      "learning_rate": 1.3407144211509014e-05,
+      "loss": 0.5402,
+      "step": 2482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5812217172693166,
+      "learning_rate": 1.3402141830769551e-05,
+      "loss": 0.6832,
+      "step": 2483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6769665312794215,
+      "learning_rate": 1.3397138487060999e-05,
+      "loss": 0.567,
+      "step": 2484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0082723461943073,
+      "learning_rate": 1.3392134181799547e-05,
+      "loss": 0.6065,
+      "step": 2485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.643472695932624,
+      "learning_rate": 1.338712891640165e-05,
+      "loss": 0.4836,
+      "step": 2486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8365973356333494,
+      "learning_rate": 1.3382122692284041e-05,
+      "loss": 0.5338,
+      "step": 2487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.119336211864276,
+      "learning_rate": 1.3377115510863716e-05,
+      "loss": 0.5629,
+      "step": 2488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.361628591840574,
+      "learning_rate": 1.3372107373557955e-05,
+      "loss": 0.557,
+      "step": 2489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5511155366068166,
+      "learning_rate": 1.336709828178429e-05,
+      "loss": 0.5483,
+      "step": 2490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8784419163558936,
+      "learning_rate": 1.3362088236960544e-05,
+      "loss": 0.5513,
+      "step": 2491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7570845418563164,
+      "learning_rate": 1.3357077240504795e-05,
+      "loss": 0.5508,
+      "step": 2492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.29768528633364,
+      "learning_rate": 1.3352065293835399e-05,
+      "loss": 0.6048,
+      "step": 2493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8987996490188348,
+      "learning_rate": 1.3347052398370969e-05,
+      "loss": 0.5844,
+      "step": 2494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7560956656853444,
+      "learning_rate": 1.3342038555530403e-05,
+      "loss": 0.5294,
+      "step": 2495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.589005562983234,
+      "learning_rate": 1.3337023766732852e-05,
+      "loss": 0.4956,
+      "step": 2496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9648016224031002,
+      "learning_rate": 1.3332008033397746e-05,
+      "loss": 0.5223,
+      "step": 2497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8368846394328258,
+      "learning_rate": 1.3326991356944776e-05,
+      "loss": 0.5616,
+      "step": 2498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6712748936988757,
+      "learning_rate": 1.33219737387939e-05,
+      "loss": 0.5395,
+      "step": 2499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7988016165904281,
+      "learning_rate": 1.3316955180365348e-05,
+      "loss": 0.5422,
+      "step": 2500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.602877881483966,
+      "learning_rate": 1.3311935683079611e-05,
+      "loss": 0.4825,
+      "step": 2501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0590548142994147,
+      "learning_rate": 1.3306915248357442e-05,
+      "loss": 0.5969,
+      "step": 2502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.700899346619194,
+      "learning_rate": 1.3301893877619874e-05,
+      "loss": 0.4407,
+      "step": 2503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6843221408158389,
+      "learning_rate": 1.3296871572288187e-05,
+      "loss": 0.5396,
+      "step": 2504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.6337852749396315,
+      "learning_rate": 1.3291848333783941e-05,
+      "loss": 0.5465,
+      "step": 2505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7264553214415639,
+      "learning_rate": 1.3286824163528948e-05,
+      "loss": 0.4085,
+      "step": 2506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.259492898398366,
+      "learning_rate": 1.3281799062945291e-05,
+      "loss": 0.523,
+      "step": 2507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6201927536228118,
+      "learning_rate": 1.3276773033455312e-05,
+      "loss": 0.5449,
+      "step": 2508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4850284455065457,
+      "learning_rate": 1.3271746076481621e-05,
+      "loss": 0.5319,
+      "step": 2509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4755127119043414,
+      "learning_rate": 1.3266718193447086e-05,
+      "loss": 0.6382,
+      "step": 2510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7305413252612087,
+      "learning_rate": 1.3261689385774839e-05,
+      "loss": 0.5864,
+      "step": 2511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3569161208287512,
+      "learning_rate": 1.3256659654888272e-05,
+      "loss": 0.4957,
+      "step": 2512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.695598288454166,
+      "learning_rate": 1.3251629002211042e-05,
+      "loss": 0.6617,
+      "step": 2513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5256932530354663,
+      "learning_rate": 1.3246597429167066e-05,
+      "loss": 0.5569,
+      "step": 2514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3657427789969914,
+      "learning_rate": 1.3241564937180513e-05,
+      "loss": 0.5365,
+      "step": 2515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5870533422480195,
+      "learning_rate": 1.3236531527675828e-05,
+      "loss": 0.6308,
+      "step": 2516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9041362481735447,
+      "learning_rate": 1.3231497202077701e-05,
+      "loss": 0.6157,
+      "step": 2517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8582489681964298,
+      "learning_rate": 1.322646196181109e-05,
+      "loss": 0.6319,
+      "step": 2518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9844304725003217,
+      "learning_rate": 1.3221425808301209e-05,
+      "loss": 0.5991,
+      "step": 2519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.861865393857314,
+      "learning_rate": 1.3216388742973532e-05,
+      "loss": 0.5271,
+      "step": 2520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9715892681113942,
+      "learning_rate": 1.3211350767253786e-05,
+      "loss": 0.6376,
+      "step": 2521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1999602875366002,
+      "learning_rate": 1.320631188256796e-05,
+      "loss": 0.5991,
+      "step": 2522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.536229177331727,
+      "learning_rate": 1.3201272090342303e-05,
+      "loss": 0.5206,
+      "step": 2523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.354772567814958,
+      "learning_rate": 1.3196231392003316e-05,
+      "loss": 0.7183,
+      "step": 2524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.591696672006161,
+      "learning_rate": 1.3191189788977758e-05,
+      "loss": 0.4691,
+      "step": 2525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.217355800273756,
+      "learning_rate": 1.3186147282692643e-05,
+      "loss": 0.6201,
+      "step": 2526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.317023850534203,
+      "learning_rate": 1.3181103874575243e-05,
+      "loss": 0.5283,
+      "step": 2527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4352265668991595,
+      "learning_rate": 1.3176059566053083e-05,
+      "loss": 0.5258,
+      "step": 2528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9676998450075778,
+      "learning_rate": 1.3171014358553946e-05,
+      "loss": 0.5958,
+      "step": 2529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.673202384963764,
+      "learning_rate": 1.3165968253505865e-05,
+      "loss": 0.6927,
+      "step": 2530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8180516199375515,
+      "learning_rate": 1.3160921252337131e-05,
+      "loss": 0.7287,
+      "step": 2531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.663579801422374,
+      "learning_rate": 1.3155873356476287e-05,
+      "loss": 0.5072,
+      "step": 2532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4329928622991708,
+      "learning_rate": 1.3150824567352128e-05,
+      "loss": 0.5207,
+      "step": 2533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.952709593391107,
+      "learning_rate": 1.3145774886393704e-05,
+      "loss": 0.5097,
+      "step": 2534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4319261106607617,
+      "learning_rate": 1.3140724315030315e-05,
+      "loss": 0.5381,
+      "step": 2535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.417611697395353,
+      "learning_rate": 1.313567285469152e-05,
+      "loss": 0.6141,
+      "step": 2536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4224904789791553,
+      "learning_rate": 1.3130620506807116e-05,
+      "loss": 0.5545,
+      "step": 2537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7998864288274803,
+      "learning_rate": 1.3125567272807167e-05,
+      "loss": 0.6591,
+      "step": 2538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.383567164288977,
+      "learning_rate": 1.3120513154121976e-05,
+      "loss": 0.6732,
+      "step": 2539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.501787841705014,
+      "learning_rate": 1.3115458152182102e-05,
+      "loss": 0.5361,
+      "step": 2540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5642490838906036,
+      "learning_rate": 1.3110402268418352e-05,
+      "loss": 0.4526,
+      "step": 2541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4126304070038085,
+      "learning_rate": 1.3105345504261781e-05,
+      "loss": 0.5356,
+      "step": 2542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7151363588059214,
+      "learning_rate": 1.3100287861143703e-05,
+      "loss": 0.5579,
+      "step": 2543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6811819649937392,
+      "learning_rate": 1.3095229340495665e-05,
+      "loss": 0.5964,
+      "step": 2544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6652441337637616,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.5238,
+      "step": 2545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.079245358204068,
+      "learning_rate": 1.3085109672337183e-05,
+      "loss": 0.6203,
+      "step": 2546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.405786049905545,
+      "learning_rate": 1.3080048527691092e-05,
+      "loss": 0.5649,
+      "step": 2547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5811789700559573,
+      "learning_rate": 1.3074986511243741e-05,
+      "loss": 0.6154,
+      "step": 2548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.98804396654207,
+      "learning_rate": 1.306992362442793e-05,
+      "loss": 0.5213,
+      "step": 2549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3514126830564686,
+      "learning_rate": 1.3064859868676694e-05,
+      "loss": 0.6176,
+      "step": 2550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5931276637143892,
+      "learning_rate": 1.3059795245423319e-05,
+      "loss": 0.5569,
+      "step": 2551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7161502083499085,
+      "learning_rate": 1.3054729756101338e-05,
+      "loss": 0.5072,
+      "step": 2552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4233358627584516,
+      "learning_rate": 1.3049663402144528e-05,
+      "loss": 0.5854,
+      "step": 2553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.783822071940793,
+      "learning_rate": 1.3044596184986906e-05,
+      "loss": 0.5264,
+      "step": 2554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4549069735992912,
+      "learning_rate": 1.303952810606274e-05,
+      "loss": 0.6404,
+      "step": 2555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5274102381373813,
+      "learning_rate": 1.3034459166806537e-05,
+      "loss": 0.6691,
+      "step": 2556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7566595083042733,
+      "learning_rate": 1.3029389368653051e-05,
+      "loss": 0.5267,
+      "step": 2557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4557039022088882,
+      "learning_rate": 1.302431871303728e-05,
+      "loss": 0.5251,
+      "step": 2558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5337194649493044,
+      "learning_rate": 1.3019247201394456e-05,
+      "loss": 0.5462,
+      "step": 2559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7661386704117712,
+      "learning_rate": 1.3014174835160065e-05,
+      "loss": 0.7034,
+      "step": 2560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9023809531763178,
+      "learning_rate": 1.300910161576983e-05,
+      "loss": 0.6294,
+      "step": 2561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5854672420521414,
+      "learning_rate": 1.3004027544659712e-05,
+      "loss": 0.4972,
+      "step": 2562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6610788093309776,
+      "learning_rate": 1.2998952623265917e-05,
+      "loss": 0.514,
+      "step": 2563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.270504815566349,
+      "learning_rate": 1.2993876853024891e-05,
+      "loss": 0.5823,
+      "step": 2564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.658209252005798,
+      "learning_rate": 1.298880023537332e-05,
+      "loss": 0.5277,
+      "step": 2565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.84991153409683,
+      "learning_rate": 1.2983722771748131e-05,
+      "loss": 0.49,
+      "step": 2566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.144597930212238,
+      "learning_rate": 1.2978644463586489e-05,
+      "loss": 0.5289,
+      "step": 2567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.456337131314162,
+      "learning_rate": 1.2973565312325798e-05,
+      "loss": 0.4732,
+      "step": 2568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9097726985370966,
+      "learning_rate": 1.29684853194037e-05,
+      "loss": 0.5669,
+      "step": 2569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7549178085406665,
+      "learning_rate": 1.296340448625808e-05,
+      "loss": 0.4903,
+      "step": 2570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6068685732849566,
+      "learning_rate": 1.2958322814327053e-05,
+      "loss": 0.5754,
+      "step": 2571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3862930224969592,
+      "learning_rate": 1.2953240305048978e-05,
+      "loss": 0.5606,
+      "step": 2572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.629882954583122,
+      "learning_rate": 1.2948156959862446e-05,
+      "loss": 0.5435,
+      "step": 2573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8058405365946169,
+      "learning_rate": 1.294307278020629e-05,
+      "loss": 0.571,
+      "step": 2574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3067711698572655,
+      "learning_rate": 1.2937987767519576e-05,
+      "loss": 0.4953,
+      "step": 2575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7102905166129976,
+      "learning_rate": 1.2932901923241603e-05,
+      "loss": 0.5988,
+      "step": 2576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.374083314974588,
+      "learning_rate": 1.2927815248811913e-05,
+      "loss": 0.5362,
+      "step": 2577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3728815149944227,
+      "learning_rate": 1.2922727745670276e-05,
+      "loss": 0.5858,
+      "step": 2578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0203377417519324,
+      "learning_rate": 1.29176394152567e-05,
+      "loss": 0.6067,
+      "step": 2579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6829199089221762,
+      "learning_rate": 1.2912550259011422e-05,
+      "loss": 0.5917,
+      "step": 2580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.922949802559928,
+      "learning_rate": 1.2907460278374925e-05,
+      "loss": 0.5253,
+      "step": 2581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.585002154848367,
+      "learning_rate": 1.2902369474787912e-05,
+      "loss": 0.489,
+      "step": 2582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8520908860994476,
+      "learning_rate": 1.2897277849691326e-05,
+      "loss": 0.5551,
+      "step": 2583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0917109214283593,
+      "learning_rate": 1.2892185404526338e-05,
+      "loss": 0.4616,
+      "step": 2584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3338908004261287,
+      "learning_rate": 1.2887092140734357e-05,
+      "loss": 0.6175,
+      "step": 2585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5375200851287538,
+      "learning_rate": 1.288199805975702e-05,
+      "loss": 0.5603,
+      "step": 2586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.501321680842173,
+      "learning_rate": 1.2876903163036194e-05,
+      "loss": 0.5375,
+      "step": 2587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5717544932047696,
+      "learning_rate": 1.2871807452013977e-05,
+      "loss": 0.5177,
+      "step": 2588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6815935844195125,
+      "learning_rate": 1.2866710928132709e-05,
+      "loss": 0.6044,
+      "step": 2589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0695900676554397,
+      "learning_rate": 1.2861613592834942e-05,
+      "loss": 0.5664,
+      "step": 2590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5431023825349823,
+      "learning_rate": 1.2856515447563467e-05,
+      "loss": 0.5188,
+      "step": 2591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.655954571024143,
+      "learning_rate": 1.2851416493761301e-05,
+      "loss": 0.6964,
+      "step": 2592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7975993761398732,
+      "learning_rate": 1.28463167328717e-05,
+      "loss": 0.6111,
+      "step": 2593
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.595291372295204,
+      "learning_rate": 1.2841216166338133e-05,
+      "loss": 0.5477,
+      "step": 2594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7372798884115173,
+      "learning_rate": 1.2836114795604309e-05,
+      "loss": 0.5242,
+      "step": 2595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5949923560368215,
+      "learning_rate": 1.2831012622114159e-05,
+      "loss": 0.6333,
+      "step": 2596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6103203739945047,
+      "learning_rate": 1.282590964731184e-05,
+      "loss": 0.6151,
+      "step": 2597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6222222374424053,
+      "learning_rate": 1.2820805872641745e-05,
+      "loss": 0.6121,
+      "step": 2598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7174563009735015,
+      "learning_rate": 1.2815701299548478e-05,
+      "loss": 0.5785,
+      "step": 2599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8841236968106665,
+      "learning_rate": 1.2810595929476884e-05,
+      "loss": 0.7105,
+      "step": 2600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6685333465558738,
+      "learning_rate": 1.2805489763872026e-05,
+      "loss": 0.4726,
+      "step": 2601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6193405011080142,
+      "learning_rate": 1.280038280417919e-05,
+      "loss": 0.4521,
+      "step": 2602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7944019033371226,
+      "learning_rate": 1.2795275051843893e-05,
+      "loss": 0.5714,
+      "step": 2603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5239672033856528,
+      "learning_rate": 1.2790166508311872e-05,
+      "loss": 0.4874,
+      "step": 2604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9050896548071201,
+      "learning_rate": 1.2785057175029092e-05,
+      "loss": 0.573,
+      "step": 2605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5470424020469826,
+      "learning_rate": 1.277994705344174e-05,
+      "loss": 0.6345,
+      "step": 2606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6133151546636268,
+      "learning_rate": 1.2774836144996222e-05,
+      "loss": 0.5392,
+      "step": 2607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9358036899379611,
+      "learning_rate": 1.276972445113917e-05,
+      "loss": 0.5887,
+      "step": 2608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5884527558523172,
+      "learning_rate": 1.276461197331744e-05,
+      "loss": 0.5447,
+      "step": 2609
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5818560146462801,
+      "learning_rate": 1.2759498712978106e-05,
+      "loss": 0.6323,
+      "step": 2610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8653257387541085,
+      "learning_rate": 1.2754384671568469e-05,
+      "loss": 0.7065,
+      "step": 2611
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6763691529518283,
+      "learning_rate": 1.2749269850536045e-05,
+      "loss": 0.5007,
+      "step": 2612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8220973231545963,
+      "learning_rate": 1.2744154251328573e-05,
+      "loss": 0.5581,
+      "step": 2613
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.812942667724676,
+      "learning_rate": 1.2739037875394013e-05,
+      "loss": 0.5972,
+      "step": 2614
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.040596349978365,
+      "learning_rate": 1.2733920724180542e-05,
+      "loss": 0.4823,
+      "step": 2615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5634337093641815,
+      "learning_rate": 1.2728802799136566e-05,
+      "loss": 0.4583,
+      "step": 2616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.646152690337211,
+      "learning_rate": 1.2723684101710696e-05,
+      "loss": 0.5847,
+      "step": 2617
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.730128129158115,
+      "learning_rate": 1.2718564633351773e-05,
+      "loss": 0.5924,
+      "step": 2618
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5727337419982275,
+      "learning_rate": 1.271344439550885e-05,
+      "loss": 0.5875,
+      "step": 2619
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8072540961351191,
+      "learning_rate": 1.2708323389631198e-05,
+      "loss": 0.5423,
+      "step": 2620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9883058926269008,
+      "learning_rate": 1.270320161716831e-05,
+      "loss": 0.6286,
+      "step": 2621
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5136707861451872,
+      "learning_rate": 1.2698079079569891e-05,
+      "loss": 0.5486,
+      "step": 2622
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8610008325942655,
+      "learning_rate": 1.2692955778285865e-05,
+      "loss": 0.6216,
+      "step": 2623
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8039352264316069,
+      "learning_rate": 1.268783171476637e-05,
+      "loss": 0.5853,
+      "step": 2624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8352506293659072,
+      "learning_rate": 1.2682706890461764e-05,
+      "loss": 0.7001,
+      "step": 2625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5305655410199857,
+      "learning_rate": 1.2677581306822613e-05,
+      "loss": 0.5375,
+      "step": 2626
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6006424001677027,
+      "learning_rate": 1.267245496529971e-05,
+      "loss": 0.6171,
+      "step": 2627
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7537569776943933,
+      "learning_rate": 1.266732786734405e-05,
+      "loss": 0.4736,
+      "step": 2628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8636329279693344,
+      "learning_rate": 1.2662200014406848e-05,
+      "loss": 0.5477,
+      "step": 2629
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6268654761822339,
+      "learning_rate": 1.2657071407939536e-05,
+      "loss": 0.4756,
+      "step": 2630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0956482826744796,
+      "learning_rate": 1.265194204939375e-05,
+      "loss": 0.6167,
+      "step": 2631
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.290706609401863,
+      "learning_rate": 1.2646811940221346e-05,
+      "loss": 0.44,
+      "step": 2632
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6380211871992496,
+      "learning_rate": 1.2641681081874394e-05,
+      "loss": 0.5592,
+      "step": 2633
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9178544524291055,
+      "learning_rate": 1.2636549475805165e-05,
+      "loss": 0.5771,
+      "step": 2634
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.565488759669598,
+      "learning_rate": 1.2631417123466154e-05,
+      "loss": 0.5723,
+      "step": 2635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7601562429874544,
+      "learning_rate": 1.2626284026310062e-05,
+      "loss": 0.6078,
+      "step": 2636
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.673645731969357,
+      "learning_rate": 1.2621150185789803e-05,
+      "loss": 0.6036,
+      "step": 2637
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8359689585981662,
+      "learning_rate": 1.2616015603358497e-05,
+      "loss": 0.6942,
+      "step": 2638
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.666001942294789,
+      "learning_rate": 1.261088028046948e-05,
+      "loss": 0.5445,
+      "step": 2639
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1133794266153347,
+      "learning_rate": 1.260574421857629e-05,
+      "loss": 0.5207,
+      "step": 2640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4784349630614841,
+      "learning_rate": 1.2600607419132685e-05,
+      "loss": 0.5636,
+      "step": 2641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6299696782245308,
+      "learning_rate": 1.2595469883592617e-05,
+      "loss": 0.6379,
+      "step": 2642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6386666058088937,
+      "learning_rate": 1.2590331613410261e-05,
+      "loss": 0.6556,
+      "step": 2643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5336982966114747,
+      "learning_rate": 1.258519261003999e-05,
+      "loss": 0.4655,
+      "step": 2644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5977002741371815,
+      "learning_rate": 1.2580052874936393e-05,
+      "loss": 0.54,
+      "step": 2645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.962639612365976,
+      "learning_rate": 1.2574912409554254e-05,
+      "loss": 0.5376,
+      "step": 2646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9083669928661595,
+      "learning_rate": 1.2569771215348576e-05,
+      "loss": 0.5945,
+      "step": 2647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.319076480943554,
+      "learning_rate": 1.2564629293774561e-05,
+      "loss": 0.6105,
+      "step": 2648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8552117767811858,
+      "learning_rate": 1.2559486646287622e-05,
+      "loss": 0.5734,
+      "step": 2649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.609139129056297,
+      "learning_rate": 1.2554343274343367e-05,
+      "loss": 0.5201,
+      "step": 2650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7404396437617273,
+      "learning_rate": 1.2549199179397627e-05,
+      "loss": 0.5125,
+      "step": 2651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4310935452242934,
+      "learning_rate": 1.2544054362906421e-05,
+      "loss": 0.5435,
+      "step": 2652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9320928325835407,
+      "learning_rate": 1.253890882632598e-05,
+      "loss": 0.5139,
+      "step": 2653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6487958915714982,
+      "learning_rate": 1.2533762571112737e-05,
+      "loss": 0.6455,
+      "step": 2654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8387407408023875,
+      "learning_rate": 1.2528615598723333e-05,
+      "loss": 0.5024,
+      "step": 2655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.572817974320171,
+      "learning_rate": 1.2523467910614597e-05,
+      "loss": 0.5906,
+      "step": 2656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.666681159832071,
+      "learning_rate": 1.2518319508243582e-05,
+      "loss": 0.5129,
+      "step": 2657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5253036767730856,
+      "learning_rate": 1.2513170393067527e-05,
+      "loss": 0.59,
+      "step": 2658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.629178580936063,
+      "learning_rate": 1.2508020566543876e-05,
+      "loss": 0.6279,
+      "step": 2659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.641281328531561,
+      "learning_rate": 1.2502870030130285e-05,
+      "loss": 0.6668,
+      "step": 2660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.519344097319205,
+      "learning_rate": 1.2497718785284594e-05,
+      "loss": 0.5505,
+      "step": 2661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1218780555186894,
+      "learning_rate": 1.2492566833464857e-05,
+      "loss": 0.5697,
+      "step": 2662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2962819413681568,
+      "learning_rate": 1.2487414176129322e-05,
+      "loss": 0.5881,
+      "step": 2663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.966077823363184,
+      "learning_rate": 1.2482260814736438e-05,
+      "loss": 0.5428,
+      "step": 2664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5207956928081976,
+      "learning_rate": 1.2477106750744852e-05,
+      "loss": 0.5074,
+      "step": 2665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6135069520696044,
+      "learning_rate": 1.2471951985613414e-05,
+      "loss": 0.4942,
+      "step": 2666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7947381821250614,
+      "learning_rate": 1.2466796520801163e-05,
+      "loss": 0.5702,
+      "step": 2667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9105488749622737,
+      "learning_rate": 1.246164035776735e-05,
+      "loss": 0.655,
+      "step": 2668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8881445856582083,
+      "learning_rate": 1.245648349797141e-05,
+      "loss": 0.7342,
+      "step": 2669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7608752099635778,
+      "learning_rate": 1.2451325942872984e-05,
+      "loss": 0.5812,
+      "step": 2670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0680229396345085,
+      "learning_rate": 1.2446167693931907e-05,
+      "loss": 0.6355,
+      "step": 2671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6246720565653074,
+      "learning_rate": 1.2441008752608212e-05,
+      "loss": 0.4495,
+      "step": 2672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8688422899872266,
+      "learning_rate": 1.2435849120362123e-05,
+      "loss": 0.5263,
+      "step": 2673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0492144670486145,
+      "learning_rate": 1.2430688798654064e-05,
+      "loss": 0.4655,
+      "step": 2674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7576184088451046,
+      "learning_rate": 1.2425527788944656e-05,
+      "loss": 0.5306,
+      "step": 2675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3132257769942721,
+      "learning_rate": 1.2420366092694713e-05,
+      "loss": 0.5425,
+      "step": 2676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6557899930149649,
+      "learning_rate": 1.2415203711365238e-05,
+      "loss": 0.6523,
+      "step": 2677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8758041008392403,
+      "learning_rate": 1.2410040646417431e-05,
+      "loss": 0.4643,
+      "step": 2678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7656196786736913,
+      "learning_rate": 1.2404876899312693e-05,
+      "loss": 0.5114,
+      "step": 2679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.083645812453246,
+      "learning_rate": 1.2399712471512607e-05,
+      "loss": 0.5425,
+      "step": 2680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6932774316501458,
+      "learning_rate": 1.239454736447895e-05,
+      "loss": 0.554,
+      "step": 2681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8493903957424156,
+      "learning_rate": 1.2389381579673704e-05,
+      "loss": 0.6166,
+      "step": 2682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2870897853203895,
+      "learning_rate": 1.2384215118559027e-05,
+      "loss": 0.526,
+      "step": 2683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.764138465549006,
+      "learning_rate": 1.2379047982597277e-05,
+      "loss": 0.5958,
+      "step": 2684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5273644621227715,
+      "learning_rate": 1.2373880173250998e-05,
+      "loss": 0.4953,
+      "step": 2685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3948142538254333,
+      "learning_rate": 1.2368711691982933e-05,
+      "loss": 0.5473,
+      "step": 2686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.741096756478754,
+      "learning_rate": 1.236354254025601e-05,
+      "loss": 0.6016,
+      "step": 2687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.865278295984042,
+      "learning_rate": 1.235837271953334e-05,
+      "loss": 0.637,
+      "step": 2688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5714086790311594,
+      "learning_rate": 1.2353202231278232e-05,
+      "loss": 0.4964,
+      "step": 2689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4811930536438649,
+      "learning_rate": 1.2348031076954186e-05,
+      "loss": 0.5249,
+      "step": 2690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.067367159003529,
+      "learning_rate": 1.2342859258024882e-05,
+      "loss": 0.614,
+      "step": 2691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6413936770143547,
+      "learning_rate": 1.2337686775954193e-05,
+      "loss": 0.6111,
+      "step": 2692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.634615891222511,
+      "learning_rate": 1.2332513632206183e-05,
+      "loss": 0.4591,
+      "step": 2693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.526853031548651,
+      "learning_rate": 1.2327339828245092e-05,
+      "loss": 0.5648,
+      "step": 2694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6145334710440793,
+      "learning_rate": 1.2322165365535364e-05,
+      "loss": 0.5694,
+      "step": 2695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5114457160155275,
+      "learning_rate": 1.2316990245541609e-05,
+      "loss": 0.5396,
+      "step": 2696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4453321770596905,
+      "learning_rate": 1.2311814469728643e-05,
+      "loss": 0.5487,
+      "step": 2697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3115063784211567,
+      "learning_rate": 1.2306638039561455e-05,
+      "loss": 0.5941,
+      "step": 2698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6929683812997371,
+      "learning_rate": 1.2301460956505225e-05,
+      "loss": 0.5705,
+      "step": 2699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.74000150203538,
+      "learning_rate": 1.229628322202531e-05,
+      "loss": 0.5772,
+      "step": 2700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8808964285876608,
+      "learning_rate": 1.229110483758726e-05,
+      "loss": 0.5317,
+      "step": 2701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.704502063460425,
+      "learning_rate": 1.2285925804656806e-05,
+      "loss": 0.7925,
+      "step": 2702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6054597703717706,
+      "learning_rate": 1.2280746124699864e-05,
+      "loss": 0.5138,
+      "step": 2703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9085731798262964,
+      "learning_rate": 1.2275565799182527e-05,
+      "loss": 0.5865,
+      "step": 2704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9566482152860176,
+      "learning_rate": 1.227038482957108e-05,
+      "loss": 0.6086,
+      "step": 2705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.577096264507488,
+      "learning_rate": 1.2265203217331982e-05,
+      "loss": 0.4581,
+      "step": 2706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6217065571459304,
+      "learning_rate": 1.226002096393188e-05,
+      "loss": 0.5292,
+      "step": 2707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5267683915761374,
+      "learning_rate": 1.2254838070837596e-05,
+      "loss": 0.5507,
+      "step": 2708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7764216002738094,
+      "learning_rate": 1.2249654539516143e-05,
+      "loss": 0.6073,
+      "step": 2709
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6052346892933498,
+      "learning_rate": 1.2244470371434705e-05,
+      "loss": 0.4969,
+      "step": 2710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6906535227849617,
+      "learning_rate": 1.2239285568060651e-05,
+      "loss": 0.6887,
+      "step": 2711
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0797290762555685,
+      "learning_rate": 1.2234100130861525e-05,
+      "loss": 0.5221,
+      "step": 2712
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8758877740420918,
+      "learning_rate": 1.2228914061305059e-05,
+      "loss": 0.5748,
+      "step": 2713
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6612769055350554,
+      "learning_rate": 1.2223727360859156e-05,
+      "loss": 0.5534,
+      "step": 2714
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.528124823310635,
+      "learning_rate": 1.2218540030991903e-05,
+      "loss": 0.5683,
+      "step": 2715
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8826673633365254,
+      "learning_rate": 1.2213352073171562e-05,
+      "loss": 0.4829,
+      "step": 2716
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.693878784196554,
+      "learning_rate": 1.2208163488866573e-05,
+      "loss": 0.5914,
+      "step": 2717
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7794899967719193,
+      "learning_rate": 1.2202974279545554e-05,
+      "loss": 0.5676,
+      "step": 2718
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7408244718558539,
+      "learning_rate": 1.2197784446677299e-05,
+      "loss": 0.5716,
+      "step": 2719
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6421150363319637,
+      "learning_rate": 1.2192593991730781e-05,
+      "loss": 0.5093,
+      "step": 2720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.941432663059238,
+      "learning_rate": 1.2187402916175146e-05,
+      "loss": 0.4516,
+      "step": 2721
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0985107843741801,
+      "learning_rate": 1.2182211221479719e-05,
+      "loss": 0.7481,
+      "step": 2722
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2316465931454874,
+      "learning_rate": 1.2177018909113994e-05,
+      "loss": 0.6233,
+      "step": 2723
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4269743055657975,
+      "learning_rate": 1.2171825980547646e-05,
+      "loss": 0.5255,
+      "step": 2724
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6823393199469514,
+      "learning_rate": 1.2166632437250527e-05,
+      "loss": 0.5054,
+      "step": 2725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.52364565516928,
+      "learning_rate": 1.2161438280692655e-05,
+      "loss": 0.3644,
+      "step": 2726
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6896257055090118,
+      "learning_rate": 1.215624351234422e-05,
+      "loss": 0.4731,
+      "step": 2727
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9859466608035767,
+      "learning_rate": 1.21510481336756e-05,
+      "loss": 0.5325,
+      "step": 2728
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2625879933116217,
+      "learning_rate": 1.214585214615733e-05,
+      "loss": 0.5758,
+      "step": 2729
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.106097076969816,
+      "learning_rate": 1.2140655551260124e-05,
+      "loss": 0.4192,
+      "step": 2730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0290911149616724,
+      "learning_rate": 1.2135458350454867e-05,
+      "loss": 0.5529,
+      "step": 2731
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7631468017245608,
+      "learning_rate": 1.2130260545212618e-05,
+      "loss": 0.6065,
+      "step": 2732
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7480827368018723,
+      "learning_rate": 1.2125062137004602e-05,
+      "loss": 0.4883,
+      "step": 2733
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6925135811013277,
+      "learning_rate": 1.2119863127302221e-05,
+      "loss": 0.6102,
+      "step": 2734
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.935764984301531,
+      "learning_rate": 1.211466351757704e-05,
+      "loss": 0.5304,
+      "step": 2735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4882151867082802,
+      "learning_rate": 1.2109463309300798e-05,
+      "loss": 0.5733,
+      "step": 2736
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.845432170369531,
+      "learning_rate": 1.2104262503945406e-05,
+      "loss": 0.5327,
+      "step": 2737
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.832811840279565,
+      "learning_rate": 1.2099061102982939e-05,
+      "loss": 0.4985,
+      "step": 2738
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.986209369415395,
+      "learning_rate": 1.2093859107885642e-05,
+      "loss": 0.5451,
+      "step": 2739
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6114339896757517,
+      "learning_rate": 1.2088656520125929e-05,
+      "loss": 0.5577,
+      "step": 2740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.968836416591233,
+      "learning_rate": 1.2083453341176386e-05,
+      "loss": 0.5261,
+      "step": 2741
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8233655333143444,
+      "learning_rate": 1.2078249572509755e-05,
+      "loss": 0.5949,
+      "step": 2742
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9068423423144556,
+      "learning_rate": 1.2073045215598953e-05,
+      "loss": 0.5108,
+      "step": 2743
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9020738042883116,
+      "learning_rate": 1.2067840271917066e-05,
+      "loss": 0.5674,
+      "step": 2744
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5901447530990052,
+      "learning_rate": 1.206263474293734e-05,
+      "loss": 0.6295,
+      "step": 2745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.699920233757064,
+      "learning_rate": 1.205742863013319e-05,
+      "loss": 0.4934,
+      "step": 2746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7137456719899584,
+      "learning_rate": 1.2052221934978197e-05,
+      "loss": 0.4309,
+      "step": 2747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4536673259051383,
+      "learning_rate": 1.20470146589461e-05,
+      "loss": 0.5986,
+      "step": 2748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6388613493108435,
+      "learning_rate": 1.2041806803510809e-05,
+      "loss": 0.6339,
+      "step": 2749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.506350709727178,
+      "learning_rate": 1.20365983701464e-05,
+      "loss": 0.5914,
+      "step": 2750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0415914834794635,
+      "learning_rate": 1.2031389360327106e-05,
+      "loss": 0.5292,
+      "step": 2751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5912173814769937,
+      "learning_rate": 1.202617977552733e-05,
+      "loss": 0.5586,
+      "step": 2752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7525248396103559,
+      "learning_rate": 1.2020969617221627e-05,
+      "loss": 0.577,
+      "step": 2753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.437870202393687,
+      "learning_rate": 1.2015758886884727e-05,
+      "loss": 0.5033,
+      "step": 2754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5060921566764878,
+      "learning_rate": 1.2010547585991516e-05,
+      "loss": 0.5138,
+      "step": 2755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5435418662140583,
+      "learning_rate": 1.200533571601704e-05,
+      "loss": 0.4613,
+      "step": 2756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8040467596015546,
+      "learning_rate": 1.2000123278436508e-05,
+      "loss": 0.5684,
+      "step": 2757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.431811336053706,
+      "learning_rate": 1.199491027472529e-05,
+      "loss": 0.5038,
+      "step": 2758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7433830303707822,
+      "learning_rate": 1.1989696706358917e-05,
+      "loss": 0.4711,
+      "step": 2759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.459738686784837,
+      "learning_rate": 1.1984482574813076e-05,
+      "loss": 0.5006,
+      "step": 2760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3003477486725914,
+      "learning_rate": 1.1979267881563618e-05,
+      "loss": 0.5428,
+      "step": 2761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3956769594774294,
+      "learning_rate": 1.197405262808655e-05,
+      "loss": 0.5839,
+      "step": 2762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4146805413763563,
+      "learning_rate": 1.1968836815858038e-05,
+      "loss": 0.6385,
+      "step": 2763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9546884018777224,
+      "learning_rate": 1.1963620446354406e-05,
+      "loss": 0.5857,
+      "step": 2764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7309601662251513,
+      "learning_rate": 1.195840352105214e-05,
+      "loss": 0.6249,
+      "step": 2765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.095768845932336,
+      "learning_rate": 1.1953186041427878e-05,
+      "loss": 0.6451,
+      "step": 2766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.409379910957918,
+      "learning_rate": 1.1947968008958414e-05,
+      "loss": 0.4052,
+      "step": 2767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.488490712589492,
+      "learning_rate": 1.1942749425120704e-05,
+      "loss": 0.4982,
+      "step": 2768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8032923132288212,
+      "learning_rate": 1.1937530291391857e-05,
+      "loss": 0.5373,
+      "step": 2769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.712829751486225,
+      "learning_rate": 1.1932310609249135e-05,
+      "loss": 0.3995,
+      "step": 2770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5478927032004424,
+      "learning_rate": 1.1927090380169963e-05,
+      "loss": 0.6386,
+      "step": 2771
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7982546830690456,
+      "learning_rate": 1.1921869605631914e-05,
+      "loss": 0.5936,
+      "step": 2772
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9096892742350738,
+      "learning_rate": 1.1916648287112714e-05,
+      "loss": 0.7044,
+      "step": 2773
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.638248544723444,
+      "learning_rate": 1.191142642609025e-05,
+      "loss": 0.6269,
+      "step": 2774
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3359957439434218,
+      "learning_rate": 1.1906204024042556e-05,
+      "loss": 0.471,
+      "step": 2775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5988554000864943,
+      "learning_rate": 1.1900981082447822e-05,
+      "loss": 0.5704,
+      "step": 2776
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.583652539147579,
+      "learning_rate": 1.1895757602784395e-05,
+      "loss": 0.6433,
+      "step": 2777
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.657637640651976,
+      "learning_rate": 1.1890533586530766e-05,
+      "loss": 0.4732,
+      "step": 2778
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4748399848419405,
+      "learning_rate": 1.1885309035165582e-05,
+      "loss": 0.4616,
+      "step": 2779
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5477391893210135,
+      "learning_rate": 1.1880083950167642e-05,
+      "loss": 0.4789,
+      "step": 2780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7728339419024615,
+      "learning_rate": 1.1874858333015895e-05,
+      "loss": 0.5087,
+      "step": 2781
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8984869882866988,
+      "learning_rate": 1.186963218518944e-05,
+      "loss": 0.654,
+      "step": 2782
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8980061282868275,
+      "learning_rate": 1.1864405508167532e-05,
+      "loss": 0.5565,
+      "step": 2783
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8183038273938331,
+      "learning_rate": 1.1859178303429566e-05,
+      "loss": 0.6565,
+      "step": 2784
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.055091845209971,
+      "learning_rate": 1.1853950572455093e-05,
+      "loss": 0.5477,
+      "step": 2785
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3974262575865166,
+      "learning_rate": 1.1848722316723809e-05,
+      "loss": 0.5635,
+      "step": 2786
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5462681515534287,
+      "learning_rate": 1.1843493537715563e-05,
+      "loss": 0.4837,
+      "step": 2787
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5989736984872178,
+      "learning_rate": 1.1838264236910348e-05,
+      "loss": 0.6426,
+      "step": 2788
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.13687426083246,
+      "learning_rate": 1.183303441578831e-05,
+      "loss": 0.5491,
+      "step": 2789
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5493911311920923,
+      "learning_rate": 1.1827804075829738e-05,
+      "loss": 0.5096,
+      "step": 2790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5678526921306508,
+      "learning_rate": 1.1822573218515068e-05,
+      "loss": 0.5411,
+      "step": 2791
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.055072751301927,
+      "learning_rate": 1.1817341845324882e-05,
+      "loss": 0.5581,
+      "step": 2792
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.616889823006585,
+      "learning_rate": 1.1812109957739907e-05,
+      "loss": 0.5568,
+      "step": 2793
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5305911619076622,
+      "learning_rate": 1.1806877557241023e-05,
+      "loss": 0.4531,
+      "step": 2794
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0127264991700833,
+      "learning_rate": 1.1801644645309252e-05,
+      "loss": 0.6005,
+      "step": 2795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6852169612580683,
+      "learning_rate": 1.179641122342575e-05,
+      "loss": 0.5625,
+      "step": 2796
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3519291682868964,
+      "learning_rate": 1.1791177293071831e-05,
+      "loss": 0.474,
+      "step": 2797
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7409842796674758,
+      "learning_rate": 1.1785942855728945e-05,
+      "loss": 0.663,
+      "step": 2798
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6646939343894271,
+      "learning_rate": 1.1780707912878693e-05,
+      "loss": 0.6145,
+      "step": 2799
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9243502416336138,
+      "learning_rate": 1.1775472466002812e-05,
+      "loss": 0.5955,
+      "step": 2800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6550079732618919,
+      "learning_rate": 1.1770236516583187e-05,
+      "loss": 0.5557,
+      "step": 2801
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4122624322862338,
+      "learning_rate": 1.176500006610184e-05,
+      "loss": 0.6087,
+      "step": 2802
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.363613820648942,
+      "learning_rate": 1.1759763116040936e-05,
+      "loss": 0.5134,
+      "step": 2803
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8593557341854666,
+      "learning_rate": 1.1754525667882786e-05,
+      "loss": 0.6679,
+      "step": 2804
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1821963477018445,
+      "learning_rate": 1.1749287723109834e-05,
+      "loss": 0.6278,
+      "step": 2805
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5633778167596428,
+      "learning_rate": 1.1744049283204677e-05,
+      "loss": 0.6185,
+      "step": 2806
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7877738050289234,
+      "learning_rate": 1.1738810349650036e-05,
+      "loss": 0.5636,
+      "step": 2807
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7163859989561419,
+      "learning_rate": 1.1733570923928785e-05,
+      "loss": 0.502,
+      "step": 2808
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.739030429020945,
+      "learning_rate": 1.1728331007523928e-05,
+      "loss": 0.69,
+      "step": 2809
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7956421708075149,
+      "learning_rate": 1.1723090601918616e-05,
+      "loss": 0.6623,
+      "step": 2810
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6210582552471293,
+      "learning_rate": 1.1717849708596136e-05,
+      "loss": 0.5851,
+      "step": 2811
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7394872589644832,
+      "learning_rate": 1.171260832903991e-05,
+      "loss": 0.59,
+      "step": 2812
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.073300758228722,
+      "learning_rate": 1.1707366464733501e-05,
+      "loss": 0.5676,
+      "step": 2813
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8770507008884283,
+      "learning_rate": 1.1702124117160603e-05,
+      "loss": 0.7333,
+      "step": 2814
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5181819451871015,
+      "learning_rate": 1.1696881287805056e-05,
+      "loss": 0.5914,
+      "step": 2815
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.520964594252851,
+      "learning_rate": 1.1691637978150831e-05,
+      "loss": 0.5123,
+      "step": 2816
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9424114903065128,
+      "learning_rate": 1.1686394189682035e-05,
+      "loss": 0.6094,
+      "step": 2817
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5518883799253531,
+      "learning_rate": 1.1681149923882913e-05,
+      "loss": 0.5392,
+      "step": 2818
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4517497261277148,
+      "learning_rate": 1.1675905182237839e-05,
+      "loss": 0.4957,
+      "step": 2819
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8344154445074144,
+      "learning_rate": 1.167065996623133e-05,
+      "loss": 0.744,
+      "step": 2820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.372651140223707,
+      "learning_rate": 1.166541427734803e-05,
+      "loss": 0.5149,
+      "step": 2821
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9249140352208132,
+      "learning_rate": 1.1660168117072725e-05,
+      "loss": 0.5174,
+      "step": 2822
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4263020348015722,
+      "learning_rate": 1.1654921486890327e-05,
+      "loss": 0.5031,
+      "step": 2823
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5363074452046572,
+      "learning_rate": 1.1649674388285883e-05,
+      "loss": 0.5387,
+      "step": 2824
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7885391690065133,
+      "learning_rate": 1.1644426822744575e-05,
+      "loss": 0.6458,
+      "step": 2825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5806303856059383,
+      "learning_rate": 1.1639178791751715e-05,
+      "loss": 0.6068,
+      "step": 2826
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5579529408305164,
+      "learning_rate": 1.1633930296792744e-05,
+      "loss": 0.5379,
+      "step": 2827
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6671484452981078,
+      "learning_rate": 1.1628681339353244e-05,
+      "loss": 0.5183,
+      "step": 2828
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8216095495142473,
+      "learning_rate": 1.1623431920918916e-05,
+      "loss": 0.6083,
+      "step": 2829
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4951007617004082,
+      "learning_rate": 1.1618182042975596e-05,
+      "loss": 0.6135,
+      "step": 2830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1015181261395943,
+      "learning_rate": 1.1612931707009253e-05,
+      "loss": 0.4994,
+      "step": 2831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.602416966013979,
+      "learning_rate": 1.1607680914505985e-05,
+      "loss": 0.6122,
+      "step": 2832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8267375776497841,
+      "learning_rate": 1.1602429666952015e-05,
+      "loss": 0.675,
+      "step": 2833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6403791786182829,
+      "learning_rate": 1.15971779658337e-05,
+      "loss": 0.6726,
+      "step": 2834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7134476032177197,
+      "learning_rate": 1.1591925812637523e-05,
+      "loss": 0.5799,
+      "step": 2835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8503557974911536,
+      "learning_rate": 1.1586673208850091e-05,
+      "loss": 0.7259,
+      "step": 2836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.563350632629,
+      "learning_rate": 1.158142015595815e-05,
+      "loss": 0.5037,
+      "step": 2837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9238041705751256,
+      "learning_rate": 1.1576166655448558e-05,
+      "loss": 0.5642,
+      "step": 2838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.050595975487935,
+      "learning_rate": 1.1570912708808311e-05,
+      "loss": 0.5491,
+      "step": 2839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6979889635134728,
+      "learning_rate": 1.1565658317524526e-05,
+      "loss": 0.6734,
+      "step": 2840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3451445410227318,
+      "learning_rate": 1.1560403483084449e-05,
+      "loss": 0.4809,
+      "step": 2841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7430721856269416,
+      "learning_rate": 1.1555148206975449e-05,
+      "loss": 0.5517,
+      "step": 2842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.237577242430862,
+      "learning_rate": 1.1549892490685018e-05,
+      "loss": 0.6405,
+      "step": 2843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7806344646164531,
+      "learning_rate": 1.1544636335700778e-05,
+      "loss": 0.5085,
+      "step": 2844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5918970218052164,
+      "learning_rate": 1.1539379743510475e-05,
+      "loss": 0.5742,
+      "step": 2845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.568125284616558,
+      "learning_rate": 1.1534122715601974e-05,
+      "loss": 0.5744,
+      "step": 2846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8123843820324685,
+      "learning_rate": 1.1528865253463266e-05,
+      "loss": 0.5292,
+      "step": 2847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6063581254772814,
+      "learning_rate": 1.1523607358582462e-05,
+      "loss": 0.5274,
+      "step": 2848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5150979372007376,
+      "learning_rate": 1.1518349032447806e-05,
+      "loss": 0.5373,
+      "step": 2849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.123204225100208,
+      "learning_rate": 1.1513090276547647e-05,
+      "loss": 0.5879,
+      "step": 2850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.413015289250622,
+      "learning_rate": 1.150783109237047e-05,
+      "loss": 0.5335,
+      "step": 2851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4999761013371091,
+      "learning_rate": 1.1502571481404873e-05,
+      "loss": 0.5307,
+      "step": 2852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2109074407321243,
+      "learning_rate": 1.149731144513958e-05,
+      "loss": 0.4505,
+      "step": 2853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7068349425405587,
+      "learning_rate": 1.1492050985063432e-05,
+      "loss": 0.564,
+      "step": 2854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6963999123592683,
+      "learning_rate": 1.1486790102665393e-05,
+      "loss": 0.6011,
+      "step": 2855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5535529479917627,
+      "learning_rate": 1.148152879943454e-05,
+      "loss": 0.5387,
+      "step": 2856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6144254952812027,
+      "learning_rate": 1.147626707686008e-05,
+      "loss": 0.5609,
+      "step": 2857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5326555113181795,
+      "learning_rate": 1.1471004936431327e-05,
+      "loss": 0.6307,
+      "step": 2858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4838821322259788,
+      "learning_rate": 1.1465742379637725e-05,
+      "loss": 0.4985,
+      "step": 2859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8403157200268199,
+      "learning_rate": 1.1460479407968827e-05,
+      "loss": 0.6109,
+      "step": 2860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5848570808225277,
+      "learning_rate": 1.1455216022914302e-05,
+      "loss": 0.5672,
+      "step": 2861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.653158415340209,
+      "learning_rate": 1.1449952225963946e-05,
+      "loss": 0.5698,
+      "step": 2862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3187237376011343,
+      "learning_rate": 1.144468801860766e-05,
+      "loss": 0.4841,
+      "step": 2863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8660494508093566,
+      "learning_rate": 1.143942340233547e-05,
+      "loss": 0.723,
+      "step": 2864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.056247853433163,
+      "learning_rate": 1.1434158378637514e-05,
+      "loss": 0.5599,
+      "step": 2865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3916470826998917,
+      "learning_rate": 1.1428892949004049e-05,
+      "loss": 0.5501,
+      "step": 2866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5720709506024506,
+      "learning_rate": 1.1423627114925434e-05,
+      "loss": 0.5959,
+      "step": 2867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.922316714065007,
+      "learning_rate": 1.1418360877892165e-05,
+      "loss": 0.4993,
+      "step": 2868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7544308493196756,
+      "learning_rate": 1.1413094239394833e-05,
+      "loss": 0.6254,
+      "step": 2869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.59699449266011,
+      "learning_rate": 1.140782720092415e-05,
+      "loss": 0.6166,
+      "step": 2870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.370616353804495,
+      "learning_rate": 1.1402559763970943e-05,
+      "loss": 0.4777,
+      "step": 2871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7488155428435532,
+      "learning_rate": 1.139729193002614e-05,
+      "loss": 0.5604,
+      "step": 2872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6045703430660077,
+      "learning_rate": 1.1392023700580796e-05,
+      "loss": 0.5412,
+      "step": 2873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.498068124438778,
+      "learning_rate": 1.1386755077126073e-05,
+      "loss": 0.4991,
+      "step": 2874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.210555088494,
+      "learning_rate": 1.1381486061153244e-05,
+      "loss": 0.5113,
+      "step": 2875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6248069062403192,
+      "learning_rate": 1.1376216654153689e-05,
+      "loss": 0.5764,
+      "step": 2876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5107197531359104,
+      "learning_rate": 1.1370946857618908e-05,
+      "loss": 0.5831,
+      "step": 2877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5939005087927844,
+      "learning_rate": 1.1365676673040502e-05,
+      "loss": 0.5631,
+      "step": 2878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.851779179718678,
+      "learning_rate": 1.1360406101910187e-05,
+      "loss": 0.7171,
+      "step": 2879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3552927727316493,
+      "learning_rate": 1.1355135145719784e-05,
+      "loss": 0.5367,
+      "step": 2880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7754418072603453,
+      "learning_rate": 1.1349863805961233e-05,
+      "loss": 0.5345,
+      "step": 2881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8179743826280796,
+      "learning_rate": 1.1344592084126573e-05,
+      "loss": 0.5139,
+      "step": 2882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.597551754360673,
+      "learning_rate": 1.133931998170795e-05,
+      "loss": 0.6067,
+      "step": 2883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5125837819581078,
+      "learning_rate": 1.1334047500197625e-05,
+      "loss": 0.5801,
+      "step": 2884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6960071667876635,
+      "learning_rate": 1.1328774641087958e-05,
+      "loss": 0.546,
+      "step": 2885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.971301460993105,
+      "learning_rate": 1.132350140587143e-05,
+      "loss": 0.5184,
+      "step": 2886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6800311381486237,
+      "learning_rate": 1.1318227796040608e-05,
+      "loss": 0.6,
+      "step": 2887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5960778890275464,
+      "learning_rate": 1.1312953813088183e-05,
+      "loss": 0.4251,
+      "step": 2888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7886081855408262,
+      "learning_rate": 1.1307679458506947e-05,
+      "loss": 0.5995,
+      "step": 2889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.747926351721227,
+      "learning_rate": 1.1302404733789787e-05,
+      "loss": 0.6044,
+      "step": 2890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7976606542502283,
+      "learning_rate": 1.1297129640429707e-05,
+      "loss": 0.5022,
+      "step": 2891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5811810125671588,
+      "learning_rate": 1.1291854179919812e-05,
+      "loss": 0.5378,
+      "step": 2892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5115518775767158,
+      "learning_rate": 1.1286578353753313e-05,
+      "loss": 0.4537,
+      "step": 2893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7824456641288025,
+      "learning_rate": 1.1281302163423515e-05,
+      "loss": 0.5816,
+      "step": 2894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6483351489272495,
+      "learning_rate": 1.1276025610423835e-05,
+      "loss": 0.5032,
+      "step": 2895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6050596790295055,
+      "learning_rate": 1.1270748696247791e-05,
+      "loss": 0.6219,
+      "step": 2896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5585829792521522,
+      "learning_rate": 1.1265471422389003e-05,
+      "loss": 0.5541,
+      "step": 2897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9017233118487284,
+      "learning_rate": 1.1260193790341186e-05,
+      "loss": 0.5379,
+      "step": 2898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6692517109677656,
+      "learning_rate": 1.1254915801598173e-05,
+      "loss": 0.6258,
+      "step": 2899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.081121249152551,
+      "learning_rate": 1.1249637457653881e-05,
+      "loss": 0.6204,
+      "step": 2900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6156171466052056,
+      "learning_rate": 1.1244358760002337e-05,
+      "loss": 0.5798,
+      "step": 2901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6080174707006538,
+      "learning_rate": 1.1239079710137659e-05,
+      "loss": 0.5018,
+      "step": 2902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8265184711167897,
+      "learning_rate": 1.1233800309554083e-05,
+      "loss": 0.6976,
+      "step": 2903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5116581007348964,
+      "learning_rate": 1.1228520559745922e-05,
+      "loss": 0.5148,
+      "step": 2904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7686726204655436,
+      "learning_rate": 1.1223240462207601e-05,
+      "loss": 0.5076,
+      "step": 2905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0440272877559362,
+      "learning_rate": 1.121796001843364e-05,
+      "loss": 0.5811,
+      "step": 2906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4624928137779813,
+      "learning_rate": 1.1212679229918657e-05,
+      "loss": 0.5674,
+      "step": 2907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6158774262285753,
+      "learning_rate": 1.1207398098157371e-05,
+      "loss": 0.453,
+      "step": 2908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8809613174289643,
+      "learning_rate": 1.1202116624644594e-05,
+      "loss": 0.6548,
+      "step": 2909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0179658113376826,
+      "learning_rate": 1.1196834810875234e-05,
+      "loss": 0.5232,
+      "step": 2910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8829045110008458,
+      "learning_rate": 1.11915526583443e-05,
+      "loss": 0.5308,
+      "step": 2911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7493388791008002,
+      "learning_rate": 1.1186270168546891e-05,
+      "loss": 0.6385,
+      "step": 2912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7633727916643955,
+      "learning_rate": 1.1180987342978209e-05,
+      "loss": 0.5678,
+      "step": 2913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.848447585282512,
+      "learning_rate": 1.1175704183133542e-05,
+      "loss": 0.5266,
+      "step": 2914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7096200492094078,
+      "learning_rate": 1.1170420690508281e-05,
+      "loss": 0.4637,
+      "step": 2915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.889843604932665,
+      "learning_rate": 1.1165136866597905e-05,
+      "loss": 0.4281,
+      "step": 2916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7741504692374862,
+      "learning_rate": 1.1159852712897989e-05,
+      "loss": 0.6851,
+      "step": 2917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.893127656658575,
+      "learning_rate": 1.1154568230904204e-05,
+      "loss": 0.4864,
+      "step": 2918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.655355810248108,
+      "learning_rate": 1.1149283422112312e-05,
+      "loss": 0.5268,
+      "step": 2919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8170725346784516,
+      "learning_rate": 1.1143998288018163e-05,
+      "loss": 0.6892,
+      "step": 2920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.420383779384179,
+      "learning_rate": 1.1138712830117706e-05,
+      "loss": 0.5352,
+      "step": 2921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.121207376658759,
+      "learning_rate": 1.1133427049906978e-05,
+      "loss": 0.6047,
+      "step": 2922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2239854596101702,
+      "learning_rate": 1.1128140948882107e-05,
+      "loss": 0.537,
+      "step": 2923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8076148797978908,
+      "learning_rate": 1.1122854528539315e-05,
+      "loss": 0.6293,
+      "step": 2924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6477877004798225,
+      "learning_rate": 1.111756779037491e-05,
+      "loss": 0.5492,
+      "step": 2925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5089472328263493,
+      "learning_rate": 1.1112280735885295e-05,
+      "loss": 0.5775,
+      "step": 2926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.275203415690187,
+      "learning_rate": 1.1106993366566957e-05,
+      "loss": 0.5525,
+      "step": 2927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7550911265310598,
+      "learning_rate": 1.1101705683916473e-05,
+      "loss": 0.4498,
+      "step": 2928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4834983303046843,
+      "learning_rate": 1.1096417689430517e-05,
+      "loss": 0.535,
+      "step": 2929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.046176374841131,
+      "learning_rate": 1.1091129384605837e-05,
+      "loss": 0.5827,
+      "step": 2930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5609417284829272,
+      "learning_rate": 1.1085840770939283e-05,
+      "loss": 0.531,
+      "step": 2931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4738703311017116,
+      "learning_rate": 1.108055184992778e-05,
+      "loss": 0.5679,
+      "step": 2932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7181607466367828,
+      "learning_rate": 1.1075262623068352e-05,
+      "loss": 0.5901,
+      "step": 2933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.685677830722803,
+      "learning_rate": 1.10699730918581e-05,
+      "loss": 0.5858,
+      "step": 2934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7654675550529497,
+      "learning_rate": 1.1064683257794216e-05,
+      "loss": 0.5006,
+      "step": 2935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.610988604940138,
+      "learning_rate": 1.1059393122373976e-05,
+      "loss": 0.5327,
+      "step": 2936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9489708589795773,
+      "learning_rate": 1.1054102687094738e-05,
+      "loss": 0.5469,
+      "step": 2937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.609897223602563,
+      "learning_rate": 1.1048811953453955e-05,
+      "loss": 0.5502,
+      "step": 2938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.945536158591971,
+      "learning_rate": 1.1043520922949156e-05,
+      "loss": 0.7288,
+      "step": 2939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7626834503333355,
+      "learning_rate": 1.1038229597077954e-05,
+      "loss": 0.59,
+      "step": 2940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8791758030649353,
+      "learning_rate": 1.1032937977338048e-05,
+      "loss": 0.5554,
+      "step": 2941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.721583446063103,
+      "learning_rate": 1.1027646065227222e-05,
+      "loss": 0.6595,
+      "step": 2942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9618113883026074,
+      "learning_rate": 1.1022353862243338e-05,
+      "loss": 0.5592,
+      "step": 2943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7128530093039571,
+      "learning_rate": 1.1017061369884345e-05,
+      "loss": 0.5614,
+      "step": 2944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5790441686976435,
+      "learning_rate": 1.101176858964827e-05,
+      "loss": 0.5426,
+      "step": 2945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5185261305382007,
+      "learning_rate": 1.1006475523033225e-05,
+      "loss": 0.5917,
+      "step": 2946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8450734135988391,
+      "learning_rate": 1.10011821715374e-05,
+      "loss": 0.6844,
+      "step": 2947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.780970693022275,
+      "learning_rate": 1.0995888536659067e-05,
+      "loss": 0.5943,
+      "step": 2948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7059542291155376,
+      "learning_rate": 1.0990594619896581e-05,
+      "loss": 0.5343,
+      "step": 2949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4569806521849618,
+      "learning_rate": 1.098530042274837e-05,
+      "loss": 0.6275,
+      "step": 2950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6919261843059368,
+      "learning_rate": 1.0980005946712949e-05,
+      "loss": 0.485,
+      "step": 2951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.071999424922584,
+      "learning_rate": 1.0974711193288906e-05,
+      "loss": 0.5902,
+      "step": 2952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9271993841648203,
+      "learning_rate": 1.096941616397491e-05,
+      "loss": 0.6531,
+      "step": 2953
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7332150944748392,
+      "learning_rate": 1.0964120860269708e-05,
+      "loss": 0.586,
+      "step": 2954
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.933201746152474,
+      "learning_rate": 1.0958825283672126e-05,
+      "loss": 0.6069,
+      "step": 2955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.037618964787883,
+      "learning_rate": 1.0953529435681063e-05,
+      "loss": 0.5487,
+      "step": 2956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.63692090375333,
+      "learning_rate": 1.09482333177955e-05,
+      "loss": 0.5407,
+      "step": 2957
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.961255565590994,
+      "learning_rate": 1.0942936931514492e-05,
+      "loss": 0.5294,
+      "step": 2958
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3782501087768673,
+      "learning_rate": 1.0937640278337167e-05,
+      "loss": 0.4898,
+      "step": 2959
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.833804235299014,
+      "learning_rate": 1.0932343359762736e-05,
+      "loss": 0.5895,
+      "step": 2960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0540947623424968,
+      "learning_rate": 1.0927046177290477e-05,
+      "loss": 0.5705,
+      "step": 2961
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8353694531187756,
+      "learning_rate": 1.092174873241975e-05,
+      "loss": 0.4782,
+      "step": 2962
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.288471580749892,
+      "learning_rate": 1.0916451026649981e-05,
+      "loss": 0.6039,
+      "step": 2963
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5919466966688622,
+      "learning_rate": 1.091115306148068e-05,
+      "loss": 0.6307,
+      "step": 2964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5519270530577924,
+      "learning_rate": 1.0905854838411418e-05,
+      "loss": 0.5422,
+      "step": 2965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8283150785871538,
+      "learning_rate": 1.0900556358941855e-05,
+      "loss": 0.5197,
+      "step": 2966
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6829968021780106,
+      "learning_rate": 1.0895257624571705e-05,
+      "loss": 0.5526,
+      "step": 2967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5383060717904002,
+      "learning_rate": 1.088995863680077e-05,
+      "loss": 0.5582,
+      "step": 2968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.745235216672498,
+      "learning_rate": 1.0884659397128911e-05,
+      "loss": 0.5596,
+      "step": 2969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.369769464372544,
+      "learning_rate": 1.0879359907056074e-05,
+      "loss": 0.5979,
+      "step": 2970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.016527090226891,
+      "learning_rate": 1.0874060168082266e-05,
+      "loss": 0.597,
+      "step": 2971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9224426282227236,
+      "learning_rate": 1.0868760181707565e-05,
+      "loss": 0.5615,
+      "step": 2972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6438409001992689,
+      "learning_rate": 1.0863459949432122e-05,
+      "loss": 0.5465,
+      "step": 2973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6150969486575681,
+      "learning_rate": 1.0858159472756157e-05,
+      "loss": 0.5144,
+      "step": 2974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.199664014053087,
+      "learning_rate": 1.085285875317996e-05,
+      "loss": 0.5932,
+      "step": 2975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.819738739857083,
+      "learning_rate": 1.0847557792203886e-05,
+      "loss": 0.5129,
+      "step": 2976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8928722311349389,
+      "learning_rate": 1.0842256591328362e-05,
+      "loss": 0.5747,
+      "step": 2977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4746456943409445,
+      "learning_rate": 1.0836955152053883e-05,
+      "loss": 0.49,
+      "step": 2978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9849739376792446,
+      "learning_rate": 1.083165347588101e-05,
+      "loss": 0.502,
+      "step": 2979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6649502969902297,
+      "learning_rate": 1.082635156431037e-05,
+      "loss": 0.5475,
+      "step": 2980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7743395450582287,
+      "learning_rate": 1.0821049418842654e-05,
+      "loss": 0.5456,
+      "step": 2981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1365924369348885,
+      "learning_rate": 1.0815747040978628e-05,
+      "loss": 0.5621,
+      "step": 2982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.341304368074102,
+      "learning_rate": 1.081044443221912e-05,
+      "loss": 0.4781,
+      "step": 2983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.260655285608248,
+      "learning_rate": 1.0805141594065022e-05,
+      "loss": 0.5592,
+      "step": 2984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.590691837905553,
+      "learning_rate": 1.0799838528017288e-05,
+      "loss": 0.5852,
+      "step": 2985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5837528694802505,
+      "learning_rate": 1.0794535235576941e-05,
+      "loss": 0.5677,
+      "step": 2986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0465314340085987,
+      "learning_rate": 1.0789231718245069e-05,
+      "loss": 0.5938,
+      "step": 2987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.703487500736417,
+      "learning_rate": 1.0783927977522819e-05,
+      "loss": 0.493,
+      "step": 2988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3261810231221967,
+      "learning_rate": 1.0778624014911403e-05,
+      "loss": 0.6085,
+      "step": 2989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7162901704506919,
+      "learning_rate": 1.0773319831912099e-05,
+      "loss": 0.5861,
+      "step": 2990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7370742855948245,
+      "learning_rate": 1.0768015430026244e-05,
+      "loss": 0.5968,
+      "step": 2991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.285857518931543,
+      "learning_rate": 1.0762710810755234e-05,
+      "loss": 0.541,
+      "step": 2992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.651463761895187,
+      "learning_rate": 1.0757405975600534e-05,
+      "loss": 0.449,
+      "step": 2993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5266826011996766,
+      "learning_rate": 1.0752100926063669e-05,
+      "loss": 0.5714,
+      "step": 2994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.61660092953963,
+      "learning_rate": 1.074679566364622e-05,
+      "loss": 0.5585,
+      "step": 2995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.02382465264863,
+      "learning_rate": 1.0741490189849826e-05,
+      "loss": 0.5941,
+      "step": 2996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9038994970567877,
+      "learning_rate": 1.0736184506176195e-05,
+      "loss": 0.6258,
+      "step": 2997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8742084613152803,
+      "learning_rate": 1.0730878614127087e-05,
+      "loss": 0.6704,
+      "step": 2998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5540042782862435,
+      "learning_rate": 1.0725572515204327e-05,
+      "loss": 0.5404,
+      "step": 2999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.974959136945086,
+      "learning_rate": 1.0720266210909793e-05,
+      "loss": 0.5107,
+      "step": 3000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.907213073659795,
+      "learning_rate": 1.0714959702745424e-05,
+      "loss": 0.5289,
+      "step": 3001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.493110627963374,
+      "learning_rate": 1.0709652992213216e-05,
+      "loss": 0.5743,
+      "step": 3002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6821163011947058,
+      "learning_rate": 1.0704346080815218e-05,
+      "loss": 0.6077,
+      "step": 3003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.939325316514344,
+      "learning_rate": 1.0699038970053544e-05,
+      "loss": 0.5502,
+      "step": 3004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7887565133067098,
+      "learning_rate": 1.069373166143036e-05,
+      "loss": 0.4994,
+      "step": 3005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.567760156867725,
+      "learning_rate": 1.068842415644789e-05,
+      "loss": 0.543,
+      "step": 3006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8833936081516223,
+      "learning_rate": 1.0683116456608411e-05,
+      "loss": 0.5693,
+      "step": 3007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.001921596369729,
+      "learning_rate": 1.0677808563414256e-05,
+      "loss": 0.4982,
+      "step": 3008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5029073128318458,
+      "learning_rate": 1.0672500478367813e-05,
+      "loss": 0.4531,
+      "step": 3009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8105791444377206,
+      "learning_rate": 1.0667192202971525e-05,
+      "loss": 0.5648,
+      "step": 3010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8565608064310117,
+      "learning_rate": 1.0661883738727888e-05,
+      "loss": 0.558,
+      "step": 3011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.521471951494628,
+      "learning_rate": 1.0656575087139452e-05,
+      "loss": 0.4901,
+      "step": 3012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.719870518770742,
+      "learning_rate": 1.0651266249708816e-05,
+      "loss": 0.5893,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6271102209304236,
+      "learning_rate": 1.064595722793864e-05,
+      "loss": 0.5674,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7126484822723427,
+      "learning_rate": 1.0640648023331625e-05,
+      "loss": 0.5661,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7352777928837653,
+      "learning_rate": 1.063533863739054e-05,
+      "loss": 0.5194,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4898829731478354,
+      "learning_rate": 1.0630029071618188e-05,
+      "loss": 0.4739,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4783328215011815,
+      "learning_rate": 1.0624719327517434e-05,
+      "loss": 0.5507,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9133520277089824,
+      "learning_rate": 1.061940940659119e-05,
+      "loss": 0.5664,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7086534823588067,
+      "learning_rate": 1.0614099310342414e-05,
+      "loss": 0.4826,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7934269298155037,
+      "learning_rate": 1.0608789040274122e-05,
+      "loss": 0.7333,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.770943890067993,
+      "learning_rate": 1.0603478597889374e-05,
+      "loss": 0.4879,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8466702113511095,
+      "learning_rate": 1.0598167984691276e-05,
+      "loss": 0.5937,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.645933491955772,
+      "learning_rate": 1.059285720218299e-05,
+      "loss": 0.4965,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.577050436919074,
+      "learning_rate": 1.058754625186772e-05,
+      "loss": 0.5316,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.634297544960071,
+      "learning_rate": 1.0582235135248718e-05,
+      "loss": 0.4881,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8162493136817117,
+      "learning_rate": 1.0576923853829284e-05,
+      "loss": 0.4992,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.65769655255232,
+      "learning_rate": 1.057161240911277e-05,
+      "loss": 0.4834,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5687854027249415,
+      "learning_rate": 1.0566300802602565e-05,
+      "loss": 0.4796,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5076619068637351,
+      "learning_rate": 1.056098903580211e-05,
+      "loss": 0.4779,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8605594944024202,
+      "learning_rate": 1.0555677110214889e-05,
+      "loss": 0.5332,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.78928315524023,
+      "learning_rate": 1.0550365027344432e-05,
+      "loss": 0.5241,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7688527097086948,
+      "learning_rate": 1.0545052788694312e-05,
+      "loss": 0.6089,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.607101376530837,
+      "learning_rate": 1.0539740395768143e-05,
+      "loss": 0.4863,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5367023051324835,
+      "learning_rate": 1.0534427850069595e-05,
+      "loss": 0.4679,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6609982286944611,
+      "learning_rate": 1.0529115153102366e-05,
+      "loss": 0.5254,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6606454483680144,
+      "learning_rate": 1.0523802306370206e-05,
+      "loss": 0.5621,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.510892436215939,
+      "learning_rate": 1.0518489311376905e-05,
+      "loss": 0.6158,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.904432643930373,
+      "learning_rate": 1.0513176169626293e-05,
+      "loss": 0.5372,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1132794707890006,
+      "learning_rate": 1.0507862882622249e-05,
+      "loss": 0.5401,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5766602914257457,
+      "learning_rate": 1.0502549451868683e-05,
+      "loss": 0.5474,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6441122570717583,
+      "learning_rate": 1.0497235878869554e-05,
+      "loss": 0.4874,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5669077408935277,
+      "learning_rate": 1.0491922165128853e-05,
+      "loss": 0.5257,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4603570126275096,
+      "learning_rate": 1.048660831215062e-05,
+      "loss": 0.4847,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.570652939715845,
+      "learning_rate": 1.0481294321438928e-05,
+      "loss": 0.4979,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.491041934065291,
+      "learning_rate": 1.0475980194497892e-05,
+      "loss": 0.5699,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6496726609755434,
+      "learning_rate": 1.0470665932831661e-05,
+      "loss": 0.6125,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6626462788685736,
+      "learning_rate": 1.0465351537944429e-05,
+      "loss": 0.5718,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4725757489325755,
+      "learning_rate": 1.0460037011340422e-05,
+      "loss": 0.5053,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.863460102281669,
+      "learning_rate": 1.0454722354523906e-05,
+      "loss": 0.5959,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5184076121753587,
+      "learning_rate": 1.0449407568999186e-05,
+      "loss": 0.4927,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0984747255881313,
+      "learning_rate": 1.04440926562706e-05,
+      "loss": 0.5706,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.40809570417057,
+      "learning_rate": 1.043877761784252e-05,
+      "loss": 0.575,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1946792552181558,
+      "learning_rate": 1.0433462455219359e-05,
+      "loss": 0.534,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9176699015199805,
+      "learning_rate": 1.0428147169905563e-05,
+      "loss": 0.4852,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5568896320795615,
+      "learning_rate": 1.0422831763405612e-05,
+      "loss": 0.5641,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8093741907358054,
+      "learning_rate": 1.0417516237224017e-05,
+      "loss": 0.539,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8237534160545363,
+      "learning_rate": 1.0412200592865331e-05,
+      "loss": 0.4902,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6507963449497098,
+      "learning_rate": 1.0406884831834133e-05,
+      "loss": 0.5073,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4276570322906832,
+      "learning_rate": 1.0401568955635042e-05,
+      "loss": 0.5575,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4835093113463171,
+      "learning_rate": 1.0396252965772702e-05,
+      "loss": 0.497,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7712724216160707,
+      "learning_rate": 1.0390936863751791e-05,
+      "loss": 0.6345,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7799046865995998,
+      "learning_rate": 1.0385620651077024e-05,
+      "loss": 0.5372,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6016679722295348,
+      "learning_rate": 1.0380304329253144e-05,
+      "loss": 0.4979,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7590416243338436,
+      "learning_rate": 1.0374987899784925e-05,
+      "loss": 0.5507,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5445721519334104,
+      "learning_rate": 1.036967136417717e-05,
+      "loss": 0.5698,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.312154940832882,
+      "learning_rate": 1.036435472393471e-05,
+      "loss": 0.5467,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.761727134652851,
+      "learning_rate": 1.0359037980562416e-05,
+      "loss": 0.5346,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.457694803484192,
+      "learning_rate": 1.0353721135565173e-05,
+      "loss": 0.5248,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7305669233827419,
+      "learning_rate": 1.034840419044791e-05,
+      "loss": 0.5387,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8813295729984526,
+      "learning_rate": 1.0343087146715573e-05,
+      "loss": 0.6682,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9262909127777008,
+      "learning_rate": 1.033777000587314e-05,
+      "loss": 0.6053,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6378868193625629,
+      "learning_rate": 1.0332452769425619e-05,
+      "loss": 0.6381,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4332950161241882,
+      "learning_rate": 1.0327135438878035e-05,
+      "loss": 0.4978,
+      "step": 3074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8368504288005418,
+      "learning_rate": 1.0321818015735459e-05,
+      "loss": 0.4502,
+      "step": 3075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.606694062886377,
+      "learning_rate": 1.031650050150297e-05,
+      "loss": 0.4703,
+      "step": 3076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4680719852288309,
+      "learning_rate": 1.0311182897685681e-05,
+      "loss": 0.5192,
+      "step": 3077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5602694465557978,
+      "learning_rate": 1.0305865205788728e-05,
+      "loss": 0.5764,
+      "step": 3078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8636746034323488,
+      "learning_rate": 1.0300547427317269e-05,
+      "loss": 0.5942,
+      "step": 3079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.492282229739771,
+      "learning_rate": 1.0295229563776494e-05,
+      "loss": 0.6116,
+      "step": 3080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7328216354218064,
+      "learning_rate": 1.0289911616671613e-05,
+      "loss": 0.6011,
+      "step": 3081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5673860990749426,
+      "learning_rate": 1.0284593587507857e-05,
+      "loss": 0.5454,
+      "step": 3082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9241111680457703,
+      "learning_rate": 1.0279275477790487e-05,
+      "loss": 0.5445,
+      "step": 3083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5009778865431305,
+      "learning_rate": 1.0273957289024778e-05,
+      "loss": 0.5142,
+      "step": 3084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6239681045457097,
+      "learning_rate": 1.0268639022716033e-05,
+      "loss": 0.5549,
+      "step": 3085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.909987678506556,
+      "learning_rate": 1.0263320680369581e-05,
+      "loss": 0.5089,
+      "step": 3086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.079710614591378,
+      "learning_rate": 1.0258002263490767e-05,
+      "loss": 0.463,
+      "step": 3087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9988301326143076,
+      "learning_rate": 1.0252683773584953e-05,
+      "loss": 0.5226,
+      "step": 3088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6009993241293554,
+      "learning_rate": 1.0247365212157527e-05,
+      "loss": 0.5868,
+      "step": 3089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5569857388604584,
+      "learning_rate": 1.02420465807139e-05,
+      "loss": 0.5923,
+      "step": 3090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8491248744563706,
+      "learning_rate": 1.0236727880759496e-05,
+      "loss": 0.527,
+      "step": 3091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.697195952922504,
+      "learning_rate": 1.0231409113799764e-05,
+      "loss": 0.5683,
+      "step": 3092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.654317373899092,
+      "learning_rate": 1.0226090281340168e-05,
+      "loss": 0.5756,
+      "step": 3093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5004602698128566,
+      "learning_rate": 1.0220771384886194e-05,
+      "loss": 0.5481,
+      "step": 3094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4842467224953537,
+      "learning_rate": 1.0215452425943346e-05,
+      "loss": 0.5416,
+      "step": 3095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.497902839890057,
+      "learning_rate": 1.021013340601714e-05,
+      "loss": 0.562,
+      "step": 3096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8242475333152814,
+      "learning_rate": 1.0204814326613115e-05,
+      "loss": 0.5917,
+      "step": 3097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.639314325503013,
+      "learning_rate": 1.0199495189236828e-05,
+      "loss": 0.5527,
+      "step": 3098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7208261893288879,
+      "learning_rate": 1.0194175995393847e-05,
+      "loss": 0.6089,
+      "step": 3099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7863556121726551,
+      "learning_rate": 1.0188856746589757e-05,
+      "loss": 0.5752,
+      "step": 3100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.090242621226699,
+      "learning_rate": 1.0183537444330165e-05,
+      "loss": 0.5227,
+      "step": 3101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6731559774436164,
+      "learning_rate": 1.0178218090120683e-05,
+      "loss": 0.5422,
+      "step": 3102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8738492074860534,
+      "learning_rate": 1.0172898685466947e-05,
+      "loss": 0.6856,
+      "step": 3103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.749937235739513,
+      "learning_rate": 1.01675792318746e-05,
+      "loss": 0.5959,
+      "step": 3104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.786285301500408,
+      "learning_rate": 1.0162259730849306e-05,
+      "loss": 0.6483,
+      "step": 3105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5018330707543317,
+      "learning_rate": 1.0156940183896737e-05,
+      "loss": 0.5744,
+      "step": 3106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6692784996804009,
+      "learning_rate": 1.0151620592522577e-05,
+      "loss": 0.545,
+      "step": 3107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.741379216360448,
+      "learning_rate": 1.0146300958232528e-05,
+      "loss": 0.6642,
+      "step": 3108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9862770003952004,
+      "learning_rate": 1.0140981282532301e-05,
+      "loss": 0.6041,
+      "step": 3109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5363151566521214,
+      "learning_rate": 1.0135661566927619e-05,
+      "loss": 0.585,
+      "step": 3110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5291342803814914,
+      "learning_rate": 1.0130341812924215e-05,
+      "loss": 0.4955,
+      "step": 3111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0901254439660115,
+      "learning_rate": 1.0125022022027834e-05,
+      "loss": 0.5961,
+      "step": 3112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3623552874347378,
+      "learning_rate": 1.0119702195744236e-05,
+      "loss": 0.6804,
+      "step": 3113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.149879093399273,
+      "learning_rate": 1.011438233557918e-05,
+      "loss": 0.538,
+      "step": 3114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.218898278972913,
+      "learning_rate": 1.0109062443038446e-05,
+      "loss": 0.5575,
+      "step": 3115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.895287753402588,
+      "learning_rate": 1.0103742519627818e-05,
+      "loss": 0.5284,
+      "step": 3116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5395842835474556,
+      "learning_rate": 1.0098422566853086e-05,
+      "loss": 0.5551,
+      "step": 3117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7346720722636386,
+      "learning_rate": 1.0093102586220056e-05,
+      "loss": 0.5281,
+      "step": 3118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5822535531416417,
+      "learning_rate": 1.0087782579234532e-05,
+      "loss": 0.5694,
+      "step": 3119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7162261983254539,
+      "learning_rate": 1.0082462547402337e-05,
+      "loss": 0.4643,
+      "step": 3120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3712800234250018,
+      "learning_rate": 1.0077142492229288e-05,
+      "loss": 0.4963,
+      "step": 3121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7248232643436225,
+      "learning_rate": 1.007182241522122e-05,
+      "loss": 0.4544,
+      "step": 3122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7514318417062922,
+      "learning_rate": 1.0066502317883969e-05,
+      "loss": 0.6327,
+      "step": 3123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5830707279048637,
+      "learning_rate": 1.0061182201723377e-05,
+      "loss": 0.534,
+      "step": 3124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5437910158063377,
+      "learning_rate": 1.005586206824529e-05,
+      "loss": 0.4633,
+      "step": 3125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7812132707993529,
+      "learning_rate": 1.0050541918955564e-05,
+      "loss": 0.5836,
+      "step": 3126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.721789355736434,
+      "learning_rate": 1.0045221755360053e-05,
+      "loss": 0.5217,
+      "step": 3127
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9125677094105056,
+      "learning_rate": 1.0039901578964619e-05,
+      "loss": 0.5881,
+      "step": 3128
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.966914192251691,
+      "learning_rate": 1.0034581391275129e-05,
+      "loss": 0.5065,
+      "step": 3129
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6222403572008401,
+      "learning_rate": 1.0029261193797446e-05,
+      "loss": 0.6019,
+      "step": 3130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.534305057743442,
+      "learning_rate": 1.0023940988037446e-05,
+      "loss": 0.6018,
+      "step": 3131
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.228198153799826,
+      "learning_rate": 1.0018620775500999e-05,
+      "loss": 0.5196,
+      "step": 3132
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5655950264647456,
+      "learning_rate": 1.0013300557693981e-05,
+      "loss": 0.5457,
+      "step": 3133
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.033457197689247,
+      "learning_rate": 1.0007980336122267e-05,
+      "loss": 0.4878,
+      "step": 3134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7092977112091063,
+      "learning_rate": 1.0002660112291736e-05,
+      "loss": 0.5062,
+      "step": 3135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9433705241406125,
+      "learning_rate": 9.997339887708269e-06,
+      "loss": 0.6324,
+      "step": 3136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9624517435912407,
+      "learning_rate": 9.992019663877738e-06,
+      "loss": 0.5567,
+      "step": 3137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.419899201802753,
+      "learning_rate": 9.986699442306025e-06,
+      "loss": 0.5167,
+      "step": 3138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8820466449353916,
+      "learning_rate": 9.981379224499006e-06,
+      "loss": 0.5527,
+      "step": 3139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6477493474792795,
+      "learning_rate": 9.976059011962557e-06,
+      "loss": 0.5344,
+      "step": 3140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5121143152413903,
+      "learning_rate": 9.970738806202557e-06,
+      "loss": 0.5308,
+      "step": 3141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0115794680859254,
+      "learning_rate": 9.965418608724875e-06,
+      "loss": 0.5843,
+      "step": 3142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5061376854081838,
+      "learning_rate": 9.960098421035383e-06,
+      "loss": 0.4843,
+      "step": 3143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3179958913107963,
+      "learning_rate": 9.95477824463995e-06,
+      "loss": 0.5257,
+      "step": 3144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2455937870952636,
+      "learning_rate": 9.94945808104444e-06,
+      "loss": 0.4677,
+      "step": 3145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4458537366305324,
+      "learning_rate": 9.944137931754712e-06,
+      "loss": 0.5279,
+      "step": 3146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4840693242320624,
+      "learning_rate": 9.938817798276627e-06,
+      "loss": 0.4969,
+      "step": 3147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5085157243562628,
+      "learning_rate": 9.933497682116035e-06,
+      "loss": 0.5089,
+      "step": 3148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3929266210216837,
+      "learning_rate": 9.928177584778783e-06,
+      "loss": 0.5253,
+      "step": 3149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6238551860201342,
+      "learning_rate": 9.922857507770716e-06,
+      "loss": 0.5615,
+      "step": 3150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.62494048126034,
+      "learning_rate": 9.917537452597667e-06,
+      "loss": 0.59,
+      "step": 3151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3634060291510757,
+      "learning_rate": 9.912217420765471e-06,
+      "loss": 0.5558,
+      "step": 3152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8371042644690334,
+      "learning_rate": 9.906897413779949e-06,
+      "loss": 0.5725,
+      "step": 3153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.330762300379894,
+      "learning_rate": 9.901577433146915e-06,
+      "loss": 0.5912,
+      "step": 3154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.424244157896965,
+      "learning_rate": 9.896257480372184e-06,
+      "loss": 0.5197,
+      "step": 3155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4294860269088328,
+      "learning_rate": 9.890937556961554e-06,
+      "loss": 0.6239,
+      "step": 3156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.011392611417451,
+      "learning_rate": 9.88561766442082e-06,
+      "loss": 0.533,
+      "step": 3157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4856038822124062,
+      "learning_rate": 9.88029780425577e-06,
+      "loss": 0.4666,
+      "step": 3158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9103451068420816,
+      "learning_rate": 9.87497797797217e-06,
+      "loss": 0.5093,
+      "step": 3159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4481258104516104,
+      "learning_rate": 9.86965818707579e-06,
+      "loss": 0.6344,
+      "step": 3160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5979995526021407,
+      "learning_rate": 9.864338433072386e-06,
+      "loss": 0.4846,
+      "step": 3161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6354582252002818,
+      "learning_rate": 9.859018717467704e-06,
+      "loss": 0.5702,
+      "step": 3162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6587139837045612,
+      "learning_rate": 9.853699041767473e-06,
+      "loss": 0.5098,
+      "step": 3163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.596098894146592,
+      "learning_rate": 9.848379407477425e-06,
+      "loss": 0.5271,
+      "step": 3164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8441312359436954,
+      "learning_rate": 9.843059816103267e-06,
+      "loss": 0.5945,
+      "step": 3165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.754059027107469,
+      "learning_rate": 9.837740269150696e-06,
+      "loss": 0.6515,
+      "step": 3166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6245618292899953,
+      "learning_rate": 9.832420768125402e-06,
+      "loss": 0.5013,
+      "step": 3167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7081453063553675,
+      "learning_rate": 9.827101314533056e-06,
+      "loss": 0.5695,
+      "step": 3168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6970246359714807,
+      "learning_rate": 9.82178190987932e-06,
+      "loss": 0.5084,
+      "step": 3169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6609512749754443,
+      "learning_rate": 9.816462555669838e-06,
+      "loss": 0.5636,
+      "step": 3170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.707367447325233,
+      "learning_rate": 9.811143253410244e-06,
+      "loss": 0.552,
+      "step": 3171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5722496068531604,
+      "learning_rate": 9.805824004606156e-06,
+      "loss": 0.4481,
+      "step": 3172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5689655605082766,
+      "learning_rate": 9.800504810763176e-06,
+      "loss": 0.5772,
+      "step": 3173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9950864927784198,
+      "learning_rate": 9.795185673386886e-06,
+      "loss": 0.5605,
+      "step": 3174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1042106111817214,
+      "learning_rate": 9.789866593982863e-06,
+      "loss": 0.5685,
+      "step": 3175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7984081849251692,
+      "learning_rate": 9.784547574056657e-06,
+      "loss": 0.608,
+      "step": 3176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6694667497836784,
+      "learning_rate": 9.779228615113808e-06,
+      "loss": 0.5189,
+      "step": 3177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4426005027143467,
+      "learning_rate": 9.773909718659831e-06,
+      "loss": 0.5186,
+      "step": 3178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.88407126558891,
+      "learning_rate": 9.768590886200241e-06,
+      "loss": 0.4845,
+      "step": 3179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9939515689785174,
+      "learning_rate": 9.76327211924051e-06,
+      "loss": 0.5649,
+      "step": 3180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0797394332885797,
+      "learning_rate": 9.757953419286107e-06,
+      "loss": 0.4948,
+      "step": 3181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5965824694865787,
+      "learning_rate": 9.752634787842478e-06,
+      "loss": 0.5419,
+      "step": 3182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.711729390858528,
+      "learning_rate": 9.747316226415052e-06,
+      "loss": 0.5308,
+      "step": 3183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.340502116771188,
+      "learning_rate": 9.741997736509238e-06,
+      "loss": 0.4805,
+      "step": 3184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8572607541574927,
+      "learning_rate": 9.73667931963042e-06,
+      "loss": 0.6794,
+      "step": 3185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4797823812360766,
+      "learning_rate": 9.731360977283969e-06,
+      "loss": 0.5938,
+      "step": 3186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6289753153736344,
+      "learning_rate": 9.726042710975224e-06,
+      "loss": 0.4822,
+      "step": 3187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.653037547938854,
+      "learning_rate": 9.720724522209518e-06,
+      "loss": 0.6071,
+      "step": 3188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.603627215644508,
+      "learning_rate": 9.715406412492145e-06,
+      "loss": 0.5316,
+      "step": 3189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.714102436417958,
+      "learning_rate": 9.710088383328392e-06,
+      "loss": 0.455,
+      "step": 3190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.460130114945431,
+      "learning_rate": 9.704770436223508e-06,
+      "loss": 0.5389,
+      "step": 3191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4551572232380092,
+      "learning_rate": 9.699452572682734e-06,
+      "loss": 0.5044,
+      "step": 3192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.755734372639096,
+      "learning_rate": 9.694134794211277e-06,
+      "loss": 0.541,
+      "step": 3193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6612937839898567,
+      "learning_rate": 9.68881710231432e-06,
+      "loss": 0.5323,
+      "step": 3194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9200347013059131,
+      "learning_rate": 9.683499498497032e-06,
+      "loss": 0.5028,
+      "step": 3195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2864330021373918,
+      "learning_rate": 9.678181984264543e-06,
+      "loss": 0.5384,
+      "step": 3196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.866876571101631,
+      "learning_rate": 9.672864561121963e-06,
+      "loss": 0.4471,
+      "step": 3197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8004828890422082,
+      "learning_rate": 9.667547230574386e-06,
+      "loss": 0.546,
+      "step": 3198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7197885879929347,
+      "learning_rate": 9.662229994126862e-06,
+      "loss": 0.4803,
+      "step": 3199
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6538466399727287,
+      "learning_rate": 9.65691285328443e-06,
+      "loss": 0.6045,
+      "step": 3200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7603909004542887,
+      "learning_rate": 9.651595809552094e-06,
+      "loss": 0.6385,
+      "step": 3201
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7151385136856308,
+      "learning_rate": 9.64627886443483e-06,
+      "loss": 0.5371,
+      "step": 3202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8572692591901205,
+      "learning_rate": 9.64096201943759e-06,
+      "loss": 0.7167,
+      "step": 3203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.098982759612629,
+      "learning_rate": 9.635645276065293e-06,
+      "loss": 0.5571,
+      "step": 3204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1952558864052625,
+      "learning_rate": 9.630328635822835e-06,
+      "loss": 0.5376,
+      "step": 3205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5015714346639317,
+      "learning_rate": 9.625012100215078e-06,
+      "loss": 0.5864,
+      "step": 3206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7369255818667335,
+      "learning_rate": 9.61969567074686e-06,
+      "loss": 0.6041,
+      "step": 3207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5671629594170011,
+      "learning_rate": 9.61437934892298e-06,
+      "loss": 0.6337,
+      "step": 3208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8658076550677958,
+      "learning_rate": 9.609063136248214e-06,
+      "loss": 0.567,
+      "step": 3209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.495232209684953,
+      "learning_rate": 9.603747034227301e-06,
+      "loss": 0.5095,
+      "step": 3210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.59906615679454,
+      "learning_rate": 9.598431044364963e-06,
+      "loss": 0.5711,
+      "step": 3211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7042279988187525,
+      "learning_rate": 9.593115168165868e-06,
+      "loss": 0.5174,
+      "step": 3212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6938917455792322,
+      "learning_rate": 9.587799407134672e-06,
+      "loss": 0.4778,
+      "step": 3213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6296935247026627,
+      "learning_rate": 9.582483762775987e-06,
+      "loss": 0.5964,
+      "step": 3214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.519647736249719,
+      "learning_rate": 9.577168236594393e-06,
+      "loss": 0.6456,
+      "step": 3215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4584167080342962,
+      "learning_rate": 9.571852830094439e-06,
+      "loss": 0.5074,
+      "step": 3216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0961441706081323,
+      "learning_rate": 9.566537544780641e-06,
+      "loss": 0.5653,
+      "step": 3217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.726411423117264,
+      "learning_rate": 9.56122238215748e-06,
+      "loss": 0.5816,
+      "step": 3218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7698989076061056,
+      "learning_rate": 9.555907343729402e-06,
+      "loss": 0.4979,
+      "step": 3219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5691581098806158,
+      "learning_rate": 9.550592431000814e-06,
+      "loss": 0.6128,
+      "step": 3220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4849414939022985,
+      "learning_rate": 9.545277645476094e-06,
+      "loss": 0.5487,
+      "step": 3221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.710678025126802,
+      "learning_rate": 9.53996298865958e-06,
+      "loss": 0.5415,
+      "step": 3222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7510032198014627,
+      "learning_rate": 9.534648462055576e-06,
+      "loss": 0.5574,
+      "step": 3223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.899814703433351,
+      "learning_rate": 9.529334067168344e-06,
+      "loss": 0.5161,
+      "step": 3224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7015274839032348,
+      "learning_rate": 9.524019805502113e-06,
+      "loss": 0.5795,
+      "step": 3225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.927620623870373,
+      "learning_rate": 9.518705678561075e-06,
+      "loss": 0.5671,
+      "step": 3226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.543692369621684,
+      "learning_rate": 9.513391687849383e-06,
+      "loss": 0.4911,
+      "step": 3227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9067214619188344,
+      "learning_rate": 9.50807783487115e-06,
+      "loss": 0.5553,
+      "step": 3228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5351544231486136,
+      "learning_rate": 9.50276412113045e-06,
+      "loss": 0.6153,
+      "step": 3229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4489275430163022,
+      "learning_rate": 9.497450548131319e-06,
+      "loss": 0.4942,
+      "step": 3230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6267414472643464,
+      "learning_rate": 9.492137117377755e-06,
+      "loss": 0.5082,
+      "step": 3231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.150879180683235,
+      "learning_rate": 9.48682383037371e-06,
+      "loss": 0.5298,
+      "step": 3232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8054336446200818,
+      "learning_rate": 9.481510688623098e-06,
+      "loss": 0.5571,
+      "step": 3233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7623403200260377,
+      "learning_rate": 9.476197693629798e-06,
+      "loss": 0.597,
+      "step": 3234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6584171602298516,
+      "learning_rate": 9.470884846897638e-06,
+      "loss": 0.5783,
+      "step": 3235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0337562395211775,
+      "learning_rate": 9.465572149930408e-06,
+      "loss": 0.4881,
+      "step": 3236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1740946876178544,
+      "learning_rate": 9.460259604231859e-06,
+      "loss": 0.5948,
+      "step": 3237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4891810822163827,
+      "learning_rate": 9.454947211305691e-06,
+      "loss": 0.5308,
+      "step": 3238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2825127774368603,
+      "learning_rate": 9.44963497265557e-06,
+      "loss": 0.4827,
+      "step": 3239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5823809918052159,
+      "learning_rate": 9.444322889785111e-06,
+      "loss": 0.5381,
+      "step": 3240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.987558518410023,
+      "learning_rate": 9.43901096419789e-06,
+      "loss": 0.5647,
+      "step": 3241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.139212624028011,
+      "learning_rate": 9.433699197397435e-06,
+      "loss": 0.524,
+      "step": 3242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6755645935896442,
+      "learning_rate": 9.42838759088723e-06,
+      "loss": 0.6714,
+      "step": 3243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9197419068615393,
+      "learning_rate": 9.423076146170718e-06,
+      "loss": 0.6699,
+      "step": 3244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5630459624113135,
+      "learning_rate": 9.417764864751287e-06,
+      "loss": 0.4335,
+      "step": 3245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7029891714709322,
+      "learning_rate": 9.412453748132286e-06,
+      "loss": 0.5721,
+      "step": 3246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7945415126681972,
+      "learning_rate": 9.407142797817014e-06,
+      "loss": 0.5454,
+      "step": 3247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8056988403587422,
+      "learning_rate": 9.401832015308728e-06,
+      "loss": 0.5606,
+      "step": 3248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0218741172733035,
+      "learning_rate": 9.39652140211063e-06,
+      "loss": 0.5348,
+      "step": 3249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.894294447698885,
+      "learning_rate": 9.39121095972588e-06,
+      "loss": 0.6041,
+      "step": 3250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8155868966456616,
+      "learning_rate": 9.385900689657588e-06,
+      "loss": 0.6942,
+      "step": 3251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.754170085682439,
+      "learning_rate": 9.380590593408813e-06,
+      "loss": 0.566,
+      "step": 3252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5527650360925715,
+      "learning_rate": 9.375280672482567e-06,
+      "loss": 0.4963,
+      "step": 3253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9415257584495524,
+      "learning_rate": 9.369970928381813e-06,
+      "loss": 0.5492,
+      "step": 3254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3682221348365369,
+      "learning_rate": 9.364661362609464e-06,
+      "loss": 0.4942,
+      "step": 3255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5352527482138425,
+      "learning_rate": 9.359351976668377e-06,
+      "loss": 0.5201,
+      "step": 3256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.013758995717571,
+      "learning_rate": 9.354042772061362e-06,
+      "loss": 0.5246,
+      "step": 3257
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8811014396380015,
+      "learning_rate": 9.348733750291186e-06,
+      "loss": 0.529,
+      "step": 3258
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.573340875594774,
+      "learning_rate": 9.343424912860552e-06,
+      "loss": 0.5629,
+      "step": 3259
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5947184532490284,
+      "learning_rate": 9.338116261272114e-06,
+      "loss": 0.5174,
+      "step": 3260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6122776316755654,
+      "learning_rate": 9.332807797028476e-06,
+      "loss": 0.5073,
+      "step": 3261
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.621083489017127,
+      "learning_rate": 9.327499521632187e-06,
+      "loss": 0.5183,
+      "step": 3262
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.880997376165174,
+      "learning_rate": 9.322191436585745e-06,
+      "loss": 0.5023,
+      "step": 3263
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6988553723592925,
+      "learning_rate": 9.316883543391589e-06,
+      "loss": 0.4955,
+      "step": 3264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8731881129831898,
+      "learning_rate": 9.31157584355211e-06,
+      "loss": 0.5377,
+      "step": 3265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6728814932299276,
+      "learning_rate": 9.306268338569643e-06,
+      "loss": 0.5656,
+      "step": 3266
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6762811543367973,
+      "learning_rate": 9.30096102994646e-06,
+      "loss": 0.5508,
+      "step": 3267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.889483021394582,
+      "learning_rate": 9.295653919184787e-06,
+      "loss": 0.6367,
+      "step": 3268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8456598028865425,
+      "learning_rate": 9.290347007786791e-06,
+      "loss": 0.5194,
+      "step": 3269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7503786483648445,
+      "learning_rate": 9.28504029725458e-06,
+      "loss": 0.452,
+      "step": 3270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6967007688715403,
+      "learning_rate": 9.27973378909021e-06,
+      "loss": 0.4595,
+      "step": 3271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.569200961287285,
+      "learning_rate": 9.274427484795676e-06,
+      "loss": 0.5823,
+      "step": 3272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.237717546181913,
+      "learning_rate": 9.269121385872915e-06,
+      "loss": 0.5068,
+      "step": 3273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6188565488619466,
+      "learning_rate": 9.263815493823808e-06,
+      "loss": 0.5533,
+      "step": 3274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9772758521868912,
+      "learning_rate": 9.258509810150177e-06,
+      "loss": 0.4692,
+      "step": 3275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6551666080262974,
+      "learning_rate": 9.253204336353786e-06,
+      "loss": 0.5342,
+      "step": 3276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9232408890218458,
+      "learning_rate": 9.247899073936334e-06,
+      "loss": 0.6086,
+      "step": 3277
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3574598291888251,
+      "learning_rate": 9.242594024399467e-06,
+      "loss": 0.5061,
+      "step": 3278
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6815311836048983,
+      "learning_rate": 9.237289189244769e-06,
+      "loss": 0.5228,
+      "step": 3279
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.182918429205142,
+      "learning_rate": 9.23198456997376e-06,
+      "loss": 0.5512,
+      "step": 3280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8944711788498716,
+      "learning_rate": 9.226680168087903e-06,
+      "loss": 0.5525,
+      "step": 3281
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.652619225930231,
+      "learning_rate": 9.221375985088597e-06,
+      "loss": 0.5805,
+      "step": 3282
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6118259593594406,
+      "learning_rate": 9.216072022477183e-06,
+      "loss": 0.5449,
+      "step": 3283
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3661850411083876,
+      "learning_rate": 9.210768281754931e-06,
+      "loss": 0.4383,
+      "step": 3284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7541516555209038,
+      "learning_rate": 9.205464764423059e-06,
+      "loss": 0.5316,
+      "step": 3285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9010789333820435,
+      "learning_rate": 9.200161471982713e-06,
+      "loss": 0.546,
+      "step": 3286
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453332725577512,
+      "learning_rate": 9.194858405934983e-06,
+      "loss": 0.5358,
+      "step": 3287
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8249845034854448,
+      "learning_rate": 9.189555567780882e-06,
+      "loss": 0.6651,
+      "step": 3288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6193465733950025,
+      "learning_rate": 9.184252959021374e-06,
+      "loss": 0.5165,
+      "step": 3289
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7405139225864776,
+      "learning_rate": 9.17895058115735e-06,
+      "loss": 0.6718,
+      "step": 3290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6090541095034165,
+      "learning_rate": 9.173648435689637e-06,
+      "loss": 0.5511,
+      "step": 3291
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7605297469507193,
+      "learning_rate": 9.168346524118994e-06,
+      "loss": 0.6073,
+      "step": 3292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.35628576911371,
+      "learning_rate": 9.16304484794612e-06,
+      "loss": 0.5411,
+      "step": 3293
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7923730051532465,
+      "learning_rate": 9.15774340867164e-06,
+      "loss": 0.5792,
+      "step": 3294
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.602777962751145,
+      "learning_rate": 9.152442207796115e-06,
+      "loss": 0.5372,
+      "step": 3295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7111703252834505,
+      "learning_rate": 9.147141246820042e-06,
+      "loss": 0.4518,
+      "step": 3296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4251631750959364,
+      "learning_rate": 9.141840527243844e-06,
+      "loss": 0.5367,
+      "step": 3297
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6132920776304853,
+      "learning_rate": 9.13654005056788e-06,
+      "loss": 0.5967,
+      "step": 3298
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5655199219646183,
+      "learning_rate": 9.131239818292438e-06,
+      "loss": 0.548,
+      "step": 3299
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.894958098321758,
+      "learning_rate": 9.125939831917738e-06,
+      "loss": 0.5734,
+      "step": 3300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3940579085150697,
+      "learning_rate": 9.120640092943929e-06,
+      "loss": 0.5117,
+      "step": 3301
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8937961915264145,
+      "learning_rate": 9.11534060287109e-06,
+      "loss": 0.6466,
+      "step": 3302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.522122303207686,
+      "learning_rate": 9.110041363199233e-06,
+      "loss": 0.4082,
+      "step": 3303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.016444365839042,
+      "learning_rate": 9.104742375428297e-06,
+      "loss": 0.5009,
+      "step": 3304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6764577514451937,
+      "learning_rate": 9.099443641058147e-06,
+      "loss": 0.5662,
+      "step": 3305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9791861875052839,
+      "learning_rate": 9.094145161588582e-06,
+      "loss": 0.5811,
+      "step": 3306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5561737157592388,
+      "learning_rate": 9.088846938519322e-06,
+      "loss": 0.6178,
+      "step": 3307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6501992327723762,
+      "learning_rate": 9.083548973350019e-06,
+      "loss": 0.5092,
+      "step": 3308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9637025914969979,
+      "learning_rate": 9.078251267580256e-06,
+      "loss": 0.5741,
+      "step": 3309
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7096925745617253,
+      "learning_rate": 9.072953822709526e-06,
+      "loss": 0.5563,
+      "step": 3310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8397807165935798,
+      "learning_rate": 9.067656640237267e-06,
+      "loss": 0.6416,
+      "step": 3311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5425838964510556,
+      "learning_rate": 9.062359721662836e-06,
+      "loss": 0.4604,
+      "step": 3312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2496903501548027,
+      "learning_rate": 9.057063068485513e-06,
+      "loss": 0.5462,
+      "step": 3313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0081668033193782,
+      "learning_rate": 9.051766682204504e-06,
+      "loss": 0.5187,
+      "step": 3314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5133730272120545,
+      "learning_rate": 9.04647056431894e-06,
+      "loss": 0.5715,
+      "step": 3315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7754489506207665,
+      "learning_rate": 9.041174716327879e-06,
+      "loss": 0.5315,
+      "step": 3316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6140462165500156,
+      "learning_rate": 9.035879139730294e-06,
+      "loss": 0.5009,
+      "step": 3317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7834063235372042,
+      "learning_rate": 9.030583836025093e-06,
+      "loss": 0.5957,
+      "step": 3318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6392959321711074,
+      "learning_rate": 9.025288806711096e-06,
+      "loss": 0.5809,
+      "step": 3319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7312037033964014,
+      "learning_rate": 9.019994053287053e-06,
+      "loss": 0.5155,
+      "step": 3320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6386732144038052,
+      "learning_rate": 9.014699577251631e-06,
+      "loss": 0.5969,
+      "step": 3321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.876489959847085,
+      "learning_rate": 9.009405380103422e-06,
+      "loss": 0.5285,
+      "step": 3322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7886889550288358,
+      "learning_rate": 9.004111463340935e-06,
+      "loss": 0.4883,
+      "step": 3323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.326042968364947,
+      "learning_rate": 8.998817828462603e-06,
+      "loss": 0.4197,
+      "step": 3324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6620703780937782,
+      "learning_rate": 8.993524476966779e-06,
+      "loss": 0.6067,
+      "step": 3325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7116737973961964,
+      "learning_rate": 8.988231410351731e-06,
+      "loss": 0.5512,
+      "step": 3326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9290568598431872,
+      "learning_rate": 8.982938630115657e-06,
+      "loss": 0.5613,
+      "step": 3327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4617105720173846,
+      "learning_rate": 8.977646137756662e-06,
+      "loss": 0.5997,
+      "step": 3328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8295296902916223,
+      "learning_rate": 8.97235393477278e-06,
+      "loss": 0.4907,
+      "step": 3329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.829575122362265,
+      "learning_rate": 8.967062022661952e-06,
+      "loss": 0.5676,
+      "step": 3330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.903001872734973,
+      "learning_rate": 8.961770402922052e-06,
+      "loss": 0.4823,
+      "step": 3331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8545586247250658,
+      "learning_rate": 8.956479077050849e-06,
+      "loss": 0.6222,
+      "step": 3332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8661197605930604,
+      "learning_rate": 8.951188046546048e-06,
+      "loss": 0.5161,
+      "step": 3333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7607582668043766,
+      "learning_rate": 8.945897312905265e-06,
+      "loss": 0.5379,
+      "step": 3334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.778000313992945,
+      "learning_rate": 8.940606877626028e-06,
+      "loss": 0.6084,
+      "step": 3335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9711614449543151,
+      "learning_rate": 8.935316742205787e-06,
+      "loss": 0.5608,
+      "step": 3336
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.860851023386395,
+      "learning_rate": 8.930026908141902e-06,
+      "loss": 0.5213,
+      "step": 3337
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.571635386288547,
+      "learning_rate": 8.924737376931651e-06,
+      "loss": 0.4704,
+      "step": 3338
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.65771253937124,
+      "learning_rate": 8.919448150072221e-06,
+      "loss": 0.5499,
+      "step": 3339
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7689837052348327,
+      "learning_rate": 8.91415922906072e-06,
+      "loss": 0.5793,
+      "step": 3340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9185454189152538,
+      "learning_rate": 8.908870615394164e-06,
+      "loss": 0.5358,
+      "step": 3341
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.365945496911761,
+      "learning_rate": 8.903582310569487e-06,
+      "loss": 0.6188,
+      "step": 3342
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8619042590200968,
+      "learning_rate": 8.898294316083529e-06,
+      "loss": 0.6592,
+      "step": 3343
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0555401460916825,
+      "learning_rate": 8.893006633433048e-06,
+      "loss": 0.5143,
+      "step": 3344
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1380170153987157,
+      "learning_rate": 8.887719264114709e-06,
+      "loss": 0.5469,
+      "step": 3345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6769058250057731,
+      "learning_rate": 8.882432209625092e-06,
+      "loss": 0.613,
+      "step": 3346
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7735909585799181,
+      "learning_rate": 8.877145471460688e-06,
+      "loss": 0.6808,
+      "step": 3347
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9786690655245478,
+      "learning_rate": 8.871859051117896e-06,
+      "loss": 0.5991,
+      "step": 3348
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4487613327030264,
+      "learning_rate": 8.866572950093026e-06,
+      "loss": 0.5105,
+      "step": 3349
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.396938166181115,
+      "learning_rate": 8.861287169882295e-06,
+      "loss": 0.5109,
+      "step": 3350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0659712957750163,
+      "learning_rate": 8.856001711981839e-06,
+      "loss": 0.5338,
+      "step": 3351
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6717207024762328,
+      "learning_rate": 8.850716577887695e-06,
+      "loss": 0.6468,
+      "step": 3352
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4959261069798924,
+      "learning_rate": 8.8454317690958e-06,
+      "loss": 0.4674,
+      "step": 3353
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6336075452253747,
+      "learning_rate": 8.840147287102016e-06,
+      "loss": 0.5199,
+      "step": 3354
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2715641457023594,
+      "learning_rate": 8.8348631334021e-06,
+      "loss": 0.4757,
+      "step": 3355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5460799332239503,
+      "learning_rate": 8.829579309491724e-06,
+      "loss": 0.4707,
+      "step": 3356
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7610486078769287,
+      "learning_rate": 8.824295816866463e-06,
+      "loss": 0.4647,
+      "step": 3357
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.962456662500166,
+      "learning_rate": 8.819012657021794e-06,
+      "loss": 0.577,
+      "step": 3358
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1193370948819803,
+      "learning_rate": 8.81372983145311e-06,
+      "loss": 0.6124,
+      "step": 3359
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.006097728507301,
+      "learning_rate": 8.808447341655703e-06,
+      "loss": 0.5663,
+      "step": 3360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6438266115233746,
+      "learning_rate": 8.803165189124768e-06,
+      "loss": 0.5284,
+      "step": 3361
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.641268687014172,
+      "learning_rate": 8.79788337535541e-06,
+      "loss": 0.4299,
+      "step": 3362
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4852241347336061,
+      "learning_rate": 8.79260190184263e-06,
+      "loss": 0.5411,
+      "step": 3363
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.645935878832186,
+      "learning_rate": 8.787320770081345e-06,
+      "loss": 0.5304,
+      "step": 3364
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.252101068132057,
+      "learning_rate": 8.782039981566364e-06,
+      "loss": 0.5487,
+      "step": 3365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.913984542416388,
+      "learning_rate": 8.776759537792402e-06,
+      "loss": 0.5139,
+      "step": 3366
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8102201669022886,
+      "learning_rate": 8.771479440254082e-06,
+      "loss": 0.534,
+      "step": 3367
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.955418257360871,
+      "learning_rate": 8.76619969044592e-06,
+      "loss": 0.5493,
+      "step": 3368
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.249706003285317,
+      "learning_rate": 8.760920289862341e-06,
+      "loss": 0.5838,
+      "step": 3369
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5805245134125003,
+      "learning_rate": 8.755641239997667e-06,
+      "loss": 0.5378,
+      "step": 3370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9319255468949066,
+      "learning_rate": 8.75036254234612e-06,
+      "loss": 0.5173,
+      "step": 3371
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5434593990516787,
+      "learning_rate": 8.745084198401828e-06,
+      "loss": 0.5579,
+      "step": 3372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5540779848075361,
+      "learning_rate": 8.739806209658812e-06,
+      "loss": 0.5174,
+      "step": 3373
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7987101783328536,
+      "learning_rate": 8.734528577611004e-06,
+      "loss": 0.4562,
+      "step": 3374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7307255254050988,
+      "learning_rate": 8.729251303752214e-06,
+      "loss": 0.6146,
+      "step": 3375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6050902973769752,
+      "learning_rate": 8.72397438957617e-06,
+      "loss": 0.4917,
+      "step": 3376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5121575562999903,
+      "learning_rate": 8.71869783657649e-06,
+      "loss": 0.6048,
+      "step": 3377
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.406316326841981,
+      "learning_rate": 8.713421646246692e-06,
+      "loss": 0.5939,
+      "step": 3378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6472641429351673,
+      "learning_rate": 8.70814582008019e-06,
+      "loss": 0.5963,
+      "step": 3379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6597719003937184,
+      "learning_rate": 8.702870359570296e-06,
+      "loss": 0.4542,
+      "step": 3380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9076279698673673,
+      "learning_rate": 8.697595266210217e-06,
+      "loss": 0.5735,
+      "step": 3381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1697302027984273,
+      "learning_rate": 8.692320541493058e-06,
+      "loss": 0.5843,
+      "step": 3382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.134872431544596,
+      "learning_rate": 8.687046186911819e-06,
+      "loss": 0.5236,
+      "step": 3383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5915816754919383,
+      "learning_rate": 8.681772203959395e-06,
+      "loss": 0.5312,
+      "step": 3384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7534524980832722,
+      "learning_rate": 8.676498594128576e-06,
+      "loss": 0.4924,
+      "step": 3385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5968324120589297,
+      "learning_rate": 8.671225358912044e-06,
+      "loss": 0.5288,
+      "step": 3386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.708071900594268,
+      "learning_rate": 8.665952499802379e-06,
+      "loss": 0.6152,
+      "step": 3387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.792109990274148,
+      "learning_rate": 8.660680018292053e-06,
+      "loss": 0.5979,
+      "step": 3388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0006176325500316,
+      "learning_rate": 8.65540791587343e-06,
+      "loss": 0.5382,
+      "step": 3389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8719756071633724,
+      "learning_rate": 8.650136194038767e-06,
+      "loss": 0.6037,
+      "step": 3390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.914575819168055,
+      "learning_rate": 8.644864854280214e-06,
+      "loss": 0.61,
+      "step": 3391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.945936613134799,
+      "learning_rate": 8.639593898089815e-06,
+      "loss": 0.5806,
+      "step": 3392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4876316162427483,
+      "learning_rate": 8.634323326959501e-06,
+      "loss": 0.4676,
+      "step": 3393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0200407294359968,
+      "learning_rate": 8.629053142381093e-06,
+      "loss": 0.5845,
+      "step": 3394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8737147793767153,
+      "learning_rate": 8.623783345846313e-06,
+      "loss": 0.5774,
+      "step": 3395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.063052533148994,
+      "learning_rate": 8.618513938846763e-06,
+      "loss": 0.5211,
+      "step": 3396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8083715444131756,
+      "learning_rate": 8.613244922873932e-06,
+      "loss": 0.5711,
+      "step": 3397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.483949160793095,
+      "learning_rate": 8.60797629941921e-06,
+      "loss": 0.4667,
+      "step": 3398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1672848222816046,
+      "learning_rate": 8.602708069973866e-06,
+      "loss": 0.5279,
+      "step": 3399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.778784798193256,
+      "learning_rate": 8.597440236029064e-06,
+      "loss": 0.5651,
+      "step": 3400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.771345927178479,
+      "learning_rate": 8.592172799075853e-06,
+      "loss": 0.4457,
+      "step": 3401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6729616871958328,
+      "learning_rate": 8.586905760605169e-06,
+      "loss": 0.5343,
+      "step": 3402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.192308577704464,
+      "learning_rate": 8.581639122107837e-06,
+      "loss": 0.5613,
+      "step": 3403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0628080109452,
+      "learning_rate": 8.576372885074567e-06,
+      "loss": 0.5215,
+      "step": 3404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8643477545770224,
+      "learning_rate": 8.571107050995955e-06,
+      "loss": 0.6952,
+      "step": 3405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6149255432110794,
+      "learning_rate": 8.565841621362488e-06,
+      "loss": 0.5514,
+      "step": 3406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0059864214289336,
+      "learning_rate": 8.560576597664533e-06,
+      "loss": 0.5322,
+      "step": 3407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5437824868770607,
+      "learning_rate": 8.555311981392342e-06,
+      "loss": 0.4937,
+      "step": 3408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.637318146866671,
+      "learning_rate": 8.550047774036058e-06,
+      "loss": 0.6439,
+      "step": 3409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5810316328392526,
+      "learning_rate": 8.5447839770857e-06,
+      "loss": 0.5983,
+      "step": 3410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.751538409621543,
+      "learning_rate": 8.539520592031176e-06,
+      "loss": 0.5084,
+      "step": 3411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6110515909151735,
+      "learning_rate": 8.534257620362277e-06,
+      "loss": 0.5311,
+      "step": 3412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6903443421611735,
+      "learning_rate": 8.528995063568673e-06,
+      "loss": 0.4906,
+      "step": 3413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.705072721314062,
+      "learning_rate": 8.523732923139922e-06,
+      "loss": 0.5176,
+      "step": 3414
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5604654799192827,
+      "learning_rate": 8.518471200565461e-06,
+      "loss": 0.5206,
+      "step": 3415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.982135391643175,
+      "learning_rate": 8.513209897334612e-06,
+      "loss": 0.5844,
+      "step": 3416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.821681824950679,
+      "learning_rate": 8.507949014936573e-06,
+      "loss": 0.7187,
+      "step": 3417
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.910025468405713,
+      "learning_rate": 8.502688554860426e-06,
+      "loss": 0.5936,
+      "step": 3418
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7680479480762976,
+      "learning_rate": 8.497428518595132e-06,
+      "loss": 0.4889,
+      "step": 3419
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.47165089471029,
+      "learning_rate": 8.492168907629534e-06,
+      "loss": 0.4243,
+      "step": 3420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8573195142506258,
+      "learning_rate": 8.486909723452356e-06,
+      "loss": 0.5376,
+      "step": 3421
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7131637315434127,
+      "learning_rate": 8.481650967552199e-06,
+      "loss": 0.605,
+      "step": 3422
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.218394453246338,
+      "learning_rate": 8.47639264141754e-06,
+      "loss": 0.5668,
+      "step": 3423
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6156326717033958,
+      "learning_rate": 8.471134746536737e-06,
+      "loss": 0.5202,
+      "step": 3424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9490268376028494,
+      "learning_rate": 8.465877284398029e-06,
+      "loss": 0.5715,
+      "step": 3425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5054576072019104,
+      "learning_rate": 8.460620256489528e-06,
+      "loss": 0.4976,
+      "step": 3426
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7013591200963036,
+      "learning_rate": 8.455363664299225e-06,
+      "loss": 0.5709,
+      "step": 3427
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7821486413904941,
+      "learning_rate": 8.450107509314983e-06,
+      "loss": 0.5857,
+      "step": 3428
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7744998006844914,
+      "learning_rate": 8.444851793024555e-06,
+      "loss": 0.5518,
+      "step": 3429
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7418535269092996,
+      "learning_rate": 8.439596516915553e-06,
+      "loss": 0.5819,
+      "step": 3430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1116832810381574,
+      "learning_rate": 8.434341682475476e-06,
+      "loss": 0.5123,
+      "step": 3431
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2979085347848125,
+      "learning_rate": 8.42908729119169e-06,
+      "loss": 0.48,
+      "step": 3432
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6086716298691839,
+      "learning_rate": 8.423833344551443e-06,
+      "loss": 0.4895,
+      "step": 3433
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2654607417118813,
+      "learning_rate": 8.418579844041852e-06,
+      "loss": 0.6031,
+      "step": 3434
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7552196906484305,
+      "learning_rate": 8.413326791149909e-06,
+      "loss": 0.5714,
+      "step": 3435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5502361443852846,
+      "learning_rate": 8.408074187362479e-06,
+      "loss": 0.559,
+      "step": 3436
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2701029333100227,
+      "learning_rate": 8.402822034166301e-06,
+      "loss": 0.5875,
+      "step": 3437
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7377878672521394,
+      "learning_rate": 8.397570333047985e-06,
+      "loss": 0.5176,
+      "step": 3438
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.022909441784619,
+      "learning_rate": 8.392319085494018e-06,
+      "loss": 0.5303,
+      "step": 3439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6100387919555392,
+      "learning_rate": 8.38706829299075e-06,
+      "loss": 0.5764,
+      "step": 3440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6557137390904924,
+      "learning_rate": 8.381817957024409e-06,
+      "loss": 0.5535,
+      "step": 3441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6041817869707817,
+      "learning_rate": 8.37656807908109e-06,
+      "loss": 0.5162,
+      "step": 3442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1175766894196446,
+      "learning_rate": 8.37131866064676e-06,
+      "loss": 0.5203,
+      "step": 3443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6750350485012122,
+      "learning_rate": 8.366069703207257e-06,
+      "loss": 0.5132,
+      "step": 3444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.949508106859869,
+      "learning_rate": 8.360821208248289e-06,
+      "loss": 0.4818,
+      "step": 3445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8633021253724802,
+      "learning_rate": 8.355573177255428e-06,
+      "loss": 0.687,
+      "step": 3446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.706227551201607,
+      "learning_rate": 8.35032561171412e-06,
+      "loss": 0.592,
+      "step": 3447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5015835744712493,
+      "learning_rate": 8.345078513109677e-06,
+      "loss": 0.5361,
+      "step": 3448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.023477560301981,
+      "learning_rate": 8.339831882927279e-06,
+      "loss": 0.4652,
+      "step": 3449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0310817248068194,
+      "learning_rate": 8.334585722651973e-06,
+      "loss": 0.6043,
+      "step": 3450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7852961856966671,
+      "learning_rate": 8.329340033768672e-06,
+      "loss": 0.7047,
+      "step": 3451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4882852350808768,
+      "learning_rate": 8.324094817762164e-06,
+      "loss": 0.5468,
+      "step": 3452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7314810119382618,
+      "learning_rate": 8.31885007611709e-06,
+      "loss": 0.4691,
+      "step": 3453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4088674496419213,
+      "learning_rate": 8.313605810317967e-06,
+      "loss": 0.5321,
+      "step": 3454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7467591016100485,
+      "learning_rate": 8.30836202184917e-06,
+      "loss": 0.4902,
+      "step": 3455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5951598803641511,
+      "learning_rate": 8.303118712194944e-06,
+      "loss": 0.5295,
+      "step": 3456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7101418091741691,
+      "learning_rate": 8.297875882839397e-06,
+      "loss": 0.6311,
+      "step": 3457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.081993921364836,
+      "learning_rate": 8.2926335352665e-06,
+      "loss": 0.4844,
+      "step": 3458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7276374503210548,
+      "learning_rate": 8.28739167096009e-06,
+      "loss": 0.5309,
+      "step": 3459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.709124751296015,
+      "learning_rate": 8.282150291403867e-06,
+      "loss": 0.5161,
+      "step": 3460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6475069097977042,
+      "learning_rate": 8.276909398081387e-06,
+      "loss": 0.568,
+      "step": 3461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8373880382945886,
+      "learning_rate": 8.271668992476077e-06,
+      "loss": 0.6748,
+      "step": 3462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8059120364876717,
+      "learning_rate": 8.266429076071221e-06,
+      "loss": 0.5096,
+      "step": 3463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.692060656478298,
+      "learning_rate": 8.261189650349969e-06,
+      "loss": 0.4236,
+      "step": 3464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.791887362191388,
+      "learning_rate": 8.255950716795328e-06,
+      "loss": 0.6147,
+      "step": 3465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7746924607997099,
+      "learning_rate": 8.250712276890168e-06,
+      "loss": 0.5607,
+      "step": 3466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.818850859894029,
+      "learning_rate": 8.245474332117219e-06,
+      "loss": 0.5052,
+      "step": 3467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6349223031169666,
+      "learning_rate": 8.240236883959067e-06,
+      "loss": 0.4904,
+      "step": 3468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.888011617941011,
+      "learning_rate": 8.234999933898164e-06,
+      "loss": 0.6347,
+      "step": 3469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15639751634192,
+      "learning_rate": 8.229763483416815e-06,
+      "loss": 0.4743,
+      "step": 3470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8530950391799448,
+      "learning_rate": 8.22452753399719e-06,
+      "loss": 0.614,
+      "step": 3471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6837890289639987,
+      "learning_rate": 8.219292087121309e-06,
+      "loss": 0.5009,
+      "step": 3472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7698491788068265,
+      "learning_rate": 8.214057144271058e-06,
+      "loss": 0.4945,
+      "step": 3473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8733405906433145,
+      "learning_rate": 8.208822706928172e-06,
+      "loss": 0.5819,
+      "step": 3474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9722872670814056,
+      "learning_rate": 8.203588776574254e-06,
+      "loss": 0.58,
+      "step": 3475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.960731203144636,
+      "learning_rate": 8.198355354690752e-06,
+      "loss": 0.6114,
+      "step": 3476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9313572992013042,
+      "learning_rate": 8.193122442758977e-06,
+      "loss": 0.5214,
+      "step": 3477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8403403733043746,
+      "learning_rate": 8.187890042260094e-06,
+      "loss": 0.6078,
+      "step": 3478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4827603307962165,
+      "learning_rate": 8.18265815467512e-06,
+      "loss": 0.6314,
+      "step": 3479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.283416241867729,
+      "learning_rate": 8.177426781484933e-06,
+      "loss": 0.5617,
+      "step": 3480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0074942960201057,
+      "learning_rate": 8.172195924170263e-06,
+      "loss": 0.5476,
+      "step": 3481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5299380817137014,
+      "learning_rate": 8.166965584211694e-06,
+      "loss": 0.5757,
+      "step": 3482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7466499389268861,
+      "learning_rate": 8.161735763089654e-06,
+      "loss": 0.5831,
+      "step": 3483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5956981259639764,
+      "learning_rate": 8.15650646228444e-06,
+      "loss": 0.5082,
+      "step": 3484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6731363143220774,
+      "learning_rate": 8.151277683276196e-06,
+      "loss": 0.4378,
+      "step": 3485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7705783866959763,
+      "learning_rate": 8.146049427544912e-06,
+      "loss": 0.4836,
+      "step": 3486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.661425664909735,
+      "learning_rate": 8.140821696570439e-06,
+      "loss": 0.5118,
+      "step": 3487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.739715549927106,
+      "learning_rate": 8.13559449183247e-06,
+      "loss": 0.5955,
+      "step": 3488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8097221351419535,
+      "learning_rate": 8.130367814810561e-06,
+      "loss": 0.5041,
+      "step": 3489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7882662172677641,
+      "learning_rate": 8.125141666984107e-06,
+      "loss": 0.5598,
+      "step": 3490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8216494262152569,
+      "learning_rate": 8.119916049832362e-06,
+      "loss": 0.6614,
+      "step": 3491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5266073450653115,
+      "learning_rate": 8.114690964834422e-06,
+      "loss": 0.4515,
+      "step": 3492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9601338933118406,
+      "learning_rate": 8.109466413469238e-06,
+      "loss": 0.5676,
+      "step": 3493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.878752188693482,
+      "learning_rate": 8.104242397215609e-06,
+      "loss": 0.5965,
+      "step": 3494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.537707484045088,
+      "learning_rate": 8.09901891755218e-06,
+      "loss": 0.4917,
+      "step": 3495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1758154627626975,
+      "learning_rate": 8.093795975957449e-06,
+      "loss": 0.4934,
+      "step": 3496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.383345385584979,
+      "learning_rate": 8.088573573909755e-06,
+      "loss": 0.5962,
+      "step": 3497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1720006573919086,
+      "learning_rate": 8.083351712887288e-06,
+      "loss": 0.532,
+      "step": 3498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8036136677306138,
+      "learning_rate": 8.078130394368088e-06,
+      "loss": 0.7017,
+      "step": 3499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.274004135255686,
+      "learning_rate": 8.072909619830037e-06,
+      "loss": 0.6391,
+      "step": 3500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0217498814172834,
+      "learning_rate": 8.067689390750863e-06,
+      "loss": 0.4781,
+      "step": 3501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6936660756269928,
+      "learning_rate": 8.062469708608144e-06,
+      "loss": 0.5307,
+      "step": 3502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.785807668339425,
+      "learning_rate": 8.057250574879296e-06,
+      "loss": 0.6847,
+      "step": 3503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8726668909760567,
+      "learning_rate": 8.052031991041591e-06,
+      "loss": 0.6219,
+      "step": 3504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.551770653314682,
+      "learning_rate": 8.046813958572129e-06,
+      "loss": 0.5882,
+      "step": 3505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.737189963283567,
+      "learning_rate": 8.041596478947862e-06,
+      "loss": 0.4883,
+      "step": 3506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203364914427533,
+      "learning_rate": 8.036379553645595e-06,
+      "loss": 0.59,
+      "step": 3507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7332755695655848,
+      "learning_rate": 8.031163184141965e-06,
+      "loss": 0.5198,
+      "step": 3508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.202271871499354,
+      "learning_rate": 8.025947371913454e-06,
+      "loss": 0.4367,
+      "step": 3509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5013212453564195,
+      "learning_rate": 8.020732118436385e-06,
+      "loss": 0.5603,
+      "step": 3510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.762089648139194,
+      "learning_rate": 8.015517425186926e-06,
+      "loss": 0.491,
+      "step": 3511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2708576194682557,
+      "learning_rate": 8.010303293641086e-06,
+      "loss": 0.4666,
+      "step": 3512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8033715357535887,
+      "learning_rate": 8.005089725274711e-06,
+      "loss": 0.5195,
+      "step": 3513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5690413173568363,
+      "learning_rate": 7.999876721563494e-06,
+      "loss": 0.5277,
+      "step": 3514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5074561914705646,
+      "learning_rate": 7.994664283982962e-06,
+      "loss": 0.4952,
+      "step": 3515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.909557172317141,
+      "learning_rate": 7.989452414008485e-06,
+      "loss": 0.5302,
+      "step": 3516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7533556664107512,
+      "learning_rate": 7.984241113115275e-06,
+      "loss": 0.5471,
+      "step": 3517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9257927191728276,
+      "learning_rate": 7.979030382778376e-06,
+      "loss": 0.7242,
+      "step": 3518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09558488051042,
+      "learning_rate": 7.973820224472675e-06,
+      "loss": 0.6181,
+      "step": 3519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7894745682896213,
+      "learning_rate": 7.968610639672896e-06,
+      "loss": 0.6838,
+      "step": 3520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0504014427280772,
+      "learning_rate": 7.9634016298536e-06,
+      "loss": 0.5635,
+      "step": 3521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9667164805245865,
+      "learning_rate": 7.958193196489191e-06,
+      "loss": 0.4898,
+      "step": 3522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8133366670293734,
+      "learning_rate": 7.952985341053902e-06,
+      "loss": 0.6287,
+      "step": 3523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9053898540065697,
+      "learning_rate": 7.947778065021805e-06,
+      "loss": 0.511,
+      "step": 3524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0155801355285496,
+      "learning_rate": 7.942571369866814e-06,
+      "loss": 0.5201,
+      "step": 3525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.465522564033468,
+      "learning_rate": 7.937365257062664e-06,
+      "loss": 0.5141,
+      "step": 3526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8003929291114593,
+      "learning_rate": 7.932159728082938e-06,
+      "loss": 0.5542,
+      "step": 3527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.759897142099997,
+      "learning_rate": 7.92695478440105e-06,
+      "loss": 0.5047,
+      "step": 3528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.911211504352529,
+      "learning_rate": 7.921750427490248e-06,
+      "loss": 0.5722,
+      "step": 3529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5021783402754705,
+      "learning_rate": 7.916546658823618e-06,
+      "loss": 0.438,
+      "step": 3530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9592646192833214,
+      "learning_rate": 7.911343479874073e-06,
+      "loss": 0.5901,
+      "step": 3531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9330002245290705,
+      "learning_rate": 7.906140892114361e-06,
+      "loss": 0.4955,
+      "step": 3532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8465960016975482,
+      "learning_rate": 7.900938897017064e-06,
+      "loss": 0.5896,
+      "step": 3533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.120580943136161,
+      "learning_rate": 7.895737496054597e-06,
+      "loss": 0.5686,
+      "step": 3534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.160638257284747,
+      "learning_rate": 7.890536690699204e-06,
+      "loss": 0.6047,
+      "step": 3535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5504104933086853,
+      "learning_rate": 7.885336482422964e-06,
+      "loss": 0.5546,
+      "step": 3536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7911314795981015,
+      "learning_rate": 7.880136872697784e-06,
+      "loss": 0.537,
+      "step": 3537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8120008568433128,
+      "learning_rate": 7.874937862995401e-06,
+      "loss": 0.5628,
+      "step": 3538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6195926092430173,
+      "learning_rate": 7.869739454787385e-06,
+      "loss": 0.5464,
+      "step": 3539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.240020526748471,
+      "learning_rate": 7.864541649545135e-06,
+      "loss": 0.541,
+      "step": 3540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8209601902183015,
+      "learning_rate": 7.85934444873988e-06,
+      "loss": 0.5385,
+      "step": 3541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.116880358115982,
+      "learning_rate": 7.854147853842672e-06,
+      "loss": 0.6379,
+      "step": 3542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.895722500049924,
+      "learning_rate": 7.848951866324402e-06,
+      "loss": 0.6396,
+      "step": 3543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.022782980794684,
+      "learning_rate": 7.84375648765578e-06,
+      "loss": 0.5518,
+      "step": 3544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1252432198957996,
+      "learning_rate": 7.838561719307346e-06,
+      "loss": 0.53,
+      "step": 3545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9067267616800012,
+      "learning_rate": 7.833367562749473e-06,
+      "loss": 0.6899,
+      "step": 3546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.652583257809706,
+      "learning_rate": 7.828174019452357e-06,
+      "loss": 0.6105,
+      "step": 3547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5768391984833636,
+      "learning_rate": 7.822981090886011e-06,
+      "loss": 0.5772,
+      "step": 3548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.093911452229709,
+      "learning_rate": 7.817788778520288e-06,
+      "loss": 0.52,
+      "step": 3549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9754021832607982,
+      "learning_rate": 7.81259708382486e-06,
+      "loss": 0.5819,
+      "step": 3550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9859684882432416,
+      "learning_rate": 7.807406008269224e-06,
+      "loss": 0.5429,
+      "step": 3551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9104098110928212,
+      "learning_rate": 7.802215553322703e-06,
+      "loss": 0.5471,
+      "step": 3552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8647407744454765,
+      "learning_rate": 7.79702572045445e-06,
+      "loss": 0.6253,
+      "step": 3553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6139166565645384,
+      "learning_rate": 7.791836511133429e-06,
+      "loss": 0.404,
+      "step": 3554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.850019886463423,
+      "learning_rate": 7.78664792682844e-06,
+      "loss": 0.7015,
+      "step": 3555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8116694923729757,
+      "learning_rate": 7.781459969008098e-06,
+      "loss": 0.5621,
+      "step": 3556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.741123202445026,
+      "learning_rate": 7.776272639140845e-06,
+      "loss": 0.658,
+      "step": 3557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1066030834157994,
+      "learning_rate": 7.771085938694943e-06,
+      "loss": 0.5495,
+      "step": 3558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1041272292194355,
+      "learning_rate": 7.765899869138478e-06,
+      "loss": 0.5356,
+      "step": 3559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.85284893998669,
+      "learning_rate": 7.760714431939354e-06,
+      "loss": 0.4957,
+      "step": 3560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4344184705915706,
+      "learning_rate": 7.755529628565298e-06,
+      "loss": 0.5352,
+      "step": 3561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7598028771136496,
+      "learning_rate": 7.750345460483859e-06,
+      "loss": 0.4588,
+      "step": 3562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8670288557862424,
+      "learning_rate": 7.745161929162405e-06,
+      "loss": 0.58,
+      "step": 3563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9017750408781471,
+      "learning_rate": 7.739979036068125e-06,
+      "loss": 0.6764,
+      "step": 3564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.878143497641812,
+      "learning_rate": 7.734796782668021e-06,
+      "loss": 0.4795,
+      "step": 3565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0543000686973905,
+      "learning_rate": 7.729615170428923e-06,
+      "loss": 0.5818,
+      "step": 3566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5473277914591004,
+      "learning_rate": 7.724434200817473e-06,
+      "loss": 0.5417,
+      "step": 3567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9948144868485294,
+      "learning_rate": 7.719253875300138e-06,
+      "loss": 0.6381,
+      "step": 3568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4837925476041662,
+      "learning_rate": 7.7140741953432e-06,
+      "loss": 0.5437,
+      "step": 3569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.75010199364121,
+      "learning_rate": 7.708895162412745e-06,
+      "loss": 0.633,
+      "step": 3570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8947124770789903,
+      "learning_rate": 7.703716777974694e-06,
+      "loss": 0.589,
+      "step": 3571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5459675977964764,
+      "learning_rate": 7.69853904349478e-06,
+      "loss": 0.5108,
+      "step": 3572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9043748688215887,
+      "learning_rate": 7.693361960438548e-06,
+      "loss": 0.5439,
+      "step": 3573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8232053779686308,
+      "learning_rate": 7.688185530271359e-06,
+      "loss": 0.6502,
+      "step": 3574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7573324899689109,
+      "learning_rate": 7.683009754458394e-06,
+      "loss": 0.6801,
+      "step": 3575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4063813990224365,
+      "learning_rate": 7.67783463446464e-06,
+      "loss": 0.5179,
+      "step": 3576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.340839525878978,
+      "learning_rate": 7.67266017175491e-06,
+      "loss": 0.5152,
+      "step": 3577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2879968410639324,
+      "learning_rate": 7.667486367793822e-06,
+      "loss": 0.5583,
+      "step": 3578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5399718729815575,
+      "learning_rate": 7.66231322404581e-06,
+      "loss": 0.5505,
+      "step": 3579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375794180668671,
+      "learning_rate": 7.657140741975121e-06,
+      "loss": 0.5009,
+      "step": 3580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5046222534429594,
+      "learning_rate": 7.651968923045817e-06,
+      "loss": 0.494,
+      "step": 3581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8017031528159795,
+      "learning_rate": 7.64679776872177e-06,
+      "loss": 0.665,
+      "step": 3582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9796170604067969,
+      "learning_rate": 7.641627280466663e-06,
+      "loss": 0.5423,
+      "step": 3583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.865430355281834,
+      "learning_rate": 7.636457459743993e-06,
+      "loss": 0.4914,
+      "step": 3584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5583811763196118,
+      "learning_rate": 7.631288308017068e-06,
+      "loss": 0.4518,
+      "step": 3585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9881807259758681,
+      "learning_rate": 7.626119826749002e-06,
+      "loss": 0.4543,
+      "step": 3586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.410804111413249,
+      "learning_rate": 7.6209520174027255e-06,
+      "loss": 0.6288,
+      "step": 3587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.714625984137669,
+      "learning_rate": 7.615784881440975e-06,
+      "loss": 0.5087,
+      "step": 3588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8390352434287205,
+      "learning_rate": 7.610618420326299e-06,
+      "loss": 0.5092,
+      "step": 3589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7488746335995324,
+      "learning_rate": 7.605452635521054e-06,
+      "loss": 0.5924,
+      "step": 3590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.557148446190701,
+      "learning_rate": 7.6002875284874e-06,
+      "loss": 0.4749,
+      "step": 3591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6417035188947158,
+      "learning_rate": 7.595123100687313e-06,
+      "loss": 0.5226,
+      "step": 3592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0907361682820413,
+      "learning_rate": 7.589959353582574e-06,
+      "loss": 0.502,
+      "step": 3593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.653045118896521,
+      "learning_rate": 7.584796288634768e-06,
+      "loss": 0.4691,
+      "step": 3594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3262325350911857,
+      "learning_rate": 7.5796339073052915e-06,
+      "loss": 0.6544,
+      "step": 3595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0325124802058676,
+      "learning_rate": 7.574472211055346e-06,
+      "loss": 0.5786,
+      "step": 3596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7101428403853538,
+      "learning_rate": 7.569311201345939e-06,
+      "loss": 0.5805,
+      "step": 3597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9707107590499582,
+      "learning_rate": 7.564150879637882e-06,
+      "loss": 0.5758,
+      "step": 3598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7049030011160693,
+      "learning_rate": 7.558991247391792e-06,
+      "loss": 0.6033,
+      "step": 3599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7346984757925854,
+      "learning_rate": 7.553832306068095e-06,
+      "loss": 0.482,
+      "step": 3600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6455723743439103,
+      "learning_rate": 7.548674057127019e-06,
+      "loss": 0.5305,
+      "step": 3601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7516596832553197,
+      "learning_rate": 7.543516502028594e-06,
+      "loss": 0.5551,
+      "step": 3602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2451971090874525,
+      "learning_rate": 7.538359642232654e-06,
+      "loss": 0.5873,
+      "step": 3603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.801881336236046,
+      "learning_rate": 7.53320347919884e-06,
+      "loss": 0.5454,
+      "step": 3604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6416150906147013,
+      "learning_rate": 7.52804801438659e-06,
+      "loss": 0.5173,
+      "step": 3605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8323871808402792,
+      "learning_rate": 7.52289324925515e-06,
+      "loss": 0.508,
+      "step": 3606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6629585418437365,
+      "learning_rate": 7.517739185263564e-06,
+      "loss": 0.5038,
+      "step": 3607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8551186067045227,
+      "learning_rate": 7.5125858238706785e-06,
+      "loss": 0.5066,
+      "step": 3608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5331792272997211,
+      "learning_rate": 7.507433166535143e-06,
+      "loss": 0.6569,
+      "step": 3609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3841801591425087,
+      "learning_rate": 7.5022812147154065e-06,
+      "loss": 0.5021,
+      "step": 3610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6098409915226113,
+      "learning_rate": 7.497129969869718e-06,
+      "loss": 0.4921,
+      "step": 3611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0615885691870455,
+      "learning_rate": 7.491979433456127e-06,
+      "loss": 0.5141,
+      "step": 3612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9256461246968266,
+      "learning_rate": 7.486829606932478e-06,
+      "loss": 0.4354,
+      "step": 3613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1265600010172374,
+      "learning_rate": 7.481680491756424e-06,
+      "loss": 0.5259,
+      "step": 3614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.759212036492166,
+      "learning_rate": 7.476532089385407e-06,
+      "loss": 0.5657,
+      "step": 3615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8339709919170107,
+      "learning_rate": 7.471384401276674e-06,
+      "loss": 0.6958,
+      "step": 3616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.697583241933579,
+      "learning_rate": 7.466237428887265e-06,
+      "loss": 0.4921,
+      "step": 3617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5361802392999562,
+      "learning_rate": 7.461091173674022e-06,
+      "loss": 0.5953,
+      "step": 3618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3258367258492463,
+      "learning_rate": 7.455945637093581e-06,
+      "loss": 0.5002,
+      "step": 3619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.644174459395446,
+      "learning_rate": 7.450800820602375e-06,
+      "loss": 0.5244,
+      "step": 3620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7509620762861924,
+      "learning_rate": 7.445656725656634e-06,
+      "loss": 0.4705,
+      "step": 3621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.783320894225278,
+      "learning_rate": 7.440513353712381e-06,
+      "loss": 0.533,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.568789006631055,
+      "learning_rate": 7.43537070622544e-06,
+      "loss": 0.5069,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.259634880642734,
+      "learning_rate": 7.430228784651426e-06,
+      "loss": 0.5944,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8296405624644945,
+      "learning_rate": 7.425087590445747e-06,
+      "loss": 0.5544,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6220045351097996,
+      "learning_rate": 7.419947125063609e-06,
+      "loss": 0.5901,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.991206944732979,
+      "learning_rate": 7.41480738996001e-06,
+      "loss": 0.5226,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.926279536396569,
+      "learning_rate": 7.40966838658974e-06,
+      "loss": 0.5381,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6214461416716055,
+      "learning_rate": 7.4045301164073834e-06,
+      "loss": 0.5355,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.104767747125721,
+      "learning_rate": 7.399392580867317e-06,
+      "loss": 0.5195,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7817868448232517,
+      "learning_rate": 7.394255781423709e-06,
+      "loss": 0.4435,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.830975054141844,
+      "learning_rate": 7.389119719530522e-06,
+      "loss": 0.6133,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.147747611067911,
+      "learning_rate": 7.383984396641506e-06,
+      "loss": 0.5456,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.502928468367558,
+      "learning_rate": 7.378849814210201e-06,
+      "loss": 0.5575,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8692456252327736,
+      "learning_rate": 7.373715973689941e-06,
+      "loss": 0.5499,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7854344606922252,
+      "learning_rate": 7.3685828765338495e-06,
+      "loss": 0.4892,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8146011394223028,
+      "learning_rate": 7.363450524194839e-06,
+      "loss": 0.6798,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8120302358075315,
+      "learning_rate": 7.358318918125613e-06,
+      "loss": 0.5359,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5744054975529393,
+      "learning_rate": 7.353188059778657e-06,
+      "loss": 0.5643,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7856849763611481,
+      "learning_rate": 7.348057950606253e-06,
+      "loss": 0.6599,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6052357461857854,
+      "learning_rate": 7.342928592060468e-06,
+      "loss": 0.5592,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.149766117477306,
+      "learning_rate": 7.337799985593152e-06,
+      "loss": 0.5813,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.435077515458113,
+      "learning_rate": 7.332672132655953e-06,
+      "loss": 0.6115,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4266322048429505,
+      "learning_rate": 7.327545034700294e-06,
+      "loss": 0.4993,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6271345639412977,
+      "learning_rate": 7.3224186931773885e-06,
+      "loss": 0.4924,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6983506651641327,
+      "learning_rate": 7.317293109538239e-06,
+      "loss": 0.4876,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5682986295452557,
+      "learning_rate": 7.312168285233633e-06,
+      "loss": 0.5052,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.025154177683025,
+      "learning_rate": 7.307044221714139e-06,
+      "loss": 0.5687,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0671595702596743,
+      "learning_rate": 7.3019209204301115e-06,
+      "loss": 0.4594,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8560320392662988,
+      "learning_rate": 7.296798382831691e-06,
+      "loss": 0.5248,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6271264961177512,
+      "learning_rate": 7.291676610368803e-06,
+      "loss": 0.515,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.785687615436632,
+      "learning_rate": 7.286555604491151e-06,
+      "loss": 0.5171,
+      "step": 3652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.040008276328225,
+      "learning_rate": 7.2814353666482276e-06,
+      "loss": 0.6647,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8602650234310514,
+      "learning_rate": 7.276315898289303e-06,
+      "loss": 0.6921,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3935116807792647,
+      "learning_rate": 7.271197200863438e-06,
+      "loss": 0.6353,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.850804138325866,
+      "learning_rate": 7.2660792758194596e-06,
+      "loss": 0.4275,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.120868525246052,
+      "learning_rate": 7.260962124605993e-06,
+      "loss": 0.4974,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5236829919106327,
+      "learning_rate": 7.2558457486714316e-06,
+      "loss": 0.4617,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4542274855218147,
+      "learning_rate": 7.2507301494639605e-06,
+      "loss": 0.4414,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9713048552680097,
+      "learning_rate": 7.245615328431535e-06,
+      "loss": 0.5,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5119725907784023,
+      "learning_rate": 7.240501287021897e-06,
+      "loss": 0.515,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9788159969949508,
+      "learning_rate": 7.2353880266825635e-06,
+      "loss": 0.5704,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0243515874223723,
+      "learning_rate": 7.230275548860833e-06,
+      "loss": 0.5506,
+      "step": 3663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.146920875580349,
+      "learning_rate": 7.225163855003781e-06,
+      "loss": 0.5159,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.130279978362854,
+      "learning_rate": 7.220052946558262e-06,
+      "loss": 0.5617,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.673044535699793,
+      "learning_rate": 7.2149428249709095e-06,
+      "loss": 0.4702,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7301902810009366,
+      "learning_rate": 7.209833491688131e-06,
+      "loss": 0.5198,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6265114540883554,
+      "learning_rate": 7.2047249481561125e-06,
+      "loss": 0.4538,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9593044442134202,
+      "learning_rate": 7.1996171958208125e-06,
+      "loss": 0.5741,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.844566768937716,
+      "learning_rate": 7.194510236127978e-06,
+      "loss": 0.6714,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8594558596943567,
+      "learning_rate": 7.189404070523118e-06,
+      "loss": 0.6456,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0423477438920585,
+      "learning_rate": 7.184298700451524e-06,
+      "loss": 0.5798,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.515375807863017,
+      "learning_rate": 7.179194127358258e-06,
+      "loss": 0.5829,
+      "step": 3673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8811657311569907,
+      "learning_rate": 7.17409035268816e-06,
+      "loss": 0.5842,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7533421389171737,
+      "learning_rate": 7.168987377885843e-06,
+      "loss": 0.6173,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.70681749343873,
+      "learning_rate": 7.163885204395692e-06,
+      "loss": 0.5149,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.461904664021777,
+      "learning_rate": 7.158783833661869e-06,
+      "loss": 0.4318,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.782256862365506,
+      "learning_rate": 7.153683267128304e-06,
+      "loss": 0.5177,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0397481351429954,
+      "learning_rate": 7.148583506238701e-06,
+      "loss": 0.5677,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8138726107488903,
+      "learning_rate": 7.143484552436537e-06,
+      "loss": 0.5717,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9169640828902461,
+      "learning_rate": 7.1383864071650635e-06,
+      "loss": 0.5597,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.67611280839457,
+      "learning_rate": 7.133289071867295e-06,
+      "loss": 0.6099,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.347920854973419,
+      "learning_rate": 7.128192547986023e-06,
+      "loss": 0.5914,
+      "step": 3683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4995952196739704,
+      "learning_rate": 7.1230968369638096e-06,
+      "loss": 0.4571,
+      "step": 3684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.71376834769949,
+      "learning_rate": 7.118001940242984e-06,
+      "loss": 0.5313,
+      "step": 3685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5240055487436972,
+      "learning_rate": 7.112907859265646e-06,
+      "loss": 0.5618,
+      "step": 3686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7957981667097282,
+      "learning_rate": 7.1078145954736655e-06,
+      "loss": 0.6447,
+      "step": 3687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8353727613130395,
+      "learning_rate": 7.102722150308678e-06,
+      "loss": 0.5683,
+      "step": 3688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.197868268043004,
+      "learning_rate": 7.097630525212091e-06,
+      "loss": 0.5187,
+      "step": 3689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0471857835385086,
+      "learning_rate": 7.092539721625078e-06,
+      "loss": 0.6121,
+      "step": 3690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3721020736124045,
+      "learning_rate": 7.087449740988579e-06,
+      "loss": 0.5359,
+      "step": 3691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9055881371946997,
+      "learning_rate": 7.082360584743302e-06,
+      "loss": 0.573,
+      "step": 3692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8463033808012874,
+      "learning_rate": 7.077272254329726e-06,
+      "loss": 0.5601,
+      "step": 3693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9411930382098508,
+      "learning_rate": 7.072184751188088e-06,
+      "loss": 0.541,
+      "step": 3694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.198998955460798,
+      "learning_rate": 7.067098076758398e-06,
+      "loss": 0.5055,
+      "step": 3695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.886790970705043,
+      "learning_rate": 7.062012232480427e-06,
+      "loss": 0.5569,
+      "step": 3696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140446313357052,
+      "learning_rate": 7.056927219793711e-06,
+      "loss": 0.4903,
+      "step": 3697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9131480315344196,
+      "learning_rate": 7.051843040137558e-06,
+      "loss": 0.598,
+      "step": 3698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.864403640004033,
+      "learning_rate": 7.046759694951029e-06,
+      "loss": 0.5423,
+      "step": 3699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7403064555999235,
+      "learning_rate": 7.04167718567295e-06,
+      "loss": 0.5935,
+      "step": 3700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.239797237386876,
+      "learning_rate": 7.036595513741924e-06,
+      "loss": 0.5829,
+      "step": 3701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.522582846361151,
+      "learning_rate": 7.0315146805963004e-06,
+      "loss": 0.5708,
+      "step": 3702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6368879368467377,
+      "learning_rate": 7.026434687674204e-06,
+      "loss": 0.5249,
+      "step": 3703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7285142804886438,
+      "learning_rate": 7.021355536413513e-06,
+      "loss": 0.625,
+      "step": 3704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9458318424774024,
+      "learning_rate": 7.016277228251871e-06,
+      "loss": 0.5371,
+      "step": 3705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.959069132193568,
+      "learning_rate": 7.011199764626682e-06,
+      "loss": 0.5643,
+      "step": 3706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6375210427134759,
+      "learning_rate": 7.006123146975112e-06,
+      "loss": 0.5109,
+      "step": 3707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8663404160528765,
+      "learning_rate": 7.001047376734087e-06,
+      "loss": 0.5933,
+      "step": 3708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8566269069359853,
+      "learning_rate": 6.995972455340292e-06,
+      "loss": 0.5493,
+      "step": 3709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7572397353235099,
+      "learning_rate": 6.990898384230174e-06,
+      "loss": 0.5145,
+      "step": 3710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9235618485123075,
+      "learning_rate": 6.985825164839937e-06,
+      "loss": 0.692,
+      "step": 3711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1075841887355757,
+      "learning_rate": 6.980752798605547e-06,
+      "loss": 0.5305,
+      "step": 3712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6935300959902517,
+      "learning_rate": 6.975681286962724e-06,
+      "loss": 0.4935,
+      "step": 3713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.014059756634064,
+      "learning_rate": 6.970610631346951e-06,
+      "loss": 0.5959,
+      "step": 3714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5532962509200339,
+      "learning_rate": 6.965540833193464e-06,
+      "loss": 0.5149,
+      "step": 3715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6637993895440895,
+      "learning_rate": 6.9604718939372615e-06,
+      "loss": 0.5031,
+      "step": 3716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6088695150910113,
+      "learning_rate": 6.9554038150130955e-06,
+      "loss": 0.4836,
+      "step": 3717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6910205646438652,
+      "learning_rate": 6.9503365978554735e-06,
+      "loss": 0.6152,
+      "step": 3718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0639530998306044,
+      "learning_rate": 6.945270243898662e-06,
+      "loss": 0.5491,
+      "step": 3719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7556661125814907,
+      "learning_rate": 6.940204754576685e-06,
+      "loss": 0.5159,
+      "step": 3720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7468565167775936,
+      "learning_rate": 6.935140131323312e-06,
+      "loss": 0.4841,
+      "step": 3721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9840402667557375,
+      "learning_rate": 6.930076375572077e-06,
+      "loss": 0.5868,
+      "step": 3722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7967272097389695,
+      "learning_rate": 6.925013488756264e-06,
+      "loss": 0.6122,
+      "step": 3723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1150458750939687,
+      "learning_rate": 6.919951472308912e-06,
+      "loss": 0.5925,
+      "step": 3724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6830728900572185,
+      "learning_rate": 6.9148903276628175e-06,
+      "loss": 0.5509,
+      "step": 3725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6921758144327734,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.5081,
+      "step": 3726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7101832847533358,
+      "learning_rate": 6.904770659504336e-06,
+      "loss": 0.5777,
+      "step": 3727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.555626751582046,
+      "learning_rate": 6.8997121388563e-06,
+      "loss": 0.4942,
+      "step": 3728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609834494521439,
+      "learning_rate": 6.89465449573822e-06,
+      "loss": 0.5424,
+      "step": 3729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.605247103445452,
+      "learning_rate": 6.889597731581652e-06,
+      "loss": 0.5211,
+      "step": 3730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4887709165863159,
+      "learning_rate": 6.8845418478179016e-06,
+      "loss": 0.4439,
+      "step": 3731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6994040781105537,
+      "learning_rate": 6.879486845878027e-06,
+      "loss": 0.5176,
+      "step": 3732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4037030131836021,
+      "learning_rate": 6.874432727192837e-06,
+      "loss": 0.4408,
+      "step": 3733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.002136496592697,
+      "learning_rate": 6.869379493192886e-06,
+      "loss": 0.4643,
+      "step": 3734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5221957258350916,
+      "learning_rate": 6.8643271453084845e-06,
+      "loss": 0.5392,
+      "step": 3735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0832547553362164,
+      "learning_rate": 6.859275684969686e-06,
+      "loss": 0.5359,
+      "step": 3736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0471298658673494,
+      "learning_rate": 6.854225113606299e-06,
+      "loss": 0.5854,
+      "step": 3737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6593909844337007,
+      "learning_rate": 6.849175432647875e-06,
+      "loss": 0.4517,
+      "step": 3738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7151240887717771,
+      "learning_rate": 6.844126643523714e-06,
+      "loss": 0.6568,
+      "step": 3739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7921255818941998,
+      "learning_rate": 6.839078747662871e-06,
+      "loss": 0.6403,
+      "step": 3740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8001008130884912,
+      "learning_rate": 6.834031746494136e-06,
+      "loss": 0.4956,
+      "step": 3741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5363600066820124,
+      "learning_rate": 6.8289856414460595e-06,
+      "loss": 0.4699,
+      "step": 3742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6902480896532968,
+      "learning_rate": 6.823940433946921e-06,
+      "loss": 0.5302,
+      "step": 3743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.894664220704163,
+      "learning_rate": 6.818896125424762e-06,
+      "loss": 0.5501,
+      "step": 3744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515895269294878,
+      "learning_rate": 6.813852717307362e-06,
+      "loss": 0.4796,
+      "step": 3745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7751043998636067,
+      "learning_rate": 6.808810211022248e-06,
+      "loss": 0.5173,
+      "step": 3746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7826495371036872,
+      "learning_rate": 6.803768607996686e-06,
+      "loss": 0.5433,
+      "step": 3747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2433936451261984,
+      "learning_rate": 6.798727909657698e-06,
+      "loss": 0.5303,
+      "step": 3748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7250589795339102,
+      "learning_rate": 6.793688117432041e-06,
+      "loss": 0.5521,
+      "step": 3749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.060728557519284,
+      "learning_rate": 6.788649232746217e-06,
+      "loss": 0.5425,
+      "step": 3750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6948911506372255,
+      "learning_rate": 6.783611257026471e-06,
+      "loss": 0.4844,
+      "step": 3751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9731405956629664,
+      "learning_rate": 6.778574191698793e-06,
+      "loss": 0.5097,
+      "step": 3752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9013600170910732,
+      "learning_rate": 6.773538038188912e-06,
+      "loss": 0.5683,
+      "step": 3753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.586175552153489,
+      "learning_rate": 6.768502797922301e-06,
+      "loss": 0.5298,
+      "step": 3754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9992476362482756,
+      "learning_rate": 6.763468472324175e-06,
+      "loss": 0.5686,
+      "step": 3755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.705886934740305,
+      "learning_rate": 6.758435062819488e-06,
+      "loss": 0.4574,
+      "step": 3756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7511143763462265,
+      "learning_rate": 6.7534025708329385e-06,
+      "loss": 0.4148,
+      "step": 3757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9795667474024514,
+      "learning_rate": 6.74837099778896e-06,
+      "loss": 0.6113,
+      "step": 3758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0922422393622,
+      "learning_rate": 6.743340345111731e-06,
+      "loss": 0.5488,
+      "step": 3759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8689936969383796,
+      "learning_rate": 6.738310614225164e-06,
+      "loss": 0.6004,
+      "step": 3760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3785497810323832,
+      "learning_rate": 6.733281806552917e-06,
+      "loss": 0.4875,
+      "step": 3761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9860323136696274,
+      "learning_rate": 6.728253923518379e-06,
+      "loss": 0.5342,
+      "step": 3762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8103018529965904,
+      "learning_rate": 6.723226966544691e-06,
+      "loss": 0.4889,
+      "step": 3763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8861988377900605,
+      "learning_rate": 6.718200937054714e-06,
+      "loss": 0.5558,
+      "step": 3764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6403248746484864,
+      "learning_rate": 6.713175836471057e-06,
+      "loss": 0.5739,
+      "step": 3765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5078463897939236,
+      "learning_rate": 6.708151666216063e-06,
+      "loss": 0.4341,
+      "step": 3766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6104866117406105,
+      "learning_rate": 6.703128427711816e-06,
+      "loss": 0.6566,
+      "step": 3767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5602921754115322,
+      "learning_rate": 6.69810612238013e-06,
+      "loss": 0.5722,
+      "step": 3768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.146705042450252,
+      "learning_rate": 6.6930847516425615e-06,
+      "loss": 0.5008,
+      "step": 3769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6082808417654897,
+      "learning_rate": 6.688064316920393e-06,
+      "loss": 0.55,
+      "step": 3770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2381798641876975,
+      "learning_rate": 6.683044819634654e-06,
+      "loss": 0.4997,
+      "step": 3771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5302428406972466,
+      "learning_rate": 6.678026261206102e-06,
+      "loss": 0.4911,
+      "step": 3772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5987415575755497,
+      "learning_rate": 6.673008643055228e-06,
+      "loss": 0.4804,
+      "step": 3773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7483394377034727,
+      "learning_rate": 6.667991966602257e-06,
+      "loss": 0.5526,
+      "step": 3774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1468716414478544,
+      "learning_rate": 6.66297623326715e-06,
+      "loss": 0.4624,
+      "step": 3775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6111801686951825,
+      "learning_rate": 6.657961444469601e-06,
+      "loss": 0.6085,
+      "step": 3776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0212042549942906,
+      "learning_rate": 6.652947601629032e-06,
+      "loss": 0.5698,
+      "step": 3777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.323603928262974,
+      "learning_rate": 6.6479347061646046e-06,
+      "loss": 0.5969,
+      "step": 3778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7013555000959877,
+      "learning_rate": 6.642922759495205e-06,
+      "loss": 0.512,
+      "step": 3779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3218950247070973,
+      "learning_rate": 6.637911763039457e-06,
+      "loss": 0.5752,
+      "step": 3780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7754423823783574,
+      "learning_rate": 6.632901718215711e-06,
+      "loss": 0.5006,
+      "step": 3781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9227634052028004,
+      "learning_rate": 6.627892626442049e-06,
+      "loss": 0.5983,
+      "step": 3782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5851219246799237,
+      "learning_rate": 6.622884489136286e-06,
+      "loss": 0.5169,
+      "step": 3783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8504655798502638,
+      "learning_rate": 6.617877307715963e-06,
+      "loss": 0.6041,
+      "step": 3784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7906057045602929,
+      "learning_rate": 6.612871083598354e-06,
+      "loss": 0.6763,
+      "step": 3785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6200881032392898,
+      "learning_rate": 6.607865818200458e-06,
+      "loss": 0.5671,
+      "step": 3786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7680420670591104,
+      "learning_rate": 6.602861512939005e-06,
+      "loss": 0.4862,
+      "step": 3787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6798745753174997,
+      "learning_rate": 6.597858169230454e-06,
+      "loss": 0.5627,
+      "step": 3788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.116063855711824,
+      "learning_rate": 6.592855788490991e-06,
+      "loss": 0.5734,
+      "step": 3789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2761126883563207,
+      "learning_rate": 6.587854372136529e-06,
+      "loss": 0.582,
+      "step": 3790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.187103758530393,
+      "learning_rate": 6.582853921582708e-06,
+      "loss": 0.5825,
+      "step": 3791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3258612074679372,
+      "learning_rate": 6.577854438244897e-06,
+      "loss": 0.4993,
+      "step": 3792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.513138986220021,
+      "learning_rate": 6.572855923538186e-06,
+      "loss": 0.5411,
+      "step": 3793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8058144765731088,
+      "learning_rate": 6.567858378877394e-06,
+      "loss": 0.4786,
+      "step": 3794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7597030303793489,
+      "learning_rate": 6.5628618056770696e-06,
+      "loss": 0.6125,
+      "step": 3795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5184164070203658,
+      "learning_rate": 6.557866205351479e-06,
+      "loss": 0.5357,
+      "step": 3796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6336387815122269,
+      "learning_rate": 6.552871579314619e-06,
+      "loss": 0.4952,
+      "step": 3797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2655217445823266,
+      "learning_rate": 6.547877928980206e-06,
+      "loss": 0.5049,
+      "step": 3798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5996643220000817,
+      "learning_rate": 6.542885255761682e-06,
+      "loss": 0.4535,
+      "step": 3799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.227059997039221,
+      "learning_rate": 6.537893561072214e-06,
+      "loss": 0.5392,
+      "step": 3800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5868912009561416,
+      "learning_rate": 6.532902846324689e-06,
+      "loss": 0.5317,
+      "step": 3801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7623984311143366,
+      "learning_rate": 6.52791311293172e-06,
+      "loss": 0.5227,
+      "step": 3802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7700076614510128,
+      "learning_rate": 6.522924362305639e-06,
+      "loss": 0.6451,
+      "step": 3803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5843893193153016,
+      "learning_rate": 6.517936595858503e-06,
+      "loss": 0.5,
+      "step": 3804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7307291473491926,
+      "learning_rate": 6.512949815002088e-06,
+      "loss": 0.644,
+      "step": 3805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.738075175697846,
+      "learning_rate": 6.50796402114789e-06,
+      "loss": 0.5689,
+      "step": 3806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.587665693524215,
+      "learning_rate": 6.502979215707133e-06,
+      "loss": 0.5371,
+      "step": 3807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6064225229944937,
+      "learning_rate": 6.497995400090748e-06,
+      "loss": 0.5714,
+      "step": 3808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3955213625074754,
+      "learning_rate": 6.4930125757094e-06,
+      "loss": 0.4016,
+      "step": 3809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.18904483788201,
+      "learning_rate": 6.488030743973463e-06,
+      "loss": 0.5629,
+      "step": 3810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5168238708286887,
+      "learning_rate": 6.483049906293035e-06,
+      "loss": 0.5431,
+      "step": 3811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.061709182256943,
+      "learning_rate": 6.478070064077933e-06,
+      "loss": 0.6077,
+      "step": 3812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6435514530292337,
+      "learning_rate": 6.4730912187376895e-06,
+      "loss": 0.518,
+      "step": 3813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.886691483919874,
+      "learning_rate": 6.468113371681557e-06,
+      "loss": 0.4737,
+      "step": 3814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.655312655505381,
+      "learning_rate": 6.463136524318503e-06,
+      "loss": 0.5488,
+      "step": 3815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4553858274245923,
+      "learning_rate": 6.4581606780572155e-06,
+      "loss": 0.4763,
+      "step": 3816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4843542363739635,
+      "learning_rate": 6.453185834306095e-06,
+      "loss": 0.4519,
+      "step": 3817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4925306204904059,
+      "learning_rate": 6.448211994473263e-06,
+      "loss": 0.463,
+      "step": 3818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6878708813152083,
+      "learning_rate": 6.443239159966556e-06,
+      "loss": 0.5759,
+      "step": 3819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5988528082635904,
+      "learning_rate": 6.438267332193519e-06,
+      "loss": 0.4898,
+      "step": 3820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.074917859029666,
+      "learning_rate": 6.4332965125614235e-06,
+      "loss": 0.5745,
+      "step": 3821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7730772147516283,
+      "learning_rate": 6.428326702477246e-06,
+      "loss": 0.4627,
+      "step": 3822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9600022351683575,
+      "learning_rate": 6.42335790334768e-06,
+      "loss": 0.6123,
+      "step": 3823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.068175631320001,
+      "learning_rate": 6.418390116579134e-06,
+      "loss": 0.5048,
+      "step": 3824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8745884824024315,
+      "learning_rate": 6.4134233435777315e-06,
+      "loss": 0.5432,
+      "step": 3825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3167426962924287,
+      "learning_rate": 6.408457585749307e-06,
+      "loss": 0.5389,
+      "step": 3826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9133279341021338,
+      "learning_rate": 6.403492844499406e-06,
+      "loss": 0.6906,
+      "step": 3827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7528179909003727,
+      "learning_rate": 6.398529121233291e-06,
+      "loss": 0.5466,
+      "step": 3828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8354681891384288,
+      "learning_rate": 6.39356641735593e-06,
+      "loss": 0.5057,
+      "step": 3829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.78238895642826,
+      "learning_rate": 6.388604734272006e-06,
+      "loss": 0.5315,
+      "step": 3830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0010161272091724,
+      "learning_rate": 6.383644073385915e-06,
+      "loss": 0.5291,
+      "step": 3831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9840563359109762,
+      "learning_rate": 6.378684436101761e-06,
+      "loss": 0.517,
+      "step": 3832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4040116240610563,
+      "learning_rate": 6.373725823823359e-06,
+      "loss": 0.5211,
+      "step": 3833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7442335607691293,
+      "learning_rate": 6.368768237954234e-06,
+      "loss": 0.5446,
+      "step": 3834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.55457598439831,
+      "learning_rate": 6.363811679897618e-06,
+      "loss": 0.517,
+      "step": 3835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.838431170847624,
+      "learning_rate": 6.358856151056458e-06,
+      "loss": 0.5022,
+      "step": 3836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.110583327521383,
+      "learning_rate": 6.353901652833403e-06,
+      "loss": 0.5912,
+      "step": 3837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9277470696341292,
+      "learning_rate": 6.348948186630815e-06,
+      "loss": 0.4876,
+      "step": 3838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0126681222087615,
+      "learning_rate": 6.343995753850762e-06,
+      "loss": 0.5759,
+      "step": 3839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.858528462546751,
+      "learning_rate": 6.339044355895016e-06,
+      "loss": 0.5421,
+      "step": 3840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8963691427875882,
+      "learning_rate": 6.334093994165067e-06,
+      "loss": 0.576,
+      "step": 3841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.042995669998301,
+      "learning_rate": 6.3291446700621e-06,
+      "loss": 0.5991,
+      "step": 3842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1376237681917254,
+      "learning_rate": 6.324196384987009e-06,
+      "loss": 0.522,
+      "step": 3843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6596236735014733,
+      "learning_rate": 6.3192491403404e-06,
+      "loss": 0.5373,
+      "step": 3844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.499466466271978,
+      "learning_rate": 6.3143029375225785e-06,
+      "loss": 0.5557,
+      "step": 3845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.629258877616758,
+      "learning_rate": 6.309357777933555e-06,
+      "loss": 0.5421,
+      "step": 3846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.038318977268545,
+      "learning_rate": 6.30441366297305e-06,
+      "loss": 0.5039,
+      "step": 3847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7829413240866454,
+      "learning_rate": 6.2994705940404825e-06,
+      "loss": 0.6547,
+      "step": 3848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.535822284656172,
+      "learning_rate": 6.294528572534977e-06,
+      "loss": 0.4843,
+      "step": 3849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7063745486488366,
+      "learning_rate": 6.289587599855367e-06,
+      "loss": 0.4744,
+      "step": 3850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9768903622021436,
+      "learning_rate": 6.284647677400177e-06,
+      "loss": 0.4285,
+      "step": 3851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8218609119234281,
+      "learning_rate": 6.279708806567646e-06,
+      "loss": 0.4103,
+      "step": 3852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6536820300891877,
+      "learning_rate": 6.274770988755712e-06,
+      "loss": 0.5024,
+      "step": 3853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2013202867928943,
+      "learning_rate": 6.2698342253620105e-06,
+      "loss": 0.5221,
+      "step": 3854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9425894790992813,
+      "learning_rate": 6.264898517783885e-06,
+      "loss": 0.5143,
+      "step": 3855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3415538858777825,
+      "learning_rate": 6.259963867418375e-06,
+      "loss": 0.4782,
+      "step": 3856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5232992375844148,
+      "learning_rate": 6.255030275662226e-06,
+      "loss": 0.4957,
+      "step": 3857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1073112639918006,
+      "learning_rate": 6.250097743911877e-06,
+      "loss": 0.4575,
+      "step": 3858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4626442379574434,
+      "learning_rate": 6.245166273563473e-06,
+      "loss": 0.5413,
+      "step": 3859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.852356861264728,
+      "learning_rate": 6.240235866012856e-06,
+      "loss": 0.6682,
+      "step": 3860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6296771261526009,
+      "learning_rate": 6.235306522655566e-06,
+      "loss": 0.5228,
+      "step": 3861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.209125938345802,
+      "learning_rate": 6.230378244886847e-06,
+      "loss": 0.5251,
+      "step": 3862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1342116807195257,
+      "learning_rate": 6.225451034101631e-06,
+      "loss": 0.4867,
+      "step": 3863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0313396148825658,
+      "learning_rate": 6.220524891694562e-06,
+      "loss": 0.5707,
+      "step": 3864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9463624626127813,
+      "learning_rate": 6.2155998190599705e-06,
+      "loss": 0.5347,
+      "step": 3865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7620860683650132,
+      "learning_rate": 6.210675817591889e-06,
+      "loss": 0.5419,
+      "step": 3866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6844160268049446,
+      "learning_rate": 6.2057528886840445e-06,
+      "loss": 0.483,
+      "step": 3867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2566054249590946,
+      "learning_rate": 6.200831033729864e-06,
+      "loss": 0.5194,
+      "step": 3868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3414726679203715,
+      "learning_rate": 6.195910254122466e-06,
+      "loss": 0.5085,
+      "step": 3869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8387092099314886,
+      "learning_rate": 6.190990551254668e-06,
+      "loss": 0.534,
+      "step": 3870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8646728492264557,
+      "learning_rate": 6.186071926518984e-06,
+      "loss": 0.5535,
+      "step": 3871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1378898972107745,
+      "learning_rate": 6.18115438130761e-06,
+      "loss": 0.4957,
+      "step": 3872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7078829181243245,
+      "learning_rate": 6.176237917012459e-06,
+      "loss": 0.5007,
+      "step": 3873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5332060626360289,
+      "learning_rate": 6.171322535025119e-06,
+      "loss": 0.5034,
+      "step": 3874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7947817732895848,
+      "learning_rate": 6.166408236736883e-06,
+      "loss": 0.5432,
+      "step": 3875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.747447248025834,
+      "learning_rate": 6.161495023538729e-06,
+      "loss": 0.5796,
+      "step": 3876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8765207259334356,
+      "learning_rate": 6.1565828968213325e-06,
+      "loss": 0.5713,
+      "step": 3877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1961149328301937,
+      "learning_rate": 6.151671857975061e-06,
+      "loss": 0.4789,
+      "step": 3878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6295464761813787,
+      "learning_rate": 6.146761908389975e-06,
+      "loss": 0.4464,
+      "step": 3879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0466811639462086,
+      "learning_rate": 6.141853049455824e-06,
+      "loss": 0.5436,
+      "step": 3880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9591857628254925,
+      "learning_rate": 6.1369452825620515e-06,
+      "loss": 0.6291,
+      "step": 3881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0230725312245106,
+      "learning_rate": 6.132038609097788e-06,
+      "loss": 0.506,
+      "step": 3882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7455409278038863,
+      "learning_rate": 6.12713303045186e-06,
+      "loss": 0.579,
+      "step": 3883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.013276403739809,
+      "learning_rate": 6.1222285480127786e-06,
+      "loss": 0.5054,
+      "step": 3884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1489440706082283,
+      "learning_rate": 6.11732516316875e-06,
+      "loss": 0.5635,
+      "step": 3885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8397531562668823,
+      "learning_rate": 6.112422877307664e-06,
+      "loss": 0.6484,
+      "step": 3886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.123942092307035,
+      "learning_rate": 6.107521691817104e-06,
+      "loss": 0.5809,
+      "step": 3887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9771294074742247,
+      "learning_rate": 6.10262160808434e-06,
+      "loss": 0.5112,
+      "step": 3888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.918779291095982,
+      "learning_rate": 6.097722627496332e-06,
+      "loss": 0.6759,
+      "step": 3889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8019158132672368,
+      "learning_rate": 6.092824751439723e-06,
+      "loss": 0.5575,
+      "step": 3890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.776682827380513,
+      "learning_rate": 6.0879279813008495e-06,
+      "loss": 0.571,
+      "step": 3891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.967440107248241,
+      "learning_rate": 6.083032318465731e-06,
+      "loss": 0.5314,
+      "step": 3892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.970351856682046,
+      "learning_rate": 6.0781377643200765e-06,
+      "loss": 0.5184,
+      "step": 3893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.061193954600697,
+      "learning_rate": 6.073244320249274e-06,
+      "loss": 0.4761,
+      "step": 3894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7636369327950874,
+      "learning_rate": 6.0683519876384034e-06,
+      "loss": 0.4744,
+      "step": 3895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.278215305644197,
+      "learning_rate": 6.063460767872233e-06,
+      "loss": 0.4758,
+      "step": 3896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6143939529095657,
+      "learning_rate": 6.05857066233521e-06,
+      "loss": 0.4714,
+      "step": 3897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7330656378329643,
+      "learning_rate": 6.053681672411471e-06,
+      "loss": 0.474,
+      "step": 3898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.574471460107547,
+      "learning_rate": 6.048793799484831e-06,
+      "loss": 0.4685,
+      "step": 3899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9289890457473569,
+      "learning_rate": 6.0439070449387924e-06,
+      "loss": 0.4816,
+      "step": 3900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8120134038909435,
+      "learning_rate": 6.039021410156542e-06,
+      "loss": 0.5143,
+      "step": 3901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9085107314654917,
+      "learning_rate": 6.03413689652095e-06,
+      "loss": 0.507,
+      "step": 3902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3106753743841864,
+      "learning_rate": 6.029253505414565e-06,
+      "loss": 0.4842,
+      "step": 3903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9638998160935168,
+      "learning_rate": 6.024371238219622e-06,
+      "loss": 0.5148,
+      "step": 3904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.215384600191575,
+      "learning_rate": 6.019490096318036e-06,
+      "loss": 0.4829,
+      "step": 3905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.585529425974857,
+      "learning_rate": 6.014610081091403e-06,
+      "loss": 0.5933,
+      "step": 3906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.584196002736741,
+      "learning_rate": 6.009731193921002e-06,
+      "loss": 0.5156,
+      "step": 3907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5202260096751705,
+      "learning_rate": 6.004853436187794e-06,
+      "loss": 0.5404,
+      "step": 3908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.876026380648518,
+      "learning_rate": 5.9999768092724145e-06,
+      "loss": 0.42,
+      "step": 3909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8375918837319636,
+      "learning_rate": 5.995101314555181e-06,
+      "loss": 0.4486,
+      "step": 3910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1612849989636436,
+      "learning_rate": 5.990226953416099e-06,
+      "loss": 0.5672,
+      "step": 3911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8328305119056958,
+      "learning_rate": 5.98535372723484e-06,
+      "loss": 0.5379,
+      "step": 3912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.189720137616625,
+      "learning_rate": 5.9804816373907625e-06,
+      "loss": 0.5301,
+      "step": 3913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.691290945907313,
+      "learning_rate": 5.975610685262902e-06,
+      "loss": 0.5143,
+      "step": 3914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7749228292920896,
+      "learning_rate": 5.970740872229974e-06,
+      "loss": 0.5345,
+      "step": 3915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6105046075390463,
+      "learning_rate": 5.965872199670362e-06,
+      "loss": 0.5275,
+      "step": 3916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3870656856218844,
+      "learning_rate": 5.961004668962136e-06,
+      "loss": 0.5884,
+      "step": 3917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7851722541285373,
+      "learning_rate": 5.956138281483039e-06,
+      "loss": 0.5743,
+      "step": 3918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5904106392153181,
+      "learning_rate": 5.951273038610496e-06,
+      "loss": 0.5272,
+      "step": 3919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8633926077680684,
+      "learning_rate": 5.946408941721602e-06,
+      "loss": 0.6422,
+      "step": 3920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8792048450117027,
+      "learning_rate": 5.941545992193129e-06,
+      "loss": 0.6987,
+      "step": 3921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0214439952595074,
+      "learning_rate": 5.936684191401525e-06,
+      "loss": 0.5707,
+      "step": 3922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5188945110645737,
+      "learning_rate": 5.931823540722912e-06,
+      "loss": 0.4772,
+      "step": 3923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6840376207479832,
+      "learning_rate": 5.9269640415330875e-06,
+      "loss": 0.5373,
+      "step": 3924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4846255564553954,
+      "learning_rate": 5.922105695207521e-06,
+      "loss": 0.6285,
+      "step": 3925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.598841121618291,
+      "learning_rate": 5.917248503121359e-06,
+      "loss": 0.477,
+      "step": 3926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.035376220648736,
+      "learning_rate": 5.912392466649419e-06,
+      "loss": 0.6004,
+      "step": 3927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8186416842169495,
+      "learning_rate": 5.907537587166191e-06,
+      "loss": 0.5841,
+      "step": 3928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0189065033397715,
+      "learning_rate": 5.90268386604584e-06,
+      "loss": 0.5444,
+      "step": 3929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.441233040900877,
+      "learning_rate": 5.897831304662201e-06,
+      "loss": 0.4729,
+      "step": 3930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.900135284902192,
+      "learning_rate": 5.892979904388781e-06,
+      "loss": 0.5601,
+      "step": 3931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1518211844431843,
+      "learning_rate": 5.888129666598756e-06,
+      "loss": 0.4998,
+      "step": 3932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.995324808460359,
+      "learning_rate": 5.883280592664979e-06,
+      "loss": 0.5044,
+      "step": 3933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.032392266253993,
+      "learning_rate": 5.878432683959972e-06,
+      "loss": 0.6025,
+      "step": 3934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7219825353458906,
+      "learning_rate": 5.8735859418559206e-06,
+      "loss": 0.5318,
+      "step": 3935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1532163681977163,
+      "learning_rate": 5.868740367724692e-06,
+      "loss": 0.5572,
+      "step": 3936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9104493028152545,
+      "learning_rate": 5.863895962937806e-06,
+      "loss": 0.511,
+      "step": 3937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7970655613635922,
+      "learning_rate": 5.859052728866468e-06,
+      "loss": 0.519,
+      "step": 3938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9044659209766448,
+      "learning_rate": 5.854210666881544e-06,
+      "loss": 0.4938,
+      "step": 3939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9069045563967078,
+      "learning_rate": 5.8493697783535665e-06,
+      "loss": 0.5434,
+      "step": 3940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7236842299557886,
+      "learning_rate": 5.844530064652742e-06,
+      "loss": 0.5434,
+      "step": 3941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9193772288496869,
+      "learning_rate": 5.839691527148938e-06,
+      "loss": 0.5663,
+      "step": 3942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6256372584827723,
+      "learning_rate": 5.834854167211699e-06,
+      "loss": 0.5618,
+      "step": 3943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8410101947794912,
+      "learning_rate": 5.8300179862102225e-06,
+      "loss": 0.5111,
+      "step": 3944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7395379045132877,
+      "learning_rate": 5.825182985513383e-06,
+      "loss": 0.5797,
+      "step": 3945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.772142372968928,
+      "learning_rate": 5.820349166489716e-06,
+      "loss": 0.5897,
+      "step": 3946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9719220298450046,
+      "learning_rate": 5.8155165305074245e-06,
+      "loss": 0.6295,
+      "step": 3947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5965061455667613,
+      "learning_rate": 5.810685078934375e-06,
+      "loss": 0.5109,
+      "step": 3948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2788946818109004,
+      "learning_rate": 5.805854813138098e-06,
+      "loss": 0.5623,
+      "step": 3949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.850210215038764,
+      "learning_rate": 5.801025734485794e-06,
+      "loss": 0.5662,
+      "step": 3950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7697763891018997,
+      "learning_rate": 5.796197844344325e-06,
+      "loss": 0.665,
+      "step": 3951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.634369527035235,
+      "learning_rate": 5.791371144080209e-06,
+      "loss": 0.5337,
+      "step": 3952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6353546360413924,
+      "learning_rate": 5.78654563505964e-06,
+      "loss": 0.4942,
+      "step": 3953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.104018005496268,
+      "learning_rate": 5.781721318648461e-06,
+      "loss": 0.6301,
+      "step": 3954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8561869162914142,
+      "learning_rate": 5.7768981962121906e-06,
+      "loss": 0.5098,
+      "step": 3955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8630049606818877,
+      "learning_rate": 5.772076269116001e-06,
+      "loss": 0.6138,
+      "step": 3956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7815523351226015,
+      "learning_rate": 5.7672555387247274e-06,
+      "loss": 0.5452,
+      "step": 3957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.500516683341203,
+      "learning_rate": 5.762436006402874e-06,
+      "loss": 0.4918,
+      "step": 3958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.528900945506693,
+      "learning_rate": 5.757617673514588e-06,
+      "loss": 0.5367,
+      "step": 3959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6891376666765927,
+      "learning_rate": 5.752800541423696e-06,
+      "loss": 0.5013,
+      "step": 3960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3278966899677003,
+      "learning_rate": 5.747984611493675e-06,
+      "loss": 0.5398,
+      "step": 3961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6958773165016172,
+      "learning_rate": 5.743169885087665e-06,
+      "loss": 0.5399,
+      "step": 3962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5628130628252987,
+      "learning_rate": 5.738356363568463e-06,
+      "loss": 0.5301,
+      "step": 3963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.619580651650061,
+      "learning_rate": 5.733544048298526e-06,
+      "loss": 0.4046,
+      "step": 3964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8388830700409752,
+      "learning_rate": 5.728732940639972e-06,
+      "loss": 0.6404,
+      "step": 3965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.361838887276079,
+      "learning_rate": 5.723923041954571e-06,
+      "loss": 0.5181,
+      "step": 3966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6110014077720822,
+      "learning_rate": 5.719114353603757e-06,
+      "loss": 0.48,
+      "step": 3967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2238964849394,
+      "learning_rate": 5.714306876948621e-06,
+      "loss": 0.5434,
+      "step": 3968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4593602516794348,
+      "learning_rate": 5.709500613349906e-06,
+      "loss": 0.4368,
+      "step": 3969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.785984442549652,
+      "learning_rate": 5.704695564168014e-06,
+      "loss": 0.4886,
+      "step": 3970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0421074797342307,
+      "learning_rate": 5.6998917307630095e-06,
+      "loss": 0.5961,
+      "step": 3971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0738591528791526,
+      "learning_rate": 5.695089114494599e-06,
+      "loss": 0.6159,
+      "step": 3972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8506086647585023,
+      "learning_rate": 5.69028771672216e-06,
+      "loss": 0.6128,
+      "step": 3973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.680761710409917,
+      "learning_rate": 5.685487538804718e-06,
+      "loss": 0.4606,
+      "step": 3974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7841197355051994,
+      "learning_rate": 5.68068858210095e-06,
+      "loss": 0.5597,
+      "step": 3975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1687558975871997,
+      "learning_rate": 5.675890847969193e-06,
+      "loss": 0.4408,
+      "step": 3976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6100465942404951,
+      "learning_rate": 5.671094337767433e-06,
+      "loss": 0.5944,
+      "step": 3977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7675960452338382,
+      "learning_rate": 5.666299052853314e-06,
+      "loss": 0.5519,
+      "step": 3978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6880528253139264,
+      "learning_rate": 5.661504994584133e-06,
+      "loss": 0.5073,
+      "step": 3979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0257428199573555,
+      "learning_rate": 5.656712164316838e-06,
+      "loss": 0.4425,
+      "step": 3980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.358803231144599,
+      "learning_rate": 5.651920563408022e-06,
+      "loss": 0.5718,
+      "step": 3981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8010041352979773,
+      "learning_rate": 5.647130193213945e-06,
+      "loss": 0.5523,
+      "step": 3982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2841019066618755,
+      "learning_rate": 5.642341055090508e-06,
+      "loss": 0.5395,
+      "step": 3983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9589831517442522,
+      "learning_rate": 5.637553150393268e-06,
+      "loss": 0.4923,
+      "step": 3984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9348682530704195,
+      "learning_rate": 5.632766480477432e-06,
+      "loss": 0.5618,
+      "step": 3985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9214619150889214,
+      "learning_rate": 5.6279810466978546e-06,
+      "loss": 0.5747,
+      "step": 3986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6427552184757264,
+      "learning_rate": 5.623196850409044e-06,
+      "loss": 0.6301,
+      "step": 3987
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3596764304402282,
+      "learning_rate": 5.618413892965158e-06,
+      "loss": 0.5489,
+      "step": 3988
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0426448695779,
+      "learning_rate": 5.613632175720001e-06,
+      "loss": 0.6251,
+      "step": 3989
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9771944726894803,
+      "learning_rate": 5.6088517000270275e-06,
+      "loss": 0.5719,
+      "step": 3990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1781772317880748,
+      "learning_rate": 5.604072467239343e-06,
+      "loss": 0.4969,
+      "step": 3991
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6451481058669248,
+      "learning_rate": 5.599294478709698e-06,
+      "loss": 0.5975,
+      "step": 3992
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.456973974547951,
+      "learning_rate": 5.5945177357904935e-06,
+      "loss": 0.6023,
+      "step": 3993
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8700933321001934,
+      "learning_rate": 5.589742239833776e-06,
+      "loss": 0.5789,
+      "step": 3994
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.989936321843003,
+      "learning_rate": 5.584967992191234e-06,
+      "loss": 0.5353,
+      "step": 3995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.593736749562601,
+      "learning_rate": 5.580194994214216e-06,
+      "loss": 0.5413,
+      "step": 3996
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.737135632640496,
+      "learning_rate": 5.5754232472537086e-06,
+      "loss": 0.5243,
+      "step": 3997
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8102374780737005,
+      "learning_rate": 5.570652752660343e-06,
+      "loss": 0.5212,
+      "step": 3998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3201710770386437,
+      "learning_rate": 5.565883511784396e-06,
+      "loss": 0.4827,
+      "step": 3999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7174433295470877,
+      "learning_rate": 5.561115525975793e-06,
+      "loss": 0.4846,
+      "step": 4000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4698141423605808,
+      "learning_rate": 5.5563487965841055e-06,
+      "loss": 0.5447,
+      "step": 4001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8247454031148241,
+      "learning_rate": 5.5515833249585385e-06,
+      "loss": 0.7072,
+      "step": 4002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8084517231722244,
+      "learning_rate": 5.546819112447952e-06,
+      "loss": 0.4932,
+      "step": 4003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2254756660959325,
+      "learning_rate": 5.542056160400848e-06,
+      "loss": 0.5394,
+      "step": 4004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.856452179975468,
+      "learning_rate": 5.537294470165369e-06,
+      "loss": 0.5118,
+      "step": 4005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.444625060246301,
+      "learning_rate": 5.532534043089302e-06,
+      "loss": 0.4944,
+      "step": 4006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0711124635836238,
+      "learning_rate": 5.527774880520073e-06,
+      "loss": 0.5523,
+      "step": 4007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9165767281869026,
+      "learning_rate": 5.523016983804759e-06,
+      "loss": 0.6046,
+      "step": 4008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.960096351088633,
+      "learning_rate": 5.518260354290066e-06,
+      "loss": 0.4989,
+      "step": 4009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.20445774709088,
+      "learning_rate": 5.513504993322352e-06,
+      "loss": 0.5226,
+      "step": 4010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7523480545706327,
+      "learning_rate": 5.508750902247612e-06,
+      "loss": 0.5464,
+      "step": 4011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.831017547190802,
+      "learning_rate": 5.503998082411479e-06,
+      "loss": 0.4598,
+      "step": 4012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0043000735723084,
+      "learning_rate": 5.499246535159231e-06,
+      "loss": 0.5897,
+      "step": 4013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.725752505608157,
+      "learning_rate": 5.494496261835781e-06,
+      "loss": 0.6311,
+      "step": 4014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0371754927968637,
+      "learning_rate": 5.489747263785687e-06,
+      "loss": 0.513,
+      "step": 4015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9597419219603391,
+      "learning_rate": 5.48499954235314e-06,
+      "loss": 0.5158,
+      "step": 4016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.311275622965174,
+      "learning_rate": 5.480253098881974e-06,
+      "loss": 0.4536,
+      "step": 4017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.956952438002974,
+      "learning_rate": 5.47550793471566e-06,
+      "loss": 0.4877,
+      "step": 4018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765385871813868,
+      "learning_rate": 5.470764051197302e-06,
+      "loss": 0.5928,
+      "step": 4019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8036530298699281,
+      "learning_rate": 5.466021449669655e-06,
+      "loss": 0.675,
+      "step": 4020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7157264748424985,
+      "learning_rate": 5.461280131475099e-06,
+      "loss": 0.5275,
+      "step": 4021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.640150243420664,
+      "learning_rate": 5.456540097955652e-06,
+      "loss": 0.5647,
+      "step": 4022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6510174810428884,
+      "learning_rate": 5.451801350452975e-06,
+      "loss": 0.5331,
+      "step": 4023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6117121611452605,
+      "learning_rate": 5.447063890308354e-06,
+      "loss": 0.4918,
+      "step": 4024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7715014242098321,
+      "learning_rate": 5.442327718862721e-06,
+      "loss": 0.665,
+      "step": 4025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8195573873343689,
+      "learning_rate": 5.4375928374566376e-06,
+      "loss": 0.6964,
+      "step": 4026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7195404057767654,
+      "learning_rate": 5.432859247430303e-06,
+      "loss": 0.5576,
+      "step": 4027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7861523211508588,
+      "learning_rate": 5.428126950123551e-06,
+      "loss": 0.5264,
+      "step": 4028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3743722394970725,
+      "learning_rate": 5.423395946875846e-06,
+      "loss": 0.6268,
+      "step": 4029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.587061817232752,
+      "learning_rate": 5.418666239026291e-06,
+      "loss": 0.4618,
+      "step": 4030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7236050981677138,
+      "learning_rate": 5.413937827913619e-06,
+      "loss": 0.6069,
+      "step": 4031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9107384669885006,
+      "learning_rate": 5.409210714876197e-06,
+      "loss": 0.4998,
+      "step": 4032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5198376879573536,
+      "learning_rate": 5.404484901252023e-06,
+      "loss": 0.5384,
+      "step": 4033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7190099109557533,
+      "learning_rate": 5.399760388378729e-06,
+      "loss": 0.518,
+      "step": 4034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7729258609586451,
+      "learning_rate": 5.395037177593579e-06,
+      "loss": 0.6252,
+      "step": 4035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6141406879938631,
+      "learning_rate": 5.390315270233469e-06,
+      "loss": 0.4894,
+      "step": 4036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4393332070691438,
+      "learning_rate": 5.385594667634923e-06,
+      "loss": 0.4959,
+      "step": 4037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.081485183586388,
+      "learning_rate": 5.3808753711341e-06,
+      "loss": 0.5192,
+      "step": 4038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.265234694543617,
+      "learning_rate": 5.376157382066784e-06,
+      "loss": 0.5036,
+      "step": 4039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1974036817416933,
+      "learning_rate": 5.371440701768394e-06,
+      "loss": 0.5172,
+      "step": 4040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8648206067598243,
+      "learning_rate": 5.366725331573974e-06,
+      "loss": 0.5149,
+      "step": 4041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.553597470654213,
+      "learning_rate": 5.3620112728182e-06,
+      "loss": 0.4452,
+      "step": 4042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0540893868523122,
+      "learning_rate": 5.357298526835381e-06,
+      "loss": 0.529,
+      "step": 4043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.842502551021806,
+      "learning_rate": 5.35258709495945e-06,
+      "loss": 0.6787,
+      "step": 4044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4998632670754062,
+      "learning_rate": 5.34787697852396e-06,
+      "loss": 0.4727,
+      "step": 4045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5041349271100766,
+      "learning_rate": 5.343168178862104e-06,
+      "loss": 0.4597,
+      "step": 4046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.235587494815972,
+      "learning_rate": 5.338460697306699e-06,
+      "loss": 0.5099,
+      "step": 4047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7822966338755881,
+      "learning_rate": 5.333754535190186e-06,
+      "loss": 0.5205,
+      "step": 4048
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5805009096627058,
+      "learning_rate": 5.329049693844635e-06,
+      "loss": 0.5434,
+      "step": 4049
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5988258460598725,
+      "learning_rate": 5.324346174601741e-06,
+      "loss": 0.5167,
+      "step": 4050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9537419840255446,
+      "learning_rate": 5.319643978792825e-06,
+      "loss": 0.5843,
+      "step": 4051
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.946913641759761,
+      "learning_rate": 5.314943107748836e-06,
+      "loss": 0.4746,
+      "step": 4052
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6500813449341156,
+      "learning_rate": 5.3102435628003435e-06,
+      "loss": 0.4764,
+      "step": 4053
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6146264275794364,
+      "learning_rate": 5.305545345277543e-06,
+      "loss": 0.5144,
+      "step": 4054
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8463815130263386,
+      "learning_rate": 5.300848456510257e-06,
+      "loss": 0.4387,
+      "step": 4055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9044721949193073,
+      "learning_rate": 5.296152897827929e-06,
+      "loss": 0.4989,
+      "step": 4056
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7574178650272556,
+      "learning_rate": 5.291458670559628e-06,
+      "loss": 0.5075,
+      "step": 4057
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7444505984545466,
+      "learning_rate": 5.286765776034044e-06,
+      "loss": 0.5131,
+      "step": 4058
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6683497493164654,
+      "learning_rate": 5.282074215579492e-06,
+      "loss": 0.6042,
+      "step": 4059
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.830161461846451,
+      "learning_rate": 5.277383990523905e-06,
+      "loss": 0.5316,
+      "step": 4060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6808063070938275,
+      "learning_rate": 5.272695102194846e-06,
+      "loss": 0.4627,
+      "step": 4061
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.569063354152936,
+      "learning_rate": 5.2680075519194926e-06,
+      "loss": 0.5563,
+      "step": 4062
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.039432992184878,
+      "learning_rate": 5.263321341024646e-06,
+      "loss": 0.5195,
+      "step": 4063
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7920950734288355,
+      "learning_rate": 5.25863647083673e-06,
+      "loss": 0.492,
+      "step": 4064
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9085522406821938,
+      "learning_rate": 5.253952942681782e-06,
+      "loss": 0.5508,
+      "step": 4065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6085320788213449,
+      "learning_rate": 5.249270757885475e-06,
+      "loss": 0.5766,
+      "step": 4066
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7775666583709516,
+      "learning_rate": 5.244589917773082e-06,
+      "loss": 0.6446,
+      "step": 4067
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.631356256268733,
+      "learning_rate": 5.239910423669509e-06,
+      "loss": 0.5334,
+      "step": 4068
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.97854754180493,
+      "learning_rate": 5.2352322768992755e-06,
+      "loss": 0.5763,
+      "step": 4069
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8229337577147757,
+      "learning_rate": 5.230555478786522e-06,
+      "loss": 0.508,
+      "step": 4070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3914746626154013,
+      "learning_rate": 5.225880030655006e-06,
+      "loss": 0.5185,
+      "step": 4071
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9294203082075647,
+      "learning_rate": 5.221205933828104e-06,
+      "loss": 0.4907,
+      "step": 4072
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.486405883960199,
+      "learning_rate": 5.216533189628808e-06,
+      "loss": 0.5985,
+      "step": 4073
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4328101771691295,
+      "learning_rate": 5.211861799379731e-06,
+      "loss": 0.5617,
+      "step": 4074
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6148565559366042,
+      "learning_rate": 5.207191764403097e-06,
+      "loss": 0.519,
+      "step": 4075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6817490709378973,
+      "learning_rate": 5.20252308602075e-06,
+      "loss": 0.4818,
+      "step": 4076
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.058916140019593,
+      "learning_rate": 5.197855765554152e-06,
+      "loss": 0.4969,
+      "step": 4077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0140690001481647,
+      "learning_rate": 5.193189804324376e-06,
+      "loss": 0.5069,
+      "step": 4078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9600191627873345,
+      "learning_rate": 5.1885252036521125e-06,
+      "loss": 0.551,
+      "step": 4079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.649179598445823,
+      "learning_rate": 5.183861964857669e-06,
+      "loss": 0.5431,
+      "step": 4080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6501916134408832,
+      "learning_rate": 5.179200089260964e-06,
+      "loss": 0.56,
+      "step": 4081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.010373551829197,
+      "learning_rate": 5.174539578181531e-06,
+      "loss": 0.5317,
+      "step": 4082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1364093309448715,
+      "learning_rate": 5.169880432938519e-06,
+      "loss": 0.5332,
+      "step": 4083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9371463496531673,
+      "learning_rate": 5.165222654850688e-06,
+      "loss": 0.5291,
+      "step": 4084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7235802825644708,
+      "learning_rate": 5.160566245236413e-06,
+      "loss": 0.5026,
+      "step": 4085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4808482595055845,
+      "learning_rate": 5.155911205413683e-06,
+      "loss": 0.433,
+      "step": 4086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5198029476975914,
+      "learning_rate": 5.151257536700094e-06,
+      "loss": 0.4904,
+      "step": 4087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9367962332526631,
+      "learning_rate": 5.146605240412859e-06,
+      "loss": 0.5998,
+      "step": 4088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6773613332848005,
+      "learning_rate": 5.141954317868798e-06,
+      "loss": 0.6062,
+      "step": 4089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7463768609469112,
+      "learning_rate": 5.137304770384348e-06,
+      "loss": 0.5315,
+      "step": 4090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9549397043074292,
+      "learning_rate": 5.132656599275554e-06,
+      "loss": 0.6047,
+      "step": 4091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7101497875139602,
+      "learning_rate": 5.128009805858067e-06,
+      "loss": 0.471,
+      "step": 4092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.278726362736307,
+      "learning_rate": 5.123364391447156e-06,
+      "loss": 0.5312,
+      "step": 4093
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.327752204535409,
+      "learning_rate": 5.118720357357696e-06,
+      "loss": 0.5078,
+      "step": 4094
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8298317351006107,
+      "learning_rate": 5.114077704904168e-06,
+      "loss": 0.6817,
+      "step": 4095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2014004734695596,
+      "learning_rate": 5.109436435400667e-06,
+      "loss": 0.5612,
+      "step": 4096
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2418824723833786,
+      "learning_rate": 5.104796550160893e-06,
+      "loss": 0.5438,
+      "step": 4097
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0708041111432554,
+      "learning_rate": 5.100158050498159e-06,
+      "loss": 0.4914,
+      "step": 4098
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8507126209343047,
+      "learning_rate": 5.095520937725378e-06,
+      "loss": 0.6604,
+      "step": 4099
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2947047442605863,
+      "learning_rate": 5.090885213155079e-06,
+      "loss": 0.5427,
+      "step": 4100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5657795132505696,
+      "learning_rate": 5.0862508780993915e-06,
+      "loss": 0.5586,
+      "step": 4101
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9940443590945756,
+      "learning_rate": 5.081617933870056e-06,
+      "loss": 0.4867,
+      "step": 4102
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7972564137226235,
+      "learning_rate": 5.076986381778417e-06,
+      "loss": 0.5162,
+      "step": 4103
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7436575100091707,
+      "learning_rate": 5.072356223135425e-06,
+      "loss": 0.5162,
+      "step": 4104
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6136532168428153,
+      "learning_rate": 5.067727459251638e-06,
+      "loss": 0.5693,
+      "step": 4105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4290557949681117,
+      "learning_rate": 5.063100091437217e-06,
+      "loss": 0.4312,
+      "step": 4106
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2194339188093757,
+      "learning_rate": 5.058474121001928e-06,
+      "loss": 0.518,
+      "step": 4107
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7550535016113227,
+      "learning_rate": 5.053849549255143e-06,
+      "loss": 0.53,
+      "step": 4108
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6910132869396168,
+      "learning_rate": 5.049226377505838e-06,
+      "loss": 0.5425,
+      "step": 4109
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4120379133065497,
+      "learning_rate": 5.044604607062591e-06,
+      "loss": 0.5915,
+      "step": 4110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.544129386565747,
+      "learning_rate": 5.0399842392335856e-06,
+      "loss": 0.5236,
+      "step": 4111
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.569143790974616,
+      "learning_rate": 5.0353652753266045e-06,
+      "loss": 0.4573,
+      "step": 4112
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.511038665099736,
+      "learning_rate": 5.03074771664904e-06,
+      "loss": 0.5574,
+      "step": 4113
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8453122474427148,
+      "learning_rate": 5.026131564507878e-06,
+      "loss": 0.5414,
+      "step": 4114
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.607026069839609,
+      "learning_rate": 5.021516820209713e-06,
+      "loss": 0.5241,
+      "step": 4115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.771524327998849,
+      "learning_rate": 5.016903485060738e-06,
+      "loss": 0.4201,
+      "step": 4116
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9359277652020246,
+      "learning_rate": 5.0122915603667485e-06,
+      "loss": 0.5458,
+      "step": 4117
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8406973258877186,
+      "learning_rate": 5.0076810474331395e-06,
+      "loss": 0.5509,
+      "step": 4118
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6308413016427585,
+      "learning_rate": 5.003071947564908e-06,
+      "loss": 0.5511,
+      "step": 4119
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6229033351453972,
+      "learning_rate": 4.998464262066648e-06,
+      "loss": 0.512,
+      "step": 4120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.997234561283974,
+      "learning_rate": 4.993857992242557e-06,
+      "loss": 0.4648,
+      "step": 4121
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8422666982601414,
+      "learning_rate": 4.9892531393964285e-06,
+      "loss": 0.5101,
+      "step": 4122
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7156096512546009,
+      "learning_rate": 4.984649704831658e-06,
+      "loss": 0.4959,
+      "step": 4123
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.091344108221678,
+      "learning_rate": 4.980047689851236e-06,
+      "loss": 0.5009,
+      "step": 4124
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8909171912767347,
+      "learning_rate": 4.975447095757755e-06,
+      "loss": 0.6728,
+      "step": 4125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9061910954757124,
+      "learning_rate": 4.970847923853404e-06,
+      "loss": 0.5284,
+      "step": 4126
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.770964816954522,
+      "learning_rate": 4.966250175439966e-06,
+      "loss": 0.5552,
+      "step": 4127
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.78140223303192,
+      "learning_rate": 4.961653851818827e-06,
+      "loss": 0.5826,
+      "step": 4128
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5120172192605281,
+      "learning_rate": 4.957058954290964e-06,
+      "loss": 0.5707,
+      "step": 4129
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5161762740415596,
+      "learning_rate": 4.952465484156956e-06,
+      "loss": 0.5313,
+      "step": 4130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.660704758727068,
+      "learning_rate": 4.947873442716972e-06,
+      "loss": 0.5177,
+      "step": 4131
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5754951078549964,
+      "learning_rate": 4.9432828312707836e-06,
+      "loss": 0.4969,
+      "step": 4132
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8092477577501012,
+      "learning_rate": 4.938693651117751e-06,
+      "loss": 0.5707,
+      "step": 4133
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6758657072741292,
+      "learning_rate": 4.934105903556831e-06,
+      "loss": 0.5424,
+      "step": 4134
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6470902414535609,
+      "learning_rate": 4.929519589886578e-06,
+      "loss": 0.5361,
+      "step": 4135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7490463791096584,
+      "learning_rate": 4.924934711405138e-06,
+      "loss": 0.5784,
+      "step": 4136
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5413504283140431,
+      "learning_rate": 4.920351269410251e-06,
+      "loss": 0.4347,
+      "step": 4137
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.548643281188768,
+      "learning_rate": 4.9157692651992495e-06,
+      "loss": 0.5949,
+      "step": 4138
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4519366659400743,
+      "learning_rate": 4.911188700069062e-06,
+      "loss": 0.4583,
+      "step": 4139
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5337774343765418,
+      "learning_rate": 4.906609575316207e-06,
+      "loss": 0.5251,
+      "step": 4140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.09028131161962,
+      "learning_rate": 4.9020318922367956e-06,
+      "loss": 0.4808,
+      "step": 4141
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9960786255328633,
+      "learning_rate": 4.897455652126533e-06,
+      "loss": 0.4909,
+      "step": 4142
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8170099637502837,
+      "learning_rate": 4.892880856280713e-06,
+      "loss": 0.5213,
+      "step": 4143
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6343868158723378,
+      "learning_rate": 4.888307505994222e-06,
+      "loss": 0.5471,
+      "step": 4144
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8263849389191067,
+      "learning_rate": 4.883735602561537e-06,
+      "loss": 0.5659,
+      "step": 4145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6793265867906504,
+      "learning_rate": 4.879165147276726e-06,
+      "loss": 0.473,
+      "step": 4146
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9133637287391057,
+      "learning_rate": 4.874596141433447e-06,
+      "loss": 0.4622,
+      "step": 4147
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7356968081751425,
+      "learning_rate": 4.870028586324947e-06,
+      "loss": 0.5342,
+      "step": 4148
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6671353842336494,
+      "learning_rate": 4.865462483244065e-06,
+      "loss": 0.512,
+      "step": 4149
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8254619325714118,
+      "learning_rate": 4.8608978334832225e-06,
+      "loss": 0.6663,
+      "step": 4150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4336665134879105,
+      "learning_rate": 4.8563346383344375e-06,
+      "loss": 0.5717,
+      "step": 4151
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6435441132345194,
+      "learning_rate": 4.851772899089312e-06,
+      "loss": 0.479,
+      "step": 4152
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7977946613075885,
+      "learning_rate": 4.847212617039037e-06,
+      "loss": 0.649,
+      "step": 4153
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5819265200485855,
+      "learning_rate": 4.842653793474389e-06,
+      "loss": 0.4762,
+      "step": 4154
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.804461069604633,
+      "learning_rate": 4.838096429685735e-06,
+      "loss": 0.5025,
+      "step": 4155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.742580916560652,
+      "learning_rate": 4.833540526963027e-06,
+      "loss": 0.5416,
+      "step": 4156
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8275409710722,
+      "learning_rate": 4.828986086595804e-06,
+      "loss": 0.5544,
+      "step": 4157
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0047611385431763,
+      "learning_rate": 4.82443310987319e-06,
+      "loss": 0.4862,
+      "step": 4158
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.13884725578916,
+      "learning_rate": 4.819881598083895e-06,
+      "loss": 0.5345,
+      "step": 4159
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.20595758424668,
+      "learning_rate": 4.815331552516217e-06,
+      "loss": 0.5252,
+      "step": 4160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5705401748893655,
+      "learning_rate": 4.810782974458035e-06,
+      "loss": 0.5149,
+      "step": 4161
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5875525256889804,
+      "learning_rate": 4.806235865196815e-06,
+      "loss": 0.5349,
+      "step": 4162
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.759145367590126,
+      "learning_rate": 4.801690226019606e-06,
+      "loss": 0.5796,
+      "step": 4163
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7985768807068565,
+      "learning_rate": 4.7971460582130425e-06,
+      "loss": 0.608,
+      "step": 4164
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6868745694843863,
+      "learning_rate": 4.792603363063342e-06,
+      "loss": 0.3585,
+      "step": 4165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6365888222146039,
+      "learning_rate": 4.7880621418563035e-06,
+      "loss": 0.4737,
+      "step": 4166
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.911856724420467,
+      "learning_rate": 4.783522395877311e-06,
+      "loss": 0.3768,
+      "step": 4167
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0312641031795486,
+      "learning_rate": 4.77898412641133e-06,
+      "loss": 0.514,
+      "step": 4168
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8219799652206197,
+      "learning_rate": 4.774447334742908e-06,
+      "loss": 0.624,
+      "step": 4169
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6542107326849833,
+      "learning_rate": 4.769912022156175e-06,
+      "loss": 0.5197,
+      "step": 4170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0724324803567553,
+      "learning_rate": 4.7653781899348395e-06,
+      "loss": 0.6462,
+      "step": 4171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8258240656333448,
+      "learning_rate": 4.760845839362196e-06,
+      "loss": 0.6742,
+      "step": 4172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8280874068093708,
+      "learning_rate": 4.756314971721115e-06,
+      "loss": 0.658,
+      "step": 4173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0118958293447244,
+      "learning_rate": 4.751785588294053e-06,
+      "loss": 0.5354,
+      "step": 4174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8790321277715708,
+      "learning_rate": 4.7472576903630314e-06,
+      "loss": 0.7022,
+      "step": 4175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.158711601529014,
+      "learning_rate": 4.742731279209674e-06,
+      "loss": 0.4968,
+      "step": 4176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.593783212149061,
+      "learning_rate": 4.738206356115167e-06,
+      "loss": 0.4231,
+      "step": 4177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9059280906842895,
+      "learning_rate": 4.733682922360282e-06,
+      "loss": 0.5667,
+      "step": 4178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8872313781869753,
+      "learning_rate": 4.729160979225365e-06,
+      "loss": 0.6937,
+      "step": 4179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.761020961844344,
+      "learning_rate": 4.724640527990345e-06,
+      "loss": 0.5912,
+      "step": 4180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4241411104379824,
+      "learning_rate": 4.720121569934726e-06,
+      "loss": 0.5262,
+      "step": 4181
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.944222511709929,
+      "learning_rate": 4.715604106337587e-06,
+      "loss": 0.5753,
+      "step": 4182
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6592726440945496,
+      "learning_rate": 4.71108813847759e-06,
+      "loss": 0.5132,
+      "step": 4183
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9934100066587621,
+      "learning_rate": 4.706573667632967e-06,
+      "loss": 0.5266,
+      "step": 4184
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.119106115877144,
+      "learning_rate": 4.702060695081532e-06,
+      "loss": 0.6274,
+      "step": 4185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8015111500618058,
+      "learning_rate": 4.69754922210067e-06,
+      "loss": 0.6662,
+      "step": 4186
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7937837631891271,
+      "learning_rate": 4.693039249967347e-06,
+      "loss": 0.5685,
+      "step": 4187
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.734797268846554,
+      "learning_rate": 4.688530779958099e-06,
+      "loss": 0.5425,
+      "step": 4188
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5709033796845642,
+      "learning_rate": 4.684023813349039e-06,
+      "loss": 0.4538,
+      "step": 4189
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0778988943881207,
+      "learning_rate": 4.679518351415855e-06,
+      "loss": 0.5793,
+      "step": 4190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8249343579017628,
+      "learning_rate": 4.675014395433808e-06,
+      "loss": 0.6453,
+      "step": 4191
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1725635708311013,
+      "learning_rate": 4.6705119466777334e-06,
+      "loss": 0.5544,
+      "step": 4192
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2415089513569435,
+      "learning_rate": 4.666011006422041e-06,
+      "loss": 0.6017,
+      "step": 4193
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6725802678443482,
+      "learning_rate": 4.661511575940712e-06,
+      "loss": 0.571,
+      "step": 4194
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8402744108204454,
+      "learning_rate": 4.657013656507299e-06,
+      "loss": 0.5096,
+      "step": 4195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4716903136236599,
+      "learning_rate": 4.6525172493949335e-06,
+      "loss": 0.5413,
+      "step": 4196
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7153644732001048,
+      "learning_rate": 4.648022355876307e-06,
+      "loss": 0.5321,
+      "step": 4197
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6700074722547253,
+      "learning_rate": 4.643528977223689e-06,
+      "loss": 0.5309,
+      "step": 4198
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9414709499110434,
+      "learning_rate": 4.63903711470893e-06,
+      "loss": 0.5851,
+      "step": 4199
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6560748860548549,
+      "learning_rate": 4.634546769603436e-06,
+      "loss": 0.5794,
+      "step": 4200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8010654280772653,
+      "learning_rate": 4.6300579431781915e-06,
+      "loss": 0.5438,
+      "step": 4201
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.896384137828404,
+      "learning_rate": 4.625570636703748e-06,
+      "loss": 0.6901,
+      "step": 4202
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8337507065456962,
+      "learning_rate": 4.621084851450229e-06,
+      "loss": 0.6171,
+      "step": 4203
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7517131855807626,
+      "learning_rate": 4.616600588687327e-06,
+      "loss": 0.5982,
+      "step": 4204
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.632966159119073,
+      "learning_rate": 4.6121178496843045e-06,
+      "loss": 0.5225,
+      "step": 4205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.967339123562816,
+      "learning_rate": 4.607636635709988e-06,
+      "loss": 0.5608,
+      "step": 4206
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7214813177800794,
+      "learning_rate": 4.603156948032776e-06,
+      "loss": 0.4403,
+      "step": 4207
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3201573272335594,
+      "learning_rate": 4.5986787879206375e-06,
+      "loss": 0.6098,
+      "step": 4208
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.953041882686842,
+      "learning_rate": 4.594202156641105e-06,
+      "loss": 0.5034,
+      "step": 4209
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1386035274669717,
+      "learning_rate": 4.589727055461278e-06,
+      "loss": 0.582,
+      "step": 4210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7794045180593125,
+      "learning_rate": 4.585253485647826e-06,
+      "loss": 0.4886,
+      "step": 4211
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.536838700047585,
+      "learning_rate": 4.5807814484669835e-06,
+      "loss": 0.5958,
+      "step": 4212
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8351149408978833,
+      "learning_rate": 4.5763109451845515e-06,
+      "loss": 0.515,
+      "step": 4213
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5180895232748348,
+      "learning_rate": 4.571841977065895e-06,
+      "loss": 0.5131,
+      "step": 4214
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2485170402612997,
+      "learning_rate": 4.567374545375948e-06,
+      "loss": 0.5604,
+      "step": 4215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.772133997225084,
+      "learning_rate": 4.562908651379206e-06,
+      "loss": 0.477,
+      "step": 4216
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8075057785340438,
+      "learning_rate": 4.558444296339731e-06,
+      "loss": 0.6681,
+      "step": 4217
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8642001827747061,
+      "learning_rate": 4.553981481521156e-06,
+      "loss": 0.4785,
+      "step": 4218
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7042075051067769,
+      "learning_rate": 4.54952020818666e-06,
+      "loss": 0.4646,
+      "step": 4219
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.733578815305826,
+      "learning_rate": 4.545060477599002e-06,
+      "loss": 0.4578,
+      "step": 4220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.792383883995807,
+      "learning_rate": 4.540602291020499e-06,
+      "loss": 0.5296,
+      "step": 4221
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5899862754544114,
+      "learning_rate": 4.536145649713029e-06,
+      "loss": 0.5061,
+      "step": 4222
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5671739964746094,
+      "learning_rate": 4.531690554938043e-06,
+      "loss": 0.568,
+      "step": 4223
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8344313325909751,
+      "learning_rate": 4.52723700795654e-06,
+      "loss": 0.5198,
+      "step": 4224
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.00195331252275,
+      "learning_rate": 4.522785010029087e-06,
+      "loss": 0.4187,
+      "step": 4225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8299349210863116,
+      "learning_rate": 4.518334562415816e-06,
+      "loss": 0.4561,
+      "step": 4226
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8061660028643746,
+      "learning_rate": 4.513885666376413e-06,
+      "loss": 0.659,
+      "step": 4227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8909425021764459,
+      "learning_rate": 4.509438323170131e-06,
+      "loss": 0.4897,
+      "step": 4228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.177810420041096,
+      "learning_rate": 4.504992534055781e-06,
+      "loss": 0.585,
+      "step": 4229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1558258963455486,
+      "learning_rate": 4.500548300291732e-06,
+      "loss": 0.5604,
+      "step": 4230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6564571678348012,
+      "learning_rate": 4.496105623135919e-06,
+      "loss": 0.4893,
+      "step": 4231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9867530139360672,
+      "learning_rate": 4.4916645038458295e-06,
+      "loss": 0.4363,
+      "step": 4232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4597411989504665,
+      "learning_rate": 4.487224943678513e-06,
+      "loss": 0.4941,
+      "step": 4233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4837176914156194,
+      "learning_rate": 4.482786943890579e-06,
+      "loss": 0.539,
+      "step": 4234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.364869604163935,
+      "learning_rate": 4.478350505738194e-06,
+      "loss": 0.507,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7915039718549868,
+      "learning_rate": 4.47391563047708e-06,
+      "loss": 0.4979,
+      "step": 4236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8381011364679926,
+      "learning_rate": 4.4694823193625225e-06,
+      "loss": 0.6518,
+      "step": 4237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9092159590432216,
+      "learning_rate": 4.465050573649359e-06,
+      "loss": 0.573,
+      "step": 4238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8064009760375536,
+      "learning_rate": 4.460620394591989e-06,
+      "loss": 0.6771,
+      "step": 4239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.935177371998135,
+      "learning_rate": 4.45619178344436e-06,
+      "loss": 0.5016,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3815046930202284,
+      "learning_rate": 4.451764741459983e-06,
+      "loss": 0.4525,
+      "step": 4241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.761405149097482,
+      "learning_rate": 4.447339269891923e-06,
+      "loss": 0.5302,
+      "step": 4242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4652293778581955,
+      "learning_rate": 4.442915369992802e-06,
+      "loss": 0.5831,
+      "step": 4243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.818305187023267,
+      "learning_rate": 4.438493043014793e-06,
+      "loss": 0.478,
+      "step": 4244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.514819071444575,
+      "learning_rate": 4.434072290209624e-06,
+      "loss": 0.5389,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.950162319096439,
+      "learning_rate": 4.429653112828589e-06,
+      "loss": 0.4941,
+      "step": 4246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6669773576381066,
+      "learning_rate": 4.4252355121225196e-06,
+      "loss": 0.5115,
+      "step": 4247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6673055245520958,
+      "learning_rate": 4.4208194893418125e-06,
+      "loss": 0.5482,
+      "step": 4248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.013775383199594,
+      "learning_rate": 4.41640504573641e-06,
+      "loss": 0.4634,
+      "step": 4249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5057526068579423,
+      "learning_rate": 4.411992182555812e-06,
+      "loss": 0.5318,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7706986826791082,
+      "learning_rate": 4.407580901049071e-06,
+      "loss": 0.5241,
+      "step": 4251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015167468470785,
+      "learning_rate": 4.403171202464791e-06,
+      "loss": 0.5347,
+      "step": 4252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.410355296287593,
+      "learning_rate": 4.398763088051127e-06,
+      "loss": 0.5459,
+      "step": 4253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8637968974063569,
+      "learning_rate": 4.394356559055787e-06,
+      "loss": 0.6857,
+      "step": 4254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0504749687390382,
+      "learning_rate": 4.389951616726029e-06,
+      "loss": 0.5337,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5916299860783554,
+      "learning_rate": 4.3855482623086645e-06,
+      "loss": 0.4737,
+      "step": 4256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7415910955374794,
+      "learning_rate": 4.381146497050053e-06,
+      "loss": 0.5029,
+      "step": 4257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8742035414581126,
+      "learning_rate": 4.3767463221961034e-06,
+      "loss": 0.6243,
+      "step": 4258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.859216467408125,
+      "learning_rate": 4.372347738992278e-06,
+      "loss": 0.5404,
+      "step": 4259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.074371531893287,
+      "learning_rate": 4.3679507486835835e-06,
+      "loss": 0.527,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7741255353248802,
+      "learning_rate": 4.363555352514587e-06,
+      "loss": 0.5121,
+      "step": 4261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5276558330846406,
+      "learning_rate": 4.359161551729385e-06,
+      "loss": 0.6011,
+      "step": 4262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5664664524824607,
+      "learning_rate": 4.354769347571638e-06,
+      "loss": 0.4984,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6015569806966188,
+      "learning_rate": 4.350378741284551e-06,
+      "loss": 0.5362,
+      "step": 4264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.243381548343782,
+      "learning_rate": 4.3459897341108756e-06,
+      "loss": 0.4976,
+      "step": 4265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9481277457926385,
+      "learning_rate": 4.341602327292912e-06,
+      "loss": 0.5174,
+      "step": 4266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.532639958568079,
+      "learning_rate": 4.3372165220725045e-06,
+      "loss": 0.5613,
+      "step": 4267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0106329407444856,
+      "learning_rate": 4.332832319691044e-06,
+      "loss": 0.5091,
+      "step": 4268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8732040494610829,
+      "learning_rate": 4.328449721389475e-06,
+      "loss": 0.5217,
+      "step": 4269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8636840376468942,
+      "learning_rate": 4.324068728408282e-06,
+      "loss": 0.5916,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9464713887785015,
+      "learning_rate": 4.319689341987493e-06,
+      "loss": 0.5346,
+      "step": 4271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8151714167075697,
+      "learning_rate": 4.315311563366686e-06,
+      "loss": 0.51,
+      "step": 4272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0888667496123716,
+      "learning_rate": 4.3109353937849815e-06,
+      "loss": 0.5667,
+      "step": 4273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.976430473787991,
+      "learning_rate": 4.306560834481045e-06,
+      "loss": 0.5384,
+      "step": 4274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5985317792537337,
+      "learning_rate": 4.302187886693087e-06,
+      "loss": 0.5283,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.891175320539059,
+      "learning_rate": 4.29781655165886e-06,
+      "loss": 0.4841,
+      "step": 4276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.035676065806014,
+      "learning_rate": 4.293446830615662e-06,
+      "loss": 0.4735,
+      "step": 4277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5658959237836942,
+      "learning_rate": 4.289078724800331e-06,
+      "loss": 0.5422,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8104934592357151,
+      "learning_rate": 4.2847122354492555e-06,
+      "loss": 0.5494,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7519612351618523,
+      "learning_rate": 4.280347363798356e-06,
+      "loss": 0.56,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.552457764379727,
+      "learning_rate": 4.275984111083102e-06,
+      "loss": 0.4793,
+      "step": 4281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.943111472764431,
+      "learning_rate": 4.2716224785385075e-06,
+      "loss": 0.536,
+      "step": 4282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4595620485155147,
+      "learning_rate": 4.267262467399114e-06,
+      "loss": 0.5536,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.806168746012722,
+      "learning_rate": 4.2629040788990205e-06,
+      "loss": 0.4922,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9090816022902093,
+      "learning_rate": 4.258547314271857e-06,
+      "loss": 0.6764,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6341226892881027,
+      "learning_rate": 4.254192174750796e-06,
+      "loss": 0.581,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7398960034092887,
+      "learning_rate": 4.249838661568554e-06,
+      "loss": 0.574,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9455699582590613,
+      "learning_rate": 4.24548677595738e-06,
+      "loss": 0.4998,
+      "step": 4288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8035411579569933,
+      "learning_rate": 4.2411365191490684e-06,
+      "loss": 0.5353,
+      "step": 4289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8324541080351848,
+      "learning_rate": 4.236787892374948e-06,
+      "loss": 0.6507,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4457149403305192,
+      "learning_rate": 4.232440896865888e-06,
+      "loss": 0.4923,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2802817352698141,
+      "learning_rate": 4.2280955338523015e-06,
+      "loss": 0.4364,
+      "step": 4292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5203324682827821,
+      "learning_rate": 4.22375180456413e-06,
+      "loss": 0.5241,
+      "step": 4293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6569065255498,
+      "learning_rate": 4.219409710230859e-06,
+      "loss": 0.5211,
+      "step": 4294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8885086143580465,
+      "learning_rate": 4.215069252081509e-06,
+      "loss": 0.7071,
+      "step": 4295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3628214928221847,
+      "learning_rate": 4.210730431344635e-06,
+      "loss": 0.5568,
+      "step": 4296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5224088349538774,
+      "learning_rate": 4.206393249248334e-06,
+      "loss": 0.4893,
+      "step": 4297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5728259333475325,
+      "learning_rate": 4.202057707020235e-06,
+      "loss": 0.4904,
+      "step": 4298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8816610659771875,
+      "learning_rate": 4.1977238058875045e-06,
+      "loss": 0.5117,
+      "step": 4299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6518810055512243,
+      "learning_rate": 4.193391547076844e-06,
+      "loss": 0.568,
+      "step": 4300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9232699874940167,
+      "learning_rate": 4.189060931814489e-06,
+      "loss": 0.5662,
+      "step": 4301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6602008154462986,
+      "learning_rate": 4.184731961326213e-06,
+      "loss": 0.5044,
+      "step": 4302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9090567052533618,
+      "learning_rate": 4.180404636837321e-06,
+      "loss": 0.598,
+      "step": 4303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.698758573521574,
+      "learning_rate": 4.176078959572656e-06,
+      "loss": 0.5404,
+      "step": 4304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.822843786998643,
+      "learning_rate": 4.171754930756586e-06,
+      "loss": 0.5254,
+      "step": 4305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.789555748210641,
+      "learning_rate": 4.167432551613021e-06,
+      "loss": 0.4815,
+      "step": 4306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7458547842571974,
+      "learning_rate": 4.163111823365403e-06,
+      "loss": 0.5589,
+      "step": 4307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6486406946533245,
+      "learning_rate": 4.158792747236702e-06,
+      "loss": 0.4377,
+      "step": 4308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.49261364304359,
+      "learning_rate": 4.154475324449425e-06,
+      "loss": 0.5225,
+      "step": 4309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5755626984820483,
+      "learning_rate": 4.1501595562256105e-06,
+      "loss": 0.5099,
+      "step": 4310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8606491315336067,
+      "learning_rate": 4.145845443786827e-06,
+      "loss": 0.5255,
+      "step": 4311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6086502040436843,
+      "learning_rate": 4.141532988354173e-06,
+      "loss": 0.5247,
+      "step": 4312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9703588212365122,
+      "learning_rate": 4.137222191148282e-06,
+      "loss": 0.5412,
+      "step": 4313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.148695307199683,
+      "learning_rate": 4.132913053389317e-06,
+      "loss": 0.5016,
+      "step": 4314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6391696859546294,
+      "learning_rate": 4.128605576296964e-06,
+      "loss": 0.4255,
+      "step": 4315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.67227946307355,
+      "learning_rate": 4.1242997610904546e-06,
+      "loss": 0.4729,
+      "step": 4316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7123167505911432,
+      "learning_rate": 4.119995608988536e-06,
+      "loss": 0.5385,
+      "step": 4317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5732828568602426,
+      "learning_rate": 4.11569312120949e-06,
+      "loss": 0.5431,
+      "step": 4318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7695447474251538,
+      "learning_rate": 4.111392298971127e-06,
+      "loss": 0.5457,
+      "step": 4319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7097619479062336,
+      "learning_rate": 4.107093143490785e-06,
+      "loss": 0.5585,
+      "step": 4320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8543136979462391,
+      "learning_rate": 4.102795655985331e-06,
+      "loss": 0.6361,
+      "step": 4321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.732746841847853,
+      "learning_rate": 4.098499837671159e-06,
+      "loss": 0.6004,
+      "step": 4322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.639076667732441,
+      "learning_rate": 4.0942056897641934e-06,
+      "loss": 0.4883,
+      "step": 4323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8405405179577656,
+      "learning_rate": 4.089913213479882e-06,
+      "loss": 0.5486,
+      "step": 4324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1978046230701054,
+      "learning_rate": 4.085622410033203e-06,
+      "loss": 0.5198,
+      "step": 4325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.792774727459377,
+      "learning_rate": 4.081333280638661e-06,
+      "loss": 0.5996,
+      "step": 4326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7010887273908653,
+      "learning_rate": 4.077045826510277e-06,
+      "loss": 0.6216,
+      "step": 4327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6180440585965425,
+      "learning_rate": 4.072760048861614e-06,
+      "loss": 0.5848,
+      "step": 4328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5506344981032751,
+      "learning_rate": 4.068475948905746e-06,
+      "loss": 0.5466,
+      "step": 4329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6424349706984507,
+      "learning_rate": 4.064193527855285e-06,
+      "loss": 0.5579,
+      "step": 4330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4884022191177233,
+      "learning_rate": 4.0599127869223565e-06,
+      "loss": 0.4707,
+      "step": 4331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7548303643166798,
+      "learning_rate": 4.055633727318617e-06,
+      "loss": 0.5861,
+      "step": 4332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9119869062858286,
+      "learning_rate": 4.051356350255246e-06,
+      "loss": 0.5503,
+      "step": 4333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6939912503982344,
+      "learning_rate": 4.047080656942943e-06,
+      "loss": 0.5389,
+      "step": 4334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8242087680103773,
+      "learning_rate": 4.042806648591938e-06,
+      "loss": 0.5677,
+      "step": 4335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.816376414246153,
+      "learning_rate": 4.038534326411978e-06,
+      "loss": 0.5759,
+      "step": 4336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9291254762977983,
+      "learning_rate": 4.0342636916123355e-06,
+      "loss": 0.5587,
+      "step": 4337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8319182832241867,
+      "learning_rate": 4.0299947454018e-06,
+      "loss": 0.6245,
+      "step": 4338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5933211395141824,
+      "learning_rate": 4.025727488988696e-06,
+      "loss": 0.4904,
+      "step": 4339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6504591024563624,
+      "learning_rate": 4.0214619235808575e-06,
+      "loss": 0.5414,
+      "step": 4340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7782113033485136,
+      "learning_rate": 4.017198050385644e-06,
+      "loss": 0.6456,
+      "step": 4341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8819442611806667,
+      "learning_rate": 4.012935870609934e-06,
+      "loss": 0.5291,
+      "step": 4342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6488349265701463,
+      "learning_rate": 4.008675385460131e-06,
+      "loss": 0.5339,
+      "step": 4343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7562219937010628,
+      "learning_rate": 4.0044165961421565e-06,
+      "loss": 0.4318,
+      "step": 4344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6955075605845247,
+      "learning_rate": 4.000159503861451e-06,
+      "loss": 0.5451,
+      "step": 4345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8859968589621399,
+      "learning_rate": 3.9959041098229735e-06,
+      "loss": 0.5352,
+      "step": 4346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5945489685291607,
+      "learning_rate": 3.991650415231211e-06,
+      "loss": 0.5377,
+      "step": 4347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6412649080436814,
+      "learning_rate": 3.987398421290155e-06,
+      "loss": 0.5285,
+      "step": 4348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3590265058481577,
+      "learning_rate": 3.983148129203326e-06,
+      "loss": 0.6015,
+      "step": 4349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.194937172795404,
+      "learning_rate": 3.978899540173759e-06,
+      "loss": 0.6,
+      "step": 4350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203069811006013,
+      "learning_rate": 3.974652655404012e-06,
+      "loss": 0.5357,
+      "step": 4351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9116053463543916,
+      "learning_rate": 3.970407476096154e-06,
+      "loss": 0.4985,
+      "step": 4352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.880652154447259,
+      "learning_rate": 3.966164003451775e-06,
+      "loss": 0.4906,
+      "step": 4353
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0036433112880463,
+      "learning_rate": 3.961922238671981e-06,
+      "loss": 0.5846,
+      "step": 4354
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5654638732856567,
+      "learning_rate": 3.957682182957394e-06,
+      "loss": 0.474,
+      "step": 4355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8057961157058378,
+      "learning_rate": 3.953443837508153e-06,
+      "loss": 0.5356,
+      "step": 4356
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6151239456580313,
+      "learning_rate": 3.949207203523913e-06,
+      "loss": 0.4835,
+      "step": 4357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2758637071141945,
+      "learning_rate": 3.944972282203844e-06,
+      "loss": 0.5046,
+      "step": 4358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.338281232114482,
+      "learning_rate": 3.940739074746632e-06,
+      "loss": 0.4471,
+      "step": 4359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5184115343284694,
+      "learning_rate": 3.936507582350479e-06,
+      "loss": 0.4545,
+      "step": 4360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9972431625870912,
+      "learning_rate": 3.932277806213093e-06,
+      "loss": 0.5695,
+      "step": 4361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7404981792912451,
+      "learning_rate": 3.928049747531711e-06,
+      "loss": 0.634,
+      "step": 4362
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.618101829282415,
+      "learning_rate": 3.923823407503076e-06,
+      "loss": 0.5452,
+      "step": 4363
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.381165679989678,
+      "learning_rate": 3.919598787323442e-06,
+      "loss": 0.4718,
+      "step": 4364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.026914058111266,
+      "learning_rate": 3.915375888188579e-06,
+      "loss": 0.6131,
+      "step": 4365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6945389525533547,
+      "learning_rate": 3.9111547112937685e-06,
+      "loss": 0.4886,
+      "step": 4366
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5092918258437114,
+      "learning_rate": 3.906935257833809e-06,
+      "loss": 0.5504,
+      "step": 4367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6907152766896716,
+      "learning_rate": 3.902717529003005e-06,
+      "loss": 0.537,
+      "step": 4368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5610445418252126,
+      "learning_rate": 3.898501525995181e-06,
+      "loss": 0.5423,
+      "step": 4369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8183448015318022,
+      "learning_rate": 3.89428725000366e-06,
+      "loss": 0.4981,
+      "step": 4370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9385833502708831,
+      "learning_rate": 3.890074702221288e-06,
+      "loss": 0.5082,
+      "step": 4371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9714794063913592,
+      "learning_rate": 3.8858638838404175e-06,
+      "loss": 0.4936,
+      "step": 4372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9389022712682815,
+      "learning_rate": 3.88165479605291e-06,
+      "loss": 0.6073,
+      "step": 4373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4883609831718507,
+      "learning_rate": 3.8774474400501415e-06,
+      "loss": 0.4421,
+      "step": 4374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.217688118044507,
+      "learning_rate": 3.873241817022996e-06,
+      "loss": 0.5283,
+      "step": 4375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.384346869012373,
+      "learning_rate": 3.869037928161863e-06,
+      "loss": 0.5371,
+      "step": 4376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8487727758849639,
+      "learning_rate": 3.8648357746566456e-06,
+      "loss": 0.6571,
+      "step": 4377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6489121992579656,
+      "learning_rate": 3.860635357696756e-06,
+      "loss": 0.543,
+      "step": 4378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1636711065043164,
+      "learning_rate": 3.8564366784711116e-06,
+      "loss": 0.5705,
+      "step": 4379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0336882044391684,
+      "learning_rate": 3.852239738168141e-06,
+      "loss": 0.5451,
+      "step": 4380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5452852548278921,
+      "learning_rate": 3.848044537975778e-06,
+      "loss": 0.5468,
+      "step": 4381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4304943118931617,
+      "learning_rate": 3.843851079081467e-06,
+      "loss": 0.4501,
+      "step": 4382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8006085684330582,
+      "learning_rate": 3.839659362672156e-06,
+      "loss": 0.5841,
+      "step": 4383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.547774955222374,
+      "learning_rate": 3.835469389934299e-06,
+      "loss": 0.5549,
+      "step": 4384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7307892385227122,
+      "learning_rate": 3.8312811620538655e-06,
+      "loss": 0.4909,
+      "step": 4385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5932724327945538,
+      "learning_rate": 3.8270946802163216e-06,
+      "loss": 0.4717,
+      "step": 4386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7094652711516556,
+      "learning_rate": 3.822909945606641e-06,
+      "loss": 0.5627,
+      "step": 4387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6582603560431985,
+      "learning_rate": 3.818726959409305e-06,
+      "loss": 0.5084,
+      "step": 4388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.156711405788705,
+      "learning_rate": 3.8145457228082995e-06,
+      "loss": 0.5009,
+      "step": 4389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7303175221701,
+      "learning_rate": 3.8103662369871143e-06,
+      "loss": 0.4487,
+      "step": 4390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7373477226759113,
+      "learning_rate": 3.806188503128746e-06,
+      "loss": 0.5266,
+      "step": 4391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.461717855266737,
+      "learning_rate": 3.802012522415689e-06,
+      "loss": 0.5882,
+      "step": 4392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7676874670829112,
+      "learning_rate": 3.7978382960299476e-06,
+      "loss": 0.4912,
+      "step": 4393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8082230301008404,
+      "learning_rate": 3.793665825153029e-06,
+      "loss": 0.4967,
+      "step": 4394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5215936534642611,
+      "learning_rate": 3.7894951109659404e-06,
+      "loss": 0.4768,
+      "step": 4395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3564040823124681,
+      "learning_rate": 3.785326154649196e-06,
+      "loss": 0.5123,
+      "step": 4396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.645404114847276,
+      "learning_rate": 3.781158957382809e-06,
+      "loss": 0.5076,
+      "step": 4397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.827494920242475,
+      "learning_rate": 3.776993520346295e-06,
+      "loss": 0.4937,
+      "step": 4398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8807057125799521,
+      "learning_rate": 3.772829844718674e-06,
+      "loss": 0.701,
+      "step": 4399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8354961952031887,
+      "learning_rate": 3.7686679316784635e-06,
+      "loss": 0.5511,
+      "step": 4400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6451920306099392,
+      "learning_rate": 3.764507782403686e-06,
+      "loss": 0.4762,
+      "step": 4401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5339154608824352,
+      "learning_rate": 3.760349398071862e-06,
+      "loss": 0.4614,
+      "step": 4402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7863068954179238,
+      "learning_rate": 3.756192779860014e-06,
+      "loss": 0.5328,
+      "step": 4403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.385386587752323,
+      "learning_rate": 3.752037928944664e-06,
+      "loss": 0.5356,
+      "step": 4404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5878973322755956,
+      "learning_rate": 3.7478848465018336e-06,
+      "loss": 0.4974,
+      "step": 4405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.053489167396226,
+      "learning_rate": 3.7437335337070445e-06,
+      "loss": 0.4761,
+      "step": 4406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1554567752332696,
+      "learning_rate": 3.739583991735316e-06,
+      "loss": 0.5461,
+      "step": 4407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.785870471324166,
+      "learning_rate": 3.7354362217611652e-06,
+      "loss": 0.6004,
+      "step": 4408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.830122562357554,
+      "learning_rate": 3.7312902249586146e-06,
+      "loss": 0.6896,
+      "step": 4409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.738127215092363,
+      "learning_rate": 3.7271460025011785e-06,
+      "loss": 0.5493,
+      "step": 4410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6726135898666328,
+      "learning_rate": 3.723003555561869e-06,
+      "loss": 0.4595,
+      "step": 4411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9094613431093408,
+      "learning_rate": 3.7188628853132023e-06,
+      "loss": 0.5915,
+      "step": 4412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.108704486651538,
+      "learning_rate": 3.714723992927177e-06,
+      "loss": 0.5889,
+      "step": 4413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6052018403621626,
+      "learning_rate": 3.710586879575302e-06,
+      "loss": 0.5123,
+      "step": 4414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.577080469841862,
+      "learning_rate": 3.70645154642858e-06,
+      "loss": 0.4705,
+      "step": 4415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6157878589002026,
+      "learning_rate": 3.702317994657506e-06,
+      "loss": 0.5107,
+      "step": 4416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4871204433109,
+      "learning_rate": 3.6981862254320757e-06,
+      "loss": 0.5421,
+      "step": 4417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2521764482181426,
+      "learning_rate": 3.694056239921776e-06,
+      "loss": 0.5243,
+      "step": 4418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7565568247181964,
+      "learning_rate": 3.689928039295592e-06,
+      "loss": 0.5412,
+      "step": 4419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9291651971877402,
+      "learning_rate": 3.6858016247219998e-06,
+      "loss": 0.5386,
+      "step": 4420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6887857765406022,
+      "learning_rate": 3.6816769973689736e-06,
+      "loss": 0.5115,
+      "step": 4421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9030600640494053,
+      "learning_rate": 3.677554158403982e-06,
+      "loss": 0.5798,
+      "step": 4422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2272957453684867,
+      "learning_rate": 3.6734331089939835e-06,
+      "loss": 0.5937,
+      "step": 4423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9392129194035348,
+      "learning_rate": 3.669313850305435e-06,
+      "loss": 0.5427,
+      "step": 4424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7396012989538312,
+      "learning_rate": 3.6651963835042813e-06,
+      "loss": 0.5536,
+      "step": 4425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.816988194752822,
+      "learning_rate": 3.6610807097559644e-06,
+      "loss": 0.6436,
+      "step": 4426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6522460990210095,
+      "learning_rate": 3.6569668302254167e-06,
+      "loss": 0.5287,
+      "step": 4427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.01833071035824,
+      "learning_rate": 3.6528547460770636e-06,
+      "loss": 0.5647,
+      "step": 4428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7292100771939485,
+      "learning_rate": 3.648744458474821e-06,
+      "loss": 0.5547,
+      "step": 4429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8834425873544847,
+      "learning_rate": 3.6446359685820974e-06,
+      "loss": 0.4974,
+      "step": 4430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8227283113817305,
+      "learning_rate": 3.6405292775617886e-06,
+      "loss": 0.5226,
+      "step": 4431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.395309157580167,
+      "learning_rate": 3.6364243865762926e-06,
+      "loss": 0.563,
+      "step": 4432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7892597494546336,
+      "learning_rate": 3.6323212967874866e-06,
+      "loss": 0.4771,
+      "step": 4433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.827874720192868,
+      "learning_rate": 3.628220009356743e-06,
+      "loss": 0.477,
+      "step": 4434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.766032618329193,
+      "learning_rate": 3.6241205254449197e-06,
+      "loss": 0.4749,
+      "step": 4435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6352778502035417,
+      "learning_rate": 3.6200228462123666e-06,
+      "loss": 0.6025,
+      "step": 4436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.684501606844399,
+      "learning_rate": 3.6159269728189237e-06,
+      "loss": 0.5129,
+      "step": 4437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.143852506360829,
+      "learning_rate": 3.6118329064239222e-06,
+      "loss": 0.5128,
+      "step": 4438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4305992344402605,
+      "learning_rate": 3.6077406481861756e-06,
+      "loss": 0.5371,
+      "step": 4439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6251139676765327,
+      "learning_rate": 3.6036501992639907e-06,
+      "loss": 0.5078,
+      "step": 4440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6758155616019588,
+      "learning_rate": 3.59956156081516e-06,
+      "loss": 0.5433,
+      "step": 4441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7853979401187023,
+      "learning_rate": 3.5954747339969653e-06,
+      "loss": 0.422,
+      "step": 4442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7049304852716773,
+      "learning_rate": 3.5913897199661716e-06,
+      "loss": 0.4499,
+      "step": 4443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.179829826234965,
+      "learning_rate": 3.587306519879037e-06,
+      "loss": 0.4981,
+      "step": 4444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.860197454547813,
+      "learning_rate": 3.5832251348912995e-06,
+      "loss": 0.4881,
+      "step": 4445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.248497936457971,
+      "learning_rate": 3.5791455661581877e-06,
+      "loss": 0.5396,
+      "step": 4446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.303936063440581,
+      "learning_rate": 3.5750678148344153e-06,
+      "loss": 0.5904,
+      "step": 4447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5925820844742564,
+      "learning_rate": 3.5709918820741816e-06,
+      "loss": 0.5558,
+      "step": 4448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7795671367282522,
+      "learning_rate": 3.5669177690311696e-06,
+      "loss": 0.5955,
+      "step": 4449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.407019938353829,
+      "learning_rate": 3.56284547685855e-06,
+      "loss": 0.4904,
+      "step": 4450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0088826996138227,
+      "learning_rate": 3.5587750067089745e-06,
+      "loss": 0.5116,
+      "step": 4451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6688120050229978,
+      "learning_rate": 3.5547063597345833e-06,
+      "loss": 0.5159,
+      "step": 4452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2596709776009045,
+      "learning_rate": 3.5506395370869963e-06,
+      "loss": 0.5196,
+      "step": 4453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5410515224566526,
+      "learning_rate": 3.546574539917317e-06,
+      "loss": 0.5329,
+      "step": 4454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8127307868608604,
+      "learning_rate": 3.5425113693761436e-06,
+      "loss": 0.6334,
+      "step": 4455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8778481843946635,
+      "learning_rate": 3.5384500266135393e-06,
+      "loss": 0.5351,
+      "step": 4456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2784384383579046,
+      "learning_rate": 3.5343905127790614e-06,
+      "loss": 0.5177,
+      "step": 4457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.65604068038394,
+      "learning_rate": 3.5303328290217453e-06,
+      "loss": 0.4846,
+      "step": 4458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.174169897313449,
+      "learning_rate": 3.526276976490112e-06,
+      "loss": 0.5517,
+      "step": 4459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5828202342576392,
+      "learning_rate": 3.52222295633216e-06,
+      "loss": 0.4582,
+      "step": 4460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5277423537645831,
+      "learning_rate": 3.5181707696953728e-06,
+      "loss": 0.5204,
+      "step": 4461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7826009354423278,
+      "learning_rate": 3.5141204177267117e-06,
+      "loss": 0.4496,
+      "step": 4462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.477482796920771,
+      "learning_rate": 3.5100719015726228e-06,
+      "loss": 0.4924,
+      "step": 4463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.826581337325004,
+      "learning_rate": 3.506025222379027e-06,
+      "loss": 0.5025,
+      "step": 4464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1495067354368076,
+      "learning_rate": 3.501980381291331e-06,
+      "loss": 0.5327,
+      "step": 4465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.544868449317594,
+      "learning_rate": 3.497937379454417e-06,
+      "loss": 0.4873,
+      "step": 4466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8329098771193582,
+      "learning_rate": 3.493896218012649e-06,
+      "loss": 0.501,
+      "step": 4467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5500502195871388,
+      "learning_rate": 3.4898568981098678e-06,
+      "loss": 0.4852,
+      "step": 4468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6893787443941743,
+      "learning_rate": 3.4858194208893967e-06,
+      "loss": 0.5336,
+      "step": 4469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8991813049159934,
+      "learning_rate": 3.481783787494033e-06,
+      "loss": 0.6898,
+      "step": 4470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.663240377155316,
+      "learning_rate": 3.477749999066056e-06,
+      "loss": 0.5349,
+      "step": 4471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7331414327811945,
+      "learning_rate": 3.4737180567472196e-06,
+      "loss": 0.576,
+      "step": 4472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.719824917378971,
+      "learning_rate": 3.469687961678757e-06,
+      "loss": 0.591,
+      "step": 4473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.85727841782826,
+      "learning_rate": 3.465659715001379e-06,
+      "loss": 0.621,
+      "step": 4474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.469336017036088,
+      "learning_rate": 3.461633317855271e-06,
+      "loss": 0.5485,
+      "step": 4475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0756135014797175,
+      "learning_rate": 3.4576087713800966e-06,
+      "loss": 0.4766,
+      "step": 4476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3765079973109633,
+      "learning_rate": 3.4535860767149963e-06,
+      "loss": 0.4847,
+      "step": 4477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6076894860479476,
+      "learning_rate": 3.4495652349985844e-06,
+      "loss": 0.4433,
+      "step": 4478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9875929380559305,
+      "learning_rate": 3.4455462473689515e-06,
+      "loss": 0.4979,
+      "step": 4479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1189770735436597,
+      "learning_rate": 3.4415291149636642e-06,
+      "loss": 0.5427,
+      "step": 4480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.663002489183957,
+      "learning_rate": 3.4375138389197627e-06,
+      "loss": 0.5049,
+      "step": 4481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8473951783914475,
+      "learning_rate": 3.433500420373763e-06,
+      "loss": 0.4542,
+      "step": 4482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9375070993786068,
+      "learning_rate": 3.429488860461655e-06,
+      "loss": 0.5149,
+      "step": 4483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7038805646781106,
+      "learning_rate": 3.425479160318902e-06,
+      "loss": 0.5427,
+      "step": 4484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6593287361481075,
+      "learning_rate": 3.421471321080441e-06,
+      "loss": 0.5438,
+      "step": 4485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.33474451618335,
+      "learning_rate": 3.4174653438806814e-06,
+      "loss": 0.4945,
+      "step": 4486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4500019030251758,
+      "learning_rate": 3.4134612298535084e-06,
+      "loss": 0.4782,
+      "step": 4487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.578044291282056,
+      "learning_rate": 3.4094589801322773e-06,
+      "loss": 0.5717,
+      "step": 4488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6497048420883385,
+      "learning_rate": 3.4054585958498177e-06,
+      "loss": 0.4746,
+      "step": 4489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0810554990322863,
+      "learning_rate": 3.401460078138428e-06,
+      "loss": 0.5234,
+      "step": 4490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0322451199748377,
+      "learning_rate": 3.3974634281298815e-06,
+      "loss": 0.4724,
+      "step": 4491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6408428701656321,
+      "learning_rate": 3.3934686469554203e-06,
+      "loss": 0.5817,
+      "step": 4492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.905386310640369,
+      "learning_rate": 3.389475735745761e-06,
+      "loss": 0.4761,
+      "step": 4493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8539524102416887,
+      "learning_rate": 3.3854846956310862e-06,
+      "loss": 0.4427,
+      "step": 4494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6142742447345237,
+      "learning_rate": 3.381495527741053e-06,
+      "loss": 0.5043,
+      "step": 4495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0181342791621555,
+      "learning_rate": 3.377508233204787e-06,
+      "loss": 0.5045,
+      "step": 4496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6579753577470562,
+      "learning_rate": 3.3735228131508824e-06,
+      "loss": 0.5781,
+      "step": 4497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8324035377017602,
+      "learning_rate": 3.3695392687074045e-06,
+      "loss": 0.5678,
+      "step": 4498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.134970383100749,
+      "learning_rate": 3.3655576010018875e-06,
+      "loss": 0.5612,
+      "step": 4499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8259567370288357,
+      "learning_rate": 3.361577811161335e-06,
+      "loss": 0.5188,
+      "step": 4500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6703294936193753,
+      "learning_rate": 3.3575999003122162e-06,
+      "loss": 0.5274,
+      "step": 4501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.558106088762108,
+      "learning_rate": 3.3536238695804713e-06,
+      "loss": 0.5958,
+      "step": 4502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8405438330300221,
+      "learning_rate": 3.3496497200915067e-06,
+      "loss": 0.4385,
+      "step": 4503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0677205016785747,
+      "learning_rate": 3.3456774529701987e-06,
+      "loss": 0.5812,
+      "step": 4504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5226409995585093,
+      "learning_rate": 3.3417070693408882e-06,
+      "loss": 0.4452,
+      "step": 4505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7931234318996274,
+      "learning_rate": 3.3377385703273835e-06,
+      "loss": 0.6132,
+      "step": 4506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8422358645954586,
+      "learning_rate": 3.3337719570529603e-06,
+      "loss": 0.6055,
+      "step": 4507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6860598894732002,
+      "learning_rate": 3.3298072306403595e-06,
+      "loss": 0.5907,
+      "step": 4508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8165251261558015,
+      "learning_rate": 3.32584439221179e-06,
+      "loss": 0.5212,
+      "step": 4509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.584558110873347,
+      "learning_rate": 3.3218834428889244e-06,
+      "loss": 0.5628,
+      "step": 4510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7584344451599583,
+      "learning_rate": 3.3179243837929e-06,
+      "loss": 0.4911,
+      "step": 4511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5889631085766653,
+      "learning_rate": 3.3139672160443215e-06,
+      "loss": 0.5442,
+      "step": 4512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7638084057753154,
+      "learning_rate": 3.3100119407632556e-06,
+      "loss": 0.5444,
+      "step": 4513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6081932213973826,
+      "learning_rate": 3.306058559069236e-06,
+      "loss": 0.546,
+      "step": 4514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7823034871872057,
+      "learning_rate": 3.3021070720812588e-06,
+      "loss": 0.4841,
+      "step": 4515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8434087579008365,
+      "learning_rate": 3.298157480917783e-06,
+      "loss": 0.4716,
+      "step": 4516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7810542647211172,
+      "learning_rate": 3.2942097866967336e-06,
+      "loss": 0.4855,
+      "step": 4517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7775669733544561,
+      "learning_rate": 3.2902639905354948e-06,
+      "loss": 0.5171,
+      "step": 4518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6793006731615003,
+      "learning_rate": 3.286320093550919e-06,
+      "loss": 0.5095,
+      "step": 4519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7637029464050722,
+      "learning_rate": 3.2823780968593156e-06,
+      "loss": 0.487,
+      "step": 4520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6999698720382463,
+      "learning_rate": 3.2784380015764596e-06,
+      "loss": 0.5072,
+      "step": 4521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9596136280443364,
+      "learning_rate": 3.274499808817586e-06,
+      "loss": 0.5173,
+      "step": 4522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6645758738744516,
+      "learning_rate": 3.2705635196973927e-06,
+      "loss": 0.5649,
+      "step": 4523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9405062985234396,
+      "learning_rate": 3.266629135330037e-06,
+      "loss": 0.4716,
+      "step": 4524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.791181988912811,
+      "learning_rate": 3.2626966568291396e-06,
+      "loss": 0.4902,
+      "step": 4525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9801226693685132,
+      "learning_rate": 3.2587660853077797e-06,
+      "loss": 0.5042,
+      "step": 4526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5961272783748524,
+      "learning_rate": 3.2548374218784963e-06,
+      "loss": 0.5688,
+      "step": 4527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6002349265724372,
+      "learning_rate": 3.2509106676532897e-06,
+      "loss": 0.5144,
+      "step": 4528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6850012635155955,
+      "learning_rate": 3.2469858237436203e-06,
+      "loss": 0.451,
+      "step": 4529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.751428190612212,
+      "learning_rate": 3.243062891260407e-06,
+      "loss": 0.5638,
+      "step": 4530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5785882920643006,
+      "learning_rate": 3.2391418713140264e-06,
+      "loss": 0.4994,
+      "step": 4531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.959458011504432,
+      "learning_rate": 3.235222765014315e-06,
+      "loss": 0.531,
+      "step": 4532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1030821758362683,
+      "learning_rate": 3.231305573470569e-06,
+      "loss": 0.6129,
+      "step": 4533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.225624177336638,
+      "learning_rate": 3.2273902977915405e-06,
+      "loss": 0.5673,
+      "step": 4534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9581934630654105,
+      "learning_rate": 3.2234769390854394e-06,
+      "loss": 0.5597,
+      "step": 4535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.008656537031282,
+      "learning_rate": 3.2195654984599334e-06,
+      "loss": 0.5339,
+      "step": 4536
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6952967953816307,
+      "learning_rate": 3.2156559770221498e-06,
+      "loss": 0.5459,
+      "step": 4537
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.60162003727249,
+      "learning_rate": 3.2117483758786683e-06,
+      "loss": 0.4399,
+      "step": 4538
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6051135819854905,
+      "learning_rate": 3.207842696135527e-06,
+      "loss": 0.4708,
+      "step": 4539
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.106059895279485,
+      "learning_rate": 3.2039389388982225e-06,
+      "loss": 0.4663,
+      "step": 4540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8314218091006139,
+      "learning_rate": 3.200037105271703e-06,
+      "loss": 0.6573,
+      "step": 4541
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8372741407500006,
+      "learning_rate": 3.1961371963603736e-06,
+      "loss": 0.4796,
+      "step": 4542
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.945888155578896,
+      "learning_rate": 3.192239213268099e-06,
+      "loss": 0.4975,
+      "step": 4543
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5894340596107046,
+      "learning_rate": 3.1883431570981917e-06,
+      "loss": 0.5879,
+      "step": 4544
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5408531207841474,
+      "learning_rate": 3.1844490289534236e-06,
+      "loss": 0.5085,
+      "step": 4545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6450555497673687,
+      "learning_rate": 3.180556829936019e-06,
+      "loss": 0.4718,
+      "step": 4546
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7600771057639353,
+      "learning_rate": 3.1766665611476566e-06,
+      "loss": 0.4791,
+      "step": 4547
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.847443189796489,
+      "learning_rate": 3.17277822368947e-06,
+      "loss": 0.5612,
+      "step": 4548
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015203857665195,
+      "learning_rate": 3.168891818662043e-06,
+      "loss": 0.6022,
+      "step": 4549
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6274906502408233,
+      "learning_rate": 3.1650073471654152e-06,
+      "loss": 0.6464,
+      "step": 4550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6877369374494142,
+      "learning_rate": 3.161124810299079e-06,
+      "loss": 0.4464,
+      "step": 4551
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.00908321492733,
+      "learning_rate": 3.157244209161977e-06,
+      "loss": 0.5777,
+      "step": 4552
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7045974291244785,
+      "learning_rate": 3.1533655448525057e-06,
+      "loss": 0.4636,
+      "step": 4553
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.049041339201241,
+      "learning_rate": 3.1494888184685134e-06,
+      "loss": 0.5178,
+      "step": 4554
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7628745660861722,
+      "learning_rate": 3.145614031107299e-06,
+      "loss": 0.5462,
+      "step": 4555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.341896307836286,
+      "learning_rate": 3.141741183865612e-06,
+      "loss": 0.5834,
+      "step": 4556
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8991138234635576,
+      "learning_rate": 3.1378702778396554e-06,
+      "loss": 0.5415,
+      "step": 4557
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.110049268252325,
+      "learning_rate": 3.134001314125079e-06,
+      "loss": 0.6077,
+      "step": 4558
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.862141507487548,
+      "learning_rate": 3.1301342938169854e-06,
+      "loss": 0.5594,
+      "step": 4559
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.060908582192064,
+      "learning_rate": 3.1262692180099285e-06,
+      "loss": 0.4813,
+      "step": 4560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.82197749191487,
+      "learning_rate": 3.1224060877979077e-06,
+      "loss": 0.5546,
+      "step": 4561
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7564100844420978,
+      "learning_rate": 3.1185449042743744e-06,
+      "loss": 0.4684,
+      "step": 4562
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8924862120120856,
+      "learning_rate": 3.114685668532229e-06,
+      "loss": 0.5318,
+      "step": 4563
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.673508487768156,
+      "learning_rate": 3.1108283816638196e-06,
+      "loss": 0.452,
+      "step": 4564
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6639680376448507,
+      "learning_rate": 3.1069730447609423e-06,
+      "loss": 0.4882,
+      "step": 4565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6432849763975377,
+      "learning_rate": 3.103119658914844e-06,
+      "loss": 0.5458,
+      "step": 4566
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.66677065282164,
+      "learning_rate": 3.0992682252162165e-06,
+      "loss": 0.4787,
+      "step": 4567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7939066979812404,
+      "learning_rate": 3.0954187447551996e-06,
+      "loss": 0.5172,
+      "step": 4568
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.571006039634687,
+      "learning_rate": 3.091571218621382e-06,
+      "loss": 0.5345,
+      "step": 4569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.624637835775342,
+      "learning_rate": 3.0877256479037952e-06,
+      "loss": 0.6292,
+      "step": 4570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.779496554794643,
+      "learning_rate": 3.0838820336909224e-06,
+      "loss": 0.4622,
+      "step": 4571
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7637403102079745,
+      "learning_rate": 3.0800403770706912e-06,
+      "loss": 0.5963,
+      "step": 4572
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7524650708331193,
+      "learning_rate": 3.076200679130471e-06,
+      "loss": 0.511,
+      "step": 4573
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1104052848160006,
+      "learning_rate": 3.072362940957083e-06,
+      "loss": 0.5833,
+      "step": 4574
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5312403197563373,
+      "learning_rate": 3.0685271636367895e-06,
+      "loss": 0.5108,
+      "step": 4575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9926576463863497,
+      "learning_rate": 3.064693348255301e-06,
+      "loss": 0.5176,
+      "step": 4576
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4581216588598758,
+      "learning_rate": 3.060861495897769e-06,
+      "loss": 0.5394,
+      "step": 4577
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8310575608018028,
+      "learning_rate": 3.0570316076487918e-06,
+      "loss": 0.6377,
+      "step": 4578
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.089108456485819,
+      "learning_rate": 3.0532036845924107e-06,
+      "loss": 0.4309,
+      "step": 4579
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9546993962722738,
+      "learning_rate": 3.049377727812113e-06,
+      "loss": 0.5661,
+      "step": 4580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1806119209620465,
+      "learning_rate": 3.0455537383908263e-06,
+      "loss": 0.5829,
+      "step": 4581
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.701959698888804,
+      "learning_rate": 3.041731717410923e-06,
+      "loss": 0.5577,
+      "step": 4582
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8657951201588083,
+      "learning_rate": 3.0379116659542186e-06,
+      "loss": 0.6679,
+      "step": 4583
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6867493452928037,
+      "learning_rate": 3.0340935851019694e-06,
+      "loss": 0.5418,
+      "step": 4584
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9608349747844884,
+      "learning_rate": 3.0302774759348797e-06,
+      "loss": 0.4937,
+      "step": 4585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5835511870850496,
+      "learning_rate": 3.0264633395330834e-06,
+      "loss": 0.4413,
+      "step": 4586
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7283066668019353,
+      "learning_rate": 3.022651176976166e-06,
+      "loss": 0.5059,
+      "step": 4587
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.858768080519354,
+      "learning_rate": 3.0188409893431556e-06,
+      "loss": 0.5329,
+      "step": 4588
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.582147290522901,
+      "learning_rate": 3.0150327777125175e-06,
+      "loss": 0.4395,
+      "step": 4589
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8968138621167072,
+      "learning_rate": 3.011226543162156e-06,
+      "loss": 0.5524,
+      "step": 4590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2460110151857244,
+      "learning_rate": 3.007422286769418e-06,
+      "loss": 0.5583,
+      "step": 4591
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7769348906756786,
+      "learning_rate": 3.003620009611091e-06,
+      "loss": 0.4991,
+      "step": 4592
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4537972092359432,
+      "learning_rate": 2.999819712763402e-06,
+      "loss": 0.5639,
+      "step": 4593
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7811778211226335,
+      "learning_rate": 2.996021397302015e-06,
+      "loss": 0.6539,
+      "step": 4594
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7068137627790445,
+      "learning_rate": 2.992225064302037e-06,
+      "loss": 0.5213,
+      "step": 4595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.563743155767783,
+      "learning_rate": 2.988430714838011e-06,
+      "loss": 0.5149,
+      "step": 4596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3995460935300925,
+      "learning_rate": 2.9846383499839205e-06,
+      "loss": 0.5318,
+      "step": 4597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.017672898383242,
+      "learning_rate": 2.9808479708131864e-06,
+      "loss": 0.4746,
+      "step": 4598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.107855399806643,
+      "learning_rate": 2.9770595783986666e-06,
+      "loss": 0.4773,
+      "step": 4599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.901474890966373,
+      "learning_rate": 2.9732731738126586e-06,
+      "loss": 0.5124,
+      "step": 4600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.669978573001662,
+      "learning_rate": 2.969488758126896e-06,
+      "loss": 0.4629,
+      "step": 4601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.898222764541306,
+      "learning_rate": 2.965706332412549e-06,
+      "loss": 0.481,
+      "step": 4602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7966344562151156,
+      "learning_rate": 2.9619258977402253e-06,
+      "loss": 0.6727,
+      "step": 4603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7762952583268574,
+      "learning_rate": 2.9581474551799703e-06,
+      "loss": 0.5456,
+      "step": 4604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.050440973619795,
+      "learning_rate": 2.9543710058012633e-06,
+      "loss": 0.5773,
+      "step": 4605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.980034167119197,
+      "learning_rate": 2.9505965506730195e-06,
+      "loss": 0.5972,
+      "step": 4606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6069337320469008,
+      "learning_rate": 2.946824090863596e-06,
+      "loss": 0.535,
+      "step": 4607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2172656787772724,
+      "learning_rate": 2.943053627440771e-06,
+      "loss": 0.4892,
+      "step": 4608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8232507897401313,
+      "learning_rate": 2.93928516147177e-06,
+      "loss": 0.5792,
+      "step": 4609
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7998870738402721,
+      "learning_rate": 2.9355186940232493e-06,
+      "loss": 0.627,
+      "step": 4610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6371535718712893,
+      "learning_rate": 2.9317542261612986e-06,
+      "loss": 0.5783,
+      "step": 4611
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2985991623713358,
+      "learning_rate": 2.927991758951445e-06,
+      "loss": 0.5974,
+      "step": 4612
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8324163542872334,
+      "learning_rate": 2.924231293458647e-06,
+      "loss": 0.5586,
+      "step": 4613
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.72497481587534,
+      "learning_rate": 2.920472830747295e-06,
+      "loss": 0.5068,
+      "step": 4614
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7274882737205781,
+      "learning_rate": 2.9167163718812143e-06,
+      "loss": 0.5853,
+      "step": 4615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6161417813284784,
+      "learning_rate": 2.9129619179236625e-06,
+      "loss": 0.4773,
+      "step": 4616
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.521280261666512,
+      "learning_rate": 2.9092094699373296e-06,
+      "loss": 0.4875,
+      "step": 4617
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.849479353328145,
+      "learning_rate": 2.90545902898434e-06,
+      "loss": 0.4612,
+      "step": 4618
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8133782209563711,
+      "learning_rate": 2.9017105961262448e-06,
+      "loss": 0.5325,
+      "step": 4619
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.832310775897283,
+      "learning_rate": 2.8979641724240324e-06,
+      "loss": 0.5286,
+      "step": 4620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.683799687161014,
+      "learning_rate": 2.8942197589381204e-06,
+      "loss": 0.5696,
+      "step": 4621
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5439195652119595,
+      "learning_rate": 2.890477356728356e-06,
+      "loss": 0.5066,
+      "step": 4622
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.967247981417357,
+      "learning_rate": 2.886736966854019e-06,
+      "loss": 0.6019,
+      "step": 4623
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6386294919351683,
+      "learning_rate": 2.8829985903738176e-06,
+      "loss": 0.5589,
+      "step": 4624
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.146562210119848,
+      "learning_rate": 2.8792622283458926e-06,
+      "loss": 0.5402,
+      "step": 4625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7065945138528738,
+      "learning_rate": 2.8755278818278143e-06,
+      "loss": 0.5714,
+      "step": 4626
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6337231816425475,
+      "learning_rate": 2.8717955518765794e-06,
+      "loss": 0.5059,
+      "step": 4627
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8087998392796891,
+      "learning_rate": 2.8680652395486198e-06,
+      "loss": 0.626,
+      "step": 4628
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0921689894041755,
+      "learning_rate": 2.864336945899788e-06,
+      "loss": 0.628,
+      "step": 4629
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1953164058757686,
+      "learning_rate": 2.860610671985371e-06,
+      "loss": 0.5764,
+      "step": 4630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.326357673395692,
+      "learning_rate": 2.856886418860083e-06,
+      "loss": 0.4405,
+      "step": 4631
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9179228494109355,
+      "learning_rate": 2.853164187578067e-06,
+      "loss": 0.5109,
+      "step": 4632
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7444058992647986,
+      "learning_rate": 2.849443979192892e-06,
+      "loss": 0.4427,
+      "step": 4633
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5423457413998338,
+      "learning_rate": 2.845725794757551e-06,
+      "loss": 0.497,
+      "step": 4634
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.169337842027456,
+      "learning_rate": 2.8420096353244763e-06,
+      "loss": 0.4455,
+      "step": 4635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3058392078152417,
+      "learning_rate": 2.838295501945516e-06,
+      "loss": 0.496,
+      "step": 4636
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7273297123819353,
+      "learning_rate": 2.834583395671947e-06,
+      "loss": 0.5522,
+      "step": 4637
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3211584249752666,
+      "learning_rate": 2.8308733175544724e-06,
+      "loss": 0.5296,
+      "step": 4638
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.60639546036655,
+      "learning_rate": 2.827165268643223e-06,
+      "loss": 0.5267,
+      "step": 4639
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0288258395586314,
+      "learning_rate": 2.8234592499877535e-06,
+      "loss": 0.5744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5695521321203947,
+      "learning_rate": 2.819755262637046e-06,
+      "loss": 0.4301,
+      "step": 4641
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4441625068651764,
+      "learning_rate": 2.8160533076395045e-06,
+      "loss": 0.5211,
+      "step": 4642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7157699194614542,
+      "learning_rate": 2.812353386042962e-06,
+      "loss": 0.4894,
+      "step": 4643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1636672136456356,
+      "learning_rate": 2.8086554988946714e-06,
+      "loss": 0.527,
+      "step": 4644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7839764501758157,
+      "learning_rate": 2.804959647241312e-06,
+      "loss": 0.6494,
+      "step": 4645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.73838303933579,
+      "learning_rate": 2.8012658321289878e-06,
+      "loss": 0.458,
+      "step": 4646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.563640534731604,
+      "learning_rate": 2.797574054603225e-06,
+      "loss": 0.4776,
+      "step": 4647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6976482993601092,
+      "learning_rate": 2.7938843157089734e-06,
+      "loss": 0.5431,
+      "step": 4648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8170651631141234,
+      "learning_rate": 2.790196616490607e-06,
+      "loss": 0.6582,
+      "step": 4649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.770366470519543,
+      "learning_rate": 2.7865109579919223e-06,
+      "loss": 0.5681,
+      "step": 4650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8910240709045962,
+      "learning_rate": 2.7828273412561324e-06,
+      "loss": 0.5144,
+      "step": 4651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.980844313176368,
+      "learning_rate": 2.7791457673258793e-06,
+      "loss": 0.5625,
+      "step": 4652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6792657849535209,
+      "learning_rate": 2.775466237243226e-06,
+      "loss": 0.4679,
+      "step": 4653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6218773560663875,
+      "learning_rate": 2.7717887520496545e-06,
+      "loss": 0.519,
+      "step": 4654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3114124354163215,
+      "learning_rate": 2.7681133127860705e-06,
+      "loss": 0.4302,
+      "step": 4655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.661828768005959,
+      "learning_rate": 2.7644399204927984e-06,
+      "loss": 0.5321,
+      "step": 4656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6731903117123619,
+      "learning_rate": 2.7607685762095825e-06,
+      "loss": 0.4861,
+      "step": 4657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.146598430494376,
+      "learning_rate": 2.7570992809755937e-06,
+      "loss": 0.4461,
+      "step": 4658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6354943286274897,
+      "learning_rate": 2.753432035829415e-06,
+      "loss": 0.4606,
+      "step": 4659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.3013077004316775,
+      "learning_rate": 2.749766841809054e-06,
+      "loss": 0.4891,
+      "step": 4660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8237650448499476,
+      "learning_rate": 2.746103699951934e-06,
+      "loss": 0.4701,
+      "step": 4661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6622659624200269,
+      "learning_rate": 2.742442611294902e-06,
+      "loss": 0.5109,
+      "step": 4662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.911612250852303,
+      "learning_rate": 2.73878357687422e-06,
+      "loss": 0.5187,
+      "step": 4663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.855021722776991,
+      "learning_rate": 2.7351265977255702e-06,
+      "loss": 0.6609,
+      "step": 4664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.451079656888799,
+      "learning_rate": 2.731471674884053e-06,
+      "loss": 0.5297,
+      "step": 4665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7562710255839773,
+      "learning_rate": 2.7278188093841874e-06,
+      "loss": 0.5234,
+      "step": 4666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.870582068698593,
+      "learning_rate": 2.7241680022599073e-06,
+      "loss": 0.4675,
+      "step": 4667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6994701197937931,
+      "learning_rate": 2.720519254544568e-06,
+      "loss": 0.4597,
+      "step": 4668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8176300683709936,
+      "learning_rate": 2.716872567270938e-06,
+      "loss": 0.4946,
+      "step": 4669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8771041442264367,
+      "learning_rate": 2.713227941471206e-06,
+      "loss": 0.499,
+      "step": 4670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.556243766630316,
+      "learning_rate": 2.7095853781769752e-06,
+      "loss": 0.4565,
+      "step": 4671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7049016219962394,
+      "learning_rate": 2.7059448784192688e-06,
+      "loss": 0.4584,
+      "step": 4672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8047321915399728,
+      "learning_rate": 2.702306443228516e-06,
+      "loss": 0.505,
+      "step": 4673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8798072702994844,
+      "learning_rate": 2.6986700736345715e-06,
+      "loss": 0.4587,
+      "step": 4674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.857974442082426,
+      "learning_rate": 2.6950357706667017e-06,
+      "loss": 0.5391,
+      "step": 4675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9447048756641956,
+      "learning_rate": 2.6914035353535897e-06,
+      "loss": 0.5276,
+      "step": 4676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5098363319067047,
+      "learning_rate": 2.68777336872333e-06,
+      "loss": 0.5119,
+      "step": 4677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6481767326155163,
+      "learning_rate": 2.6841452718034343e-06,
+      "loss": 0.5253,
+      "step": 4678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6444103892495576,
+      "learning_rate": 2.6805192456208297e-06,
+      "loss": 0.5875,
+      "step": 4679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6587604547682173,
+      "learning_rate": 2.6768952912018498e-06,
+      "loss": 0.5845,
+      "step": 4680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0829614226018425,
+      "learning_rate": 2.6732734095722545e-06,
+      "loss": 0.4574,
+      "step": 4681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6600183289201758,
+      "learning_rate": 2.6696536017572074e-06,
+      "loss": 0.492,
+      "step": 4682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3656654666729007,
+      "learning_rate": 2.666035868781285e-06,
+      "loss": 0.5799,
+      "step": 4683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7432134412141325,
+      "learning_rate": 2.6624202116684816e-06,
+      "loss": 0.5825,
+      "step": 4684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.074952921007968,
+      "learning_rate": 2.6588066314422e-06,
+      "loss": 0.4889,
+      "step": 4685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6170617861076273,
+      "learning_rate": 2.6551951291252576e-06,
+      "loss": 0.4501,
+      "step": 4686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8527004469117931,
+      "learning_rate": 2.651585705739881e-06,
+      "loss": 0.5179,
+      "step": 4687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2461528390832197,
+      "learning_rate": 2.6479783623077105e-06,
+      "loss": 0.5469,
+      "step": 4688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5551883758094218,
+      "learning_rate": 2.6443730998497985e-06,
+      "loss": 0.5716,
+      "step": 4689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7440166235785302,
+      "learning_rate": 2.6407699193866045e-06,
+      "loss": 0.4891,
+      "step": 4690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8247310351671647,
+      "learning_rate": 2.6371688219380032e-06,
+      "loss": 0.6513,
+      "step": 4691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0510593029960753,
+      "learning_rate": 2.6335698085232764e-06,
+      "loss": 0.5468,
+      "step": 4692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6568956837741822,
+      "learning_rate": 2.62997288016112e-06,
+      "loss": 0.4766,
+      "step": 4693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7269459429525194,
+      "learning_rate": 2.6263780378696324e-06,
+      "loss": 0.5236,
+      "step": 4694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9980366788692485,
+      "learning_rate": 2.6227852826663294e-06,
+      "loss": 0.5498,
+      "step": 4695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6576990406047916,
+      "learning_rate": 2.6191946155681303e-06,
+      "loss": 0.4761,
+      "step": 4696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9038184312555195,
+      "learning_rate": 2.6156060375913685e-06,
+      "loss": 0.4934,
+      "step": 4697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.049955509895633,
+      "learning_rate": 2.6120195497517818e-06,
+      "loss": 0.5593,
+      "step": 4698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0668363412142248,
+      "learning_rate": 2.608435153064519e-06,
+      "loss": 0.4909,
+      "step": 4699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1641833883356494,
+      "learning_rate": 2.6048528485441347e-06,
+      "loss": 0.4423,
+      "step": 4700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0852207260907716,
+      "learning_rate": 2.601272637204595e-06,
+      "loss": 0.5568,
+      "step": 4701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7275523843662814,
+      "learning_rate": 2.5976945200592683e-06,
+      "loss": 0.527,
+      "step": 4702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6981566435609008,
+      "learning_rate": 2.5941184981209354e-06,
+      "loss": 0.5778,
+      "step": 4703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6409673891161243,
+      "learning_rate": 2.5905445724017786e-06,
+      "loss": 0.6031,
+      "step": 4704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.811475181436531,
+      "learning_rate": 2.586972743913394e-06,
+      "loss": 0.5666,
+      "step": 4705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8116472621173674,
+      "learning_rate": 2.5834030136667796e-06,
+      "loss": 0.69,
+      "step": 4706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.715794519171324,
+      "learning_rate": 2.579835382672339e-06,
+      "loss": 0.5364,
+      "step": 4707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.677334624125083,
+      "learning_rate": 2.5762698519398832e-06,
+      "loss": 0.5113,
+      "step": 4708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0193553854892854,
+      "learning_rate": 2.5727064224786267e-06,
+      "loss": 0.5347,
+      "step": 4709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6673715586205116,
+      "learning_rate": 2.569145095297192e-06,
+      "loss": 0.5077,
+      "step": 4710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.16205387703063,
+      "learning_rate": 2.5655858714036054e-06,
+      "loss": 0.5269,
+      "step": 4711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7510742158918697,
+      "learning_rate": 2.5620287518052967e-06,
+      "loss": 0.5037,
+      "step": 4712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.117478780364901,
+      "learning_rate": 2.5584737375091016e-06,
+      "loss": 0.505,
+      "step": 4713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7757902921422428,
+      "learning_rate": 2.554920829521259e-06,
+      "loss": 0.4382,
+      "step": 4714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6808565044247754,
+      "learning_rate": 2.551370028847416e-06,
+      "loss": 0.51,
+      "step": 4715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7819131604997762,
+      "learning_rate": 2.547821336492614e-06,
+      "loss": 0.556,
+      "step": 4716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6860457399601343,
+      "learning_rate": 2.544274753461303e-06,
+      "loss": 0.5235,
+      "step": 4717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5317012202243596,
+      "learning_rate": 2.5407302807573387e-06,
+      "loss": 0.5398,
+      "step": 4718
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7569047641592161,
+      "learning_rate": 2.5371879193839756e-06,
+      "loss": 0.5108,
+      "step": 4719
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.016503635745383,
+      "learning_rate": 2.5336476703438705e-06,
+      "loss": 0.5373,
+      "step": 4720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8937809039370306,
+      "learning_rate": 2.530109534639085e-06,
+      "loss": 0.5034,
+      "step": 4721
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9984114906588022,
+      "learning_rate": 2.5265735132710802e-06,
+      "loss": 0.5001,
+      "step": 4722
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.045756515233843,
+      "learning_rate": 2.5230396072407204e-06,
+      "loss": 0.4691,
+      "step": 4723
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9489391451486022,
+      "learning_rate": 2.519507817548269e-06,
+      "loss": 0.4757,
+      "step": 4724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6823116302646663,
+      "learning_rate": 2.515978145193393e-06,
+      "loss": 0.5003,
+      "step": 4725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8184589752107418,
+      "learning_rate": 2.512450591175157e-06,
+      "loss": 0.451,
+      "step": 4726
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8572646733810427,
+      "learning_rate": 2.508925156492027e-06,
+      "loss": 0.4792,
+      "step": 4727
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4543357052081243,
+      "learning_rate": 2.5054018421418737e-06,
+      "loss": 0.5532,
+      "step": 4728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9764546930797129,
+      "learning_rate": 2.5018806491219627e-06,
+      "loss": 0.4905,
+      "step": 4729
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.032926888813982,
+      "learning_rate": 2.4983615784289585e-06,
+      "loss": 0.523,
+      "step": 4730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.778983947648959,
+      "learning_rate": 2.494844631058927e-06,
+      "loss": 0.5267,
+      "step": 4731
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.100484421213782,
+      "learning_rate": 2.4913298080073344e-06,
+      "loss": 0.4958,
+      "step": 4732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.879425786564652,
+      "learning_rate": 2.487817110269042e-06,
+      "loss": 0.5175,
+      "step": 4733
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.061559090497519,
+      "learning_rate": 2.4843065388383126e-06,
+      "loss": 0.5368,
+      "step": 4734
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7818517627894516,
+      "learning_rate": 2.480798094708805e-06,
+      "loss": 0.5955,
+      "step": 4735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1339457634266754,
+      "learning_rate": 2.4772917788735786e-06,
+      "loss": 0.562,
+      "step": 4736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7214710929390697,
+      "learning_rate": 2.473787592325091e-06,
+      "loss": 0.513,
+      "step": 4737
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7944819197721604,
+      "learning_rate": 2.470285536055188e-06,
+      "loss": 0.5768,
+      "step": 4738
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0618441885824232,
+      "learning_rate": 2.4667856110551235e-06,
+      "loss": 0.5816,
+      "step": 4739
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.078150734788772,
+      "learning_rate": 2.463287818315543e-06,
+      "loss": 0.5084,
+      "step": 4740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0920760982286004,
+      "learning_rate": 2.4597921588264893e-06,
+      "loss": 0.5087,
+      "step": 4741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5929787736876955,
+      "learning_rate": 2.456298633577402e-06,
+      "loss": 0.5334,
+      "step": 4742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.30884903632308,
+      "learning_rate": 2.4528072435571158e-06,
+      "loss": 0.4696,
+      "step": 4743
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6054509390566747,
+      "learning_rate": 2.449317989753862e-06,
+      "loss": 0.5097,
+      "step": 4744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6579965385119517,
+      "learning_rate": 2.445830873155266e-06,
+      "loss": 0.5047,
+      "step": 4745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9569764021876217,
+      "learning_rate": 2.4423458947483482e-06,
+      "loss": 0.5484,
+      "step": 4746
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6907424016314427,
+      "learning_rate": 2.4388630555195247e-06,
+      "loss": 0.4593,
+      "step": 4747
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2364909722158046,
+      "learning_rate": 2.4353823564546064e-06,
+      "loss": 0.5213,
+      "step": 4748
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5422673725377358,
+      "learning_rate": 2.4319037985387985e-06,
+      "loss": 0.4797,
+      "step": 4749
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7345524682533575,
+      "learning_rate": 2.428427382756695e-06,
+      "loss": 0.4425,
+      "step": 4750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.021768993448649,
+      "learning_rate": 2.424953110092294e-06,
+      "loss": 0.6012,
+      "step": 4751
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8890885681799636,
+      "learning_rate": 2.4214809815289797e-06,
+      "loss": 0.5722,
+      "step": 4752
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5313444754091527,
+      "learning_rate": 2.4180109980495293e-06,
+      "loss": 0.4451,
+      "step": 4753
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7780948767429403,
+      "learning_rate": 2.4145431606361148e-06,
+      "loss": 0.6097,
+      "step": 4754
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.590339897604809,
+      "learning_rate": 2.4110774702703e-06,
+      "loss": 0.517,
+      "step": 4755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5327697218073597,
+      "learning_rate": 2.4076139279330414e-06,
+      "loss": 0.5755,
+      "step": 4756
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7766061525971221,
+      "learning_rate": 2.4041525346046877e-06,
+      "loss": 0.6414,
+      "step": 4757
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7163866087259383,
+      "learning_rate": 2.4006932912649816e-06,
+      "loss": 0.4436,
+      "step": 4758
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7208298791640084,
+      "learning_rate": 2.3972361988930505e-06,
+      "loss": 0.604,
+      "step": 4759
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7367760013062783,
+      "learning_rate": 2.3937812584674168e-06,
+      "loss": 0.5608,
+      "step": 4760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0483420125061063,
+      "learning_rate": 2.3903284709659957e-06,
+      "loss": 0.5732,
+      "step": 4761
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8853049313438561,
+      "learning_rate": 2.3868778373660927e-06,
+      "loss": 0.4798,
+      "step": 4762
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6269333585687593,
+      "learning_rate": 2.3834293586444e-06,
+      "loss": 0.4741,
+      "step": 4763
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8748578697295496,
+      "learning_rate": 2.379983035777005e-06,
+      "loss": 0.6844,
+      "step": 4764
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.080409316071463,
+      "learning_rate": 2.37653886973938e-06,
+      "loss": 0.5194,
+      "step": 4765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.120214379742731,
+      "learning_rate": 2.3730968615063886e-06,
+      "loss": 0.569,
+      "step": 4766
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1227249194303175,
+      "learning_rate": 2.3696570120522868e-06,
+      "loss": 0.5183,
+      "step": 4767
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.515123744724309,
+      "learning_rate": 2.3662193223507135e-06,
+      "loss": 0.398,
+      "step": 4768
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7814540771708973,
+      "learning_rate": 2.362783793374701e-06,
+      "loss": 0.4899,
+      "step": 4769
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4141720731847514,
+      "learning_rate": 2.3593504260966695e-06,
+      "loss": 0.4637,
+      "step": 4770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7559460231451955,
+      "learning_rate": 2.355919221488424e-06,
+      "loss": 0.5316,
+      "step": 4771
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.079846505409417,
+      "learning_rate": 2.352490180521162e-06,
+      "loss": 0.5327,
+      "step": 4772
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6041693827074706,
+      "learning_rate": 2.349063304165462e-06,
+      "loss": 0.5657,
+      "step": 4773
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.055109673895342,
+      "learning_rate": 2.345638593391302e-06,
+      "loss": 0.5567,
+      "step": 4774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.229994623911119,
+      "learning_rate": 2.3422160491680334e-06,
+      "loss": 0.5216,
+      "step": 4775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8262741880172517,
+      "learning_rate": 2.3387956724644014e-06,
+      "loss": 0.6661,
+      "step": 4776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.794426264052474,
+      "learning_rate": 2.3353774642485374e-06,
+      "loss": 0.5448,
+      "step": 4777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7371486571972075,
+      "learning_rate": 2.331961425487956e-06,
+      "loss": 0.6161,
+      "step": 4778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.243796069425476,
+      "learning_rate": 2.3285475571495617e-06,
+      "loss": 0.4235,
+      "step": 4779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7836813937375826,
+      "learning_rate": 2.3251358601996453e-06,
+      "loss": 0.5471,
+      "step": 4780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7998647548915219,
+      "learning_rate": 2.3217263356038744e-06,
+      "loss": 0.5507,
+      "step": 4781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6223650480497804,
+      "learning_rate": 2.31831898432731e-06,
+      "loss": 0.5769,
+      "step": 4782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6989209856940244,
+      "learning_rate": 2.3149138073343958e-06,
+      "loss": 0.5191,
+      "step": 4783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6269691341814156,
+      "learning_rate": 2.3115108055889614e-06,
+      "loss": 0.5175,
+      "step": 4784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0489077128912716,
+      "learning_rate": 2.3081099800542183e-06,
+      "loss": 0.5254,
+      "step": 4785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5989383583491772,
+      "learning_rate": 2.3047113316927627e-06,
+      "loss": 0.432,
+      "step": 4786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.485802066724422,
+      "learning_rate": 2.301314861466575e-06,
+      "loss": 0.4429,
+      "step": 4787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9186238890906029,
+      "learning_rate": 2.297920570337019e-06,
+      "loss": 0.4693,
+      "step": 4788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7512704034003566,
+      "learning_rate": 2.294528459264842e-06,
+      "loss": 0.5297,
+      "step": 4789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2475718619184937,
+      "learning_rate": 2.291138529210174e-06,
+      "loss": 0.5893,
+      "step": 4790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0357209947559927,
+      "learning_rate": 2.287750781132527e-06,
+      "loss": 0.5007,
+      "step": 4791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6631047097214255,
+      "learning_rate": 2.284365215990797e-06,
+      "loss": 0.5577,
+      "step": 4792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.588290048986684,
+      "learning_rate": 2.2809818347432598e-06,
+      "loss": 0.5961,
+      "step": 4793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.888912664310801,
+      "learning_rate": 2.2776006383475745e-06,
+      "loss": 0.5331,
+      "step": 4794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7715813778545242,
+      "learning_rate": 2.274221627760782e-06,
+      "loss": 0.5327,
+      "step": 4795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5819739222329465,
+      "learning_rate": 2.270844803939305e-06,
+      "loss": 0.4904,
+      "step": 4796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7205988455342525,
+      "learning_rate": 2.2674701678389423e-06,
+      "loss": 0.4775,
+      "step": 4797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8307163315473899,
+      "learning_rate": 2.2640977204148838e-06,
+      "loss": 0.6117,
+      "step": 4798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7459148143295562,
+      "learning_rate": 2.26072746262169e-06,
+      "loss": 0.4824,
+      "step": 4799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.128556713004545,
+      "learning_rate": 2.2573593954133067e-06,
+      "loss": 0.5903,
+      "step": 4800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7611195906043757,
+      "learning_rate": 2.2539935197430574e-06,
+      "loss": 0.4684,
+      "step": 4801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8343214314849834,
+      "learning_rate": 2.2506298365636482e-06,
+      "loss": 0.565,
+      "step": 4802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.878203328139911,
+      "learning_rate": 2.2472683468271584e-06,
+      "loss": 0.5457,
+      "step": 4803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8183824543456759,
+      "learning_rate": 2.2439090514850527e-06,
+      "loss": 0.6499,
+      "step": 4804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5084249353876047,
+      "learning_rate": 2.2405519514881723e-06,
+      "loss": 0.4981,
+      "step": 4805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9015361384515457,
+      "learning_rate": 2.2371970477867377e-06,
+      "loss": 0.4637,
+      "step": 4806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.197214541988391,
+      "learning_rate": 2.2338443413303466e-06,
+      "loss": 0.494,
+      "step": 4807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0189612376830564,
+      "learning_rate": 2.230493833067977e-06,
+      "loss": 0.5397,
+      "step": 4808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9502551675956263,
+      "learning_rate": 2.2271455239479822e-06,
+      "loss": 0.5108,
+      "step": 4809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6595214313898872,
+      "learning_rate": 2.2237994149180943e-06,
+      "loss": 0.5212,
+      "step": 4810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.461737274545426,
+      "learning_rate": 2.220455506925422e-06,
+      "loss": 0.5982,
+      "step": 4811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.304646551452546,
+      "learning_rate": 2.2171138009164515e-06,
+      "loss": 0.4797,
+      "step": 4812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.738718935770548,
+      "learning_rate": 2.213774297837047e-06,
+      "loss": 0.5306,
+      "step": 4813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.316785621669269,
+      "learning_rate": 2.210436998632446e-06,
+      "loss": 0.5494,
+      "step": 4814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7766664895926378,
+      "learning_rate": 2.2071019042472643e-06,
+      "loss": 0.5795,
+      "step": 4815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.372696804714982,
+      "learning_rate": 2.2037690156254944e-06,
+      "loss": 0.5228,
+      "step": 4816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6688439004991134,
+      "learning_rate": 2.2004383337105016e-06,
+      "loss": 0.5069,
+      "step": 4817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7765948283976245,
+      "learning_rate": 2.1971098594450315e-06,
+      "loss": 0.517,
+      "step": 4818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8405926861462009,
+      "learning_rate": 2.1937835937711995e-06,
+      "loss": 0.5983,
+      "step": 4819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7638746979451692,
+      "learning_rate": 2.190459537630495e-06,
+      "loss": 0.6155,
+      "step": 4820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1598377830421702,
+      "learning_rate": 2.187137691963791e-06,
+      "loss": 0.5458,
+      "step": 4821
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.03681953246126,
+      "learning_rate": 2.1838180577113268e-06,
+      "loss": 0.5773,
+      "step": 4822
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.826909756997338,
+      "learning_rate": 2.1805006358127213e-06,
+      "loss": 0.4663,
+      "step": 4823
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9968245595782734,
+      "learning_rate": 2.177185427206956e-06,
+      "loss": 0.4687,
+      "step": 4824
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8448802165939293,
+      "learning_rate": 2.173872432832398e-06,
+      "loss": 0.4914,
+      "step": 4825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7321882052866953,
+      "learning_rate": 2.1705616536267838e-06,
+      "loss": 0.4839,
+      "step": 4826
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8140822365033764,
+      "learning_rate": 2.1672530905272215e-06,
+      "loss": 0.5347,
+      "step": 4827
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6067901325443419,
+      "learning_rate": 2.1639467444701934e-06,
+      "loss": 0.4621,
+      "step": 4828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.486798025692315,
+      "learning_rate": 2.160642616391553e-06,
+      "loss": 0.5373,
+      "step": 4829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7669000599493152,
+      "learning_rate": 2.1573407072265284e-06,
+      "loss": 0.555,
+      "step": 4830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6996998876340257,
+      "learning_rate": 2.1540410179097173e-06,
+      "loss": 0.5272,
+      "step": 4831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8183958944320424,
+      "learning_rate": 2.1507435493750885e-06,
+      "loss": 0.5261,
+      "step": 4832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6651646306786458,
+      "learning_rate": 2.1474483025559857e-06,
+      "loss": 0.4903,
+      "step": 4833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.855896123242796,
+      "learning_rate": 2.1441552783851195e-06,
+      "loss": 0.4676,
+      "step": 4834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9765699151387157,
+      "learning_rate": 2.1408644777945753e-06,
+      "loss": 0.4793,
+      "step": 4835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0362225062346124,
+      "learning_rate": 2.137575901715806e-06,
+      "loss": 0.5283,
+      "step": 4836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4767715143872486,
+      "learning_rate": 2.1342895510796367e-06,
+      "loss": 0.5489,
+      "step": 4837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7729250051960117,
+      "learning_rate": 2.1310054268162628e-06,
+      "loss": 0.5712,
+      "step": 4838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.920286902674208,
+      "learning_rate": 2.127723529855248e-06,
+      "loss": 0.4903,
+      "step": 4839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7800992804220916,
+      "learning_rate": 2.124443861125525e-06,
+      "loss": 0.5117,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.169081776081325,
+      "learning_rate": 2.1211664215553997e-06,
+      "loss": 0.5375,
+      "step": 4841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.728175472329314,
+      "learning_rate": 2.1178912120725416e-06,
+      "loss": 0.493,
+      "step": 4842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.746956216804649,
+      "learning_rate": 2.114618233603992e-06,
+      "loss": 0.5131,
+      "step": 4843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6870853791899327,
+      "learning_rate": 2.111347487076164e-06,
+      "loss": 0.4677,
+      "step": 4844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.906478894221574,
+      "learning_rate": 2.1080789734148366e-06,
+      "loss": 0.4581,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6965458372578373,
+      "learning_rate": 2.1048126935451495e-06,
+      "loss": 0.5619,
+      "step": 4846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.028174656118157,
+      "learning_rate": 2.10154864839162e-06,
+      "loss": 0.5647,
+      "step": 4847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9800218513666492,
+      "learning_rate": 2.0982868388781286e-06,
+      "loss": 0.5401,
+      "step": 4848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6933531827870836,
+      "learning_rate": 2.0950272659279246e-06,
+      "loss": 0.5248,
+      "step": 4849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4890586317200325,
+      "learning_rate": 2.091769930463621e-06,
+      "loss": 0.5827,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9981121240950221,
+      "learning_rate": 2.0885148334072013e-06,
+      "loss": 0.5511,
+      "step": 4851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6567467514208933,
+      "learning_rate": 2.085261975680014e-06,
+      "loss": 0.5161,
+      "step": 4852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.149920583624905,
+      "learning_rate": 2.0820113582027734e-06,
+      "loss": 0.6108,
+      "step": 4853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1800108859605896,
+      "learning_rate": 2.07876298189556e-06,
+      "loss": 0.5263,
+      "step": 4854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.592867544569354,
+      "learning_rate": 2.07551684767782e-06,
+      "loss": 0.5808,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7972197758579647,
+      "learning_rate": 2.072272956468364e-06,
+      "loss": 0.5187,
+      "step": 4856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.1994237619829775,
+      "learning_rate": 2.0690313091853697e-06,
+      "loss": 0.571,
+      "step": 4857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.100326895108673,
+      "learning_rate": 2.0657919067463773e-06,
+      "loss": 0.5357,
+      "step": 4858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.196806837967913,
+      "learning_rate": 2.062554750068294e-06,
+      "loss": 0.5801,
+      "step": 4859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8600714698374348,
+      "learning_rate": 2.05931984006739e-06,
+      "loss": 0.5929,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4665952818135761,
+      "learning_rate": 2.0560871776592996e-06,
+      "loss": 0.5171,
+      "step": 4861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.21394404642647,
+      "learning_rate": 2.0528567637590214e-06,
+      "loss": 0.5169,
+      "step": 4862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.925881603100552,
+      "learning_rate": 2.0496285992809163e-06,
+      "loss": 0.6428,
+      "step": 4863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9215817335390961,
+      "learning_rate": 2.0464026851387096e-06,
+      "loss": 0.5262,
+      "step": 4864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7880203865490594,
+      "learning_rate": 2.0431790222454906e-06,
+      "loss": 0.5405,
+      "step": 4865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.802959855178915,
+      "learning_rate": 2.03995761151371e-06,
+      "loss": 0.5067,
+      "step": 4866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0121396459574528,
+      "learning_rate": 2.0367384538551805e-06,
+      "loss": 0.5382,
+      "step": 4867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3072560669591353,
+      "learning_rate": 2.033521550181078e-06,
+      "loss": 0.4999,
+      "step": 4868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7547460591922628,
+      "learning_rate": 2.0303069014019415e-06,
+      "loss": 0.5485,
+      "step": 4869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9212921249197723,
+      "learning_rate": 2.0270945084276695e-06,
+      "loss": 0.4768,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7363925999585534,
+      "learning_rate": 2.0238843721675226e-06,
+      "loss": 0.4851,
+      "step": 4871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2192644385167815,
+      "learning_rate": 2.020676493530126e-06,
+      "loss": 0.5505,
+      "step": 4872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6241815011931031,
+      "learning_rate": 2.0174708734234596e-06,
+      "loss": 0.5425,
+      "step": 4873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1834801932382373,
+      "learning_rate": 2.0142675127548684e-06,
+      "loss": 0.5047,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.241527392659044,
+      "learning_rate": 2.0110664124310574e-06,
+      "loss": 0.4584,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6281399210994998,
+      "learning_rate": 2.007867573358091e-06,
+      "loss": 0.5905,
+      "step": 4876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.737574714236872,
+      "learning_rate": 2.0046709964413947e-06,
+      "loss": 0.4363,
+      "step": 4877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7901079112544352,
+      "learning_rate": 2.0014766825857514e-06,
+      "loss": 0.6004,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8103820041732432,
+      "learning_rate": 1.9982846326953066e-06,
+      "loss": 0.5299,
+      "step": 4879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3204279651100803,
+      "learning_rate": 1.995094847673561e-06,
+      "loss": 0.5139,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8465486811482705,
+      "learning_rate": 1.991907328423379e-06,
+      "loss": 0.4617,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6872180038962048,
+      "learning_rate": 1.9887220758469794e-06,
+      "loss": 0.4757,
+      "step": 4882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7722244901348632,
+      "learning_rate": 1.985539090845943e-06,
+      "loss": 0.4853,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.17760429148881,
+      "learning_rate": 1.982358374321205e-06,
+      "loss": 0.5756,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3051050296552167,
+      "learning_rate": 1.9791799271730626e-06,
+      "loss": 0.4996,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8426605568219305,
+      "learning_rate": 1.9760037503011664e-06,
+      "loss": 0.5366,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9381453205068448,
+      "learning_rate": 1.972829844604528e-06,
+      "loss": 0.4541,
+      "step": 4887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.911405297062591,
+      "learning_rate": 1.9696582109815145e-06,
+      "loss": 0.4848,
+      "step": 4888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0693471230056715,
+      "learning_rate": 1.966488850329851e-06,
+      "loss": 0.5156,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9903693664047324,
+      "learning_rate": 1.9633217635466164e-06,
+      "loss": 0.5371,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8057172698902741,
+      "learning_rate": 1.960156951528248e-06,
+      "loss": 0.5411,
+      "step": 4891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9065206300417215,
+      "learning_rate": 1.9569944151705423e-06,
+      "loss": 0.4676,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.929069947737278,
+      "learning_rate": 1.9538341553686446e-06,
+      "loss": 0.5584,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8489158351455703,
+      "learning_rate": 1.950676173017062e-06,
+      "loss": 0.4829,
+      "step": 4894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0721575757797424,
+      "learning_rate": 1.947520469009655e-06,
+      "loss": 0.4756,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7986477189600867,
+      "learning_rate": 1.9443670442396378e-06,
+      "loss": 0.4998,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0753155403713315,
+      "learning_rate": 1.941215899599581e-06,
+      "loss": 0.4717,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.426817800075875,
+      "learning_rate": 1.93806703598141e-06,
+      "loss": 0.4881,
+      "step": 4898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6228770442557878,
+      "learning_rate": 1.9349204542764044e-06,
+      "loss": 0.5273,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.247153052748518,
+      "learning_rate": 1.9317761553751957e-06,
+      "loss": 0.5368,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.776733890748408,
+      "learning_rate": 1.928634140167772e-06,
+      "loss": 0.522,
+      "step": 4901
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0000316890729897,
+      "learning_rate": 1.9254944095434745e-06,
+      "loss": 0.5276,
+      "step": 4902
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7973278977942502,
+      "learning_rate": 1.9223569643909978e-06,
+      "loss": 0.4736,
+      "step": 4903
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.462598045735337,
+      "learning_rate": 1.919221805598388e-06,
+      "loss": 0.5064,
+      "step": 4904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5256433384253747,
+      "learning_rate": 1.9160889340530455e-06,
+      "loss": 0.5328,
+      "step": 4905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6533790118125644,
+      "learning_rate": 1.9129583506417236e-06,
+      "loss": 0.5311,
+      "step": 4906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.266168335853361,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.5746,
+      "step": 4907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6104359565125876,
+      "learning_rate": 1.9067040517649115e-06,
+      "loss": 0.5401,
+      "step": 4908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.880764308691986,
+      "learning_rate": 1.9035803380696883e-06,
+      "loss": 0.4984,
+      "step": 4909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8128362681790577,
+      "learning_rate": 1.9004589160490173e-06,
+      "loss": 0.4395,
+      "step": 4910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.614041257992648,
+      "learning_rate": 1.8973397865864095e-06,
+      "loss": 0.5615,
+      "step": 4911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.560993053229691,
+      "learning_rate": 1.8942229505647292e-06,
+      "loss": 0.4662,
+      "step": 4912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5904711961855644,
+      "learning_rate": 1.8911084088661903e-06,
+      "loss": 0.4576,
+      "step": 4913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2212270068485656,
+      "learning_rate": 1.8879961623723553e-06,
+      "loss": 0.5012,
+      "step": 4914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4921693115644192,
+      "learning_rate": 1.884886211964141e-06,
+      "loss": 0.4934,
+      "step": 4915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8901106917988315,
+      "learning_rate": 1.8817785585218118e-06,
+      "loss": 0.4825,
+      "step": 4916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8311958219405533,
+      "learning_rate": 1.878673202924982e-06,
+      "loss": 0.6574,
+      "step": 4917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5668842652669688,
+      "learning_rate": 1.8755701460526166e-06,
+      "loss": 0.5414,
+      "step": 4918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5850864411235206,
+      "learning_rate": 1.8724693887830292e-06,
+      "loss": 0.5628,
+      "step": 4919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.179003468783936,
+      "learning_rate": 1.8693709319938824e-06,
+      "loss": 0.4588,
+      "step": 4920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5414519846905903,
+      "learning_rate": 1.866274776562188e-06,
+      "loss": 0.542,
+      "step": 4921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.031519631914586,
+      "learning_rate": 1.863180923364306e-06,
+      "loss": 0.5671,
+      "step": 4922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.742738760889124,
+      "learning_rate": 1.860089373275945e-06,
+      "loss": 0.4845,
+      "step": 4923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.640404668555194,
+      "learning_rate": 1.8570001271721627e-06,
+      "loss": 0.4983,
+      "step": 4924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9586757633187037,
+      "learning_rate": 1.8539131859273628e-06,
+      "loss": 0.496,
+      "step": 4925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9220286412741159,
+      "learning_rate": 1.8508285504152979e-06,
+      "loss": 0.4567,
+      "step": 4926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0082085071432356,
+      "learning_rate": 1.847746221509067e-06,
+      "loss": 0.5707,
+      "step": 4927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7303740513197088,
+      "learning_rate": 1.8446662000811177e-06,
+      "loss": 0.5194,
+      "step": 4928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.047113165172449,
+      "learning_rate": 1.841588487003243e-06,
+      "loss": 0.5452,
+      "step": 4929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.084116646221239,
+      "learning_rate": 1.8385130831465837e-06,
+      "loss": 0.5708,
+      "step": 4930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0950580780682633,
+      "learning_rate": 1.8354399893816255e-06,
+      "loss": 0.493,
+      "step": 4931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8919342243880934,
+      "learning_rate": 1.8323692065782018e-06,
+      "loss": 0.6223,
+      "step": 4932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8199696446957525,
+      "learning_rate": 1.8293007356054903e-06,
+      "loss": 0.4649,
+      "step": 4933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.698647612495492,
+      "learning_rate": 1.8262345773320167e-06,
+      "loss": 0.459,
+      "step": 4934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6990941956266017,
+      "learning_rate": 1.8231707326256498e-06,
+      "loss": 0.5398,
+      "step": 4935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7811212543873634,
+      "learning_rate": 1.8201092023536048e-06,
+      "loss": 0.5527,
+      "step": 4936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6881102414024791,
+      "learning_rate": 1.817049987382441e-06,
+      "loss": 0.5361,
+      "step": 4937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5020667093075475,
+      "learning_rate": 1.8139930885780621e-06,
+      "loss": 0.4191,
+      "step": 4938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8538096400475168,
+      "learning_rate": 1.8109385068057183e-06,
+      "loss": 0.4871,
+      "step": 4939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4243309199429617,
+      "learning_rate": 1.8078862429300015e-06,
+      "loss": 0.5378,
+      "step": 4940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7266739369492696,
+      "learning_rate": 1.8048362978148492e-06,
+      "loss": 0.5374,
+      "step": 4941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9397268344337137,
+      "learning_rate": 1.8017886723235423e-06,
+      "loss": 0.5183,
+      "step": 4942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1983946019032548,
+      "learning_rate": 1.7987433673187026e-06,
+      "loss": 0.5435,
+      "step": 4943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6643422954127889,
+      "learning_rate": 1.7957003836623e-06,
+      "loss": 0.5168,
+      "step": 4944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8958033540425716,
+      "learning_rate": 1.7926597222156438e-06,
+      "loss": 0.5693,
+      "step": 4945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4969600995872887,
+      "learning_rate": 1.789621383839386e-06,
+      "loss": 0.5112,
+      "step": 4946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5847728248933044,
+      "learning_rate": 1.786585369393522e-06,
+      "loss": 0.4871,
+      "step": 4947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6885368669014185,
+      "learning_rate": 1.7835516797373908e-06,
+      "loss": 0.5516,
+      "step": 4948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.10845741343853,
+      "learning_rate": 1.7805203157296692e-06,
+      "loss": 0.4841,
+      "step": 4949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2306338785056767,
+      "learning_rate": 1.7774912782283815e-06,
+      "loss": 0.521,
+      "step": 4950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.066722432869018,
+      "learning_rate": 1.7744645680908878e-06,
+      "loss": 0.5137,
+      "step": 4951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8725179567977365,
+      "learning_rate": 1.771440186173894e-06,
+      "loss": 0.6448,
+      "step": 4952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2912001667281707,
+      "learning_rate": 1.7684181333334437e-06,
+      "loss": 0.5722,
+      "step": 4953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.593402868484713,
+      "learning_rate": 1.7653984104249221e-06,
+      "loss": 0.4989,
+      "step": 4954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7247326891824768,
+      "learning_rate": 1.7623810183030576e-06,
+      "loss": 0.5279,
+      "step": 4955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5695136706475192,
+      "learning_rate": 1.7593659578219147e-06,
+      "loss": 0.451,
+      "step": 4956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6220651412720477,
+      "learning_rate": 1.7563532298349018e-06,
+      "loss": 0.5307,
+      "step": 4957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6482656586233375,
+      "learning_rate": 1.7533428351947634e-06,
+      "loss": 0.5277,
+      "step": 4958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1923184539321494,
+      "learning_rate": 1.7503347747535859e-06,
+      "loss": 0.5335,
+      "step": 4959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.393908498139969,
+      "learning_rate": 1.7473290493627948e-06,
+      "loss": 0.5002,
+      "step": 4960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8815211162849108,
+      "learning_rate": 1.744325659873154e-06,
+      "loss": 0.4671,
+      "step": 4961
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.92371686972131,
+      "learning_rate": 1.7413246071347667e-06,
+      "loss": 0.5167,
+      "step": 4962
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7342718392057404,
+      "learning_rate": 1.7383258919970746e-06,
+      "loss": 0.484,
+      "step": 4963
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.763665981726944,
+      "learning_rate": 1.7353295153088578e-06,
+      "loss": 0.4274,
+      "step": 4964
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6297340243039615,
+      "learning_rate": 1.7323354779182345e-06,
+      "loss": 0.5522,
+      "step": 4965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.87227439826092,
+      "learning_rate": 1.729343780672661e-06,
+      "loss": 0.5344,
+      "step": 4966
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8416212899580628,
+      "learning_rate": 1.7263544244189302e-06,
+      "loss": 0.5321,
+      "step": 4967
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8044538898537963,
+      "learning_rate": 1.7233674100031728e-06,
+      "loss": 0.4641,
+      "step": 4968
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5082648424087117,
+      "learning_rate": 1.7203827382708582e-06,
+      "loss": 0.4938,
+      "step": 4969
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8977609316557993,
+      "learning_rate": 1.7174004100667907e-06,
+      "loss": 0.5816,
+      "step": 4970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2411096685458687,
+      "learning_rate": 1.7144204262351116e-06,
+      "loss": 0.5352,
+      "step": 4971
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8544067852941317,
+      "learning_rate": 1.7114427876192996e-06,
+      "loss": 0.5374,
+      "step": 4972
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5841161078420114,
+      "learning_rate": 1.7084674950621694e-06,
+      "loss": 0.5273,
+      "step": 4973
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8115761257733687,
+      "learning_rate": 1.7054945494058705e-06,
+      "loss": 0.6951,
+      "step": 4974
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.028255762304728,
+      "learning_rate": 1.7025239514918913e-06,
+      "loss": 0.4812,
+      "step": 4975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5000713289617629,
+      "learning_rate": 1.6995557021610477e-06,
+      "loss": 0.4614,
+      "step": 4976
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4803520081299926,
+      "learning_rate": 1.696589802253501e-06,
+      "loss": 0.5126,
+      "step": 4977
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7009031997259974,
+      "learning_rate": 1.6936262526087432e-06,
+      "loss": 0.5196,
+      "step": 4978
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7140563443968984,
+      "learning_rate": 1.690665054065599e-06,
+      "loss": 0.5504,
+      "step": 4979
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0247143493765347,
+      "learning_rate": 1.6877062074622296e-06,
+      "loss": 0.4731,
+      "step": 4980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6618411897790388,
+      "learning_rate": 1.6847497136361312e-06,
+      "loss": 0.5311,
+      "step": 4981
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.71529562644815,
+      "learning_rate": 1.6817955734241321e-06,
+      "loss": 0.459,
+      "step": 4982
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.414770796235784,
+      "learning_rate": 1.6788437876623963e-06,
+      "loss": 0.3954,
+      "step": 4983
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8423244292370542,
+      "learning_rate": 1.6758943571864206e-06,
+      "loss": 0.4688,
+      "step": 4984
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7783254262263768,
+      "learning_rate": 1.6729472828310334e-06,
+      "loss": 0.6095,
+      "step": 4985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5806904857797506,
+      "learning_rate": 1.670002565430401e-06,
+      "loss": 0.5004,
+      "step": 4986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.80246312631853,
+      "learning_rate": 1.6670602058180164e-06,
+      "loss": 0.641,
+      "step": 4987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7323860515097196,
+      "learning_rate": 1.6641202048267102e-06,
+      "loss": 0.5157,
+      "step": 4988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8800116808197844,
+      "learning_rate": 1.6611825632886424e-06,
+      "loss": 0.4939,
+      "step": 4989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7936698260329802,
+      "learning_rate": 1.658247282035307e-06,
+      "loss": 0.6518,
+      "step": 4990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9254299254060236,
+      "learning_rate": 1.6553143618975288e-06,
+      "loss": 0.522,
+      "step": 4991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7163223596115464,
+      "learning_rate": 1.6523838037054652e-06,
+      "loss": 0.5026,
+      "step": 4992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.011833890933118,
+      "learning_rate": 1.6494556082886038e-06,
+      "loss": 0.5206,
+      "step": 4993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7194968138493654,
+      "learning_rate": 1.646529776475765e-06,
+      "loss": 0.5523,
+      "step": 4994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8091457512004703,
+      "learning_rate": 1.6436063090950982e-06,
+      "loss": 0.4855,
+      "step": 4995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9968156825667618,
+      "learning_rate": 1.6406852069740876e-06,
+      "loss": 0.641,
+      "step": 4996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8820676434187373,
+      "learning_rate": 1.6377664709395403e-06,
+      "loss": 0.5294,
+      "step": 4997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8261920599849357,
+      "learning_rate": 1.634850101817601e-06,
+      "loss": 0.5126,
+      "step": 4998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.673556063805365,
+      "learning_rate": 1.631936100433742e-06,
+      "loss": 0.5105,
+      "step": 4999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.97230543979478,
+      "learning_rate": 1.629024467612762e-06,
+      "loss": 0.6392,
+      "step": 5000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6330907313445577,
+      "learning_rate": 1.6261152041787986e-06,
+      "loss": 0.4536,
+      "step": 5001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.848928142206014,
+      "learning_rate": 1.6232083109553088e-06,
+      "loss": 0.5408,
+      "step": 5002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.616066713151909,
+      "learning_rate": 1.6203037887650842e-06,
+      "loss": 0.5172,
+      "step": 5003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.468820410188697,
+      "learning_rate": 1.6174016384302415e-06,
+      "loss": 0.4957,
+      "step": 5004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.734288481234387,
+      "learning_rate": 1.6145018607722308e-06,
+      "loss": 0.5142,
+      "step": 5005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8307668098013805,
+      "learning_rate": 1.6116044566118261e-06,
+      "loss": 0.4701,
+      "step": 5006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7042165564644178,
+      "learning_rate": 1.6087094267691329e-06,
+      "loss": 0.3841,
+      "step": 5007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7070544789206097,
+      "learning_rate": 1.6058167720635832e-06,
+      "loss": 0.4575,
+      "step": 5008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6433227461654882,
+      "learning_rate": 1.6029264933139354e-06,
+      "loss": 0.5339,
+      "step": 5009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.924791407168595,
+      "learning_rate": 1.6000385913382777e-06,
+      "loss": 0.5675,
+      "step": 5010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.636439628371117,
+      "learning_rate": 1.5971530669540258e-06,
+      "loss": 0.5014,
+      "step": 5011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.016139972727754,
+      "learning_rate": 1.5942699209779189e-06,
+      "loss": 0.5154,
+      "step": 5012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5928975120504283,
+      "learning_rate": 1.5913891542260284e-06,
+      "loss": 0.4355,
+      "step": 5013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.444777652052721,
+      "learning_rate": 1.5885107675137468e-06,
+      "loss": 0.5137,
+      "step": 5014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5865900965516493,
+      "learning_rate": 1.5856347616557955e-06,
+      "loss": 0.5197,
+      "step": 5015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0239183833101237,
+      "learning_rate": 1.5827611374662233e-06,
+      "loss": 0.5187,
+      "step": 5016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8801081141129119,
+      "learning_rate": 1.579889895758403e-06,
+      "loss": 0.6266,
+      "step": 5017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8068258387987048,
+      "learning_rate": 1.5770210373450356e-06,
+      "loss": 0.531,
+      "step": 5018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.811660927783212,
+      "learning_rate": 1.5741545630381405e-06,
+      "loss": 0.5168,
+      "step": 5019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.87946223359963,
+      "learning_rate": 1.571290473649071e-06,
+      "loss": 0.447,
+      "step": 5020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8359252431543591,
+      "learning_rate": 1.5684287699884993e-06,
+      "loss": 0.4809,
+      "step": 5021
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7936281400146667,
+      "learning_rate": 1.5655694528664266e-06,
+      "loss": 0.5786,
+      "step": 5022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3971291987415286,
+      "learning_rate": 1.5627125230921725e-06,
+      "loss": 0.5657,
+      "step": 5023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.937835758767534,
+      "learning_rate": 1.5598579814743918e-06,
+      "loss": 0.4651,
+      "step": 5024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3587011342470863,
+      "learning_rate": 1.5570058288210522e-06,
+      "loss": 0.544,
+      "step": 5025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8475061122672076,
+      "learning_rate": 1.5541560659394507e-06,
+      "loss": 0.5321,
+      "step": 5026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6038732479155324,
+      "learning_rate": 1.5513086936362053e-06,
+      "loss": 0.551,
+      "step": 5027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1193936856501856,
+      "learning_rate": 1.5484637127172609e-06,
+      "loss": 0.5896,
+      "step": 5028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.820476016973179,
+      "learning_rate": 1.5456211239878815e-06,
+      "loss": 0.4195,
+      "step": 5029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3891006859174864,
+      "learning_rate": 1.5427809282526563e-06,
+      "loss": 0.4862,
+      "step": 5030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9514067864731706,
+      "learning_rate": 1.5399431263154973e-06,
+      "loss": 0.5275,
+      "step": 5031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7847109584549485,
+      "learning_rate": 1.537107718979638e-06,
+      "loss": 0.5227,
+      "step": 5032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.830203295746685,
+      "learning_rate": 1.5342747070476339e-06,
+      "loss": 0.521,
+      "step": 5033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6093107433368663,
+      "learning_rate": 1.5314440913213645e-06,
+      "loss": 0.4087,
+      "step": 5034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.69180617805871,
+      "learning_rate": 1.5286158726020283e-06,
+      "loss": 0.477,
+      "step": 5035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9546526756763767,
+      "learning_rate": 1.5257900516901469e-06,
+      "loss": 0.5055,
+      "step": 5036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7760572261642475,
+      "learning_rate": 1.522966629385564e-06,
+      "loss": 0.4972,
+      "step": 5037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8180021252564913,
+      "learning_rate": 1.520145606487442e-06,
+      "loss": 0.5822,
+      "step": 5038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2347877992975627,
+      "learning_rate": 1.5173269837942683e-06,
+      "loss": 0.5399,
+      "step": 5039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.180065211032639,
+      "learning_rate": 1.5145107621038447e-06,
+      "loss": 0.6354,
+      "step": 5040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.60187702257472,
+      "learning_rate": 1.5116969422132966e-06,
+      "loss": 0.4012,
+      "step": 5041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9365831006341843,
+      "learning_rate": 1.5088855249190715e-06,
+      "loss": 0.5431,
+      "step": 5042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5123992127813053,
+      "learning_rate": 1.506076511016935e-06,
+      "loss": 0.5357,
+      "step": 5043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8750932725900131,
+      "learning_rate": 1.5032699013019724e-06,
+      "loss": 0.5039,
+      "step": 5044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3188106222793796,
+      "learning_rate": 1.5004656965685892e-06,
+      "loss": 0.5447,
+      "step": 5045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9783059221775405,
+      "learning_rate": 1.4976638976105063e-06,
+      "loss": 0.4522,
+      "step": 5046
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5129917410979332,
+      "learning_rate": 1.4948645052207721e-06,
+      "loss": 0.5386,
+      "step": 5047
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1817799299323535,
+      "learning_rate": 1.4920675201917467e-06,
+      "loss": 0.5405,
+      "step": 5048
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.720425862641818,
+      "learning_rate": 1.4892729433151099e-06,
+      "loss": 0.5185,
+      "step": 5049
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4986377329533587,
+      "learning_rate": 1.4864807753818623e-06,
+      "loss": 0.4609,
+      "step": 5050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7290282067973597,
+      "learning_rate": 1.4836910171823205e-06,
+      "loss": 0.4796,
+      "step": 5051
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8517390986204245,
+      "learning_rate": 1.4809036695061184e-06,
+      "loss": 0.5297,
+      "step": 5052
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.244286682682886,
+      "learning_rate": 1.4781187331422109e-06,
+      "loss": 0.4951,
+      "step": 5053
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7595923768266863,
+      "learning_rate": 1.475336208878868e-06,
+      "loss": 0.5237,
+      "step": 5054
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7924015890493352,
+      "learning_rate": 1.4725560975036756e-06,
+      "loss": 0.5191,
+      "step": 5055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.027951418347729,
+      "learning_rate": 1.469778399803541e-06,
+      "loss": 0.4132,
+      "step": 5056
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1840564518740333,
+      "learning_rate": 1.4670031165646836e-06,
+      "loss": 0.4969,
+      "step": 5057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.036273516414469,
+      "learning_rate": 1.4642302485726423e-06,
+      "loss": 0.5533,
+      "step": 5058
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5461395717286757,
+      "learning_rate": 1.4614597966122712e-06,
+      "loss": 0.5211,
+      "step": 5059
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8841591051189739,
+      "learning_rate": 1.4586917614677409e-06,
+      "loss": 0.5006,
+      "step": 5060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7979059929182946,
+      "learning_rate": 1.4559261439225415e-06,
+      "loss": 0.5248,
+      "step": 5061
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7628072119628593,
+      "learning_rate": 1.4531629447594685e-06,
+      "loss": 0.4168,
+      "step": 5062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8471181065902216,
+      "learning_rate": 1.4504021647606448e-06,
+      "loss": 0.688,
+      "step": 5063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5824525644209788,
+      "learning_rate": 1.4476438047075004e-06,
+      "loss": 0.5043,
+      "step": 5064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5739212063863546,
+      "learning_rate": 1.4448878653807841e-06,
+      "loss": 0.4656,
+      "step": 5065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6618296306682958,
+      "learning_rate": 1.4421343475605597e-06,
+      "loss": 0.4678,
+      "step": 5066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4446098608671596,
+      "learning_rate": 1.4393832520262042e-06,
+      "loss": 0.6074,
+      "step": 5067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9172106627415002,
+      "learning_rate": 1.4366345795564084e-06,
+      "loss": 0.4819,
+      "step": 5068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7136776300025125,
+      "learning_rate": 1.4338883309291773e-06,
+      "loss": 0.4562,
+      "step": 5069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.181291224346496,
+      "learning_rate": 1.4311445069218355e-06,
+      "loss": 0.5132,
+      "step": 5070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6737398948226008,
+      "learning_rate": 1.4284031083110127e-06,
+      "loss": 0.4339,
+      "step": 5071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6460465100227946,
+      "learning_rate": 1.4256641358726586e-06,
+      "loss": 0.4653,
+      "step": 5072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7721013472772476,
+      "learning_rate": 1.4229275903820306e-06,
+      "loss": 0.5248,
+      "step": 5073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1086098266300315,
+      "learning_rate": 1.4201934726137046e-06,
+      "loss": 0.5401,
+      "step": 5074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9795603903459378,
+      "learning_rate": 1.417461783341565e-06,
+      "loss": 0.4951,
+      "step": 5075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8606139703371425,
+      "learning_rate": 1.414732523338812e-06,
+      "loss": 0.4719,
+      "step": 5076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8914813802893427,
+      "learning_rate": 1.4120056933779558e-06,
+      "loss": 0.5227,
+      "step": 5077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7847853078936804,
+      "learning_rate": 1.409281294230821e-06,
+      "loss": 0.6013,
+      "step": 5078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6262497912583507,
+      "learning_rate": 1.4065593266685428e-06,
+      "loss": 0.4807,
+      "step": 5079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.339617682607038,
+      "learning_rate": 1.4038397914615686e-06,
+      "loss": 0.5664,
+      "step": 5080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7913254029211472,
+      "learning_rate": 1.4011226893796559e-06,
+      "loss": 0.5182,
+      "step": 5081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.742625846738827,
+      "learning_rate": 1.398408021191876e-06,
+      "loss": 0.5078,
+      "step": 5082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.076915666245757,
+      "learning_rate": 1.395695787666611e-06,
+      "loss": 0.6155,
+      "step": 5083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0308173179880917,
+      "learning_rate": 1.3929859895715502e-06,
+      "loss": 0.548,
+      "step": 5084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7728802199159044,
+      "learning_rate": 1.3902786276736969e-06,
+      "loss": 0.5889,
+      "step": 5085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0787478477742005,
+      "learning_rate": 1.3875737027393655e-06,
+      "loss": 0.5793,
+      "step": 5086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6801962701542277,
+      "learning_rate": 1.384871215534177e-06,
+      "loss": 0.465,
+      "step": 5087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8579355419427432,
+      "learning_rate": 1.3821711668230675e-06,
+      "loss": 0.5807,
+      "step": 5088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8725234529138788,
+      "learning_rate": 1.379473557370279e-06,
+      "loss": 0.5815,
+      "step": 5089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.755624060080873,
+      "learning_rate": 1.3767783879393637e-06,
+      "loss": 0.5274,
+      "step": 5090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.942279880634414,
+      "learning_rate": 1.374085659293184e-06,
+      "loss": 0.519,
+      "step": 5091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7835758985744542,
+      "learning_rate": 1.3713953721939122e-06,
+      "loss": 0.6038,
+      "step": 5092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8345492206297827,
+      "learning_rate": 1.3687075274030238e-06,
+      "loss": 0.4685,
+      "step": 5093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.756630973984486,
+      "learning_rate": 1.3660221256813133e-06,
+      "loss": 0.505,
+      "step": 5094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.003049434694826,
+      "learning_rate": 1.3633391677888763e-06,
+      "loss": 0.5589,
+      "step": 5095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8384459748559607,
+      "learning_rate": 1.3606586544851185e-06,
+      "loss": 0.6693,
+      "step": 5096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8156488708206489,
+      "learning_rate": 1.3579805865287521e-06,
+      "loss": 0.5339,
+      "step": 5097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6568983942313786,
+      "learning_rate": 1.3553049646777993e-06,
+      "loss": 0.429,
+      "step": 5098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5332971885027842,
+      "learning_rate": 1.3526317896895903e-06,
+      "loss": 0.4721,
+      "step": 5099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8508114465968464,
+      "learning_rate": 1.3499610623207604e-06,
+      "loss": 0.4621,
+      "step": 5100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5976559625205593,
+      "learning_rate": 1.3472927833272541e-06,
+      "loss": 0.4828,
+      "step": 5101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8152619416398459,
+      "learning_rate": 1.3446269534643209e-06,
+      "loss": 0.5277,
+      "step": 5102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7139765940091807,
+      "learning_rate": 1.34196357348652e-06,
+      "loss": 0.5281,
+      "step": 5103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9477938484116302,
+      "learning_rate": 1.3393026441477163e-06,
+      "loss": 0.4739,
+      "step": 5104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.692641557181617,
+      "learning_rate": 1.3366441662010765e-06,
+      "loss": 0.5069,
+      "step": 5105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1094265929509715,
+      "learning_rate": 1.3339881403990796e-06,
+      "loss": 0.6038,
+      "step": 5106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9109003926345178,
+      "learning_rate": 1.3313345674935085e-06,
+      "loss": 0.5054,
+      "step": 5107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5370219061712518,
+      "learning_rate": 1.3286834482354506e-06,
+      "loss": 0.5643,
+      "step": 5108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8387641127947565,
+      "learning_rate": 1.3260347833752995e-06,
+      "loss": 0.6291,
+      "step": 5109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6352801586014,
+      "learning_rate": 1.3233885736627538e-06,
+      "loss": 0.5786,
+      "step": 5110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5951140900923717,
+      "learning_rate": 1.3207448198468199e-06,
+      "loss": 0.4953,
+      "step": 5111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.135934944301686,
+      "learning_rate": 1.3181035226758044e-06,
+      "loss": 0.5862,
+      "step": 5112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9131364416287586,
+      "learning_rate": 1.3154646828973217e-06,
+      "loss": 0.5877,
+      "step": 5113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8989433410787075,
+      "learning_rate": 1.3128283012582887e-06,
+      "loss": 0.5749,
+      "step": 5114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.822560411883798,
+      "learning_rate": 1.3101943785049309e-06,
+      "loss": 0.5517,
+      "step": 5115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6712266334103842,
+      "learning_rate": 1.3075629153827685e-06,
+      "loss": 0.4679,
+      "step": 5116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9379635456605249,
+      "learning_rate": 1.3049339126366388e-06,
+      "loss": 0.6678,
+      "step": 5117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7033727509912866,
+      "learning_rate": 1.3023073710106726e-06,
+      "loss": 0.5492,
+      "step": 5118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9994550262475264,
+      "learning_rate": 1.2996832912483059e-06,
+      "loss": 0.5281,
+      "step": 5119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6856098195068068,
+      "learning_rate": 1.2970616740922804e-06,
+      "loss": 0.4786,
+      "step": 5120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.103961094453419,
+      "learning_rate": 1.294442520284639e-06,
+      "loss": 0.5566,
+      "step": 5121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2326539111347614,
+      "learning_rate": 1.2918258305667264e-06,
+      "loss": 0.4707,
+      "step": 5122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.605541532261124,
+      "learning_rate": 1.2892116056791927e-06,
+      "loss": 0.5043,
+      "step": 5123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.362332019793204,
+      "learning_rate": 1.2865998463619878e-06,
+      "loss": 0.5054,
+      "step": 5124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.253291923858368,
+      "learning_rate": 1.283990553354365e-06,
+      "loss": 0.4658,
+      "step": 5125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8074872336294796,
+      "learning_rate": 1.2813837273948815e-06,
+      "loss": 0.5794,
+      "step": 5126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0235277821711044,
+      "learning_rate": 1.2787793692213902e-06,
+      "loss": 0.4832,
+      "step": 5127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1496218951088295,
+      "learning_rate": 1.2761774795710502e-06,
+      "loss": 0.4966,
+      "step": 5128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5852946683446774,
+      "learning_rate": 1.2735780591803205e-06,
+      "loss": 0.5019,
+      "step": 5129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.026115213240645,
+      "learning_rate": 1.2709811087849633e-06,
+      "loss": 0.4967,
+      "step": 5130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8330725461854414,
+      "learning_rate": 1.2683866291200386e-06,
+      "loss": 0.5695,
+      "step": 5131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.689992837638145,
+      "learning_rate": 1.2657946209199078e-06,
+      "loss": 0.5414,
+      "step": 5132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9566700436395936,
+      "learning_rate": 1.2632050849182365e-06,
+      "loss": 0.5129,
+      "step": 5133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8906013532032,
+      "learning_rate": 1.2606180218479836e-06,
+      "loss": 0.5254,
+      "step": 5134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3816960275181103,
+      "learning_rate": 1.2580334324414145e-06,
+      "loss": 0.5332,
+      "step": 5135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7130715734619462,
+      "learning_rate": 1.255451317430091e-06,
+      "loss": 0.5631,
+      "step": 5136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2567779111989554,
+      "learning_rate": 1.2528716775448747e-06,
+      "loss": 0.4193,
+      "step": 5137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.671408985078462,
+      "learning_rate": 1.2502945135159272e-06,
+      "loss": 0.5423,
+      "step": 5138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9493884022789354,
+      "learning_rate": 1.2477198260727085e-06,
+      "loss": 0.5419,
+      "step": 5139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1960427028358818,
+      "learning_rate": 1.2451476159439824e-06,
+      "loss": 0.4992,
+      "step": 5140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8789671460527115,
+      "learning_rate": 1.242577883857804e-06,
+      "loss": 0.5534,
+      "step": 5141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8289307253578402,
+      "learning_rate": 1.2400106305415317e-06,
+      "loss": 0.5288,
+      "step": 5142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2094304702831526,
+      "learning_rate": 1.2374458567218217e-06,
+      "loss": 0.4858,
+      "step": 5143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8218672055884557,
+      "learning_rate": 1.234883563124627e-06,
+      "loss": 0.4889,
+      "step": 5144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0335487190089845,
+      "learning_rate": 1.2323237504751995e-06,
+      "loss": 0.6166,
+      "step": 5145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.695783373478773,
+      "learning_rate": 1.2297664194980885e-06,
+      "loss": 0.5322,
+      "step": 5146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6192410963625512,
+      "learning_rate": 1.2272115709171418e-06,
+      "loss": 0.5716,
+      "step": 5147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.899105075537213,
+      "learning_rate": 1.2246592054555062e-06,
+      "loss": 0.5567,
+      "step": 5148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8145597196723433,
+      "learning_rate": 1.2221093238356185e-06,
+      "loss": 0.4845,
+      "step": 5149
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8220124539874192,
+      "learning_rate": 1.2195619267792192e-06,
+      "loss": 0.6416,
+      "step": 5150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.709014362923482,
+      "learning_rate": 1.2170170150073446e-06,
+      "loss": 0.5298,
+      "step": 5151
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7550896984674695,
+      "learning_rate": 1.2144745892403253e-06,
+      "loss": 0.5005,
+      "step": 5152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6856334703334788,
+      "learning_rate": 1.2119346501977914e-06,
+      "loss": 0.4995,
+      "step": 5153
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9158138178242967,
+      "learning_rate": 1.2093971985986653e-06,
+      "loss": 0.5704,
+      "step": 5154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0943848657041704,
+      "learning_rate": 1.2068622351611691e-06,
+      "loss": 0.4765,
+      "step": 5155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.727781165296325,
+      "learning_rate": 1.2043297606028181e-06,
+      "loss": 0.4693,
+      "step": 5156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6140846907027393,
+      "learning_rate": 1.2017997756404232e-06,
+      "loss": 0.5155,
+      "step": 5157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0995424196745622,
+      "learning_rate": 1.1992722809900925e-06,
+      "loss": 0.5367,
+      "step": 5158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8225506933728797,
+      "learning_rate": 1.1967472773672284e-06,
+      "loss": 0.5573,
+      "step": 5159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7259340784892927,
+      "learning_rate": 1.1942247654865259e-06,
+      "loss": 0.5118,
+      "step": 5160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.086841506640954,
+      "learning_rate": 1.1917047460619778e-06,
+      "loss": 0.5231,
+      "step": 5161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.900071692638027,
+      "learning_rate": 1.1891872198068689e-06,
+      "loss": 0.522,
+      "step": 5162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6348231195730094,
+      "learning_rate": 1.1866721874337827e-06,
+      "loss": 0.5216,
+      "step": 5163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3979420034523313,
+      "learning_rate": 1.1841596496545927e-06,
+      "loss": 0.503,
+      "step": 5164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.377491124679425,
+      "learning_rate": 1.1816496071804662e-06,
+      "loss": 0.5429,
+      "step": 5165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8742522723957227,
+      "learning_rate": 1.1791420607218673e-06,
+      "loss": 0.4982,
+      "step": 5166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9025517265451866,
+      "learning_rate": 1.1766370109885506e-06,
+      "loss": 0.5694,
+      "step": 5167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6712161937155536,
+      "learning_rate": 1.1741344586895642e-06,
+      "loss": 0.5397,
+      "step": 5168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7376241759567208,
+      "learning_rate": 1.171634404533254e-06,
+      "loss": 0.4898,
+      "step": 5169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8917353647736783,
+      "learning_rate": 1.1691368492272504e-06,
+      "loss": 0.6945,
+      "step": 5170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.815499629569795,
+      "learning_rate": 1.1666417934784824e-06,
+      "loss": 0.5868,
+      "step": 5171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3095035552932344,
+      "learning_rate": 1.1641492379931717e-06,
+      "loss": 0.5366,
+      "step": 5172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6771545398193433,
+      "learning_rate": 1.1616591834768299e-06,
+      "loss": 0.52,
+      "step": 5173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7499162888819983,
+      "learning_rate": 1.1591716306342615e-06,
+      "loss": 0.5523,
+      "step": 5174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9644943423873569,
+      "learning_rate": 1.1566865801695649e-06,
+      "loss": 0.5388,
+      "step": 5175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0234525582644203,
+      "learning_rate": 1.1542040327861259e-06,
+      "loss": 0.5938,
+      "step": 5176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.874571849029857,
+      "learning_rate": 1.1517239891866261e-06,
+      "loss": 0.5461,
+      "step": 5177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6686730271435308,
+      "learning_rate": 1.149246450073036e-06,
+      "loss": 0.4022,
+      "step": 5178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6498363922448553,
+      "learning_rate": 1.146771416146617e-06,
+      "loss": 0.5285,
+      "step": 5179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7305657694621686,
+      "learning_rate": 1.1442988881079243e-06,
+      "loss": 0.5566,
+      "step": 5180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.365178449648462,
+      "learning_rate": 1.1418288666568e-06,
+      "loss": 0.5819,
+      "step": 5181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6559135401662564,
+      "learning_rate": 1.1393613524923797e-06,
+      "loss": 0.5033,
+      "step": 5182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0099144285541226,
+      "learning_rate": 1.1368963463130866e-06,
+      "loss": 0.545,
+      "step": 5183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7622253844108904,
+      "learning_rate": 1.1344338488166362e-06,
+      "loss": 0.5423,
+      "step": 5184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7934535850506377,
+      "learning_rate": 1.1319738607000342e-06,
+      "loss": 0.4703,
+      "step": 5185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6527646734241868,
+      "learning_rate": 1.1295163826595724e-06,
+      "loss": 0.5031,
+      "step": 5186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0401532088373284,
+      "learning_rate": 1.1270614153908376e-06,
+      "loss": 0.527,
+      "step": 5187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2755705831791917,
+      "learning_rate": 1.1246089595887023e-06,
+      "loss": 0.4195,
+      "step": 5188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8116781943864253,
+      "learning_rate": 1.1221590159473294e-06,
+      "loss": 0.4997,
+      "step": 5189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6466695408108516,
+      "learning_rate": 1.119711585160169e-06,
+      "loss": 0.4944,
+      "step": 5190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.800611906141071,
+      "learning_rate": 1.117266667919965e-06,
+      "loss": 0.5318,
+      "step": 5191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5182392118996162,
+      "learning_rate": 1.1148242649187403e-06,
+      "loss": 0.5019,
+      "step": 5192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2233835955485475,
+      "learning_rate": 1.1123843768478148e-06,
+      "loss": 0.5186,
+      "step": 5193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8095459580855782,
+      "learning_rate": 1.1099470043977933e-06,
+      "loss": 0.5392,
+      "step": 5194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5608408191388272,
+      "learning_rate": 1.1075121482585704e-06,
+      "loss": 0.5425,
+      "step": 5195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6726470236439297,
+      "learning_rate": 1.1050798091193249e-06,
+      "loss": 0.5311,
+      "step": 5196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7899810516994883,
+      "learning_rate": 1.1026499876685258e-06,
+      "loss": 0.4494,
+      "step": 5197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0326943906176274,
+      "learning_rate": 1.100222684593929e-06,
+      "loss": 0.5415,
+      "step": 5198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0081576088247592,
+      "learning_rate": 1.097797900582579e-06,
+      "loss": 0.4491,
+      "step": 5199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9240666558775097,
+      "learning_rate": 1.0953756363208034e-06,
+      "loss": 0.4795,
+      "step": 5200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9174967098595161,
+      "learning_rate": 1.0929558924942217e-06,
+      "loss": 0.6177,
+      "step": 5201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.594296120199087,
+      "learning_rate": 1.090538669787735e-06,
+      "loss": 0.5254,
+      "step": 5202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6700812658425377,
+      "learning_rate": 1.088123968885534e-06,
+      "loss": 0.5406,
+      "step": 5203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8683717350893911,
+      "learning_rate": 1.0857117904710946e-06,
+      "loss": 0.556,
+      "step": 5204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7809489848110969,
+      "learning_rate": 1.0833021352271788e-06,
+      "loss": 0.4446,
+      "step": 5205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7388158331788843,
+      "learning_rate": 1.0808950038358357e-06,
+      "loss": 0.5087,
+      "step": 5206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1368779536851146,
+      "learning_rate": 1.0784903969783966e-06,
+      "loss": 0.4609,
+      "step": 5207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8073681186479131,
+      "learning_rate": 1.0760883153354818e-06,
+      "loss": 0.545,
+      "step": 5208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.072679811370707,
+      "learning_rate": 1.0736887595869927e-06,
+      "loss": 0.64,
+      "step": 5209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8289944912635068,
+      "learning_rate": 1.071291730412123e-06,
+      "loss": 0.5066,
+      "step": 5210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8727069651224963,
+      "learning_rate": 1.068897228489344e-06,
+      "loss": 0.4719,
+      "step": 5211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.082635995197506,
+      "learning_rate": 1.0665052544964184e-06,
+      "loss": 0.5972,
+      "step": 5212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.122242474520823,
+      "learning_rate": 1.0641158091103832e-06,
+      "loss": 0.5936,
+      "step": 5213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2425168653351197,
+      "learning_rate": 1.0617288930075697e-06,
+      "loss": 0.5801,
+      "step": 5214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.966666584576513,
+      "learning_rate": 1.0593445068635876e-06,
+      "loss": 0.6281,
+      "step": 5215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8127351278921406,
+      "learning_rate": 1.0569626513533338e-06,
+      "loss": 0.7006,
+      "step": 5216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.12709803441382,
+      "learning_rate": 1.0545833271509864e-06,
+      "loss": 0.4771,
+      "step": 5217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7355264210652722,
+      "learning_rate": 1.0522065349300103e-06,
+      "loss": 0.6501,
+      "step": 5218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0803342724104734,
+      "learning_rate": 1.0498322753631484e-06,
+      "loss": 0.5458,
+      "step": 5219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7997909430395516,
+      "learning_rate": 1.0474605491224321e-06,
+      "loss": 0.4939,
+      "step": 5220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7497370683788545,
+      "learning_rate": 1.0450913568791731e-06,
+      "loss": 0.6239,
+      "step": 5221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7363334207554824,
+      "learning_rate": 1.0427246993039653e-06,
+      "loss": 0.4994,
+      "step": 5222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1015349972631445,
+      "learning_rate": 1.040360577066688e-06,
+      "loss": 0.5091,
+      "step": 5223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.685971392438554,
+      "learning_rate": 1.0379989908364997e-06,
+      "loss": 0.4692,
+      "step": 5224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7329612648541084,
+      "learning_rate": 1.0356399412818418e-06,
+      "loss": 0.4524,
+      "step": 5225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7111335767994824,
+      "learning_rate": 1.0332834290704397e-06,
+      "loss": 0.5371,
+      "step": 5226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7074629116332267,
+      "learning_rate": 1.0309294548692994e-06,
+      "loss": 0.5103,
+      "step": 5227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8096125004311864,
+      "learning_rate": 1.028578019344706e-06,
+      "loss": 0.641,
+      "step": 5228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7467133141202027,
+      "learning_rate": 1.0262291231622313e-06,
+      "loss": 0.5137,
+      "step": 5229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.646112972024216,
+      "learning_rate": 1.0238827669867235e-06,
+      "loss": 0.4774,
+      "step": 5230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8217971439832121,
+      "learning_rate": 1.0215389514823148e-06,
+      "loss": 0.4597,
+      "step": 5231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.143975144202015,
+      "learning_rate": 1.0191976773124157e-06,
+      "loss": 0.5033,
+      "step": 5232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7886138983961486,
+      "learning_rate": 1.0168589451397204e-06,
+      "loss": 0.6485,
+      "step": 5233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.84559814372103,
+      "learning_rate": 1.014522755626205e-06,
+      "loss": 0.5478,
+      "step": 5234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.175256342045875,
+      "learning_rate": 1.0121891094331172e-06,
+      "loss": 0.5568,
+      "step": 5235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9469544136173578,
+      "learning_rate": 1.009858007220994e-06,
+      "loss": 0.5374,
+      "step": 5236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0540634095130983,
+      "learning_rate": 1.0075294496496479e-06,
+      "loss": 0.5973,
+      "step": 5237
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.233591641443994,
+      "learning_rate": 1.0052034373781716e-06,
+      "loss": 0.5647,
+      "step": 5238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6872358093985638,
+      "learning_rate": 1.0028799710649406e-06,
+      "loss": 0.5653,
+      "step": 5239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6765275414935994,
+      "learning_rate": 1.0005590513676045e-06,
+      "loss": 0.5314,
+      "step": 5240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8738442596259778,
+      "learning_rate": 9.982406789430964e-07,
+      "loss": 0.6521,
+      "step": 5241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0239246099119903,
+      "learning_rate": 9.959248544476252e-07,
+      "loss": 0.444,
+      "step": 5242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.561377809108836,
+      "learning_rate": 9.936115785366817e-07,
+      "loss": 0.4766,
+      "step": 5243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.796086595195485,
+      "learning_rate": 9.913008518650336e-07,
+      "loss": 0.554,
+      "step": 5244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8504693730335595,
+      "learning_rate": 9.889926750867251e-07,
+      "loss": 0.5706,
+      "step": 5245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.66106399801966,
+      "learning_rate": 9.866870488550838e-07,
+      "loss": 0.5376,
+      "step": 5246
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.732770873267827,
+      "learning_rate": 9.843839738227113e-07,
+      "loss": 0.5263,
+      "step": 5247
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8319236798312235,
+      "learning_rate": 9.820834506414866e-07,
+      "loss": 0.4627,
+      "step": 5248
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9066484753758912,
+      "learning_rate": 9.797854799625706e-07,
+      "loss": 0.574,
+      "step": 5249
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9054011451668458,
+      "learning_rate": 9.77490062436396e-07,
+      "loss": 0.5199,
+      "step": 5250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.748370521586706,
+      "learning_rate": 9.751971987126785e-07,
+      "loss": 0.5747,
+      "step": 5251
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8437125070605271,
+      "learning_rate": 9.72906889440407e-07,
+      "loss": 0.5425,
+      "step": 5252
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7996297299019355,
+      "learning_rate": 9.706191352678495e-07,
+      "loss": 0.5863,
+      "step": 5253
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9797445814339092,
+      "learning_rate": 9.683339368425494e-07,
+      "loss": 0.4442,
+      "step": 5254
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.301360728157937,
+      "learning_rate": 9.660512948113276e-07,
+      "loss": 0.5093,
+      "step": 5255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9741875236267497,
+      "learning_rate": 9.637712098202812e-07,
+      "loss": 0.6047,
+      "step": 5256
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8286259522747137,
+      "learning_rate": 9.614936825147836e-07,
+      "loss": 0.4197,
+      "step": 5257
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9753449749492242,
+      "learning_rate": 9.592187135394826e-07,
+      "loss": 0.5964,
+      "step": 5258
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.662970289562128,
+      "learning_rate": 9.569463035383054e-07,
+      "loss": 0.5274,
+      "step": 5259
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7947871182110071,
+      "learning_rate": 9.546764531544528e-07,
+      "loss": 0.5182,
+      "step": 5260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.967234683559804,
+      "learning_rate": 9.524091630304e-07,
+      "loss": 0.4778,
+      "step": 5261
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.035944974031275,
+      "learning_rate": 9.50144433807898e-07,
+      "loss": 0.5527,
+      "step": 5262
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.622583251935238,
+      "learning_rate": 9.478822661279763e-07,
+      "loss": 0.5644,
+      "step": 5263
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.999788774876618,
+      "learning_rate": 9.456226606309338e-07,
+      "loss": 0.5488,
+      "step": 5264
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0404384719333524,
+      "learning_rate": 9.4336561795635e-07,
+      "loss": 0.5029,
+      "step": 5265
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.585923598381144,
+      "learning_rate": 9.411111387430738e-07,
+      "loss": 0.4405,
+      "step": 5266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.80992472432184,
+      "learning_rate": 9.388592236292316e-07,
+      "loss": 0.5214,
+      "step": 5267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2809310290636335,
+      "learning_rate": 9.366098732522233e-07,
+      "loss": 0.5564,
+      "step": 5268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.129480916305318,
+      "learning_rate": 9.343630882487221e-07,
+      "loss": 0.5677,
+      "step": 5269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3516855244991968,
+      "learning_rate": 9.321188692546767e-07,
+      "loss": 0.4861,
+      "step": 5270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7339292162925721,
+      "learning_rate": 9.298772169053083e-07,
+      "loss": 0.5336,
+      "step": 5271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3734983396401192,
+      "learning_rate": 9.276381318351124e-07,
+      "loss": 0.5913,
+      "step": 5272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9427667570393463,
+      "learning_rate": 9.254016146778555e-07,
+      "loss": 0.4558,
+      "step": 5273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8050725305870916,
+      "learning_rate": 9.231676660665812e-07,
+      "loss": 0.5108,
+      "step": 5274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5035685237469003,
+      "learning_rate": 9.209362866336024e-07,
+      "loss": 0.5083,
+      "step": 5275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9222617148758074,
+      "learning_rate": 9.187074770105076e-07,
+      "loss": 0.5192,
+      "step": 5276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6874430845768307,
+      "learning_rate": 9.164812378281562e-07,
+      "loss": 0.5235,
+      "step": 5277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.021595878648317,
+      "learning_rate": 9.1425756971668e-07,
+      "loss": 0.4929,
+      "step": 5278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8483119433298313,
+      "learning_rate": 9.120364733054843e-07,
+      "loss": 0.6652,
+      "step": 5279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7914011710969091,
+      "learning_rate": 9.098179492232451e-07,
+      "loss": 0.6407,
+      "step": 5280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6118244689642303,
+      "learning_rate": 9.076019980979111e-07,
+      "loss": 0.4607,
+      "step": 5281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3566805398053354,
+      "learning_rate": 9.05388620556703e-07,
+      "loss": 0.4352,
+      "step": 5282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5532202185607165,
+      "learning_rate": 9.03177817226113e-07,
+      "loss": 0.4479,
+      "step": 5283
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8779308357139086,
+      "learning_rate": 9.009695887319026e-07,
+      "loss": 0.5102,
+      "step": 5284
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9971967628463545,
+      "learning_rate": 8.987639356991085e-07,
+      "loss": 0.5326,
+      "step": 5285
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8313354481827557,
+      "learning_rate": 8.965608587520347e-07,
+      "loss": 0.4561,
+      "step": 5286
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7675946829945728,
+      "learning_rate": 8.94360358514258e-07,
+      "loss": 0.4392,
+      "step": 5287
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2582388195007588,
+      "learning_rate": 8.921624356086256e-07,
+      "loss": 0.4684,
+      "step": 5288
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7727714692704795,
+      "learning_rate": 8.899670906572544e-07,
+      "loss": 0.5085,
+      "step": 5289
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.268744031839971,
+      "learning_rate": 8.877743242815318e-07,
+      "loss": 0.4687,
+      "step": 5290
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.037624107399918,
+      "learning_rate": 8.855841371021168e-07,
+      "loss": 0.4734,
+      "step": 5291
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2803054571862442,
+      "learning_rate": 8.833965297389368e-07,
+      "loss": 0.5849,
+      "step": 5292
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.802156147901091,
+      "learning_rate": 8.81211502811189e-07,
+      "loss": 0.4952,
+      "step": 5293
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.038344613873767,
+      "learning_rate": 8.790290569373416e-07,
+      "loss": 0.5012,
+      "step": 5294
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7030541889938395,
+      "learning_rate": 8.768491927351308e-07,
+      "loss": 0.5988,
+      "step": 5295
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7662467322430102,
+      "learning_rate": 8.746719108215617e-07,
+      "loss": 0.4867,
+      "step": 5296
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7629173201174753,
+      "learning_rate": 8.724972118129116e-07,
+      "loss": 0.4561,
+      "step": 5297
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6496331822895667,
+      "learning_rate": 8.703250963247223e-07,
+      "loss": 0.5032,
+      "step": 5298
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0684177070026664,
+      "learning_rate": 8.681555649718076e-07,
+      "loss": 0.5031,
+      "step": 5299
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9396000170723604,
+      "learning_rate": 8.659886183682475e-07,
+      "loss": 0.5368,
+      "step": 5300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4742340863210406,
+      "learning_rate": 8.638242571273935e-07,
+      "loss": 0.4614,
+      "step": 5301
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7909608233149192,
+      "learning_rate": 8.616624818618635e-07,
+      "loss": 0.4924,
+      "step": 5302
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1255905431663535,
+      "learning_rate": 8.595032931835423e-07,
+      "loss": 0.6011,
+      "step": 5303
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7443159683210419,
+      "learning_rate": 8.573466917035834e-07,
+      "loss": 0.4909,
+      "step": 5304
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7661138895740278,
+      "learning_rate": 8.551926780324094e-07,
+      "loss": 0.5421,
+      "step": 5305
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.127461105595029,
+      "learning_rate": 8.530412527797083e-07,
+      "loss": 0.5018,
+      "step": 5306
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0740059082407782,
+      "learning_rate": 8.508924165544375e-07,
+      "loss": 0.5724,
+      "step": 5307
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9159477014825794,
+      "learning_rate": 8.487461699648203e-07,
+      "loss": 0.4652,
+      "step": 5308
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0518633861192073,
+      "learning_rate": 8.466025136183476e-07,
+      "loss": 0.4339,
+      "step": 5309
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2584939314805896,
+      "learning_rate": 8.44461448121775e-07,
+      "loss": 0.5335,
+      "step": 5310
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8525837529330154,
+      "learning_rate": 8.42322974081129e-07,
+      "loss": 0.5315,
+      "step": 5311
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7416474001349338,
+      "learning_rate": 8.401870921016996e-07,
+      "loss": 0.5718,
+      "step": 5312
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5656168509636352,
+      "learning_rate": 8.380538027880425e-07,
+      "loss": 0.5043,
+      "step": 5313
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8962757521534575,
+      "learning_rate": 8.35923106743981e-07,
+      "loss": 0.5004,
+      "step": 5314
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4684896828799994,
+      "learning_rate": 8.337950045726051e-07,
+      "loss": 0.4268,
+      "step": 5315
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6143754571937265,
+      "learning_rate": 8.316694968762696e-07,
+      "loss": 0.4387,
+      "step": 5316
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7205627901751872,
+      "learning_rate": 8.295465842565942e-07,
+      "loss": 0.5207,
+      "step": 5317
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6101000457239085,
+      "learning_rate": 8.274262673144651e-07,
+      "loss": 0.4012,
+      "step": 5318
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8631517802030988,
+      "learning_rate": 8.253085466500332e-07,
+      "loss": 0.4167,
+      "step": 5319
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8617712550174828,
+      "learning_rate": 8.231934228627158e-07,
+      "loss": 0.4654,
+      "step": 5320
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7406993206254793,
+      "learning_rate": 8.210808965511941e-07,
+      "loss": 0.5538,
+      "step": 5321
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.688562905965698,
+      "learning_rate": 8.189709683134139e-07,
+      "loss": 0.6148,
+      "step": 5322
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7486525219805609,
+      "learning_rate": 8.168636387465856e-07,
+      "loss": 0.5031,
+      "step": 5323
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4878900888381195,
+      "learning_rate": 8.147589084471851e-07,
+      "loss": 0.4103,
+      "step": 5324
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5238706559820416,
+      "learning_rate": 8.126567780109506e-07,
+      "loss": 0.469,
+      "step": 5325
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.832689522719165,
+      "learning_rate": 8.105572480328872e-07,
+      "loss": 0.562,
+      "step": 5326
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8677477275617091,
+      "learning_rate": 8.084603191072615e-07,
+      "loss": 0.6159,
+      "step": 5327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.558336188904265,
+      "learning_rate": 8.063659918276056e-07,
+      "loss": 0.443,
+      "step": 5328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5927317732182045,
+      "learning_rate": 8.042742667867143e-07,
+      "loss": 0.4524,
+      "step": 5329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.80662189687871,
+      "learning_rate": 8.02185144576646e-07,
+      "loss": 0.4577,
+      "step": 5330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8845391332655346,
+      "learning_rate": 8.000986257887211e-07,
+      "loss": 0.529,
+      "step": 5331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.971951297035259,
+      "learning_rate": 7.98014711013525e-07,
+      "loss": 0.6146,
+      "step": 5332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8362163835859945,
+      "learning_rate": 7.95933400840907e-07,
+      "loss": 0.6721,
+      "step": 5333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7415461370416703,
+      "learning_rate": 7.938546958599747e-07,
+      "loss": 0.5372,
+      "step": 5334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1355715060404474,
+      "learning_rate": 7.91778596659104e-07,
+      "loss": 0.5074,
+      "step": 5335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.189742693690816,
+      "learning_rate": 7.897051038259285e-07,
+      "loss": 0.5383,
+      "step": 5336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.658342078186364,
+      "learning_rate": 7.876342179473473e-07,
+      "loss": 0.5923,
+      "step": 5337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.84626005714495,
+      "learning_rate": 7.855659396095183e-07,
+      "loss": 0.5081,
+      "step": 5338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5803540975414554,
+      "learning_rate": 7.835002693978655e-07,
+      "loss": 0.426,
+      "step": 5339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0851633855825566,
+      "learning_rate": 7.814372078970711e-07,
+      "loss": 0.4834,
+      "step": 5340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.833877657297366,
+      "learning_rate": 7.793767556910814e-07,
+      "loss": 0.511,
+      "step": 5341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6776541830210088,
+      "learning_rate": 7.773189133631032e-07,
+      "loss": 0.4503,
+      "step": 5342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.6452117906352095,
+      "learning_rate": 7.752636814956027e-07,
+      "loss": 0.4768,
+      "step": 5343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.592449822629031,
+      "learning_rate": 7.732110606703103e-07,
+      "loss": 0.6107,
+      "step": 5344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8240343412526525,
+      "learning_rate": 7.711610514682155e-07,
+      "loss": 0.6039,
+      "step": 5345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9280047948855819,
+      "learning_rate": 7.691136544695699e-07,
+      "loss": 0.5112,
+      "step": 5346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1592446586247354,
+      "learning_rate": 7.670688702538831e-07,
+      "loss": 0.4893,
+      "step": 5347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6416861593202112,
+      "learning_rate": 7.6502669939993e-07,
+      "loss": 0.5266,
+      "step": 5348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9464951944432747,
+      "learning_rate": 7.629871424857394e-07,
+      "loss": 0.4888,
+      "step": 5349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6683099534336128,
+      "learning_rate": 7.60950200088606e-07,
+      "loss": 0.5118,
+      "step": 5350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0595497562326113,
+      "learning_rate": 7.58915872785081e-07,
+      "loss": 0.485,
+      "step": 5351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.001759366171111,
+      "learning_rate": 7.56884161150977e-07,
+      "loss": 0.5051,
+      "step": 5352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.38308920656847,
+      "learning_rate": 7.548550657613651e-07,
+      "loss": 0.5405,
+      "step": 5353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3621802050339555,
+      "learning_rate": 7.52828587190576e-07,
+      "loss": 0.4993,
+      "step": 5354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8895450838758623,
+      "learning_rate": 7.50804726012202e-07,
+      "loss": 0.5604,
+      "step": 5355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.586049347282142,
+      "learning_rate": 7.487834827990915e-07,
+      "loss": 0.3744,
+      "step": 5356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6255950562497592,
+      "learning_rate": 7.467648581233533e-07,
+      "loss": 0.498,
+      "step": 5357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4126469696561712,
+      "learning_rate": 7.447488525563551e-07,
+      "loss": 0.4276,
+      "step": 5358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0358818862625365,
+      "learning_rate": 7.427354666687226e-07,
+      "loss": 0.5388,
+      "step": 5359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6955014608532497,
+      "learning_rate": 7.407247010303409e-07,
+      "loss": 0.5106,
+      "step": 5360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.427713014276837,
+      "learning_rate": 7.387165562103526e-07,
+      "loss": 0.5716,
+      "step": 5361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.742362911780383,
+      "learning_rate": 7.367110327771587e-07,
+      "loss": 0.5466,
+      "step": 5362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.826515175340725,
+      "learning_rate": 7.347081312984194e-07,
+      "loss": 0.6959,
+      "step": 5363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.606673045189989,
+      "learning_rate": 7.327078523410525e-07,
+      "loss": 0.5987,
+      "step": 5364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6221345987446592,
+      "learning_rate": 7.307101964712271e-07,
+      "loss": 0.5323,
+      "step": 5365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.934706036433136,
+      "learning_rate": 7.287151642543822e-07,
+      "loss": 0.547,
+      "step": 5366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9931753809730988,
+      "learning_rate": 7.267227562552048e-07,
+      "loss": 0.4808,
+      "step": 5367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.193308405613236,
+      "learning_rate": 7.247329730376429e-07,
+      "loss": 0.5557,
+      "step": 5368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9327478878437128,
+      "learning_rate": 7.227458151648992e-07,
+      "loss": 0.5507,
+      "step": 5369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4580166615877204,
+      "learning_rate": 7.207612831994337e-07,
+      "loss": 0.5762,
+      "step": 5370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6603708712513139,
+      "learning_rate": 7.187793777029661e-07,
+      "loss": 0.4423,
+      "step": 5371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.989665193738245,
+      "learning_rate": 7.168000992364699e-07,
+      "loss": 0.5108,
+      "step": 5372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.031688155445486,
+      "learning_rate": 7.148234483601746e-07,
+      "loss": 0.5648,
+      "step": 5373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4723938322497405,
+      "learning_rate": 7.128494256335694e-07,
+      "loss": 0.5166,
+      "step": 5374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.214679068859399,
+      "learning_rate": 7.108780316153951e-07,
+      "loss": 0.5911,
+      "step": 5375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7382051041774333,
+      "learning_rate": 7.08909266863651e-07,
+      "loss": 0.5497,
+      "step": 5376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8347194585062205,
+      "learning_rate": 7.069431319355924e-07,
+      "loss": 0.5034,
+      "step": 5377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.382621950668795,
+      "learning_rate": 7.049796273877297e-07,
+      "loss": 0.4834,
+      "step": 5378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.955842734436443,
+      "learning_rate": 7.030187537758282e-07,
+      "loss": 0.6084,
+      "step": 5379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5112777143163436,
+      "learning_rate": 7.010605116549085e-07,
+      "loss": 0.5019,
+      "step": 5380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1291816092671514,
+      "learning_rate": 6.991049015792483e-07,
+      "loss": 0.5534,
+      "step": 5381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0657944158015757,
+      "learning_rate": 6.971519241023795e-07,
+      "loss": 0.5068,
+      "step": 5382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8690281719285682,
+      "learning_rate": 6.952015797770862e-07,
+      "loss": 0.5363,
+      "step": 5383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8711731293771976,
+      "learning_rate": 6.932538691554103e-07,
+      "loss": 0.4706,
+      "step": 5384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2523210013320267,
+      "learning_rate": 6.913087927886464e-07,
+      "loss": 0.5782,
+      "step": 5385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.630972160655108,
+      "learning_rate": 6.893663512273474e-07,
+      "loss": 0.4784,
+      "step": 5386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9043910291761166,
+      "learning_rate": 6.874265450213124e-07,
+      "loss": 0.4469,
+      "step": 5387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6047705550720108,
+      "learning_rate": 6.854893747196034e-07,
+      "loss": 0.5104,
+      "step": 5388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8060261005771412,
+      "learning_rate": 6.835548408705284e-07,
+      "loss": 0.4876,
+      "step": 5389
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6084050740447393,
+      "learning_rate": 6.816229440216571e-07,
+      "loss": 0.3958,
+      "step": 5390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6221745221449155,
+      "learning_rate": 6.796936847198066e-07,
+      "loss": 0.5324,
+      "step": 5391
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6341119834743885,
+      "learning_rate": 6.777670635110523e-07,
+      "loss": 0.5297,
+      "step": 5392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.71207480637078,
+      "learning_rate": 6.758430809407169e-07,
+      "loss": 0.5325,
+      "step": 5393
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.160759474780141,
+      "learning_rate": 6.739217375533813e-07,
+      "loss": 0.5981,
+      "step": 5394
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7155233291002396,
+      "learning_rate": 6.720030338928785e-07,
+      "loss": 0.5519,
+      "step": 5395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.153020589110243,
+      "learning_rate": 6.700869705022916e-07,
+      "loss": 0.51,
+      "step": 5396
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9077729031749033,
+      "learning_rate": 6.681735479239593e-07,
+      "loss": 0.5815,
+      "step": 5397
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.210278391803528,
+      "learning_rate": 6.662627666994725e-07,
+      "loss": 0.5374,
+      "step": 5398
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8140199913680666,
+      "learning_rate": 6.643546273696732e-07,
+      "loss": 0.5176,
+      "step": 5399
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.593855802308051,
+      "learning_rate": 6.624491304746561e-07,
+      "loss": 0.5277,
+      "step": 5400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7788031734779763,
+      "learning_rate": 6.605462765537674e-07,
+      "loss": 0.5151,
+      "step": 5401
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9356811447980207,
+      "learning_rate": 6.586460661456074e-07,
+      "loss": 0.5193,
+      "step": 5402
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6604127891129614,
+      "learning_rate": 6.567484997880247e-07,
+      "loss": 0.509,
+      "step": 5403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0733473533714966,
+      "learning_rate": 6.548535780181242e-07,
+      "loss": 0.5028,
+      "step": 5404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.709616704534783,
+      "learning_rate": 6.529613013722568e-07,
+      "loss": 0.5069,
+      "step": 5405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6608701550003162,
+      "learning_rate": 6.510716703860298e-07,
+      "loss": 0.5224,
+      "step": 5406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.393087426807728,
+      "learning_rate": 6.491846855942984e-07,
+      "loss": 0.5252,
+      "step": 5407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5144912009457663,
+      "learning_rate": 6.47300347531169e-07,
+      "loss": 0.4424,
+      "step": 5408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5319355328154383,
+      "learning_rate": 6.454186567299992e-07,
+      "loss": 0.5199,
+      "step": 5409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7952159055665065,
+      "learning_rate": 6.435396137233985e-07,
+      "loss": 0.6443,
+      "step": 5410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5635385875348329,
+      "learning_rate": 6.416632190432259e-07,
+      "loss": 0.3673,
+      "step": 5411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.602378202629224,
+      "learning_rate": 6.397894732205889e-07,
+      "loss": 0.4961,
+      "step": 5412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.80149707269951,
+      "learning_rate": 6.37918376785851e-07,
+      "loss": 0.4908,
+      "step": 5413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0436292491641486,
+      "learning_rate": 6.360499302686207e-07,
+      "loss": 0.5246,
+      "step": 5414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6821650697392168,
+      "learning_rate": 6.341841341977584e-07,
+      "loss": 0.579,
+      "step": 5415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8154347098846444,
+      "learning_rate": 6.323209891013715e-07,
+      "loss": 0.5611,
+      "step": 5416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8794708176086857,
+      "learning_rate": 6.304604955068216e-07,
+      "loss": 0.6687,
+      "step": 5417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.555262523392295,
+      "learning_rate": 6.286026539407164e-07,
+      "loss": 0.4749,
+      "step": 5418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9378997705623406,
+      "learning_rate": 6.267474649289152e-07,
+      "loss": 0.4966,
+      "step": 5419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.051389328612246,
+      "learning_rate": 6.248949289965234e-07,
+      "loss": 0.5252,
+      "step": 5420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0385796396251994,
+      "learning_rate": 6.230450466678995e-07,
+      "loss": 0.5882,
+      "step": 5421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.681684352479866,
+      "learning_rate": 6.211978184666468e-07,
+      "loss": 0.4725,
+      "step": 5422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.885416161391408,
+      "learning_rate": 6.193532449156203e-07,
+      "loss": 0.5172,
+      "step": 5423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.708843656796631,
+      "learning_rate": 6.175113265369237e-07,
+      "loss": 0.5448,
+      "step": 5424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2660579050610585,
+      "learning_rate": 6.156720638519054e-07,
+      "loss": 0.5316,
+      "step": 5425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.085066366878838,
+      "learning_rate": 6.138354573811678e-07,
+      "loss": 0.5336,
+      "step": 5426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.746861686063354,
+      "learning_rate": 6.120015076445573e-07,
+      "loss": 0.5473,
+      "step": 5427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.596916276500172,
+      "learning_rate": 6.101702151611688e-07,
+      "loss": 0.4391,
+      "step": 5428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.127852313619974,
+      "learning_rate": 6.083415804493487e-07,
+      "loss": 0.5427,
+      "step": 5429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8424357879460207,
+      "learning_rate": 6.06515604026684e-07,
+      "loss": 0.5136,
+      "step": 5430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.956970869581323,
+      "learning_rate": 6.046922864100158e-07,
+      "loss": 0.4561,
+      "step": 5431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6993587718250447,
+      "learning_rate": 6.02871628115429e-07,
+      "loss": 0.5392,
+      "step": 5432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.859755507481992,
+      "learning_rate": 6.010536296582592e-07,
+      "loss": 0.5281,
+      "step": 5433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7848401161749068,
+      "learning_rate": 5.992382915530848e-07,
+      "loss": 0.5026,
+      "step": 5434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6118331801005796,
+      "learning_rate": 5.974256143137335e-07,
+      "loss": 0.4735,
+      "step": 5435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5630477999520747,
+      "learning_rate": 5.95615598453283e-07,
+      "loss": 0.4496,
+      "step": 5436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8604408332156526,
+      "learning_rate": 5.938082444840521e-07,
+      "loss": 0.4847,
+      "step": 5437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3358933698341224,
+      "learning_rate": 5.920035529176082e-07,
+      "loss": 0.4567,
+      "step": 5438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.389043948150227,
+      "learning_rate": 5.902015242647651e-07,
+      "loss": 0.4755,
+      "step": 5439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.953536506021114,
+      "learning_rate": 5.884021590355859e-07,
+      "loss": 0.5265,
+      "step": 5440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7782773899922157,
+      "learning_rate": 5.866054577393742e-07,
+      "loss": 0.4807,
+      "step": 5441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3451091734957,
+      "learning_rate": 5.848114208846834e-07,
+      "loss": 0.5058,
+      "step": 5442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9351081850853404,
+      "learning_rate": 5.830200489793136e-07,
+      "loss": 0.6106,
+      "step": 5443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6181850962036872,
+      "learning_rate": 5.812313425303062e-07,
+      "loss": 0.5104,
+      "step": 5444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5411046480835573,
+      "learning_rate": 5.794453020439517e-07,
+      "loss": 0.5759,
+      "step": 5445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5677883328204243,
+      "learning_rate": 5.776619280257855e-07,
+      "loss": 0.5427,
+      "step": 5446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8806280920649232,
+      "learning_rate": 5.758812209805887e-07,
+      "loss": 0.6145,
+      "step": 5447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9250974013224746,
+      "learning_rate": 5.741031814123843e-07,
+      "loss": 0.4347,
+      "step": 5448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0649664495225735,
+      "learning_rate": 5.723278098244455e-07,
+      "loss": 0.5444,
+      "step": 5449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7844849236028417,
+      "learning_rate": 5.705551067192871e-07,
+      "loss": 0.5574,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1678059969339705,
+      "learning_rate": 5.687850725986654e-07,
+      "loss": 0.4959,
+      "step": 5451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9244080765000289,
+      "learning_rate": 5.670177079635886e-07,
+      "loss": 0.5977,
+      "step": 5452
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6991182100101492,
+      "learning_rate": 5.652530133143042e-07,
+      "loss": 0.5049,
+      "step": 5453
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.646226316633901,
+      "learning_rate": 5.634909891503048e-07,
+      "loss": 0.5287,
+      "step": 5454
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.339778778583666,
+      "learning_rate": 5.617316359703284e-07,
+      "loss": 0.4825,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6381948213913933,
+      "learning_rate": 5.599749542723565e-07,
+      "loss": 0.4819,
+      "step": 5456
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8419798675812633,
+      "learning_rate": 5.582209445536135e-07,
+      "loss": 0.527,
+      "step": 5457
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.820462337869772,
+      "learning_rate": 5.564696073105669e-07,
+      "loss": 0.5301,
+      "step": 5458
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6325865693016453,
+      "learning_rate": 5.547209430389322e-07,
+      "loss": 0.5026,
+      "step": 5459
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9392881454065862,
+      "learning_rate": 5.529749522336625e-07,
+      "loss": 0.481,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6437087735628504,
+      "learning_rate": 5.512316353889591e-07,
+      "loss": 0.4342,
+      "step": 5461
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7727986370687553,
+      "learning_rate": 5.494909929982617e-07,
+      "loss": 0.6165,
+      "step": 5462
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.155581624308307,
+      "learning_rate": 5.477530255542573e-07,
+      "loss": 0.536,
+      "step": 5463
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1837918087756765,
+      "learning_rate": 5.460177335488736e-07,
+      "loss": 0.515,
+      "step": 5464
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.806801989880357,
+      "learning_rate": 5.442851174732799e-07,
+      "loss": 0.4649,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0922672360545276,
+      "learning_rate": 5.425551778178917e-07,
+      "loss": 0.5476,
+      "step": 5466
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7487106930414764,
+      "learning_rate": 5.408279150723628e-07,
+      "loss": 0.6097,
+      "step": 5467
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.030707475119805,
+      "learning_rate": 5.391033297255932e-07,
+      "loss": 0.4931,
+      "step": 5468
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8220176648365916,
+      "learning_rate": 5.373814222657214e-07,
+      "loss": 0.5166,
+      "step": 5469
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7998599189868485,
+      "learning_rate": 5.356621931801309e-07,
+      "loss": 0.5775,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0085398945970616,
+      "learning_rate": 5.339456429554446e-07,
+      "loss": 0.5716,
+      "step": 5471
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8631104089111434,
+      "learning_rate": 5.322317720775316e-07,
+      "loss": 0.4365,
+      "step": 5472
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8274556874937138,
+      "learning_rate": 5.305205810314951e-07,
+      "loss": 0.4296,
+      "step": 5473
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4362644562531,
+      "learning_rate": 5.288120703016863e-07,
+      "loss": 0.4705,
+      "step": 5474
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8193550739946693,
+      "learning_rate": 5.271062403716953e-07,
+      "loss": 0.5435,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.396314248734287,
+      "learning_rate": 5.254030917243535e-07,
+      "loss": 0.5817,
+      "step": 5476
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7334765661261313,
+      "learning_rate": 5.237026248417343e-07,
+      "loss": 0.5053,
+      "step": 5477
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.51028067734458,
+      "learning_rate": 5.220048402051503e-07,
+      "loss": 0.4156,
+      "step": 5478
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5416704962205503,
+      "learning_rate": 5.203097382951572e-07,
+      "loss": 0.4281,
+      "step": 5479
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.097045241334041,
+      "learning_rate": 5.186173195915478e-07,
+      "loss": 0.567,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8295064843228168,
+      "learning_rate": 5.169275845733601e-07,
+      "loss": 0.5599,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6583305741021612,
+      "learning_rate": 5.15240533718867e-07,
+      "loss": 0.4677,
+      "step": 5482
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.778491200236851,
+      "learning_rate": 5.135561675055889e-07,
+      "loss": 0.5663,
+      "step": 5483
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.092199781743835,
+      "learning_rate": 5.118744864102787e-07,
+      "loss": 0.5105,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9922804620273344,
+      "learning_rate": 5.101954909089346e-07,
+      "loss": 0.5141,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.494537958768569,
+      "learning_rate": 5.08519181476792e-07,
+      "loss": 0.6082,
+      "step": 5486
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6339695856932839,
+      "learning_rate": 5.06845558588327e-07,
+      "loss": 0.4996,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.407747469687187,
+      "learning_rate": 5.051746227172538e-07,
+      "loss": 0.4993,
+      "step": 5488
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3349905701564326,
+      "learning_rate": 5.035063743365299e-07,
+      "loss": 0.4859,
+      "step": 5489
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9462879942799758,
+      "learning_rate": 5.018408139183462e-07,
+      "loss": 0.5295,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8455185530727872,
+      "learning_rate": 5.001779419341391e-07,
+      "loss": 0.6616,
+      "step": 5491
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.058220724281517,
+      "learning_rate": 4.985177588545786e-07,
+      "loss": 0.5732,
+      "step": 5492
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8733386760807167,
+      "learning_rate": 4.96860265149578e-07,
+      "loss": 0.5732,
+      "step": 5493
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1923482858943735,
+      "learning_rate": 4.952054612882873e-07,
+      "loss": 0.5186,
+      "step": 5494
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7830985137779314,
+      "learning_rate": 4.93553347739093e-07,
+      "loss": 0.5182,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.138579318896787,
+      "learning_rate": 4.919039249696233e-07,
+      "loss": 0.5083,
+      "step": 5496
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.260112837998369,
+      "learning_rate": 4.90257193446746e-07,
+      "loss": 0.6124,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6665266210208696,
+      "learning_rate": 4.886131536365623e-07,
+      "loss": 0.4615,
+      "step": 5498
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1542056410864885,
+      "learning_rate": 4.869718060044148e-07,
+      "loss": 0.5065,
+      "step": 5499
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2903191957307127,
+      "learning_rate": 4.853331510148851e-07,
+      "loss": 0.5075,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0844566703088407,
+      "learning_rate": 4.836971891317898e-07,
+      "loss": 0.4334,
+      "step": 5501
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9519011247799045,
+      "learning_rate": 4.820639208181832e-07,
+      "loss": 0.523,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.018413833299174,
+      "learning_rate": 4.804333465363609e-07,
+      "loss": 0.4904,
+      "step": 5503
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8008759380904364,
+      "learning_rate": 4.788054667478526e-07,
+      "loss": 0.4093,
+      "step": 5504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.833976677128723,
+      "learning_rate": 4.771802819134253e-07,
+      "loss": 0.6672,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.582876342271089,
+      "learning_rate": 4.7555779249308545e-07,
+      "loss": 0.5705,
+      "step": 5506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8648545340887331,
+      "learning_rate": 4.739379989460746e-07,
+      "loss": 0.6868,
+      "step": 5507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.041653320755163,
+      "learning_rate": 4.723209017308727e-07,
+      "loss": 0.4551,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4988726817066682,
+      "learning_rate": 4.707065013051948e-07,
+      "loss": 0.5254,
+      "step": 5509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8241186068141817,
+      "learning_rate": 4.69094798125993e-07,
+      "loss": 0.5395,
+      "step": 5510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7386793557022002,
+      "learning_rate": 4.6748579264945806e-07,
+      "loss": 0.532,
+      "step": 5511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5254280669633937,
+      "learning_rate": 4.658794853310156e-07,
+      "loss": 0.5159,
+      "step": 5512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5765166677315572,
+      "learning_rate": 4.6427587662532636e-07,
+      "loss": 0.364,
+      "step": 5513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.042494746006534,
+      "learning_rate": 4.6267496698628846e-07,
+      "loss": 0.4728,
+      "step": 5514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.782064634237988,
+      "learning_rate": 4.6107675686703715e-07,
+      "loss": 0.535,
+      "step": 5515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6684911333367665,
+      "learning_rate": 4.5948124671994164e-07,
+      "loss": 0.4918,
+      "step": 5516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.05712337819604,
+      "learning_rate": 4.5788843699660745e-07,
+      "loss": 0.6678,
+      "step": 5517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7268008880614005,
+      "learning_rate": 4.562983281478761e-07,
+      "loss": 0.589,
+      "step": 5518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0780818365385083,
+      "learning_rate": 4.547109206238243e-07,
+      "loss": 0.3836,
+      "step": 5519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.89552245406834,
+      "learning_rate": 4.5312621487376585e-07,
+      "loss": 0.5277,
+      "step": 5520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6383246639182985,
+      "learning_rate": 4.515442113462465e-07,
+      "loss": 0.4596,
+      "step": 5521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0979601956128975,
+      "learning_rate": 4.499649104890502e-07,
+      "loss": 0.4812,
+      "step": 5522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.954691899807524,
+      "learning_rate": 4.4838831274919505e-07,
+      "loss": 0.5445,
+      "step": 5523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.700503975365014,
+      "learning_rate": 4.468144185729328e-07,
+      "loss": 0.5246,
+      "step": 5524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0140063678502953,
+      "learning_rate": 4.4524322840575175e-07,
+      "loss": 0.4577,
+      "step": 5525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5497913158744379,
+      "learning_rate": 4.4367474269237267e-07,
+      "loss": 0.451,
+      "step": 5526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4700819204913502,
+      "learning_rate": 4.4210896187675266e-07,
+      "loss": 0.3839,
+      "step": 5527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6031805869488278,
+      "learning_rate": 4.4054588640208285e-07,
+      "loss": 0.5362,
+      "step": 5528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.887948356405248,
+      "learning_rate": 4.3898551671078926e-07,
+      "loss": 0.4675,
+      "step": 5529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8612170495874447,
+      "learning_rate": 4.3742785324453086e-07,
+      "loss": 0.5181,
+      "step": 5530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3855648564492027,
+      "learning_rate": 4.358728964442005e-07,
+      "loss": 0.5765,
+      "step": 5531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5856911962043025,
+      "learning_rate": 4.34320646749925e-07,
+      "loss": 0.5496,
+      "step": 5532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6620395821544467,
+      "learning_rate": 4.327711046010663e-07,
+      "loss": 0.5092,
+      "step": 5533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6056072201154907,
+      "learning_rate": 4.3122427043621905e-07,
+      "loss": 0.4929,
+      "step": 5534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7734318369040638,
+      "learning_rate": 4.2968014469321194e-07,
+      "loss": 0.4772,
+      "step": 5535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5831910145683477,
+      "learning_rate": 4.2813872780910425e-07,
+      "loss": 0.5596,
+      "step": 5536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.892979810157702,
+      "learning_rate": 4.266000202201948e-07,
+      "loss": 0.4885,
+      "step": 5537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.432385088935785,
+      "learning_rate": 4.2506402236200616e-07,
+      "loss": 0.5124,
+      "step": 5538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7943844896579842,
+      "learning_rate": 4.2353073466930404e-07,
+      "loss": 0.6415,
+      "step": 5539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.032948817636901,
+      "learning_rate": 4.2200015757607905e-07,
+      "loss": 0.4569,
+      "step": 5540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4935812974029716,
+      "learning_rate": 4.20472291515559e-07,
+      "loss": 0.4345,
+      "step": 5541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.171577322540812,
+      "learning_rate": 4.189471369202036e-07,
+      "loss": 0.53,
+      "step": 5542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9293797225368792,
+      "learning_rate": 4.1742469422170417e-07,
+      "loss": 0.4945,
+      "step": 5543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6368575547536892,
+      "learning_rate": 4.15904963850986e-07,
+      "loss": 0.4638,
+      "step": 5544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7524597615556021,
+      "learning_rate": 4.143879462382039e-07,
+      "loss": 0.5096,
+      "step": 5545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1715153478969205,
+      "learning_rate": 4.128736418127477e-07,
+      "loss": 0.4535,
+      "step": 5546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1082635335677993,
+      "learning_rate": 4.113620510032368e-07,
+      "loss": 0.4389,
+      "step": 5547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.637298459235742,
+      "learning_rate": 4.0985317423752557e-07,
+      "loss": 0.5048,
+      "step": 5548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.633860303923681,
+      "learning_rate": 4.083470119426969e-07,
+      "loss": 0.5722,
+      "step": 5549
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5763248228282514,
+      "learning_rate": 4.0684356454506747e-07,
+      "loss": 0.4272,
+      "step": 5550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6522954387380382,
+      "learning_rate": 4.053428324701836e-07,
+      "loss": 0.4902,
+      "step": 5551
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9537808833703756,
+      "learning_rate": 4.0384481614282764e-07,
+      "loss": 0.5418,
+      "step": 5552
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.778741529668083,
+      "learning_rate": 4.0234951598700725e-07,
+      "loss": 0.5184,
+      "step": 5553
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.653168972740757,
+      "learning_rate": 4.008569324259648e-07,
+      "loss": 0.5685,
+      "step": 5554
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.809000034029341,
+      "learning_rate": 3.9936706588217243e-07,
+      "loss": 0.6556,
+      "step": 5555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6619250105088161,
+      "learning_rate": 3.9787991677733397e-07,
+      "loss": 0.5006,
+      "step": 5556
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7394853006675532,
+      "learning_rate": 3.9639548553238483e-07,
+      "loss": 0.5417,
+      "step": 5557
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9233091679584446,
+      "learning_rate": 3.94913772567489e-07,
+      "loss": 0.5558,
+      "step": 5558
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.813501375789264,
+      "learning_rate": 3.9343477830204424e-07,
+      "loss": 0.4918,
+      "step": 5559
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.162774094460681,
+      "learning_rate": 3.9195850315467244e-07,
+      "loss": 0.5251,
+      "step": 5560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.098464669508243,
+      "learning_rate": 3.904849475432337e-07,
+      "loss": 0.532,
+      "step": 5561
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.941872033049104,
+      "learning_rate": 3.8901411188481453e-07,
+      "loss": 0.6112,
+      "step": 5562
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 22.927106066859036,
+      "learning_rate": 3.875459965957307e-07,
+      "loss": 0.4826,
+      "step": 5563
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9958532424456348,
+      "learning_rate": 3.860806020915286e-07,
+      "loss": 0.5343,
+      "step": 5564
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8255715031396351,
+      "learning_rate": 3.846179287869878e-07,
+      "loss": 0.4927,
+      "step": 5565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7632629559132088,
+      "learning_rate": 3.831579770961125e-07,
+      "loss": 0.6075,
+      "step": 5566
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3147310706089375,
+      "learning_rate": 3.817007474321399e-07,
+      "loss": 0.4376,
+      "step": 5567
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.056378589973051,
+      "learning_rate": 3.802462402075358e-07,
+      "loss": 0.5254,
+      "step": 5568
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.617772143279416,
+      "learning_rate": 3.787944558339951e-07,
+      "loss": 0.4424,
+      "step": 5569
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1939939271834183,
+      "learning_rate": 3.773453947224426e-07,
+      "loss": 0.5187,
+      "step": 5570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.956958578245797,
+      "learning_rate": 3.7589905728303123e-07,
+      "loss": 0.5497,
+      "step": 5571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.778566715127792,
+      "learning_rate": 3.7445544392514465e-07,
+      "loss": 0.4879,
+      "step": 5572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6535265026507302,
+      "learning_rate": 3.7301455505739494e-07,
+      "loss": 0.5247,
+      "step": 5573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5589484126552224,
+      "learning_rate": 3.7157639108762136e-07,
+      "loss": 0.5096,
+      "step": 5574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8907137237240486,
+      "learning_rate": 3.701409524228927e-07,
+      "loss": 0.5386,
+      "step": 5575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9439111831138112,
+      "learning_rate": 3.687082394695096e-07,
+      "loss": 0.5031,
+      "step": 5576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4844103495512235,
+      "learning_rate": 3.6727825263299656e-07,
+      "loss": 0.5069,
+      "step": 5577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8045519557126708,
+      "learning_rate": 3.6585099231810863e-07,
+      "loss": 0.5073,
+      "step": 5578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9982728512919428,
+      "learning_rate": 3.644264589288282e-07,
+      "loss": 0.5896,
+      "step": 5579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.544418153605649,
+      "learning_rate": 3.630046528683695e-07,
+      "loss": 0.4519,
+      "step": 5580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6203013835068178,
+      "learning_rate": 3.615855745391683e-07,
+      "loss": 0.5729,
+      "step": 5581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.95692316772664,
+      "learning_rate": 3.6016922434289113e-07,
+      "loss": 0.4322,
+      "step": 5582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8849142803118357,
+      "learning_rate": 3.587556026804362e-07,
+      "loss": 0.4973,
+      "step": 5583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.121672764948359,
+      "learning_rate": 3.573447099519245e-07,
+      "loss": 0.5307,
+      "step": 5584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0670949191673,
+      "learning_rate": 3.559365465567055e-07,
+      "loss": 0.4759,
+      "step": 5585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9541272813893749,
+      "learning_rate": 3.545311128933582e-07,
+      "loss": 0.4493,
+      "step": 5586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9224704823253531,
+      "learning_rate": 3.531284093596865e-07,
+      "loss": 0.5409,
+      "step": 5587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.152940274184901,
+      "learning_rate": 3.5172843635272403e-07,
+      "loss": 0.4631,
+      "step": 5588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.050467848480177,
+      "learning_rate": 3.503311942687293e-07,
+      "loss": 0.4661,
+      "step": 5589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7805195563331617,
+      "learning_rate": 3.489366835031882e-07,
+      "loss": 0.4063,
+      "step": 5590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6160048609347182,
+      "learning_rate": 3.47544904450815e-07,
+      "loss": 0.517,
+      "step": 5591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3393254205560954,
+      "learning_rate": 3.4615585750555016e-07,
+      "loss": 0.4856,
+      "step": 5592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6517630752575734,
+      "learning_rate": 3.4476954306056023e-07,
+      "loss": 0.5455,
+      "step": 5593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7460034493424519,
+      "learning_rate": 3.43385961508238e-07,
+      "loss": 0.5977,
+      "step": 5594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8839885815707271,
+      "learning_rate": 3.420051132402036e-07,
+      "loss": 0.5537,
+      "step": 5595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1794853146887125,
+      "learning_rate": 3.406269986473032e-07,
+      "loss": 0.5482,
+      "step": 5596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.572296485898109,
+      "learning_rate": 3.392516181196093e-07,
+      "loss": 0.4812,
+      "step": 5597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8857899175029467,
+      "learning_rate": 3.378789720464193e-07,
+      "loss": 0.4833,
+      "step": 5598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.452259550281556,
+      "learning_rate": 3.365090608162591e-07,
+      "loss": 0.5284,
+      "step": 5599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4727400055952322,
+      "learning_rate": 3.351418848168808e-07,
+      "loss": 0.463,
+      "step": 5600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4281582985626238,
+      "learning_rate": 3.3377744443525816e-07,
+      "loss": 0.5539,
+      "step": 5601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.23792780422691,
+      "learning_rate": 3.324157400575945e-07,
+      "loss": 0.454,
+      "step": 5602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7688576413136963,
+      "learning_rate": 3.31056772069317e-07,
+      "loss": 0.4243,
+      "step": 5603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8496667815258787,
+      "learning_rate": 3.2970054085507795e-07,
+      "loss": 0.4483,
+      "step": 5604
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0115254035848356,
+      "learning_rate": 3.2834704679875596e-07,
+      "loss": 0.5753,
+      "step": 5605
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8939506564016722,
+      "learning_rate": 3.269962902834545e-07,
+      "loss": 0.4633,
+      "step": 5606
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4087334748925102,
+      "learning_rate": 3.256482716915044e-07,
+      "loss": 0.4545,
+      "step": 5607
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4742038955905254,
+      "learning_rate": 3.2430299140445597e-07,
+      "loss": 0.5314,
+      "step": 5608
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9049964882633597,
+      "learning_rate": 3.229604498030914e-07,
+      "loss": 0.6021,
+      "step": 5609
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3263233081313563,
+      "learning_rate": 3.216206472674122e-07,
+      "loss": 0.4542,
+      "step": 5610
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0680344273328752,
+      "learning_rate": 3.202835841766483e-07,
+      "loss": 0.57,
+      "step": 5611
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.117439303900396,
+      "learning_rate": 3.1894926090925037e-07,
+      "loss": 0.4974,
+      "step": 5612
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7175153033339705,
+      "learning_rate": 3.176176778428974e-07,
+      "loss": 0.4687,
+      "step": 5613
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.051808349014429,
+      "learning_rate": 3.1628883535449127e-07,
+      "loss": 0.4744,
+      "step": 5614
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3936775600641025,
+      "learning_rate": 3.149627338201566e-07,
+      "loss": 0.5803,
+      "step": 5615
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8134783521452127,
+      "learning_rate": 3.1363937361524545e-07,
+      "loss": 0.5799,
+      "step": 5616
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.487988523987755,
+      "learning_rate": 3.123187551143314e-07,
+      "loss": 0.6053,
+      "step": 5617
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6984205599805344,
+      "learning_rate": 3.110008786912122e-07,
+      "loss": 0.5126,
+      "step": 5618
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.272631572260754,
+      "learning_rate": 3.0968574471891057e-07,
+      "loss": 0.542,
+      "step": 5619
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9044958448447087,
+      "learning_rate": 3.08373353569672e-07,
+      "loss": 0.4604,
+      "step": 5620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.860191914303617,
+      "learning_rate": 3.07063705614965e-07,
+      "loss": 0.4806,
+      "step": 5621
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0739500647590736,
+      "learning_rate": 3.0575680122548525e-07,
+      "loss": 0.5661,
+      "step": 5622
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9714246411792535,
+      "learning_rate": 3.044526407711501e-07,
+      "loss": 0.4993,
+      "step": 5623
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2365108561865887,
+      "learning_rate": 3.031512246210955e-07,
+      "loss": 0.5227,
+      "step": 5624
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8056259438682738,
+      "learning_rate": 3.0185255314368555e-07,
+      "loss": 0.4453,
+      "step": 5625
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5917801728469472,
+      "learning_rate": 3.005566267065085e-07,
+      "loss": 0.536,
+      "step": 5626
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9510512797295725,
+      "learning_rate": 2.992634456763721e-07,
+      "loss": 0.6083,
+      "step": 5627
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7569929976939416,
+      "learning_rate": 2.97973010419309e-07,
+      "loss": 0.5015,
+      "step": 5628
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.47564054110847,
+      "learning_rate": 2.9668532130057384e-07,
+      "loss": 0.5957,
+      "step": 5629
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.882788708693888,
+      "learning_rate": 2.954003786846449e-07,
+      "loss": 0.5011,
+      "step": 5630
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7454876444377672,
+      "learning_rate": 2.941181829352213e-07,
+      "loss": 0.5264,
+      "step": 5631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.720837136711613,
+      "learning_rate": 2.9283873441522723e-07,
+      "loss": 0.5307,
+      "step": 5632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.345199934138379,
+      "learning_rate": 2.915620334868074e-07,
+      "loss": 0.5835,
+      "step": 5633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.180448449994201,
+      "learning_rate": 2.902880805113284e-07,
+      "loss": 0.53,
+      "step": 5634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.347680328468108,
+      "learning_rate": 2.8901687584938163e-07,
+      "loss": 0.5385,
+      "step": 5635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.739993355423697,
+      "learning_rate": 2.877484198607783e-07,
+      "loss": 0.522,
+      "step": 5636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.488043613413681,
+      "learning_rate": 2.864827129045511e-07,
+      "loss": 0.5673,
+      "step": 5637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.61352986894745,
+      "learning_rate": 2.852197553389568e-07,
+      "loss": 0.4706,
+      "step": 5638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7857630743028219,
+      "learning_rate": 2.8395954752147293e-07,
+      "loss": 0.4051,
+      "step": 5639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7136581425326916,
+      "learning_rate": 2.8270208980879843e-07,
+      "loss": 0.5403,
+      "step": 5640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.586878768749608,
+      "learning_rate": 2.8144738255685423e-07,
+      "loss": 0.5496,
+      "step": 5641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7607999354088124,
+      "learning_rate": 2.801954261207818e-07,
+      "loss": 0.5479,
+      "step": 5642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.868793497129673,
+      "learning_rate": 2.789462208549454e-07,
+      "loss": 0.527,
+      "step": 5643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5338021907534796,
+      "learning_rate": 2.7769976711293e-07,
+      "loss": 0.479,
+      "step": 5644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4521959762504708,
+      "learning_rate": 2.764560652475412e-07,
+      "loss": 0.5075,
+      "step": 5645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7304591080826419,
+      "learning_rate": 2.752151156108074e-07,
+      "loss": 0.6165,
+      "step": 5646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6520688734572802,
+      "learning_rate": 2.7397691855397534e-07,
+      "loss": 0.4131,
+      "step": 5647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8331808830373454,
+      "learning_rate": 2.727414744275147e-07,
+      "loss": 0.6202,
+      "step": 5648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1700663115921053,
+      "learning_rate": 2.7150878358111585e-07,
+      "loss": 0.5959,
+      "step": 5649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8577071582318834,
+      "learning_rate": 2.7027884636368853e-07,
+      "loss": 0.5893,
+      "step": 5650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6101912250685864,
+      "learning_rate": 2.6905166312336525e-07,
+      "loss": 0.5207,
+      "step": 5651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5161926510143877,
+      "learning_rate": 2.67827234207495e-07,
+      "loss": 0.4996,
+      "step": 5652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0195658693988054,
+      "learning_rate": 2.666055599626527e-07,
+      "loss": 0.5387,
+      "step": 5653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7978710548921606,
+      "learning_rate": 2.653866407346284e-07,
+      "loss": 0.4698,
+      "step": 5654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1953552750988408,
+      "learning_rate": 2.641704768684361e-07,
+      "loss": 0.5744,
+      "step": 5655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7061947054244628,
+      "learning_rate": 2.629570687083083e-07,
+      "loss": 0.5624,
+      "step": 5656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7974812091259467,
+      "learning_rate": 2.6174641659769683e-07,
+      "loss": 0.5968,
+      "step": 5657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3941992123204954,
+      "learning_rate": 2.6053852087927436e-07,
+      "loss": 0.525,
+      "step": 5658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6012198752935967,
+      "learning_rate": 2.5933338189493395e-07,
+      "loss": 0.5044,
+      "step": 5659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.041410683149231,
+      "learning_rate": 2.581309999857873e-07,
+      "loss": 0.5279,
+      "step": 5660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.794242368659679,
+      "learning_rate": 2.569313754921665e-07,
+      "loss": 0.495,
+      "step": 5661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.951794126027265,
+      "learning_rate": 2.5573450875362117e-07,
+      "loss": 0.5081,
+      "step": 5662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.159289251570974,
+      "learning_rate": 2.5454040010892354e-07,
+      "loss": 0.4904,
+      "step": 5663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.366793525231874,
+      "learning_rate": 2.5334904989606336e-07,
+      "loss": 0.5813,
+      "step": 5664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9625030497988338,
+      "learning_rate": 2.5216045845224854e-07,
+      "loss": 0.5333,
+      "step": 5665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7435094058140683,
+      "learning_rate": 2.5097462611390897e-07,
+      "loss": 0.5229,
+      "step": 5666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.132375867982676,
+      "learning_rate": 2.4979155321669166e-07,
+      "loss": 0.5043,
+      "step": 5667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.238222786690634,
+      "learning_rate": 2.486112400954621e-07,
+      "loss": 0.4535,
+      "step": 5668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9033522897397254,
+      "learning_rate": 2.474336870843064e-07,
+      "loss": 0.7114,
+      "step": 5669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8814047522471904,
+      "learning_rate": 2.462588945165267e-07,
+      "loss": 0.5813,
+      "step": 5670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7573810139687767,
+      "learning_rate": 2.450868627246483e-07,
+      "loss": 0.5384,
+      "step": 5671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8240040867557372,
+      "learning_rate": 2.439175920404102e-07,
+      "loss": 0.5124,
+      "step": 5672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1473860047290225,
+      "learning_rate": 2.427510827947721e-07,
+      "loss": 0.6432,
+      "step": 5673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.897227800983536,
+      "learning_rate": 2.415873353179132e-07,
+      "loss": 0.5358,
+      "step": 5674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.225117313834344,
+      "learning_rate": 2.404263499392301e-07,
+      "loss": 0.5087,
+      "step": 5675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0501739036707,
+      "learning_rate": 2.3926812698733535e-07,
+      "loss": 0.5123,
+      "step": 5676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7384507207343924,
+      "learning_rate": 2.381126667900624e-07,
+      "loss": 0.4721,
+      "step": 5677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9179908755015405,
+      "learning_rate": 2.3695996967446178e-07,
+      "loss": 0.4843,
+      "step": 5678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.944086121991037,
+      "learning_rate": 2.3581003596680252e-07,
+      "loss": 0.551,
+      "step": 5679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6944520919557107,
+      "learning_rate": 2.3466286599257094e-07,
+      "loss": 0.5837,
+      "step": 5680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.179501922743825,
+      "learning_rate": 2.335184600764695e-07,
+      "loss": 0.5719,
+      "step": 5681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.672218660987619,
+      "learning_rate": 2.3237681854242245e-07,
+      "loss": 0.4244,
+      "step": 5682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.671856865608342,
+      "learning_rate": 2.3123794171356683e-07,
+      "loss": 0.4777,
+      "step": 5683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7636914256893084,
+      "learning_rate": 2.3010182991226038e-07,
+      "loss": 0.6166,
+      "step": 5684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8025700418910301,
+      "learning_rate": 2.2896848346007584e-07,
+      "loss": 0.6153,
+      "step": 5685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0143572257953672,
+      "learning_rate": 2.2783790267780658e-07,
+      "loss": 0.567,
+      "step": 5686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.598344335264216,
+      "learning_rate": 2.2671008788546e-07,
+      "loss": 0.4619,
+      "step": 5687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.849376390313886,
+      "learning_rate": 2.2558503940226296e-07,
+      "loss": 0.4964,
+      "step": 5688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.234645696729217,
+      "learning_rate": 2.2446275754665514e-07,
+      "loss": 0.5658,
+      "step": 5689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6855409837593587,
+      "learning_rate": 2.2334324263629914e-07,
+      "loss": 0.4686,
+      "step": 5690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6881894650699683,
+      "learning_rate": 2.2222649498806924e-07,
+      "loss": 0.4768,
+      "step": 5691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8624299501138348,
+      "learning_rate": 2.211125149180604e-07,
+      "loss": 0.4979,
+      "step": 5692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.059555269290884,
+      "learning_rate": 2.2000130274158039e-07,
+      "loss": 0.4972,
+      "step": 5693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8498307554546196,
+      "learning_rate": 2.1889285877315647e-07,
+      "loss": 0.6714,
+      "step": 5694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8249672891097877,
+      "learning_rate": 2.1778718332653103e-07,
+      "loss": 0.4721,
+      "step": 5695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.365681556319147,
+      "learning_rate": 2.1668427671466375e-07,
+      "loss": 0.49,
+      "step": 5696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6938551405702111,
+      "learning_rate": 2.1558413924972933e-07,
+      "loss": 0.5092,
+      "step": 5697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8806121866673333,
+      "learning_rate": 2.144867712431198e-07,
+      "loss": 0.5187,
+      "step": 5698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8414640614447735,
+      "learning_rate": 2.1339217300544335e-07,
+      "loss": 0.6245,
+      "step": 5699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8493071657739333,
+      "learning_rate": 2.1230034484652218e-07,
+      "loss": 0.4991,
+      "step": 5700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9094670927563961,
+      "learning_rate": 2.1121128707539796e-07,
+      "loss": 0.5142,
+      "step": 5701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0929353936428097,
+      "learning_rate": 2.101250000003241e-07,
+      "loss": 0.5473,
+      "step": 5702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.399600658695943,
+      "learning_rate": 2.0904148392877354e-07,
+      "loss": 0.4882,
+      "step": 5703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8309014005504234,
+      "learning_rate": 2.079607391674321e-07,
+      "loss": 0.5205,
+      "step": 5704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.254634230032327,
+      "learning_rate": 2.0688276602220392e-07,
+      "loss": 0.5045,
+      "step": 5705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.268846570966351,
+      "learning_rate": 2.0580756479820496e-07,
+      "loss": 0.5319,
+      "step": 5706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9330688359935873,
+      "learning_rate": 2.0473513579976957e-07,
+      "loss": 0.5995,
+      "step": 5707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.200434521304202,
+      "learning_rate": 2.0366547933044712e-07,
+      "loss": 0.5082,
+      "step": 5708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5192135252132446,
+      "learning_rate": 2.0259859569299989e-07,
+      "loss": 0.4444,
+      "step": 5709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1280806005274453,
+      "learning_rate": 2.0153448518940967e-07,
+      "loss": 0.5067,
+      "step": 5710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8481568685904608,
+      "learning_rate": 2.0047314812086883e-07,
+      "loss": 0.4719,
+      "step": 5711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.978109503968233,
+      "learning_rate": 1.9941458478778597e-07,
+      "loss": 0.5929,
+      "step": 5712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2715443652672693,
+      "learning_rate": 1.983587954897881e-07,
+      "loss": 0.5123,
+      "step": 5713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9080264349128129,
+      "learning_rate": 1.9730578052571058e-07,
+      "loss": 0.5273,
+      "step": 5714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2525770354019397,
+      "learning_rate": 1.962555401936106e-07,
+      "loss": 0.4558,
+      "step": 5715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8710098363808425,
+      "learning_rate": 1.9520807479075366e-07,
+      "loss": 0.5349,
+      "step": 5716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8381609758425717,
+      "learning_rate": 1.941633846136237e-07,
+      "loss": 0.529,
+      "step": 5717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5650900223184803,
+      "learning_rate": 1.931214699579176e-07,
+      "loss": 0.4337,
+      "step": 5718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.81681202560634,
+      "learning_rate": 1.9208233111854824e-07,
+      "loss": 0.5152,
+      "step": 5719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.842218775385713,
+      "learning_rate": 1.910459683896415e-07,
+      "loss": 0.5318,
+      "step": 5720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7847863062803504,
+      "learning_rate": 1.900123820645361e-07,
+      "loss": 0.6669,
+      "step": 5721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5837016478683639,
+      "learning_rate": 1.8898157243578797e-07,
+      "loss": 0.4744,
+      "step": 5722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0032246231496194,
+      "learning_rate": 1.8795353979516596e-07,
+      "loss": 0.4992,
+      "step": 5723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6133110387627856,
+      "learning_rate": 1.8692828443365063e-07,
+      "loss": 0.5014,
+      "step": 5724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7946245231506093,
+      "learning_rate": 1.8590580664143987e-07,
+      "loss": 0.5444,
+      "step": 5725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.72738490198111,
+      "learning_rate": 1.8488610670794327e-07,
+      "loss": 0.5171,
+      "step": 5726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.220544938880382,
+      "learning_rate": 1.838691849217855e-07,
+      "loss": 0.5056,
+      "step": 5727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9156246727626514,
+      "learning_rate": 1.8285504157080414e-07,
+      "loss": 0.5239,
+      "step": 5728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.558220607057536,
+      "learning_rate": 1.818436769420484e-07,
+      "loss": 0.5213,
+      "step": 5729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1241486260552693,
+      "learning_rate": 1.80835091321786e-07,
+      "loss": 0.557,
+      "step": 5730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9732188539545448,
+      "learning_rate": 1.798292849954919e-07,
+      "loss": 0.5967,
+      "step": 5731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7773686477687811,
+      "learning_rate": 1.788262582478595e-07,
+      "loss": 0.4482,
+      "step": 5732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.064939672236274,
+      "learning_rate": 1.7782601136279277e-07,
+      "loss": 0.581,
+      "step": 5733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3276318900988966,
+      "learning_rate": 1.768285446234086e-07,
+      "loss": 0.5038,
+      "step": 5734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5866058993143588,
+      "learning_rate": 1.7583385831204003e-07,
+      "loss": 0.5503,
+      "step": 5735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2737136617844773,
+      "learning_rate": 1.7484195271022853e-07,
+      "loss": 0.5416,
+      "step": 5736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9016505693388819,
+      "learning_rate": 1.7385282809873283e-07,
+      "loss": 0.6189,
+      "step": 5737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1433650961505415,
+      "learning_rate": 1.7286648475752122e-07,
+      "loss": 0.4651,
+      "step": 5738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5766038186872717,
+      "learning_rate": 1.7188292296577703e-07,
+      "loss": 0.4194,
+      "step": 5739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9029765739656677,
+      "learning_rate": 1.7090214300189534e-07,
+      "loss": 0.5127,
+      "step": 5740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1194874196803535,
+      "learning_rate": 1.69924145143483e-07,
+      "loss": 0.5443,
+      "step": 5741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6455632281983206,
+      "learning_rate": 1.6894892966736076e-07,
+      "loss": 0.5088,
+      "step": 5742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.622834995129754,
+      "learning_rate": 1.6797649684956118e-07,
+      "loss": 0.4714,
+      "step": 5743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.037232829231243,
+      "learning_rate": 1.6700684696532854e-07,
+      "loss": 0.5026,
+      "step": 5744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1106015280314323,
+      "learning_rate": 1.6603998028912215e-07,
+      "loss": 0.4492,
+      "step": 5745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2602733652290876,
+      "learning_rate": 1.6507589709461092e-07,
+      "loss": 0.5293,
+      "step": 5746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.658360472980002,
+      "learning_rate": 1.6411459765467542e-07,
+      "loss": 0.4892,
+      "step": 5747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.487916442643535,
+      "learning_rate": 1.6315608224141023e-07,
+      "loss": 0.4825,
+      "step": 5748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.407962565705222,
+      "learning_rate": 1.6220035112612166e-07,
+      "loss": 0.6076,
+      "step": 5749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1661907447491444,
+      "learning_rate": 1.6124740457932553e-07,
+      "loss": 0.5034,
+      "step": 5750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.459907077191516,
+      "learning_rate": 1.6029724287075387e-07,
+      "loss": 0.4911,
+      "step": 5751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7833008347757215,
+      "learning_rate": 1.5934986626934602e-07,
+      "loss": 0.5406,
+      "step": 5752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.351696711411307,
+      "learning_rate": 1.5840527504325632e-07,
+      "loss": 0.4544,
+      "step": 5753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6074511236762872,
+      "learning_rate": 1.574634694598476e-07,
+      "loss": 0.5245,
+      "step": 5754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8568410975818401,
+      "learning_rate": 1.565244497856977e-07,
+      "loss": 0.499,
+      "step": 5755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.607708328624799,
+      "learning_rate": 1.5558821628659293e-07,
+      "loss": 0.5157,
+      "step": 5756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6523354252328244,
+      "learning_rate": 1.5465476922753352e-07,
+      "loss": 0.5472,
+      "step": 5757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.832596008979178,
+      "learning_rate": 1.5372410887272814e-07,
+      "loss": 0.5552,
+      "step": 5758
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.694863936269125,
+      "learning_rate": 1.527962354855994e-07,
+      "loss": 0.4607,
+      "step": 5759
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5612385570579812,
+      "learning_rate": 1.5187114932877945e-07,
+      "loss": 0.4279,
+      "step": 5760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3040804665158254,
+      "learning_rate": 1.5094885066411213e-07,
+      "loss": 0.5685,
+      "step": 5761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.79934698185178,
+      "learning_rate": 1.50029339752652e-07,
+      "loss": 0.538,
+      "step": 5762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8025842584801408,
+      "learning_rate": 1.4911261685466416e-07,
+      "loss": 0.5218,
+      "step": 5763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5485336329769788,
+      "learning_rate": 1.4819868222962664e-07,
+      "loss": 0.5056,
+      "step": 5764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6585864962874797,
+      "learning_rate": 1.472875361362247e-07,
+      "loss": 0.5036,
+      "step": 5765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8775312942016613,
+      "learning_rate": 1.4637917883235874e-07,
+      "loss": 0.4972,
+      "step": 5766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9286381618484247,
+      "learning_rate": 1.454736105751353e-07,
+      "loss": 0.5264,
+      "step": 5767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8402855026074088,
+      "learning_rate": 1.4457083162087383e-07,
+      "loss": 0.6609,
+      "step": 5768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2815021795894683,
+      "learning_rate": 1.436708422251054e-07,
+      "loss": 0.5038,
+      "step": 5769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7214990191551582,
+      "learning_rate": 1.4277364264256854e-07,
+      "loss": 0.4955,
+      "step": 5770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7177698906167869,
+      "learning_rate": 1.4187923312721452e-07,
+      "loss": 0.4573,
+      "step": 5771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.10352607278353,
+      "learning_rate": 1.4098761393220305e-07,
+      "loss": 0.514,
+      "step": 5772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1854230658857885,
+      "learning_rate": 1.4009878530990784e-07,
+      "loss": 0.457,
+      "step": 5773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8442974259788028,
+      "learning_rate": 1.392127475119065e-07,
+      "loss": 0.4462,
+      "step": 5774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3788270598122767,
+      "learning_rate": 1.3832950078899288e-07,
+      "loss": 0.558,
+      "step": 5775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7599611216295172,
+      "learning_rate": 1.3744904539116588e-07,
+      "loss": 0.5228,
+      "step": 5776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8657460378338655,
+      "learning_rate": 1.3657138156763838e-07,
+      "loss": 0.4647,
+      "step": 5777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.064102270384013,
+      "learning_rate": 1.3569650956682944e-07,
+      "loss": 0.4899,
+      "step": 5778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7295293241354924,
+      "learning_rate": 1.34824429636371e-07,
+      "loss": 0.6118,
+      "step": 5779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7685657698898165,
+      "learning_rate": 1.3395514202310443e-07,
+      "loss": 0.4568,
+      "step": 5780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7634470235251414,
+      "learning_rate": 1.3308864697307745e-07,
+      "loss": 0.4913,
+      "step": 5781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.417786056499222,
+      "learning_rate": 1.3222494473155156e-07,
+      "loss": 0.4741,
+      "step": 5782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9909920726022206,
+      "learning_rate": 1.313640355429946e-07,
+      "loss": 0.5246,
+      "step": 5783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8847889944588136,
+      "learning_rate": 1.3050591965108605e-07,
+      "loss": 0.6915,
+      "step": 5784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8516854421327147,
+      "learning_rate": 1.2965059729871275e-07,
+      "loss": 0.4702,
+      "step": 5785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.457456689364434,
+      "learning_rate": 1.287980687279722e-07,
+      "loss": 0.5067,
+      "step": 5786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0748825343585153,
+      "learning_rate": 1.2794833418017237e-07,
+      "loss": 0.5297,
+      "step": 5787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6355633783062749,
+      "learning_rate": 1.2710139389582654e-07,
+      "loss": 0.5549,
+      "step": 5788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0828526387336974,
+      "learning_rate": 1.2625724811466067e-07,
+      "loss": 0.5821,
+      "step": 5789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.224557599136987,
+      "learning_rate": 1.254158970756092e-07,
+      "loss": 0.5774,
+      "step": 5790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7003195376509936,
+      "learning_rate": 1.2457734101681274e-07,
+      "loss": 0.506,
+      "step": 5791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9039283725465752,
+      "learning_rate": 1.2374158017562366e-07,
+      "loss": 0.4872,
+      "step": 5792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.516454340285755,
+      "learning_rate": 1.2290861478860272e-07,
+      "loss": 0.5085,
+      "step": 5793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8422632654665942,
+      "learning_rate": 1.2207844509151912e-07,
+      "loss": 0.5614,
+      "step": 5794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0208227531190666,
+      "learning_rate": 1.2125107131935042e-07,
+      "loss": 0.4478,
+      "step": 5795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.760065814150823,
+      "learning_rate": 1.2042649370628268e-07,
+      "loss": 0.5295,
+      "step": 5796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8766269901302202,
+      "learning_rate": 1.1960471248571138e-07,
+      "loss": 0.5676,
+      "step": 5797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.704519547197977,
+      "learning_rate": 1.1878572789023935e-07,
+      "loss": 0.5251,
+      "step": 5798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.19187772996013,
+      "learning_rate": 1.1796954015167895e-07,
+      "loss": 0.5719,
+      "step": 5799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.889495880577242,
+      "learning_rate": 1.1715614950104981e-07,
+      "loss": 0.5768,
+      "step": 5800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5529730807004736,
+      "learning_rate": 1.1634555616858112e-07,
+      "loss": 0.5307,
+      "step": 5801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8631060362031411,
+      "learning_rate": 1.1553776038370934e-07,
+      "loss": 0.6784,
+      "step": 5802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.314735046098452,
+      "learning_rate": 1.1473276237507935e-07,
+      "loss": 0.5832,
+      "step": 5803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6770399574968615,
+      "learning_rate": 1.1393056237054445e-07,
+      "loss": 0.4165,
+      "step": 5804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.976462877187105,
+      "learning_rate": 1.1313116059716522e-07,
+      "loss": 0.529,
+      "step": 5805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.006570646499599,
+      "learning_rate": 1.1233455728121179e-07,
+      "loss": 0.6487,
+      "step": 5806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6611582091873567,
+      "learning_rate": 1.1154075264815822e-07,
+      "loss": 0.4901,
+      "step": 5807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1531374935774954,
+      "learning_rate": 1.1074974692269258e-07,
+      "loss": 0.5141,
+      "step": 5808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7306397104167693,
+      "learning_rate": 1.0996154032870576e-07,
+      "loss": 0.6579,
+      "step": 5809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.216636568446214,
+      "learning_rate": 1.091761330892982e-07,
+      "loss": 0.5158,
+      "step": 5810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9266384818202562,
+      "learning_rate": 1.0839352542677761e-07,
+      "loss": 0.5257,
+      "step": 5811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7865112430440062,
+      "learning_rate": 1.0761371756265903e-07,
+      "loss": 0.6383,
+      "step": 5812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6441057617649604,
+      "learning_rate": 1.068367097176659e-07,
+      "loss": 0.5781,
+      "step": 5813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.21908179599252,
+      "learning_rate": 1.0606250211172897e-07,
+      "loss": 0.5827,
+      "step": 5814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5337565373758872,
+      "learning_rate": 1.0529109496398626e-07,
+      "loss": 0.4738,
+      "step": 5815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4673425256127222,
+      "learning_rate": 1.0452248849278201e-07,
+      "loss": 0.4587,
+      "step": 5816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.034661920731525,
+      "learning_rate": 1.0375668291566998e-07,
+      "loss": 0.5322,
+      "step": 5817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6741977179861478,
+      "learning_rate": 1.02993678449409e-07,
+      "loss": 0.4669,
+      "step": 5818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.475169741070677,
+      "learning_rate": 1.022334753099652e-07,
+      "loss": 0.6306,
+      "step": 5819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7376306802670318,
+      "learning_rate": 1.0147607371251312e-07,
+      "loss": 0.5507,
+      "step": 5820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9135866445791503,
+      "learning_rate": 1.0072147387143349e-07,
+      "loss": 0.4491,
+      "step": 5821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2577446425527814,
+      "learning_rate": 9.996967600031437e-08,
+      "loss": 0.5323,
+      "step": 5822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.060989755923159,
+      "learning_rate": 9.92206803119511e-08,
+      "loss": 0.4971,
+      "step": 5823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1733404933207923,
+      "learning_rate": 9.847448701834405e-08,
+      "loss": 0.5263,
+      "step": 5824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6881051028862264,
+      "learning_rate": 9.773109633070322e-08,
+      "loss": 0.4585,
+      "step": 5825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3288831109069266,
+      "learning_rate": 9.699050845944357e-08,
+      "loss": 0.5928,
+      "step": 5826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8306160535829766,
+      "learning_rate": 9.625272361418525e-08,
+      "loss": 0.6075,
+      "step": 5827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8265741417226973,
+      "learning_rate": 9.551774200375896e-08,
+      "loss": 0.4739,
+      "step": 5828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0994160992331614,
+      "learning_rate": 9.478556383619719e-08,
+      "loss": 0.6395,
+      "step": 5829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.853024589952663,
+      "learning_rate": 9.405618931874417e-08,
+      "loss": 0.5402,
+      "step": 5830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9033846863831596,
+      "learning_rate": 9.332961865784584e-08,
+      "loss": 0.5699,
+      "step": 5831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.075476716060726,
+      "learning_rate": 9.260585205915773e-08,
+      "loss": 0.5942,
+      "step": 5832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8994500364145113,
+      "learning_rate": 9.188488972753928e-08,
+      "loss": 0.5545,
+      "step": 5833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.431558028571724,
+      "learning_rate": 9.116673186705838e-08,
+      "loss": 0.5217,
+      "step": 5834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.797095376262693,
+      "learning_rate": 9.045137868098686e-08,
+      "loss": 0.5388,
+      "step": 5835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0447917652101903,
+      "learning_rate": 8.973883037180609e-08,
+      "loss": 0.5585,
+      "step": 5836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.89916343090045,
+      "learning_rate": 8.902908714119918e-08,
+      "loss": 0.629,
+      "step": 5837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.966634199198327,
+      "learning_rate": 8.832214919005877e-08,
+      "loss": 0.5519,
+      "step": 5838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9747781321844748,
+      "learning_rate": 8.761801671848036e-08,
+      "loss": 0.5825,
+      "step": 5839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.881859196326343,
+      "learning_rate": 8.691668992577007e-08,
+      "loss": 0.4728,
+      "step": 5840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.261003623672291,
+      "learning_rate": 8.621816901043578e-08,
+      "loss": 0.5426,
+      "step": 5841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7382278623847516,
+      "learning_rate": 8.552245417019045e-08,
+      "loss": 0.6194,
+      "step": 5842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9536818933796178,
+      "learning_rate": 8.482954560195655e-08,
+      "loss": 0.4845,
+      "step": 5843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.358564978850815,
+      "learning_rate": 8.413944350186054e-08,
+      "loss": 0.5432,
+      "step": 5844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8375261844057094,
+      "learning_rate": 8.345214806523394e-08,
+      "loss": 0.5737,
+      "step": 5845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7585352742644391,
+      "learning_rate": 8.276765948661447e-08,
+      "loss": 0.5102,
+      "step": 5846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7195140714716992,
+      "learning_rate": 8.208597795974382e-08,
+      "loss": 0.5032,
+      "step": 5847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.690926825371246,
+      "learning_rate": 8.14071036775721e-08,
+      "loss": 0.4646,
+      "step": 5848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7885273808368207,
+      "learning_rate": 8.073103683225337e-08,
+      "loss": 0.4737,
+      "step": 5849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.944810899202194,
+      "learning_rate": 8.005777761514677e-08,
+      "loss": 0.5348,
+      "step": 5850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6254805602251445,
+      "learning_rate": 7.938732621681545e-08,
+      "loss": 0.5023,
+      "step": 5851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.878911830480318,
+      "learning_rate": 7.87196828270309e-08,
+      "loss": 0.5319,
+      "step": 5852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.285351684638002,
+      "learning_rate": 7.805484763476756e-08,
+      "loss": 0.4628,
+      "step": 5853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.876197830486094,
+      "learning_rate": 7.739282082820488e-08,
+      "loss": 0.5999,
+      "step": 5854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8159419511075751,
+      "learning_rate": 7.673360259472962e-08,
+      "loss": 0.6646,
+      "step": 5855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0206635829278996,
+      "learning_rate": 7.607719312093142e-08,
+      "loss": 0.5036,
+      "step": 5856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7246949482585126,
+      "learning_rate": 7.542359259260612e-08,
+      "loss": 0.5832,
+      "step": 5857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6300882931051999,
+      "learning_rate": 7.477280119475239e-08,
+      "loss": 0.4564,
+      "step": 5858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9730395113453176,
+      "learning_rate": 7.412481911157732e-08,
+      "loss": 0.5569,
+      "step": 5859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6972190448724627,
+      "learning_rate": 7.347964652648976e-08,
+      "loss": 0.5363,
+      "step": 5860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1586585911848397,
+      "learning_rate": 7.283728362210474e-08,
+      "loss": 0.4558,
+      "step": 5861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9607000064610938,
+      "learning_rate": 7.219773058024127e-08,
+      "loss": 0.5543,
+      "step": 5862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6393119028150938,
+      "learning_rate": 7.156098758192453e-08,
+      "loss": 0.5575,
+      "step": 5863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8346915811946721,
+      "learning_rate": 7.09270548073826e-08,
+      "loss": 0.6411,
+      "step": 5864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6544145993673887,
+      "learning_rate": 7.029593243604859e-08,
+      "loss": 0.5069,
+      "step": 5865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1414113999051114,
+      "learning_rate": 6.966762064656075e-08,
+      "loss": 0.5214,
+      "step": 5866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9658928191196066,
+      "learning_rate": 6.904211961676122e-08,
+      "loss": 0.5397,
+      "step": 5867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5715717216378153,
+      "learning_rate": 6.841942952369618e-08,
+      "loss": 0.5561,
+      "step": 5868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.070433748584896,
+      "learning_rate": 6.779955054361797e-08,
+      "loss": 0.4872,
+      "step": 5869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5734723972214832,
+      "learning_rate": 6.718248285198182e-08,
+      "loss": 0.3955,
+      "step": 5870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5110291555526216,
+      "learning_rate": 6.656822662344686e-08,
+      "loss": 0.5691,
+      "step": 5871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6820266125197736,
+      "learning_rate": 6.595678203187739e-08,
+      "loss": 0.5553,
+      "step": 5872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.250308617576051,
+      "learning_rate": 6.53481492503405e-08,
+      "loss": 0.5273,
+      "step": 5873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.667010298311093,
+      "learning_rate": 6.474232845111062e-08,
+      "loss": 0.5409,
+      "step": 5874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6057152763185423,
+      "learning_rate": 6.413931980566168e-08,
+      "loss": 0.4474,
+      "step": 5875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2141746891052594,
+      "learning_rate": 6.353912348467606e-08,
+      "loss": 0.4561,
+      "step": 5876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0188995223161696,
+      "learning_rate": 6.294173965803563e-08,
+      "loss": 0.4617,
+      "step": 5877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.510134222287849,
+      "learning_rate": 6.23471684948318e-08,
+      "loss": 0.539,
+      "step": 5878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.269355323950165,
+      "learning_rate": 6.175541016335329e-08,
+      "loss": 0.5013,
+      "step": 5879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.488079882118401,
+      "learning_rate": 6.116646483109945e-08,
+      "loss": 0.5614,
+      "step": 5880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7707812470965902,
+      "learning_rate": 6.058033266476693e-08,
+      "loss": 0.4965,
+      "step": 5881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0868793904716645,
+      "learning_rate": 5.99970138302608e-08,
+      "loss": 0.4707,
+      "step": 5882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9790795635522382,
+      "learning_rate": 5.9416508492688986e-08,
+      "loss": 0.4846,
+      "step": 5883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7706237441173696,
+      "learning_rate": 5.8838816816360054e-08,
+      "loss": 0.5135,
+      "step": 5884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3573883256596653,
+      "learning_rate": 5.826393896478988e-08,
+      "loss": 0.5615,
+      "step": 5885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1442639081810593,
+      "learning_rate": 5.7691875100697184e-08,
+      "loss": 0.4453,
+      "step": 5886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5727633799229666,
+      "learning_rate": 5.712262538600133e-08,
+      "loss": 0.4683,
+      "step": 5887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8822999529542175,
+      "learning_rate": 5.655618998182899e-08,
+      "loss": 0.5084,
+      "step": 5888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.141777427515292,
+      "learning_rate": 5.5992569048507475e-08,
+      "loss": 0.6176,
+      "step": 5889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3238889506068667,
+      "learning_rate": 5.543176274556916e-08,
+      "loss": 0.4803,
+      "step": 5890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6985188518235368,
+      "learning_rate": 5.487377123174931e-08,
+      "loss": 0.4699,
+      "step": 5891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4711245340437635,
+      "learning_rate": 5.43185946649849e-08,
+      "loss": 0.5608,
+      "step": 5892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5412267918214804,
+      "learning_rate": 5.376623320241914e-08,
+      "loss": 0.4727,
+      "step": 5893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.467833696112235,
+      "learning_rate": 5.321668700039584e-08,
+      "loss": 0.4427,
+      "step": 5894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9917583896159072,
+      "learning_rate": 5.26699562144628e-08,
+      "loss": 0.4971,
+      "step": 5895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.257680770335551,
+      "learning_rate": 5.2126040999371796e-08,
+      "loss": 0.5826,
+      "step": 5896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9067917337181202,
+      "learning_rate": 5.1584941509075225e-08,
+      "loss": 0.5536,
+      "step": 5897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0340335638532205,
+      "learning_rate": 5.10466578967328e-08,
+      "loss": 0.5309,
+      "step": 5898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7851302534639313,
+      "learning_rate": 5.051119031470264e-08,
+      "loss": 0.5199,
+      "step": 5899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9946970026928457,
+      "learning_rate": 4.997853891454796e-08,
+      "loss": 0.497,
+      "step": 5900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7955078519954508,
+      "learning_rate": 4.9448703847034816e-08,
+      "loss": 0.5503,
+      "step": 5901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.148170410148587,
+      "learning_rate": 4.892168526213215e-08,
+      "loss": 0.541,
+      "step": 5902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8276212119572157,
+      "learning_rate": 4.8397483309011726e-08,
+      "loss": 0.5014,
+      "step": 5903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8950287741961453,
+      "learning_rate": 4.78760981360471e-08,
+      "loss": 0.5731,
+      "step": 5904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9136598953046344,
+      "learning_rate": 4.735752989081577e-08,
+      "loss": 0.4963,
+      "step": 5905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.448381457563135,
+      "learning_rate": 4.6841778720095874e-08,
+      "loss": 0.5048,
+      "step": 5906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.072631410288573,
+      "learning_rate": 4.632884476987176e-08,
+      "loss": 0.525,
+      "step": 5907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9681742231270831,
+      "learning_rate": 4.581872818532729e-08,
+      "loss": 0.4439,
+      "step": 5908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2873271379332216,
+      "learning_rate": 4.531142911085029e-08,
+      "loss": 0.542,
+      "step": 5909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.078482911850069,
+      "learning_rate": 4.480694769003035e-08,
+      "loss": 0.5171,
+      "step": 5910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8362402797860806,
+      "learning_rate": 4.4305284065659926e-08,
+      "loss": 0.4717,
+      "step": 5911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9730715373083025,
+      "learning_rate": 4.38064383797332e-08,
+      "loss": 0.6388,
+      "step": 5912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7281852696973548,
+      "learning_rate": 4.331041077344944e-08,
+      "loss": 0.4905,
+      "step": 5913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5580824920321685,
+      "learning_rate": 4.281720138720524e-08,
+      "loss": 0.4742,
+      "step": 5914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5935631098586465,
+      "learning_rate": 4.2326810360605594e-08,
+      "loss": 0.4494,
+      "step": 5915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2925266866154907,
+      "learning_rate": 4.1839237832452805e-08,
+      "loss": 0.5532,
+      "step": 5916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8016032451120296,
+      "learning_rate": 4.1354483940753145e-08,
+      "loss": 0.6796,
+      "step": 5917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8344522317967995,
+      "learning_rate": 4.087254882271574e-08,
+      "loss": 0.6461,
+      "step": 5918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.785019475054023,
+      "learning_rate": 4.039343261475148e-08,
+      "loss": 0.6415,
+      "step": 5919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7430746574422216,
+      "learning_rate": 3.9917135452473e-08,
+      "loss": 0.4702,
+      "step": 5920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.479129168305605,
+      "learning_rate": 3.944365747069467e-08,
+      "loss": 0.49,
+      "step": 5921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7978448144795363,
+      "learning_rate": 3.897299880343486e-08,
+      "loss": 0.4663,
+      "step": 5922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8325766600697626,
+      "learning_rate": 3.8505159583911434e-08,
+      "loss": 0.5468,
+      "step": 5923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6980204184842043,
+      "learning_rate": 3.8040139944545143e-08,
+      "loss": 0.4641,
+      "step": 5924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.94939873333579,
+      "learning_rate": 3.757794001695847e-08,
+      "loss": 0.5317,
+      "step": 5925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.060777489231123,
+      "learning_rate": 3.711855993197788e-08,
+      "loss": 0.4542,
+      "step": 5926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2728019729843516,
+      "learning_rate": 3.666199981962826e-08,
+      "loss": 0.5125,
+      "step": 5927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3138351315497725,
+      "learning_rate": 3.6208259809139554e-08,
+      "loss": 0.4643,
+      "step": 5928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.432905460475706,
+      "learning_rate": 3.575734002893905e-08,
+      "loss": 0.6132,
+      "step": 5929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.90460750253291,
+      "learning_rate": 3.5309240606662406e-08,
+      "loss": 0.4957,
+      "step": 5930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6643522995486784,
+      "learning_rate": 3.486396166914041e-08,
+      "loss": 0.5348,
+      "step": 5931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.865765603955716,
+      "learning_rate": 3.44215033424089e-08,
+      "loss": 0.4896,
+      "step": 5932
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8981051268588158,
+      "learning_rate": 3.3981865751705477e-08,
+      "loss": 0.4885,
+      "step": 5933
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.149895028899569,
+      "learning_rate": 3.354504902146838e-08,
+      "loss": 0.5647,
+      "step": 5934
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8550190968785492,
+      "learning_rate": 3.31110532753387e-08,
+      "loss": 0.4555,
+      "step": 5935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4933083200228383,
+      "learning_rate": 3.267987863615485e-08,
+      "loss": 0.542,
+      "step": 5936
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0161118711989054,
+      "learning_rate": 3.225152522596364e-08,
+      "loss": 0.4565,
+      "step": 5937
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.495882447272534,
+      "learning_rate": 3.182599316600699e-08,
+      "loss": 0.5915,
+      "step": 5938
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7743637277905421,
+      "learning_rate": 3.140328257673187e-08,
+      "loss": 0.6336,
+      "step": 5939
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8024080447890714,
+      "learning_rate": 3.0983393577785906e-08,
+      "loss": 0.5002,
+      "step": 5940
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7285313456176725,
+      "learning_rate": 3.056632628801737e-08,
+      "loss": 0.4236,
+      "step": 5941
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6131260099113782,
+      "learning_rate": 3.015208082547627e-08,
+      "loss": 0.5624,
+      "step": 5942
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8879601869151617,
+      "learning_rate": 2.974065730741327e-08,
+      "loss": 0.51,
+      "step": 5943
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5172864070536853,
+      "learning_rate": 2.933205585028298e-08,
+      "loss": 0.5423,
+      "step": 5944
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.467854931749885,
+      "learning_rate": 2.8926276569736233e-08,
+      "loss": 0.4922,
+      "step": 5945
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8154603732716557,
+      "learning_rate": 2.852331958063115e-08,
+      "loss": 0.5331,
+      "step": 5946
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9068177348348498,
+      "learning_rate": 2.8123184997022046e-08,
+      "loss": 0.4912,
+      "step": 5947
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8838508765342459,
+      "learning_rate": 2.772587293216611e-08,
+      "loss": 0.6388,
+      "step": 5948
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7005016090722815,
+      "learning_rate": 2.7331383498522268e-08,
+      "loss": 0.5262,
+      "step": 5949
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8656387314603482,
+      "learning_rate": 2.6939716807748982e-08,
+      "loss": 0.6309,
+      "step": 5950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6835403244013276,
+      "learning_rate": 2.6550872970707576e-08,
+      "loss": 0.4508,
+      "step": 5951
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8674451313147884,
+      "learning_rate": 2.6164852097460003e-08,
+      "loss": 0.5817,
+      "step": 5952
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7358819711199256,
+      "learning_rate": 2.5781654297267756e-08,
+      "loss": 0.5137,
+      "step": 5953
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8716689023151645,
+      "learning_rate": 2.540127967859407e-08,
+      "loss": 0.4726,
+      "step": 5954
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7604176114115797,
+      "learning_rate": 2.5023728349102828e-08,
+      "loss": 0.5198,
+      "step": 5955
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7888636552810037,
+      "learning_rate": 2.464900041566076e-08,
+      "loss": 0.6682,
+      "step": 5956
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.738737788003692,
+      "learning_rate": 2.427709598433303e-08,
+      "loss": 0.4448,
+      "step": 5957
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7625834869238701,
+      "learning_rate": 2.3908015160385433e-08,
+      "loss": 0.5047,
+      "step": 5958
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5607446357844712,
+      "learning_rate": 2.3541758048286624e-08,
+      "loss": 0.5283,
+      "step": 5959
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8930537518064612,
+      "learning_rate": 2.3178324751704784e-08,
+      "loss": 0.5415,
+      "step": 5960
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2761780722803326,
+      "learning_rate": 2.2817715373507633e-08,
+      "loss": 0.4552,
+      "step": 5961
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7748507924612442,
+      "learning_rate": 2.245993001576685e-08,
+      "loss": 0.5552,
+      "step": 5962
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6082668775803284,
+      "learning_rate": 2.2104968779752546e-08,
+      "loss": 0.5657,
+      "step": 5963
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9976077724561967,
+      "learning_rate": 2.1752831765934346e-08,
+      "loss": 0.4843,
+      "step": 5964
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1298853639784574,
+      "learning_rate": 2.1403519073984748e-08,
+      "loss": 0.4296,
+      "step": 5965
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2060693321460003,
+      "learning_rate": 2.105703080277688e-08,
+      "loss": 0.5126,
+      "step": 5966
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.541420706898163,
+      "learning_rate": 2.0713367050381185e-08,
+      "loss": 0.4951,
+      "step": 5967
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.42653782735051,
+      "learning_rate": 2.037252791407318e-08,
+      "loss": 0.5054,
+      "step": 5968
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8832835163776709,
+      "learning_rate": 2.0034513490325702e-08,
+      "loss": 0.4792,
+      "step": 5969
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.552375919565958,
+      "learning_rate": 1.9699323874814437e-08,
+      "loss": 0.5272,
+      "step": 5970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.495688252152424,
+      "learning_rate": 1.936695916241127e-08,
+      "loss": 0.4678,
+      "step": 5971
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.65203886151888,
+      "learning_rate": 1.9037419447193174e-08,
+      "loss": 0.5144,
+      "step": 5972
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.031460646802275,
+      "learning_rate": 1.8710704822435534e-08,
+      "loss": 0.4922,
+      "step": 5973
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.098376272743303,
+      "learning_rate": 1.8386815380614376e-08,
+      "loss": 0.5242,
+      "step": 5974
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0621762594416393,
+      "learning_rate": 1.8065751213405257e-08,
+      "loss": 0.5417,
+      "step": 5975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8422424871345163,
+      "learning_rate": 1.7747512411685486e-08,
+      "loss": 0.6316,
+      "step": 5976
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8611185433061277,
+      "learning_rate": 1.7432099065531894e-08,
+      "loss": 0.4972,
+      "step": 5977
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.480645009058756,
+      "learning_rate": 1.7119511264220844e-08,
+      "loss": 0.478,
+      "step": 5978
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.601337935041734,
+      "learning_rate": 1.6809749096229344e-08,
+      "loss": 0.5952,
+      "step": 5979
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.55256071037857,
+      "learning_rate": 1.6502812649236145e-08,
+      "loss": 0.5188,
+      "step": 5980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.967344117296934,
+      "learning_rate": 1.6198702010118417e-08,
+      "loss": 0.4801,
+      "step": 5981
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3729922785878927,
+      "learning_rate": 1.5897417264953974e-08,
+      "loss": 0.4796,
+      "step": 5982
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7701566996589087,
+      "learning_rate": 1.5598958499021265e-08,
+      "loss": 0.6094,
+      "step": 5983
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.887084479484916,
+      "learning_rate": 1.5303325796797163e-08,
+      "loss": 0.5025,
+      "step": 5984
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.326302459805063,
+      "learning_rate": 1.5010519241961397e-08,
+      "loss": 0.5599,
+      "step": 5985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.601109066705509,
+      "learning_rate": 1.4720538917391003e-08,
+      "loss": 0.4887,
+      "step": 5986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9638803039546344,
+      "learning_rate": 1.4433384905164771e-08,
+      "loss": 0.6018,
+      "step": 5987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9606316153069678,
+      "learning_rate": 1.4149057286562128e-08,
+      "loss": 0.5183,
+      "step": 5988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7939366969822725,
+      "learning_rate": 1.3867556142059813e-08,
+      "loss": 0.5123,
+      "step": 5989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9879157093953252,
+      "learning_rate": 1.3588881551337418e-08,
+      "loss": 0.5239,
+      "step": 5990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8641073202789538,
+      "learning_rate": 1.331303359327185e-08,
+      "loss": 0.4618,
+      "step": 5991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6458447544998993,
+      "learning_rate": 1.3040012345941765e-08,
+      "loss": 0.5109,
+      "step": 5992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5293467793254272,
+      "learning_rate": 1.2769817886626456e-08,
+      "loss": 0.4404,
+      "step": 5993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6047548186449463,
+      "learning_rate": 1.2502450291802526e-08,
+      "loss": 0.4377,
+      "step": 5994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.271572431964419,
+      "learning_rate": 1.2237909637147217e-08,
+      "loss": 0.5512,
+      "step": 5995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.676784451880704,
+      "learning_rate": 1.1976195997540629e-08,
+      "loss": 0.5042,
+      "step": 5996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6213598206242292,
+      "learning_rate": 1.1717309447057957e-08,
+      "loss": 0.5539,
+      "step": 5997
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0761851755276037,
+      "learning_rate": 1.1461250058977248e-08,
+      "loss": 0.5572,
+      "step": 5998
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7699285038970038,
+      "learning_rate": 1.120801790577608e-08,
+      "loss": 0.6407,
+      "step": 5999
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8684826500102942,
+      "learning_rate": 1.0957613059131567e-08,
+      "loss": 0.5019,
+      "step": 6000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.552835189030329,
+      "learning_rate": 1.0710035589918122e-08,
+      "loss": 0.5794,
+      "step": 6001
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9614688399999574,
+      "learning_rate": 1.0465285568214134e-08,
+      "loss": 0.5405,
+      "step": 6002
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.726210705263834,
+      "learning_rate": 1.02233630632953e-08,
+      "loss": 0.5353,
+      "step": 6003
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7985217939726623,
+      "learning_rate": 9.984268143636844e-09,
+      "loss": 0.5102,
+      "step": 6004
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5904313622927868,
+      "learning_rate": 9.74800087691352e-09,
+      "loss": 0.4563,
+      "step": 6005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1269539568433875,
+      "learning_rate": 9.514561330001836e-09,
+      "loss": 0.5649,
+      "step": 6006
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8373999814821,
+      "learning_rate": 9.283949568974493e-09,
+      "loss": 0.4324,
+      "step": 6007
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6780806552461915,
+      "learning_rate": 9.056165659107053e-09,
+      "loss": 0.444,
+      "step": 6008
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.045000694048385,
+      "learning_rate": 8.831209664872386e-09,
+      "loss": 0.4941,
+      "step": 6009
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8633122051280686,
+      "learning_rate": 8.609081649945116e-09,
+      "loss": 0.523,
+      "step": 6010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7417056082660591,
+      "learning_rate": 8.389781677196062e-09,
+      "loss": 0.5244,
+      "step": 6011
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6737642438160858,
+      "learning_rate": 8.173309808700014e-09,
+      "loss": 0.5312,
+      "step": 6012
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8544428616877225,
+      "learning_rate": 7.959666105727959e-09,
+      "loss": 0.6475,
+      "step": 6013
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3556287018550464,
+      "learning_rate": 7.748850628749305e-09,
+      "loss": 0.5902,
+      "step": 6014
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.04058674313181,
+      "learning_rate": 7.54086343743854e-09,
+      "loss": 0.5478,
+      "step": 6015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6536767694829522,
+      "learning_rate": 7.335704590663017e-09,
+      "loss": 0.4735,
+      "step": 6016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1151383630942235,
+      "learning_rate": 7.1333741464940655e-09,
+      "loss": 0.4824,
+      "step": 6017
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.23593569980348,
+      "learning_rate": 6.933872162199207e-09,
+      "loss": 0.5378,
+      "step": 6018
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8375063438308903,
+      "learning_rate": 6.737198694249936e-09,
+      "loss": 0.4559,
+      "step": 6019
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8107006493269413,
+      "learning_rate": 6.543353798311725e-09,
+      "loss": 0.48,
+      "step": 6020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6684699109672598,
+      "learning_rate": 6.352337529252905e-09,
+      "loss": 0.5087,
+      "step": 6021
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7984607443605114,
+      "learning_rate": 6.164149941140229e-09,
+      "loss": 0.5235,
+      "step": 6022
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6652540290018503,
+      "learning_rate": 5.978791087239977e-09,
+      "loss": 0.4859,
+      "step": 6023
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6947746409604256,
+      "learning_rate": 5.796261020016847e-09,
+      "loss": 0.4949,
+      "step": 6024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.245859915121618,
+      "learning_rate": 5.616559791136178e-09,
+      "loss": 0.5334,
+      "step": 6025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6820846648585666,
+      "learning_rate": 5.439687451461728e-09,
+      "loss": 0.5331,
+      "step": 6026
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9202294315392776,
+      "learning_rate": 5.265644051057894e-09,
+      "loss": 0.4797,
+      "step": 6027
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8365468521108848,
+      "learning_rate": 5.0944296391863825e-09,
+      "loss": 0.492,
+      "step": 6028
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9051230897738378,
+      "learning_rate": 4.926044264308427e-09,
+      "loss": 0.5332,
+      "step": 6029
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.426039311954394,
+      "learning_rate": 4.760487974085903e-09,
+      "loss": 0.3799,
+      "step": 6030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9231004078138703,
+      "learning_rate": 4.597760815379104e-09,
+      "loss": 0.5255,
+      "step": 6031
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7440411551247066,
+      "learning_rate": 4.437862834247852e-09,
+      "loss": 0.4691,
+      "step": 6032
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.525012092439888,
+      "learning_rate": 4.2807940759515e-09,
+      "loss": 0.5857,
+      "step": 6033
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7941351777893728,
+      "learning_rate": 4.126554584946707e-09,
+      "loss": 0.5231,
+      "step": 6034
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5491614908236913,
+      "learning_rate": 3.975144404890774e-09,
+      "loss": 0.4747,
+      "step": 6035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.549198850516288,
+      "learning_rate": 3.82656357864053e-09,
+      "loss": 0.4581,
+      "step": 6036
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2501559963895086,
+      "learning_rate": 3.680812148251223e-09,
+      "loss": 0.5464,
+      "step": 6037
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8084576978719638,
+      "learning_rate": 3.53789015497763e-09,
+      "loss": 0.5521,
+      "step": 6038
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4398327279225236,
+      "learning_rate": 3.3977976392740587e-09,
+      "loss": 0.524,
+      "step": 6039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8136494565678047,
+      "learning_rate": 3.2605346407932336e-09,
+      "loss": 0.4449,
+      "step": 6040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.768264511641396,
+      "learning_rate": 3.1261011983851898e-09,
+      "loss": 0.4706,
+      "step": 6041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1213878258075916,
+      "learning_rate": 2.994497350103931e-09,
+      "loss": 0.5287,
+      "step": 6042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5841810900013638,
+      "learning_rate": 2.8657231331985503e-09,
+      "loss": 0.5635,
+      "step": 6043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9208736374569297,
+      "learning_rate": 2.7397785841176692e-09,
+      "loss": 0.5681,
+      "step": 6044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.684086835314751,
+      "learning_rate": 2.6166637385094396e-09,
+      "loss": 0.4728,
+      "step": 6045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7683237285898605,
+      "learning_rate": 2.496378631222651e-09,
+      "loss": 0.6258,
+      "step": 6046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6385352959148323,
+      "learning_rate": 2.3789232963034036e-09,
+      "loss": 0.5677,
+      "step": 6047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9807319686771834,
+      "learning_rate": 2.264297766995105e-09,
+      "loss": 0.521,
+      "step": 6048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8017487807469054,
+      "learning_rate": 2.152502075745133e-09,
+      "loss": 0.6381,
+      "step": 6049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.226885351241074,
+      "learning_rate": 2.0435362541959545e-09,
+      "loss": 0.4281,
+      "step": 6050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.446887245499019,
+      "learning_rate": 1.9374003331895654e-09,
+      "loss": 0.4561,
+      "step": 6051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1130937084356183,
+      "learning_rate": 1.8340943427685997e-09,
+      "loss": 0.4983,
+      "step": 6052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.989375443835961,
+      "learning_rate": 1.7336183121730022e-09,
+      "loss": 0.4706,
+      "step": 6053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7301289031982313,
+      "learning_rate": 1.635972269841135e-09,
+      "loss": 0.6234,
+      "step": 6054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7034976235449508,
+      "learning_rate": 1.541156243413111e-09,
+      "loss": 0.4839,
+      "step": 6055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.901447396209108,
+      "learning_rate": 1.4491702597263513e-09,
+      "loss": 0.5013,
+      "step": 6056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8258500162709057,
+      "learning_rate": 1.360014344816696e-09,
+      "loss": 0.4691,
+      "step": 6057
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6825650318678937,
+      "learning_rate": 1.273688523919514e-09,
+      "loss": 0.5027,
+      "step": 6058
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8015963690841894,
+      "learning_rate": 1.1901928214685942e-09,
+      "loss": 0.5381,
+      "step": 6059
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9561930441517035,
+      "learning_rate": 1.1095272610994746e-09,
+      "loss": 0.6109,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9111956809614574,
+      "learning_rate": 1.0316918656416708e-09,
+      "loss": 0.4977,
+      "step": 6061
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.777822634051183,
+      "learning_rate": 9.56686657127559e-10,
+      "loss": 0.4701,
+      "step": 6062
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9017687275022053,
+      "learning_rate": 8.845116567879342e-10,
+      "loss": 0.51,
+      "step": 6063
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.515118450242749,
+      "learning_rate": 8.151668850508998e-10,
+      "loss": 0.5149,
+      "step": 6064
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.660513635162216,
+      "learning_rate": 7.486523615440888e-10,
+      "loss": 0.4495,
+      "step": 6065
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2496162538575986,
+      "learning_rate": 6.849681050946633e-10,
+      "loss": 0.5656,
+      "step": 6066
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9707400740411671,
+      "learning_rate": 6.241141337282042e-10,
+      "loss": 0.5144,
+      "step": 6067
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9934444586586146,
+      "learning_rate": 5.660904646698217e-10,
+      "loss": 0.5149,
+      "step": 6068
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.0351938035534545,
+      "learning_rate": 5.10897114341935e-10,
+      "loss": 0.4379,
+      "step": 6069
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.820480887025477,
+      "learning_rate": 4.585340983676023e-10,
+      "loss": 0.5357,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.28583060795893,
+      "learning_rate": 4.090014315694113e-10,
+      "loss": 0.4443,
+      "step": 6071
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8994080206466093,
+      "learning_rate": 3.6229912796392763e-10,
+      "loss": 0.4638,
+      "step": 6072
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8172487187500708,
+      "learning_rate": 3.1842720077390754e-10,
+      "loss": 0.5264,
+      "step": 6073
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8591896026449006,
+      "learning_rate": 2.773856624149751e-10,
+      "loss": 0.5166,
+      "step": 6074
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3349573646952306,
+      "learning_rate": 2.391745245045041e-10,
+      "loss": 0.6745,
+      "step": 6075
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7779817456955365,
+      "learning_rate": 2.0379379785828712e-10,
+      "loss": 0.5211,
+      "step": 6076
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8476235979903926,
+      "learning_rate": 1.7124349249053596e-10,
+      "loss": 0.4637,
+      "step": 6077
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0124106618774764,
+      "learning_rate": 1.4152361761388122e-10,
+      "loss": 0.5014,
+      "step": 6078
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9938550658870382,
+      "learning_rate": 1.1463418164159301e-10,
+      "loss": 0.5286,
+      "step": 6079
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8458356895666863,
+      "learning_rate": 9.057519218425015e-11,
+      "loss": 0.6786,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5375338304789894,
+      "learning_rate": 6.934665605196068e-11,
+      "loss": 0.505,
+      "step": 6081
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.845936835800302,
+      "learning_rate": 5.094857925214136e-11,
+      "loss": 0.5415,
+      "step": 6082
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.068088805016753,
+      "learning_rate": 3.5380966993958655e-11,
+      "loss": 0.4726,
+      "step": 6083
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7331096911985349,
+      "learning_rate": 2.264382368277751e-11,
+      "loss": 0.4586,
+      "step": 6084
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3482204216424822,
+      "learning_rate": 1.2737152924602313e-11,
+      "loss": 0.5521,
+      "step": 6085
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8728179085571117,
+      "learning_rate": 5.660957522746202e-12,
+      "loss": 0.648,
+      "step": 6086
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7055100732431538,
+      "learning_rate": 1.4152394811617343e-12,
+      "loss": 0.4564,
+      "step": 6087
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.083319644650456,
+      "learning_rate": 0.0,
+      "loss": 0.5634,
+      "step": 6088
+    },
+    {
+      "epoch": 1.0,
+      "step": 6088,
+      "total_flos": 2846846128226304.0,
+      "train_loss": 0.571346512345815,
+      "train_runtime": 142962.4661,
+      "train_samples_per_second": 5.451,
+      "train_steps_per_second": 0.043
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6088,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "total_flos": 2846846128226304.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}