diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,81608 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 11654,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.855755805969238,
+      "learning_rate": 2.9997425776557404e-05,
+      "loss": 6.0374,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.109065532684326,
+      "learning_rate": 2.999485155311481e-05,
+      "loss": 5.84,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.335865497589111,
+      "learning_rate": 2.9992277329672214e-05,
+      "loss": 5.6446,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.389212608337402,
+      "learning_rate": 2.998970310622962e-05,
+      "loss": 5.5062,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.956500053405762,
+      "learning_rate": 2.9987128882787027e-05,
+      "loss": 5.2977,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.355259895324707,
+      "learning_rate": 2.9984554659344434e-05,
+      "loss": 5.0181,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3592987060546875,
+      "learning_rate": 2.9981980435901837e-05,
+      "loss": 4.8524,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.215114593505859,
+      "learning_rate": 2.9979406212459244e-05,
+      "loss": 4.7503,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.257913589477539,
+      "learning_rate": 2.9976831989016647e-05,
+      "loss": 4.5238,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.579934597015381,
+      "learning_rate": 2.9974257765574054e-05,
+      "loss": 4.353,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.536452293395996,
+      "learning_rate": 2.9971683542131457e-05,
+      "loss": 4.2689,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.942359447479248,
+      "learning_rate": 2.996910931868886e-05,
+      "loss": 4.1416,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.103983402252197,
+      "learning_rate": 2.9966535095246267e-05,
+      "loss": 3.9087,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.696576118469238,
+      "learning_rate": 2.996396087180367e-05,
+      "loss": 3.8292,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.177977561950684,
+      "learning_rate": 2.996138664836108e-05,
+      "loss": 3.5344,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.678316593170166,
+      "learning_rate": 2.9958812424918484e-05,
+      "loss": 3.3705,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.075394630432129,
+      "learning_rate": 2.995623820147589e-05,
+      "loss": 3.7041,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.564874649047852,
+      "learning_rate": 2.9953663978033294e-05,
+      "loss": 2.9878,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.051210403442383,
+      "learning_rate": 2.99510897545907e-05,
+      "loss": 3.2233,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.445047378540039,
+      "learning_rate": 2.9948515531148104e-05,
+      "loss": 3.0284,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.75167179107666,
+      "learning_rate": 2.9945941307705507e-05,
+      "loss": 3.2145,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.302085876464844,
+      "learning_rate": 2.9943367084262914e-05,
+      "loss": 3.022,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.62187385559082,
+      "learning_rate": 2.9940792860820317e-05,
+      "loss": 2.8868,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.765596389770508,
+      "learning_rate": 2.9938218637377727e-05,
+      "loss": 2.7575,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.492706298828125,
+      "learning_rate": 2.993564441393513e-05,
+      "loss": 2.9801,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.508520126342773,
+      "learning_rate": 2.9933070190492537e-05,
+      "loss": 2.9401,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.079510688781738,
+      "learning_rate": 2.993049596704994e-05,
+      "loss": 2.8228,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.55045509338379,
+      "learning_rate": 2.9927921743607347e-05,
+      "loss": 2.5193,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.096270561218262,
+      "learning_rate": 2.992534752016475e-05,
+      "loss": 2.612,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.282428741455078,
+      "learning_rate": 2.9922773296722157e-05,
+      "loss": 2.8728,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.487578392028809,
+      "learning_rate": 2.992019907327956e-05,
+      "loss": 2.8666,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.594644546508789,
+      "learning_rate": 2.9917624849836964e-05,
+      "loss": 2.4412,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.025802612304688,
+      "learning_rate": 2.991505062639437e-05,
+      "loss": 2.6776,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.0786771774292,
+      "learning_rate": 2.9912476402951777e-05,
+      "loss": 2.0032,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.964434623718262,
+      "learning_rate": 2.9909902179509184e-05,
+      "loss": 2.7171,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.710619926452637,
+      "learning_rate": 2.9907327956066587e-05,
+      "loss": 2.3159,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.789759635925293,
+      "learning_rate": 2.9904753732623994e-05,
+      "loss": 2.3495,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.272398948669434,
+      "learning_rate": 2.9902179509181397e-05,
+      "loss": 2.4484,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.568193435668945,
+      "learning_rate": 2.9899605285738804e-05,
+      "loss": 2.3863,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.755990028381348,
+      "learning_rate": 2.9897031062296207e-05,
+      "loss": 2.1977,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.962193489074707,
+      "learning_rate": 2.9894456838853614e-05,
+      "loss": 2.4596,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.54276180267334,
+      "learning_rate": 2.9891882615411017e-05,
+      "loss": 2.0344,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.461871147155762,
+      "learning_rate": 2.9889308391968424e-05,
+      "loss": 2.5114,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.544355392456055,
+      "learning_rate": 2.988673416852583e-05,
+      "loss": 2.2582,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.856222152709961,
+      "learning_rate": 2.9884159945083234e-05,
+      "loss": 2.5919,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.033227920532227,
+      "learning_rate": 2.988158572164064e-05,
+      "loss": 2.117,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.805606842041016,
+      "learning_rate": 2.9879011498198044e-05,
+      "loss": 2.5082,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.737322807312012,
+      "learning_rate": 2.987643727475545e-05,
+      "loss": 2.3978,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.552938461303711,
+      "learning_rate": 2.9873863051312854e-05,
+      "loss": 2.1329,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.492465019226074,
+      "learning_rate": 2.987128882787026e-05,
+      "loss": 2.2769,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.515401840209961,
+      "learning_rate": 2.9868714604427664e-05,
+      "loss": 2.6346,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.540853500366211,
+      "learning_rate": 2.986614038098507e-05,
+      "loss": 2.3172,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.621674537658691,
+      "learning_rate": 2.9863566157542477e-05,
+      "loss": 2.3729,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.57535457611084,
+      "learning_rate": 2.986099193409988e-05,
+      "loss": 1.7151,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.841571807861328,
+      "learning_rate": 2.9858417710657287e-05,
+      "loss": 2.0466,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.51070499420166,
+      "learning_rate": 2.985584348721469e-05,
+      "loss": 2.4203,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.48379135131836,
+      "learning_rate": 2.9853269263772097e-05,
+      "loss": 2.1071,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.393445014953613,
+      "learning_rate": 2.98506950403295e-05,
+      "loss": 2.014,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.898478507995605,
+      "learning_rate": 2.9848120816886907e-05,
+      "loss": 2.4063,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.223072052001953,
+      "learning_rate": 2.984554659344431e-05,
+      "loss": 2.1862,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.349495887756348,
+      "learning_rate": 2.9842972370001717e-05,
+      "loss": 1.829,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.530231475830078,
+      "learning_rate": 2.9840398146559124e-05,
+      "loss": 2.2932,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.20637321472168,
+      "learning_rate": 2.9837823923116527e-05,
+      "loss": 1.7466,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.135061264038086,
+      "learning_rate": 2.9835249699673934e-05,
+      "loss": 1.9535,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.84786319732666,
+      "learning_rate": 2.9832675476231337e-05,
+      "loss": 2.3064,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.094093322753906,
+      "learning_rate": 2.9830101252788743e-05,
+      "loss": 1.9594,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.481359481811523,
+      "learning_rate": 2.9827527029346147e-05,
+      "loss": 2.3698,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.151596069335938,
+      "learning_rate": 2.9824952805903553e-05,
+      "loss": 2.1976,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.80447769165039,
+      "learning_rate": 2.9822378582460957e-05,
+      "loss": 2.2426,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.807653427124023,
+      "learning_rate": 2.9819804359018363e-05,
+      "loss": 1.8431,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.624991416931152,
+      "learning_rate": 2.981723013557577e-05,
+      "loss": 1.9657,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.18333911895752,
+      "learning_rate": 2.9814655912133177e-05,
+      "loss": 1.9263,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.231605529785156,
+      "learning_rate": 2.981208168869058e-05,
+      "loss": 2.0695,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.867193222045898,
+      "learning_rate": 2.9809507465247983e-05,
+      "loss": 1.9182,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.365927696228027,
+      "learning_rate": 2.980693324180539e-05,
+      "loss": 1.9339,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.1305570602417,
+      "learning_rate": 2.9804359018362793e-05,
+      "loss": 2.1178,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.789942741394043,
+      "learning_rate": 2.98017847949202e-05,
+      "loss": 1.8076,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.080610275268555,
+      "learning_rate": 2.9799210571477603e-05,
+      "loss": 2.0405,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.912185668945312,
+      "learning_rate": 2.979663634803501e-05,
+      "loss": 1.9494,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.001837730407715,
+      "learning_rate": 2.9794062124592413e-05,
+      "loss": 2.0811,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.438093185424805,
+      "learning_rate": 2.9791487901149823e-05,
+      "loss": 2.081,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.761951446533203,
+      "learning_rate": 2.9788913677707227e-05,
+      "loss": 1.9545,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.435627937316895,
+      "learning_rate": 2.9786339454264633e-05,
+      "loss": 1.9778,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.932748794555664,
+      "learning_rate": 2.9783765230822037e-05,
+      "loss": 2.2337,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.060193061828613,
+      "learning_rate": 2.978119100737944e-05,
+      "loss": 1.9319,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.534597396850586,
+      "learning_rate": 2.9778616783936847e-05,
+      "loss": 2.0977,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.02622127532959,
+      "learning_rate": 2.977604256049425e-05,
+      "loss": 2.1778,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.81395149230957,
+      "learning_rate": 2.9773468337051657e-05,
+      "loss": 1.8546,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.430489540100098,
+      "learning_rate": 2.977089411360906e-05,
+      "loss": 2.142,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.661596298217773,
+      "learning_rate": 2.976831989016647e-05,
+      "loss": 1.9116,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.51032829284668,
+      "learning_rate": 2.9765745666723873e-05,
+      "loss": 1.8959,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.853291511535645,
+      "learning_rate": 2.976317144328128e-05,
+      "loss": 1.9983,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.778831481933594,
+      "learning_rate": 2.9760597219838683e-05,
+      "loss": 1.8415,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.720138549804688,
+      "learning_rate": 2.9758022996396087e-05,
+      "loss": 2.322,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.475862503051758,
+      "learning_rate": 2.9755448772953493e-05,
+      "loss": 2.1695,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.582047462463379,
+      "learning_rate": 2.9752874549510896e-05,
+      "loss": 2.1263,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.227659225463867,
+      "learning_rate": 2.9750300326068303e-05,
+      "loss": 1.7871,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.419367790222168,
+      "learning_rate": 2.9747726102625706e-05,
+      "loss": 1.8047,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.549893379211426,
+      "learning_rate": 2.9745151879183113e-05,
+      "loss": 2.0386,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.57759666442871,
+      "learning_rate": 2.974257765574052e-05,
+      "loss": 1.9337,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.29601764678955,
+      "learning_rate": 2.9740003432297927e-05,
+      "loss": 1.6766,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.312577247619629,
+      "learning_rate": 2.973742920885533e-05,
+      "loss": 2.0605,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.877716064453125,
+      "learning_rate": 2.9734854985412736e-05,
+      "loss": 2.0258,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.774568557739258,
+      "learning_rate": 2.973228076197014e-05,
+      "loss": 1.8771,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.98751449584961,
+      "learning_rate": 2.9729706538527543e-05,
+      "loss": 2.1093,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.158930778503418,
+      "learning_rate": 2.972713231508495e-05,
+      "loss": 2.0506,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.581437110900879,
+      "learning_rate": 2.9724558091642353e-05,
+      "loss": 1.721,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.814128875732422,
+      "learning_rate": 2.972198386819976e-05,
+      "loss": 1.9709,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.382211685180664,
+      "learning_rate": 2.9719409644757166e-05,
+      "loss": 1.759,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.167340278625488,
+      "learning_rate": 2.9716835421314573e-05,
+      "loss": 1.9925,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.374336242675781,
+      "learning_rate": 2.9714261197871976e-05,
+      "loss": 2.5769,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.552855491638184,
+      "learning_rate": 2.9711686974429383e-05,
+      "loss": 2.045,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.009696006774902,
+      "learning_rate": 2.9709112750986786e-05,
+      "loss": 2.0438,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.150276184082031,
+      "learning_rate": 2.9706538527544193e-05,
+      "loss": 1.7308,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.456452369689941,
+      "learning_rate": 2.9703964304101596e-05,
+      "loss": 1.8647,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.520078659057617,
+      "learning_rate": 2.9701390080659e-05,
+      "loss": 1.7663,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.088191032409668,
+      "learning_rate": 2.9698815857216406e-05,
+      "loss": 1.6428,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.780815124511719,
+      "learning_rate": 2.969624163377381e-05,
+      "loss": 2.1887,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.029000282287598,
+      "learning_rate": 2.969366741033122e-05,
+      "loss": 1.9132,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.509160995483398,
+      "learning_rate": 2.9691093186888623e-05,
+      "loss": 1.9838,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.37309741973877,
+      "learning_rate": 2.968851896344603e-05,
+      "loss": 1.9265,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.40943717956543,
+      "learning_rate": 2.9685944740003433e-05,
+      "loss": 1.9344,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.383122444152832,
+      "learning_rate": 2.968337051656084e-05,
+      "loss": 2.1259,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.58685302734375,
+      "learning_rate": 2.9680796293118243e-05,
+      "loss": 1.7329,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.93775749206543,
+      "learning_rate": 2.9678222069675646e-05,
+      "loss": 1.7628,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.446601867675781,
+      "learning_rate": 2.9675647846233053e-05,
+      "loss": 1.9658,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.44230842590332,
+      "learning_rate": 2.9673073622790456e-05,
+      "loss": 2.2834,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.857770919799805,
+      "learning_rate": 2.9670499399347866e-05,
+      "loss": 1.8792,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.195531845092773,
+      "learning_rate": 2.966792517590527e-05,
+      "loss": 1.8233,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.215664863586426,
+      "learning_rate": 2.9665350952462676e-05,
+      "loss": 1.6713,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.225882530212402,
+      "learning_rate": 2.966277672902008e-05,
+      "loss": 2.454,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.208149909973145,
+      "learning_rate": 2.9660202505577486e-05,
+      "loss": 1.9905,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.276273727416992,
+      "learning_rate": 2.965762828213489e-05,
+      "loss": 1.9253,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.66161823272705,
+      "learning_rate": 2.9655054058692296e-05,
+      "loss": 1.4329,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.97610855102539,
+      "learning_rate": 2.96524798352497e-05,
+      "loss": 1.9271,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.867330551147461,
+      "learning_rate": 2.9649905611807103e-05,
+      "loss": 1.4213,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.628447532653809,
+      "learning_rate": 2.964733138836451e-05,
+      "loss": 2.0446,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.068493843078613,
+      "learning_rate": 2.9644757164921916e-05,
+      "loss": 1.8949,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.425756454467773,
+      "learning_rate": 2.9642182941479323e-05,
+      "loss": 1.8883,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.106772422790527,
+      "learning_rate": 2.9639608718036726e-05,
+      "loss": 1.6391,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.371277809143066,
+      "learning_rate": 2.9637034494594133e-05,
+      "loss": 1.836,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.329122543334961,
+      "learning_rate": 2.9634460271151536e-05,
+      "loss": 1.7353,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.681215286254883,
+      "learning_rate": 2.9631886047708943e-05,
+      "loss": 1.5368,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.243221282958984,
+      "learning_rate": 2.9629311824266346e-05,
+      "loss": 2.144,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.58792781829834,
+      "learning_rate": 2.9626737600823753e-05,
+      "loss": 1.6091,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.012331008911133,
+      "learning_rate": 2.9624163377381156e-05,
+      "loss": 1.7524,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.462698936462402,
+      "learning_rate": 2.9621589153938563e-05,
+      "loss": 2.0098,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.41120433807373,
+      "learning_rate": 2.961901493049597e-05,
+      "loss": 2.0257,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.883111000061035,
+      "learning_rate": 2.9616440707053373e-05,
+      "loss": 1.8638,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.62372875213623,
+      "learning_rate": 2.961386648361078e-05,
+      "loss": 2.052,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.91905403137207,
+      "learning_rate": 2.9611292260168183e-05,
+      "loss": 2.2381,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.769111633300781,
+      "learning_rate": 2.960871803672559e-05,
+      "loss": 1.7367,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.132026672363281,
+      "learning_rate": 2.9606143813282993e-05,
+      "loss": 1.9155,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.776375770568848,
+      "learning_rate": 2.96035695898404e-05,
+      "loss": 2.0896,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.961363792419434,
+      "learning_rate": 2.9600995366397803e-05,
+      "loss": 2.0811,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.618287086486816,
+      "learning_rate": 2.959842114295521e-05,
+      "loss": 2.0352,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.396384239196777,
+      "learning_rate": 2.9595846919512616e-05,
+      "loss": 1.677,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.013529777526855,
+      "learning_rate": 2.959327269607002e-05,
+      "loss": 1.8562,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.963462829589844,
+      "learning_rate": 2.9590698472627426e-05,
+      "loss": 1.9417,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.495522499084473,
+      "learning_rate": 2.958812424918483e-05,
+      "loss": 1.5038,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.643668174743652,
+      "learning_rate": 2.9585550025742236e-05,
+      "loss": 1.6444,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.734110832214355,
+      "learning_rate": 2.958297580229964e-05,
+      "loss": 1.7968,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.823417663574219,
+      "learning_rate": 2.9580401578857046e-05,
+      "loss": 1.6856,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.06626033782959,
+      "learning_rate": 2.957782735541445e-05,
+      "loss": 1.6747,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.7233304977417,
+      "learning_rate": 2.9575253131971856e-05,
+      "loss": 1.5999,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.944986343383789,
+      "learning_rate": 2.9572678908529263e-05,
+      "loss": 1.79,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.763249397277832,
+      "learning_rate": 2.9570104685086666e-05,
+      "loss": 1.6137,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.439078330993652,
+      "learning_rate": 2.9567530461644073e-05,
+      "loss": 1.5438,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.983046531677246,
+      "learning_rate": 2.9564956238201476e-05,
+      "loss": 1.9339,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.30870532989502,
+      "learning_rate": 2.9562382014758882e-05,
+      "loss": 1.6122,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.260065078735352,
+      "learning_rate": 2.9559807791316286e-05,
+      "loss": 2.0262,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.255437850952148,
+      "learning_rate": 2.9557233567873692e-05,
+      "loss": 1.545,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.572342872619629,
+      "learning_rate": 2.9554659344431096e-05,
+      "loss": 1.9368,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.593486785888672,
+      "learning_rate": 2.9552085120988502e-05,
+      "loss": 1.9177,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.95742416381836,
+      "learning_rate": 2.954951089754591e-05,
+      "loss": 2.0184,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.451391220092773,
+      "learning_rate": 2.9546936674103316e-05,
+      "loss": 1.7796,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.525118827819824,
+      "learning_rate": 2.954436245066072e-05,
+      "loss": 1.854,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.262989044189453,
+      "learning_rate": 2.9541788227218122e-05,
+      "loss": 1.7874,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.921760559082031,
+      "learning_rate": 2.953921400377553e-05,
+      "loss": 1.959,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.572196006774902,
+      "learning_rate": 2.9536639780332932e-05,
+      "loss": 1.523,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.206779479980469,
+      "learning_rate": 2.953406555689034e-05,
+      "loss": 2.1411,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.821024894714355,
+      "learning_rate": 2.9531491333447742e-05,
+      "loss": 2.0197,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.028032302856445,
+      "learning_rate": 2.952891711000515e-05,
+      "loss": 1.457,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.02061653137207,
+      "learning_rate": 2.9526342886562552e-05,
+      "loss": 1.7391,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.09890079498291,
+      "learning_rate": 2.9523768663119962e-05,
+      "loss": 1.9001,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.106844902038574,
+      "learning_rate": 2.9521194439677366e-05,
+      "loss": 1.7622,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.670572280883789,
+      "learning_rate": 2.9518620216234772e-05,
+      "loss": 2.1797,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.153185844421387,
+      "learning_rate": 2.9516045992792176e-05,
+      "loss": 1.986,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.073293685913086,
+      "learning_rate": 2.951347176934958e-05,
+      "loss": 1.8458,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.061956405639648,
+      "learning_rate": 2.9510897545906986e-05,
+      "loss": 1.8076,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.938712120056152,
+      "learning_rate": 2.950832332246439e-05,
+      "loss": 1.7217,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.107938766479492,
+      "learning_rate": 2.9505749099021796e-05,
+      "loss": 1.8765,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.666096687316895,
+      "learning_rate": 2.95031748755792e-05,
+      "loss": 1.7959,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.081636428833008,
+      "learning_rate": 2.950060065213661e-05,
+      "loss": 1.705,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.058525085449219,
+      "learning_rate": 2.9498026428694012e-05,
+      "loss": 1.236,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.916040420532227,
+      "learning_rate": 2.949545220525142e-05,
+      "loss": 1.6955,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.877468109130859,
+      "learning_rate": 2.9492877981808822e-05,
+      "loss": 1.6584,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.352563858032227,
+      "learning_rate": 2.9490303758366226e-05,
+      "loss": 1.9344,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.303910255432129,
+      "learning_rate": 2.9487729534923632e-05,
+      "loss": 1.662,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.003342628479004,
+      "learning_rate": 2.9485155311481035e-05,
+      "loss": 1.9813,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.470848083496094,
+      "learning_rate": 2.9482581088038442e-05,
+      "loss": 2.0612,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.331122398376465,
+      "learning_rate": 2.9480006864595845e-05,
+      "loss": 1.6463,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.50797176361084,
+      "learning_rate": 2.9477432641153252e-05,
+      "loss": 1.3334,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.60113525390625,
+      "learning_rate": 2.947485841771066e-05,
+      "loss": 1.7248,
+      "step": 204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.610932350158691,
+      "learning_rate": 2.9472284194268066e-05,
+      "loss": 1.6779,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.794546127319336,
+      "learning_rate": 2.946970997082547e-05,
+      "loss": 1.7341,
+      "step": 206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.368892669677734,
+      "learning_rate": 2.9467135747382875e-05,
+      "loss": 1.9062,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.087629318237305,
+      "learning_rate": 2.946456152394028e-05,
+      "loss": 1.9048,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.668496131896973,
+      "learning_rate": 2.9461987300497682e-05,
+      "loss": 1.7576,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.209294319152832,
+      "learning_rate": 2.945941307705509e-05,
+      "loss": 1.9772,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.424263954162598,
+      "learning_rate": 2.9456838853612492e-05,
+      "loss": 1.6316,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.347501754760742,
+      "learning_rate": 2.94542646301699e-05,
+      "loss": 1.6466,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.764115333557129,
+      "learning_rate": 2.9451690406727305e-05,
+      "loss": 1.7222,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.490241050720215,
+      "learning_rate": 2.9449116183284712e-05,
+      "loss": 2.0773,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.803098678588867,
+      "learning_rate": 2.9446541959842115e-05,
+      "loss": 1.6786,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.36636734008789,
+      "learning_rate": 2.9443967736399522e-05,
+      "loss": 1.9252,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.605925559997559,
+      "learning_rate": 2.9441393512956925e-05,
+      "loss": 1.5526,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.705082893371582,
+      "learning_rate": 2.9438819289514332e-05,
+      "loss": 1.4581,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.624661445617676,
+      "learning_rate": 2.9436245066071735e-05,
+      "loss": 2.0116,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.22999382019043,
+      "learning_rate": 2.943367084262914e-05,
+      "loss": 1.5265,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.914277076721191,
+      "learning_rate": 2.9431096619186545e-05,
+      "loss": 1.633,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.711101531982422,
+      "learning_rate": 2.942852239574395e-05,
+      "loss": 1.7177,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.378776550292969,
+      "learning_rate": 2.942594817230136e-05,
+      "loss": 1.7416,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.701340675354004,
+      "learning_rate": 2.9423373948858762e-05,
+      "loss": 1.9424,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.773192405700684,
+      "learning_rate": 2.942079972541617e-05,
+      "loss": 1.6253,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.05998706817627,
+      "learning_rate": 2.9418225501973572e-05,
+      "loss": 1.4659,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.306830406188965,
+      "learning_rate": 2.941565127853098e-05,
+      "loss": 1.7184,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.687745094299316,
+      "learning_rate": 2.9413077055088382e-05,
+      "loss": 1.5897,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.609406471252441,
+      "learning_rate": 2.9410502831645785e-05,
+      "loss": 2.0312,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.43237018585205,
+      "learning_rate": 2.9407928608203192e-05,
+      "loss": 1.9321,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.16401195526123,
+      "learning_rate": 2.9405354384760595e-05,
+      "loss": 1.7838,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.92502212524414,
+      "learning_rate": 2.9402780161318005e-05,
+      "loss": 1.3907,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.83435344696045,
+      "learning_rate": 2.940020593787541e-05,
+      "loss": 1.6566,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.208948135375977,
+      "learning_rate": 2.9397631714432815e-05,
+      "loss": 2.1311,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.743586540222168,
+      "learning_rate": 2.939505749099022e-05,
+      "loss": 1.9852,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.744789123535156,
+      "learning_rate": 2.9392483267547625e-05,
+      "loss": 1.4427,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.780157089233398,
+      "learning_rate": 2.938990904410503e-05,
+      "loss": 1.4846,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.985099792480469,
+      "learning_rate": 2.9387334820662435e-05,
+      "loss": 1.3918,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.792673587799072,
+      "learning_rate": 2.938476059721984e-05,
+      "loss": 1.5816,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.847929954528809,
+      "learning_rate": 2.9382186373777242e-05,
+      "loss": 1.9881,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.681547164916992,
+      "learning_rate": 2.937961215033465e-05,
+      "loss": 1.5669,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.897395133972168,
+      "learning_rate": 2.9377037926892055e-05,
+      "loss": 2.1757,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.938973426818848,
+      "learning_rate": 2.9374463703449462e-05,
+      "loss": 1.4375,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.108646392822266,
+      "learning_rate": 2.9371889480006865e-05,
+      "loss": 1.6926,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.904474258422852,
+      "learning_rate": 2.9369315256564272e-05,
+      "loss": 1.3446,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.004955291748047,
+      "learning_rate": 2.9366741033121675e-05,
+      "loss": 1.4969,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.158055305480957,
+      "learning_rate": 2.9364166809679082e-05,
+      "loss": 2.0364,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.525121688842773,
+      "learning_rate": 2.9361592586236485e-05,
+      "loss": 1.7299,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.56649398803711,
+      "learning_rate": 2.9359018362793892e-05,
+      "loss": 1.7901,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.06497573852539,
+      "learning_rate": 2.9356444139351295e-05,
+      "loss": 2.3591,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.082416534423828,
+      "learning_rate": 2.9353869915908702e-05,
+      "loss": 1.9583,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.90018892288208,
+      "learning_rate": 2.935129569246611e-05,
+      "loss": 1.4908,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.350253105163574,
+      "learning_rate": 2.934872146902351e-05,
+      "loss": 1.7432,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.411694526672363,
+      "learning_rate": 2.934614724558092e-05,
+      "loss": 1.7825,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.787676811218262,
+      "learning_rate": 2.934357302213832e-05,
+      "loss": 1.9696,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.568073749542236,
+      "learning_rate": 2.934099879869573e-05,
+      "loss": 1.5693,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.90737247467041,
+      "learning_rate": 2.933842457525313e-05,
+      "loss": 1.56,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.960163116455078,
+      "learning_rate": 2.933585035181054e-05,
+      "loss": 1.6016,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.698493957519531,
+      "learning_rate": 2.933327612836794e-05,
+      "loss": 1.5217,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.947373390197754,
+      "learning_rate": 2.9330701904925348e-05,
+      "loss": 1.9468,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.384953498840332,
+      "learning_rate": 2.9328127681482755e-05,
+      "loss": 1.3466,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.936210632324219,
+      "learning_rate": 2.9325553458040158e-05,
+      "loss": 2.0577,
+      "step": 262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.217910766601562,
+      "learning_rate": 2.9322979234597565e-05,
+      "loss": 1.7381,
+      "step": 263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.802103996276855,
+      "learning_rate": 2.9320405011154968e-05,
+      "loss": 1.6338,
+      "step": 264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.834925651550293,
+      "learning_rate": 2.9317830787712375e-05,
+      "loss": 1.7873,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.63886547088623,
+      "learning_rate": 2.9315256564269778e-05,
+      "loss": 1.5055,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.555371284484863,
+      "learning_rate": 2.9312682340827185e-05,
+      "loss": 1.7096,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.70984172821045,
+      "learning_rate": 2.9310108117384588e-05,
+      "loss": 1.76,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.787165641784668,
+      "learning_rate": 2.9307533893941995e-05,
+      "loss": 1.5348,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.777557373046875,
+      "learning_rate": 2.93049596704994e-05,
+      "loss": 1.8856,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.017600059509277,
+      "learning_rate": 2.9302385447056805e-05,
+      "loss": 1.8646,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.701334953308105,
+      "learning_rate": 2.929981122361421e-05,
+      "loss": 1.531,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.639413833618164,
+      "learning_rate": 2.9297237000171615e-05,
+      "loss": 1.962,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.560542106628418,
+      "learning_rate": 2.929466277672902e-05,
+      "loss": 1.7777,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.680726051330566,
+      "learning_rate": 2.9292088553286425e-05,
+      "loss": 1.5241,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.093884468078613,
+      "learning_rate": 2.928951432984383e-05,
+      "loss": 1.6674,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.894911289215088,
+      "learning_rate": 2.9286940106401235e-05,
+      "loss": 1.5643,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.18982982635498,
+      "learning_rate": 2.928436588295864e-05,
+      "loss": 1.4667,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.561273574829102,
+      "learning_rate": 2.9281791659516045e-05,
+      "loss": 1.617,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.593453407287598,
+      "learning_rate": 2.9279217436073455e-05,
+      "loss": 1.6699,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.71127986907959,
+      "learning_rate": 2.9276643212630858e-05,
+      "loss": 1.6506,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.987764358520508,
+      "learning_rate": 2.927406898918826e-05,
+      "loss": 1.4503,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.146617889404297,
+      "learning_rate": 2.9271494765745668e-05,
+      "loss": 1.686,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.64239501953125,
+      "learning_rate": 2.926892054230307e-05,
+      "loss": 1.9325,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.928953170776367,
+      "learning_rate": 2.9266346318860478e-05,
+      "loss": 1.6489,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.513096809387207,
+      "learning_rate": 2.926377209541788e-05,
+      "loss": 1.7969,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.037224769592285,
+      "learning_rate": 2.9261197871975288e-05,
+      "loss": 1.6694,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.013139724731445,
+      "learning_rate": 2.925862364853269e-05,
+      "loss": 1.7193,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.478838920593262,
+      "learning_rate": 2.92560494250901e-05,
+      "loss": 1.4296,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.771774291992188,
+      "learning_rate": 2.9253475201647505e-05,
+      "loss": 1.8696,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.5445556640625,
+      "learning_rate": 2.925090097820491e-05,
+      "loss": 1.4061,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.020146369934082,
+      "learning_rate": 2.9248326754762315e-05,
+      "loss": 1.6945,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.573962211608887,
+      "learning_rate": 2.9245752531319718e-05,
+      "loss": 1.6359,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.782485008239746,
+      "learning_rate": 2.9243178307877125e-05,
+      "loss": 1.6955,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.277192115783691,
+      "learning_rate": 2.9240604084434528e-05,
+      "loss": 1.5417,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.727743148803711,
+      "learning_rate": 2.9238029860991935e-05,
+      "loss": 1.4512,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.91127872467041,
+      "learning_rate": 2.9235455637549338e-05,
+      "loss": 1.7006,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.304641723632812,
+      "learning_rate": 2.9232881414106748e-05,
+      "loss": 1.4382,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.38481616973877,
+      "learning_rate": 2.923030719066415e-05,
+      "loss": 1.5406,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.092000007629395,
+      "learning_rate": 2.9227732967221558e-05,
+      "loss": 1.5547,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.8936767578125,
+      "learning_rate": 2.922515874377896e-05,
+      "loss": 1.9512,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.751150131225586,
+      "learning_rate": 2.9222584520336365e-05,
+      "loss": 1.6481,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.386673927307129,
+      "learning_rate": 2.922001029689377e-05,
+      "loss": 1.7778,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.260998725891113,
+      "learning_rate": 2.9217436073451175e-05,
+      "loss": 1.4244,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.2503080368042,
+      "learning_rate": 2.921486185000858e-05,
+      "loss": 1.575,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.5574369430542,
+      "learning_rate": 2.9212287626565984e-05,
+      "loss": 1.6031,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.495893478393555,
+      "learning_rate": 2.920971340312339e-05,
+      "loss": 1.7797,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.136427879333496,
+      "learning_rate": 2.9207139179680798e-05,
+      "loss": 1.4174,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.032553672790527,
+      "learning_rate": 2.9204564956238205e-05,
+      "loss": 1.6757,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.711871147155762,
+      "learning_rate": 2.9201990732795608e-05,
+      "loss": 1.8516,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.710490226745605,
+      "learning_rate": 2.9199416509353015e-05,
+      "loss": 1.6837,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.25143814086914,
+      "learning_rate": 2.9196842285910418e-05,
+      "loss": 1.5361,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.08812427520752,
+      "learning_rate": 2.919426806246782e-05,
+      "loss": 1.3348,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.16710090637207,
+      "learning_rate": 2.9191693839025228e-05,
+      "loss": 1.3417,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.723457336425781,
+      "learning_rate": 2.918911961558263e-05,
+      "loss": 1.7711,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.169995307922363,
+      "learning_rate": 2.9186545392140038e-05,
+      "loss": 1.6362,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.744826316833496,
+      "learning_rate": 2.9183971168697444e-05,
+      "loss": 2.0765,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.706212997436523,
+      "learning_rate": 2.918139694525485e-05,
+      "loss": 1.6454,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.484731674194336,
+      "learning_rate": 2.9178822721812254e-05,
+      "loss": 1.7635,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.310269355773926,
+      "learning_rate": 2.917624849836966e-05,
+      "loss": 1.519,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.98282527923584,
+      "learning_rate": 2.9173674274927064e-05,
+      "loss": 2.0391,
+      "step": 321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.362037658691406,
+      "learning_rate": 2.917110005148447e-05,
+      "loss": 1.8951,
+      "step": 322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.465320587158203,
+      "learning_rate": 2.9168525828041874e-05,
+      "loss": 1.6288,
+      "step": 323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.646321296691895,
+      "learning_rate": 2.9165951604599278e-05,
+      "loss": 1.6041,
+      "step": 324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.398006439208984,
+      "learning_rate": 2.9163377381156684e-05,
+      "loss": 1.7292,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.411280632019043,
+      "learning_rate": 2.9160803157714088e-05,
+      "loss": 1.7537,
+      "step": 326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.08098316192627,
+      "learning_rate": 2.9158228934271498e-05,
+      "loss": 2.0472,
+      "step": 327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.14523983001709,
+      "learning_rate": 2.91556547108289e-05,
+      "loss": 1.6499,
+      "step": 328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.763137817382812,
+      "learning_rate": 2.9153080487386308e-05,
+      "loss": 1.5005,
+      "step": 329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.802804946899414,
+      "learning_rate": 2.915050626394371e-05,
+      "loss": 1.2987,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.917118072509766,
+      "learning_rate": 2.9147932040501118e-05,
+      "loss": 1.7462,
+      "step": 331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.044041633605957,
+      "learning_rate": 2.914535781705852e-05,
+      "loss": 1.9712,
+      "step": 332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.21104621887207,
+      "learning_rate": 2.9142783593615928e-05,
+      "loss": 1.5639,
+      "step": 333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.971760749816895,
+      "learning_rate": 2.914020937017333e-05,
+      "loss": 1.4879,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.787034034729004,
+      "learning_rate": 2.9137635146730734e-05,
+      "loss": 1.481,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.895040512084961,
+      "learning_rate": 2.9135060923288144e-05,
+      "loss": 1.4038,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.115073204040527,
+      "learning_rate": 2.9132486699845548e-05,
+      "loss": 1.4964,
+      "step": 337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.226001739501953,
+      "learning_rate": 2.9129912476402954e-05,
+      "loss": 1.5353,
+      "step": 338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.958303451538086,
+      "learning_rate": 2.9127338252960358e-05,
+      "loss": 1.6507,
+      "step": 339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.76336669921875,
+      "learning_rate": 2.9124764029517764e-05,
+      "loss": 1.7615,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.854474067687988,
+      "learning_rate": 2.9122189806075168e-05,
+      "loss": 1.675,
+      "step": 341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.663541793823242,
+      "learning_rate": 2.9119615582632574e-05,
+      "loss": 1.5917,
+      "step": 342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.632105350494385,
+      "learning_rate": 2.9117041359189977e-05,
+      "loss": 1.8109,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.303193092346191,
+      "learning_rate": 2.911446713574738e-05,
+      "loss": 1.5036,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.947858810424805,
+      "learning_rate": 2.9111892912304787e-05,
+      "loss": 1.5195,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.142110824584961,
+      "learning_rate": 2.9109318688862194e-05,
+      "loss": 1.5374,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.125122547149658,
+      "learning_rate": 2.91067444654196e-05,
+      "loss": 1.1923,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.325124740600586,
+      "learning_rate": 2.9104170241977004e-05,
+      "loss": 1.8373,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.541301727294922,
+      "learning_rate": 2.910159601853441e-05,
+      "loss": 1.428,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.802016258239746,
+      "learning_rate": 2.9099021795091814e-05,
+      "loss": 1.5451,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.415572166442871,
+      "learning_rate": 2.909644757164922e-05,
+      "loss": 1.5812,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.448362350463867,
+      "learning_rate": 2.9093873348206624e-05,
+      "loss": 1.7371,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.94411849975586,
+      "learning_rate": 2.909129912476403e-05,
+      "loss": 1.5187,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.400604248046875,
+      "learning_rate": 2.9088724901321434e-05,
+      "loss": 1.6639,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.481334686279297,
+      "learning_rate": 2.908615067787884e-05,
+      "loss": 1.515,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.073720932006836,
+      "learning_rate": 2.9083576454436247e-05,
+      "loss": 1.6674,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.867945671081543,
+      "learning_rate": 2.908100223099365e-05,
+      "loss": 1.9646,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.329514503479004,
+      "learning_rate": 2.9078428007551057e-05,
+      "loss": 1.8233,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.168272972106934,
+      "learning_rate": 2.907585378410846e-05,
+      "loss": 1.6868,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.724711418151855,
+      "learning_rate": 2.9073279560665867e-05,
+      "loss": 1.644,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.938602447509766,
+      "learning_rate": 2.907070533722327e-05,
+      "loss": 1.5774,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.528223037719727,
+      "learning_rate": 2.9068131113780677e-05,
+      "loss": 1.4291,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.6717529296875,
+      "learning_rate": 2.906555689033808e-05,
+      "loss": 1.5272,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.576579093933105,
+      "learning_rate": 2.9062982666895487e-05,
+      "loss": 1.3728,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.873361587524414,
+      "learning_rate": 2.9060408443452894e-05,
+      "loss": 1.8679,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.140089988708496,
+      "learning_rate": 2.9057834220010297e-05,
+      "loss": 1.4819,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.3444185256958,
+      "learning_rate": 2.9055259996567704e-05,
+      "loss": 1.7938,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.555766105651855,
+      "learning_rate": 2.9052685773125107e-05,
+      "loss": 1.586,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.24820613861084,
+      "learning_rate": 2.9050111549682514e-05,
+      "loss": 1.4957,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.031588554382324,
+      "learning_rate": 2.9047537326239917e-05,
+      "loss": 1.7064,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.380867958068848,
+      "learning_rate": 2.9044963102797324e-05,
+      "loss": 1.4169,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.075194358825684,
+      "learning_rate": 2.9042388879354727e-05,
+      "loss": 1.5741,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.459732055664062,
+      "learning_rate": 2.9039814655912134e-05,
+      "loss": 1.3803,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.930313110351562,
+      "learning_rate": 2.903724043246954e-05,
+      "loss": 1.6806,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.009334564208984,
+      "learning_rate": 2.9034666209026944e-05,
+      "loss": 1.8382,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.107048988342285,
+      "learning_rate": 2.903209198558435e-05,
+      "loss": 1.706,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.72223949432373,
+      "learning_rate": 2.9029517762141754e-05,
+      "loss": 1.5575,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.12093448638916,
+      "learning_rate": 2.902694353869916e-05,
+      "loss": 1.8506,
+      "step": 378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.984782218933105,
+      "learning_rate": 2.9024369315256564e-05,
+      "loss": 1.6462,
+      "step": 379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.998955726623535,
+      "learning_rate": 2.902179509181397e-05,
+      "loss": 1.5175,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.353002548217773,
+      "learning_rate": 2.9019220868371374e-05,
+      "loss": 1.3723,
+      "step": 381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.063586235046387,
+      "learning_rate": 2.901664664492878e-05,
+      "loss": 1.5693,
+      "step": 382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.760736465454102,
+      "learning_rate": 2.9014072421486184e-05,
+      "loss": 1.4859,
+      "step": 383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.812806129455566,
+      "learning_rate": 2.9011498198043594e-05,
+      "loss": 1.4088,
+      "step": 384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.350480079650879,
+      "learning_rate": 2.9008923974600997e-05,
+      "loss": 1.5996,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.718733787536621,
+      "learning_rate": 2.90063497511584e-05,
+      "loss": 1.232,
+      "step": 386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.795390605926514,
+      "learning_rate": 2.9003775527715807e-05,
+      "loss": 1.2631,
+      "step": 387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.643641471862793,
+      "learning_rate": 2.900120130427321e-05,
+      "loss": 1.6076,
+      "step": 388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.053388595581055,
+      "learning_rate": 2.8998627080830617e-05,
+      "loss": 1.6208,
+      "step": 389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.285635948181152,
+      "learning_rate": 2.899605285738802e-05,
+      "loss": 1.9027,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.512655258178711,
+      "learning_rate": 2.8993478633945427e-05,
+      "loss": 1.6828,
+      "step": 391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.464376449584961,
+      "learning_rate": 2.899090441050283e-05,
+      "loss": 1.2634,
+      "step": 392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.082260131835938,
+      "learning_rate": 2.898833018706024e-05,
+      "loss": 1.3558,
+      "step": 393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.868090629577637,
+      "learning_rate": 2.8985755963617644e-05,
+      "loss": 1.8094,
+      "step": 394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.074893951416016,
+      "learning_rate": 2.898318174017505e-05,
+      "loss": 1.5454,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.83586597442627,
+      "learning_rate": 2.8980607516732454e-05,
+      "loss": 1.6831,
+      "step": 396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.7890043258667,
+      "learning_rate": 2.8978033293289857e-05,
+      "loss": 1.3461,
+      "step": 397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.520912170410156,
+      "learning_rate": 2.8975459069847264e-05,
+      "loss": 1.6014,
+      "step": 398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.111466407775879,
+      "learning_rate": 2.8972884846404667e-05,
+      "loss": 1.8845,
+      "step": 399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.061944007873535,
+      "learning_rate": 2.8970310622962074e-05,
+      "loss": 1.7318,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.755029678344727,
+      "learning_rate": 2.8967736399519477e-05,
+      "loss": 1.425,
+      "step": 401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.505181789398193,
+      "learning_rate": 2.8965162176076884e-05,
+      "loss": 1.4305,
+      "step": 402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.357077598571777,
+      "learning_rate": 2.896258795263429e-05,
+      "loss": 1.6816,
+      "step": 403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.416088104248047,
+      "learning_rate": 2.8960013729191697e-05,
+      "loss": 1.7239,
+      "step": 404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.75820541381836,
+      "learning_rate": 2.89574395057491e-05,
+      "loss": 1.6036,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.285774230957031,
+      "learning_rate": 2.8954865282306504e-05,
+      "loss": 1.2902,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.048407554626465,
+      "learning_rate": 2.895229105886391e-05,
+      "loss": 1.5071,
+      "step": 407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.721714019775391,
+      "learning_rate": 2.8949716835421314e-05,
+      "loss": 1.7365,
+      "step": 408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.985815048217773,
+      "learning_rate": 2.894714261197872e-05,
+      "loss": 1.3089,
+      "step": 409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.920326232910156,
+      "learning_rate": 2.8944568388536123e-05,
+      "loss": 1.8378,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.613027572631836,
+      "learning_rate": 2.894199416509353e-05,
+      "loss": 1.2882,
+      "step": 411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.50846004486084,
+      "learning_rate": 2.8939419941650937e-05,
+      "loss": 1.7531,
+      "step": 412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.982809066772461,
+      "learning_rate": 2.8936845718208344e-05,
+      "loss": 1.6533,
+      "step": 413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.472440719604492,
+      "learning_rate": 2.8934271494765747e-05,
+      "loss": 1.4576,
+      "step": 414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.862814903259277,
+      "learning_rate": 2.8931697271323154e-05,
+      "loss": 1.2296,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.792984962463379,
+      "learning_rate": 2.8929123047880557e-05,
+      "loss": 1.4784,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.533136367797852,
+      "learning_rate": 2.892654882443796e-05,
+      "loss": 1.3715,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.728937149047852,
+      "learning_rate": 2.8923974600995367e-05,
+      "loss": 1.2633,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.582226753234863,
+      "learning_rate": 2.892140037755277e-05,
+      "loss": 1.4999,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.515429496765137,
+      "learning_rate": 2.8918826154110177e-05,
+      "loss": 1.8531,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.933034896850586,
+      "learning_rate": 2.8916251930667583e-05,
+      "loss": 2.1722,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.477798461914062,
+      "learning_rate": 2.891367770722499e-05,
+      "loss": 1.4052,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.258628845214844,
+      "learning_rate": 2.8911103483782393e-05,
+      "loss": 1.707,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.19088077545166,
+      "learning_rate": 2.89085292603398e-05,
+      "loss": 1.489,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.1199312210083,
+      "learning_rate": 2.8905955036897203e-05,
+      "loss": 1.4829,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.810800552368164,
+      "learning_rate": 2.890338081345461e-05,
+      "loss": 1.5978,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.57884407043457,
+      "learning_rate": 2.8900806590012013e-05,
+      "loss": 1.4805,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.866064071655273,
+      "learning_rate": 2.8898232366569417e-05,
+      "loss": 1.7046,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.306712627410889,
+      "learning_rate": 2.8895658143126823e-05,
+      "loss": 1.1738,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.592244148254395,
+      "learning_rate": 2.8893083919684227e-05,
+      "loss": 1.5762,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.549958229064941,
+      "learning_rate": 2.8890509696241637e-05,
+      "loss": 1.4738,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.925586700439453,
+      "learning_rate": 2.888793547279904e-05,
+      "loss": 1.6632,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.961637496948242,
+      "learning_rate": 2.8885361249356447e-05,
+      "loss": 1.6991,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.438910484313965,
+      "learning_rate": 2.888278702591385e-05,
+      "loss": 1.7705,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.384795188903809,
+      "learning_rate": 2.8880212802471257e-05,
+      "loss": 1.411,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.856956481933594,
+      "learning_rate": 2.887763857902866e-05,
+      "loss": 2.1184,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.427003860473633,
+      "learning_rate": 2.8875064355586067e-05,
+      "loss": 1.4459,
+      "step": 437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.227859497070312,
+      "learning_rate": 2.887249013214347e-05,
+      "loss": 1.6471,
+      "step": 438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.303044319152832,
+      "learning_rate": 2.8869915908700873e-05,
+      "loss": 1.3036,
+      "step": 439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.696854591369629,
+      "learning_rate": 2.8867341685258283e-05,
+      "loss": 1.5718,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.387533187866211,
+      "learning_rate": 2.8864767461815687e-05,
+      "loss": 1.596,
+      "step": 441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.539954662322998,
+      "learning_rate": 2.8862193238373093e-05,
+      "loss": 1.1893,
+      "step": 442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.909122467041016,
+      "learning_rate": 2.8859619014930497e-05,
+      "loss": 1.4737,
+      "step": 443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.117277145385742,
+      "learning_rate": 2.8857044791487903e-05,
+      "loss": 1.3799,
+      "step": 444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.359807968139648,
+      "learning_rate": 2.8854470568045307e-05,
+      "loss": 1.7931,
+      "step": 445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.885112762451172,
+      "learning_rate": 2.8851896344602713e-05,
+      "loss": 1.5852,
+      "step": 446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.114673614501953,
+      "learning_rate": 2.8849322121160116e-05,
+      "loss": 1.4774,
+      "step": 447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.589859962463379,
+      "learning_rate": 2.884674789771752e-05,
+      "loss": 1.6173,
+      "step": 448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.234646797180176,
+      "learning_rate": 2.8844173674274926e-05,
+      "loss": 1.2883,
+      "step": 449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.6058931350708,
+      "learning_rate": 2.8841599450832333e-05,
+      "loss": 1.2674,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.516396522521973,
+      "learning_rate": 2.883902522738974e-05,
+      "loss": 1.7395,
+      "step": 451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.401927947998047,
+      "learning_rate": 2.8836451003947143e-05,
+      "loss": 1.6409,
+      "step": 452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.46969223022461,
+      "learning_rate": 2.883387678050455e-05,
+      "loss": 0.9971,
+      "step": 453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.885191917419434,
+      "learning_rate": 2.8831302557061953e-05,
+      "loss": 1.2535,
+      "step": 454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.642657279968262,
+      "learning_rate": 2.882872833361936e-05,
+      "loss": 1.7177,
+      "step": 455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.326385498046875,
+      "learning_rate": 2.8826154110176763e-05,
+      "loss": 1.5923,
+      "step": 456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.094305992126465,
+      "learning_rate": 2.882357988673417e-05,
+      "loss": 1.5446,
+      "step": 457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.640613555908203,
+      "learning_rate": 2.8821005663291573e-05,
+      "loss": 1.4332,
+      "step": 458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.753673553466797,
+      "learning_rate": 2.881843143984898e-05,
+      "loss": 1.7025,
+      "step": 459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.737837791442871,
+      "learning_rate": 2.8815857216406386e-05,
+      "loss": 2.0308,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.723050117492676,
+      "learning_rate": 2.881328299296379e-05,
+      "loss": 1.5147,
+      "step": 461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.3440523147583,
+      "learning_rate": 2.8810708769521196e-05,
+      "loss": 1.5351,
+      "step": 462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.674872398376465,
+      "learning_rate": 2.88081345460786e-05,
+      "loss": 1.7827,
+      "step": 463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.43867301940918,
+      "learning_rate": 2.8805560322636006e-05,
+      "loss": 1.6664,
+      "step": 464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.447357177734375,
+      "learning_rate": 2.880298609919341e-05,
+      "loss": 1.5267,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.519311904907227,
+      "learning_rate": 2.8800411875750816e-05,
+      "loss": 1.6896,
+      "step": 466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.392447471618652,
+      "learning_rate": 2.879783765230822e-05,
+      "loss": 1.2946,
+      "step": 467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.106021881103516,
+      "learning_rate": 2.8795263428865626e-05,
+      "loss": 1.5775,
+      "step": 468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.461575508117676,
+      "learning_rate": 2.8792689205423033e-05,
+      "loss": 1.204,
+      "step": 469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.577383518218994,
+      "learning_rate": 2.8790114981980436e-05,
+      "loss": 1.5391,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.269767761230469,
+      "learning_rate": 2.8787540758537843e-05,
+      "loss": 2.0634,
+      "step": 471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.887826919555664,
+      "learning_rate": 2.8784966535095246e-05,
+      "loss": 1.4488,
+      "step": 472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.82689094543457,
+      "learning_rate": 2.8782392311652653e-05,
+      "loss": 2.0102,
+      "step": 473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.734695911407471,
+      "learning_rate": 2.8779818088210056e-05,
+      "loss": 1.5332,
+      "step": 474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.526495933532715,
+      "learning_rate": 2.8777243864767463e-05,
+      "loss": 1.5298,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.68381118774414,
+      "learning_rate": 2.8774669641324866e-05,
+      "loss": 1.8056,
+      "step": 476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.388910293579102,
+      "learning_rate": 2.8772095417882273e-05,
+      "loss": 1.6498,
+      "step": 477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.723257064819336,
+      "learning_rate": 2.876952119443968e-05,
+      "loss": 1.5671,
+      "step": 478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.094700813293457,
+      "learning_rate": 2.8766946970997083e-05,
+      "loss": 1.5728,
+      "step": 479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.833842754364014,
+      "learning_rate": 2.876437274755449e-05,
+      "loss": 1.2576,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.016430854797363,
+      "learning_rate": 2.8761798524111893e-05,
+      "loss": 1.7658,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.015495300292969,
+      "learning_rate": 2.87592243006693e-05,
+      "loss": 1.6063,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.249187469482422,
+      "learning_rate": 2.8756650077226703e-05,
+      "loss": 1.4659,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.637657165527344,
+      "learning_rate": 2.875407585378411e-05,
+      "loss": 1.4088,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.466156005859375,
+      "learning_rate": 2.8751501630341513e-05,
+      "loss": 1.4845,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.58089828491211,
+      "learning_rate": 2.874892740689892e-05,
+      "loss": 1.5108,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.993913650512695,
+      "learning_rate": 2.8746353183456323e-05,
+      "loss": 1.3799,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.493663787841797,
+      "learning_rate": 2.8743778960013733e-05,
+      "loss": 1.7722,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.064780235290527,
+      "learning_rate": 2.8741204736571136e-05,
+      "loss": 1.5231,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.072488784790039,
+      "learning_rate": 2.873863051312854e-05,
+      "loss": 1.385,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.39823055267334,
+      "learning_rate": 2.8736056289685946e-05,
+      "loss": 1.6999,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.070072174072266,
+      "learning_rate": 2.873348206624335e-05,
+      "loss": 1.579,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.037577629089355,
+      "learning_rate": 2.8730907842800756e-05,
+      "loss": 1.7168,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.199870109558105,
+      "learning_rate": 2.872833361935816e-05,
+      "loss": 1.4653,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.702022552490234,
+      "learning_rate": 2.8725759395915566e-05,
+      "loss": 1.2119,
+      "step": 495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.548079013824463,
+      "learning_rate": 2.872318517247297e-05,
+      "loss": 1.398,
+      "step": 496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.74194622039795,
+      "learning_rate": 2.872061094903038e-05,
+      "loss": 1.3153,
+      "step": 497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.39633560180664,
+      "learning_rate": 2.8718036725587783e-05,
+      "loss": 1.1622,
+      "step": 498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.131392478942871,
+      "learning_rate": 2.871546250214519e-05,
+      "loss": 1.4024,
+      "step": 499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.805859565734863,
+      "learning_rate": 2.8712888278702593e-05,
+      "loss": 1.2898,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.168538093566895,
+      "learning_rate": 2.8710314055259996e-05,
+      "loss": 1.7268,
+      "step": 501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.682873725891113,
+      "learning_rate": 2.8707739831817403e-05,
+      "loss": 1.4385,
+      "step": 502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.96394157409668,
+      "learning_rate": 2.8705165608374806e-05,
+      "loss": 1.3478,
+      "step": 503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.533995628356934,
+      "learning_rate": 2.8702591384932213e-05,
+      "loss": 1.664,
+      "step": 504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.12846851348877,
+      "learning_rate": 2.8700017161489616e-05,
+      "loss": 1.2708,
+      "step": 505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.074721336364746,
+      "learning_rate": 2.8697442938047023e-05,
+      "loss": 1.283,
+      "step": 506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.145303726196289,
+      "learning_rate": 2.869486871460443e-05,
+      "loss": 0.968,
+      "step": 507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.491410255432129,
+      "learning_rate": 2.8692294491161836e-05,
+      "loss": 1.7349,
+      "step": 508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.441530227661133,
+      "learning_rate": 2.868972026771924e-05,
+      "loss": 1.3327,
+      "step": 509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.813334465026855,
+      "learning_rate": 2.8687146044276643e-05,
+      "loss": 1.4541,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.257591247558594,
+      "learning_rate": 2.868457182083405e-05,
+      "loss": 1.3805,
+      "step": 511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.843669891357422,
+      "learning_rate": 2.8681997597391453e-05,
+      "loss": 1.6476,
+      "step": 512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.05841064453125,
+      "learning_rate": 2.867942337394886e-05,
+      "loss": 1.5867,
+      "step": 513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.566347122192383,
+      "learning_rate": 2.8676849150506263e-05,
+      "loss": 1.1905,
+      "step": 514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.416635513305664,
+      "learning_rate": 2.867427492706367e-05,
+      "loss": 1.5511,
+      "step": 515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.213944435119629,
+      "learning_rate": 2.8671700703621076e-05,
+      "loss": 1.5191,
+      "step": 516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.978853225708008,
+      "learning_rate": 2.8669126480178483e-05,
+      "loss": 1.5472,
+      "step": 517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.227170944213867,
+      "learning_rate": 2.8666552256735886e-05,
+      "loss": 1.5105,
+      "step": 518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.5068464279174805,
+      "learning_rate": 2.8663978033293293e-05,
+      "loss": 1.5845,
+      "step": 519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.044169425964355,
+      "learning_rate": 2.8661403809850696e-05,
+      "loss": 1.2448,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.292195320129395,
+      "learning_rate": 2.86588295864081e-05,
+      "loss": 1.5076,
+      "step": 521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.377120018005371,
+      "learning_rate": 2.8656255362965506e-05,
+      "loss": 1.4297,
+      "step": 522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.932923316955566,
+      "learning_rate": 2.865368113952291e-05,
+      "loss": 1.3458,
+      "step": 523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.574246406555176,
+      "learning_rate": 2.8651106916080316e-05,
+      "loss": 1.2264,
+      "step": 524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.966174125671387,
+      "learning_rate": 2.8648532692637722e-05,
+      "loss": 1.4927,
+      "step": 525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.199288368225098,
+      "learning_rate": 2.864595846919513e-05,
+      "loss": 1.7801,
+      "step": 526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.957565307617188,
+      "learning_rate": 2.8643384245752532e-05,
+      "loss": 1.4717,
+      "step": 527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.930829048156738,
+      "learning_rate": 2.864081002230994e-05,
+      "loss": 1.345,
+      "step": 528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.972787857055664,
+      "learning_rate": 2.8638235798867342e-05,
+      "loss": 1.5369,
+      "step": 529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.50515365600586,
+      "learning_rate": 2.863566157542475e-05,
+      "loss": 1.5552,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.620457649230957,
+      "learning_rate": 2.8633087351982152e-05,
+      "loss": 1.2706,
+      "step": 531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.689730644226074,
+      "learning_rate": 2.8630513128539556e-05,
+      "loss": 1.4852,
+      "step": 532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.077995300292969,
+      "learning_rate": 2.8627938905096962e-05,
+      "loss": 1.68,
+      "step": 533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.704209804534912,
+      "learning_rate": 2.8625364681654366e-05,
+      "loss": 1.5412,
+      "step": 534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.958584785461426,
+      "learning_rate": 2.8622790458211776e-05,
+      "loss": 1.772,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.211577415466309,
+      "learning_rate": 2.862021623476918e-05,
+      "loss": 1.6424,
+      "step": 536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.039546966552734,
+      "learning_rate": 2.8617642011326586e-05,
+      "loss": 1.4005,
+      "step": 537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.478930473327637,
+      "learning_rate": 2.861506778788399e-05,
+      "loss": 1.602,
+      "step": 538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.348405838012695,
+      "learning_rate": 2.8612493564441396e-05,
+      "loss": 1.4291,
+      "step": 539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.137516975402832,
+      "learning_rate": 2.86099193409988e-05,
+      "loss": 1.5088,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.610368728637695,
+      "learning_rate": 2.8607345117556206e-05,
+      "loss": 1.6462,
+      "step": 541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.408897399902344,
+      "learning_rate": 2.860477089411361e-05,
+      "loss": 1.5357,
+      "step": 542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.426880836486816,
+      "learning_rate": 2.8602196670671012e-05,
+      "loss": 1.4384,
+      "step": 543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.32236099243164,
+      "learning_rate": 2.8599622447228422e-05,
+      "loss": 1.5944,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.563417434692383,
+      "learning_rate": 2.8597048223785826e-05,
+      "loss": 1.4292,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.363273620605469,
+      "learning_rate": 2.8594474000343232e-05,
+      "loss": 1.3923,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.25059700012207,
+      "learning_rate": 2.8591899776900636e-05,
+      "loss": 1.7713,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.885030746459961,
+      "learning_rate": 2.8589325553458042e-05,
+      "loss": 1.7659,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.823013305664062,
+      "learning_rate": 2.8586751330015446e-05,
+      "loss": 1.6639,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.12160873413086,
+      "learning_rate": 2.8584177106572852e-05,
+      "loss": 1.5859,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.695314407348633,
+      "learning_rate": 2.8581602883130256e-05,
+      "loss": 1.666,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.679184913635254,
+      "learning_rate": 2.857902865968766e-05,
+      "loss": 1.3786,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.873235702514648,
+      "learning_rate": 2.8576454436245065e-05,
+      "loss": 1.3557,
+      "step": 553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.05200481414795,
+      "learning_rate": 2.8573880212802472e-05,
+      "loss": 1.8113,
+      "step": 554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.330602645874023,
+      "learning_rate": 2.857130598935988e-05,
+      "loss": 1.405,
+      "step": 555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.589398384094238,
+      "learning_rate": 2.8568731765917282e-05,
+      "loss": 1.4438,
+      "step": 556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.285298347473145,
+      "learning_rate": 2.856615754247469e-05,
+      "loss": 1.4108,
+      "step": 557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.805575847625732,
+      "learning_rate": 2.8563583319032092e-05,
+      "loss": 1.339,
+      "step": 558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.822475433349609,
+      "learning_rate": 2.85610090955895e-05,
+      "loss": 1.2177,
+      "step": 559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.784849166870117,
+      "learning_rate": 2.8558434872146902e-05,
+      "loss": 1.4748,
+      "step": 560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.376548767089844,
+      "learning_rate": 2.855586064870431e-05,
+      "loss": 1.126,
+      "step": 561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.582365989685059,
+      "learning_rate": 2.8553286425261712e-05,
+      "loss": 1.3181,
+      "step": 562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.676347732543945,
+      "learning_rate": 2.855071220181912e-05,
+      "loss": 1.5478,
+      "step": 563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.534523010253906,
+      "learning_rate": 2.8548137978376525e-05,
+      "loss": 1.3806,
+      "step": 564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.204734802246094,
+      "learning_rate": 2.854556375493393e-05,
+      "loss": 1.6416,
+      "step": 565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.165093421936035,
+      "learning_rate": 2.8542989531491335e-05,
+      "loss": 1.8188,
+      "step": 566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.71523380279541,
+      "learning_rate": 2.854041530804874e-05,
+      "loss": 1.5907,
+      "step": 567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.430357933044434,
+      "learning_rate": 2.8537841084606145e-05,
+      "loss": 1.1473,
+      "step": 568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.172737121582031,
+      "learning_rate": 2.853526686116355e-05,
+      "loss": 1.3425,
+      "step": 569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.882026672363281,
+      "learning_rate": 2.8532692637720955e-05,
+      "loss": 1.394,
+      "step": 570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.96976375579834,
+      "learning_rate": 2.853011841427836e-05,
+      "loss": 1.7068,
+      "step": 571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.174710273742676,
+      "learning_rate": 2.8527544190835765e-05,
+      "loss": 1.5241,
+      "step": 572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.107884407043457,
+      "learning_rate": 2.8524969967393172e-05,
+      "loss": 1.2171,
+      "step": 573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.180644035339355,
+      "learning_rate": 2.8522395743950575e-05,
+      "loss": 1.522,
+      "step": 574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.132983207702637,
+      "learning_rate": 2.8519821520507982e-05,
+      "loss": 1.273,
+      "step": 575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.64364767074585,
+      "learning_rate": 2.8517247297065385e-05,
+      "loss": 1.2571,
+      "step": 576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.831807136535645,
+      "learning_rate": 2.8514673073622792e-05,
+      "loss": 1.7455,
+      "step": 577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.58155345916748,
+      "learning_rate": 2.8512098850180195e-05,
+      "loss": 1.4143,
+      "step": 578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.167181015014648,
+      "learning_rate": 2.8509524626737602e-05,
+      "loss": 1.4688,
+      "step": 579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.433991432189941,
+      "learning_rate": 2.8506950403295005e-05,
+      "loss": 1.4494,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.626629829406738,
+      "learning_rate": 2.8504376179852412e-05,
+      "loss": 1.4249,
+      "step": 581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.549328804016113,
+      "learning_rate": 2.850180195640982e-05,
+      "loss": 1.7137,
+      "step": 582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.940505981445312,
+      "learning_rate": 2.8499227732967222e-05,
+      "loss": 1.3205,
+      "step": 583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.76657485961914,
+      "learning_rate": 2.849665350952463e-05,
+      "loss": 1.4745,
+      "step": 584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.353049278259277,
+      "learning_rate": 2.8494079286082032e-05,
+      "loss": 1.6701,
+      "step": 585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.746253967285156,
+      "learning_rate": 2.849150506263944e-05,
+      "loss": 1.3818,
+      "step": 586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.59052562713623,
+      "learning_rate": 2.8488930839196842e-05,
+      "loss": 1.7664,
+      "step": 587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.615872383117676,
+      "learning_rate": 2.848635661575425e-05,
+      "loss": 1.5378,
+      "step": 588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.799748420715332,
+      "learning_rate": 2.8483782392311652e-05,
+      "loss": 1.5329,
+      "step": 589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.83292293548584,
+      "learning_rate": 2.848120816886906e-05,
+      "loss": 1.3462,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.456145286560059,
+      "learning_rate": 2.8478633945426462e-05,
+      "loss": 1.4992,
+      "step": 591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.450884819030762,
+      "learning_rate": 2.8476059721983872e-05,
+      "loss": 1.2677,
+      "step": 592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.532412528991699,
+      "learning_rate": 2.8473485498541275e-05,
+      "loss": 1.5072,
+      "step": 593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.4922709465026855,
+      "learning_rate": 2.847091127509868e-05,
+      "loss": 1.2154,
+      "step": 594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.205540657043457,
+      "learning_rate": 2.8468337051656085e-05,
+      "loss": 1.3805,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.42839241027832,
+      "learning_rate": 2.846576282821349e-05,
+      "loss": 1.3325,
+      "step": 596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.487515449523926,
+      "learning_rate": 2.8463188604770895e-05,
+      "loss": 1.2221,
+      "step": 597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.484041213989258,
+      "learning_rate": 2.84606143813283e-05,
+      "loss": 1.8827,
+      "step": 598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.353645324707031,
+      "learning_rate": 2.8458040157885705e-05,
+      "loss": 1.607,
+      "step": 599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.193730354309082,
+      "learning_rate": 2.845546593444311e-05,
+      "loss": 1.2073,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.85962963104248,
+      "learning_rate": 2.845289171100052e-05,
+      "loss": 1.4088,
+      "step": 601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.017230987548828,
+      "learning_rate": 2.8450317487557922e-05,
+      "loss": 1.4605,
+      "step": 602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.847349643707275,
+      "learning_rate": 2.844774326411533e-05,
+      "loss": 1.1767,
+      "step": 603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.358426094055176,
+      "learning_rate": 2.8445169040672732e-05,
+      "loss": 1.5359,
+      "step": 604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.367965698242188,
+      "learning_rate": 2.8442594817230135e-05,
+      "loss": 1.5171,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.827372550964355,
+      "learning_rate": 2.844002059378754e-05,
+      "loss": 1.3408,
+      "step": 606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.997843265533447,
+      "learning_rate": 2.8437446370344945e-05,
+      "loss": 1.3004,
+      "step": 607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.360730171203613,
+      "learning_rate": 2.843487214690235e-05,
+      "loss": 1.48,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.979563236236572,
+      "learning_rate": 2.8432297923459755e-05,
+      "loss": 1.3397,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.846831321716309,
+      "learning_rate": 2.842972370001716e-05,
+      "loss": 1.2568,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.71766185760498,
+      "learning_rate": 2.8427149476574568e-05,
+      "loss": 1.5845,
+      "step": 611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.213266372680664,
+      "learning_rate": 2.8424575253131975e-05,
+      "loss": 1.2543,
+      "step": 612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.464890480041504,
+      "learning_rate": 2.8422001029689378e-05,
+      "loss": 1.2864,
+      "step": 613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.840841293334961,
+      "learning_rate": 2.841942680624678e-05,
+      "loss": 1.521,
+      "step": 614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.37797737121582,
+      "learning_rate": 2.8416852582804188e-05,
+      "loss": 1.5883,
+      "step": 615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.5304765701293945,
+      "learning_rate": 2.841427835936159e-05,
+      "loss": 1.089,
+      "step": 616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.18742847442627,
+      "learning_rate": 2.8411704135918998e-05,
+      "loss": 1.2754,
+      "step": 617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.496501922607422,
+      "learning_rate": 2.84091299124764e-05,
+      "loss": 1.7009,
+      "step": 618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.113269805908203,
+      "learning_rate": 2.8406555689033808e-05,
+      "loss": 1.5481,
+      "step": 619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.157917022705078,
+      "learning_rate": 2.8403981465591215e-05,
+      "loss": 1.4719,
+      "step": 620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.711053848266602,
+      "learning_rate": 2.840140724214862e-05,
+      "loss": 1.5226,
+      "step": 621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.997692108154297,
+      "learning_rate": 2.8398833018706025e-05,
+      "loss": 1.3378,
+      "step": 622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.8751749992370605,
+      "learning_rate": 2.839625879526343e-05,
+      "loss": 1.4806,
+      "step": 623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.335694313049316,
+      "learning_rate": 2.8393684571820835e-05,
+      "loss": 1.5422,
+      "step": 624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.491147994995117,
+      "learning_rate": 2.8391110348378238e-05,
+      "loss": 1.6276,
+      "step": 625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.00670337677002,
+      "learning_rate": 2.8388536124935645e-05,
+      "loss": 1.2147,
+      "step": 626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.42430305480957,
+      "learning_rate": 2.8385961901493048e-05,
+      "loss": 1.3524,
+      "step": 627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.35117244720459,
+      "learning_rate": 2.8383387678050455e-05,
+      "loss": 1.2318,
+      "step": 628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.867499351501465,
+      "learning_rate": 2.8380813454607858e-05,
+      "loss": 1.5234,
+      "step": 629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.456783294677734,
+      "learning_rate": 2.8378239231165268e-05,
+      "loss": 1.4583,
+      "step": 630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.594684600830078,
+      "learning_rate": 2.837566500772267e-05,
+      "loss": 1.7668,
+      "step": 631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.295198440551758,
+      "learning_rate": 2.8373090784280078e-05,
+      "loss": 1.3842,
+      "step": 632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.227025985717773,
+      "learning_rate": 2.837051656083748e-05,
+      "loss": 1.3941,
+      "step": 633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.663003921508789,
+      "learning_rate": 2.8367942337394888e-05,
+      "loss": 1.7119,
+      "step": 634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.245357513427734,
+      "learning_rate": 2.836536811395229e-05,
+      "loss": 1.3433,
+      "step": 635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.337903022766113,
+      "learning_rate": 2.8362793890509695e-05,
+      "loss": 1.3921,
+      "step": 636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.937525749206543,
+      "learning_rate": 2.83602196670671e-05,
+      "loss": 1.6586,
+      "step": 637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.402658462524414,
+      "learning_rate": 2.8357645443624505e-05,
+      "loss": 1.2618,
+      "step": 638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.879486083984375,
+      "learning_rate": 2.8355071220181915e-05,
+      "loss": 1.6364,
+      "step": 639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.75072956085205,
+      "learning_rate": 2.8352496996739318e-05,
+      "loss": 1.7883,
+      "step": 640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.510144233703613,
+      "learning_rate": 2.8349922773296725e-05,
+      "loss": 1.5349,
+      "step": 641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.616477966308594,
+      "learning_rate": 2.8347348549854128e-05,
+      "loss": 1.7636,
+      "step": 642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.672359466552734,
+      "learning_rate": 2.8344774326411535e-05,
+      "loss": 1.5705,
+      "step": 643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.690023422241211,
+      "learning_rate": 2.8342200102968938e-05,
+      "loss": 1.615,
+      "step": 644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.147051811218262,
+      "learning_rate": 2.8339625879526345e-05,
+      "loss": 1.3188,
+      "step": 645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.338275909423828,
+      "learning_rate": 2.8337051656083748e-05,
+      "loss": 1.4246,
+      "step": 646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.594513416290283,
+      "learning_rate": 2.833447743264115e-05,
+      "loss": 1.5287,
+      "step": 647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.40947151184082,
+      "learning_rate": 2.833190320919856e-05,
+      "loss": 1.4517,
+      "step": 648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.413890838623047,
+      "learning_rate": 2.8329328985755965e-05,
+      "loss": 1.5831,
+      "step": 649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.007394790649414,
+      "learning_rate": 2.832675476231337e-05,
+      "loss": 1.4636,
+      "step": 650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.903453350067139,
+      "learning_rate": 2.8324180538870775e-05,
+      "loss": 1.1872,
+      "step": 651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.605274200439453,
+      "learning_rate": 2.832160631542818e-05,
+      "loss": 1.1718,
+      "step": 652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.102076053619385,
+      "learning_rate": 2.8319032091985585e-05,
+      "loss": 1.1408,
+      "step": 653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.305570602416992,
+      "learning_rate": 2.831645786854299e-05,
+      "loss": 1.6147,
+      "step": 654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.515800476074219,
+      "learning_rate": 2.8313883645100395e-05,
+      "loss": 1.5238,
+      "step": 655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.747093200683594,
+      "learning_rate": 2.8311309421657798e-05,
+      "loss": 1.2978,
+      "step": 656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.113810539245605,
+      "learning_rate": 2.8308735198215204e-05,
+      "loss": 1.8871,
+      "step": 657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.422607421875,
+      "learning_rate": 2.830616097477261e-05,
+      "loss": 1.7156,
+      "step": 658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.17238998413086,
+      "learning_rate": 2.8303586751330018e-05,
+      "loss": 1.7009,
+      "step": 659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.30120849609375,
+      "learning_rate": 2.830101252788742e-05,
+      "loss": 1.8961,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.465618133544922,
+      "learning_rate": 2.8298438304444828e-05,
+      "loss": 1.2713,
+      "step": 661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.053711891174316,
+      "learning_rate": 2.829586408100223e-05,
+      "loss": 1.3993,
+      "step": 662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.97900104522705,
+      "learning_rate": 2.8293289857559638e-05,
+      "loss": 1.736,
+      "step": 663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.041558265686035,
+      "learning_rate": 2.829071563411704e-05,
+      "loss": 1.4241,
+      "step": 664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.299877166748047,
+      "learning_rate": 2.8288141410674448e-05,
+      "loss": 1.4024,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.503842353820801,
+      "learning_rate": 2.828556718723185e-05,
+      "loss": 1.2853,
+      "step": 666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.113901138305664,
+      "learning_rate": 2.8282992963789258e-05,
+      "loss": 1.6806,
+      "step": 667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.860781669616699,
+      "learning_rate": 2.8280418740346664e-05,
+      "loss": 1.3881,
+      "step": 668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.9158501625061035,
+      "learning_rate": 2.8277844516904068e-05,
+      "loss": 1.2087,
+      "step": 669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.270143508911133,
+      "learning_rate": 2.8275270293461474e-05,
+      "loss": 1.2132,
+      "step": 670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.712434768676758,
+      "learning_rate": 2.8272696070018878e-05,
+      "loss": 1.2317,
+      "step": 671
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.592971801757812,
+      "learning_rate": 2.8270121846576284e-05,
+      "loss": 1.8483,
+      "step": 672
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.637419700622559,
+      "learning_rate": 2.8267547623133688e-05,
+      "loss": 1.2322,
+      "step": 673
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.837871551513672,
+      "learning_rate": 2.8264973399691094e-05,
+      "loss": 1.562,
+      "step": 674
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.74756908416748,
+      "learning_rate": 2.8262399176248498e-05,
+      "loss": 1.522,
+      "step": 675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.316798210144043,
+      "learning_rate": 2.8259824952805904e-05,
+      "loss": 1.3224,
+      "step": 676
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.852266788482666,
+      "learning_rate": 2.825725072936331e-05,
+      "loss": 1.4292,
+      "step": 677
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.70324993133545,
+      "learning_rate": 2.8254676505920714e-05,
+      "loss": 1.3933,
+      "step": 678
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.499561309814453,
+      "learning_rate": 2.825210228247812e-05,
+      "loss": 1.5622,
+      "step": 679
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.421582221984863,
+      "learning_rate": 2.8249528059035524e-05,
+      "loss": 1.3405,
+      "step": 680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.83010196685791,
+      "learning_rate": 2.824695383559293e-05,
+      "loss": 1.3413,
+      "step": 681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.317943572998047,
+      "learning_rate": 2.8244379612150334e-05,
+      "loss": 1.5046,
+      "step": 682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.324114799499512,
+      "learning_rate": 2.824180538870774e-05,
+      "loss": 1.2111,
+      "step": 683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.69671630859375,
+      "learning_rate": 2.8239231165265144e-05,
+      "loss": 1.6075,
+      "step": 684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.922121047973633,
+      "learning_rate": 2.823665694182255e-05,
+      "loss": 1.401,
+      "step": 685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.156432151794434,
+      "learning_rate": 2.8234082718379958e-05,
+      "loss": 1.0052,
+      "step": 686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.44300365447998,
+      "learning_rate": 2.823150849493736e-05,
+      "loss": 1.5392,
+      "step": 687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.195891380310059,
+      "learning_rate": 2.8228934271494768e-05,
+      "loss": 1.3723,
+      "step": 688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.78988265991211,
+      "learning_rate": 2.822636004805217e-05,
+      "loss": 1.5812,
+      "step": 689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.798638343811035,
+      "learning_rate": 2.8223785824609578e-05,
+      "loss": 1.7575,
+      "step": 690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.687625885009766,
+      "learning_rate": 2.822121160116698e-05,
+      "loss": 1.1519,
+      "step": 691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.919822692871094,
+      "learning_rate": 2.8218637377724388e-05,
+      "loss": 1.331,
+      "step": 692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.755447387695312,
+      "learning_rate": 2.821606315428179e-05,
+      "loss": 1.4046,
+      "step": 693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.944352149963379,
+      "learning_rate": 2.8213488930839197e-05,
+      "loss": 1.5211,
+      "step": 694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.388495445251465,
+      "learning_rate": 2.82109147073966e-05,
+      "loss": 1.4173,
+      "step": 695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.420978546142578,
+      "learning_rate": 2.820834048395401e-05,
+      "loss": 1.1827,
+      "step": 696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.561078071594238,
+      "learning_rate": 2.8205766260511414e-05,
+      "loss": 1.5684,
+      "step": 697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.48249340057373,
+      "learning_rate": 2.8203192037068817e-05,
+      "loss": 1.533,
+      "step": 698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.367238998413086,
+      "learning_rate": 2.8200617813626224e-05,
+      "loss": 1.4733,
+      "step": 699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.245125770568848,
+      "learning_rate": 2.8198043590183627e-05,
+      "loss": 1.4795,
+      "step": 700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.23082160949707,
+      "learning_rate": 2.8195469366741034e-05,
+      "loss": 1.5353,
+      "step": 701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.436493396759033,
+      "learning_rate": 2.8192895143298437e-05,
+      "loss": 1.2041,
+      "step": 702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.462575435638428,
+      "learning_rate": 2.8190320919855844e-05,
+      "loss": 1.2031,
+      "step": 703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.718941688537598,
+      "learning_rate": 2.8187746696413247e-05,
+      "loss": 1.2735,
+      "step": 704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.902457237243652,
+      "learning_rate": 2.8185172472970657e-05,
+      "loss": 1.3001,
+      "step": 705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.40603256225586,
+      "learning_rate": 2.818259824952806e-05,
+      "loss": 1.3974,
+      "step": 706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.115034103393555,
+      "learning_rate": 2.8180024026085467e-05,
+      "loss": 1.3445,
+      "step": 707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.21794605255127,
+      "learning_rate": 2.817744980264287e-05,
+      "loss": 1.5452,
+      "step": 708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.31137752532959,
+      "learning_rate": 2.8174875579200274e-05,
+      "loss": 1.4253,
+      "step": 709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.634665489196777,
+      "learning_rate": 2.817230135575768e-05,
+      "loss": 1.789,
+      "step": 710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.329280853271484,
+      "learning_rate": 2.8169727132315084e-05,
+      "loss": 1.4738,
+      "step": 711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.350824356079102,
+      "learning_rate": 2.816715290887249e-05,
+      "loss": 1.4322,
+      "step": 712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.817314624786377,
+      "learning_rate": 2.8164578685429894e-05,
+      "loss": 1.3692,
+      "step": 713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.686982154846191,
+      "learning_rate": 2.81620044619873e-05,
+      "loss": 1.4641,
+      "step": 714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.947505950927734,
+      "learning_rate": 2.8159430238544707e-05,
+      "loss": 1.5755,
+      "step": 715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.84663200378418,
+      "learning_rate": 2.8156856015102114e-05,
+      "loss": 1.5036,
+      "step": 716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.863555908203125,
+      "learning_rate": 2.8154281791659517e-05,
+      "loss": 1.6515,
+      "step": 717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.05031967163086,
+      "learning_rate": 2.815170756821692e-05,
+      "loss": 1.6212,
+      "step": 718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.781850814819336,
+      "learning_rate": 2.8149133344774327e-05,
+      "loss": 1.5527,
+      "step": 719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.855267524719238,
+      "learning_rate": 2.814655912133173e-05,
+      "loss": 1.5061,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.520151138305664,
+      "learning_rate": 2.8143984897889137e-05,
+      "loss": 1.2851,
+      "step": 721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.402628898620605,
+      "learning_rate": 2.814141067444654e-05,
+      "loss": 1.458,
+      "step": 722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.520021438598633,
+      "learning_rate": 2.8138836451003947e-05,
+      "loss": 1.3035,
+      "step": 723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.891894340515137,
+      "learning_rate": 2.8136262227561354e-05,
+      "loss": 1.5014,
+      "step": 724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.058806419372559,
+      "learning_rate": 2.813368800411876e-05,
+      "loss": 1.4258,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.435503959655762,
+      "learning_rate": 2.8131113780676164e-05,
+      "loss": 1.0866,
+      "step": 726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.342641830444336,
+      "learning_rate": 2.812853955723357e-05,
+      "loss": 1.8102,
+      "step": 727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.731269836425781,
+      "learning_rate": 2.8125965333790974e-05,
+      "loss": 1.5781,
+      "step": 728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.534306526184082,
+      "learning_rate": 2.8123391110348377e-05,
+      "loss": 1.5534,
+      "step": 729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.31511402130127,
+      "learning_rate": 2.8120816886905784e-05,
+      "loss": 0.9475,
+      "step": 730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.601847648620605,
+      "learning_rate": 2.8118242663463187e-05,
+      "loss": 1.2074,
+      "step": 731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.2041015625,
+      "learning_rate": 2.8115668440020594e-05,
+      "loss": 1.5293,
+      "step": 732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.243498802185059,
+      "learning_rate": 2.8113094216577997e-05,
+      "loss": 1.8844,
+      "step": 733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.502090454101562,
+      "learning_rate": 2.8110519993135407e-05,
+      "loss": 1.4478,
+      "step": 734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.382709503173828,
+      "learning_rate": 2.810794576969281e-05,
+      "loss": 1.3572,
+      "step": 735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.930559158325195,
+      "learning_rate": 2.8105371546250217e-05,
+      "loss": 1.0489,
+      "step": 736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.150694847106934,
+      "learning_rate": 2.810279732280762e-05,
+      "loss": 1.4539,
+      "step": 737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.979594230651855,
+      "learning_rate": 2.8100223099365027e-05,
+      "loss": 1.6143,
+      "step": 738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.314689636230469,
+      "learning_rate": 2.809764887592243e-05,
+      "loss": 1.5458,
+      "step": 739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.491994857788086,
+      "learning_rate": 2.8095074652479834e-05,
+      "loss": 1.3965,
+      "step": 740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.279489517211914,
+      "learning_rate": 2.809250042903724e-05,
+      "loss": 1.1503,
+      "step": 741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.607050895690918,
+      "learning_rate": 2.8089926205594644e-05,
+      "loss": 1.2468,
+      "step": 742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.813472747802734,
+      "learning_rate": 2.8087351982152054e-05,
+      "loss": 1.2145,
+      "step": 743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.898530960083008,
+      "learning_rate": 2.8084777758709457e-05,
+      "loss": 1.3393,
+      "step": 744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.19594955444336,
+      "learning_rate": 2.8082203535266864e-05,
+      "loss": 1.5255,
+      "step": 745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.30061149597168,
+      "learning_rate": 2.8079629311824267e-05,
+      "loss": 1.2362,
+      "step": 746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.08763313293457,
+      "learning_rate": 2.8077055088381674e-05,
+      "loss": 1.2331,
+      "step": 747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.105565071105957,
+      "learning_rate": 2.8074480864939077e-05,
+      "loss": 1.1464,
+      "step": 748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.09921646118164,
+      "learning_rate": 2.8071906641496484e-05,
+      "loss": 1.5649,
+      "step": 749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.6478853225708,
+      "learning_rate": 2.8069332418053887e-05,
+      "loss": 1.4615,
+      "step": 750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.303277969360352,
+      "learning_rate": 2.806675819461129e-05,
+      "loss": 1.3727,
+      "step": 751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.733452796936035,
+      "learning_rate": 2.8064183971168697e-05,
+      "loss": 1.4761,
+      "step": 752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.374414443969727,
+      "learning_rate": 2.8061609747726104e-05,
+      "loss": 1.5373,
+      "step": 753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.646998405456543,
+      "learning_rate": 2.805903552428351e-05,
+      "loss": 1.4541,
+      "step": 754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.969915390014648,
+      "learning_rate": 2.8056461300840914e-05,
+      "loss": 1.4849,
+      "step": 755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.019808769226074,
+      "learning_rate": 2.805388707739832e-05,
+      "loss": 1.1678,
+      "step": 756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.369609832763672,
+      "learning_rate": 2.8051312853955724e-05,
+      "loss": 1.2665,
+      "step": 757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.37863826751709,
+      "learning_rate": 2.804873863051313e-05,
+      "loss": 1.4877,
+      "step": 758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.627493858337402,
+      "learning_rate": 2.8046164407070534e-05,
+      "loss": 1.5047,
+      "step": 759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.57814884185791,
+      "learning_rate": 2.8043590183627937e-05,
+      "loss": 1.356,
+      "step": 760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.7247514724731445,
+      "learning_rate": 2.8041015960185344e-05,
+      "loss": 1.6163,
+      "step": 761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.711459159851074,
+      "learning_rate": 2.803844173674275e-05,
+      "loss": 1.4718,
+      "step": 762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.438310623168945,
+      "learning_rate": 2.8035867513300157e-05,
+      "loss": 1.4829,
+      "step": 763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.3248929977417,
+      "learning_rate": 2.803329328985756e-05,
+      "loss": 1.5657,
+      "step": 764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.526896476745605,
+      "learning_rate": 2.8030719066414967e-05,
+      "loss": 1.3642,
+      "step": 765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.918802261352539,
+      "learning_rate": 2.802814484297237e-05,
+      "loss": 1.489,
+      "step": 766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.489181518554688,
+      "learning_rate": 2.8025570619529777e-05,
+      "loss": 1.5363,
+      "step": 767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.939236640930176,
+      "learning_rate": 2.802299639608718e-05,
+      "loss": 1.5765,
+      "step": 768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.207650184631348,
+      "learning_rate": 2.8020422172644587e-05,
+      "loss": 1.4685,
+      "step": 769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.779074668884277,
+      "learning_rate": 2.801784794920199e-05,
+      "loss": 1.5115,
+      "step": 770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.131409645080566,
+      "learning_rate": 2.8015273725759397e-05,
+      "loss": 1.466,
+      "step": 771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.781936645507812,
+      "learning_rate": 2.8012699502316803e-05,
+      "loss": 1.561,
+      "step": 772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.795979499816895,
+      "learning_rate": 2.8010125278874207e-05,
+      "loss": 1.5802,
+      "step": 773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.044179916381836,
+      "learning_rate": 2.8007551055431613e-05,
+      "loss": 1.5705,
+      "step": 774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.744950294494629,
+      "learning_rate": 2.8004976831989017e-05,
+      "loss": 1.3674,
+      "step": 775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.9183454513549805,
+      "learning_rate": 2.8002402608546423e-05,
+      "loss": 1.1358,
+      "step": 776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.269308090209961,
+      "learning_rate": 2.7999828385103827e-05,
+      "loss": 1.5533,
+      "step": 777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.909619331359863,
+      "learning_rate": 2.7997254161661233e-05,
+      "loss": 1.4723,
+      "step": 778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.315204620361328,
+      "learning_rate": 2.7994679938218637e-05,
+      "loss": 1.2354,
+      "step": 779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.495931625366211,
+      "learning_rate": 2.7992105714776043e-05,
+      "loss": 1.5534,
+      "step": 780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.21084213256836,
+      "learning_rate": 2.798953149133345e-05,
+      "loss": 1.4437,
+      "step": 781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.249410629272461,
+      "learning_rate": 2.7986957267890853e-05,
+      "loss": 1.6794,
+      "step": 782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.520964622497559,
+      "learning_rate": 2.798438304444826e-05,
+      "loss": 1.5549,
+      "step": 783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.072175979614258,
+      "learning_rate": 2.7981808821005663e-05,
+      "loss": 1.2691,
+      "step": 784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.172987937927246,
+      "learning_rate": 2.797923459756307e-05,
+      "loss": 1.1972,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.787796020507812,
+      "learning_rate": 2.7976660374120473e-05,
+      "loss": 1.3193,
+      "step": 786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.934823036193848,
+      "learning_rate": 2.797408615067788e-05,
+      "loss": 1.0993,
+      "step": 787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.232659339904785,
+      "learning_rate": 2.7971511927235283e-05,
+      "loss": 1.3508,
+      "step": 788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.57980728149414,
+      "learning_rate": 2.796893770379269e-05,
+      "loss": 1.1617,
+      "step": 789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.604145050048828,
+      "learning_rate": 2.7966363480350097e-05,
+      "loss": 1.7157,
+      "step": 790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.965630531311035,
+      "learning_rate": 2.79637892569075e-05,
+      "loss": 1.3628,
+      "step": 791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.321697235107422,
+      "learning_rate": 2.7961215033464907e-05,
+      "loss": 1.4101,
+      "step": 792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.196749687194824,
+      "learning_rate": 2.795864081002231e-05,
+      "loss": 1.0958,
+      "step": 793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.979670524597168,
+      "learning_rate": 2.7956066586579717e-05,
+      "loss": 1.225,
+      "step": 794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.82540225982666,
+      "learning_rate": 2.795349236313712e-05,
+      "loss": 1.5192,
+      "step": 795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.932717323303223,
+      "learning_rate": 2.7950918139694527e-05,
+      "loss": 1.0531,
+      "step": 796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.442691802978516,
+      "learning_rate": 2.794834391625193e-05,
+      "loss": 1.261,
+      "step": 797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.675254821777344,
+      "learning_rate": 2.7945769692809337e-05,
+      "loss": 1.2304,
+      "step": 798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.10240650177002,
+      "learning_rate": 2.794319546936674e-05,
+      "loss": 1.4201,
+      "step": 799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.222650527954102,
+      "learning_rate": 2.794062124592415e-05,
+      "loss": 1.245,
+      "step": 800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.896738052368164,
+      "learning_rate": 2.7938047022481553e-05,
+      "loss": 1.4258,
+      "step": 801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.333304405212402,
+      "learning_rate": 2.7935472799038956e-05,
+      "loss": 1.4705,
+      "step": 802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.350672721862793,
+      "learning_rate": 2.7932898575596363e-05,
+      "loss": 1.659,
+      "step": 803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.64991283416748,
+      "learning_rate": 2.7930324352153766e-05,
+      "loss": 1.1396,
+      "step": 804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.883172988891602,
+      "learning_rate": 2.7927750128711173e-05,
+      "loss": 1.3913,
+      "step": 805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.036734580993652,
+      "learning_rate": 2.7925175905268576e-05,
+      "loss": 1.8256,
+      "step": 806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.359436988830566,
+      "learning_rate": 2.7922601681825983e-05,
+      "loss": 1.5282,
+      "step": 807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.453700065612793,
+      "learning_rate": 2.7920027458383386e-05,
+      "loss": 1.657,
+      "step": 808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.343161582946777,
+      "learning_rate": 2.7917453234940796e-05,
+      "loss": 1.5153,
+      "step": 809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.08825969696045,
+      "learning_rate": 2.79148790114982e-05,
+      "loss": 1.165,
+      "step": 810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.013175010681152,
+      "learning_rate": 2.7912304788055606e-05,
+      "loss": 1.6793,
+      "step": 811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.553650856018066,
+      "learning_rate": 2.790973056461301e-05,
+      "loss": 1.4495,
+      "step": 812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.575993537902832,
+      "learning_rate": 2.7907156341170413e-05,
+      "loss": 1.6111,
+      "step": 813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.336498260498047,
+      "learning_rate": 2.790458211772782e-05,
+      "loss": 1.2413,
+      "step": 814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.19493579864502,
+      "learning_rate": 2.7902007894285223e-05,
+      "loss": 1.5179,
+      "step": 815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.755159378051758,
+      "learning_rate": 2.789943367084263e-05,
+      "loss": 1.2592,
+      "step": 816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.641608238220215,
+      "learning_rate": 2.7896859447400033e-05,
+      "loss": 1.1667,
+      "step": 817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.296409606933594,
+      "learning_rate": 2.789428522395744e-05,
+      "loss": 1.2261,
+      "step": 818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.423913955688477,
+      "learning_rate": 2.7891711000514846e-05,
+      "loss": 2.0305,
+      "step": 819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.940492630004883,
+      "learning_rate": 2.7889136777072253e-05,
+      "loss": 1.6239,
+      "step": 820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.727612495422363,
+      "learning_rate": 2.7886562553629656e-05,
+      "loss": 1.3816,
+      "step": 821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.049920082092285,
+      "learning_rate": 2.788398833018706e-05,
+      "loss": 1.2178,
+      "step": 822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.606457710266113,
+      "learning_rate": 2.7881414106744466e-05,
+      "loss": 1.4069,
+      "step": 823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.72132682800293,
+      "learning_rate": 2.787883988330187e-05,
+      "loss": 1.4682,
+      "step": 824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.023664474487305,
+      "learning_rate": 2.7876265659859276e-05,
+      "loss": 1.4439,
+      "step": 825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.507688522338867,
+      "learning_rate": 2.787369143641668e-05,
+      "loss": 1.184,
+      "step": 826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.646373748779297,
+      "learning_rate": 2.7871117212974086e-05,
+      "loss": 1.5025,
+      "step": 827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.352978706359863,
+      "learning_rate": 2.7868542989531493e-05,
+      "loss": 1.3281,
+      "step": 828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.009771347045898,
+      "learning_rate": 2.78659687660889e-05,
+      "loss": 1.5252,
+      "step": 829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.628283500671387,
+      "learning_rate": 2.7863394542646303e-05,
+      "loss": 1.3438,
+      "step": 830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.474054336547852,
+      "learning_rate": 2.786082031920371e-05,
+      "loss": 1.7123,
+      "step": 831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.158101081848145,
+      "learning_rate": 2.7858246095761113e-05,
+      "loss": 1.6228,
+      "step": 832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.310938835144043,
+      "learning_rate": 2.7855671872318516e-05,
+      "loss": 1.3533,
+      "step": 833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.569671630859375,
+      "learning_rate": 2.7853097648875923e-05,
+      "loss": 1.3526,
+      "step": 834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.98193359375,
+      "learning_rate": 2.7850523425433326e-05,
+      "loss": 1.1168,
+      "step": 835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.391763687133789,
+      "learning_rate": 2.7847949201990733e-05,
+      "loss": 1.3365,
+      "step": 836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.450783729553223,
+      "learning_rate": 2.7845374978548136e-05,
+      "loss": 1.6182,
+      "step": 837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.776142120361328,
+      "learning_rate": 2.7842800755105546e-05,
+      "loss": 1.6554,
+      "step": 838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.17233943939209,
+      "learning_rate": 2.784022653166295e-05,
+      "loss": 1.3684,
+      "step": 839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.352310180664062,
+      "learning_rate": 2.7837652308220356e-05,
+      "loss": 1.2635,
+      "step": 840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.076722145080566,
+      "learning_rate": 2.783507808477776e-05,
+      "loss": 1.344,
+      "step": 841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.454001426696777,
+      "learning_rate": 2.7832503861335166e-05,
+      "loss": 1.5776,
+      "step": 842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.781198501586914,
+      "learning_rate": 2.782992963789257e-05,
+      "loss": 1.4146,
+      "step": 843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.85221004486084,
+      "learning_rate": 2.7827355414449973e-05,
+      "loss": 1.1327,
+      "step": 844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.658119201660156,
+      "learning_rate": 2.782478119100738e-05,
+      "loss": 1.5778,
+      "step": 845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.849745750427246,
+      "learning_rate": 2.7822206967564783e-05,
+      "loss": 1.3338,
+      "step": 846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.33973503112793,
+      "learning_rate": 2.7819632744122193e-05,
+      "loss": 1.4931,
+      "step": 847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.998106002807617,
+      "learning_rate": 2.7817058520679596e-05,
+      "loss": 1.3042,
+      "step": 848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.173836708068848,
+      "learning_rate": 2.7814484297237003e-05,
+      "loss": 1.2464,
+      "step": 849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.28589916229248,
+      "learning_rate": 2.7811910073794406e-05,
+      "loss": 1.884,
+      "step": 850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.372815132141113,
+      "learning_rate": 2.7809335850351813e-05,
+      "loss": 1.2398,
+      "step": 851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.412333488464355,
+      "learning_rate": 2.7806761626909216e-05,
+      "loss": 1.5797,
+      "step": 852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.245325088500977,
+      "learning_rate": 2.7804187403466623e-05,
+      "loss": 1.3753,
+      "step": 853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.252384185791016,
+      "learning_rate": 2.7801613180024026e-05,
+      "loss": 1.2435,
+      "step": 854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.715044975280762,
+      "learning_rate": 2.779903895658143e-05,
+      "loss": 1.9103,
+      "step": 855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.821426868438721,
+      "learning_rate": 2.7796464733138836e-05,
+      "loss": 1.2186,
+      "step": 856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.929902076721191,
+      "learning_rate": 2.7793890509696243e-05,
+      "loss": 1.4858,
+      "step": 857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.203032493591309,
+      "learning_rate": 2.779131628625365e-05,
+      "loss": 1.4556,
+      "step": 858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.393985748291016,
+      "learning_rate": 2.7788742062811053e-05,
+      "loss": 1.5507,
+      "step": 859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.583320617675781,
+      "learning_rate": 2.778616783936846e-05,
+      "loss": 1.2571,
+      "step": 860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.82917594909668,
+      "learning_rate": 2.7783593615925863e-05,
+      "loss": 1.2594,
+      "step": 861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.128918647766113,
+      "learning_rate": 2.778101939248327e-05,
+      "loss": 1.4016,
+      "step": 862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.211956977844238,
+      "learning_rate": 2.7778445169040673e-05,
+      "loss": 1.4521,
+      "step": 863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.162777900695801,
+      "learning_rate": 2.7775870945598076e-05,
+      "loss": 1.1528,
+      "step": 864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.554145812988281,
+      "learning_rate": 2.7773296722155483e-05,
+      "loss": 1.02,
+      "step": 865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.833945274353027,
+      "learning_rate": 2.777072249871289e-05,
+      "loss": 1.167,
+      "step": 866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.881366729736328,
+      "learning_rate": 2.7768148275270296e-05,
+      "loss": 1.4133,
+      "step": 867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.62943172454834,
+      "learning_rate": 2.77655740518277e-05,
+      "loss": 1.2017,
+      "step": 868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.090350151062012,
+      "learning_rate": 2.7762999828385106e-05,
+      "loss": 1.022,
+      "step": 869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.169464111328125,
+      "learning_rate": 2.776042560494251e-05,
+      "loss": 1.5532,
+      "step": 870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.9035005569458,
+      "learning_rate": 2.7757851381499916e-05,
+      "loss": 1.5007,
+      "step": 871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.399003028869629,
+      "learning_rate": 2.775527715805732e-05,
+      "loss": 1.3543,
+      "step": 872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.287327766418457,
+      "learning_rate": 2.7752702934614726e-05,
+      "loss": 1.3413,
+      "step": 873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.973354339599609,
+      "learning_rate": 2.775012871117213e-05,
+      "loss": 1.2202,
+      "step": 874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.43429946899414,
+      "learning_rate": 2.7747554487729536e-05,
+      "loss": 1.2334,
+      "step": 875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.133187294006348,
+      "learning_rate": 2.7744980264286942e-05,
+      "loss": 1.505,
+      "step": 876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.597518920898438,
+      "learning_rate": 2.7742406040844346e-05,
+      "loss": 1.3289,
+      "step": 877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.42212963104248,
+      "learning_rate": 2.7739831817401752e-05,
+      "loss": 1.3873,
+      "step": 878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.565329551696777,
+      "learning_rate": 2.7737257593959156e-05,
+      "loss": 1.2618,
+      "step": 879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.424479484558105,
+      "learning_rate": 2.7734683370516562e-05,
+      "loss": 1.3611,
+      "step": 880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.964366912841797,
+      "learning_rate": 2.7732109147073966e-05,
+      "loss": 1.4917,
+      "step": 881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.830155372619629,
+      "learning_rate": 2.7729534923631372e-05,
+      "loss": 1.6664,
+      "step": 882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.698051929473877,
+      "learning_rate": 2.7726960700188776e-05,
+      "loss": 1.2856,
+      "step": 883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.606985092163086,
+      "learning_rate": 2.7724386476746182e-05,
+      "loss": 1.0351,
+      "step": 884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.696028709411621,
+      "learning_rate": 2.772181225330359e-05,
+      "loss": 1.3133,
+      "step": 885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.831123352050781,
+      "learning_rate": 2.7719238029860992e-05,
+      "loss": 1.4761,
+      "step": 886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.008023262023926,
+      "learning_rate": 2.77166638064184e-05,
+      "loss": 1.4388,
+      "step": 887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.56207275390625,
+      "learning_rate": 2.7714089582975802e-05,
+      "loss": 1.5946,
+      "step": 888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.514052391052246,
+      "learning_rate": 2.771151535953321e-05,
+      "loss": 1.4256,
+      "step": 889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.17920970916748,
+      "learning_rate": 2.7708941136090612e-05,
+      "loss": 1.1741,
+      "step": 890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.985677719116211,
+      "learning_rate": 2.770636691264802e-05,
+      "loss": 1.3246,
+      "step": 891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.475797653198242,
+      "learning_rate": 2.7703792689205422e-05,
+      "loss": 1.4087,
+      "step": 892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.537306785583496,
+      "learning_rate": 2.770121846576283e-05,
+      "loss": 1.2527,
+      "step": 893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.671489715576172,
+      "learning_rate": 2.7698644242320236e-05,
+      "loss": 1.3167,
+      "step": 894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.936307907104492,
+      "learning_rate": 2.769607001887764e-05,
+      "loss": 1.576,
+      "step": 895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.451406478881836,
+      "learning_rate": 2.7693495795435046e-05,
+      "loss": 1.4639,
+      "step": 896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.476738929748535,
+      "learning_rate": 2.769092157199245e-05,
+      "loss": 1.3466,
+      "step": 897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.354731559753418,
+      "learning_rate": 2.7688347348549856e-05,
+      "loss": 1.3035,
+      "step": 898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.596505165100098,
+      "learning_rate": 2.768577312510726e-05,
+      "loss": 1.6087,
+      "step": 899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.208419799804688,
+      "learning_rate": 2.7683198901664666e-05,
+      "loss": 1.2007,
+      "step": 900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.790543556213379,
+      "learning_rate": 2.768062467822207e-05,
+      "loss": 1.3654,
+      "step": 901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.729399681091309,
+      "learning_rate": 2.7678050454779476e-05,
+      "loss": 1.1288,
+      "step": 902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.655479431152344,
+      "learning_rate": 2.767547623133688e-05,
+      "loss": 1.4537,
+      "step": 903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.013383865356445,
+      "learning_rate": 2.767290200789429e-05,
+      "loss": 1.302,
+      "step": 904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.9959077835083,
+      "learning_rate": 2.7670327784451692e-05,
+      "loss": 1.4639,
+      "step": 905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.561423301696777,
+      "learning_rate": 2.7667753561009095e-05,
+      "loss": 1.1788,
+      "step": 906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.05864429473877,
+      "learning_rate": 2.7665179337566502e-05,
+      "loss": 1.217,
+      "step": 907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.962160110473633,
+      "learning_rate": 2.7662605114123905e-05,
+      "loss": 1.3522,
+      "step": 908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.301092147827148,
+      "learning_rate": 2.7660030890681312e-05,
+      "loss": 1.3572,
+      "step": 909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.414505004882812,
+      "learning_rate": 2.7657456667238715e-05,
+      "loss": 1.8806,
+      "step": 910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.82504653930664,
+      "learning_rate": 2.7654882443796122e-05,
+      "loss": 1.4387,
+      "step": 911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.63746452331543,
+      "learning_rate": 2.7652308220353525e-05,
+      "loss": 1.1837,
+      "step": 912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.7293620109558105,
+      "learning_rate": 2.7649733996910935e-05,
+      "loss": 1.3675,
+      "step": 913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.78594970703125,
+      "learning_rate": 2.764715977346834e-05,
+      "loss": 1.345,
+      "step": 914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.400517463684082,
+      "learning_rate": 2.7644585550025745e-05,
+      "loss": 1.3403,
+      "step": 915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.140812873840332,
+      "learning_rate": 2.764201132658315e-05,
+      "loss": 1.3672,
+      "step": 916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.761661052703857,
+      "learning_rate": 2.7639437103140552e-05,
+      "loss": 1.2969,
+      "step": 917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.45772933959961,
+      "learning_rate": 2.763686287969796e-05,
+      "loss": 1.5179,
+      "step": 918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.847336769104004,
+      "learning_rate": 2.7634288656255362e-05,
+      "loss": 1.2688,
+      "step": 919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.579736709594727,
+      "learning_rate": 2.763171443281277e-05,
+      "loss": 1.5179,
+      "step": 920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.5482964515686035,
+      "learning_rate": 2.7629140209370172e-05,
+      "loss": 0.9854,
+      "step": 921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.172941207885742,
+      "learning_rate": 2.762656598592758e-05,
+      "loss": 1.5952,
+      "step": 922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.809013366699219,
+      "learning_rate": 2.7623991762484985e-05,
+      "loss": 1.326,
+      "step": 923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.597607612609863,
+      "learning_rate": 2.7621417539042392e-05,
+      "loss": 1.5772,
+      "step": 924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.929927825927734,
+      "learning_rate": 2.7618843315599795e-05,
+      "loss": 1.5638,
+      "step": 925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.811972618103027,
+      "learning_rate": 2.7616269092157202e-05,
+      "loss": 1.2806,
+      "step": 926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.242873191833496,
+      "learning_rate": 2.7613694868714605e-05,
+      "loss": 1.7581,
+      "step": 927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.977294921875,
+      "learning_rate": 2.761112064527201e-05,
+      "loss": 1.4192,
+      "step": 928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.7886962890625,
+      "learning_rate": 2.7608546421829415e-05,
+      "loss": 1.6531,
+      "step": 929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.794466972351074,
+      "learning_rate": 2.760597219838682e-05,
+      "loss": 1.662,
+      "step": 930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.725576400756836,
+      "learning_rate": 2.7603397974944225e-05,
+      "loss": 1.4014,
+      "step": 931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.456843376159668,
+      "learning_rate": 2.7600823751501632e-05,
+      "loss": 1.5779,
+      "step": 932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.219736099243164,
+      "learning_rate": 2.759824952805904e-05,
+      "loss": 1.3011,
+      "step": 933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.357527732849121,
+      "learning_rate": 2.7595675304616442e-05,
+      "loss": 1.3028,
+      "step": 934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.564726829528809,
+      "learning_rate": 2.759310108117385e-05,
+      "loss": 1.6339,
+      "step": 935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.923804759979248,
+      "learning_rate": 2.7590526857731252e-05,
+      "loss": 1.463,
+      "step": 936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.143631935119629,
+      "learning_rate": 2.7587952634288655e-05,
+      "loss": 1.2058,
+      "step": 937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.371817588806152,
+      "learning_rate": 2.7585378410846062e-05,
+      "loss": 1.4673,
+      "step": 938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.691073417663574,
+      "learning_rate": 2.7582804187403465e-05,
+      "loss": 1.1621,
+      "step": 939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.7957763671875,
+      "learning_rate": 2.7580229963960872e-05,
+      "loss": 1.3863,
+      "step": 940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.425039291381836,
+      "learning_rate": 2.7577655740518275e-05,
+      "loss": 1.4881,
+      "step": 941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.161550521850586,
+      "learning_rate": 2.7575081517075685e-05,
+      "loss": 1.3773,
+      "step": 942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.596711158752441,
+      "learning_rate": 2.757250729363309e-05,
+      "loss": 1.1722,
+      "step": 943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.607233047485352,
+      "learning_rate": 2.7569933070190495e-05,
+      "loss": 1.3246,
+      "step": 944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.256855010986328,
+      "learning_rate": 2.75673588467479e-05,
+      "loss": 1.4224,
+      "step": 945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.440996170043945,
+      "learning_rate": 2.7564784623305305e-05,
+      "loss": 1.0807,
+      "step": 946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.548016548156738,
+      "learning_rate": 2.756221039986271e-05,
+      "loss": 1.3143,
+      "step": 947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.569645881652832,
+      "learning_rate": 2.7559636176420112e-05,
+      "loss": 1.1236,
+      "step": 948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.226103782653809,
+      "learning_rate": 2.755706195297752e-05,
+      "loss": 1.0959,
+      "step": 949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.283565521240234,
+      "learning_rate": 2.755448772953492e-05,
+      "loss": 1.0641,
+      "step": 950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.367616653442383,
+      "learning_rate": 2.7551913506092332e-05,
+      "loss": 1.6593,
+      "step": 951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.412775039672852,
+      "learning_rate": 2.7549339282649735e-05,
+      "loss": 1.2759,
+      "step": 952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.963075160980225,
+      "learning_rate": 2.7546765059207142e-05,
+      "loss": 1.4865,
+      "step": 953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.61840534210205,
+      "learning_rate": 2.7544190835764545e-05,
+      "loss": 1.5278,
+      "step": 954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.823490142822266,
+      "learning_rate": 2.7541616612321952e-05,
+      "loss": 1.5876,
+      "step": 955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.646615028381348,
+      "learning_rate": 2.7539042388879355e-05,
+      "loss": 1.6518,
+      "step": 956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.296534538269043,
+      "learning_rate": 2.753646816543676e-05,
+      "loss": 1.5283,
+      "step": 957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.436152458190918,
+      "learning_rate": 2.7533893941994165e-05,
+      "loss": 1.2958,
+      "step": 958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.631742000579834,
+      "learning_rate": 2.7531319718551568e-05,
+      "loss": 1.1314,
+      "step": 959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.731194496154785,
+      "learning_rate": 2.7528745495108975e-05,
+      "loss": 1.7553,
+      "step": 960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.916111946105957,
+      "learning_rate": 2.752617127166638e-05,
+      "loss": 1.3605,
+      "step": 961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.284040451049805,
+      "learning_rate": 2.752359704822379e-05,
+      "loss": 1.361,
+      "step": 962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.29102611541748,
+      "learning_rate": 2.752102282478119e-05,
+      "loss": 1.5038,
+      "step": 963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.322530746459961,
+      "learning_rate": 2.7518448601338598e-05,
+      "loss": 1.3679,
+      "step": 964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.11369800567627,
+      "learning_rate": 2.7515874377896e-05,
+      "loss": 1.5415,
+      "step": 965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.6457438468933105,
+      "learning_rate": 2.7513300154453408e-05,
+      "loss": 1.2653,
+      "step": 966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.037080764770508,
+      "learning_rate": 2.751072593101081e-05,
+      "loss": 0.9422,
+      "step": 967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.72690486907959,
+      "learning_rate": 2.7508151707568215e-05,
+      "loss": 1.3183,
+      "step": 968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.159808158874512,
+      "learning_rate": 2.750557748412562e-05,
+      "loss": 1.3254,
+      "step": 969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.896674156188965,
+      "learning_rate": 2.7503003260683028e-05,
+      "loss": 1.3447,
+      "step": 970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.429150581359863,
+      "learning_rate": 2.7500429037240435e-05,
+      "loss": 1.3015,
+      "step": 971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.94772720336914,
+      "learning_rate": 2.7497854813797838e-05,
+      "loss": 1.3144,
+      "step": 972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.137228012084961,
+      "learning_rate": 2.7495280590355245e-05,
+      "loss": 1.0446,
+      "step": 973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.031691551208496,
+      "learning_rate": 2.7492706366912648e-05,
+      "loss": 0.9954,
+      "step": 974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.93708324432373,
+      "learning_rate": 2.7490132143470055e-05,
+      "loss": 1.2037,
+      "step": 975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.269675254821777,
+      "learning_rate": 2.7487557920027458e-05,
+      "loss": 1.4372,
+      "step": 976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.201359748840332,
+      "learning_rate": 2.7484983696584865e-05,
+      "loss": 1.1643,
+      "step": 977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.586709976196289,
+      "learning_rate": 2.7482409473142268e-05,
+      "loss": 1.5917,
+      "step": 978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.20718002319336,
+      "learning_rate": 2.747983524969967e-05,
+      "loss": 1.4842,
+      "step": 979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.088601112365723,
+      "learning_rate": 2.747726102625708e-05,
+      "loss": 1.4403,
+      "step": 980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.692402839660645,
+      "learning_rate": 2.7474686802814485e-05,
+      "loss": 1.2648,
+      "step": 981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.76282787322998,
+      "learning_rate": 2.747211257937189e-05,
+      "loss": 1.4759,
+      "step": 982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.340919494628906,
+      "learning_rate": 2.7469538355929295e-05,
+      "loss": 1.4664,
+      "step": 983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.029942512512207,
+      "learning_rate": 2.74669641324867e-05,
+      "loss": 1.4747,
+      "step": 984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.550951957702637,
+      "learning_rate": 2.7464389909044105e-05,
+      "loss": 1.8357,
+      "step": 985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.380844116210938,
+      "learning_rate": 2.746181568560151e-05,
+      "loss": 1.2763,
+      "step": 986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.784636497497559,
+      "learning_rate": 2.7459241462158915e-05,
+      "loss": 1.2095,
+      "step": 987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.628052711486816,
+      "learning_rate": 2.745666723871632e-05,
+      "loss": 1.3114,
+      "step": 988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.085103988647461,
+      "learning_rate": 2.7454093015273728e-05,
+      "loss": 1.2096,
+      "step": 989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.746047973632812,
+      "learning_rate": 2.745151879183113e-05,
+      "loss": 1.7489,
+      "step": 990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.280878067016602,
+      "learning_rate": 2.7448944568388538e-05,
+      "loss": 1.3324,
+      "step": 991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.545253753662109,
+      "learning_rate": 2.744637034494594e-05,
+      "loss": 1.3374,
+      "step": 992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.984289169311523,
+      "learning_rate": 2.7443796121503348e-05,
+      "loss": 1.4667,
+      "step": 993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.125088691711426,
+      "learning_rate": 2.744122189806075e-05,
+      "loss": 1.2929,
+      "step": 994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.825432777404785,
+      "learning_rate": 2.7438647674618158e-05,
+      "loss": 1.2813,
+      "step": 995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.433895111083984,
+      "learning_rate": 2.743607345117556e-05,
+      "loss": 1.3678,
+      "step": 996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.616389274597168,
+      "learning_rate": 2.7433499227732968e-05,
+      "loss": 1.4376,
+      "step": 997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.003437042236328,
+      "learning_rate": 2.7430925004290375e-05,
+      "loss": 1.6973,
+      "step": 998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.557788848876953,
+      "learning_rate": 2.7428350780847778e-05,
+      "loss": 1.4694,
+      "step": 999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.035713195800781,
+      "learning_rate": 2.7425776557405185e-05,
+      "loss": 1.4349,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.870453834533691,
+      "learning_rate": 2.7423202333962588e-05,
+      "loss": 1.3485,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.851469993591309,
+      "learning_rate": 2.7420628110519995e-05,
+      "loss": 1.3463,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.11250114440918,
+      "learning_rate": 2.7418053887077398e-05,
+      "loss": 1.8346,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.808446884155273,
+      "learning_rate": 2.7415479663634805e-05,
+      "loss": 1.4318,
+      "step": 1004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.631911277770996,
+      "learning_rate": 2.7412905440192208e-05,
+      "loss": 1.1731,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.999469757080078,
+      "learning_rate": 2.7410331216749615e-05,
+      "loss": 1.2488,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.889474868774414,
+      "learning_rate": 2.7407756993307018e-05,
+      "loss": 1.5142,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.910511016845703,
+      "learning_rate": 2.7405182769864428e-05,
+      "loss": 1.2691,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.317187309265137,
+      "learning_rate": 2.740260854642183e-05,
+      "loss": 1.14,
+      "step": 1009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.15993881225586,
+      "learning_rate": 2.7400034322979234e-05,
+      "loss": 1.1762,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.653911590576172,
+      "learning_rate": 2.739746009953664e-05,
+      "loss": 1.2432,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.123514175415039,
+      "learning_rate": 2.7394885876094044e-05,
+      "loss": 1.0913,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.117807388305664,
+      "learning_rate": 2.739231165265145e-05,
+      "loss": 1.3455,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.247596740722656,
+      "learning_rate": 2.7389737429208854e-05,
+      "loss": 1.2397,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.6008939743042,
+      "learning_rate": 2.738716320576626e-05,
+      "loss": 1.3644,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.20018482208252,
+      "learning_rate": 2.7384588982323664e-05,
+      "loss": 1.4907,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.3839693069458,
+      "learning_rate": 2.7382014758881074e-05,
+      "loss": 1.3098,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.762544631958008,
+      "learning_rate": 2.7379440535438478e-05,
+      "loss": 1.2336,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.597867965698242,
+      "learning_rate": 2.7376866311995884e-05,
+      "loss": 1.428,
+      "step": 1019
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.393948554992676,
+      "learning_rate": 2.7374292088553288e-05,
+      "loss": 1.2987,
+      "step": 1020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.504260063171387,
+      "learning_rate": 2.737171786511069e-05,
+      "loss": 1.1499,
+      "step": 1021
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.080394744873047,
+      "learning_rate": 2.7369143641668098e-05,
+      "loss": 1.3871,
+      "step": 1022
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.832898139953613,
+      "learning_rate": 2.73665694182255e-05,
+      "loss": 1.1188,
+      "step": 1023
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.266690254211426,
+      "learning_rate": 2.7363995194782908e-05,
+      "loss": 1.4037,
+      "step": 1024
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.53109359741211,
+      "learning_rate": 2.736142097134031e-05,
+      "loss": 1.3645,
+      "step": 1025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.360075950622559,
+      "learning_rate": 2.7358846747897718e-05,
+      "loss": 1.4885,
+      "step": 1026
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.781167030334473,
+      "learning_rate": 2.7356272524455124e-05,
+      "loss": 1.673,
+      "step": 1027
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.556243896484375,
+      "learning_rate": 2.735369830101253e-05,
+      "loss": 1.4541,
+      "step": 1028
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.666065216064453,
+      "learning_rate": 2.7351124077569934e-05,
+      "loss": 1.5686,
+      "step": 1029
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.713845252990723,
+      "learning_rate": 2.734854985412734e-05,
+      "loss": 1.2677,
+      "step": 1030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.593574523925781,
+      "learning_rate": 2.7345975630684744e-05,
+      "loss": 1.4152,
+      "step": 1031
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.018454551696777,
+      "learning_rate": 2.7343401407242148e-05,
+      "loss": 1.4421,
+      "step": 1032
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.887699127197266,
+      "learning_rate": 2.7340827183799554e-05,
+      "loss": 1.085,
+      "step": 1033
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.83116626739502,
+      "learning_rate": 2.7338252960356958e-05,
+      "loss": 1.101,
+      "step": 1034
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.687261581420898,
+      "learning_rate": 2.7335678736914364e-05,
+      "loss": 1.158,
+      "step": 1035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.233701705932617,
+      "learning_rate": 2.733310451347177e-05,
+      "loss": 1.1735,
+      "step": 1036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.780478477478027,
+      "learning_rate": 2.7330530290029178e-05,
+      "loss": 1.4419,
+      "step": 1037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.909387588500977,
+      "learning_rate": 2.732795606658658e-05,
+      "loss": 1.2245,
+      "step": 1038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.03641128540039,
+      "learning_rate": 2.7325381843143988e-05,
+      "loss": 1.5364,
+      "step": 1039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.964493274688721,
+      "learning_rate": 2.732280761970139e-05,
+      "loss": 1.0429,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.043766021728516,
+      "learning_rate": 2.7320233396258794e-05,
+      "loss": 1.6543,
+      "step": 1041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.70139217376709,
+      "learning_rate": 2.73176591728162e-05,
+      "loss": 0.9705,
+      "step": 1042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.825154304504395,
+      "learning_rate": 2.7315084949373604e-05,
+      "loss": 1.4729,
+      "step": 1043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.802508354187012,
+      "learning_rate": 2.731251072593101e-05,
+      "loss": 1.2117,
+      "step": 1044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.77186393737793,
+      "learning_rate": 2.7309936502488414e-05,
+      "loss": 1.604,
+      "step": 1045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.28923225402832,
+      "learning_rate": 2.7307362279045824e-05,
+      "loss": 1.3641,
+      "step": 1046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.963648796081543,
+      "learning_rate": 2.7304788055603227e-05,
+      "loss": 1.1925,
+      "step": 1047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.663580894470215,
+      "learning_rate": 2.7302213832160634e-05,
+      "loss": 1.2647,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.14822006225586,
+      "learning_rate": 2.7299639608718037e-05,
+      "loss": 1.2401,
+      "step": 1049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.369800567626953,
+      "learning_rate": 2.7297065385275444e-05,
+      "loss": 1.3942,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.311555862426758,
+      "learning_rate": 2.7294491161832847e-05,
+      "loss": 1.2202,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.417490005493164,
+      "learning_rate": 2.729191693839025e-05,
+      "loss": 1.0955,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.044454574584961,
+      "learning_rate": 2.7289342714947657e-05,
+      "loss": 1.2672,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.748496055603027,
+      "learning_rate": 2.728676849150506e-05,
+      "loss": 1.1653,
+      "step": 1054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.002110481262207,
+      "learning_rate": 2.728419426806247e-05,
+      "loss": 1.1476,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.822497367858887,
+      "learning_rate": 2.7281620044619874e-05,
+      "loss": 1.367,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.774259567260742,
+      "learning_rate": 2.727904582117728e-05,
+      "loss": 1.242,
+      "step": 1057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.742218017578125,
+      "learning_rate": 2.7276471597734684e-05,
+      "loss": 1.5533,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.503838539123535,
+      "learning_rate": 2.727389737429209e-05,
+      "loss": 1.2968,
+      "step": 1059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.95584774017334,
+      "learning_rate": 2.7271323150849494e-05,
+      "loss": 1.3682,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.960376739501953,
+      "learning_rate": 2.72687489274069e-05,
+      "loss": 1.4881,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.651063919067383,
+      "learning_rate": 2.7266174703964304e-05,
+      "loss": 1.1785,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.194541931152344,
+      "learning_rate": 2.7263600480521707e-05,
+      "loss": 1.172,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.579036712646484,
+      "learning_rate": 2.7261026257079114e-05,
+      "loss": 1.563,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.894886493682861,
+      "learning_rate": 2.725845203363652e-05,
+      "loss": 1.2226,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.038219451904297,
+      "learning_rate": 2.7255877810193927e-05,
+      "loss": 1.1319,
+      "step": 1066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.268056869506836,
+      "learning_rate": 2.725330358675133e-05,
+      "loss": 1.0023,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.772714614868164,
+      "learning_rate": 2.7250729363308737e-05,
+      "loss": 1.3525,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.555374145507812,
+      "learning_rate": 2.724815513986614e-05,
+      "loss": 1.1983,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.176958084106445,
+      "learning_rate": 2.7245580916423547e-05,
+      "loss": 1.3963,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.416862487792969,
+      "learning_rate": 2.724300669298095e-05,
+      "loss": 1.2034,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.942066192626953,
+      "learning_rate": 2.7240432469538354e-05,
+      "loss": 1.3266,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.036736488342285,
+      "learning_rate": 2.723785824609576e-05,
+      "loss": 1.4365,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.972512245178223,
+      "learning_rate": 2.7235284022653167e-05,
+      "loss": 1.4161,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.268491744995117,
+      "learning_rate": 2.7232709799210574e-05,
+      "loss": 1.2924,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.488449096679688,
+      "learning_rate": 2.7230135575767977e-05,
+      "loss": 1.0855,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.027210235595703,
+      "learning_rate": 2.7227561352325384e-05,
+      "loss": 1.3078,
+      "step": 1077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.640469551086426,
+      "learning_rate": 2.7224987128882787e-05,
+      "loss": 1.3335,
+      "step": 1078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.25822114944458,
+      "learning_rate": 2.7222412905440194e-05,
+      "loss": 0.8048,
+      "step": 1079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.746109962463379,
+      "learning_rate": 2.7219838681997597e-05,
+      "loss": 1.454,
+      "step": 1080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.298506736755371,
+      "learning_rate": 2.7217264458555004e-05,
+      "loss": 1.1325,
+      "step": 1081
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.089698791503906,
+      "learning_rate": 2.7214690235112407e-05,
+      "loss": 1.4946,
+      "step": 1082
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.399145126342773,
+      "learning_rate": 2.721211601166981e-05,
+      "loss": 1.3796,
+      "step": 1083
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.360230445861816,
+      "learning_rate": 2.720954178822722e-05,
+      "loss": 1.7057,
+      "step": 1084
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.048979759216309,
+      "learning_rate": 2.7206967564784624e-05,
+      "loss": 1.2736,
+      "step": 1085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.581949234008789,
+      "learning_rate": 2.720439334134203e-05,
+      "loss": 1.2437,
+      "step": 1086
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.216153144836426,
+      "learning_rate": 2.7201819117899434e-05,
+      "loss": 1.1414,
+      "step": 1087
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.193462371826172,
+      "learning_rate": 2.719924489445684e-05,
+      "loss": 1.5273,
+      "step": 1088
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.907831192016602,
+      "learning_rate": 2.7196670671014244e-05,
+      "loss": 1.4671,
+      "step": 1089
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.600893020629883,
+      "learning_rate": 2.719409644757165e-05,
+      "loss": 1.3231,
+      "step": 1090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.562541961669922,
+      "learning_rate": 2.7191522224129054e-05,
+      "loss": 1.4549,
+      "step": 1091
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.27472972869873,
+      "learning_rate": 2.718894800068646e-05,
+      "loss": 1.2719,
+      "step": 1092
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.14704418182373,
+      "learning_rate": 2.7186373777243867e-05,
+      "loss": 1.0528,
+      "step": 1093
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.050322532653809,
+      "learning_rate": 2.718379955380127e-05,
+      "loss": 1.5463,
+      "step": 1094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.295217514038086,
+      "learning_rate": 2.7181225330358677e-05,
+      "loss": 1.1819,
+      "step": 1095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.834641456604004,
+      "learning_rate": 2.717865110691608e-05,
+      "loss": 1.4474,
+      "step": 1096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.026323318481445,
+      "learning_rate": 2.7176076883473487e-05,
+      "loss": 1.1722,
+      "step": 1097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.28368854522705,
+      "learning_rate": 2.717350266003089e-05,
+      "loss": 1.2893,
+      "step": 1098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.067283630371094,
+      "learning_rate": 2.7170928436588297e-05,
+      "loss": 1.0889,
+      "step": 1099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.283278465270996,
+      "learning_rate": 2.71683542131457e-05,
+      "loss": 1.1066,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.266766548156738,
+      "learning_rate": 2.7165779989703107e-05,
+      "loss": 1.5371,
+      "step": 1101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.837309837341309,
+      "learning_rate": 2.716320576626051e-05,
+      "loss": 1.2524,
+      "step": 1102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.806564331054688,
+      "learning_rate": 2.7160631542817917e-05,
+      "loss": 1.7349,
+      "step": 1103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.308426856994629,
+      "learning_rate": 2.7158057319375324e-05,
+      "loss": 1.8968,
+      "step": 1104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.350577354431152,
+      "learning_rate": 2.7155483095932727e-05,
+      "loss": 1.2429,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.256535530090332,
+      "learning_rate": 2.7152908872490134e-05,
+      "loss": 1.2191,
+      "step": 1106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.345229148864746,
+      "learning_rate": 2.7150334649047537e-05,
+      "loss": 1.1011,
+      "step": 1107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.522645950317383,
+      "learning_rate": 2.7147760425604944e-05,
+      "loss": 1.2915,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.150025367736816,
+      "learning_rate": 2.7145186202162347e-05,
+      "loss": 1.2009,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.376709938049316,
+      "learning_rate": 2.7142611978719754e-05,
+      "loss": 1.5594,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.121232032775879,
+      "learning_rate": 2.7140037755277157e-05,
+      "loss": 1.3129,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.032523155212402,
+      "learning_rate": 2.7137463531834567e-05,
+      "loss": 1.1689,
+      "step": 1112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.9547905921936035,
+      "learning_rate": 2.713488930839197e-05,
+      "loss": 1.2583,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.238072395324707,
+      "learning_rate": 2.7132315084949373e-05,
+      "loss": 1.1958,
+      "step": 1114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.16339111328125,
+      "learning_rate": 2.712974086150678e-05,
+      "loss": 1.2236,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.557994842529297,
+      "learning_rate": 2.7127166638064183e-05,
+      "loss": 1.1426,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.909612655639648,
+      "learning_rate": 2.712459241462159e-05,
+      "loss": 1.2599,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.249550819396973,
+      "learning_rate": 2.7122018191178993e-05,
+      "loss": 1.4662,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.27196979522705,
+      "learning_rate": 2.71194439677364e-05,
+      "loss": 1.472,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.640380859375,
+      "learning_rate": 2.7116869744293803e-05,
+      "loss": 1.4584,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.803701400756836,
+      "learning_rate": 2.7114295520851213e-05,
+      "loss": 1.0802,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.972593307495117,
+      "learning_rate": 2.7111721297408617e-05,
+      "loss": 1.3308,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.79111385345459,
+      "learning_rate": 2.7109147073966023e-05,
+      "loss": 1.2226,
+      "step": 1123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.142602920532227,
+      "learning_rate": 2.7106572850523427e-05,
+      "loss": 1.5817,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.296048641204834,
+      "learning_rate": 2.710399862708083e-05,
+      "loss": 1.0926,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.158304214477539,
+      "learning_rate": 2.7101424403638237e-05,
+      "loss": 1.4636,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.532355308532715,
+      "learning_rate": 2.709885018019564e-05,
+      "loss": 1.4941,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.275026798248291,
+      "learning_rate": 2.7096275956753047e-05,
+      "loss": 0.9197,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.894339561462402,
+      "learning_rate": 2.709370173331045e-05,
+      "loss": 1.5351,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.644191265106201,
+      "learning_rate": 2.7091127509867857e-05,
+      "loss": 1.028,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.101517677307129,
+      "learning_rate": 2.7088553286425263e-05,
+      "loss": 1.4001,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.999380588531494,
+      "learning_rate": 2.708597906298267e-05,
+      "loss": 1.015,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.905345916748047,
+      "learning_rate": 2.7083404839540073e-05,
+      "loss": 1.6007,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.128106117248535,
+      "learning_rate": 2.708083061609748e-05,
+      "loss": 1.4566,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.673796653747559,
+      "learning_rate": 2.7078256392654883e-05,
+      "loss": 1.6084,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.865784168243408,
+      "learning_rate": 2.7075682169212287e-05,
+      "loss": 1.2059,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.854172706604004,
+      "learning_rate": 2.7073107945769693e-05,
+      "loss": 1.2551,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.900066375732422,
+      "learning_rate": 2.7070533722327097e-05,
+      "loss": 1.1971,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.985684394836426,
+      "learning_rate": 2.7067959498884503e-05,
+      "loss": 1.4171,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.223388671875,
+      "learning_rate": 2.706538527544191e-05,
+      "loss": 1.502,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.067261695861816,
+      "learning_rate": 2.7062811051999317e-05,
+      "loss": 1.5211,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.43513298034668,
+      "learning_rate": 2.706023682855672e-05,
+      "loss": 1.1767,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.506795883178711,
+      "learning_rate": 2.7057662605114127e-05,
+      "loss": 1.4128,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.250001430511475,
+      "learning_rate": 2.705508838167153e-05,
+      "loss": 1.2489,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.432572841644287,
+      "learning_rate": 2.7052514158228933e-05,
+      "loss": 1.1841,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.109137535095215,
+      "learning_rate": 2.704993993478634e-05,
+      "loss": 1.2777,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.922881126403809,
+      "learning_rate": 2.7047365711343743e-05,
+      "loss": 1.4648,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.688481330871582,
+      "learning_rate": 2.704479148790115e-05,
+      "loss": 1.358,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.354059219360352,
+      "learning_rate": 2.7042217264458553e-05,
+      "loss": 0.9972,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.341496467590332,
+      "learning_rate": 2.7039643041015963e-05,
+      "loss": 1.2205,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.882163047790527,
+      "learning_rate": 2.7037068817573366e-05,
+      "loss": 1.2021,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.504749298095703,
+      "learning_rate": 2.7034494594130773e-05,
+      "loss": 1.3326,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.217513084411621,
+      "learning_rate": 2.7031920370688176e-05,
+      "loss": 1.3535,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.810968399047852,
+      "learning_rate": 2.7029346147245583e-05,
+      "loss": 1.2546,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.84786319732666,
+      "learning_rate": 2.7026771923802986e-05,
+      "loss": 1.4116,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.833187103271484,
+      "learning_rate": 2.702419770036039e-05,
+      "loss": 1.4681,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.529823303222656,
+      "learning_rate": 2.7021623476917796e-05,
+      "loss": 1.2409,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.504322052001953,
+      "learning_rate": 2.70190492534752e-05,
+      "loss": 1.1238,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.834338188171387,
+      "learning_rate": 2.701647503003261e-05,
+      "loss": 1.3571,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.878161430358887,
+      "learning_rate": 2.7013900806590013e-05,
+      "loss": 1.19,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.034382820129395,
+      "learning_rate": 2.701132658314742e-05,
+      "loss": 1.0617,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.727060317993164,
+      "learning_rate": 2.7008752359704823e-05,
+      "loss": 1.061,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.908223152160645,
+      "learning_rate": 2.700617813626223e-05,
+      "loss": 0.8979,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.940628051757812,
+      "learning_rate": 2.7003603912819633e-05,
+      "loss": 1.3193,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.10629653930664,
+      "learning_rate": 2.700102968937704e-05,
+      "loss": 1.7784,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.900351524353027,
+      "learning_rate": 2.6998455465934443e-05,
+      "loss": 1.2081,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.000555038452148,
+      "learning_rate": 2.6995881242491846e-05,
+      "loss": 1.3022,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.327181816101074,
+      "learning_rate": 2.6993307019049253e-05,
+      "loss": 1.2732,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.021303176879883,
+      "learning_rate": 2.699073279560666e-05,
+      "loss": 1.4867,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.396635055541992,
+      "learning_rate": 2.6988158572164066e-05,
+      "loss": 1.8524,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.448280334472656,
+      "learning_rate": 2.698558434872147e-05,
+      "loss": 1.3967,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.023969650268555,
+      "learning_rate": 2.6983010125278876e-05,
+      "loss": 1.1958,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.381837844848633,
+      "learning_rate": 2.698043590183628e-05,
+      "loss": 1.2659,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.306741714477539,
+      "learning_rate": 2.6977861678393686e-05,
+      "loss": 1.1591,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.948760032653809,
+      "learning_rate": 2.697528745495109e-05,
+      "loss": 1.0862,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.335007667541504,
+      "learning_rate": 2.6972713231508493e-05,
+      "loss": 0.8766,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.226988792419434,
+      "learning_rate": 2.69701390080659e-05,
+      "loss": 1.2587,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.792378902435303,
+      "learning_rate": 2.6967564784623306e-05,
+      "loss": 1.2425,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.548218727111816,
+      "learning_rate": 2.6964990561180713e-05,
+      "loss": 1.2892,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.905377388000488,
+      "learning_rate": 2.6962416337738116e-05,
+      "loss": 1.5361,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.688945770263672,
+      "learning_rate": 2.6959842114295523e-05,
+      "loss": 1.3123,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.875213146209717,
+      "learning_rate": 2.6957267890852926e-05,
+      "loss": 0.9015,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.691976547241211,
+      "learning_rate": 2.6954693667410333e-05,
+      "loss": 1.4515,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.65355396270752,
+      "learning_rate": 2.6952119443967736e-05,
+      "loss": 1.9028,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.477837562561035,
+      "learning_rate": 2.6949545220525143e-05,
+      "loss": 1.2233,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.239665985107422,
+      "learning_rate": 2.6946970997082546e-05,
+      "loss": 1.2302,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.567435264587402,
+      "learning_rate": 2.694439677363995e-05,
+      "loss": 1.456,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.737056732177734,
+      "learning_rate": 2.694182255019736e-05,
+      "loss": 1.0246,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.440693855285645,
+      "learning_rate": 2.6939248326754763e-05,
+      "loss": 1.1833,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.651070594787598,
+      "learning_rate": 2.693667410331217e-05,
+      "loss": 1.2408,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.181726455688477,
+      "learning_rate": 2.6934099879869573e-05,
+      "loss": 1.4955,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.472039222717285,
+      "learning_rate": 2.693152565642698e-05,
+      "loss": 1.4052,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.136848449707031,
+      "learning_rate": 2.6928951432984383e-05,
+      "loss": 1.0829,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.888769626617432,
+      "learning_rate": 2.692637720954179e-05,
+      "loss": 1.1699,
+      "step": 1194
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.18743896484375,
+      "learning_rate": 2.6923802986099193e-05,
+      "loss": 1.0537,
+      "step": 1195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.397427558898926,
+      "learning_rate": 2.69212287626566e-05,
+      "loss": 1.2303,
+      "step": 1196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.45164966583252,
+      "learning_rate": 2.6918654539214006e-05,
+      "loss": 1.5722,
+      "step": 1197
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.499549865722656,
+      "learning_rate": 2.691608031577141e-05,
+      "loss": 1.4703,
+      "step": 1198
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.246573448181152,
+      "learning_rate": 2.6913506092328816e-05,
+      "loss": 1.5224,
+      "step": 1199
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.132889747619629,
+      "learning_rate": 2.691093186888622e-05,
+      "loss": 1.2977,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.937135696411133,
+      "learning_rate": 2.6908357645443626e-05,
+      "loss": 1.3967,
+      "step": 1201
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.2399263381958,
+      "learning_rate": 2.690578342200103e-05,
+      "loss": 1.0836,
+      "step": 1202
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.38232135772705,
+      "learning_rate": 2.6903209198558436e-05,
+      "loss": 1.2621,
+      "step": 1203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.263805389404297,
+      "learning_rate": 2.690063497511584e-05,
+      "loss": 1.382,
+      "step": 1204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.532333374023438,
+      "learning_rate": 2.6898060751673246e-05,
+      "loss": 1.0092,
+      "step": 1205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.846914291381836,
+      "learning_rate": 2.689548652823065e-05,
+      "loss": 1.2845,
+      "step": 1206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.730926990509033,
+      "learning_rate": 2.6892912304788056e-05,
+      "loss": 0.9868,
+      "step": 1207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.525662422180176,
+      "learning_rate": 2.6890338081345463e-05,
+      "loss": 1.4252,
+      "step": 1208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.073384284973145,
+      "learning_rate": 2.6887763857902866e-05,
+      "loss": 1.4963,
+      "step": 1209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.478918075561523,
+      "learning_rate": 2.6885189634460273e-05,
+      "loss": 1.3453,
+      "step": 1210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.732980728149414,
+      "learning_rate": 2.6882615411017676e-05,
+      "loss": 1.4029,
+      "step": 1211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.388533592224121,
+      "learning_rate": 2.6880041187575083e-05,
+      "loss": 1.1371,
+      "step": 1212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.94727611541748,
+      "learning_rate": 2.6877466964132486e-05,
+      "loss": 1.2883,
+      "step": 1213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.437952995300293,
+      "learning_rate": 2.6874892740689893e-05,
+      "loss": 1.311,
+      "step": 1214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.140856742858887,
+      "learning_rate": 2.6872318517247296e-05,
+      "loss": 1.1271,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.91845417022705,
+      "learning_rate": 2.6869744293804706e-05,
+      "loss": 1.1375,
+      "step": 1216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.56423568725586,
+      "learning_rate": 2.686717007036211e-05,
+      "loss": 1.4121,
+      "step": 1217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.286993026733398,
+      "learning_rate": 2.6864595846919513e-05,
+      "loss": 1.304,
+      "step": 1218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.56916332244873,
+      "learning_rate": 2.686202162347692e-05,
+      "loss": 1.4004,
+      "step": 1219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.579012870788574,
+      "learning_rate": 2.6859447400034322e-05,
+      "loss": 1.4633,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.460388660430908,
+      "learning_rate": 2.685687317659173e-05,
+      "loss": 0.9385,
+      "step": 1221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.638516426086426,
+      "learning_rate": 2.6854298953149132e-05,
+      "loss": 1.6112,
+      "step": 1222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.712013244628906,
+      "learning_rate": 2.685172472970654e-05,
+      "loss": 1.4429,
+      "step": 1223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.578434944152832,
+      "learning_rate": 2.6849150506263942e-05,
+      "loss": 1.4887,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.140835762023926,
+      "learning_rate": 2.6846576282821352e-05,
+      "loss": 1.1822,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.051060676574707,
+      "learning_rate": 2.6844002059378756e-05,
+      "loss": 1.2473,
+      "step": 1226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.668966293334961,
+      "learning_rate": 2.6841427835936162e-05,
+      "loss": 1.2553,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.400277137756348,
+      "learning_rate": 2.6838853612493566e-05,
+      "loss": 1.2029,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.655241012573242,
+      "learning_rate": 2.683627938905097e-05,
+      "loss": 1.5658,
+      "step": 1229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.049247741699219,
+      "learning_rate": 2.6833705165608376e-05,
+      "loss": 1.293,
+      "step": 1230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.07247543334961,
+      "learning_rate": 2.683113094216578e-05,
+      "loss": 1.3902,
+      "step": 1231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.127625465393066,
+      "learning_rate": 2.6828556718723186e-05,
+      "loss": 1.2842,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.649299144744873,
+      "learning_rate": 2.682598249528059e-05,
+      "loss": 0.8758,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.730408668518066,
+      "learning_rate": 2.6823408271837996e-05,
+      "loss": 1.4391,
+      "step": 1234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.480378150939941,
+      "learning_rate": 2.6820834048395402e-05,
+      "loss": 1.5429,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.526721954345703,
+      "learning_rate": 2.681825982495281e-05,
+      "loss": 1.2032,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.1389799118042,
+      "learning_rate": 2.6815685601510212e-05,
+      "loss": 1.0992,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.146998405456543,
+      "learning_rate": 2.681311137806762e-05,
+      "loss": 1.4844,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.229116439819336,
+      "learning_rate": 2.6810537154625022e-05,
+      "loss": 1.2092,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.03791332244873,
+      "learning_rate": 2.6807962931182426e-05,
+      "loss": 1.2726,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.248305320739746,
+      "learning_rate": 2.6805388707739832e-05,
+      "loss": 1.4621,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.154080390930176,
+      "learning_rate": 2.6802814484297236e-05,
+      "loss": 1.3475,
+      "step": 1242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.43741512298584,
+      "learning_rate": 2.6800240260854642e-05,
+      "loss": 1.3909,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.604585647583008,
+      "learning_rate": 2.679766603741205e-05,
+      "loss": 1.4298,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.446846961975098,
+      "learning_rate": 2.6795091813969456e-05,
+      "loss": 1.1632,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.354076862335205,
+      "learning_rate": 2.679251759052686e-05,
+      "loss": 1.257,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.221611976623535,
+      "learning_rate": 2.6789943367084266e-05,
+      "loss": 1.4602,
+      "step": 1247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.221695899963379,
+      "learning_rate": 2.678736914364167e-05,
+      "loss": 1.3385,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.068194389343262,
+      "learning_rate": 2.6784794920199072e-05,
+      "loss": 1.226,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.30330753326416,
+      "learning_rate": 2.678222069675648e-05,
+      "loss": 1.3546,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.008672714233398,
+      "learning_rate": 2.6779646473313882e-05,
+      "loss": 1.4961,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.799113273620605,
+      "learning_rate": 2.677707224987129e-05,
+      "loss": 1.419,
+      "step": 1252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.145889282226562,
+      "learning_rate": 2.6774498026428692e-05,
+      "loss": 0.9806,
+      "step": 1253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.050820350646973,
+      "learning_rate": 2.6771923802986102e-05,
+      "loss": 1.3412,
+      "step": 1254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.461099624633789,
+      "learning_rate": 2.6769349579543506e-05,
+      "loss": 1.1918,
+      "step": 1255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.203243255615234,
+      "learning_rate": 2.6766775356100912e-05,
+      "loss": 1.2476,
+      "step": 1256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.571125984191895,
+      "learning_rate": 2.6764201132658315e-05,
+      "loss": 1.1585,
+      "step": 1257
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.301223754882812,
+      "learning_rate": 2.6761626909215722e-05,
+      "loss": 1.5106,
+      "step": 1258
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.19857406616211,
+      "learning_rate": 2.6759052685773125e-05,
+      "loss": 1.2659,
+      "step": 1259
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.036355018615723,
+      "learning_rate": 2.675647846233053e-05,
+      "loss": 1.137,
+      "step": 1260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.456526756286621,
+      "learning_rate": 2.6753904238887935e-05,
+      "loss": 1.4946,
+      "step": 1261
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.323201179504395,
+      "learning_rate": 2.675133001544534e-05,
+      "loss": 1.1494,
+      "step": 1262
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.239011764526367,
+      "learning_rate": 2.674875579200275e-05,
+      "loss": 1.3713,
+      "step": 1263
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.527006149291992,
+      "learning_rate": 2.6746181568560152e-05,
+      "loss": 1.1823,
+      "step": 1264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.94394063949585,
+      "learning_rate": 2.674360734511756e-05,
+      "loss": 1.322,
+      "step": 1265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.312918663024902,
+      "learning_rate": 2.6741033121674962e-05,
+      "loss": 1.1758,
+      "step": 1266
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.257575988769531,
+      "learning_rate": 2.673845889823237e-05,
+      "loss": 1.2889,
+      "step": 1267
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.413891792297363,
+      "learning_rate": 2.6735884674789772e-05,
+      "loss": 1.515,
+      "step": 1268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.223206043243408,
+      "learning_rate": 2.673331045134718e-05,
+      "loss": 1.0707,
+      "step": 1269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.151848793029785,
+      "learning_rate": 2.6730736227904582e-05,
+      "loss": 1.0929,
+      "step": 1270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.26773452758789,
+      "learning_rate": 2.6728162004461985e-05,
+      "loss": 1.1119,
+      "step": 1271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.067112922668457,
+      "learning_rate": 2.6725587781019392e-05,
+      "loss": 1.4869,
+      "step": 1272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.439190864562988,
+      "learning_rate": 2.67230135575768e-05,
+      "loss": 1.4908,
+      "step": 1273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.8590669631958,
+      "learning_rate": 2.6720439334134205e-05,
+      "loss": 1.3083,
+      "step": 1274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.895794868469238,
+      "learning_rate": 2.671786511069161e-05,
+      "loss": 1.2789,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.065464973449707,
+      "learning_rate": 2.6715290887249015e-05,
+      "loss": 1.3152,
+      "step": 1276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.821928977966309,
+      "learning_rate": 2.671271666380642e-05,
+      "loss": 1.3512,
+      "step": 1277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.471129417419434,
+      "learning_rate": 2.6710142440363825e-05,
+      "loss": 1.0972,
+      "step": 1278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.04776668548584,
+      "learning_rate": 2.670756821692123e-05,
+      "loss": 1.248,
+      "step": 1279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.509843826293945,
+      "learning_rate": 2.6704993993478632e-05,
+      "loss": 1.2877,
+      "step": 1280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.483633041381836,
+      "learning_rate": 2.670241977003604e-05,
+      "loss": 1.217,
+      "step": 1281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.813414573669434,
+      "learning_rate": 2.6699845546593445e-05,
+      "loss": 1.2106,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.751136779785156,
+      "learning_rate": 2.6697271323150852e-05,
+      "loss": 1.1097,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.629862785339355,
+      "learning_rate": 2.6694697099708255e-05,
+      "loss": 1.1052,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.738077640533447,
+      "learning_rate": 2.6692122876265662e-05,
+      "loss": 0.9371,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.272834777832031,
+      "learning_rate": 2.6689548652823065e-05,
+      "loss": 1.178,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.922650337219238,
+      "learning_rate": 2.6686974429380472e-05,
+      "loss": 1.1148,
+      "step": 1287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.045205116271973,
+      "learning_rate": 2.6684400205937875e-05,
+      "loss": 1.4301,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.679380416870117,
+      "learning_rate": 2.6681825982495282e-05,
+      "loss": 1.3469,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.931136131286621,
+      "learning_rate": 2.6679251759052685e-05,
+      "loss": 1.4528,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.9316534996032715,
+      "learning_rate": 2.667667753561009e-05,
+      "loss": 0.8932,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.66825008392334,
+      "learning_rate": 2.66741033121675e-05,
+      "loss": 1.2996,
+      "step": 1292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.578680515289307,
+      "learning_rate": 2.6671529088724902e-05,
+      "loss": 0.929,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.219220161437988,
+      "learning_rate": 2.666895486528231e-05,
+      "loss": 1.1454,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.822579860687256,
+      "learning_rate": 2.6666380641839712e-05,
+      "loss": 1.1635,
+      "step": 1295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.75918197631836,
+      "learning_rate": 2.666380641839712e-05,
+      "loss": 1.3699,
+      "step": 1296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.776763916015625,
+      "learning_rate": 2.6661232194954522e-05,
+      "loss": 1.425,
+      "step": 1297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.903858184814453,
+      "learning_rate": 2.665865797151193e-05,
+      "loss": 1.2982,
+      "step": 1298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.8899312019348145,
+      "learning_rate": 2.6656083748069332e-05,
+      "loss": 1.0836,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.65578556060791,
+      "learning_rate": 2.665350952462674e-05,
+      "loss": 1.2516,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.424769401550293,
+      "learning_rate": 2.6650935301184145e-05,
+      "loss": 1.6553,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.16710090637207,
+      "learning_rate": 2.664836107774155e-05,
+      "loss": 1.1495,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.541897773742676,
+      "learning_rate": 2.6645786854298955e-05,
+      "loss": 1.2282,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.607604026794434,
+      "learning_rate": 2.664321263085636e-05,
+      "loss": 1.4688,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.082883834838867,
+      "learning_rate": 2.6640638407413765e-05,
+      "loss": 0.9886,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.696791648864746,
+      "learning_rate": 2.663806418397117e-05,
+      "loss": 1.0391,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.231827735900879,
+      "learning_rate": 2.6635489960528575e-05,
+      "loss": 1.4162,
+      "step": 1307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.852787971496582,
+      "learning_rate": 2.6632915737085978e-05,
+      "loss": 1.0936,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.198777198791504,
+      "learning_rate": 2.6630341513643385e-05,
+      "loss": 1.2517,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.10767650604248,
+      "learning_rate": 2.6627767290200788e-05,
+      "loss": 1.2609,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.866680145263672,
+      "learning_rate": 2.6625193066758195e-05,
+      "loss": 1.2264,
+      "step": 1311
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.617739677429199,
+      "learning_rate": 2.66226188433156e-05,
+      "loss": 1.0467,
+      "step": 1312
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.443436622619629,
+      "learning_rate": 2.6620044619873005e-05,
+      "loss": 0.9794,
+      "step": 1313
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.602325439453125,
+      "learning_rate": 2.661747039643041e-05,
+      "loss": 1.3684,
+      "step": 1314
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.663492202758789,
+      "learning_rate": 2.6614896172987815e-05,
+      "loss": 1.2373,
+      "step": 1315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.530584335327148,
+      "learning_rate": 2.661232194954522e-05,
+      "loss": 1.4406,
+      "step": 1316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.34512996673584,
+      "learning_rate": 2.6609747726102625e-05,
+      "loss": 1.0497,
+      "step": 1317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.65249490737915,
+      "learning_rate": 2.660717350266003e-05,
+      "loss": 1.0519,
+      "step": 1318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.023550987243652,
+      "learning_rate": 2.6604599279217435e-05,
+      "loss": 1.3027,
+      "step": 1319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.111237525939941,
+      "learning_rate": 2.6602025055774845e-05,
+      "loss": 1.4954,
+      "step": 1320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.713436126708984,
+      "learning_rate": 2.6599450832332248e-05,
+      "loss": 1.6837,
+      "step": 1321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.139344215393066,
+      "learning_rate": 2.659687660888965e-05,
+      "loss": 1.4476,
+      "step": 1322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.681056022644043,
+      "learning_rate": 2.6594302385447058e-05,
+      "loss": 1.2565,
+      "step": 1323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.738880157470703,
+      "learning_rate": 2.659172816200446e-05,
+      "loss": 1.2536,
+      "step": 1324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.696887969970703,
+      "learning_rate": 2.6589153938561868e-05,
+      "loss": 1.6081,
+      "step": 1325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.451643943786621,
+      "learning_rate": 2.658657971511927e-05,
+      "loss": 1.1937,
+      "step": 1326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.157116889953613,
+      "learning_rate": 2.6584005491676678e-05,
+      "loss": 1.1614,
+      "step": 1327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.9005765914917,
+      "learning_rate": 2.658143126823408e-05,
+      "loss": 1.3988,
+      "step": 1328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.618021011352539,
+      "learning_rate": 2.6578857044791488e-05,
+      "loss": 1.0125,
+      "step": 1329
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.41938591003418,
+      "learning_rate": 2.6576282821348895e-05,
+      "loss": 1.1074,
+      "step": 1330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.549199104309082,
+      "learning_rate": 2.65737085979063e-05,
+      "loss": 1.3725,
+      "step": 1331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.936274528503418,
+      "learning_rate": 2.6571134374463705e-05,
+      "loss": 1.4184,
+      "step": 1332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.33743667602539,
+      "learning_rate": 2.6568560151021108e-05,
+      "loss": 1.3609,
+      "step": 1333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.436013221740723,
+      "learning_rate": 2.6565985927578515e-05,
+      "loss": 1.3234,
+      "step": 1334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.227049827575684,
+      "learning_rate": 2.6563411704135918e-05,
+      "loss": 1.455,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.859749794006348,
+      "learning_rate": 2.6560837480693325e-05,
+      "loss": 1.381,
+      "step": 1336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.653039932250977,
+      "learning_rate": 2.6558263257250728e-05,
+      "loss": 1.2519,
+      "step": 1337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.58700942993164,
+      "learning_rate": 2.6555689033808135e-05,
+      "loss": 1.5164,
+      "step": 1338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.98537826538086,
+      "learning_rate": 2.655311481036554e-05,
+      "loss": 1.0812,
+      "step": 1339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.802996635437012,
+      "learning_rate": 2.6550540586922948e-05,
+      "loss": 1.4409,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.917903900146484,
+      "learning_rate": 2.654796636348035e-05,
+      "loss": 0.9767,
+      "step": 1341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.221455574035645,
+      "learning_rate": 2.6545392140037758e-05,
+      "loss": 1.2497,
+      "step": 1342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.750977516174316,
+      "learning_rate": 2.654281791659516e-05,
+      "loss": 1.1992,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.34138298034668,
+      "learning_rate": 2.6540243693152565e-05,
+      "loss": 1.3582,
+      "step": 1344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.008773803710938,
+      "learning_rate": 2.653766946970997e-05,
+      "loss": 1.5271,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.252571105957031,
+      "learning_rate": 2.6535095246267375e-05,
+      "loss": 1.1353,
+      "step": 1346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.196686744689941,
+      "learning_rate": 2.653252102282478e-05,
+      "loss": 1.2903,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.113208770751953,
+      "learning_rate": 2.6529946799382188e-05,
+      "loss": 1.1973,
+      "step": 1348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.13182258605957,
+      "learning_rate": 2.6527372575939595e-05,
+      "loss": 1.1939,
+      "step": 1349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.174057960510254,
+      "learning_rate": 2.6524798352496998e-05,
+      "loss": 1.3718,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.287139892578125,
+      "learning_rate": 2.6522224129054405e-05,
+      "loss": 1.4532,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.905390739440918,
+      "learning_rate": 2.6519649905611808e-05,
+      "loss": 1.2902,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.484794616699219,
+      "learning_rate": 2.651707568216921e-05,
+      "loss": 1.2279,
+      "step": 1353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.344484329223633,
+      "learning_rate": 2.6514501458726618e-05,
+      "loss": 1.0599,
+      "step": 1354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.926910400390625,
+      "learning_rate": 2.651192723528402e-05,
+      "loss": 1.246,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.935715675354004,
+      "learning_rate": 2.6509353011841428e-05,
+      "loss": 1.2552,
+      "step": 1356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.152778625488281,
+      "learning_rate": 2.650677878839883e-05,
+      "loss": 1.1893,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.60451602935791,
+      "learning_rate": 2.650420456495624e-05,
+      "loss": 1.1447,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.021504402160645,
+      "learning_rate": 2.6501630341513645e-05,
+      "loss": 1.0769,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.891911506652832,
+      "learning_rate": 2.649905611807105e-05,
+      "loss": 0.9558,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.601040363311768,
+      "learning_rate": 2.6496481894628454e-05,
+      "loss": 1.0276,
+      "step": 1361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.832829475402832,
+      "learning_rate": 2.649390767118586e-05,
+      "loss": 1.287,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.218730926513672,
+      "learning_rate": 2.6491333447743264e-05,
+      "loss": 1.0422,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.534462928771973,
+      "learning_rate": 2.6488759224300668e-05,
+      "loss": 1.4117,
+      "step": 1364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.883268356323242,
+      "learning_rate": 2.6486185000858074e-05,
+      "loss": 1.2663,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.62038803100586,
+      "learning_rate": 2.6483610777415478e-05,
+      "loss": 1.1101,
+      "step": 1366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.019786834716797,
+      "learning_rate": 2.6481036553972888e-05,
+      "loss": 1.3157,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.632616996765137,
+      "learning_rate": 2.647846233053029e-05,
+      "loss": 1.1233,
+      "step": 1368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.388335227966309,
+      "learning_rate": 2.6475888107087698e-05,
+      "loss": 1.3165,
+      "step": 1369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.935334205627441,
+      "learning_rate": 2.64733138836451e-05,
+      "loss": 1.1324,
+      "step": 1370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.972235679626465,
+      "learning_rate": 2.6470739660202508e-05,
+      "loss": 1.2516,
+      "step": 1371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.370040893554688,
+      "learning_rate": 2.646816543675991e-05,
+      "loss": 1.1454,
+      "step": 1372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.459895610809326,
+      "learning_rate": 2.6465591213317318e-05,
+      "loss": 1.247,
+      "step": 1373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.179786682128906,
+      "learning_rate": 2.646301698987472e-05,
+      "loss": 1.3015,
+      "step": 1374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.019957542419434,
+      "learning_rate": 2.6460442766432124e-05,
+      "loss": 1.12,
+      "step": 1375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.15821647644043,
+      "learning_rate": 2.645786854298953e-05,
+      "loss": 1.3423,
+      "step": 1376
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.62690258026123,
+      "learning_rate": 2.6455294319546938e-05,
+      "loss": 1.3008,
+      "step": 1377
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.900343894958496,
+      "learning_rate": 2.6452720096104344e-05,
+      "loss": 1.6182,
+      "step": 1378
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.850994110107422,
+      "learning_rate": 2.6450145872661748e-05,
+      "loss": 1.2545,
+      "step": 1379
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.256598472595215,
+      "learning_rate": 2.6447571649219154e-05,
+      "loss": 0.8649,
+      "step": 1380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.804770469665527,
+      "learning_rate": 2.6444997425776558e-05,
+      "loss": 1.3885,
+      "step": 1381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.699951171875,
+      "learning_rate": 2.6442423202333964e-05,
+      "loss": 1.0952,
+      "step": 1382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.45407485961914,
+      "learning_rate": 2.6439848978891368e-05,
+      "loss": 1.3435,
+      "step": 1383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.335962295532227,
+      "learning_rate": 2.6437274755448774e-05,
+      "loss": 1.4324,
+      "step": 1384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.221283912658691,
+      "learning_rate": 2.6434700532006178e-05,
+      "loss": 1.1905,
+      "step": 1385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.06550407409668,
+      "learning_rate": 2.6432126308563584e-05,
+      "loss": 1.444,
+      "step": 1386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.470441818237305,
+      "learning_rate": 2.642955208512099e-05,
+      "loss": 1.2508,
+      "step": 1387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.260307312011719,
+      "learning_rate": 2.6426977861678394e-05,
+      "loss": 0.9873,
+      "step": 1388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.299884796142578,
+      "learning_rate": 2.64244036382358e-05,
+      "loss": 1.4163,
+      "step": 1389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.187042236328125,
+      "learning_rate": 2.6421829414793204e-05,
+      "loss": 1.1896,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.056967735290527,
+      "learning_rate": 2.641925519135061e-05,
+      "loss": 0.9773,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.281644821166992,
+      "learning_rate": 2.6416680967908014e-05,
+      "loss": 1.4031,
+      "step": 1392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.758089065551758,
+      "learning_rate": 2.641410674446542e-05,
+      "loss": 1.3926,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.50987720489502,
+      "learning_rate": 2.6411532521022824e-05,
+      "loss": 1.2747,
+      "step": 1394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.121760368347168,
+      "learning_rate": 2.6408958297580227e-05,
+      "loss": 1.3597,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.474566459655762,
+      "learning_rate": 2.6406384074137638e-05,
+      "loss": 1.4541,
+      "step": 1396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.610142707824707,
+      "learning_rate": 2.640380985069504e-05,
+      "loss": 1.3285,
+      "step": 1397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.047139167785645,
+      "learning_rate": 2.6401235627252447e-05,
+      "loss": 1.5605,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.22934341430664,
+      "learning_rate": 2.639866140380985e-05,
+      "loss": 1.3207,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.720802307128906,
+      "learning_rate": 2.6396087180367257e-05,
+      "loss": 1.3976,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.501425743103027,
+      "learning_rate": 2.639351295692466e-05,
+      "loss": 1.3132,
+      "step": 1401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.377949714660645,
+      "learning_rate": 2.6390938733482067e-05,
+      "loss": 1.3306,
+      "step": 1402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.295886039733887,
+      "learning_rate": 2.638836451003947e-05,
+      "loss": 1.4345,
+      "step": 1403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.783322811126709,
+      "learning_rate": 2.6385790286596877e-05,
+      "loss": 1.2414,
+      "step": 1404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.674589157104492,
+      "learning_rate": 2.6383216063154284e-05,
+      "loss": 1.4647,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.309123039245605,
+      "learning_rate": 2.6380641839711687e-05,
+      "loss": 1.1609,
+      "step": 1406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.010190963745117,
+      "learning_rate": 2.6378067616269094e-05,
+      "loss": 1.2418,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.536715507507324,
+      "learning_rate": 2.6375493392826497e-05,
+      "loss": 1.3751,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.624482154846191,
+      "learning_rate": 2.6372919169383904e-05,
+      "loss": 1.3539,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.700296401977539,
+      "learning_rate": 2.6370344945941307e-05,
+      "loss": 1.249,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.026456832885742,
+      "learning_rate": 2.6367770722498714e-05,
+      "loss": 0.8828,
+      "step": 1411
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.169084548950195,
+      "learning_rate": 2.6365196499056117e-05,
+      "loss": 1.102,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.534589767456055,
+      "learning_rate": 2.6362622275613524e-05,
+      "loss": 1.3815,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.347875595092773,
+      "learning_rate": 2.6360048052170927e-05,
+      "loss": 1.5063,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.11861515045166,
+      "learning_rate": 2.6357473828728334e-05,
+      "loss": 1.1653,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.822431564331055,
+      "learning_rate": 2.635489960528574e-05,
+      "loss": 1.5637,
+      "step": 1416
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.840667724609375,
+      "learning_rate": 2.6352325381843144e-05,
+      "loss": 1.0735,
+      "step": 1417
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.409327507019043,
+      "learning_rate": 2.634975115840055e-05,
+      "loss": 1.3696,
+      "step": 1418
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.566048622131348,
+      "learning_rate": 2.6347176934957954e-05,
+      "loss": 1.4983,
+      "step": 1419
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.011872291564941,
+      "learning_rate": 2.634460271151536e-05,
+      "loss": 1.4012,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.427055358886719,
+      "learning_rate": 2.6342028488072764e-05,
+      "loss": 0.9848,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.475969314575195,
+      "learning_rate": 2.633945426463017e-05,
+      "loss": 1.1438,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.266942501068115,
+      "learning_rate": 2.6336880041187574e-05,
+      "loss": 1.326,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.881961822509766,
+      "learning_rate": 2.6334305817744984e-05,
+      "loss": 1.2736,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.330427646636963,
+      "learning_rate": 2.6331731594302387e-05,
+      "loss": 1.2111,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.54010009765625,
+      "learning_rate": 2.632915737085979e-05,
+      "loss": 1.7535,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.168648719787598,
+      "learning_rate": 2.6326583147417197e-05,
+      "loss": 1.1077,
+      "step": 1427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.138731956481934,
+      "learning_rate": 2.63240089239746e-05,
+      "loss": 1.2385,
+      "step": 1428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.944056510925293,
+      "learning_rate": 2.6321434700532007e-05,
+      "loss": 1.1677,
+      "step": 1429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.671627044677734,
+      "learning_rate": 2.631886047708941e-05,
+      "loss": 1.2365,
+      "step": 1430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.128754615783691,
+      "learning_rate": 2.6316286253646817e-05,
+      "loss": 1.1243,
+      "step": 1431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.254434585571289,
+      "learning_rate": 2.631371203020422e-05,
+      "loss": 1.4588,
+      "step": 1432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.757545471191406,
+      "learning_rate": 2.6311137806761627e-05,
+      "loss": 1.2052,
+      "step": 1433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.631169319152832,
+      "learning_rate": 2.6308563583319034e-05,
+      "loss": 1.4379,
+      "step": 1434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.045636177062988,
+      "learning_rate": 2.630598935987644e-05,
+      "loss": 1.083,
+      "step": 1435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.049459457397461,
+      "learning_rate": 2.6303415136433844e-05,
+      "loss": 1.5538,
+      "step": 1436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.972939491271973,
+      "learning_rate": 2.6300840912991247e-05,
+      "loss": 1.223,
+      "step": 1437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.741581916809082,
+      "learning_rate": 2.6298266689548654e-05,
+      "loss": 1.4703,
+      "step": 1438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.267253875732422,
+      "learning_rate": 2.6295692466106057e-05,
+      "loss": 1.1549,
+      "step": 1439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.986956596374512,
+      "learning_rate": 2.6293118242663464e-05,
+      "loss": 1.5899,
+      "step": 1440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.246567726135254,
+      "learning_rate": 2.6290544019220867e-05,
+      "loss": 1.2799,
+      "step": 1441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.742122650146484,
+      "learning_rate": 2.6287969795778274e-05,
+      "loss": 1.2074,
+      "step": 1442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.831069946289062,
+      "learning_rate": 2.628539557233568e-05,
+      "loss": 1.1914,
+      "step": 1443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.123552322387695,
+      "learning_rate": 2.6282821348893087e-05,
+      "loss": 1.1744,
+      "step": 1444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.420841217041016,
+      "learning_rate": 2.628024712545049e-05,
+      "loss": 1.1882,
+      "step": 1445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.566805839538574,
+      "learning_rate": 2.6277672902007897e-05,
+      "loss": 1.1515,
+      "step": 1446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.217146873474121,
+      "learning_rate": 2.62750986785653e-05,
+      "loss": 1.154,
+      "step": 1447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.855069637298584,
+      "learning_rate": 2.6272524455122704e-05,
+      "loss": 0.9249,
+      "step": 1448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.529651641845703,
+      "learning_rate": 2.626995023168011e-05,
+      "loss": 0.9164,
+      "step": 1449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.14615249633789,
+      "learning_rate": 2.6267376008237514e-05,
+      "loss": 1.1964,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.80463981628418,
+      "learning_rate": 2.626480178479492e-05,
+      "loss": 1.1795,
+      "step": 1451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.780580520629883,
+      "learning_rate": 2.6262227561352324e-05,
+      "loss": 1.3118,
+      "step": 1452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.311175346374512,
+      "learning_rate": 2.6259653337909734e-05,
+      "loss": 1.2455,
+      "step": 1453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.386007308959961,
+      "learning_rate": 2.6257079114467137e-05,
+      "loss": 1.4528,
+      "step": 1454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.25583553314209,
+      "learning_rate": 2.6254504891024544e-05,
+      "loss": 1.3275,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.80235767364502,
+      "learning_rate": 2.6251930667581947e-05,
+      "loss": 1.5025,
+      "step": 1456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.165590286254883,
+      "learning_rate": 2.624935644413935e-05,
+      "loss": 1.1918,
+      "step": 1457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.783446311950684,
+      "learning_rate": 2.6246782220696757e-05,
+      "loss": 1.5329,
+      "step": 1458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.697442054748535,
+      "learning_rate": 2.624420799725416e-05,
+      "loss": 1.1324,
+      "step": 1459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.200699806213379,
+      "learning_rate": 2.6241633773811567e-05,
+      "loss": 1.028,
+      "step": 1460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.565408706665039,
+      "learning_rate": 2.623905955036897e-05,
+      "loss": 0.9422,
+      "step": 1461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.901795864105225,
+      "learning_rate": 2.623648532692638e-05,
+      "loss": 1.1371,
+      "step": 1462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.764126777648926,
+      "learning_rate": 2.6233911103483784e-05,
+      "loss": 1.1865,
+      "step": 1463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.938124656677246,
+      "learning_rate": 2.623133688004119e-05,
+      "loss": 1.2255,
+      "step": 1464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.876240730285645,
+      "learning_rate": 2.6228762656598594e-05,
+      "loss": 1.1229,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.972127914428711,
+      "learning_rate": 2.6226188433156e-05,
+      "loss": 1.4471,
+      "step": 1466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.517899513244629,
+      "learning_rate": 2.6223614209713403e-05,
+      "loss": 1.3616,
+      "step": 1467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.41218376159668,
+      "learning_rate": 2.6221039986270807e-05,
+      "loss": 1.1269,
+      "step": 1468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.990455627441406,
+      "learning_rate": 2.6218465762828213e-05,
+      "loss": 1.179,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.675251960754395,
+      "learning_rate": 2.6215891539385617e-05,
+      "loss": 1.1038,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.660655975341797,
+      "learning_rate": 2.6213317315943027e-05,
+      "loss": 0.9937,
+      "step": 1471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.12274169921875,
+      "learning_rate": 2.621074309250043e-05,
+      "loss": 1.1449,
+      "step": 1472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.534292697906494,
+      "learning_rate": 2.6208168869057837e-05,
+      "loss": 0.9945,
+      "step": 1473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.044416427612305,
+      "learning_rate": 2.620559464561524e-05,
+      "loss": 1.1414,
+      "step": 1474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.75588321685791,
+      "learning_rate": 2.6203020422172647e-05,
+      "loss": 1.5593,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.109524726867676,
+      "learning_rate": 2.620044619873005e-05,
+      "loss": 1.3625,
+      "step": 1476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.501923561096191,
+      "learning_rate": 2.6197871975287457e-05,
+      "loss": 1.2569,
+      "step": 1477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.203754425048828,
+      "learning_rate": 2.619529775184486e-05,
+      "loss": 1.412,
+      "step": 1478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.245170593261719,
+      "learning_rate": 2.6192723528402263e-05,
+      "loss": 1.1217,
+      "step": 1479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.047924995422363,
+      "learning_rate": 2.619014930495967e-05,
+      "loss": 0.9712,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.3880033493042,
+      "learning_rate": 2.6187575081517077e-05,
+      "loss": 1.1291,
+      "step": 1481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.66402530670166,
+      "learning_rate": 2.6185000858074483e-05,
+      "loss": 1.4886,
+      "step": 1482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.70718002319336,
+      "learning_rate": 2.6182426634631887e-05,
+      "loss": 1.0842,
+      "step": 1483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.335410118103027,
+      "learning_rate": 2.6179852411189293e-05,
+      "loss": 1.4253,
+      "step": 1484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.54088020324707,
+      "learning_rate": 2.6177278187746697e-05,
+      "loss": 1.2456,
+      "step": 1485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.075895309448242,
+      "learning_rate": 2.6174703964304103e-05,
+      "loss": 1.403,
+      "step": 1486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.078447341918945,
+      "learning_rate": 2.6172129740861507e-05,
+      "loss": 1.0031,
+      "step": 1487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.30174446105957,
+      "learning_rate": 2.6169555517418913e-05,
+      "loss": 1.1542,
+      "step": 1488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.88626480102539,
+      "learning_rate": 2.6166981293976317e-05,
+      "loss": 1.6623,
+      "step": 1489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.433023452758789,
+      "learning_rate": 2.6164407070533723e-05,
+      "loss": 0.9823,
+      "step": 1490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.406853675842285,
+      "learning_rate": 2.616183284709113e-05,
+      "loss": 1.1836,
+      "step": 1491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.292494773864746,
+      "learning_rate": 2.6159258623648533e-05,
+      "loss": 1.2547,
+      "step": 1492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.658616065979004,
+      "learning_rate": 2.615668440020594e-05,
+      "loss": 1.0207,
+      "step": 1493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.94169807434082,
+      "learning_rate": 2.6154110176763343e-05,
+      "loss": 1.0582,
+      "step": 1494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.028471946716309,
+      "learning_rate": 2.615153595332075e-05,
+      "loss": 1.3083,
+      "step": 1495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.022738456726074,
+      "learning_rate": 2.6148961729878153e-05,
+      "loss": 1.2612,
+      "step": 1496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.101866722106934,
+      "learning_rate": 2.614638750643556e-05,
+      "loss": 1.2579,
+      "step": 1497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.523536682128906,
+      "learning_rate": 2.6143813282992963e-05,
+      "loss": 1.1771,
+      "step": 1498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.217952728271484,
+      "learning_rate": 2.6141239059550366e-05,
+      "loss": 1.1969,
+      "step": 1499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.444476127624512,
+      "learning_rate": 2.6138664836107777e-05,
+      "loss": 1.7069,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.153836250305176,
+      "learning_rate": 2.613609061266518e-05,
+      "loss": 1.3156,
+      "step": 1501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.480145454406738,
+      "learning_rate": 2.6133516389222587e-05,
+      "loss": 1.0281,
+      "step": 1502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.882759094238281,
+      "learning_rate": 2.613094216577999e-05,
+      "loss": 1.6403,
+      "step": 1503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.847168922424316,
+      "learning_rate": 2.6128367942337396e-05,
+      "loss": 1.1885,
+      "step": 1504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.725817680358887,
+      "learning_rate": 2.61257937188948e-05,
+      "loss": 1.132,
+      "step": 1505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.618316650390625,
+      "learning_rate": 2.6123219495452206e-05,
+      "loss": 1.1898,
+      "step": 1506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.863554000854492,
+      "learning_rate": 2.612064527200961e-05,
+      "loss": 1.3372,
+      "step": 1507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.243571281433105,
+      "learning_rate": 2.6118071048567016e-05,
+      "loss": 1.1867,
+      "step": 1508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.022723197937012,
+      "learning_rate": 2.6115496825124423e-05,
+      "loss": 1.3467,
+      "step": 1509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.887679100036621,
+      "learning_rate": 2.6112922601681826e-05,
+      "loss": 1.2277,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.539658069610596,
+      "learning_rate": 2.6110348378239233e-05,
+      "loss": 1.161,
+      "step": 1511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.810113906860352,
+      "learning_rate": 2.6107774154796636e-05,
+      "loss": 1.3282,
+      "step": 1512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.964231967926025,
+      "learning_rate": 2.6105199931354043e-05,
+      "loss": 1.3472,
+      "step": 1513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.931862831115723,
+      "learning_rate": 2.6102625707911446e-05,
+      "loss": 1.1696,
+      "step": 1514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.483572959899902,
+      "learning_rate": 2.6100051484468853e-05,
+      "loss": 1.3194,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.26704216003418,
+      "learning_rate": 2.6097477261026256e-05,
+      "loss": 1.0045,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.326842308044434,
+      "learning_rate": 2.6094903037583663e-05,
+      "loss": 1.4534,
+      "step": 1517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.104940414428711,
+      "learning_rate": 2.6092328814141066e-05,
+      "loss": 1.2621,
+      "step": 1518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.792030334472656,
+      "learning_rate": 2.6089754590698476e-05,
+      "loss": 1.1254,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.367633819580078,
+      "learning_rate": 2.608718036725588e-05,
+      "loss": 1.1915,
+      "step": 1520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.074049949645996,
+      "learning_rate": 2.6084606143813283e-05,
+      "loss": 1.2102,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.285543441772461,
+      "learning_rate": 2.608203192037069e-05,
+      "loss": 1.3275,
+      "step": 1522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.799038887023926,
+      "learning_rate": 2.6079457696928093e-05,
+      "loss": 1.3434,
+      "step": 1523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.813976287841797,
+      "learning_rate": 2.60768834734855e-05,
+      "loss": 1.3586,
+      "step": 1524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.80359172821045,
+      "learning_rate": 2.6074309250042903e-05,
+      "loss": 1.1103,
+      "step": 1525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.167686462402344,
+      "learning_rate": 2.607173502660031e-05,
+      "loss": 1.1062,
+      "step": 1526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.026318550109863,
+      "learning_rate": 2.6069160803157713e-05,
+      "loss": 0.9971,
+      "step": 1527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.773980140686035,
+      "learning_rate": 2.6066586579715123e-05,
+      "loss": 1.4267,
+      "step": 1528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.633464813232422,
+      "learning_rate": 2.6064012356272526e-05,
+      "loss": 1.5509,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.131473541259766,
+      "learning_rate": 2.606143813282993e-05,
+      "loss": 1.0474,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.265179634094238,
+      "learning_rate": 2.6058863909387336e-05,
+      "loss": 1.294,
+      "step": 1531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.328397750854492,
+      "learning_rate": 2.605628968594474e-05,
+      "loss": 1.2807,
+      "step": 1532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.06914234161377,
+      "learning_rate": 2.6053715462502146e-05,
+      "loss": 1.2215,
+      "step": 1533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.479385375976562,
+      "learning_rate": 2.605114123905955e-05,
+      "loss": 1.2997,
+      "step": 1534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.318657875061035,
+      "learning_rate": 2.6048567015616956e-05,
+      "loss": 1.342,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.627128601074219,
+      "learning_rate": 2.604599279217436e-05,
+      "loss": 1.2197,
+      "step": 1536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.470630645751953,
+      "learning_rate": 2.6043418568731766e-05,
+      "loss": 1.3896,
+      "step": 1537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.276989936828613,
+      "learning_rate": 2.6040844345289173e-05,
+      "loss": 1.3584,
+      "step": 1538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.971053600311279,
+      "learning_rate": 2.603827012184658e-05,
+      "loss": 1.0443,
+      "step": 1539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.825150489807129,
+      "learning_rate": 2.6035695898403983e-05,
+      "loss": 1.3234,
+      "step": 1540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.27294921875,
+      "learning_rate": 2.6033121674961386e-05,
+      "loss": 1.2559,
+      "step": 1541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.06420612335205,
+      "learning_rate": 2.6030547451518793e-05,
+      "loss": 1.1736,
+      "step": 1542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.466202735900879,
+      "learning_rate": 2.6027973228076196e-05,
+      "loss": 1.3382,
+      "step": 1543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.974349975585938,
+      "learning_rate": 2.6025399004633603e-05,
+      "loss": 1.451,
+      "step": 1544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.396931648254395,
+      "learning_rate": 2.6022824781191006e-05,
+      "loss": 1.2475,
+      "step": 1545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.682100296020508,
+      "learning_rate": 2.6020250557748413e-05,
+      "loss": 0.9821,
+      "step": 1546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.19713306427002,
+      "learning_rate": 2.601767633430582e-05,
+      "loss": 1.1506,
+      "step": 1547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.293886184692383,
+      "learning_rate": 2.6015102110863226e-05,
+      "loss": 1.0679,
+      "step": 1548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.456722259521484,
+      "learning_rate": 2.601252788742063e-05,
+      "loss": 1.0798,
+      "step": 1549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.706494331359863,
+      "learning_rate": 2.6009953663978036e-05,
+      "loss": 1.4401,
+      "step": 1550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.215545654296875,
+      "learning_rate": 2.600737944053544e-05,
+      "loss": 0.8901,
+      "step": 1551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.596031188964844,
+      "learning_rate": 2.6004805217092843e-05,
+      "loss": 1.2893,
+      "step": 1552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.22983455657959,
+      "learning_rate": 2.600223099365025e-05,
+      "loss": 1.275,
+      "step": 1553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.633913040161133,
+      "learning_rate": 2.5999656770207653e-05,
+      "loss": 0.9564,
+      "step": 1554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.765681266784668,
+      "learning_rate": 2.599708254676506e-05,
+      "loss": 1.1862,
+      "step": 1555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.760149955749512,
+      "learning_rate": 2.5994508323322463e-05,
+      "loss": 1.1313,
+      "step": 1556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.333000183105469,
+      "learning_rate": 2.5991934099879873e-05,
+      "loss": 1.1004,
+      "step": 1557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.956636428833008,
+      "learning_rate": 2.5989359876437276e-05,
+      "loss": 1.691,
+      "step": 1558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.976466178894043,
+      "learning_rate": 2.5986785652994683e-05,
+      "loss": 1.0736,
+      "step": 1559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.870991230010986,
+      "learning_rate": 2.5984211429552086e-05,
+      "loss": 0.9734,
+      "step": 1560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.663421630859375,
+      "learning_rate": 2.598163720610949e-05,
+      "loss": 1.0836,
+      "step": 1561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.073293685913086,
+      "learning_rate": 2.5979062982666896e-05,
+      "loss": 1.2648,
+      "step": 1562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.528361320495605,
+      "learning_rate": 2.59764887592243e-05,
+      "loss": 1.0944,
+      "step": 1563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.58749008178711,
+      "learning_rate": 2.5973914535781706e-05,
+      "loss": 1.2068,
+      "step": 1564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.19024658203125,
+      "learning_rate": 2.597134031233911e-05,
+      "loss": 1.0406,
+      "step": 1565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.314001083374023,
+      "learning_rate": 2.596876608889652e-05,
+      "loss": 1.1504,
+      "step": 1566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.663982391357422,
+      "learning_rate": 2.5966191865453923e-05,
+      "loss": 0.9644,
+      "step": 1567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.71829605102539,
+      "learning_rate": 2.596361764201133e-05,
+      "loss": 0.8445,
+      "step": 1568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.616984367370605,
+      "learning_rate": 2.5961043418568733e-05,
+      "loss": 1.2033,
+      "step": 1569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.283010482788086,
+      "learning_rate": 2.595846919512614e-05,
+      "loss": 1.1312,
+      "step": 1570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.120360374450684,
+      "learning_rate": 2.5955894971683542e-05,
+      "loss": 1.0336,
+      "step": 1571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.295063972473145,
+      "learning_rate": 2.5953320748240946e-05,
+      "loss": 1.0758,
+      "step": 1572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.565741539001465,
+      "learning_rate": 2.5950746524798352e-05,
+      "loss": 1.3195,
+      "step": 1573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.618343353271484,
+      "learning_rate": 2.5948172301355756e-05,
+      "loss": 1.3803,
+      "step": 1574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.500362396240234,
+      "learning_rate": 2.5945598077913166e-05,
+      "loss": 1.1424,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.92062759399414,
+      "learning_rate": 2.594302385447057e-05,
+      "loss": 1.189,
+      "step": 1576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.251409530639648,
+      "learning_rate": 2.5940449631027976e-05,
+      "loss": 1.1997,
+      "step": 1577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.031352043151855,
+      "learning_rate": 2.593787540758538e-05,
+      "loss": 1.1386,
+      "step": 1578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.570923805236816,
+      "learning_rate": 2.5935301184142786e-05,
+      "loss": 1.2136,
+      "step": 1579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.8291654586792,
+      "learning_rate": 2.593272696070019e-05,
+      "loss": 1.0956,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.069900512695312,
+      "learning_rate": 2.5930152737257596e-05,
+      "loss": 1.3491,
+      "step": 1581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.534960746765137,
+      "learning_rate": 2.5927578513815e-05,
+      "loss": 1.3644,
+      "step": 1582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.055986404418945,
+      "learning_rate": 2.5925004290372402e-05,
+      "loss": 1.1657,
+      "step": 1583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.447620391845703,
+      "learning_rate": 2.592243006692981e-05,
+      "loss": 1.1843,
+      "step": 1584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.631956100463867,
+      "learning_rate": 2.5919855843487216e-05,
+      "loss": 1.0351,
+      "step": 1585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.020038604736328,
+      "learning_rate": 2.5917281620044622e-05,
+      "loss": 0.9871,
+      "step": 1586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.377992630004883,
+      "learning_rate": 2.5914707396602026e-05,
+      "loss": 1.4819,
+      "step": 1587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.236485481262207,
+      "learning_rate": 2.5912133173159432e-05,
+      "loss": 1.2268,
+      "step": 1588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.214728355407715,
+      "learning_rate": 2.5909558949716836e-05,
+      "loss": 1.075,
+      "step": 1589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.975008964538574,
+      "learning_rate": 2.5906984726274242e-05,
+      "loss": 0.9712,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.329899787902832,
+      "learning_rate": 2.5904410502831646e-05,
+      "loss": 1.2087,
+      "step": 1591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.313892364501953,
+      "learning_rate": 2.5901836279389052e-05,
+      "loss": 1.2299,
+      "step": 1592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.25578784942627,
+      "learning_rate": 2.5899262055946456e-05,
+      "loss": 1.1505,
+      "step": 1593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.274794578552246,
+      "learning_rate": 2.5896687832503862e-05,
+      "loss": 0.9779,
+      "step": 1594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.605783462524414,
+      "learning_rate": 2.589411360906127e-05,
+      "loss": 1.0434,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.504106521606445,
+      "learning_rate": 2.5891539385618672e-05,
+      "loss": 1.0292,
+      "step": 1596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.182815551757812,
+      "learning_rate": 2.588896516217608e-05,
+      "loss": 1.0919,
+      "step": 1597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.316807746887207,
+      "learning_rate": 2.5886390938733482e-05,
+      "loss": 1.0491,
+      "step": 1598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.9618000984191895,
+      "learning_rate": 2.588381671529089e-05,
+      "loss": 1.0251,
+      "step": 1599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.987043380737305,
+      "learning_rate": 2.5881242491848292e-05,
+      "loss": 1.3196,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.570785522460938,
+      "learning_rate": 2.58786682684057e-05,
+      "loss": 1.1278,
+      "step": 1601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.415283203125,
+      "learning_rate": 2.5876094044963102e-05,
+      "loss": 1.3667,
+      "step": 1602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.775483131408691,
+      "learning_rate": 2.5873519821520505e-05,
+      "loss": 1.266,
+      "step": 1603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.52068042755127,
+      "learning_rate": 2.5870945598077916e-05,
+      "loss": 1.5216,
+      "step": 1604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.699344635009766,
+      "learning_rate": 2.586837137463532e-05,
+      "loss": 1.1651,
+      "step": 1605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.356231689453125,
+      "learning_rate": 2.5865797151192726e-05,
+      "loss": 1.2763,
+      "step": 1606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.245226860046387,
+      "learning_rate": 2.586322292775013e-05,
+      "loss": 1.2878,
+      "step": 1607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.9049530029296875,
+      "learning_rate": 2.5860648704307535e-05,
+      "loss": 1.4374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.493533611297607,
+      "learning_rate": 2.585807448086494e-05,
+      "loss": 1.0598,
+      "step": 1609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.12313175201416,
+      "learning_rate": 2.5855500257422345e-05,
+      "loss": 1.3791,
+      "step": 1610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.579049587249756,
+      "learning_rate": 2.585292603397975e-05,
+      "loss": 1.0437,
+      "step": 1611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.188210487365723,
+      "learning_rate": 2.5850351810537155e-05,
+      "loss": 1.2966,
+      "step": 1612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.327075004577637,
+      "learning_rate": 2.5847777587094562e-05,
+      "loss": 1.2624,
+      "step": 1613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.03768539428711,
+      "learning_rate": 2.5845203363651965e-05,
+      "loss": 1.2136,
+      "step": 1614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.479057312011719,
+      "learning_rate": 2.5842629140209372e-05,
+      "loss": 1.2015,
+      "step": 1615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.179889678955078,
+      "learning_rate": 2.5840054916766775e-05,
+      "loss": 1.265,
+      "step": 1616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.410205841064453,
+      "learning_rate": 2.5837480693324182e-05,
+      "loss": 1.1442,
+      "step": 1617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.869892120361328,
+      "learning_rate": 2.5834906469881585e-05,
+      "loss": 1.3213,
+      "step": 1618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.214839935302734,
+      "learning_rate": 2.5832332246438992e-05,
+      "loss": 0.9812,
+      "step": 1619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.514745712280273,
+      "learning_rate": 2.5829758022996395e-05,
+      "loss": 1.3735,
+      "step": 1620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.812685489654541,
+      "learning_rate": 2.5827183799553802e-05,
+      "loss": 0.9847,
+      "step": 1621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.392012119293213,
+      "learning_rate": 2.5824609576111205e-05,
+      "loss": 1.0244,
+      "step": 1622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.04334545135498,
+      "learning_rate": 2.5822035352668615e-05,
+      "loss": 1.0548,
+      "step": 1623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.246118545532227,
+      "learning_rate": 2.581946112922602e-05,
+      "loss": 1.3623,
+      "step": 1624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.836997985839844,
+      "learning_rate": 2.5816886905783422e-05,
+      "loss": 1.1976,
+      "step": 1625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.115344047546387,
+      "learning_rate": 2.581431268234083e-05,
+      "loss": 1.3798,
+      "step": 1626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.662017822265625,
+      "learning_rate": 2.5811738458898232e-05,
+      "loss": 1.1289,
+      "step": 1627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.208451271057129,
+      "learning_rate": 2.580916423545564e-05,
+      "loss": 1.1622,
+      "step": 1628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.290624618530273,
+      "learning_rate": 2.5806590012013042e-05,
+      "loss": 1.1739,
+      "step": 1629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.374139785766602,
+      "learning_rate": 2.580401578857045e-05,
+      "loss": 1.1181,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.209418296813965,
+      "learning_rate": 2.5801441565127852e-05,
+      "loss": 1.1703,
+      "step": 1631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.328374862670898,
+      "learning_rate": 2.5798867341685262e-05,
+      "loss": 1.2326,
+      "step": 1632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.27452278137207,
+      "learning_rate": 2.5796293118242665e-05,
+      "loss": 1.2253,
+      "step": 1633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.137526512145996,
+      "learning_rate": 2.579371889480007e-05,
+      "loss": 1.0306,
+      "step": 1634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.130615234375,
+      "learning_rate": 2.5791144671357475e-05,
+      "loss": 1.0341,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.105877876281738,
+      "learning_rate": 2.578857044791488e-05,
+      "loss": 1.3181,
+      "step": 1636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.853985786437988,
+      "learning_rate": 2.5785996224472285e-05,
+      "loss": 1.3231,
+      "step": 1637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.044734954833984,
+      "learning_rate": 2.578342200102969e-05,
+      "loss": 1.1382,
+      "step": 1638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.615414142608643,
+      "learning_rate": 2.5780847777587095e-05,
+      "loss": 0.9365,
+      "step": 1639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.614706039428711,
+      "learning_rate": 2.57782735541445e-05,
+      "loss": 0.9821,
+      "step": 1640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.8402838706970215,
+      "learning_rate": 2.5775699330701905e-05,
+      "loss": 0.92,
+      "step": 1641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.08004093170166,
+      "learning_rate": 2.5773125107259312e-05,
+      "loss": 1.236,
+      "step": 1642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.8212308883667,
+      "learning_rate": 2.577055088381672e-05,
+      "loss": 1.1995,
+      "step": 1643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.835196495056152,
+      "learning_rate": 2.5767976660374122e-05,
+      "loss": 1.1684,
+      "step": 1644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.536199569702148,
+      "learning_rate": 2.5765402436931525e-05,
+      "loss": 1.1275,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.259613990783691,
+      "learning_rate": 2.5762828213488932e-05,
+      "loss": 0.9788,
+      "step": 1646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.785974502563477,
+      "learning_rate": 2.5760253990046335e-05,
+      "loss": 1.0851,
+      "step": 1647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.26859188079834,
+      "learning_rate": 2.5757679766603742e-05,
+      "loss": 1.1746,
+      "step": 1648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.363548278808594,
+      "learning_rate": 2.5755105543161145e-05,
+      "loss": 1.0277,
+      "step": 1649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.938871383666992,
+      "learning_rate": 2.5752531319718552e-05,
+      "loss": 1.199,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.887062072753906,
+      "learning_rate": 2.574995709627596e-05,
+      "loss": 0.7206,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.923095703125,
+      "learning_rate": 2.5747382872833365e-05,
+      "loss": 1.2589,
+      "step": 1652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.889364242553711,
+      "learning_rate": 2.574480864939077e-05,
+      "loss": 1.1392,
+      "step": 1653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.161477088928223,
+      "learning_rate": 2.5742234425948175e-05,
+      "loss": 1.312,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.382858276367188,
+      "learning_rate": 2.573966020250558e-05,
+      "loss": 1.2759,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.420166015625,
+      "learning_rate": 2.573708597906298e-05,
+      "loss": 1.1455,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.9227294921875,
+      "learning_rate": 2.573451175562039e-05,
+      "loss": 1.2758,
+      "step": 1657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.996256351470947,
+      "learning_rate": 2.573193753217779e-05,
+      "loss": 1.275,
+      "step": 1658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.737109184265137,
+      "learning_rate": 2.57293633087352e-05,
+      "loss": 1.4071,
+      "step": 1659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.454028129577637,
+      "learning_rate": 2.57267890852926e-05,
+      "loss": 0.8918,
+      "step": 1660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.383586883544922,
+      "learning_rate": 2.572421486185001e-05,
+      "loss": 1.1086,
+      "step": 1661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.172550201416016,
+      "learning_rate": 2.5721640638407415e-05,
+      "loss": 1.3437,
+      "step": 1662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.78988265991211,
+      "learning_rate": 2.571906641496482e-05,
+      "loss": 1.4695,
+      "step": 1663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.235658645629883,
+      "learning_rate": 2.5716492191522225e-05,
+      "loss": 1.36,
+      "step": 1664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.791393756866455,
+      "learning_rate": 2.5713917968079628e-05,
+      "loss": 1.2821,
+      "step": 1665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.439427375793457,
+      "learning_rate": 2.5711343744637035e-05,
+      "loss": 1.2101,
+      "step": 1666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.998271942138672,
+      "learning_rate": 2.5708769521194438e-05,
+      "loss": 1.2884,
+      "step": 1667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.0204439163208,
+      "learning_rate": 2.5706195297751845e-05,
+      "loss": 1.1096,
+      "step": 1668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.507468223571777,
+      "learning_rate": 2.5703621074309248e-05,
+      "loss": 1.2128,
+      "step": 1669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.155778884887695,
+      "learning_rate": 2.5701046850866658e-05,
+      "loss": 1.1895,
+      "step": 1670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.207258224487305,
+      "learning_rate": 2.569847262742406e-05,
+      "loss": 1.2311,
+      "step": 1671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.249384880065918,
+      "learning_rate": 2.5695898403981468e-05,
+      "loss": 1.2064,
+      "step": 1672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.2927937507629395,
+      "learning_rate": 2.569332418053887e-05,
+      "loss": 0.9132,
+      "step": 1673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.482983589172363,
+      "learning_rate": 2.5690749957096278e-05,
+      "loss": 1.0854,
+      "step": 1674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.173653602600098,
+      "learning_rate": 2.568817573365368e-05,
+      "loss": 1.2891,
+      "step": 1675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.37519645690918,
+      "learning_rate": 2.5685601510211085e-05,
+      "loss": 0.9443,
+      "step": 1676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.192808151245117,
+      "learning_rate": 2.568302728676849e-05,
+      "loss": 1.1677,
+      "step": 1677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.935910701751709,
+      "learning_rate": 2.5680453063325895e-05,
+      "loss": 1.0965,
+      "step": 1678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.470190048217773,
+      "learning_rate": 2.56778788398833e-05,
+      "loss": 1.2907,
+      "step": 1679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.00965690612793,
+      "learning_rate": 2.5675304616440708e-05,
+      "loss": 1.3334,
+      "step": 1680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.875039100646973,
+      "learning_rate": 2.5672730392998115e-05,
+      "loss": 1.1551,
+      "step": 1681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.871654033660889,
+      "learning_rate": 2.5670156169555518e-05,
+      "loss": 1.0475,
+      "step": 1682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.366538047790527,
+      "learning_rate": 2.5667581946112925e-05,
+      "loss": 1.1854,
+      "step": 1683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.781619548797607,
+      "learning_rate": 2.5665007722670328e-05,
+      "loss": 1.0515,
+      "step": 1684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.574727058410645,
+      "learning_rate": 2.5662433499227735e-05,
+      "loss": 1.1163,
+      "step": 1685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.076858520507812,
+      "learning_rate": 2.5659859275785138e-05,
+      "loss": 1.151,
+      "step": 1686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.47282600402832,
+      "learning_rate": 2.565728505234254e-05,
+      "loss": 1.091,
+      "step": 1687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.974745750427246,
+      "learning_rate": 2.5654710828899948e-05,
+      "loss": 1.1764,
+      "step": 1688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.930011749267578,
+      "learning_rate": 2.5652136605457355e-05,
+      "loss": 1.2829,
+      "step": 1689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.917031288146973,
+      "learning_rate": 2.564956238201476e-05,
+      "loss": 1.5465,
+      "step": 1690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.354314804077148,
+      "learning_rate": 2.5646988158572165e-05,
+      "loss": 1.4153,
+      "step": 1691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.695298671722412,
+      "learning_rate": 2.564441393512957e-05,
+      "loss": 0.8676,
+      "step": 1692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.762816429138184,
+      "learning_rate": 2.5641839711686975e-05,
+      "loss": 1.0837,
+      "step": 1693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.785277366638184,
+      "learning_rate": 2.563926548824438e-05,
+      "loss": 1.0689,
+      "step": 1694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.864838600158691,
+      "learning_rate": 2.5636691264801785e-05,
+      "loss": 1.176,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.101361274719238,
+      "learning_rate": 2.563411704135919e-05,
+      "loss": 1.0014,
+      "step": 1696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.913434028625488,
+      "learning_rate": 2.5631542817916595e-05,
+      "loss": 1.0631,
+      "step": 1697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.819299697875977,
+      "learning_rate": 2.5628968594474e-05,
+      "loss": 1.2241,
+      "step": 1698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.142266273498535,
+      "learning_rate": 2.5626394371031408e-05,
+      "loss": 0.9767,
+      "step": 1699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.407238960266113,
+      "learning_rate": 2.562382014758881e-05,
+      "loss": 1.2434,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.397313117980957,
+      "learning_rate": 2.5621245924146218e-05,
+      "loss": 1.0554,
+      "step": 1701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.991972923278809,
+      "learning_rate": 2.561867170070362e-05,
+      "loss": 1.0965,
+      "step": 1702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.07584285736084,
+      "learning_rate": 2.5616097477261028e-05,
+      "loss": 0.9142,
+      "step": 1703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.226771354675293,
+      "learning_rate": 2.561352325381843e-05,
+      "loss": 0.9517,
+      "step": 1704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.419251441955566,
+      "learning_rate": 2.5610949030375838e-05,
+      "loss": 1.5665,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.380773544311523,
+      "learning_rate": 2.560837480693324e-05,
+      "loss": 1.1231,
+      "step": 1706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.461137771606445,
+      "learning_rate": 2.5605800583490644e-05,
+      "loss": 1.1659,
+      "step": 1707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.826033592224121,
+      "learning_rate": 2.5603226360048055e-05,
+      "loss": 1.2141,
+      "step": 1708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.870662689208984,
+      "learning_rate": 2.5600652136605458e-05,
+      "loss": 1.3264,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.239055633544922,
+      "learning_rate": 2.5598077913162865e-05,
+      "loss": 1.2444,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.364582061767578,
+      "learning_rate": 2.5595503689720268e-05,
+      "loss": 1.0615,
+      "step": 1711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.995655059814453,
+      "learning_rate": 2.5592929466277675e-05,
+      "loss": 1.1462,
+      "step": 1712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.26231575012207,
+      "learning_rate": 2.5590355242835078e-05,
+      "loss": 0.9843,
+      "step": 1713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.783554077148438,
+      "learning_rate": 2.5587781019392484e-05,
+      "loss": 1.1091,
+      "step": 1714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.424436569213867,
+      "learning_rate": 2.5585206795949888e-05,
+      "loss": 1.2304,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.603520393371582,
+      "learning_rate": 2.5582632572507294e-05,
+      "loss": 1.3525,
+      "step": 1716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.819580078125,
+      "learning_rate": 2.55800583490647e-05,
+      "loss": 1.5046,
+      "step": 1717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.068258285522461,
+      "learning_rate": 2.5577484125622104e-05,
+      "loss": 1.2893,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.862461090087891,
+      "learning_rate": 2.557490990217951e-05,
+      "loss": 1.0952,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.446170806884766,
+      "learning_rate": 2.5572335678736914e-05,
+      "loss": 1.3306,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.586013793945312,
+      "learning_rate": 2.556976145529432e-05,
+      "loss": 1.1739,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.612980842590332,
+      "learning_rate": 2.5567187231851724e-05,
+      "loss": 1.215,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.444388389587402,
+      "learning_rate": 2.556461300840913e-05,
+      "loss": 1.1309,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.63848876953125,
+      "learning_rate": 2.5562038784966534e-05,
+      "loss": 1.3445,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.80753231048584,
+      "learning_rate": 2.555946456152394e-05,
+      "loss": 1.0968,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.618228435516357,
+      "learning_rate": 2.5556890338081344e-05,
+      "loss": 0.9463,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.364997863769531,
+      "learning_rate": 2.5554316114638754e-05,
+      "loss": 1.1908,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.8157958984375,
+      "learning_rate": 2.5551741891196158e-05,
+      "loss": 1.0125,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.717795372009277,
+      "learning_rate": 2.554916766775356e-05,
+      "loss": 1.2761,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.9950590133667,
+      "learning_rate": 2.5546593444310968e-05,
+      "loss": 1.376,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.8765716552734375,
+      "learning_rate": 2.554401922086837e-05,
+      "loss": 1.1086,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.276371002197266,
+      "learning_rate": 2.5541444997425778e-05,
+      "loss": 1.3216,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.059179306030273,
+      "learning_rate": 2.553887077398318e-05,
+      "loss": 0.9557,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.337965965270996,
+      "learning_rate": 2.5536296550540588e-05,
+      "loss": 0.8447,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.218817710876465,
+      "learning_rate": 2.553372232709799e-05,
+      "loss": 1.3292,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.164855003356934,
+      "learning_rate": 2.55311481036554e-05,
+      "loss": 0.9875,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.308812141418457,
+      "learning_rate": 2.5528573880212804e-05,
+      "loss": 1.1649,
+      "step": 1737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.199872970581055,
+      "learning_rate": 2.5525999656770208e-05,
+      "loss": 1.5198,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.293773174285889,
+      "learning_rate": 2.5523425433327614e-05,
+      "loss": 0.9806,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.174756050109863,
+      "learning_rate": 2.5520851209885018e-05,
+      "loss": 1.1593,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.54172420501709,
+      "learning_rate": 2.5518276986442424e-05,
+      "loss": 1.1001,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.815516471862793,
+      "learning_rate": 2.5515702762999828e-05,
+      "loss": 1.2876,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.003947257995605,
+      "learning_rate": 2.5513128539557234e-05,
+      "loss": 1.1743,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.989374160766602,
+      "learning_rate": 2.5510554316114637e-05,
+      "loss": 1.1333,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.674088478088379,
+      "learning_rate": 2.5507980092672044e-05,
+      "loss": 1.0615,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.114914894104004,
+      "learning_rate": 2.550540586922945e-05,
+      "loss": 0.9093,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.078839302062988,
+      "learning_rate": 2.5502831645786858e-05,
+      "loss": 1.2257,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.610944747924805,
+      "learning_rate": 2.550025742234426e-05,
+      "loss": 1.3549,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.474281311035156,
+      "learning_rate": 2.5497683198901664e-05,
+      "loss": 1.3019,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.139394760131836,
+      "learning_rate": 2.549510897545907e-05,
+      "loss": 1.0164,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.196832656860352,
+      "learning_rate": 2.5492534752016474e-05,
+      "loss": 0.9659,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.259092330932617,
+      "learning_rate": 2.548996052857388e-05,
+      "loss": 1.2302,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.978081703186035,
+      "learning_rate": 2.5487386305131284e-05,
+      "loss": 1.5391,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.210874557495117,
+      "learning_rate": 2.548481208168869e-05,
+      "loss": 1.0691,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.688666343688965,
+      "learning_rate": 2.5482237858246097e-05,
+      "loss": 1.291,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.124794960021973,
+      "learning_rate": 2.5479663634803504e-05,
+      "loss": 1.1149,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.819000244140625,
+      "learning_rate": 2.5477089411360907e-05,
+      "loss": 1.4636,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.510807991027832,
+      "learning_rate": 2.5474515187918314e-05,
+      "loss": 1.155,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.002891540527344,
+      "learning_rate": 2.5471940964475717e-05,
+      "loss": 1.1307,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.713222026824951,
+      "learning_rate": 2.546936674103312e-05,
+      "loss": 0.823,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.20630168914795,
+      "learning_rate": 2.5466792517590527e-05,
+      "loss": 1.0278,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.881501197814941,
+      "learning_rate": 2.546421829414793e-05,
+      "loss": 1.1759,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.865151405334473,
+      "learning_rate": 2.5461644070705337e-05,
+      "loss": 1.0436,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.163095474243164,
+      "learning_rate": 2.545906984726274e-05,
+      "loss": 1.1672,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.597827911376953,
+      "learning_rate": 2.545649562382015e-05,
+      "loss": 0.9883,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.688282012939453,
+      "learning_rate": 2.5453921400377554e-05,
+      "loss": 1.1817,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.824360847473145,
+      "learning_rate": 2.545134717693496e-05,
+      "loss": 1.0302,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.243977546691895,
+      "learning_rate": 2.5448772953492364e-05,
+      "loss": 1.3014,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.205143928527832,
+      "learning_rate": 2.5446198730049767e-05,
+      "loss": 1.1799,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.626049995422363,
+      "learning_rate": 2.5443624506607174e-05,
+      "loss": 1.2875,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.9079484939575195,
+      "learning_rate": 2.5441050283164577e-05,
+      "loss": 0.9806,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.699675559997559,
+      "learning_rate": 2.5438476059721984e-05,
+      "loss": 1.2925,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.091042518615723,
+      "learning_rate": 2.5435901836279387e-05,
+      "loss": 1.1246,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.246540069580078,
+      "learning_rate": 2.5433327612836797e-05,
+      "loss": 1.3969,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.48653507232666,
+      "learning_rate": 2.54307533893942e-05,
+      "loss": 1.5656,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.698553085327148,
+      "learning_rate": 2.5428179165951607e-05,
+      "loss": 1.3264,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.113031387329102,
+      "learning_rate": 2.542560494250901e-05,
+      "loss": 1.349,
+      "step": 1777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.13812255859375,
+      "learning_rate": 2.5423030719066417e-05,
+      "loss": 1.2943,
+      "step": 1778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.228736877441406,
+      "learning_rate": 2.542045649562382e-05,
+      "loss": 1.1007,
+      "step": 1779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.414812088012695,
+      "learning_rate": 2.5417882272181224e-05,
+      "loss": 1.3554,
+      "step": 1780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.727490425109863,
+      "learning_rate": 2.541530804873863e-05,
+      "loss": 1.0629,
+      "step": 1781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.751081466674805,
+      "learning_rate": 2.5412733825296034e-05,
+      "loss": 1.0592,
+      "step": 1782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.771555423736572,
+      "learning_rate": 2.541015960185344e-05,
+      "loss": 1.2257,
+      "step": 1783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.105414390563965,
+      "learning_rate": 2.5407585378410847e-05,
+      "loss": 1.0661,
+      "step": 1784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.473969459533691,
+      "learning_rate": 2.5405011154968254e-05,
+      "loss": 1.3305,
+      "step": 1785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.59005355834961,
+      "learning_rate": 2.5402436931525657e-05,
+      "loss": 1.2249,
+      "step": 1786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.714675903320312,
+      "learning_rate": 2.5399862708083064e-05,
+      "loss": 1.05,
+      "step": 1787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.160584449768066,
+      "learning_rate": 2.5397288484640467e-05,
+      "loss": 1.1466,
+      "step": 1788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.179819107055664,
+      "learning_rate": 2.5394714261197874e-05,
+      "loss": 1.1973,
+      "step": 1789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.159802436828613,
+      "learning_rate": 2.5392140037755277e-05,
+      "loss": 1.049,
+      "step": 1790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.245061874389648,
+      "learning_rate": 2.538956581431268e-05,
+      "loss": 1.1937,
+      "step": 1791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.878827095031738,
+      "learning_rate": 2.5386991590870087e-05,
+      "loss": 1.2525,
+      "step": 1792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.605012893676758,
+      "learning_rate": 2.5384417367427494e-05,
+      "loss": 1.3157,
+      "step": 1793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.687931060791016,
+      "learning_rate": 2.53818431439849e-05,
+      "loss": 1.1184,
+      "step": 1794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.726444244384766,
+      "learning_rate": 2.5379268920542304e-05,
+      "loss": 1.0972,
+      "step": 1795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.861693382263184,
+      "learning_rate": 2.537669469709971e-05,
+      "loss": 1.4293,
+      "step": 1796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.042244911193848,
+      "learning_rate": 2.5374120473657114e-05,
+      "loss": 1.2176,
+      "step": 1797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.745487213134766,
+      "learning_rate": 2.537154625021452e-05,
+      "loss": 1.0754,
+      "step": 1798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.054213523864746,
+      "learning_rate": 2.5368972026771924e-05,
+      "loss": 1.2724,
+      "step": 1799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.261024475097656,
+      "learning_rate": 2.536639780332933e-05,
+      "loss": 1.2302,
+      "step": 1800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.732303619384766,
+      "learning_rate": 2.5363823579886734e-05,
+      "loss": 1.4895,
+      "step": 1801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.90101957321167,
+      "learning_rate": 2.5361249356444137e-05,
+      "loss": 0.7279,
+      "step": 1802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.161076545715332,
+      "learning_rate": 2.5358675133001547e-05,
+      "loss": 1.0414,
+      "step": 1803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.7752532958984375,
+      "learning_rate": 2.535610090955895e-05,
+      "loss": 1.0087,
+      "step": 1804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.58970832824707,
+      "learning_rate": 2.5353526686116357e-05,
+      "loss": 1.0037,
+      "step": 1805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.076090812683105,
+      "learning_rate": 2.535095246267376e-05,
+      "loss": 1.2012,
+      "step": 1806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.196894645690918,
+      "learning_rate": 2.5348378239231167e-05,
+      "loss": 1.153,
+      "step": 1807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.644584655761719,
+      "learning_rate": 2.534580401578857e-05,
+      "loss": 0.9725,
+      "step": 1808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.094841003417969,
+      "learning_rate": 2.5343229792345977e-05,
+      "loss": 1.504,
+      "step": 1809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.794973373413086,
+      "learning_rate": 2.534065556890338e-05,
+      "loss": 1.2477,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.357816696166992,
+      "learning_rate": 2.5338081345460783e-05,
+      "loss": 1.0704,
+      "step": 1811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.8945417404174805,
+      "learning_rate": 2.5335507122018194e-05,
+      "loss": 1.0943,
+      "step": 1812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.366178512573242,
+      "learning_rate": 2.5332932898575597e-05,
+      "loss": 1.0265,
+      "step": 1813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.090495109558105,
+      "learning_rate": 2.5330358675133004e-05,
+      "loss": 1.2477,
+      "step": 1814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.481625556945801,
+      "learning_rate": 2.5327784451690407e-05,
+      "loss": 1.0535,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.137459754943848,
+      "learning_rate": 2.5325210228247814e-05,
+      "loss": 1.3704,
+      "step": 1816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.223336219787598,
+      "learning_rate": 2.5322636004805217e-05,
+      "loss": 1.2863,
+      "step": 1817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.916875839233398,
+      "learning_rate": 2.5320061781362623e-05,
+      "loss": 1.3063,
+      "step": 1818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.86544418334961,
+      "learning_rate": 2.5317487557920027e-05,
+      "loss": 1.1277,
+      "step": 1819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.277205467224121,
+      "learning_rate": 2.5314913334477433e-05,
+      "loss": 1.3575,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.45832347869873,
+      "learning_rate": 2.531233911103484e-05,
+      "loss": 1.416,
+      "step": 1821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.004668235778809,
+      "learning_rate": 2.5309764887592243e-05,
+      "loss": 1.2249,
+      "step": 1822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.491814136505127,
+      "learning_rate": 2.530719066414965e-05,
+      "loss": 0.8599,
+      "step": 1823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.60008430480957,
+      "learning_rate": 2.5304616440707053e-05,
+      "loss": 1.1542,
+      "step": 1824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.817724227905273,
+      "learning_rate": 2.530204221726446e-05,
+      "loss": 1.3333,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.774344444274902,
+      "learning_rate": 2.5299467993821863e-05,
+      "loss": 1.4904,
+      "step": 1826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.157302379608154,
+      "learning_rate": 2.529689377037927e-05,
+      "loss": 1.0191,
+      "step": 1827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.208282470703125,
+      "learning_rate": 2.5294319546936673e-05,
+      "loss": 1.0935,
+      "step": 1828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.650165557861328,
+      "learning_rate": 2.529174532349408e-05,
+      "loss": 1.1832,
+      "step": 1829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.321013450622559,
+      "learning_rate": 2.5289171100051483e-05,
+      "loss": 1.3029,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.568893432617188,
+      "learning_rate": 2.5286596876608893e-05,
+      "loss": 0.9535,
+      "step": 1831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.806543350219727,
+      "learning_rate": 2.5284022653166297e-05,
+      "loss": 0.7877,
+      "step": 1832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.580751419067383,
+      "learning_rate": 2.52814484297237e-05,
+      "loss": 1.1122,
+      "step": 1833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.052247047424316,
+      "learning_rate": 2.5278874206281107e-05,
+      "loss": 1.1541,
+      "step": 1834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.40207290649414,
+      "learning_rate": 2.527629998283851e-05,
+      "loss": 1.3442,
+      "step": 1835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.848437309265137,
+      "learning_rate": 2.5273725759395917e-05,
+      "loss": 1.3851,
+      "step": 1836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.450674057006836,
+      "learning_rate": 2.527115153595332e-05,
+      "loss": 1.0922,
+      "step": 1837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.358359336853027,
+      "learning_rate": 2.5268577312510727e-05,
+      "loss": 1.1509,
+      "step": 1838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.33609390258789,
+      "learning_rate": 2.526600308906813e-05,
+      "loss": 1.2894,
+      "step": 1839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.63636589050293,
+      "learning_rate": 2.526342886562554e-05,
+      "loss": 0.9445,
+      "step": 1840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.046621322631836,
+      "learning_rate": 2.5260854642182943e-05,
+      "loss": 1.2915,
+      "step": 1841
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.787140846252441,
+      "learning_rate": 2.5258280418740347e-05,
+      "loss": 1.0923,
+      "step": 1842
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.000146865844727,
+      "learning_rate": 2.5255706195297753e-05,
+      "loss": 1.2291,
+      "step": 1843
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.815464973449707,
+      "learning_rate": 2.5253131971855157e-05,
+      "loss": 1.1863,
+      "step": 1844
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.100621223449707,
+      "learning_rate": 2.5250557748412563e-05,
+      "loss": 1.186,
+      "step": 1845
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.029321670532227,
+      "learning_rate": 2.5247983524969967e-05,
+      "loss": 1.1971,
+      "step": 1846
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.647441864013672,
+      "learning_rate": 2.5245409301527373e-05,
+      "loss": 0.9327,
+      "step": 1847
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.917750358581543,
+      "learning_rate": 2.5242835078084776e-05,
+      "loss": 1.3352,
+      "step": 1848
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.372392654418945,
+      "learning_rate": 2.5240260854642183e-05,
+      "loss": 1.2961,
+      "step": 1849
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.865127563476562,
+      "learning_rate": 2.523768663119959e-05,
+      "loss": 1.1148,
+      "step": 1850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.053229331970215,
+      "learning_rate": 2.5235112407756997e-05,
+      "loss": 1.089,
+      "step": 1851
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.55601692199707,
+      "learning_rate": 2.52325381843144e-05,
+      "loss": 0.9104,
+      "step": 1852
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.748366355895996,
+      "learning_rate": 2.5229963960871803e-05,
+      "loss": 1.2221,
+      "step": 1853
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.148165702819824,
+      "learning_rate": 2.522738973742921e-05,
+      "loss": 0.9479,
+      "step": 1854
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.006891250610352,
+      "learning_rate": 2.5224815513986613e-05,
+      "loss": 1.0907,
+      "step": 1855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.19072437286377,
+      "learning_rate": 2.522224129054402e-05,
+      "loss": 1.0549,
+      "step": 1856
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.527332305908203,
+      "learning_rate": 2.5219667067101423e-05,
+      "loss": 1.3011,
+      "step": 1857
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.543940544128418,
+      "learning_rate": 2.521709284365883e-05,
+      "loss": 1.231,
+      "step": 1858
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.198790550231934,
+      "learning_rate": 2.5214518620216236e-05,
+      "loss": 0.9201,
+      "step": 1859
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.62169361114502,
+      "learning_rate": 2.5211944396773643e-05,
+      "loss": 1.3424,
+      "step": 1860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.031617164611816,
+      "learning_rate": 2.5209370173331046e-05,
+      "loss": 1.4437,
+      "step": 1861
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.64932632446289,
+      "learning_rate": 2.5206795949888453e-05,
+      "loss": 1.3224,
+      "step": 1862
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.811517715454102,
+      "learning_rate": 2.5204221726445856e-05,
+      "loss": 1.0686,
+      "step": 1863
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.884221076965332,
+      "learning_rate": 2.520164750300326e-05,
+      "loss": 1.07,
+      "step": 1864
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.0543694496154785,
+      "learning_rate": 2.5199073279560666e-05,
+      "loss": 0.9218,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.33708381652832,
+      "learning_rate": 2.519649905611807e-05,
+      "loss": 1.3307,
+      "step": 1866
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.017557144165039,
+      "learning_rate": 2.5193924832675476e-05,
+      "loss": 1.0409,
+      "step": 1867
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.777966499328613,
+      "learning_rate": 2.519135060923288e-05,
+      "loss": 1.1855,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.000447273254395,
+      "learning_rate": 2.518877638579029e-05,
+      "loss": 0.8718,
+      "step": 1869
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.26295280456543,
+      "learning_rate": 2.5186202162347693e-05,
+      "loss": 1.3592,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.32453727722168,
+      "learning_rate": 2.51836279389051e-05,
+      "loss": 1.0087,
+      "step": 1871
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.395368576049805,
+      "learning_rate": 2.5181053715462503e-05,
+      "loss": 0.9962,
+      "step": 1872
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.196331977844238,
+      "learning_rate": 2.5178479492019906e-05,
+      "loss": 1.1135,
+      "step": 1873
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.52117919921875,
+      "learning_rate": 2.5175905268577313e-05,
+      "loss": 1.1214,
+      "step": 1874
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.943233489990234,
+      "learning_rate": 2.5173331045134716e-05,
+      "loss": 1.4006,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.03986930847168,
+      "learning_rate": 2.5170756821692123e-05,
+      "loss": 0.9299,
+      "step": 1876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.859893798828125,
+      "learning_rate": 2.5168182598249526e-05,
+      "loss": 1.413,
+      "step": 1877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.069256782531738,
+      "learning_rate": 2.5165608374806936e-05,
+      "loss": 1.2064,
+      "step": 1878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.76870346069336,
+      "learning_rate": 2.516303415136434e-05,
+      "loss": 1.116,
+      "step": 1879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.318020820617676,
+      "learning_rate": 2.5160459927921746e-05,
+      "loss": 1.0757,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.963784217834473,
+      "learning_rate": 2.515788570447915e-05,
+      "loss": 1.2655,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.540380477905273,
+      "learning_rate": 2.5155311481036556e-05,
+      "loss": 1.0673,
+      "step": 1882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.368696212768555,
+      "learning_rate": 2.515273725759396e-05,
+      "loss": 1.072,
+      "step": 1883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.880315780639648,
+      "learning_rate": 2.5150163034151363e-05,
+      "loss": 1.3764,
+      "step": 1884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.173974990844727,
+      "learning_rate": 2.514758881070877e-05,
+      "loss": 1.2389,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.06794261932373,
+      "learning_rate": 2.5145014587266173e-05,
+      "loss": 1.4193,
+      "step": 1886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.367919921875,
+      "learning_rate": 2.514244036382358e-05,
+      "loss": 1.0134,
+      "step": 1887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.085611343383789,
+      "learning_rate": 2.5139866140380986e-05,
+      "loss": 1.2374,
+      "step": 1888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.164103507995605,
+      "learning_rate": 2.5137291916938393e-05,
+      "loss": 0.9791,
+      "step": 1889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.86156940460205,
+      "learning_rate": 2.5134717693495796e-05,
+      "loss": 1.2038,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.802731513977051,
+      "learning_rate": 2.5132143470053203e-05,
+      "loss": 0.9789,
+      "step": 1891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.810247898101807,
+      "learning_rate": 2.5129569246610606e-05,
+      "loss": 1.0637,
+      "step": 1892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.733325004577637,
+      "learning_rate": 2.5126995023168013e-05,
+      "loss": 1.2844,
+      "step": 1893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.678900718688965,
+      "learning_rate": 2.5124420799725416e-05,
+      "loss": 0.9974,
+      "step": 1894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.847447395324707,
+      "learning_rate": 2.512184657628282e-05,
+      "loss": 1.1089,
+      "step": 1895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.557967185974121,
+      "learning_rate": 2.5119272352840226e-05,
+      "loss": 1.1357,
+      "step": 1896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.160346984863281,
+      "learning_rate": 2.5116698129397633e-05,
+      "loss": 1.2265,
+      "step": 1897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.898780822753906,
+      "learning_rate": 2.511412390595504e-05,
+      "loss": 1.3529,
+      "step": 1898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.820026397705078,
+      "learning_rate": 2.5111549682512443e-05,
+      "loss": 1.1643,
+      "step": 1899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.765425682067871,
+      "learning_rate": 2.510897545906985e-05,
+      "loss": 0.9822,
+      "step": 1900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.185986518859863,
+      "learning_rate": 2.5106401235627253e-05,
+      "loss": 1.3487,
+      "step": 1901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.548151969909668,
+      "learning_rate": 2.510382701218466e-05,
+      "loss": 1.4098,
+      "step": 1902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.726716041564941,
+      "learning_rate": 2.5101252788742063e-05,
+      "loss": 1.2139,
+      "step": 1903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.740248680114746,
+      "learning_rate": 2.509867856529947e-05,
+      "loss": 0.8818,
+      "step": 1904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.98962688446045,
+      "learning_rate": 2.5096104341856873e-05,
+      "loss": 1.2039,
+      "step": 1905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.169604301452637,
+      "learning_rate": 2.5093530118414276e-05,
+      "loss": 1.1749,
+      "step": 1906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.317733764648438,
+      "learning_rate": 2.5090955894971686e-05,
+      "loss": 1.0668,
+      "step": 1907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.362645149230957,
+      "learning_rate": 2.508838167152909e-05,
+      "loss": 0.8885,
+      "step": 1908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.150195121765137,
+      "learning_rate": 2.5085807448086496e-05,
+      "loss": 1.0103,
+      "step": 1909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.825246810913086,
+      "learning_rate": 2.50832332246439e-05,
+      "loss": 0.9991,
+      "step": 1910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.028517723083496,
+      "learning_rate": 2.5080659001201306e-05,
+      "loss": 1.1897,
+      "step": 1911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.390515327453613,
+      "learning_rate": 2.507808477775871e-05,
+      "loss": 1.0032,
+      "step": 1912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.067804336547852,
+      "learning_rate": 2.5075510554316116e-05,
+      "loss": 1.1421,
+      "step": 1913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.462785720825195,
+      "learning_rate": 2.507293633087352e-05,
+      "loss": 1.1792,
+      "step": 1914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.050780296325684,
+      "learning_rate": 2.5070362107430923e-05,
+      "loss": 1.0274,
+      "step": 1915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.930495262145996,
+      "learning_rate": 2.5067787883988333e-05,
+      "loss": 1.1926,
+      "step": 1916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.409160614013672,
+      "learning_rate": 2.5065213660545736e-05,
+      "loss": 1.2423,
+      "step": 1917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.221443176269531,
+      "learning_rate": 2.5062639437103143e-05,
+      "loss": 1.0732,
+      "step": 1918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.196868896484375,
+      "learning_rate": 2.5060065213660546e-05,
+      "loss": 0.9071,
+      "step": 1919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.157271385192871,
+      "learning_rate": 2.5057490990217953e-05,
+      "loss": 1.0358,
+      "step": 1920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.6808500289917,
+      "learning_rate": 2.5054916766775356e-05,
+      "loss": 1.1404,
+      "step": 1921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.573707580566406,
+      "learning_rate": 2.5052342543332763e-05,
+      "loss": 1.2259,
+      "step": 1922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.4099702835083,
+      "learning_rate": 2.5049768319890166e-05,
+      "loss": 0.9453,
+      "step": 1923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.807967185974121,
+      "learning_rate": 2.5047194096447572e-05,
+      "loss": 1.1408,
+      "step": 1924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.258343696594238,
+      "learning_rate": 2.504461987300498e-05,
+      "loss": 1.0049,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.263693809509277,
+      "learning_rate": 2.5042045649562382e-05,
+      "loss": 1.1866,
+      "step": 1926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.111815452575684,
+      "learning_rate": 2.503947142611979e-05,
+      "loss": 1.3251,
+      "step": 1927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.936982154846191,
+      "learning_rate": 2.5036897202677192e-05,
+      "loss": 1.181,
+      "step": 1928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.280904769897461,
+      "learning_rate": 2.50343229792346e-05,
+      "loss": 0.9024,
+      "step": 1929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.838895797729492,
+      "learning_rate": 2.5031748755792002e-05,
+      "loss": 1.4286,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.120624542236328,
+      "learning_rate": 2.502917453234941e-05,
+      "loss": 1.2885,
+      "step": 1931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.158381462097168,
+      "learning_rate": 2.5026600308906812e-05,
+      "loss": 1.1287,
+      "step": 1932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.586956024169922,
+      "learning_rate": 2.502402608546422e-05,
+      "loss": 1.1301,
+      "step": 1933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.403854370117188,
+      "learning_rate": 2.5021451862021622e-05,
+      "loss": 1.3198,
+      "step": 1934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.893551349639893,
+      "learning_rate": 2.5018877638579032e-05,
+      "loss": 1.3,
+      "step": 1935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.152052879333496,
+      "learning_rate": 2.5016303415136436e-05,
+      "loss": 1.2528,
+      "step": 1936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.34033203125,
+      "learning_rate": 2.501372919169384e-05,
+      "loss": 1.2822,
+      "step": 1937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.420156478881836,
+      "learning_rate": 2.5011154968251246e-05,
+      "loss": 1.0191,
+      "step": 1938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.021442413330078,
+      "learning_rate": 2.500858074480865e-05,
+      "loss": 1.2212,
+      "step": 1939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.50693130493164,
+      "learning_rate": 2.5006006521366056e-05,
+      "loss": 1.3093,
+      "step": 1940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.51550579071045,
+      "learning_rate": 2.500343229792346e-05,
+      "loss": 1.0646,
+      "step": 1941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.394984245300293,
+      "learning_rate": 2.5000858074480866e-05,
+      "loss": 1.1632,
+      "step": 1942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.698156356811523,
+      "learning_rate": 2.499828385103827e-05,
+      "loss": 1.3627,
+      "step": 1943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.20472240447998,
+      "learning_rate": 2.499570962759568e-05,
+      "loss": 1.2292,
+      "step": 1944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.285568237304688,
+      "learning_rate": 2.4993135404153082e-05,
+      "loss": 1.4758,
+      "step": 1945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.318853378295898,
+      "learning_rate": 2.4990561180710486e-05,
+      "loss": 1.2371,
+      "step": 1946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.862863540649414,
+      "learning_rate": 2.4987986957267892e-05,
+      "loss": 1.0279,
+      "step": 1947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.340022087097168,
+      "learning_rate": 2.4985412733825296e-05,
+      "loss": 1.1188,
+      "step": 1948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.183253765106201,
+      "learning_rate": 2.4982838510382702e-05,
+      "loss": 0.9651,
+      "step": 1949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.86628532409668,
+      "learning_rate": 2.4980264286940106e-05,
+      "loss": 1.2317,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.645370483398438,
+      "learning_rate": 2.4977690063497512e-05,
+      "loss": 1.209,
+      "step": 1951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.665984153747559,
+      "learning_rate": 2.4975115840054916e-05,
+      "loss": 0.9434,
+      "step": 1952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.246528625488281,
+      "learning_rate": 2.4972541616612322e-05,
+      "loss": 1.4181,
+      "step": 1953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.156112670898438,
+      "learning_rate": 2.496996739316973e-05,
+      "loss": 1.1267,
+      "step": 1954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.022880554199219,
+      "learning_rate": 2.4967393169727136e-05,
+      "loss": 1.1695,
+      "step": 1955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.308635711669922,
+      "learning_rate": 2.496481894628454e-05,
+      "loss": 1.2813,
+      "step": 1956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.928953170776367,
+      "learning_rate": 2.4962244722841942e-05,
+      "loss": 1.0572,
+      "step": 1957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.274496078491211,
+      "learning_rate": 2.495967049939935e-05,
+      "loss": 0.9021,
+      "step": 1958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.08324909210205,
+      "learning_rate": 2.4957096275956752e-05,
+      "loss": 1.0692,
+      "step": 1959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.38923454284668,
+      "learning_rate": 2.495452205251416e-05,
+      "loss": 1.103,
+      "step": 1960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.054323196411133,
+      "learning_rate": 2.4951947829071562e-05,
+      "loss": 1.2061,
+      "step": 1961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.091486930847168,
+      "learning_rate": 2.494937360562897e-05,
+      "loss": 1.1214,
+      "step": 1962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.852002143859863,
+      "learning_rate": 2.4946799382186375e-05,
+      "loss": 0.8592,
+      "step": 1963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.171123504638672,
+      "learning_rate": 2.4944225158743782e-05,
+      "loss": 0.8868,
+      "step": 1964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.343501091003418,
+      "learning_rate": 2.4941650935301185e-05,
+      "loss": 1.3639,
+      "step": 1965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.870195388793945,
+      "learning_rate": 2.4939076711858592e-05,
+      "loss": 1.1631,
+      "step": 1966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.815661430358887,
+      "learning_rate": 2.4936502488415995e-05,
+      "loss": 1.1268,
+      "step": 1967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.165377616882324,
+      "learning_rate": 2.49339282649734e-05,
+      "loss": 0.9369,
+      "step": 1968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.087328910827637,
+      "learning_rate": 2.4931354041530805e-05,
+      "loss": 1.0966,
+      "step": 1969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.71306324005127,
+      "learning_rate": 2.492877981808821e-05,
+      "loss": 1.432,
+      "step": 1970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.951461791992188,
+      "learning_rate": 2.4926205594645615e-05,
+      "loss": 1.2412,
+      "step": 1971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.967318534851074,
+      "learning_rate": 2.492363137120302e-05,
+      "loss": 1.2419,
+      "step": 1972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.283565521240234,
+      "learning_rate": 2.492105714776043e-05,
+      "loss": 1.1954,
+      "step": 1973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.887789726257324,
+      "learning_rate": 2.4918482924317832e-05,
+      "loss": 1.6527,
+      "step": 1974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.543041229248047,
+      "learning_rate": 2.491590870087524e-05,
+      "loss": 1.1091,
+      "step": 1975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.84188461303711,
+      "learning_rate": 2.4913334477432642e-05,
+      "loss": 1.2852,
+      "step": 1976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.968687057495117,
+      "learning_rate": 2.491076025399005e-05,
+      "loss": 1.5774,
+      "step": 1977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.033377647399902,
+      "learning_rate": 2.4908186030547452e-05,
+      "loss": 0.9287,
+      "step": 1978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.477993965148926,
+      "learning_rate": 2.4905611807104855e-05,
+      "loss": 1.3667,
+      "step": 1979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.788491249084473,
+      "learning_rate": 2.4903037583662262e-05,
+      "loss": 1.4238,
+      "step": 1980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.339115142822266,
+      "learning_rate": 2.4900463360219665e-05,
+      "loss": 1.0565,
+      "step": 1981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.48612117767334,
+      "learning_rate": 2.4897889136777075e-05,
+      "loss": 1.3312,
+      "step": 1982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.537652015686035,
+      "learning_rate": 2.489531491333448e-05,
+      "loss": 1.2015,
+      "step": 1983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.066079139709473,
+      "learning_rate": 2.4892740689891885e-05,
+      "loss": 1.1343,
+      "step": 1984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.819148063659668,
+      "learning_rate": 2.489016646644929e-05,
+      "loss": 1.1123,
+      "step": 1985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.840709686279297,
+      "learning_rate": 2.4887592243006695e-05,
+      "loss": 1.1617,
+      "step": 1986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.1939115524292,
+      "learning_rate": 2.48850180195641e-05,
+      "loss": 1.3644,
+      "step": 1987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.358109474182129,
+      "learning_rate": 2.4882443796121502e-05,
+      "loss": 1.3645,
+      "step": 1988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.732720375061035,
+      "learning_rate": 2.487986957267891e-05,
+      "loss": 1.033,
+      "step": 1989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.227747917175293,
+      "learning_rate": 2.4877295349236312e-05,
+      "loss": 1.1426,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.248335838317871,
+      "learning_rate": 2.487472112579372e-05,
+      "loss": 1.0489,
+      "step": 1991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.124897956848145,
+      "learning_rate": 2.4872146902351125e-05,
+      "loss": 0.9307,
+      "step": 1992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.019804954528809,
+      "learning_rate": 2.4869572678908532e-05,
+      "loss": 1.4164,
+      "step": 1993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.667664527893066,
+      "learning_rate": 2.4866998455465935e-05,
+      "loss": 1.0404,
+      "step": 1994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.261445045471191,
+      "learning_rate": 2.4864424232023342e-05,
+      "loss": 1.0281,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.9775824546813965,
+      "learning_rate": 2.4861850008580745e-05,
+      "loss": 0.8719,
+      "step": 1996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.73702621459961,
+      "learning_rate": 2.4859275785138152e-05,
+      "loss": 1.1269,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.8519062995910645,
+      "learning_rate": 2.4856701561695555e-05,
+      "loss": 1.0876,
+      "step": 1998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.00592041015625,
+      "learning_rate": 2.485412733825296e-05,
+      "loss": 0.7832,
+      "step": 1999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.4707441329956055,
+      "learning_rate": 2.4851553114810365e-05,
+      "loss": 1.2687,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.627842903137207,
+      "learning_rate": 2.4848978891367772e-05,
+      "loss": 1.1093,
+      "step": 2001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.832741737365723,
+      "learning_rate": 2.484640466792518e-05,
+      "loss": 1.2518,
+      "step": 2002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.39503288269043,
+      "learning_rate": 2.4843830444482582e-05,
+      "loss": 1.2994,
+      "step": 2003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.167266845703125,
+      "learning_rate": 2.484125622103999e-05,
+      "loss": 0.9416,
+      "step": 2004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.453071117401123,
+      "learning_rate": 2.4838681997597392e-05,
+      "loss": 0.8271,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.479333400726318,
+      "learning_rate": 2.48361077741548e-05,
+      "loss": 1.0426,
+      "step": 2006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.295331001281738,
+      "learning_rate": 2.48335335507122e-05,
+      "loss": 1.0965,
+      "step": 2007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.796544075012207,
+      "learning_rate": 2.483095932726961e-05,
+      "loss": 1.4362,
+      "step": 2008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.850000381469727,
+      "learning_rate": 2.482838510382701e-05,
+      "loss": 1.3482,
+      "step": 2009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.764659881591797,
+      "learning_rate": 2.4825810880384415e-05,
+      "loss": 1.2939,
+      "step": 2010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.835600852966309,
+      "learning_rate": 2.4823236656941825e-05,
+      "loss": 1.2095,
+      "step": 2011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.962176322937012,
+      "learning_rate": 2.4820662433499228e-05,
+      "loss": 1.0814,
+      "step": 2012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.528059959411621,
+      "learning_rate": 2.4818088210056635e-05,
+      "loss": 0.8885,
+      "step": 2013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.620846748352051,
+      "learning_rate": 2.4815513986614038e-05,
+      "loss": 1.1178,
+      "step": 2014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.787505626678467,
+      "learning_rate": 2.4812939763171445e-05,
+      "loss": 1.0418,
+      "step": 2015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.11175537109375,
+      "learning_rate": 2.4810365539728848e-05,
+      "loss": 1.3541,
+      "step": 2016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.326554298400879,
+      "learning_rate": 2.4807791316286255e-05,
+      "loss": 1.4547,
+      "step": 2017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.435914039611816,
+      "learning_rate": 2.4805217092843658e-05,
+      "loss": 1.3384,
+      "step": 2018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.599921226501465,
+      "learning_rate": 2.480264286940106e-05,
+      "loss": 0.9495,
+      "step": 2019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.479714393615723,
+      "learning_rate": 2.480006864595847e-05,
+      "loss": 1.3211,
+      "step": 2020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.65898609161377,
+      "learning_rate": 2.4797494422515875e-05,
+      "loss": 1.1568,
+      "step": 2021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.756569862365723,
+      "learning_rate": 2.479492019907328e-05,
+      "loss": 1.0404,
+      "step": 2022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.057429313659668,
+      "learning_rate": 2.4792345975630685e-05,
+      "loss": 1.0557,
+      "step": 2023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.154457092285156,
+      "learning_rate": 2.478977175218809e-05,
+      "loss": 1.3369,
+      "step": 2024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.76572322845459,
+      "learning_rate": 2.4787197528745495e-05,
+      "loss": 1.0762,
+      "step": 2025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.778600692749023,
+      "learning_rate": 2.47846233053029e-05,
+      "loss": 1.3628,
+      "step": 2026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.858481407165527,
+      "learning_rate": 2.4782049081860305e-05,
+      "loss": 1.1526,
+      "step": 2027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.333586692810059,
+      "learning_rate": 2.477947485841771e-05,
+      "loss": 1.0982,
+      "step": 2028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.965689659118652,
+      "learning_rate": 2.4776900634975115e-05,
+      "loss": 0.9712,
+      "step": 2029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.018839836120605,
+      "learning_rate": 2.477432641153252e-05,
+      "loss": 1.3437,
+      "step": 2030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.67626953125,
+      "learning_rate": 2.4771752188089928e-05,
+      "loss": 1.0218,
+      "step": 2031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.6708984375,
+      "learning_rate": 2.476917796464733e-05,
+      "loss": 1.0562,
+      "step": 2032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.617552757263184,
+      "learning_rate": 2.4766603741204738e-05,
+      "loss": 0.9068,
+      "step": 2033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.860594749450684,
+      "learning_rate": 2.476402951776214e-05,
+      "loss": 1.108,
+      "step": 2034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.098282814025879,
+      "learning_rate": 2.4761455294319548e-05,
+      "loss": 0.9153,
+      "step": 2035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.727445602416992,
+      "learning_rate": 2.475888107087695e-05,
+      "loss": 1.2464,
+      "step": 2036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.118781089782715,
+      "learning_rate": 2.4756306847434358e-05,
+      "loss": 1.2765,
+      "step": 2037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.542216300964355,
+      "learning_rate": 2.475373262399176e-05,
+      "loss": 0.8026,
+      "step": 2038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.281174659729004,
+      "learning_rate": 2.475115840054917e-05,
+      "loss": 0.855,
+      "step": 2039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.210792541503906,
+      "learning_rate": 2.4748584177106575e-05,
+      "loss": 1.0354,
+      "step": 2040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.555898666381836,
+      "learning_rate": 2.4746009953663978e-05,
+      "loss": 0.6123,
+      "step": 2041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.490297317504883,
+      "learning_rate": 2.4743435730221385e-05,
+      "loss": 1.1734,
+      "step": 2042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.4724702835083,
+      "learning_rate": 2.4740861506778788e-05,
+      "loss": 1.2586,
+      "step": 2043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.335028648376465,
+      "learning_rate": 2.4738287283336195e-05,
+      "loss": 1.2043,
+      "step": 2044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.547687530517578,
+      "learning_rate": 2.4735713059893598e-05,
+      "loss": 1.2341,
+      "step": 2045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.621874809265137,
+      "learning_rate": 2.4733138836451005e-05,
+      "loss": 1.0011,
+      "step": 2046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.322978019714355,
+      "learning_rate": 2.4730564613008408e-05,
+      "loss": 1.2744,
+      "step": 2047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.387273788452148,
+      "learning_rate": 2.4727990389565818e-05,
+      "loss": 1.2313,
+      "step": 2048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.519750595092773,
+      "learning_rate": 2.472541616612322e-05,
+      "loss": 1.2329,
+      "step": 2049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.675507545471191,
+      "learning_rate": 2.4722841942680625e-05,
+      "loss": 1.3132,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.72492504119873,
+      "learning_rate": 2.472026771923803e-05,
+      "loss": 1.1005,
+      "step": 2051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.680103302001953,
+      "learning_rate": 2.4717693495795435e-05,
+      "loss": 1.2857,
+      "step": 2052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.613232135772705,
+      "learning_rate": 2.471511927235284e-05,
+      "loss": 1.1919,
+      "step": 2053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.367504119873047,
+      "learning_rate": 2.4712545048910245e-05,
+      "loss": 1.0231,
+      "step": 2054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.910067081451416,
+      "learning_rate": 2.470997082546765e-05,
+      "loss": 0.9046,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.293525695800781,
+      "learning_rate": 2.4707396602025055e-05,
+      "loss": 1.2671,
+      "step": 2056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.532302856445312,
+      "learning_rate": 2.470482237858246e-05,
+      "loss": 1.4923,
+      "step": 2057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.526620864868164,
+      "learning_rate": 2.4702248155139868e-05,
+      "loss": 1.2102,
+      "step": 2058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.08618450164795,
+      "learning_rate": 2.4699673931697275e-05,
+      "loss": 1.1245,
+      "step": 2059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.348907470703125,
+      "learning_rate": 2.4697099708254678e-05,
+      "loss": 1.041,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.697951316833496,
+      "learning_rate": 2.469452548481208e-05,
+      "loss": 1.3327,
+      "step": 2061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.185370922088623,
+      "learning_rate": 2.4691951261369488e-05,
+      "loss": 0.9218,
+      "step": 2062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.989435195922852,
+      "learning_rate": 2.468937703792689e-05,
+      "loss": 0.8377,
+      "step": 2063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.306659698486328,
+      "learning_rate": 2.4686802814484298e-05,
+      "loss": 0.9904,
+      "step": 2064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.163015365600586,
+      "learning_rate": 2.46842285910417e-05,
+      "loss": 1.2342,
+      "step": 2065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.481640815734863,
+      "learning_rate": 2.4681654367599108e-05,
+      "loss": 1.0154,
+      "step": 2066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.87024211883545,
+      "learning_rate": 2.4679080144156514e-05,
+      "loss": 1.123,
+      "step": 2067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.469578742980957,
+      "learning_rate": 2.467650592071392e-05,
+      "loss": 1.4519,
+      "step": 2068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.79355239868164,
+      "learning_rate": 2.4673931697271324e-05,
+      "loss": 1.1126,
+      "step": 2069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.070310592651367,
+      "learning_rate": 2.467135747382873e-05,
+      "loss": 1.2017,
+      "step": 2070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.575870513916016,
+      "learning_rate": 2.4668783250386134e-05,
+      "loss": 1.3892,
+      "step": 2071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.681638717651367,
+      "learning_rate": 2.4666209026943538e-05,
+      "loss": 1.4273,
+      "step": 2072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.094696044921875,
+      "learning_rate": 2.4663634803500944e-05,
+      "loss": 1.1297,
+      "step": 2073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.590570449829102,
+      "learning_rate": 2.4661060580058348e-05,
+      "loss": 1.0709,
+      "step": 2074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.218464851379395,
+      "learning_rate": 2.4658486356615754e-05,
+      "loss": 1.2096,
+      "step": 2075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.500626564025879,
+      "learning_rate": 2.4655912133173158e-05,
+      "loss": 1.3741,
+      "step": 2076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.936196327209473,
+      "learning_rate": 2.4653337909730568e-05,
+      "loss": 1.0692,
+      "step": 2077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.131011009216309,
+      "learning_rate": 2.465076368628797e-05,
+      "loss": 1.255,
+      "step": 2078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.559833526611328,
+      "learning_rate": 2.4648189462845378e-05,
+      "loss": 1.0649,
+      "step": 2079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.875581741333008,
+      "learning_rate": 2.464561523940278e-05,
+      "loss": 1.0418,
+      "step": 2080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.681763172149658,
+      "learning_rate": 2.4643041015960188e-05,
+      "loss": 1.1694,
+      "step": 2081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.073424339294434,
+      "learning_rate": 2.464046679251759e-05,
+      "loss": 1.0565,
+      "step": 2082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.330310821533203,
+      "learning_rate": 2.4637892569074994e-05,
+      "loss": 1.1167,
+      "step": 2083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.346527099609375,
+      "learning_rate": 2.46353183456324e-05,
+      "loss": 1.3122,
+      "step": 2084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.09740924835205,
+      "learning_rate": 2.4632744122189804e-05,
+      "loss": 1.0779,
+      "step": 2085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.799961566925049,
+      "learning_rate": 2.4630169898747214e-05,
+      "loss": 1.0795,
+      "step": 2086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.641633987426758,
+      "learning_rate": 2.4627595675304618e-05,
+      "loss": 1.0811,
+      "step": 2087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.578268051147461,
+      "learning_rate": 2.4625021451862024e-05,
+      "loss": 1.0085,
+      "step": 2088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.661127090454102,
+      "learning_rate": 2.4622447228419428e-05,
+      "loss": 1.0115,
+      "step": 2089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.416635513305664,
+      "learning_rate": 2.4619873004976834e-05,
+      "loss": 1.3297,
+      "step": 2090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.808404922485352,
+      "learning_rate": 2.4617298781534238e-05,
+      "loss": 1.2156,
+      "step": 2091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.63053035736084,
+      "learning_rate": 2.461472455809164e-05,
+      "loss": 1.1493,
+      "step": 2092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.865976333618164,
+      "learning_rate": 2.4612150334649048e-05,
+      "loss": 1.2221,
+      "step": 2093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.930879592895508,
+      "learning_rate": 2.460957611120645e-05,
+      "loss": 1.1561,
+      "step": 2094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.664885520935059,
+      "learning_rate": 2.4607001887763857e-05,
+      "loss": 1.2096,
+      "step": 2095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.464540481567383,
+      "learning_rate": 2.4604427664321264e-05,
+      "loss": 0.7703,
+      "step": 2096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.401766777038574,
+      "learning_rate": 2.460185344087867e-05,
+      "loss": 1.0532,
+      "step": 2097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.557998657226562,
+      "learning_rate": 2.4599279217436074e-05,
+      "loss": 1.0615,
+      "step": 2098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.52142333984375,
+      "learning_rate": 2.459670499399348e-05,
+      "loss": 1.373,
+      "step": 2099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.780794143676758,
+      "learning_rate": 2.4594130770550884e-05,
+      "loss": 1.3402,
+      "step": 2100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.523842811584473,
+      "learning_rate": 2.459155654710829e-05,
+      "loss": 1.0245,
+      "step": 2101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.324873924255371,
+      "learning_rate": 2.4588982323665694e-05,
+      "loss": 0.7332,
+      "step": 2102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.999865531921387,
+      "learning_rate": 2.4586408100223097e-05,
+      "loss": 1.238,
+      "step": 2103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.391741752624512,
+      "learning_rate": 2.4583833876780504e-05,
+      "loss": 0.992,
+      "step": 2104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.214604377746582,
+      "learning_rate": 2.458125965333791e-05,
+      "loss": 0.9706,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.142857551574707,
+      "learning_rate": 2.4578685429895317e-05,
+      "loss": 1.2553,
+      "step": 2106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.071646690368652,
+      "learning_rate": 2.457611120645272e-05,
+      "loss": 0.8326,
+      "step": 2107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.613232612609863,
+      "learning_rate": 2.4573536983010127e-05,
+      "loss": 0.7947,
+      "step": 2108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.005168914794922,
+      "learning_rate": 2.457096275956753e-05,
+      "loss": 1.0563,
+      "step": 2109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.119216918945312,
+      "learning_rate": 2.4568388536124937e-05,
+      "loss": 0.992,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.742559432983398,
+      "learning_rate": 2.456581431268234e-05,
+      "loss": 1.1163,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.28771686553955,
+      "learning_rate": 2.4563240089239747e-05,
+      "loss": 1.2394,
+      "step": 2112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.60543155670166,
+      "learning_rate": 2.456066586579715e-05,
+      "loss": 1.2264,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.486471176147461,
+      "learning_rate": 2.4558091642354554e-05,
+      "loss": 1.3191,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.7002534866333,
+      "learning_rate": 2.4555517418911964e-05,
+      "loss": 1.0788,
+      "step": 2115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.65157413482666,
+      "learning_rate": 2.4552943195469367e-05,
+      "loss": 1.0665,
+      "step": 2116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.475146293640137,
+      "learning_rate": 2.4550368972026774e-05,
+      "loss": 1.131,
+      "step": 2117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.699040412902832,
+      "learning_rate": 2.4547794748584177e-05,
+      "loss": 1.0983,
+      "step": 2118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.982903480529785,
+      "learning_rate": 2.4545220525141584e-05,
+      "loss": 1.1224,
+      "step": 2119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.772966384887695,
+      "learning_rate": 2.4542646301698987e-05,
+      "loss": 1.167,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.30195426940918,
+      "learning_rate": 2.4540072078256394e-05,
+      "loss": 1.1701,
+      "step": 2121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.401288032531738,
+      "learning_rate": 2.4537497854813797e-05,
+      "loss": 1.1435,
+      "step": 2122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.261945724487305,
+      "learning_rate": 2.45349236313712e-05,
+      "loss": 1.1808,
+      "step": 2123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.20970630645752,
+      "learning_rate": 2.453234940792861e-05,
+      "loss": 1.2177,
+      "step": 2124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.664337158203125,
+      "learning_rate": 2.4529775184486014e-05,
+      "loss": 1.0365,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.273855209350586,
+      "learning_rate": 2.452720096104342e-05,
+      "loss": 1.1114,
+      "step": 2126
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.142509460449219,
+      "learning_rate": 2.4524626737600824e-05,
+      "loss": 1.1827,
+      "step": 2127
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.21065902709961,
+      "learning_rate": 2.452205251415823e-05,
+      "loss": 1.1838,
+      "step": 2128
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.823779106140137,
+      "learning_rate": 2.4519478290715634e-05,
+      "loss": 1.3231,
+      "step": 2129
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.876160621643066,
+      "learning_rate": 2.451690406727304e-05,
+      "loss": 1.041,
+      "step": 2130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.582056999206543,
+      "learning_rate": 2.4514329843830444e-05,
+      "loss": 1.3357,
+      "step": 2131
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.97353744506836,
+      "learning_rate": 2.451175562038785e-05,
+      "loss": 0.9429,
+      "step": 2132
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.752605438232422,
+      "learning_rate": 2.4509181396945254e-05,
+      "loss": 1.0655,
+      "step": 2133
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.494879722595215,
+      "learning_rate": 2.450660717350266e-05,
+      "loss": 1.2365,
+      "step": 2134
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.78968620300293,
+      "learning_rate": 2.4504032950060067e-05,
+      "loss": 1.3128,
+      "step": 2135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.948899269104004,
+      "learning_rate": 2.450145872661747e-05,
+      "loss": 1.0771,
+      "step": 2136
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.892141342163086,
+      "learning_rate": 2.4498884503174877e-05,
+      "loss": 1.4076,
+      "step": 2137
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.983065605163574,
+      "learning_rate": 2.449631027973228e-05,
+      "loss": 1.0138,
+      "step": 2138
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.955356121063232,
+      "learning_rate": 2.4493736056289687e-05,
+      "loss": 0.8654,
+      "step": 2139
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.503843307495117,
+      "learning_rate": 2.449116183284709e-05,
+      "loss": 1.051,
+      "step": 2140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.763766288757324,
+      "learning_rate": 2.4488587609404497e-05,
+      "loss": 1.3045,
+      "step": 2141
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.153258323669434,
+      "learning_rate": 2.44860133859619e-05,
+      "loss": 1.0647,
+      "step": 2142
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.2325439453125,
+      "learning_rate": 2.448343916251931e-05,
+      "loss": 1.2145,
+      "step": 2143
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.163873672485352,
+      "learning_rate": 2.4480864939076714e-05,
+      "loss": 1.0033,
+      "step": 2144
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.364749908447266,
+      "learning_rate": 2.4478290715634117e-05,
+      "loss": 1.3655,
+      "step": 2145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.249857902526855,
+      "learning_rate": 2.4475716492191524e-05,
+      "loss": 1.2007,
+      "step": 2146
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.793172836303711,
+      "learning_rate": 2.4473142268748927e-05,
+      "loss": 0.8409,
+      "step": 2147
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.970074653625488,
+      "learning_rate": 2.4470568045306334e-05,
+      "loss": 1.1618,
+      "step": 2148
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.811893463134766,
+      "learning_rate": 2.4467993821863737e-05,
+      "loss": 0.9153,
+      "step": 2149
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.616089820861816,
+      "learning_rate": 2.4465419598421144e-05,
+      "loss": 1.1202,
+      "step": 2150
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.34910774230957,
+      "learning_rate": 2.4462845374978547e-05,
+      "loss": 1.6053,
+      "step": 2151
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.412336349487305,
+      "learning_rate": 2.4460271151535954e-05,
+      "loss": 1.3831,
+      "step": 2152
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.091730117797852,
+      "learning_rate": 2.445769692809336e-05,
+      "loss": 1.0023,
+      "step": 2153
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.222784042358398,
+      "learning_rate": 2.4455122704650764e-05,
+      "loss": 1.1603,
+      "step": 2154
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.691291809082031,
+      "learning_rate": 2.445254848120817e-05,
+      "loss": 1.3121,
+      "step": 2155
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.616677284240723,
+      "learning_rate": 2.4449974257765574e-05,
+      "loss": 1.2645,
+      "step": 2156
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.994897842407227,
+      "learning_rate": 2.444740003432298e-05,
+      "loss": 1.1881,
+      "step": 2157
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.053680419921875,
+      "learning_rate": 2.4444825810880384e-05,
+      "loss": 1.1504,
+      "step": 2158
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.763884544372559,
+      "learning_rate": 2.444225158743779e-05,
+      "loss": 1.1584,
+      "step": 2159
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.906607627868652,
+      "learning_rate": 2.4439677363995194e-05,
+      "loss": 1.4657,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.504964351654053,
+      "learning_rate": 2.44371031405526e-05,
+      "loss": 0.8196,
+      "step": 2161
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.369091510772705,
+      "learning_rate": 2.4434528917110007e-05,
+      "loss": 0.7597,
+      "step": 2162
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.04252815246582,
+      "learning_rate": 2.4431954693667414e-05,
+      "loss": 1.0252,
+      "step": 2163
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.198497772216797,
+      "learning_rate": 2.4429380470224817e-05,
+      "loss": 1.0809,
+      "step": 2164
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.258764266967773,
+      "learning_rate": 2.442680624678222e-05,
+      "loss": 0.9795,
+      "step": 2165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.062975883483887,
+      "learning_rate": 2.4424232023339627e-05,
+      "loss": 1.0093,
+      "step": 2166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.119672775268555,
+      "learning_rate": 2.442165779989703e-05,
+      "loss": 0.8311,
+      "step": 2167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.630653381347656,
+      "learning_rate": 2.4419083576454437e-05,
+      "loss": 1.0723,
+      "step": 2168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.558615684509277,
+      "learning_rate": 2.441650935301184e-05,
+      "loss": 1.5513,
+      "step": 2169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.386838912963867,
+      "learning_rate": 2.4413935129569247e-05,
+      "loss": 1.3053,
+      "step": 2170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.350480079650879,
+      "learning_rate": 2.4411360906126653e-05,
+      "loss": 1.13,
+      "step": 2171
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.781171798706055,
+      "learning_rate": 2.440878668268406e-05,
+      "loss": 0.9033,
+      "step": 2172
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.367735862731934,
+      "learning_rate": 2.4406212459241463e-05,
+      "loss": 1.1802,
+      "step": 2173
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.530876636505127,
+      "learning_rate": 2.440363823579887e-05,
+      "loss": 0.9118,
+      "step": 2174
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.17416000366211,
+      "learning_rate": 2.4401064012356273e-05,
+      "loss": 1.0697,
+      "step": 2175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.217011451721191,
+      "learning_rate": 2.4398489788913677e-05,
+      "loss": 1.2457,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.692070007324219,
+      "learning_rate": 2.4395915565471083e-05,
+      "loss": 1.2613,
+      "step": 2177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.750065803527832,
+      "learning_rate": 2.4393341342028487e-05,
+      "loss": 1.2005,
+      "step": 2178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.901527404785156,
+      "learning_rate": 2.4390767118585893e-05,
+      "loss": 1.3635,
+      "step": 2179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.055781364440918,
+      "learning_rate": 2.4388192895143297e-05,
+      "loss": 1.1935,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.499872207641602,
+      "learning_rate": 2.4385618671700707e-05,
+      "loss": 0.9131,
+      "step": 2181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.323253631591797,
+      "learning_rate": 2.438304444825811e-05,
+      "loss": 1.2284,
+      "step": 2182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.72081470489502,
+      "learning_rate": 2.4380470224815517e-05,
+      "loss": 1.2344,
+      "step": 2183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.034542083740234,
+      "learning_rate": 2.437789600137292e-05,
+      "loss": 1.2939,
+      "step": 2184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.844457626342773,
+      "learning_rate": 2.4375321777930327e-05,
+      "loss": 0.8724,
+      "step": 2185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.399674415588379,
+      "learning_rate": 2.437274755448773e-05,
+      "loss": 1.0371,
+      "step": 2186
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.00466537475586,
+      "learning_rate": 2.4370173331045133e-05,
+      "loss": 1.0131,
+      "step": 2187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.96268081665039,
+      "learning_rate": 2.436759910760254e-05,
+      "loss": 1.0904,
+      "step": 2188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.096373558044434,
+      "learning_rate": 2.4365024884159943e-05,
+      "loss": 1.0727,
+      "step": 2189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.763750076293945,
+      "learning_rate": 2.4362450660717353e-05,
+      "loss": 1.1843,
+      "step": 2190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.026836395263672,
+      "learning_rate": 2.4359876437274757e-05,
+      "loss": 1.0918,
+      "step": 2191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.820173263549805,
+      "learning_rate": 2.4357302213832163e-05,
+      "loss": 1.0206,
+      "step": 2192
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.105895042419434,
+      "learning_rate": 2.4354727990389567e-05,
+      "loss": 1.1331,
+      "step": 2193
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.734973907470703,
+      "learning_rate": 2.4352153766946973e-05,
+      "loss": 1.0124,
+      "step": 2194
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.105067253112793,
+      "learning_rate": 2.4349579543504377e-05,
+      "loss": 1.0347,
+      "step": 2195
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.364143371582031,
+      "learning_rate": 2.434700532006178e-05,
+      "loss": 1.2429,
+      "step": 2196
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.511385917663574,
+      "learning_rate": 2.4344431096619187e-05,
+      "loss": 1.1924,
+      "step": 2197
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.025053024291992,
+      "learning_rate": 2.434185687317659e-05,
+      "loss": 1.1453,
+      "step": 2198
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.77366828918457,
+      "learning_rate": 2.4339282649733997e-05,
+      "loss": 1.1927,
+      "step": 2199
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.097921371459961,
+      "learning_rate": 2.4336708426291403e-05,
+      "loss": 1.1981,
+      "step": 2200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.000970840454102,
+      "learning_rate": 2.433413420284881e-05,
+      "loss": 1.0992,
+      "step": 2201
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.089954376220703,
+      "learning_rate": 2.4331559979406213e-05,
+      "loss": 1.192,
+      "step": 2202
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.446324348449707,
+      "learning_rate": 2.432898575596362e-05,
+      "loss": 1.102,
+      "step": 2203
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.706451416015625,
+      "learning_rate": 2.4326411532521023e-05,
+      "loss": 1.1809,
+      "step": 2204
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.350419998168945,
+      "learning_rate": 2.432383730907843e-05,
+      "loss": 1.292,
+      "step": 2205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.564776420593262,
+      "learning_rate": 2.4321263085635833e-05,
+      "loss": 0.9112,
+      "step": 2206
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.59047794342041,
+      "learning_rate": 2.4318688862193236e-05,
+      "loss": 0.85,
+      "step": 2207
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.712336540222168,
+      "learning_rate": 2.4316114638750643e-05,
+      "loss": 1.1457,
+      "step": 2208
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.856526374816895,
+      "learning_rate": 2.431354041530805e-05,
+      "loss": 1.2405,
+      "step": 2209
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.182084083557129,
+      "learning_rate": 2.4310966191865456e-05,
+      "loss": 0.9503,
+      "step": 2210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.732983589172363,
+      "learning_rate": 2.430839196842286e-05,
+      "loss": 1.1711,
+      "step": 2211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.850709915161133,
+      "learning_rate": 2.4305817744980266e-05,
+      "loss": 1.2098,
+      "step": 2212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.524785995483398,
+      "learning_rate": 2.430324352153767e-05,
+      "loss": 0.9561,
+      "step": 2213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.560460090637207,
+      "learning_rate": 2.4300669298095076e-05,
+      "loss": 1.1777,
+      "step": 2214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.62714958190918,
+      "learning_rate": 2.429809507465248e-05,
+      "loss": 1.1986,
+      "step": 2215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.657073020935059,
+      "learning_rate": 2.4295520851209886e-05,
+      "loss": 1.5808,
+      "step": 2216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.352283477783203,
+      "learning_rate": 2.429294662776729e-05,
+      "loss": 1.2082,
+      "step": 2217
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.29774284362793,
+      "learning_rate": 2.4290372404324693e-05,
+      "loss": 1.0112,
+      "step": 2218
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.176529884338379,
+      "learning_rate": 2.4287798180882103e-05,
+      "loss": 1.2053,
+      "step": 2219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.629195213317871,
+      "learning_rate": 2.4285223957439506e-05,
+      "loss": 1.4169,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.850541114807129,
+      "learning_rate": 2.4282649733996913e-05,
+      "loss": 1.3332,
+      "step": 2221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.906332015991211,
+      "learning_rate": 2.4280075510554316e-05,
+      "loss": 0.9488,
+      "step": 2222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.36665153503418,
+      "learning_rate": 2.4277501287111723e-05,
+      "loss": 0.9325,
+      "step": 2223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.38780689239502,
+      "learning_rate": 2.4274927063669126e-05,
+      "loss": 1.0467,
+      "step": 2224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.8286662101745605,
+      "learning_rate": 2.4272352840226533e-05,
+      "loss": 0.9573,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.623621940612793,
+      "learning_rate": 2.4269778616783936e-05,
+      "loss": 0.9291,
+      "step": 2226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.773752212524414,
+      "learning_rate": 2.426720439334134e-05,
+      "loss": 1.127,
+      "step": 2227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.005531311035156,
+      "learning_rate": 2.426463016989875e-05,
+      "loss": 1.2232,
+      "step": 2228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.158402442932129,
+      "learning_rate": 2.4262055946456153e-05,
+      "loss": 1.0502,
+      "step": 2229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.19765853881836,
+      "learning_rate": 2.425948172301356e-05,
+      "loss": 0.9553,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.267332077026367,
+      "learning_rate": 2.4256907499570963e-05,
+      "loss": 0.8448,
+      "step": 2231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.809307098388672,
+      "learning_rate": 2.425433327612837e-05,
+      "loss": 0.9016,
+      "step": 2232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.3159818649292,
+      "learning_rate": 2.4251759052685773e-05,
+      "loss": 1.0381,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.401824951171875,
+      "learning_rate": 2.424918482924318e-05,
+      "loss": 1.15,
+      "step": 2234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.003650665283203,
+      "learning_rate": 2.4246610605800583e-05,
+      "loss": 0.8876,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.029595375061035,
+      "learning_rate": 2.424403638235799e-05,
+      "loss": 0.981,
+      "step": 2236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.966934204101562,
+      "learning_rate": 2.4241462158915393e-05,
+      "loss": 1.181,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.154629707336426,
+      "learning_rate": 2.42388879354728e-05,
+      "loss": 1.3361,
+      "step": 2238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.800402641296387,
+      "learning_rate": 2.4236313712030206e-05,
+      "loss": 1.1646,
+      "step": 2239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.770148277282715,
+      "learning_rate": 2.423373948858761e-05,
+      "loss": 1.1305,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.895112037658691,
+      "learning_rate": 2.4231165265145016e-05,
+      "loss": 1.1084,
+      "step": 2241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.0063605308532715,
+      "learning_rate": 2.422859104170242e-05,
+      "loss": 0.6602,
+      "step": 2242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.963147163391113,
+      "learning_rate": 2.4226016818259826e-05,
+      "loss": 0.9376,
+      "step": 2243
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.689924240112305,
+      "learning_rate": 2.422344259481723e-05,
+      "loss": 0.9693,
+      "step": 2244
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.47618579864502,
+      "learning_rate": 2.4220868371374636e-05,
+      "loss": 1.0734,
+      "step": 2245
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.367137908935547,
+      "learning_rate": 2.421829414793204e-05,
+      "loss": 1.1059,
+      "step": 2246
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.199468612670898,
+      "learning_rate": 2.421571992448945e-05,
+      "loss": 1.0812,
+      "step": 2247
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.501741409301758,
+      "learning_rate": 2.4213145701046853e-05,
+      "loss": 1.1653,
+      "step": 2248
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.0991792678833,
+      "learning_rate": 2.4210571477604256e-05,
+      "loss": 1.2308,
+      "step": 2249
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.76620864868164,
+      "learning_rate": 2.4207997254161663e-05,
+      "loss": 1.2191,
+      "step": 2250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.74355697631836,
+      "learning_rate": 2.4205423030719066e-05,
+      "loss": 1.1048,
+      "step": 2251
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.557761192321777,
+      "learning_rate": 2.4202848807276473e-05,
+      "loss": 1.1524,
+      "step": 2252
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.952515602111816,
+      "learning_rate": 2.4200274583833876e-05,
+      "loss": 1.2135,
+      "step": 2253
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.356393814086914,
+      "learning_rate": 2.4197700360391283e-05,
+      "loss": 0.8388,
+      "step": 2254
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.512103080749512,
+      "learning_rate": 2.4195126136948686e-05,
+      "loss": 0.9181,
+      "step": 2255
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.448094367980957,
+      "learning_rate": 2.4192551913506093e-05,
+      "loss": 0.994,
+      "step": 2256
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.41581916809082,
+      "learning_rate": 2.41899776900635e-05,
+      "loss": 1.4013,
+      "step": 2257
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.822907447814941,
+      "learning_rate": 2.4187403466620903e-05,
+      "loss": 0.9049,
+      "step": 2258
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.646187782287598,
+      "learning_rate": 2.418482924317831e-05,
+      "loss": 0.9902,
+      "step": 2259
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.040304183959961,
+      "learning_rate": 2.4182255019735713e-05,
+      "loss": 0.9898,
+      "step": 2260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.171103477478027,
+      "learning_rate": 2.417968079629312e-05,
+      "loss": 1.1363,
+      "step": 2261
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.7490553855896,
+      "learning_rate": 2.4177106572850523e-05,
+      "loss": 1.0245,
+      "step": 2262
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.144671440124512,
+      "learning_rate": 2.417453234940793e-05,
+      "loss": 1.1563,
+      "step": 2263
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.781315803527832,
+      "learning_rate": 2.4171958125965333e-05,
+      "loss": 1.0074,
+      "step": 2264
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.80992603302002,
+      "learning_rate": 2.416938390252274e-05,
+      "loss": 1.0563,
+      "step": 2265
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.283391952514648,
+      "learning_rate": 2.4166809679080146e-05,
+      "loss": 1.0983,
+      "step": 2266
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.335240364074707,
+      "learning_rate": 2.4164235455637553e-05,
+      "loss": 0.9964,
+      "step": 2267
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.979182243347168,
+      "learning_rate": 2.4161661232194956e-05,
+      "loss": 1.0521,
+      "step": 2268
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.834734916687012,
+      "learning_rate": 2.415908700875236e-05,
+      "loss": 1.3008,
+      "step": 2269
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.852832794189453,
+      "learning_rate": 2.4156512785309766e-05,
+      "loss": 1.1512,
+      "step": 2270
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.862049102783203,
+      "learning_rate": 2.415393856186717e-05,
+      "loss": 1.3111,
+      "step": 2271
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.671493530273438,
+      "learning_rate": 2.4151364338424576e-05,
+      "loss": 0.907,
+      "step": 2272
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.944957733154297,
+      "learning_rate": 2.414879011498198e-05,
+      "loss": 1.2896,
+      "step": 2273
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.786681175231934,
+      "learning_rate": 2.4146215891539386e-05,
+      "loss": 1.1686,
+      "step": 2274
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.018754005432129,
+      "learning_rate": 2.4143641668096792e-05,
+      "loss": 1.1924,
+      "step": 2275
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.193449974060059,
+      "learning_rate": 2.41410674446542e-05,
+      "loss": 1.1071,
+      "step": 2276
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.369192123413086,
+      "learning_rate": 2.4138493221211602e-05,
+      "loss": 1.2097,
+      "step": 2277
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.837075233459473,
+      "learning_rate": 2.413591899776901e-05,
+      "loss": 1.1033,
+      "step": 2278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.052380561828613,
+      "learning_rate": 2.4133344774326412e-05,
+      "loss": 1.0338,
+      "step": 2279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.481216430664062,
+      "learning_rate": 2.4130770550883816e-05,
+      "loss": 1.1987,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.217461585998535,
+      "learning_rate": 2.4128196327441222e-05,
+      "loss": 1.4332,
+      "step": 2281
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.972358703613281,
+      "learning_rate": 2.4125622103998626e-05,
+      "loss": 0.7925,
+      "step": 2282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.535296440124512,
+      "learning_rate": 2.4123047880556032e-05,
+      "loss": 1.5184,
+      "step": 2283
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.570453643798828,
+      "learning_rate": 2.4120473657113436e-05,
+      "loss": 1.0196,
+      "step": 2284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.724305152893066,
+      "learning_rate": 2.4117899433670846e-05,
+      "loss": 1.0131,
+      "step": 2285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.26595687866211,
+      "learning_rate": 2.411532521022825e-05,
+      "loss": 1.1198,
+      "step": 2286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.108465194702148,
+      "learning_rate": 2.4112750986785656e-05,
+      "loss": 0.8719,
+      "step": 2287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.542659759521484,
+      "learning_rate": 2.411017676334306e-05,
+      "loss": 1.0413,
+      "step": 2288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.4580659866333,
+      "learning_rate": 2.4107602539900466e-05,
+      "loss": 1.186,
+      "step": 2289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.495659828186035,
+      "learning_rate": 2.410502831645787e-05,
+      "loss": 0.9637,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.1148042678833,
+      "learning_rate": 2.4102454093015272e-05,
+      "loss": 1.3322,
+      "step": 2291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.231887817382812,
+      "learning_rate": 2.409987986957268e-05,
+      "loss": 1.1449,
+      "step": 2292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.8561177253723145,
+      "learning_rate": 2.4097305646130082e-05,
+      "loss": 1.1552,
+      "step": 2293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.66640853881836,
+      "learning_rate": 2.4094731422687492e-05,
+      "loss": 1.3027,
+      "step": 2294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.01046371459961,
+      "learning_rate": 2.4092157199244896e-05,
+      "loss": 1.2658,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.291804313659668,
+      "learning_rate": 2.4089582975802302e-05,
+      "loss": 1.3789,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.136520385742188,
+      "learning_rate": 2.4087008752359706e-05,
+      "loss": 1.2063,
+      "step": 2297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.098982810974121,
+      "learning_rate": 2.4084434528917112e-05,
+      "loss": 1.0684,
+      "step": 2298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.366686820983887,
+      "learning_rate": 2.4081860305474516e-05,
+      "loss": 0.8837,
+      "step": 2299
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.365204811096191,
+      "learning_rate": 2.407928608203192e-05,
+      "loss": 1.1671,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.287782669067383,
+      "learning_rate": 2.4076711858589326e-05,
+      "loss": 0.9609,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.816483497619629,
+      "learning_rate": 2.407413763514673e-05,
+      "loss": 1.068,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.215349197387695,
+      "learning_rate": 2.4071563411704136e-05,
+      "loss": 0.9013,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.46806812286377,
+      "learning_rate": 2.4068989188261542e-05,
+      "loss": 1.2445,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.204339981079102,
+      "learning_rate": 2.406641496481895e-05,
+      "loss": 1.0809,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.090676307678223,
+      "learning_rate": 2.4063840741376352e-05,
+      "loss": 1.0355,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.966399192810059,
+      "learning_rate": 2.406126651793376e-05,
+      "loss": 1.3129,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.648179054260254,
+      "learning_rate": 2.4058692294491162e-05,
+      "loss": 1.1391,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.937187194824219,
+      "learning_rate": 2.405611807104857e-05,
+      "loss": 1.1248,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.938966751098633,
+      "learning_rate": 2.4053543847605972e-05,
+      "loss": 0.9587,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.952705383300781,
+      "learning_rate": 2.4050969624163375e-05,
+      "loss": 1.0574,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.699490547180176,
+      "learning_rate": 2.4048395400720782e-05,
+      "loss": 1.1071,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.558082580566406,
+      "learning_rate": 2.404582117727819e-05,
+      "loss": 1.023,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.829763412475586,
+      "learning_rate": 2.4043246953835595e-05,
+      "loss": 1.1763,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.652128219604492,
+      "learning_rate": 2.4040672730393e-05,
+      "loss": 0.9157,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.757291793823242,
+      "learning_rate": 2.4038098506950405e-05,
+      "loss": 1.2526,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.581213474273682,
+      "learning_rate": 2.403552428350781e-05,
+      "loss": 0.9489,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.630509376525879,
+      "learning_rate": 2.4032950060065215e-05,
+      "loss": 0.9891,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.893701553344727,
+      "learning_rate": 2.403037583662262e-05,
+      "loss": 0.8387,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.503652572631836,
+      "learning_rate": 2.4027801613180025e-05,
+      "loss": 1.1652,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.663091659545898,
+      "learning_rate": 2.402522738973743e-05,
+      "loss": 1.0581,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.012900352478027,
+      "learning_rate": 2.4022653166294832e-05,
+      "loss": 1.1648,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.778726577758789,
+      "learning_rate": 2.4020078942852242e-05,
+      "loss": 1.1431,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.202214241027832,
+      "learning_rate": 2.4017504719409645e-05,
+      "loss": 1.0112,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.303224563598633,
+      "learning_rate": 2.4014930495967052e-05,
+      "loss": 1.2525,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.540024757385254,
+      "learning_rate": 2.4012356272524455e-05,
+      "loss": 1.4424,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.628973007202148,
+      "learning_rate": 2.4009782049081862e-05,
+      "loss": 0.8955,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.897580623626709,
+      "learning_rate": 2.4007207825639265e-05,
+      "loss": 1.3406,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.044696807861328,
+      "learning_rate": 2.4004633602196672e-05,
+      "loss": 1.0739,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.221452713012695,
+      "learning_rate": 2.4002059378754075e-05,
+      "loss": 1.1278,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.01343822479248,
+      "learning_rate": 2.399948515531148e-05,
+      "loss": 1.2545,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.265592575073242,
+      "learning_rate": 2.399691093186889e-05,
+      "loss": 1.0236,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.603173732757568,
+      "learning_rate": 2.3994336708426292e-05,
+      "loss": 0.8957,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.266589164733887,
+      "learning_rate": 2.39917624849837e-05,
+      "loss": 1.4813,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.32174015045166,
+      "learning_rate": 2.3989188261541102e-05,
+      "loss": 1.3589,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.692406177520752,
+      "learning_rate": 2.398661403809851e-05,
+      "loss": 1.0378,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.303297996520996,
+      "learning_rate": 2.3984039814655912e-05,
+      "loss": 1.4373,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.728151321411133,
+      "learning_rate": 2.398146559121332e-05,
+      "loss": 1.121,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.557869911193848,
+      "learning_rate": 2.3978891367770722e-05,
+      "loss": 1.101,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.885689735412598,
+      "learning_rate": 2.397631714432813e-05,
+      "loss": 0.8533,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.080730438232422,
+      "learning_rate": 2.3973742920885532e-05,
+      "loss": 1.2499,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.192622184753418,
+      "learning_rate": 2.397116869744294e-05,
+      "loss": 0.8441,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.07056999206543,
+      "learning_rate": 2.3968594474000345e-05,
+      "loss": 1.0561,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.36457347869873,
+      "learning_rate": 2.396602025055775e-05,
+      "loss": 1.0302,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.434206008911133,
+      "learning_rate": 2.3963446027115155e-05,
+      "loss": 1.065,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.678196907043457,
+      "learning_rate": 2.396087180367256e-05,
+      "loss": 1.1172,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.543830871582031,
+      "learning_rate": 2.3958297580229965e-05,
+      "loss": 0.9344,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.137594223022461,
+      "learning_rate": 2.395572335678737e-05,
+      "loss": 0.9821,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.554171562194824,
+      "learning_rate": 2.3953149133344775e-05,
+      "loss": 0.8165,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.486687660217285,
+      "learning_rate": 2.395057490990218e-05,
+      "loss": 1.0482,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.66871166229248,
+      "learning_rate": 2.394800068645959e-05,
+      "loss": 0.936,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.195672035217285,
+      "learning_rate": 2.3945426463016992e-05,
+      "loss": 1.0888,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.308737754821777,
+      "learning_rate": 2.3942852239574395e-05,
+      "loss": 1.037,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.782885551452637,
+      "learning_rate": 2.3940278016131802e-05,
+      "loss": 1.0355,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.690522193908691,
+      "learning_rate": 2.3937703792689205e-05,
+      "loss": 1.1222,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.021770477294922,
+      "learning_rate": 2.3935129569246612e-05,
+      "loss": 1.0931,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.61075210571289,
+      "learning_rate": 2.3932555345804015e-05,
+      "loss": 0.9862,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.538852214813232,
+      "learning_rate": 2.392998112236142e-05,
+      "loss": 0.8301,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.294318199157715,
+      "learning_rate": 2.3927406898918825e-05,
+      "loss": 1.1885,
+      "step": 2359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.791847229003906,
+      "learning_rate": 2.392483267547623e-05,
+      "loss": 0.931,
+      "step": 2360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.116369247436523,
+      "learning_rate": 2.392225845203364e-05,
+      "loss": 0.9284,
+      "step": 2361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.191892623901367,
+      "learning_rate": 2.391968422859104e-05,
+      "loss": 0.647,
+      "step": 2362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.659319877624512,
+      "learning_rate": 2.391711000514845e-05,
+      "loss": 1.1507,
+      "step": 2363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.853606224060059,
+      "learning_rate": 2.391453578170585e-05,
+      "loss": 0.8916,
+      "step": 2364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.060551643371582,
+      "learning_rate": 2.3911961558263258e-05,
+      "loss": 0.9975,
+      "step": 2365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.454354286193848,
+      "learning_rate": 2.390938733482066e-05,
+      "loss": 1.3043,
+      "step": 2366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.785755157470703,
+      "learning_rate": 2.3906813111378068e-05,
+      "loss": 1.52,
+      "step": 2367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.606592178344727,
+      "learning_rate": 2.390423888793547e-05,
+      "loss": 0.897,
+      "step": 2368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.56293773651123,
+      "learning_rate": 2.3901664664492878e-05,
+      "loss": 1.0566,
+      "step": 2369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.36727523803711,
+      "learning_rate": 2.3899090441050285e-05,
+      "loss": 1.3117,
+      "step": 2370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.051411628723145,
+      "learning_rate": 2.389651621760769e-05,
+      "loss": 1.1332,
+      "step": 2371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.34546947479248,
+      "learning_rate": 2.3893941994165095e-05,
+      "loss": 1.2682,
+      "step": 2372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.438085556030273,
+      "learning_rate": 2.3891367770722498e-05,
+      "loss": 0.9392,
+      "step": 2373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.597270011901855,
+      "learning_rate": 2.3888793547279905e-05,
+      "loss": 1.0658,
+      "step": 2374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.972052574157715,
+      "learning_rate": 2.3886219323837308e-05,
+      "loss": 1.2095,
+      "step": 2375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.842507362365723,
+      "learning_rate": 2.3883645100394715e-05,
+      "loss": 1.201,
+      "step": 2376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.305310249328613,
+      "learning_rate": 2.3881070876952118e-05,
+      "loss": 1.0585,
+      "step": 2377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.667527198791504,
+      "learning_rate": 2.3878496653509525e-05,
+      "loss": 1.1337,
+      "step": 2378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.113306045532227,
+      "learning_rate": 2.3875922430066928e-05,
+      "loss": 1.1791,
+      "step": 2379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.53714370727539,
+      "learning_rate": 2.3873348206624338e-05,
+      "loss": 1.1967,
+      "step": 2380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.78366470336914,
+      "learning_rate": 2.387077398318174e-05,
+      "loss": 1.4728,
+      "step": 2381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.642723083496094,
+      "learning_rate": 2.3868199759739148e-05,
+      "loss": 1.2941,
+      "step": 2382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.167917251586914,
+      "learning_rate": 2.386562553629655e-05,
+      "loss": 1.1403,
+      "step": 2383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.521174430847168,
+      "learning_rate": 2.3863051312853955e-05,
+      "loss": 1.1917,
+      "step": 2384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.093134880065918,
+      "learning_rate": 2.386047708941136e-05,
+      "loss": 1.3338,
+      "step": 2385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.815524101257324,
+      "learning_rate": 2.3857902865968765e-05,
+      "loss": 1.1557,
+      "step": 2386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.5319743156433105,
+      "learning_rate": 2.385532864252617e-05,
+      "loss": 0.9474,
+      "step": 2387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.921976089477539,
+      "learning_rate": 2.3852754419083575e-05,
+      "loss": 0.8523,
+      "step": 2388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.433771133422852,
+      "learning_rate": 2.3850180195640985e-05,
+      "loss": 1.0348,
+      "step": 2389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.832609176635742,
+      "learning_rate": 2.3847605972198388e-05,
+      "loss": 0.9005,
+      "step": 2390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.235001564025879,
+      "learning_rate": 2.3845031748755795e-05,
+      "loss": 1.3815,
+      "step": 2391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.518238067626953,
+      "learning_rate": 2.3842457525313198e-05,
+      "loss": 0.8948,
+      "step": 2392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.102937698364258,
+      "learning_rate": 2.3839883301870605e-05,
+      "loss": 1.2714,
+      "step": 2393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.932148933410645,
+      "learning_rate": 2.3837309078428008e-05,
+      "loss": 1.3764,
+      "step": 2394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.55225658416748,
+      "learning_rate": 2.383473485498541e-05,
+      "loss": 1.0773,
+      "step": 2395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.794912338256836,
+      "learning_rate": 2.3832160631542818e-05,
+      "loss": 1.009,
+      "step": 2396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.654611587524414,
+      "learning_rate": 2.382958640810022e-05,
+      "loss": 0.9229,
+      "step": 2397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.335205078125,
+      "learning_rate": 2.382701218465763e-05,
+      "loss": 1.0753,
+      "step": 2398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.122414588928223,
+      "learning_rate": 2.3824437961215035e-05,
+      "loss": 1.0487,
+      "step": 2399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.104586601257324,
+      "learning_rate": 2.382186373777244e-05,
+      "loss": 1.1819,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.469476699829102,
+      "learning_rate": 2.3819289514329845e-05,
+      "loss": 1.165,
+      "step": 2401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.234419822692871,
+      "learning_rate": 2.381671529088725e-05,
+      "loss": 1.0652,
+      "step": 2402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.224913597106934,
+      "learning_rate": 2.3814141067444655e-05,
+      "loss": 1.156,
+      "step": 2403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.427159309387207,
+      "learning_rate": 2.3811566844002058e-05,
+      "loss": 1.1087,
+      "step": 2404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.388914108276367,
+      "learning_rate": 2.3808992620559465e-05,
+      "loss": 0.994,
+      "step": 2405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.564704895019531,
+      "learning_rate": 2.3806418397116868e-05,
+      "loss": 1.1732,
+      "step": 2406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.740270614624023,
+      "learning_rate": 2.3803844173674275e-05,
+      "loss": 1.0191,
+      "step": 2407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.566525459289551,
+      "learning_rate": 2.380126995023168e-05,
+      "loss": 0.9126,
+      "step": 2408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.765815258026123,
+      "learning_rate": 2.3798695726789088e-05,
+      "loss": 1.1143,
+      "step": 2409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.17686653137207,
+      "learning_rate": 2.379612150334649e-05,
+      "loss": 0.9496,
+      "step": 2410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.54204273223877,
+      "learning_rate": 2.3793547279903898e-05,
+      "loss": 1.2374,
+      "step": 2411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.409787654876709,
+      "learning_rate": 2.37909730564613e-05,
+      "loss": 1.0006,
+      "step": 2412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.912604331970215,
+      "learning_rate": 2.3788398833018708e-05,
+      "loss": 1.0432,
+      "step": 2413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.363262176513672,
+      "learning_rate": 2.378582460957611e-05,
+      "loss": 1.0378,
+      "step": 2414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.565895080566406,
+      "learning_rate": 2.3783250386133514e-05,
+      "loss": 1.1047,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.706425666809082,
+      "learning_rate": 2.378067616269092e-05,
+      "loss": 1.3295,
+      "step": 2416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.505475997924805,
+      "learning_rate": 2.3778101939248328e-05,
+      "loss": 1.1323,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.23418140411377,
+      "learning_rate": 2.3775527715805734e-05,
+      "loss": 1.0101,
+      "step": 2418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.298528671264648,
+      "learning_rate": 2.3772953492363138e-05,
+      "loss": 1.049,
+      "step": 2419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.873881340026855,
+      "learning_rate": 2.3770379268920544e-05,
+      "loss": 0.8916,
+      "step": 2420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.962735652923584,
+      "learning_rate": 2.3767805045477948e-05,
+      "loss": 0.8387,
+      "step": 2421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.635884284973145,
+      "learning_rate": 2.3765230822035354e-05,
+      "loss": 1.1091,
+      "step": 2422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.053528785705566,
+      "learning_rate": 2.3762656598592758e-05,
+      "loss": 1.2137,
+      "step": 2423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.316943168640137,
+      "learning_rate": 2.3760082375150164e-05,
+      "loss": 1.2033,
+      "step": 2424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.887518882751465,
+      "learning_rate": 2.3757508151707568e-05,
+      "loss": 1.2407,
+      "step": 2425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.951789855957031,
+      "learning_rate": 2.375493392826497e-05,
+      "loss": 1.1114,
+      "step": 2426
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.072427749633789,
+      "learning_rate": 2.375235970482238e-05,
+      "loss": 0.9998,
+      "step": 2427
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.388823509216309,
+      "learning_rate": 2.3749785481379784e-05,
+      "loss": 0.9954,
+      "step": 2428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.042737007141113,
+      "learning_rate": 2.374721125793719e-05,
+      "loss": 0.9225,
+      "step": 2429
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.438971042633057,
+      "learning_rate": 2.3744637034494594e-05,
+      "loss": 0.7363,
+      "step": 2430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.267683029174805,
+      "learning_rate": 2.3742062811052e-05,
+      "loss": 0.9942,
+      "step": 2431
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.633086204528809,
+      "learning_rate": 2.3739488587609404e-05,
+      "loss": 0.8846,
+      "step": 2432
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.981773376464844,
+      "learning_rate": 2.373691436416681e-05,
+      "loss": 1.3067,
+      "step": 2433
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.61812973022461,
+      "learning_rate": 2.3734340140724214e-05,
+      "loss": 1.0108,
+      "step": 2434
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.247781753540039,
+      "learning_rate": 2.373176591728162e-05,
+      "loss": 1.3639,
+      "step": 2435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.135536193847656,
+      "learning_rate": 2.3729191693839028e-05,
+      "loss": 0.9578,
+      "step": 2436
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.909148216247559,
+      "learning_rate": 2.372661747039643e-05,
+      "loss": 1.023,
+      "step": 2437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.051506996154785,
+      "learning_rate": 2.3724043246953838e-05,
+      "loss": 1.3859,
+      "step": 2438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.185151100158691,
+      "learning_rate": 2.372146902351124e-05,
+      "loss": 1.1907,
+      "step": 2439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.996024131774902,
+      "learning_rate": 2.3718894800068648e-05,
+      "loss": 1.3278,
+      "step": 2440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.550662994384766,
+      "learning_rate": 2.371632057662605e-05,
+      "loss": 1.0441,
+      "step": 2441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.213271141052246,
+      "learning_rate": 2.3713746353183458e-05,
+      "loss": 0.8978,
+      "step": 2442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.587554931640625,
+      "learning_rate": 2.371117212974086e-05,
+      "loss": 1.001,
+      "step": 2443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.72619915008545,
+      "learning_rate": 2.3708597906298268e-05,
+      "loss": 1.0332,
+      "step": 2444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.2136812210083,
+      "learning_rate": 2.370602368285567e-05,
+      "loss": 0.957,
+      "step": 2445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.504052639007568,
+      "learning_rate": 2.3703449459413078e-05,
+      "loss": 0.7831,
+      "step": 2446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.622686386108398,
+      "learning_rate": 2.3700875235970484e-05,
+      "loss": 1.1063,
+      "step": 2447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.420857429504395,
+      "learning_rate": 2.3698301012527887e-05,
+      "loss": 1.5009,
+      "step": 2448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.327728271484375,
+      "learning_rate": 2.3695726789085294e-05,
+      "loss": 1.0195,
+      "step": 2449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.8270902633667,
+      "learning_rate": 2.3693152565642697e-05,
+      "loss": 1.2884,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.860349655151367,
+      "learning_rate": 2.3690578342200104e-05,
+      "loss": 1.3965,
+      "step": 2451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.761497497558594,
+      "learning_rate": 2.3688004118757507e-05,
+      "loss": 0.9062,
+      "step": 2452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.600850105285645,
+      "learning_rate": 2.3685429895314914e-05,
+      "loss": 1.1713,
+      "step": 2453
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.563528060913086,
+      "learning_rate": 2.3682855671872317e-05,
+      "loss": 1.2687,
+      "step": 2454
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.841002464294434,
+      "learning_rate": 2.3680281448429727e-05,
+      "loss": 0.7598,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.881284713745117,
+      "learning_rate": 2.367770722498713e-05,
+      "loss": 0.9367,
+      "step": 2456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.2709784507751465,
+      "learning_rate": 2.3675133001544534e-05,
+      "loss": 0.8366,
+      "step": 2457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.911156177520752,
+      "learning_rate": 2.367255877810194e-05,
+      "loss": 0.7566,
+      "step": 2458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.620280265808105,
+      "learning_rate": 2.3669984554659344e-05,
+      "loss": 1.492,
+      "step": 2459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.352581024169922,
+      "learning_rate": 2.366741033121675e-05,
+      "loss": 0.9107,
+      "step": 2460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.401845932006836,
+      "learning_rate": 2.3664836107774154e-05,
+      "loss": 1.0071,
+      "step": 2461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.157147407531738,
+      "learning_rate": 2.366226188433156e-05,
+      "loss": 1.1496,
+      "step": 2462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.345260620117188,
+      "learning_rate": 2.3659687660888964e-05,
+      "loss": 0.8295,
+      "step": 2463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.7211527824401855,
+      "learning_rate": 2.365711343744637e-05,
+      "loss": 0.9211,
+      "step": 2464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.277822494506836,
+      "learning_rate": 2.3654539214003777e-05,
+      "loss": 1.1629,
+      "step": 2465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.909029960632324,
+      "learning_rate": 2.365196499056118e-05,
+      "loss": 1.3003,
+      "step": 2466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.170978546142578,
+      "learning_rate": 2.3649390767118587e-05,
+      "loss": 1.1131,
+      "step": 2467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.749275207519531,
+      "learning_rate": 2.364681654367599e-05,
+      "loss": 1.1058,
+      "step": 2468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.033509254455566,
+      "learning_rate": 2.3644242320233397e-05,
+      "loss": 0.9819,
+      "step": 2469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.735815048217773,
+      "learning_rate": 2.36416680967908e-05,
+      "loss": 1.0819,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.044149398803711,
+      "learning_rate": 2.3639093873348207e-05,
+      "loss": 1.0405,
+      "step": 2471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.202775001525879,
+      "learning_rate": 2.363651964990561e-05,
+      "loss": 1.0634,
+      "step": 2472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.201421737670898,
+      "learning_rate": 2.3633945426463017e-05,
+      "loss": 0.991,
+      "step": 2473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.920056343078613,
+      "learning_rate": 2.3631371203020424e-05,
+      "loss": 0.7986,
+      "step": 2474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.732706069946289,
+      "learning_rate": 2.362879697957783e-05,
+      "loss": 1.3758,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.223851203918457,
+      "learning_rate": 2.3626222756135234e-05,
+      "loss": 1.053,
+      "step": 2476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.358149528503418,
+      "learning_rate": 2.3623648532692637e-05,
+      "loss": 1.0756,
+      "step": 2477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.53659439086914,
+      "learning_rate": 2.3621074309250044e-05,
+      "loss": 0.8786,
+      "step": 2478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.722784042358398,
+      "learning_rate": 2.3618500085807447e-05,
+      "loss": 1.3966,
+      "step": 2479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.544040679931641,
+      "learning_rate": 2.3615925862364854e-05,
+      "loss": 0.9188,
+      "step": 2480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.28165864944458,
+      "learning_rate": 2.3613351638922257e-05,
+      "loss": 0.7604,
+      "step": 2481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.316466331481934,
+      "learning_rate": 2.3610777415479664e-05,
+      "loss": 0.9873,
+      "step": 2482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.219130516052246,
+      "learning_rate": 2.3608203192037067e-05,
+      "loss": 1.2108,
+      "step": 2483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.904192924499512,
+      "learning_rate": 2.3605628968594477e-05,
+      "loss": 1.0583,
+      "step": 2484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.60234546661377,
+      "learning_rate": 2.360305474515188e-05,
+      "loss": 1.2682,
+      "step": 2485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.382621765136719,
+      "learning_rate": 2.3600480521709287e-05,
+      "loss": 1.0676,
+      "step": 2486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.308059692382812,
+      "learning_rate": 2.359790629826669e-05,
+      "loss": 0.9831,
+      "step": 2487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.96368408203125,
+      "learning_rate": 2.3595332074824094e-05,
+      "loss": 0.9276,
+      "step": 2488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.224778175354004,
+      "learning_rate": 2.35927578513815e-05,
+      "loss": 1.1833,
+      "step": 2489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.238512992858887,
+      "learning_rate": 2.3590183627938904e-05,
+      "loss": 1.0521,
+      "step": 2490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.882194519042969,
+      "learning_rate": 2.358760940449631e-05,
+      "loss": 0.9314,
+      "step": 2491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.799914360046387,
+      "learning_rate": 2.3585035181053714e-05,
+      "loss": 1.2231,
+      "step": 2492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.082046508789062,
+      "learning_rate": 2.3582460957611124e-05,
+      "loss": 1.2219,
+      "step": 2493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.452569007873535,
+      "learning_rate": 2.3579886734168527e-05,
+      "loss": 0.8429,
+      "step": 2494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.005644798278809,
+      "learning_rate": 2.3577312510725934e-05,
+      "loss": 1.3801,
+      "step": 2495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.329168319702148,
+      "learning_rate": 2.3574738287283337e-05,
+      "loss": 1.0627,
+      "step": 2496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.755752563476562,
+      "learning_rate": 2.3572164063840744e-05,
+      "loss": 1.2499,
+      "step": 2497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.486258506774902,
+      "learning_rate": 2.3569589840398147e-05,
+      "loss": 1.0468,
+      "step": 2498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.862729549407959,
+      "learning_rate": 2.356701561695555e-05,
+      "loss": 0.9382,
+      "step": 2499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.966972351074219,
+      "learning_rate": 2.3564441393512957e-05,
+      "loss": 1.1468,
+      "step": 2500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.419432640075684,
+      "learning_rate": 2.356186717007036e-05,
+      "loss": 1.0783,
+      "step": 2501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.90982723236084,
+      "learning_rate": 2.3559292946627767e-05,
+      "loss": 1.0103,
+      "step": 2502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.468710899353027,
+      "learning_rate": 2.3556718723185174e-05,
+      "loss": 1.0959,
+      "step": 2503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.415430068969727,
+      "learning_rate": 2.355414449974258e-05,
+      "loss": 1.0021,
+      "step": 2504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.645242691040039,
+      "learning_rate": 2.3551570276299984e-05,
+      "loss": 0.9629,
+      "step": 2505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.518164157867432,
+      "learning_rate": 2.354899605285739e-05,
+      "loss": 0.7111,
+      "step": 2506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.395265579223633,
+      "learning_rate": 2.3546421829414794e-05,
+      "loss": 0.9499,
+      "step": 2507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.945225715637207,
+      "learning_rate": 2.3543847605972197e-05,
+      "loss": 0.8378,
+      "step": 2508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.525571823120117,
+      "learning_rate": 2.3541273382529604e-05,
+      "loss": 1.2957,
+      "step": 2509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.679350852966309,
+      "learning_rate": 2.3538699159087007e-05,
+      "loss": 1.0082,
+      "step": 2510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.790804862976074,
+      "learning_rate": 2.3536124935644414e-05,
+      "loss": 1.0925,
+      "step": 2511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.276101112365723,
+      "learning_rate": 2.353355071220182e-05,
+      "loss": 0.9825,
+      "step": 2512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.411559104919434,
+      "learning_rate": 2.3530976488759227e-05,
+      "loss": 0.8942,
+      "step": 2513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.060784339904785,
+      "learning_rate": 2.352840226531663e-05,
+      "loss": 1.1793,
+      "step": 2514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.921730041503906,
+      "learning_rate": 2.3525828041874037e-05,
+      "loss": 0.9212,
+      "step": 2515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.655002593994141,
+      "learning_rate": 2.352325381843144e-05,
+      "loss": 0.6529,
+      "step": 2516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.469806671142578,
+      "learning_rate": 2.3520679594988847e-05,
+      "loss": 1.0394,
+      "step": 2517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.596567153930664,
+      "learning_rate": 2.351810537154625e-05,
+      "loss": 1.147,
+      "step": 2518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.936881065368652,
+      "learning_rate": 2.3515531148103653e-05,
+      "loss": 1.1531,
+      "step": 2519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.614805221557617,
+      "learning_rate": 2.351295692466106e-05,
+      "loss": 1.0526,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.913834571838379,
+      "learning_rate": 2.3510382701218467e-05,
+      "loss": 1.2466,
+      "step": 2521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.67934513092041,
+      "learning_rate": 2.3507808477775873e-05,
+      "loss": 1.1262,
+      "step": 2522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.341279983520508,
+      "learning_rate": 2.3505234254333277e-05,
+      "loss": 0.9325,
+      "step": 2523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.471835136413574,
+      "learning_rate": 2.3502660030890683e-05,
+      "loss": 1.223,
+      "step": 2524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.702771186828613,
+      "learning_rate": 2.3500085807448087e-05,
+      "loss": 1.1022,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.640060424804688,
+      "learning_rate": 2.3497511584005493e-05,
+      "loss": 0.9489,
+      "step": 2526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.58908462524414,
+      "learning_rate": 2.3494937360562897e-05,
+      "loss": 1.3934,
+      "step": 2527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.46060848236084,
+      "learning_rate": 2.3492363137120303e-05,
+      "loss": 0.9935,
+      "step": 2528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.000808715820312,
+      "learning_rate": 2.3489788913677707e-05,
+      "loss": 0.8257,
+      "step": 2529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.350912094116211,
+      "learning_rate": 2.348721469023511e-05,
+      "loss": 1.2273,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.099184036254883,
+      "learning_rate": 2.348464046679252e-05,
+      "loss": 0.9647,
+      "step": 2531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.951770782470703,
+      "learning_rate": 2.3482066243349923e-05,
+      "loss": 1.0582,
+      "step": 2532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.116839408874512,
+      "learning_rate": 2.347949201990733e-05,
+      "loss": 1.1364,
+      "step": 2533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.103583335876465,
+      "learning_rate": 2.3476917796464733e-05,
+      "loss": 1.1859,
+      "step": 2534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.171732902526855,
+      "learning_rate": 2.347434357302214e-05,
+      "loss": 1.0032,
+      "step": 2535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.168376922607422,
+      "learning_rate": 2.3471769349579543e-05,
+      "loss": 0.9084,
+      "step": 2536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.724564552307129,
+      "learning_rate": 2.346919512613695e-05,
+      "loss": 0.8318,
+      "step": 2537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.293848037719727,
+      "learning_rate": 2.3466620902694353e-05,
+      "loss": 1.2379,
+      "step": 2538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.937259674072266,
+      "learning_rate": 2.346404667925176e-05,
+      "loss": 0.8578,
+      "step": 2539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.461071014404297,
+      "learning_rate": 2.3461472455809167e-05,
+      "loss": 0.9782,
+      "step": 2540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.833611488342285,
+      "learning_rate": 2.345889823236657e-05,
+      "loss": 1.1425,
+      "step": 2541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.43184757232666,
+      "learning_rate": 2.3456324008923977e-05,
+      "loss": 0.9369,
+      "step": 2542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.954643249511719,
+      "learning_rate": 2.345374978548138e-05,
+      "loss": 0.893,
+      "step": 2543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.736577033996582,
+      "learning_rate": 2.3451175562038787e-05,
+      "loss": 0.8636,
+      "step": 2544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.804381370544434,
+      "learning_rate": 2.344860133859619e-05,
+      "loss": 1.1972,
+      "step": 2545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.188714027404785,
+      "learning_rate": 2.3446027115153597e-05,
+      "loss": 0.6603,
+      "step": 2546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.234238624572754,
+      "learning_rate": 2.3443452891711e-05,
+      "loss": 1.5362,
+      "step": 2547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.846101760864258,
+      "learning_rate": 2.3440878668268407e-05,
+      "loss": 0.9585,
+      "step": 2548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.942909240722656,
+      "learning_rate": 2.343830444482581e-05,
+      "loss": 1.2047,
+      "step": 2549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.016656875610352,
+      "learning_rate": 2.3435730221383217e-05,
+      "loss": 1.1495,
+      "step": 2550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.28046703338623,
+      "learning_rate": 2.3433155997940623e-05,
+      "loss": 1.0153,
+      "step": 2551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.70443058013916,
+      "learning_rate": 2.3430581774498026e-05,
+      "loss": 1.107,
+      "step": 2552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.147012710571289,
+      "learning_rate": 2.3428007551055433e-05,
+      "loss": 0.8575,
+      "step": 2553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.779894828796387,
+      "learning_rate": 2.3425433327612836e-05,
+      "loss": 0.9142,
+      "step": 2554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.599183082580566,
+      "learning_rate": 2.3422859104170243e-05,
+      "loss": 0.7884,
+      "step": 2555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.4769868850708,
+      "learning_rate": 2.3420284880727646e-05,
+      "loss": 1.2565,
+      "step": 2556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.201024055480957,
+      "learning_rate": 2.3417710657285053e-05,
+      "loss": 1.0752,
+      "step": 2557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.123604774475098,
+      "learning_rate": 2.3415136433842456e-05,
+      "loss": 0.8538,
+      "step": 2558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.135371208190918,
+      "learning_rate": 2.3412562210399866e-05,
+      "loss": 0.981,
+      "step": 2559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.122720241546631,
+      "learning_rate": 2.340998798695727e-05,
+      "loss": 1.1322,
+      "step": 2560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.272409439086914,
+      "learning_rate": 2.3407413763514673e-05,
+      "loss": 1.2099,
+      "step": 2561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.999232769012451,
+      "learning_rate": 2.340483954007208e-05,
+      "loss": 1.0271,
+      "step": 2562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.023630619049072,
+      "learning_rate": 2.3402265316629483e-05,
+      "loss": 0.7494,
+      "step": 2563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.72968578338623,
+      "learning_rate": 2.339969109318689e-05,
+      "loss": 1.1202,
+      "step": 2564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.302896499633789,
+      "learning_rate": 2.3397116869744293e-05,
+      "loss": 0.7018,
+      "step": 2565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.548614501953125,
+      "learning_rate": 2.33945426463017e-05,
+      "loss": 1.174,
+      "step": 2566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.122110366821289,
+      "learning_rate": 2.3391968422859103e-05,
+      "loss": 1.2374,
+      "step": 2567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.153768539428711,
+      "learning_rate": 2.338939419941651e-05,
+      "loss": 1.1996,
+      "step": 2568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.598315238952637,
+      "learning_rate": 2.3386819975973916e-05,
+      "loss": 1.1503,
+      "step": 2569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.490194320678711,
+      "learning_rate": 2.3384245752531323e-05,
+      "loss": 1.1182,
+      "step": 2570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.928894996643066,
+      "learning_rate": 2.3381671529088726e-05,
+      "loss": 0.9613,
+      "step": 2571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.012547492980957,
+      "learning_rate": 2.337909730564613e-05,
+      "loss": 1.117,
+      "step": 2572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.221946716308594,
+      "learning_rate": 2.3376523082203536e-05,
+      "loss": 1.391,
+      "step": 2573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.530134201049805,
+      "learning_rate": 2.337394885876094e-05,
+      "loss": 1.1089,
+      "step": 2574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.09863567352295,
+      "learning_rate": 2.3371374635318346e-05,
+      "loss": 0.9242,
+      "step": 2575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.62304973602295,
+      "learning_rate": 2.336880041187575e-05,
+      "loss": 0.9879,
+      "step": 2576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.571710586547852,
+      "learning_rate": 2.3366226188433156e-05,
+      "loss": 1.1416,
+      "step": 2577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.05125904083252,
+      "learning_rate": 2.3363651964990563e-05,
+      "loss": 1.3324,
+      "step": 2578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.33615779876709,
+      "learning_rate": 2.336107774154797e-05,
+      "loss": 0.8008,
+      "step": 2579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.51756763458252,
+      "learning_rate": 2.3358503518105373e-05,
+      "loss": 0.9045,
+      "step": 2580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.906956672668457,
+      "learning_rate": 2.3355929294662776e-05,
+      "loss": 0.8163,
+      "step": 2581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.578678131103516,
+      "learning_rate": 2.3353355071220183e-05,
+      "loss": 1.1397,
+      "step": 2582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.404522895812988,
+      "learning_rate": 2.3350780847777586e-05,
+      "loss": 0.8252,
+      "step": 2583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.912517547607422,
+      "learning_rate": 2.3348206624334993e-05,
+      "loss": 0.9024,
+      "step": 2584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.505443572998047,
+      "learning_rate": 2.3345632400892396e-05,
+      "loss": 1.0416,
+      "step": 2585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.26039981842041,
+      "learning_rate": 2.3343058177449803e-05,
+      "loss": 0.9666,
+      "step": 2586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.800965309143066,
+      "learning_rate": 2.3340483954007206e-05,
+      "loss": 0.759,
+      "step": 2587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.022990226745605,
+      "learning_rate": 2.3337909730564616e-05,
+      "loss": 1.054,
+      "step": 2588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.547623634338379,
+      "learning_rate": 2.333533550712202e-05,
+      "loss": 0.9252,
+      "step": 2589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.132389068603516,
+      "learning_rate": 2.3332761283679426e-05,
+      "loss": 0.9291,
+      "step": 2590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.922917366027832,
+      "learning_rate": 2.333018706023683e-05,
+      "loss": 0.9224,
+      "step": 2591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.583481788635254,
+      "learning_rate": 2.3327612836794233e-05,
+      "loss": 1.0704,
+      "step": 2592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.341734886169434,
+      "learning_rate": 2.332503861335164e-05,
+      "loss": 1.1663,
+      "step": 2593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.68163013458252,
+      "learning_rate": 2.3322464389909043e-05,
+      "loss": 1.1501,
+      "step": 2594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.966670036315918,
+      "learning_rate": 2.331989016646645e-05,
+      "loss": 1.4174,
+      "step": 2595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.94918441772461,
+      "learning_rate": 2.3317315943023853e-05,
+      "loss": 1.1149,
+      "step": 2596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.16345500946045,
+      "learning_rate": 2.3314741719581263e-05,
+      "loss": 1.2118,
+      "step": 2597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.790206432342529,
+      "learning_rate": 2.3312167496138666e-05,
+      "loss": 0.864,
+      "step": 2598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.023547172546387,
+      "learning_rate": 2.3309593272696073e-05,
+      "loss": 1.086,
+      "step": 2599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.443239688873291,
+      "learning_rate": 2.3307019049253476e-05,
+      "loss": 0.7738,
+      "step": 2600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.190574645996094,
+      "learning_rate": 2.3304444825810883e-05,
+      "loss": 1.0625,
+      "step": 2601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.739475250244141,
+      "learning_rate": 2.3301870602368286e-05,
+      "loss": 1.162,
+      "step": 2602
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.268941879272461,
+      "learning_rate": 2.329929637892569e-05,
+      "loss": 1.0953,
+      "step": 2603
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.179390907287598,
+      "learning_rate": 2.3296722155483096e-05,
+      "loss": 0.9851,
+      "step": 2604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.926496505737305,
+      "learning_rate": 2.32941479320405e-05,
+      "loss": 1.1368,
+      "step": 2605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.089879035949707,
+      "learning_rate": 2.3291573708597906e-05,
+      "loss": 1.1087,
+      "step": 2606
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.267629623413086,
+      "learning_rate": 2.3288999485155313e-05,
+      "loss": 0.8739,
+      "step": 2607
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.331788063049316,
+      "learning_rate": 2.328642526171272e-05,
+      "loss": 0.9835,
+      "step": 2608
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.474777221679688,
+      "learning_rate": 2.3283851038270123e-05,
+      "loss": 1.0124,
+      "step": 2609
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.057835578918457,
+      "learning_rate": 2.328127681482753e-05,
+      "loss": 1.4244,
+      "step": 2610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.820352554321289,
+      "learning_rate": 2.3278702591384933e-05,
+      "loss": 0.902,
+      "step": 2611
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.378910064697266,
+      "learning_rate": 2.3276128367942336e-05,
+      "loss": 1.0497,
+      "step": 2612
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.737698554992676,
+      "learning_rate": 2.3273554144499743e-05,
+      "loss": 1.1132,
+      "step": 2613
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.998089790344238,
+      "learning_rate": 2.3270979921057146e-05,
+      "loss": 1.1106,
+      "step": 2614
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.101094245910645,
+      "learning_rate": 2.3268405697614553e-05,
+      "loss": 1.4158,
+      "step": 2615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.936396598815918,
+      "learning_rate": 2.326583147417196e-05,
+      "loss": 0.9709,
+      "step": 2616
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.512072563171387,
+      "learning_rate": 2.3263257250729366e-05,
+      "loss": 1.1137,
+      "step": 2617
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.632737159729004,
+      "learning_rate": 2.326068302728677e-05,
+      "loss": 0.9908,
+      "step": 2618
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.253110885620117,
+      "learning_rate": 2.3258108803844176e-05,
+      "loss": 1.0179,
+      "step": 2619
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.720448017120361,
+      "learning_rate": 2.325553458040158e-05,
+      "loss": 0.8714,
+      "step": 2620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.655285835266113,
+      "learning_rate": 2.3252960356958986e-05,
+      "loss": 0.8205,
+      "step": 2621
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.039182662963867,
+      "learning_rate": 2.325038613351639e-05,
+      "loss": 0.9435,
+      "step": 2622
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.007562637329102,
+      "learning_rate": 2.3247811910073792e-05,
+      "loss": 0.8717,
+      "step": 2623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.206671714782715,
+      "learning_rate": 2.32452376866312e-05,
+      "loss": 1.0214,
+      "step": 2624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.041656494140625,
+      "learning_rate": 2.3242663463188606e-05,
+      "loss": 0.8245,
+      "step": 2625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.8939790725708,
+      "learning_rate": 2.3240089239746013e-05,
+      "loss": 0.9214,
+      "step": 2626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.08059310913086,
+      "learning_rate": 2.3237515016303416e-05,
+      "loss": 1.0139,
+      "step": 2627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.726590156555176,
+      "learning_rate": 2.3234940792860822e-05,
+      "loss": 0.7992,
+      "step": 2628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.939742088317871,
+      "learning_rate": 2.3232366569418226e-05,
+      "loss": 0.9317,
+      "step": 2629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.064697265625,
+      "learning_rate": 2.3229792345975632e-05,
+      "loss": 1.1421,
+      "step": 2630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.365456581115723,
+      "learning_rate": 2.3227218122533036e-05,
+      "loss": 0.8567,
+      "step": 2631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.208319664001465,
+      "learning_rate": 2.3224643899090442e-05,
+      "loss": 1.1267,
+      "step": 2632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.146499633789062,
+      "learning_rate": 2.3222069675647846e-05,
+      "loss": 1.2663,
+      "step": 2633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.190032958984375,
+      "learning_rate": 2.321949545220525e-05,
+      "loss": 1.2884,
+      "step": 2634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.034303665161133,
+      "learning_rate": 2.321692122876266e-05,
+      "loss": 1.0279,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.336434364318848,
+      "learning_rate": 2.3214347005320062e-05,
+      "loss": 1.0477,
+      "step": 2636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.87071704864502,
+      "learning_rate": 2.321177278187747e-05,
+      "loss": 1.2337,
+      "step": 2637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.137707710266113,
+      "learning_rate": 2.3209198558434872e-05,
+      "loss": 1.255,
+      "step": 2638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.236285209655762,
+      "learning_rate": 2.320662433499228e-05,
+      "loss": 1.3126,
+      "step": 2639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.853729248046875,
+      "learning_rate": 2.3204050111549682e-05,
+      "loss": 1.1931,
+      "step": 2640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.575400352478027,
+      "learning_rate": 2.320147588810709e-05,
+      "loss": 1.0255,
+      "step": 2641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.494661331176758,
+      "learning_rate": 2.3198901664664492e-05,
+      "loss": 0.8133,
+      "step": 2642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.102679252624512,
+      "learning_rate": 2.31963274412219e-05,
+      "loss": 0.9565,
+      "step": 2643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.377016067504883,
+      "learning_rate": 2.3193753217779306e-05,
+      "loss": 1.2388,
+      "step": 2644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.880106449127197,
+      "learning_rate": 2.319117899433671e-05,
+      "loss": 0.8306,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.2524518966674805,
+      "learning_rate": 2.3188604770894116e-05,
+      "loss": 0.9023,
+      "step": 2646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.944124221801758,
+      "learning_rate": 2.318603054745152e-05,
+      "loss": 1.0252,
+      "step": 2647
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.78840160369873,
+      "learning_rate": 2.3183456324008926e-05,
+      "loss": 1.0816,
+      "step": 2648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.999737739562988,
+      "learning_rate": 2.318088210056633e-05,
+      "loss": 0.7798,
+      "step": 2649
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.514599800109863,
+      "learning_rate": 2.3178307877123736e-05,
+      "loss": 1.0205,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.377599716186523,
+      "learning_rate": 2.317573365368114e-05,
+      "loss": 1.173,
+      "step": 2651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.738985061645508,
+      "learning_rate": 2.3173159430238546e-05,
+      "loss": 1.08,
+      "step": 2652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.516234397888184,
+      "learning_rate": 2.317058520679595e-05,
+      "loss": 0.8602,
+      "step": 2653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.325765609741211,
+      "learning_rate": 2.3168010983353356e-05,
+      "loss": 1.1015,
+      "step": 2654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.299615859985352,
+      "learning_rate": 2.3165436759910762e-05,
+      "loss": 0.9153,
+      "step": 2655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.902812957763672,
+      "learning_rate": 2.3162862536468166e-05,
+      "loss": 0.8305,
+      "step": 2656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.841736793518066,
+      "learning_rate": 2.3160288313025572e-05,
+      "loss": 1.0197,
+      "step": 2657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.85616683959961,
+      "learning_rate": 2.3157714089582975e-05,
+      "loss": 1.2513,
+      "step": 2658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.059075355529785,
+      "learning_rate": 2.3155139866140382e-05,
+      "loss": 1.2462,
+      "step": 2659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.764070510864258,
+      "learning_rate": 2.3152565642697785e-05,
+      "loss": 1.0946,
+      "step": 2660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.843402862548828,
+      "learning_rate": 2.3149991419255192e-05,
+      "loss": 1.2516,
+      "step": 2661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.12812328338623,
+      "learning_rate": 2.3147417195812595e-05,
+      "loss": 0.9564,
+      "step": 2662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.824984550476074,
+      "learning_rate": 2.3144842972370006e-05,
+      "loss": 1.1269,
+      "step": 2663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.98025894165039,
+      "learning_rate": 2.314226874892741e-05,
+      "loss": 1.4424,
+      "step": 2664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.523545265197754,
+      "learning_rate": 2.3139694525484812e-05,
+      "loss": 0.9476,
+      "step": 2665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.977849006652832,
+      "learning_rate": 2.313712030204222e-05,
+      "loss": 0.8876,
+      "step": 2666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.086889266967773,
+      "learning_rate": 2.3134546078599622e-05,
+      "loss": 0.9347,
+      "step": 2667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.332056999206543,
+      "learning_rate": 2.313197185515703e-05,
+      "loss": 0.9366,
+      "step": 2668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.346678733825684,
+      "learning_rate": 2.3129397631714432e-05,
+      "loss": 1.0269,
+      "step": 2669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.478102684020996,
+      "learning_rate": 2.312682340827184e-05,
+      "loss": 1.1438,
+      "step": 2670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.203483581542969,
+      "learning_rate": 2.3124249184829242e-05,
+      "loss": 0.9656,
+      "step": 2671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.374492645263672,
+      "learning_rate": 2.312167496138665e-05,
+      "loss": 1.0915,
+      "step": 2672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.016454696655273,
+      "learning_rate": 2.3119100737944055e-05,
+      "loss": 1.0329,
+      "step": 2673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.490626335144043,
+      "learning_rate": 2.3116526514501462e-05,
+      "loss": 1.3229,
+      "step": 2674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.105101585388184,
+      "learning_rate": 2.3113952291058865e-05,
+      "loss": 0.871,
+      "step": 2675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.454776763916016,
+      "learning_rate": 2.311137806761627e-05,
+      "loss": 1.1395,
+      "step": 2676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.295034408569336,
+      "learning_rate": 2.3108803844173675e-05,
+      "loss": 0.9053,
+      "step": 2677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.13992977142334,
+      "learning_rate": 2.310622962073108e-05,
+      "loss": 1.3587,
+      "step": 2678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.177966117858887,
+      "learning_rate": 2.3103655397288485e-05,
+      "loss": 0.9886,
+      "step": 2679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.235572814941406,
+      "learning_rate": 2.310108117384589e-05,
+      "loss": 0.953,
+      "step": 2680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.598236083984375,
+      "learning_rate": 2.3098506950403295e-05,
+      "loss": 0.9892,
+      "step": 2681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.04853343963623,
+      "learning_rate": 2.3095932726960702e-05,
+      "loss": 1.2259,
+      "step": 2682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.717796325683594,
+      "learning_rate": 2.309335850351811e-05,
+      "loss": 1.0625,
+      "step": 2683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.189144134521484,
+      "learning_rate": 2.3090784280075512e-05,
+      "loss": 1.2453,
+      "step": 2684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.803874969482422,
+      "learning_rate": 2.3088210056632915e-05,
+      "loss": 0.787,
+      "step": 2685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.659193992614746,
+      "learning_rate": 2.3085635833190322e-05,
+      "loss": 0.8775,
+      "step": 2686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.041272163391113,
+      "learning_rate": 2.3083061609747725e-05,
+      "loss": 1.0143,
+      "step": 2687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.787315368652344,
+      "learning_rate": 2.3080487386305132e-05,
+      "loss": 1.1311,
+      "step": 2688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.955000877380371,
+      "learning_rate": 2.3077913162862535e-05,
+      "loss": 1.021,
+      "step": 2689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.984403610229492,
+      "learning_rate": 2.3075338939419942e-05,
+      "loss": 1.1738,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.193453788757324,
+      "learning_rate": 2.3072764715977345e-05,
+      "loss": 0.9614,
+      "step": 2691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.679840087890625,
+      "learning_rate": 2.3070190492534755e-05,
+      "loss": 1.1948,
+      "step": 2692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.818044662475586,
+      "learning_rate": 2.306761626909216e-05,
+      "loss": 1.0516,
+      "step": 2693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.360944747924805,
+      "learning_rate": 2.3065042045649565e-05,
+      "loss": 0.7857,
+      "step": 2694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.251326560974121,
+      "learning_rate": 2.306246782220697e-05,
+      "loss": 1.0349,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.83499526977539,
+      "learning_rate": 2.3059893598764372e-05,
+      "loss": 0.9429,
+      "step": 2696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.595775604248047,
+      "learning_rate": 2.305731937532178e-05,
+      "loss": 0.9833,
+      "step": 2697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.570494651794434,
+      "learning_rate": 2.3054745151879182e-05,
+      "loss": 1.0116,
+      "step": 2698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.598008155822754,
+      "learning_rate": 2.305217092843659e-05,
+      "loss": 0.9391,
+      "step": 2699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.88185977935791,
+      "learning_rate": 2.3049596704993992e-05,
+      "loss": 1.01,
+      "step": 2700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.411881446838379,
+      "learning_rate": 2.3047022481551402e-05,
+      "loss": 0.9149,
+      "step": 2701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.78014850616455,
+      "learning_rate": 2.3044448258108805e-05,
+      "loss": 0.8936,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.988958835601807,
+      "learning_rate": 2.3041874034666212e-05,
+      "loss": 0.6614,
+      "step": 2703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.601667404174805,
+      "learning_rate": 2.3039299811223615e-05,
+      "loss": 1.1616,
+      "step": 2704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.891489028930664,
+      "learning_rate": 2.3036725587781022e-05,
+      "loss": 1.3806,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.160721778869629,
+      "learning_rate": 2.3034151364338425e-05,
+      "loss": 1.0335,
+      "step": 2706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.962326526641846,
+      "learning_rate": 2.303157714089583e-05,
+      "loss": 0.7447,
+      "step": 2707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.290735244750977,
+      "learning_rate": 2.3029002917453235e-05,
+      "loss": 1.0803,
+      "step": 2708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.305471420288086,
+      "learning_rate": 2.302642869401064e-05,
+      "loss": 1.1649,
+      "step": 2709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.52600383758545,
+      "learning_rate": 2.3023854470568045e-05,
+      "loss": 1.2812,
+      "step": 2710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.067754745483398,
+      "learning_rate": 2.302128024712545e-05,
+      "loss": 1.0728,
+      "step": 2711
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.478352546691895,
+      "learning_rate": 2.301870602368286e-05,
+      "loss": 0.9654,
+      "step": 2712
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.699317932128906,
+      "learning_rate": 2.301613180024026e-05,
+      "loss": 0.9325,
+      "step": 2713
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.604063987731934,
+      "learning_rate": 2.301355757679767e-05,
+      "loss": 1.0163,
+      "step": 2714
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.880563259124756,
+      "learning_rate": 2.301098335335507e-05,
+      "loss": 0.9194,
+      "step": 2715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.372673034667969,
+      "learning_rate": 2.3008409129912475e-05,
+      "loss": 1.118,
+      "step": 2716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.623473167419434,
+      "learning_rate": 2.300583490646988e-05,
+      "loss": 0.8165,
+      "step": 2717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.650661945343018,
+      "learning_rate": 2.3003260683027285e-05,
+      "loss": 0.9933,
+      "step": 2718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.055195808410645,
+      "learning_rate": 2.300068645958469e-05,
+      "loss": 1.1341,
+      "step": 2719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.060074806213379,
+      "learning_rate": 2.2998112236142098e-05,
+      "loss": 0.9825,
+      "step": 2720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.135762214660645,
+      "learning_rate": 2.2995538012699505e-05,
+      "loss": 1.2304,
+      "step": 2721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.723862648010254,
+      "learning_rate": 2.2992963789256908e-05,
+      "loss": 1.0414,
+      "step": 2722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.84394645690918,
+      "learning_rate": 2.2990389565814315e-05,
+      "loss": 1.1302,
+      "step": 2723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.975812911987305,
+      "learning_rate": 2.2987815342371718e-05,
+      "loss": 0.8251,
+      "step": 2724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.302264213562012,
+      "learning_rate": 2.2985241118929125e-05,
+      "loss": 1.1593,
+      "step": 2725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.009163856506348,
+      "learning_rate": 2.2982666895486528e-05,
+      "loss": 1.0301,
+      "step": 2726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.274458885192871,
+      "learning_rate": 2.298009267204393e-05,
+      "loss": 1.1649,
+      "step": 2727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.925476551055908,
+      "learning_rate": 2.2977518448601338e-05,
+      "loss": 0.9727,
+      "step": 2728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.69002914428711,
+      "learning_rate": 2.297494422515874e-05,
+      "loss": 1.0547,
+      "step": 2729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.099454879760742,
+      "learning_rate": 2.297237000171615e-05,
+      "loss": 1.1742,
+      "step": 2730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.325862884521484,
+      "learning_rate": 2.2969795778273555e-05,
+      "loss": 1.0888,
+      "step": 2731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.108589172363281,
+      "learning_rate": 2.296722155483096e-05,
+      "loss": 0.9231,
+      "step": 2732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.083965301513672,
+      "learning_rate": 2.2964647331388365e-05,
+      "loss": 1.1818,
+      "step": 2733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.379091262817383,
+      "learning_rate": 2.296207310794577e-05,
+      "loss": 1.1966,
+      "step": 2734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.886136054992676,
+      "learning_rate": 2.2959498884503175e-05,
+      "loss": 1.2887,
+      "step": 2735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.736721992492676,
+      "learning_rate": 2.295692466106058e-05,
+      "loss": 1.0544,
+      "step": 2736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.164587020874023,
+      "learning_rate": 2.2954350437617985e-05,
+      "loss": 0.9271,
+      "step": 2737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.921046257019043,
+      "learning_rate": 2.2951776214175388e-05,
+      "loss": 0.8671,
+      "step": 2738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.995047569274902,
+      "learning_rate": 2.2949201990732798e-05,
+      "loss": 1.0532,
+      "step": 2739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.59412670135498,
+      "learning_rate": 2.29466277672902e-05,
+      "loss": 0.8976,
+      "step": 2740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.083028793334961,
+      "learning_rate": 2.2944053543847608e-05,
+      "loss": 1.1602,
+      "step": 2741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.787667274475098,
+      "learning_rate": 2.294147932040501e-05,
+      "loss": 1.0648,
+      "step": 2742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.504075050354004,
+      "learning_rate": 2.2938905096962418e-05,
+      "loss": 1.009,
+      "step": 2743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.056138038635254,
+      "learning_rate": 2.293633087351982e-05,
+      "loss": 1.0141,
+      "step": 2744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.839750289916992,
+      "learning_rate": 2.2933756650077228e-05,
+      "loss": 0.873,
+      "step": 2745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.619115829467773,
+      "learning_rate": 2.293118242663463e-05,
+      "loss": 0.9876,
+      "step": 2746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.747673034667969,
+      "learning_rate": 2.2928608203192038e-05,
+      "loss": 1.2646,
+      "step": 2747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.004373550415039,
+      "learning_rate": 2.2926033979749445e-05,
+      "loss": 1.4506,
+      "step": 2748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.066360473632812,
+      "learning_rate": 2.2923459756306848e-05,
+      "loss": 0.8475,
+      "step": 2749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.013704299926758,
+      "learning_rate": 2.2920885532864255e-05,
+      "loss": 1.0386,
+      "step": 2750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.597650527954102,
+      "learning_rate": 2.2918311309421658e-05,
+      "loss": 1.4272,
+      "step": 2751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.709369659423828,
+      "learning_rate": 2.2915737085979065e-05,
+      "loss": 1.3146,
+      "step": 2752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.2885942459106445,
+      "learning_rate": 2.2913162862536468e-05,
+      "loss": 0.7658,
+      "step": 2753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.996438980102539,
+      "learning_rate": 2.2910588639093875e-05,
+      "loss": 1.0493,
+      "step": 2754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.672294616699219,
+      "learning_rate": 2.2908014415651278e-05,
+      "loss": 1.4401,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.644652366638184,
+      "learning_rate": 2.2905440192208685e-05,
+      "loss": 1.0895,
+      "step": 2756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.453187942504883,
+      "learning_rate": 2.2902865968766088e-05,
+      "loss": 1.0532,
+      "step": 2757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.38924789428711,
+      "learning_rate": 2.2900291745323495e-05,
+      "loss": 1.0041,
+      "step": 2758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.288237571716309,
+      "learning_rate": 2.28977175218809e-05,
+      "loss": 1.269,
+      "step": 2759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.722000122070312,
+      "learning_rate": 2.2895143298438305e-05,
+      "loss": 1.1034,
+      "step": 2760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.730677604675293,
+      "learning_rate": 2.289256907499571e-05,
+      "loss": 1.0287,
+      "step": 2761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.569239616394043,
+      "learning_rate": 2.2889994851553114e-05,
+      "loss": 1.1965,
+      "step": 2762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.154129028320312,
+      "learning_rate": 2.288742062811052e-05,
+      "loss": 1.194,
+      "step": 2763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.696370124816895,
+      "learning_rate": 2.2884846404667924e-05,
+      "loss": 1.3635,
+      "step": 2764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.712349891662598,
+      "learning_rate": 2.288227218122533e-05,
+      "loss": 0.8794,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.65976619720459,
+      "learning_rate": 2.2879697957782734e-05,
+      "loss": 1.0708,
+      "step": 2766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.705913543701172,
+      "learning_rate": 2.2877123734340145e-05,
+      "loss": 0.8806,
+      "step": 2767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.749248504638672,
+      "learning_rate": 2.2874549510897548e-05,
+      "loss": 1.1793,
+      "step": 2768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.434737205505371,
+      "learning_rate": 2.287197528745495e-05,
+      "loss": 1.2681,
+      "step": 2769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.738548278808594,
+      "learning_rate": 2.2869401064012358e-05,
+      "loss": 0.9724,
+      "step": 2770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.327269554138184,
+      "learning_rate": 2.286682684056976e-05,
+      "loss": 1.1014,
+      "step": 2771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.601359367370605,
+      "learning_rate": 2.2864252617127168e-05,
+      "loss": 1.1573,
+      "step": 2772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.368884086608887,
+      "learning_rate": 2.286167839368457e-05,
+      "loss": 0.8289,
+      "step": 2773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.068943977355957,
+      "learning_rate": 2.2859104170241978e-05,
+      "loss": 1.1792,
+      "step": 2774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.891889572143555,
+      "learning_rate": 2.285652994679938e-05,
+      "loss": 1.2894,
+      "step": 2775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.454261779785156,
+      "learning_rate": 2.2853955723356788e-05,
+      "loss": 1.4133,
+      "step": 2776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.031497955322266,
+      "learning_rate": 2.2851381499914194e-05,
+      "loss": 1.0014,
+      "step": 2777
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.623445987701416,
+      "learning_rate": 2.28488072764716e-05,
+      "loss": 0.78,
+      "step": 2778
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.759760856628418,
+      "learning_rate": 2.2846233053029004e-05,
+      "loss": 1.2732,
+      "step": 2779
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.171009063720703,
+      "learning_rate": 2.2843658829586408e-05,
+      "loss": 0.856,
+      "step": 2780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.820732116699219,
+      "learning_rate": 2.2841084606143814e-05,
+      "loss": 1.0889,
+      "step": 2781
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.587847709655762,
+      "learning_rate": 2.2838510382701218e-05,
+      "loss": 1.0345,
+      "step": 2782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.194660186767578,
+      "learning_rate": 2.2835936159258624e-05,
+      "loss": 1.1592,
+      "step": 2783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.72214126586914,
+      "learning_rate": 2.2833361935816028e-05,
+      "loss": 0.9592,
+      "step": 2784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.25465202331543,
+      "learning_rate": 2.2830787712373434e-05,
+      "loss": 1.1084,
+      "step": 2785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.497455596923828,
+      "learning_rate": 2.282821348893084e-05,
+      "loss": 0.9393,
+      "step": 2786
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.359732627868652,
+      "learning_rate": 2.2825639265488248e-05,
+      "loss": 1.0535,
+      "step": 2787
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.600067138671875,
+      "learning_rate": 2.282306504204565e-05,
+      "loss": 0.7851,
+      "step": 2788
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.811413764953613,
+      "learning_rate": 2.2820490818603054e-05,
+      "loss": 1.1524,
+      "step": 2789
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.236885070800781,
+      "learning_rate": 2.281791659516046e-05,
+      "loss": 1.0283,
+      "step": 2790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.800850868225098,
+      "learning_rate": 2.2815342371717864e-05,
+      "loss": 1.1249,
+      "step": 2791
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.862846851348877,
+      "learning_rate": 2.281276814827527e-05,
+      "loss": 0.7767,
+      "step": 2792
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.416818618774414,
+      "learning_rate": 2.2810193924832674e-05,
+      "loss": 1.0211,
+      "step": 2793
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.245564460754395,
+      "learning_rate": 2.280761970139008e-05,
+      "loss": 1.3034,
+      "step": 2794
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.16881275177002,
+      "learning_rate": 2.2805045477947484e-05,
+      "loss": 1.2065,
+      "step": 2795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.259509086608887,
+      "learning_rate": 2.2802471254504894e-05,
+      "loss": 0.9956,
+      "step": 2796
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.27147388458252,
+      "learning_rate": 2.2799897031062298e-05,
+      "loss": 0.9276,
+      "step": 2797
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.915818214416504,
+      "learning_rate": 2.2797322807619704e-05,
+      "loss": 0.9996,
+      "step": 2798
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.080011367797852,
+      "learning_rate": 2.2794748584177107e-05,
+      "loss": 0.8308,
+      "step": 2799
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.386871337890625,
+      "learning_rate": 2.279217436073451e-05,
+      "loss": 1.4012,
+      "step": 2800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.126442909240723,
+      "learning_rate": 2.2789600137291917e-05,
+      "loss": 1.0585,
+      "step": 2801
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.754056930541992,
+      "learning_rate": 2.278702591384932e-05,
+      "loss": 1.2098,
+      "step": 2802
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.785674095153809,
+      "learning_rate": 2.2784451690406727e-05,
+      "loss": 1.4384,
+      "step": 2803
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.46529483795166,
+      "learning_rate": 2.278187746696413e-05,
+      "loss": 0.7423,
+      "step": 2804
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.653386116027832,
+      "learning_rate": 2.277930324352154e-05,
+      "loss": 1.4328,
+      "step": 2805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.566006660461426,
+      "learning_rate": 2.2776729020078944e-05,
+      "loss": 1.0497,
+      "step": 2806
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.502777099609375,
+      "learning_rate": 2.277415479663635e-05,
+      "loss": 0.8336,
+      "step": 2807
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.23631477355957,
+      "learning_rate": 2.2771580573193754e-05,
+      "loss": 0.9385,
+      "step": 2808
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.036250114440918,
+      "learning_rate": 2.276900634975116e-05,
+      "loss": 0.9397,
+      "step": 2809
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.418767929077148,
+      "learning_rate": 2.2766432126308564e-05,
+      "loss": 0.8866,
+      "step": 2810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.312596321105957,
+      "learning_rate": 2.2763857902865967e-05,
+      "loss": 1.059,
+      "step": 2811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.346514701843262,
+      "learning_rate": 2.2761283679423374e-05,
+      "loss": 0.8439,
+      "step": 2812
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.678810119628906,
+      "learning_rate": 2.2758709455980777e-05,
+      "loss": 0.8421,
+      "step": 2813
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.164229393005371,
+      "learning_rate": 2.2756135232538184e-05,
+      "loss": 1.0234,
+      "step": 2814
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.911999702453613,
+      "learning_rate": 2.275356100909559e-05,
+      "loss": 1.0253,
+      "step": 2815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.797465324401855,
+      "learning_rate": 2.2750986785652997e-05,
+      "loss": 1.1644,
+      "step": 2816
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.00514030456543,
+      "learning_rate": 2.27484125622104e-05,
+      "loss": 1.4585,
+      "step": 2817
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.886373996734619,
+      "learning_rate": 2.2745838338767807e-05,
+      "loss": 0.9078,
+      "step": 2818
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.170586585998535,
+      "learning_rate": 2.274326411532521e-05,
+      "loss": 1.3247,
+      "step": 2819
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.58406925201416,
+      "learning_rate": 2.2740689891882614e-05,
+      "loss": 1.162,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.046093940734863,
+      "learning_rate": 2.273811566844002e-05,
+      "loss": 0.9556,
+      "step": 2821
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.614295959472656,
+      "learning_rate": 2.2735541444997424e-05,
+      "loss": 1.2942,
+      "step": 2822
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.412017345428467,
+      "learning_rate": 2.273296722155483e-05,
+      "loss": 0.7609,
+      "step": 2823
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.07819938659668,
+      "learning_rate": 2.2730392998112237e-05,
+      "loss": 0.7727,
+      "step": 2824
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.380146980285645,
+      "learning_rate": 2.2727818774669644e-05,
+      "loss": 1.11,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.714298248291016,
+      "learning_rate": 2.2725244551227047e-05,
+      "loss": 1.2154,
+      "step": 2826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.598361015319824,
+      "learning_rate": 2.2722670327784454e-05,
+      "loss": 0.9784,
+      "step": 2827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.480252265930176,
+      "learning_rate": 2.2720096104341857e-05,
+      "loss": 0.9387,
+      "step": 2828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.496916770935059,
+      "learning_rate": 2.2717521880899264e-05,
+      "loss": 1.0954,
+      "step": 2829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.717202186584473,
+      "learning_rate": 2.2714947657456667e-05,
+      "loss": 1.0192,
+      "step": 2830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.959425926208496,
+      "learning_rate": 2.271237343401407e-05,
+      "loss": 1.0376,
+      "step": 2831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.882668495178223,
+      "learning_rate": 2.2709799210571477e-05,
+      "loss": 0.9194,
+      "step": 2832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.36072301864624,
+      "learning_rate": 2.270722498712888e-05,
+      "loss": 0.7008,
+      "step": 2833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.445476531982422,
+      "learning_rate": 2.270465076368629e-05,
+      "loss": 1.0435,
+      "step": 2834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.398406982421875,
+      "learning_rate": 2.2702076540243694e-05,
+      "loss": 1.0314,
+      "step": 2835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.824779987335205,
+      "learning_rate": 2.26995023168011e-05,
+      "loss": 0.8664,
+      "step": 2836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.393730163574219,
+      "learning_rate": 2.2696928093358504e-05,
+      "loss": 0.9419,
+      "step": 2837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.884232521057129,
+      "learning_rate": 2.269435386991591e-05,
+      "loss": 0.9628,
+      "step": 2838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.899201393127441,
+      "learning_rate": 2.2691779646473314e-05,
+      "loss": 0.804,
+      "step": 2839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.749068260192871,
+      "learning_rate": 2.268920542303072e-05,
+      "loss": 1.3074,
+      "step": 2840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.67462158203125,
+      "learning_rate": 2.2686631199588124e-05,
+      "loss": 1.093,
+      "step": 2841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.70548152923584,
+      "learning_rate": 2.2684056976145527e-05,
+      "loss": 1.016,
+      "step": 2842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.614361763000488,
+      "learning_rate": 2.2681482752702937e-05,
+      "loss": 0.9766,
+      "step": 2843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.241039276123047,
+      "learning_rate": 2.267890852926034e-05,
+      "loss": 0.9755,
+      "step": 2844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.387537002563477,
+      "learning_rate": 2.2676334305817747e-05,
+      "loss": 0.8501,
+      "step": 2845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.643594741821289,
+      "learning_rate": 2.267376008237515e-05,
+      "loss": 1.5157,
+      "step": 2846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.009003639221191,
+      "learning_rate": 2.2671185858932557e-05,
+      "loss": 1.1499,
+      "step": 2847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.391432762145996,
+      "learning_rate": 2.266861163548996e-05,
+      "loss": 1.2275,
+      "step": 2848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.121082305908203,
+      "learning_rate": 2.2666037412047367e-05,
+      "loss": 1.1788,
+      "step": 2849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.744053840637207,
+      "learning_rate": 2.266346318860477e-05,
+      "loss": 0.9646,
+      "step": 2850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.824888229370117,
+      "learning_rate": 2.2660888965162177e-05,
+      "loss": 1.0316,
+      "step": 2851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.28082275390625,
+      "learning_rate": 2.265831474171958e-05,
+      "loss": 0.9807,
+      "step": 2852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.869412422180176,
+      "learning_rate": 2.2655740518276987e-05,
+      "loss": 0.9285,
+      "step": 2853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.541510105133057,
+      "learning_rate": 2.2653166294834394e-05,
+      "loss": 0.9391,
+      "step": 2854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.450916290283203,
+      "learning_rate": 2.2650592071391797e-05,
+      "loss": 0.9292,
+      "step": 2855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.403661727905273,
+      "learning_rate": 2.2648017847949204e-05,
+      "loss": 1.1493,
+      "step": 2856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.383374214172363,
+      "learning_rate": 2.2645443624506607e-05,
+      "loss": 0.9382,
+      "step": 2857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.462475299835205,
+      "learning_rate": 2.2642869401064014e-05,
+      "loss": 0.6428,
+      "step": 2858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.96205472946167,
+      "learning_rate": 2.2640295177621417e-05,
+      "loss": 0.8163,
+      "step": 2859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.264400482177734,
+      "learning_rate": 2.2637720954178824e-05,
+      "loss": 0.9747,
+      "step": 2860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.79320240020752,
+      "learning_rate": 2.2635146730736227e-05,
+      "loss": 1.5273,
+      "step": 2861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.706038475036621,
+      "learning_rate": 2.2632572507293634e-05,
+      "loss": 1.2014,
+      "step": 2862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.922988891601562,
+      "learning_rate": 2.262999828385104e-05,
+      "loss": 1.104,
+      "step": 2863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.83487606048584,
+      "learning_rate": 2.2627424060408444e-05,
+      "loss": 1.2641,
+      "step": 2864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.853885650634766,
+      "learning_rate": 2.262484983696585e-05,
+      "loss": 1.1656,
+      "step": 2865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.23297119140625,
+      "learning_rate": 2.2622275613523254e-05,
+      "loss": 0.6008,
+      "step": 2866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.741523742675781,
+      "learning_rate": 2.261970139008066e-05,
+      "loss": 1.371,
+      "step": 2867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.03989315032959,
+      "learning_rate": 2.2617127166638063e-05,
+      "loss": 1.0317,
+      "step": 2868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.759997367858887,
+      "learning_rate": 2.261455294319547e-05,
+      "loss": 1.3011,
+      "step": 2869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.567174911499023,
+      "learning_rate": 2.2611978719752873e-05,
+      "loss": 1.3222,
+      "step": 2870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.156388282775879,
+      "learning_rate": 2.2609404496310284e-05,
+      "loss": 0.7418,
+      "step": 2871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.658440589904785,
+      "learning_rate": 2.2606830272867687e-05,
+      "loss": 1.2242,
+      "step": 2872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.034728050231934,
+      "learning_rate": 2.260425604942509e-05,
+      "loss": 0.7375,
+      "step": 2873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.8517427444458,
+      "learning_rate": 2.2601681825982497e-05,
+      "loss": 1.0272,
+      "step": 2874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.360161781311035,
+      "learning_rate": 2.25991076025399e-05,
+      "loss": 0.8941,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.058897018432617,
+      "learning_rate": 2.2596533379097307e-05,
+      "loss": 0.9598,
+      "step": 2876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.791024208068848,
+      "learning_rate": 2.259395915565471e-05,
+      "loss": 1.1843,
+      "step": 2877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.726823806762695,
+      "learning_rate": 2.2591384932212117e-05,
+      "loss": 0.8995,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.622626304626465,
+      "learning_rate": 2.258881070876952e-05,
+      "loss": 0.896,
+      "step": 2879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.012248992919922,
+      "learning_rate": 2.2586236485326927e-05,
+      "loss": 1.0955,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.1736478805542,
+      "learning_rate": 2.2583662261884333e-05,
+      "loss": 0.9293,
+      "step": 2881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.817082405090332,
+      "learning_rate": 2.258108803844174e-05,
+      "loss": 1.0724,
+      "step": 2882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.296951293945312,
+      "learning_rate": 2.2578513814999143e-05,
+      "loss": 1.0913,
+      "step": 2883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.798802375793457,
+      "learning_rate": 2.2575939591556547e-05,
+      "loss": 1.1309,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.366511344909668,
+      "learning_rate": 2.2573365368113953e-05,
+      "loss": 0.8792,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.514967918395996,
+      "learning_rate": 2.2570791144671357e-05,
+      "loss": 1.0338,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.1610746383667,
+      "learning_rate": 2.2568216921228763e-05,
+      "loss": 1.0468,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.265742301940918,
+      "learning_rate": 2.2565642697786167e-05,
+      "loss": 1.1355,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.74632453918457,
+      "learning_rate": 2.2563068474343573e-05,
+      "loss": 1.0831,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.062989234924316,
+      "learning_rate": 2.256049425090098e-05,
+      "loss": 1.3055,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.476349830627441,
+      "learning_rate": 2.2557920027458387e-05,
+      "loss": 1.2305,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.062787055969238,
+      "learning_rate": 2.255534580401579e-05,
+      "loss": 1.3284,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.596938133239746,
+      "learning_rate": 2.2552771580573193e-05,
+      "loss": 1.0379,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.208369255065918,
+      "learning_rate": 2.25501973571306e-05,
+      "loss": 0.8495,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.72470474243164,
+      "learning_rate": 2.2547623133688003e-05,
+      "loss": 0.9906,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.03720760345459,
+      "learning_rate": 2.254504891024541e-05,
+      "loss": 1.004,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.569757461547852,
+      "learning_rate": 2.2542474686802813e-05,
+      "loss": 1.1559,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.373429298400879,
+      "learning_rate": 2.253990046336022e-05,
+      "loss": 0.8029,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.130106449127197,
+      "learning_rate": 2.2537326239917623e-05,
+      "loss": 0.8941,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.858506202697754,
+      "learning_rate": 2.2534752016475033e-05,
+      "loss": 0.6541,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.609339714050293,
+      "learning_rate": 2.2532177793032437e-05,
+      "loss": 1.0596,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.439431190490723,
+      "learning_rate": 2.2529603569589843e-05,
+      "loss": 1.0829,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.836402893066406,
+      "learning_rate": 2.2527029346147247e-05,
+      "loss": 1.3006,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.604819297790527,
+      "learning_rate": 2.252445512270465e-05,
+      "loss": 1.1105,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.094498634338379,
+      "learning_rate": 2.2521880899262056e-05,
+      "loss": 1.2208,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.871305465698242,
+      "learning_rate": 2.251930667581946e-05,
+      "loss": 0.8997,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.772424697875977,
+      "learning_rate": 2.2516732452376866e-05,
+      "loss": 0.9788,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.608298301696777,
+      "learning_rate": 2.251415822893427e-05,
+      "loss": 1.1644,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.14294719696045,
+      "learning_rate": 2.251158400549168e-05,
+      "loss": 1.2377,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.051589012145996,
+      "learning_rate": 2.2509009782049083e-05,
+      "loss": 1.0112,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.835176467895508,
+      "learning_rate": 2.250643555860649e-05,
+      "loss": 1.1839,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.873061180114746,
+      "learning_rate": 2.2503861335163893e-05,
+      "loss": 0.8594,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.312557697296143,
+      "learning_rate": 2.25012871117213e-05,
+      "loss": 0.8498,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.777792930603027,
+      "learning_rate": 2.2498712888278703e-05,
+      "loss": 0.9957,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.795259475708008,
+      "learning_rate": 2.2496138664836106e-05,
+      "loss": 1.1728,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.050849914550781,
+      "learning_rate": 2.2493564441393513e-05,
+      "loss": 1.4074,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.401333808898926,
+      "learning_rate": 2.2490990217950916e-05,
+      "loss": 0.826,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.290714263916016,
+      "learning_rate": 2.2488415994508323e-05,
+      "loss": 0.9411,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.823974609375,
+      "learning_rate": 2.248584177106573e-05,
+      "loss": 1.1492,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.005365371704102,
+      "learning_rate": 2.2483267547623136e-05,
+      "loss": 0.8701,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.801459312438965,
+      "learning_rate": 2.248069332418054e-05,
+      "loss": 0.8018,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.824800491333008,
+      "learning_rate": 2.2478119100737946e-05,
+      "loss": 0.7951,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.079216003417969,
+      "learning_rate": 2.247554487729535e-05,
+      "loss": 0.8694,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.32170295715332,
+      "learning_rate": 2.2472970653852753e-05,
+      "loss": 0.9289,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.865641593933105,
+      "learning_rate": 2.247039643041016e-05,
+      "loss": 0.9216,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.041830062866211,
+      "learning_rate": 2.2467822206967563e-05,
+      "loss": 0.7588,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.602577209472656,
+      "learning_rate": 2.246524798352497e-05,
+      "loss": 0.8429,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.01832389831543,
+      "learning_rate": 2.2462673760082376e-05,
+      "loss": 0.9837,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.765362739562988,
+      "learning_rate": 2.2460099536639783e-05,
+      "loss": 1.2726,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.35781478881836,
+      "learning_rate": 2.2457525313197186e-05,
+      "loss": 0.7913,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.378927230834961,
+      "learning_rate": 2.2454951089754593e-05,
+      "loss": 0.9463,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.902555465698242,
+      "learning_rate": 2.2452376866311996e-05,
+      "loss": 0.8122,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.180845260620117,
+      "learning_rate": 2.2449802642869403e-05,
+      "loss": 0.8902,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.628933906555176,
+      "learning_rate": 2.2447228419426806e-05,
+      "loss": 1.2149,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.003393173217773,
+      "learning_rate": 2.244465419598421e-05,
+      "loss": 0.8691,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.806173324584961,
+      "learning_rate": 2.2442079972541616e-05,
+      "loss": 1.2196,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.46871280670166,
+      "learning_rate": 2.243950574909902e-05,
+      "loss": 0.9806,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.911006927490234,
+      "learning_rate": 2.243693152565643e-05,
+      "loss": 0.8281,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.352621078491211,
+      "learning_rate": 2.2434357302213833e-05,
+      "loss": 1.0753,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.792496681213379,
+      "learning_rate": 2.243178307877124e-05,
+      "loss": 0.963,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.439682006835938,
+      "learning_rate": 2.2429208855328643e-05,
+      "loss": 0.9795,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.610797882080078,
+      "learning_rate": 2.242663463188605e-05,
+      "loss": 0.8668,
+      "step": 2942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.38074016571045,
+      "learning_rate": 2.2424060408443453e-05,
+      "loss": 1.3201,
+      "step": 2943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.588486671447754,
+      "learning_rate": 2.242148618500086e-05,
+      "loss": 1.0096,
+      "step": 2944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.158686637878418,
+      "learning_rate": 2.2418911961558263e-05,
+      "loss": 0.9543,
+      "step": 2945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.06312084197998,
+      "learning_rate": 2.2416337738115666e-05,
+      "loss": 0.8876,
+      "step": 2946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.975364685058594,
+      "learning_rate": 2.2413763514673076e-05,
+      "loss": 1.313,
+      "step": 2947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.51795482635498,
+      "learning_rate": 2.241118929123048e-05,
+      "loss": 1.2252,
+      "step": 2948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.76899528503418,
+      "learning_rate": 2.2408615067787886e-05,
+      "loss": 0.8506,
+      "step": 2949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.216376304626465,
+      "learning_rate": 2.240604084434529e-05,
+      "loss": 0.9271,
+      "step": 2950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.150128364562988,
+      "learning_rate": 2.2403466620902696e-05,
+      "loss": 1.004,
+      "step": 2951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.28695297241211,
+      "learning_rate": 2.24008923974601e-05,
+      "loss": 0.9094,
+      "step": 2952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.283291816711426,
+      "learning_rate": 2.2398318174017506e-05,
+      "loss": 1.1637,
+      "step": 2953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.857816696166992,
+      "learning_rate": 2.239574395057491e-05,
+      "loss": 1.0686,
+      "step": 2954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.769723892211914,
+      "learning_rate": 2.2393169727132316e-05,
+      "loss": 1.1265,
+      "step": 2955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.864490032196045,
+      "learning_rate": 2.239059550368972e-05,
+      "loss": 0.9829,
+      "step": 2956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.275936126708984,
+      "learning_rate": 2.2388021280247126e-05,
+      "loss": 0.9762,
+      "step": 2957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.359052658081055,
+      "learning_rate": 2.2385447056804533e-05,
+      "loss": 0.8706,
+      "step": 2958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.31600570678711,
+      "learning_rate": 2.2382872833361936e-05,
+      "loss": 1.0673,
+      "step": 2959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.782825469970703,
+      "learning_rate": 2.2380298609919343e-05,
+      "loss": 1.0843,
+      "step": 2960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.320279121398926,
+      "learning_rate": 2.2377724386476746e-05,
+      "loss": 0.922,
+      "step": 2961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.702962875366211,
+      "learning_rate": 2.2375150163034153e-05,
+      "loss": 1.0546,
+      "step": 2962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.960341453552246,
+      "learning_rate": 2.2372575939591556e-05,
+      "loss": 1.0943,
+      "step": 2963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.947541236877441,
+      "learning_rate": 2.2370001716148963e-05,
+      "loss": 1.1067,
+      "step": 2964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.1266508102417,
+      "learning_rate": 2.2367427492706366e-05,
+      "loss": 0.9991,
+      "step": 2965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.976988792419434,
+      "learning_rate": 2.2364853269263773e-05,
+      "loss": 0.8191,
+      "step": 2966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.975715637207031,
+      "learning_rate": 2.236227904582118e-05,
+      "loss": 1.0675,
+      "step": 2967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.150272369384766,
+      "learning_rate": 2.2359704822378583e-05,
+      "loss": 0.9077,
+      "step": 2968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.799205780029297,
+      "learning_rate": 2.235713059893599e-05,
+      "loss": 1.3587,
+      "step": 2969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.644124031066895,
+      "learning_rate": 2.2354556375493393e-05,
+      "loss": 1.458,
+      "step": 2970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.589526176452637,
+      "learning_rate": 2.23519821520508e-05,
+      "loss": 1.245,
+      "step": 2971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.116387367248535,
+      "learning_rate": 2.2349407928608202e-05,
+      "loss": 1.0382,
+      "step": 2972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.83653450012207,
+      "learning_rate": 2.234683370516561e-05,
+      "loss": 1.156,
+      "step": 2973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.72766399383545,
+      "learning_rate": 2.2344259481723012e-05,
+      "loss": 0.9832,
+      "step": 2974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.677478790283203,
+      "learning_rate": 2.2341685258280423e-05,
+      "loss": 0.8163,
+      "step": 2975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.391899108886719,
+      "learning_rate": 2.2339111034837826e-05,
+      "loss": 0.9602,
+      "step": 2976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.288619041442871,
+      "learning_rate": 2.233653681139523e-05,
+      "loss": 0.8514,
+      "step": 2977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.410719871520996,
+      "learning_rate": 2.2333962587952636e-05,
+      "loss": 0.9772,
+      "step": 2978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.934706687927246,
+      "learning_rate": 2.233138836451004e-05,
+      "loss": 0.9284,
+      "step": 2979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.661917686462402,
+      "learning_rate": 2.2328814141067446e-05,
+      "loss": 0.9782,
+      "step": 2980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.239506721496582,
+      "learning_rate": 2.232623991762485e-05,
+      "loss": 0.7205,
+      "step": 2981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.699063301086426,
+      "learning_rate": 2.2323665694182256e-05,
+      "loss": 1.3547,
+      "step": 2982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.53893756866455,
+      "learning_rate": 2.232109147073966e-05,
+      "loss": 0.9747,
+      "step": 2983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.555697441101074,
+      "learning_rate": 2.2318517247297066e-05,
+      "loss": 0.7239,
+      "step": 2984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.867225170135498,
+      "learning_rate": 2.2315943023854472e-05,
+      "loss": 0.836,
+      "step": 2985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.933558464050293,
+      "learning_rate": 2.231336880041188e-05,
+      "loss": 0.9002,
+      "step": 2986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.113739013671875,
+      "learning_rate": 2.2310794576969282e-05,
+      "loss": 0.844,
+      "step": 2987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.667778968811035,
+      "learning_rate": 2.2308220353526686e-05,
+      "loss": 1.0439,
+      "step": 2988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.227678298950195,
+      "learning_rate": 2.2305646130084092e-05,
+      "loss": 0.9577,
+      "step": 2989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.674300193786621,
+      "learning_rate": 2.2303071906641496e-05,
+      "loss": 1.0516,
+      "step": 2990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.962779998779297,
+      "learning_rate": 2.2300497683198902e-05,
+      "loss": 1.0495,
+      "step": 2991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.408998489379883,
+      "learning_rate": 2.2297923459756306e-05,
+      "loss": 0.6937,
+      "step": 2992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.951751708984375,
+      "learning_rate": 2.2295349236313712e-05,
+      "loss": 0.9929,
+      "step": 2993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.693395614624023,
+      "learning_rate": 2.229277501287112e-05,
+      "loss": 1.1388,
+      "step": 2994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.005244255065918,
+      "learning_rate": 2.2290200789428526e-05,
+      "loss": 1.2366,
+      "step": 2995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.858135223388672,
+      "learning_rate": 2.228762656598593e-05,
+      "loss": 0.7901,
+      "step": 2996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.465733528137207,
+      "learning_rate": 2.2285052342543332e-05,
+      "loss": 1.0484,
+      "step": 2997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.89690113067627,
+      "learning_rate": 2.228247811910074e-05,
+      "loss": 1.2352,
+      "step": 2998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.465445518493652,
+      "learning_rate": 2.2279903895658142e-05,
+      "loss": 1.2596,
+      "step": 2999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.890931129455566,
+      "learning_rate": 2.227732967221555e-05,
+      "loss": 1.2102,
+      "step": 3000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.394305229187012,
+      "learning_rate": 2.2274755448772952e-05,
+      "loss": 1.0521,
+      "step": 3001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.572005271911621,
+      "learning_rate": 2.227218122533036e-05,
+      "loss": 0.915,
+      "step": 3002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.06268310546875,
+      "learning_rate": 2.2269607001887762e-05,
+      "loss": 0.8785,
+      "step": 3003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.87403392791748,
+      "learning_rate": 2.2267032778445172e-05,
+      "loss": 1.2461,
+      "step": 3004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.28782844543457,
+      "learning_rate": 2.2264458555002576e-05,
+      "loss": 0.8862,
+      "step": 3005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.314935684204102,
+      "learning_rate": 2.2261884331559982e-05,
+      "loss": 0.6601,
+      "step": 3006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.152556419372559,
+      "learning_rate": 2.2259310108117386e-05,
+      "loss": 1.1001,
+      "step": 3007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.843362808227539,
+      "learning_rate": 2.225673588467479e-05,
+      "loss": 1.2651,
+      "step": 3008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.01810359954834,
+      "learning_rate": 2.2254161661232195e-05,
+      "loss": 1.1865,
+      "step": 3009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.529011726379395,
+      "learning_rate": 2.22515874377896e-05,
+      "loss": 0.9467,
+      "step": 3010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.114763259887695,
+      "learning_rate": 2.2249013214347005e-05,
+      "loss": 0.8459,
+      "step": 3011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.628227233886719,
+      "learning_rate": 2.224643899090441e-05,
+      "loss": 1.033,
+      "step": 3012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.931683540344238,
+      "learning_rate": 2.224386476746182e-05,
+      "loss": 1.0218,
+      "step": 3013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.078429222106934,
+      "learning_rate": 2.2241290544019222e-05,
+      "loss": 1.13,
+      "step": 3014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.351057052612305,
+      "learning_rate": 2.223871632057663e-05,
+      "loss": 0.8109,
+      "step": 3015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.0247802734375,
+      "learning_rate": 2.2236142097134032e-05,
+      "loss": 1.0613,
+      "step": 3016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.838735580444336,
+      "learning_rate": 2.223356787369144e-05,
+      "loss": 1.2196,
+      "step": 3017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.077594757080078,
+      "learning_rate": 2.2230993650248842e-05,
+      "loss": 0.9813,
+      "step": 3018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.985511779785156,
+      "learning_rate": 2.2228419426806245e-05,
+      "loss": 1.0313,
+      "step": 3019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.833869934082031,
+      "learning_rate": 2.2225845203363652e-05,
+      "loss": 1.0435,
+      "step": 3020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.254064559936523,
+      "learning_rate": 2.2223270979921055e-05,
+      "loss": 1.0493,
+      "step": 3021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.48665714263916,
+      "learning_rate": 2.2220696756478462e-05,
+      "loss": 0.8993,
+      "step": 3022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.376240730285645,
+      "learning_rate": 2.221812253303587e-05,
+      "loss": 1.0113,
+      "step": 3023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.388966083526611,
+      "learning_rate": 2.2215548309593275e-05,
+      "loss": 0.6854,
+      "step": 3024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.38918399810791,
+      "learning_rate": 2.221297408615068e-05,
+      "loss": 0.946,
+      "step": 3025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.641005516052246,
+      "learning_rate": 2.2210399862708085e-05,
+      "loss": 1.0363,
+      "step": 3026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.025074005126953,
+      "learning_rate": 2.220782563926549e-05,
+      "loss": 0.8921,
+      "step": 3027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.211996078491211,
+      "learning_rate": 2.2205251415822892e-05,
+      "loss": 1.0066,
+      "step": 3028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.35977840423584,
+      "learning_rate": 2.22026771923803e-05,
+      "loss": 0.9677,
+      "step": 3029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.429361343383789,
+      "learning_rate": 2.2200102968937702e-05,
+      "loss": 1.0247,
+      "step": 3030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.093644142150879,
+      "learning_rate": 2.219752874549511e-05,
+      "loss": 0.8323,
+      "step": 3031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.507508277893066,
+      "learning_rate": 2.2194954522052515e-05,
+      "loss": 0.979,
+      "step": 3032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.718098640441895,
+      "learning_rate": 2.2192380298609922e-05,
+      "loss": 1.0606,
+      "step": 3033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.544224739074707,
+      "learning_rate": 2.2189806075167325e-05,
+      "loss": 0.9891,
+      "step": 3034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.190923690795898,
+      "learning_rate": 2.2187231851724732e-05,
+      "loss": 1.1908,
+      "step": 3035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.675702095031738,
+      "learning_rate": 2.2184657628282135e-05,
+      "loss": 0.9008,
+      "step": 3036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.517020225524902,
+      "learning_rate": 2.2182083404839542e-05,
+      "loss": 0.9227,
+      "step": 3037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.844659805297852,
+      "learning_rate": 2.2179509181396945e-05,
+      "loss": 1.1884,
+      "step": 3038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.5204596519470215,
+      "learning_rate": 2.217693495795435e-05,
+      "loss": 0.7349,
+      "step": 3039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.043553352355957,
+      "learning_rate": 2.2174360734511755e-05,
+      "loss": 1.1814,
+      "step": 3040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.549355506896973,
+      "learning_rate": 2.217178651106916e-05,
+      "loss": 1.0007,
+      "step": 3041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.698515892028809,
+      "learning_rate": 2.216921228762657e-05,
+      "loss": 0.8188,
+      "step": 3042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.914580345153809,
+      "learning_rate": 2.2166638064183972e-05,
+      "loss": 0.9664,
+      "step": 3043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.412646293640137,
+      "learning_rate": 2.216406384074138e-05,
+      "loss": 0.8891,
+      "step": 3044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.946002960205078,
+      "learning_rate": 2.2161489617298782e-05,
+      "loss": 1.3124,
+      "step": 3045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.516925811767578,
+      "learning_rate": 2.215891539385619e-05,
+      "loss": 0.9214,
+      "step": 3046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.866408348083496,
+      "learning_rate": 2.2156341170413592e-05,
+      "loss": 0.8427,
+      "step": 3047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.414539337158203,
+      "learning_rate": 2.2153766946971e-05,
+      "loss": 0.7524,
+      "step": 3048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.338263511657715,
+      "learning_rate": 2.2151192723528402e-05,
+      "loss": 0.8296,
+      "step": 3049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.56003475189209,
+      "learning_rate": 2.2148618500085805e-05,
+      "loss": 1.0169,
+      "step": 3050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.801880836486816,
+      "learning_rate": 2.2146044276643215e-05,
+      "loss": 0.9856,
+      "step": 3051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.434285163879395,
+      "learning_rate": 2.214347005320062e-05,
+      "loss": 1.2144,
+      "step": 3052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.876259803771973,
+      "learning_rate": 2.2140895829758025e-05,
+      "loss": 1.2534,
+      "step": 3053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.42397403717041,
+      "learning_rate": 2.213832160631543e-05,
+      "loss": 0.8078,
+      "step": 3054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.448081016540527,
+      "learning_rate": 2.2135747382872835e-05,
+      "loss": 0.8845,
+      "step": 3055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.044166564941406,
+      "learning_rate": 2.213317315943024e-05,
+      "loss": 0.9541,
+      "step": 3056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.51616382598877,
+      "learning_rate": 2.2130598935987645e-05,
+      "loss": 1.1335,
+      "step": 3057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.724225997924805,
+      "learning_rate": 2.212802471254505e-05,
+      "loss": 0.9727,
+      "step": 3058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.272114753723145,
+      "learning_rate": 2.2125450489102455e-05,
+      "loss": 0.9316,
+      "step": 3059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.649346351623535,
+      "learning_rate": 2.212287626565986e-05,
+      "loss": 0.8534,
+      "step": 3060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.92440414428711,
+      "learning_rate": 2.2120302042217265e-05,
+      "loss": 0.8505,
+      "step": 3061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.025346755981445,
+      "learning_rate": 2.211772781877467e-05,
+      "loss": 0.8607,
+      "step": 3062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.599574089050293,
+      "learning_rate": 2.2115153595332075e-05,
+      "loss": 0.8316,
+      "step": 3063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.571738243103027,
+      "learning_rate": 2.211257937188948e-05,
+      "loss": 0.9319,
+      "step": 3064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.115510940551758,
+      "learning_rate": 2.2110005148446885e-05,
+      "loss": 0.7957,
+      "step": 3065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.370075225830078,
+      "learning_rate": 2.210743092500429e-05,
+      "loss": 0.7935,
+      "step": 3066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.367995262145996,
+      "learning_rate": 2.2104856701561695e-05,
+      "loss": 0.7663,
+      "step": 3067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.3759126663208,
+      "learning_rate": 2.21022824781191e-05,
+      "loss": 0.8605,
+      "step": 3068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.688425064086914,
+      "learning_rate": 2.2099708254676505e-05,
+      "loss": 0.8383,
+      "step": 3069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.046122550964355,
+      "learning_rate": 2.209713403123391e-05,
+      "loss": 1.0199,
+      "step": 3070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.270439147949219,
+      "learning_rate": 2.2094559807791318e-05,
+      "loss": 1.112,
+      "step": 3071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.894763946533203,
+      "learning_rate": 2.209198558434872e-05,
+      "loss": 0.8632,
+      "step": 3072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.473116874694824,
+      "learning_rate": 2.2089411360906128e-05,
+      "loss": 1.1591,
+      "step": 3073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.77759075164795,
+      "learning_rate": 2.208683713746353e-05,
+      "loss": 1.1484,
+      "step": 3074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.56604290008545,
+      "learning_rate": 2.2084262914020938e-05,
+      "loss": 1.3491,
+      "step": 3075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.549914360046387,
+      "learning_rate": 2.208168869057834e-05,
+      "loss": 0.894,
+      "step": 3076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.53173828125,
+      "learning_rate": 2.2079114467135748e-05,
+      "loss": 1.1102,
+      "step": 3077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.941644668579102,
+      "learning_rate": 2.207654024369315e-05,
+      "loss": 1.0642,
+      "step": 3078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.05459976196289,
+      "learning_rate": 2.2073966020250558e-05,
+      "loss": 0.8512,
+      "step": 3079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.281445503234863,
+      "learning_rate": 2.2071391796807965e-05,
+      "loss": 1.159,
+      "step": 3080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.477843284606934,
+      "learning_rate": 2.2068817573365368e-05,
+      "loss": 0.8448,
+      "step": 3081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.021671295166016,
+      "learning_rate": 2.2066243349922775e-05,
+      "loss": 0.7585,
+      "step": 3082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.576374053955078,
+      "learning_rate": 2.2063669126480178e-05,
+      "loss": 0.9537,
+      "step": 3083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.444153308868408,
+      "learning_rate": 2.2061094903037585e-05,
+      "loss": 0.9178,
+      "step": 3084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.676690101623535,
+      "learning_rate": 2.2058520679594988e-05,
+      "loss": 1.1081,
+      "step": 3085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.60732650756836,
+      "learning_rate": 2.2055946456152395e-05,
+      "loss": 1.2603,
+      "step": 3086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.08698844909668,
+      "learning_rate": 2.2053372232709798e-05,
+      "loss": 0.9504,
+      "step": 3087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.596535682678223,
+      "learning_rate": 2.2050798009267205e-05,
+      "loss": 1.0693,
+      "step": 3088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.122248649597168,
+      "learning_rate": 2.204822378582461e-05,
+      "loss": 1.0534,
+      "step": 3089
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.22197437286377,
+      "learning_rate": 2.2045649562382018e-05,
+      "loss": 0.8598,
+      "step": 3090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.440376281738281,
+      "learning_rate": 2.204307533893942e-05,
+      "loss": 1.1198,
+      "step": 3091
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.762871742248535,
+      "learning_rate": 2.2040501115496825e-05,
+      "loss": 1.0391,
+      "step": 3092
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.070852279663086,
+      "learning_rate": 2.203792689205423e-05,
+      "loss": 1.0754,
+      "step": 3093
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.561709403991699,
+      "learning_rate": 2.2035352668611635e-05,
+      "loss": 0.7504,
+      "step": 3094
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.179771423339844,
+      "learning_rate": 2.203277844516904e-05,
+      "loss": 0.8929,
+      "step": 3095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.326729774475098,
+      "learning_rate": 2.2030204221726445e-05,
+      "loss": 1.0587,
+      "step": 3096
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.242788314819336,
+      "learning_rate": 2.202762999828385e-05,
+      "loss": 1.0423,
+      "step": 3097
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.7095365524292,
+      "learning_rate": 2.2025055774841258e-05,
+      "loss": 0.9249,
+      "step": 3098
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.872523307800293,
+      "learning_rate": 2.2022481551398665e-05,
+      "loss": 1.2675,
+      "step": 3099
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.039074897766113,
+      "learning_rate": 2.2019907327956068e-05,
+      "loss": 0.9968,
+      "step": 3100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.447712898254395,
+      "learning_rate": 2.201733310451347e-05,
+      "loss": 0.9513,
+      "step": 3101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.730088233947754,
+      "learning_rate": 2.2014758881070878e-05,
+      "loss": 1.0489,
+      "step": 3102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.814104080200195,
+      "learning_rate": 2.201218465762828e-05,
+      "loss": 1.0052,
+      "step": 3103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.879773139953613,
+      "learning_rate": 2.2009610434185688e-05,
+      "loss": 0.799,
+      "step": 3104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.684769630432129,
+      "learning_rate": 2.200703621074309e-05,
+      "loss": 0.8019,
+      "step": 3105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.733920097351074,
+      "learning_rate": 2.2004461987300498e-05,
+      "loss": 0.8762,
+      "step": 3106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.692930221557617,
+      "learning_rate": 2.20018877638579e-05,
+      "loss": 1.3131,
+      "step": 3107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.138765335083008,
+      "learning_rate": 2.199931354041531e-05,
+      "loss": 0.8082,
+      "step": 3108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.816980361938477,
+      "learning_rate": 2.1996739316972715e-05,
+      "loss": 1.0204,
+      "step": 3109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.869331359863281,
+      "learning_rate": 2.199416509353012e-05,
+      "loss": 0.952,
+      "step": 3110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.218375205993652,
+      "learning_rate": 2.1991590870087525e-05,
+      "loss": 1.1501,
+      "step": 3111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.711645126342773,
+      "learning_rate": 2.1989016646644928e-05,
+      "loss": 1.1254,
+      "step": 3112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.05109977722168,
+      "learning_rate": 2.1986442423202335e-05,
+      "loss": 1.2439,
+      "step": 3113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.715018272399902,
+      "learning_rate": 2.1983868199759738e-05,
+      "loss": 0.9996,
+      "step": 3114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.492490768432617,
+      "learning_rate": 2.1981293976317144e-05,
+      "loss": 0.8824,
+      "step": 3115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.98580265045166,
+      "learning_rate": 2.1978719752874548e-05,
+      "loss": 0.9619,
+      "step": 3116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.428837776184082,
+      "learning_rate": 2.1976145529431958e-05,
+      "loss": 0.8728,
+      "step": 3117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.138510704040527,
+      "learning_rate": 2.197357130598936e-05,
+      "loss": 1.1656,
+      "step": 3118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.214956283569336,
+      "learning_rate": 2.1970997082546768e-05,
+      "loss": 0.8116,
+      "step": 3119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.222757339477539,
+      "learning_rate": 2.196842285910417e-05,
+      "loss": 0.9675,
+      "step": 3120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.356564998626709,
+      "learning_rate": 2.1965848635661578e-05,
+      "loss": 0.8123,
+      "step": 3121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.394486427307129,
+      "learning_rate": 2.196327441221898e-05,
+      "loss": 1.2035,
+      "step": 3122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.59044361114502,
+      "learning_rate": 2.1960700188776384e-05,
+      "loss": 1.2356,
+      "step": 3123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.820239067077637,
+      "learning_rate": 2.195812596533379e-05,
+      "loss": 1.1329,
+      "step": 3124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.050148010253906,
+      "learning_rate": 2.1955551741891194e-05,
+      "loss": 1.2123,
+      "step": 3125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.765544891357422,
+      "learning_rate": 2.19529775184486e-05,
+      "loss": 1.2961,
+      "step": 3126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.465819358825684,
+      "learning_rate": 2.1950403295006008e-05,
+      "loss": 0.8429,
+      "step": 3127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.164430618286133,
+      "learning_rate": 2.1947829071563414e-05,
+      "loss": 0.9639,
+      "step": 3128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.604117393493652,
+      "learning_rate": 2.1945254848120818e-05,
+      "loss": 1.1961,
+      "step": 3129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.781293869018555,
+      "learning_rate": 2.1942680624678224e-05,
+      "loss": 1.1247,
+      "step": 3130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.708008766174316,
+      "learning_rate": 2.1940106401235628e-05,
+      "loss": 0.839,
+      "step": 3131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.764863967895508,
+      "learning_rate": 2.1937532177793034e-05,
+      "loss": 0.9776,
+      "step": 3132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.481902122497559,
+      "learning_rate": 2.1934957954350438e-05,
+      "loss": 0.9819,
+      "step": 3133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.671924591064453,
+      "learning_rate": 2.193238373090784e-05,
+      "loss": 0.8807,
+      "step": 3134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.681611061096191,
+      "learning_rate": 2.1929809507465248e-05,
+      "loss": 0.8351,
+      "step": 3135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.935581684112549,
+      "learning_rate": 2.1927235284022654e-05,
+      "loss": 0.7053,
+      "step": 3136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.76996898651123,
+      "learning_rate": 2.192466106058006e-05,
+      "loss": 1.1438,
+      "step": 3137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.040933609008789,
+      "learning_rate": 2.1922086837137464e-05,
+      "loss": 1.1028,
+      "step": 3138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.696728706359863,
+      "learning_rate": 2.191951261369487e-05,
+      "loss": 1.2261,
+      "step": 3139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.489141464233398,
+      "learning_rate": 2.1916938390252274e-05,
+      "loss": 1.1004,
+      "step": 3140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.115700721740723,
+      "learning_rate": 2.191436416680968e-05,
+      "loss": 1.0205,
+      "step": 3141
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.014881134033203,
+      "learning_rate": 2.1911789943367084e-05,
+      "loss": 1.0217,
+      "step": 3142
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.941526889801025,
+      "learning_rate": 2.1909215719924488e-05,
+      "loss": 0.7474,
+      "step": 3143
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.077324867248535,
+      "learning_rate": 2.1906641496481894e-05,
+      "loss": 0.7121,
+      "step": 3144
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.537542343139648,
+      "learning_rate": 2.1904067273039297e-05,
+      "loss": 1.0168,
+      "step": 3145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.554915428161621,
+      "learning_rate": 2.1901493049596708e-05,
+      "loss": 0.97,
+      "step": 3146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.603519439697266,
+      "learning_rate": 2.189891882615411e-05,
+      "loss": 0.9573,
+      "step": 3147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.304115295410156,
+      "learning_rate": 2.1896344602711518e-05,
+      "loss": 1.0942,
+      "step": 3148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.811833381652832,
+      "learning_rate": 2.189377037926892e-05,
+      "loss": 0.897,
+      "step": 3149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.509538650512695,
+      "learning_rate": 2.1891196155826328e-05,
+      "loss": 0.6555,
+      "step": 3150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.038629531860352,
+      "learning_rate": 2.188862193238373e-05,
+      "loss": 1.0459,
+      "step": 3151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.320324897766113,
+      "learning_rate": 2.1886047708941137e-05,
+      "loss": 0.9769,
+      "step": 3152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.822304725646973,
+      "learning_rate": 2.188347348549854e-05,
+      "loss": 0.7505,
+      "step": 3153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.76353645324707,
+      "learning_rate": 2.1880899262055944e-05,
+      "loss": 1.0906,
+      "step": 3154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.874078750610352,
+      "learning_rate": 2.1878325038613354e-05,
+      "loss": 1.54,
+      "step": 3155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.56296443939209,
+      "learning_rate": 2.1875750815170757e-05,
+      "loss": 1.1177,
+      "step": 3156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.449142456054688,
+      "learning_rate": 2.1873176591728164e-05,
+      "loss": 0.9417,
+      "step": 3157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.855195999145508,
+      "learning_rate": 2.1870602368285567e-05,
+      "loss": 1.0936,
+      "step": 3158
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.562729358673096,
+      "learning_rate": 2.1868028144842974e-05,
+      "loss": 0.7096,
+      "step": 3159
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.778154373168945,
+      "learning_rate": 2.1865453921400377e-05,
+      "loss": 0.9019,
+      "step": 3160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.041037559509277,
+      "learning_rate": 2.1862879697957784e-05,
+      "loss": 0.962,
+      "step": 3161
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.064740180969238,
+      "learning_rate": 2.1860305474515187e-05,
+      "loss": 1.0679,
+      "step": 3162
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.525568008422852,
+      "learning_rate": 2.1857731251072594e-05,
+      "loss": 1.0027,
+      "step": 3163
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.927902221679688,
+      "learning_rate": 2.1855157027629997e-05,
+      "loss": 0.9911,
+      "step": 3164
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.060417175292969,
+      "learning_rate": 2.1852582804187404e-05,
+      "loss": 1.1202,
+      "step": 3165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.134319305419922,
+      "learning_rate": 2.185000858074481e-05,
+      "loss": 0.9622,
+      "step": 3166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.938206672668457,
+      "learning_rate": 2.1847434357302214e-05,
+      "loss": 0.9234,
+      "step": 3167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.985689163208008,
+      "learning_rate": 2.184486013385962e-05,
+      "loss": 1.2295,
+      "step": 3168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.237380981445312,
+      "learning_rate": 2.1842285910417024e-05,
+      "loss": 0.9057,
+      "step": 3169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.169559478759766,
+      "learning_rate": 2.183971168697443e-05,
+      "loss": 0.6954,
+      "step": 3170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.447514533996582,
+      "learning_rate": 2.1837137463531834e-05,
+      "loss": 1.0324,
+      "step": 3171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.671792984008789,
+      "learning_rate": 2.183456324008924e-05,
+      "loss": 0.8003,
+      "step": 3172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.219077110290527,
+      "learning_rate": 2.1831989016646644e-05,
+      "loss": 0.6495,
+      "step": 3173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.128629684448242,
+      "learning_rate": 2.182941479320405e-05,
+      "loss": 0.7984,
+      "step": 3174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.862685680389404,
+      "learning_rate": 2.1826840569761457e-05,
+      "loss": 0.791,
+      "step": 3175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.163856506347656,
+      "learning_rate": 2.182426634631886e-05,
+      "loss": 0.8417,
+      "step": 3176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.783720016479492,
+      "learning_rate": 2.1821692122876267e-05,
+      "loss": 1.0774,
+      "step": 3177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.313047409057617,
+      "learning_rate": 2.181911789943367e-05,
+      "loss": 0.8129,
+      "step": 3178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.134902000427246,
+      "learning_rate": 2.1816543675991077e-05,
+      "loss": 0.7819,
+      "step": 3179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.417607307434082,
+      "learning_rate": 2.181396945254848e-05,
+      "loss": 1.0525,
+      "step": 3180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.567964553833008,
+      "learning_rate": 2.1811395229105887e-05,
+      "loss": 0.7269,
+      "step": 3181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.554522514343262,
+      "learning_rate": 2.180882100566329e-05,
+      "loss": 0.7956,
+      "step": 3182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.361642837524414,
+      "learning_rate": 2.1806246782220697e-05,
+      "loss": 0.7218,
+      "step": 3183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.764366149902344,
+      "learning_rate": 2.1803672558778104e-05,
+      "loss": 1.27,
+      "step": 3184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.85014533996582,
+      "learning_rate": 2.1801098335335507e-05,
+      "loss": 0.9879,
+      "step": 3185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.578740119934082,
+      "learning_rate": 2.1798524111892914e-05,
+      "loss": 0.884,
+      "step": 3186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.373249053955078,
+      "learning_rate": 2.1795949888450317e-05,
+      "loss": 0.8658,
+      "step": 3187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.385355949401855,
+      "learning_rate": 2.1793375665007724e-05,
+      "loss": 1.0733,
+      "step": 3188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.13532829284668,
+      "learning_rate": 2.1790801441565127e-05,
+      "loss": 1.1228,
+      "step": 3189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.906203269958496,
+      "learning_rate": 2.1788227218122534e-05,
+      "loss": 1.0137,
+      "step": 3190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.440940856933594,
+      "learning_rate": 2.1785652994679937e-05,
+      "loss": 0.9276,
+      "step": 3191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.625938415527344,
+      "learning_rate": 2.1783078771237344e-05,
+      "loss": 1.3305,
+      "step": 3192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.186883926391602,
+      "learning_rate": 2.178050454779475e-05,
+      "loss": 0.9193,
+      "step": 3193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.901379585266113,
+      "learning_rate": 2.1777930324352157e-05,
+      "loss": 0.9429,
+      "step": 3194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.108570098876953,
+      "learning_rate": 2.177535610090956e-05,
+      "loss": 0.8455,
+      "step": 3195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.209932327270508,
+      "learning_rate": 2.1772781877466964e-05,
+      "loss": 0.8453,
+      "step": 3196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.383414268493652,
+      "learning_rate": 2.177020765402437e-05,
+      "loss": 0.847,
+      "step": 3197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.280423164367676,
+      "learning_rate": 2.1767633430581774e-05,
+      "loss": 0.9467,
+      "step": 3198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.35180377960205,
+      "learning_rate": 2.176505920713918e-05,
+      "loss": 1.2009,
+      "step": 3199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.709630966186523,
+      "learning_rate": 2.1762484983696584e-05,
+      "loss": 1.105,
+      "step": 3200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.566025733947754,
+      "learning_rate": 2.175991076025399e-05,
+      "loss": 0.919,
+      "step": 3201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.236313819885254,
+      "learning_rate": 2.1757336536811394e-05,
+      "loss": 1.142,
+      "step": 3202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.696102142333984,
+      "learning_rate": 2.1754762313368804e-05,
+      "loss": 0.8391,
+      "step": 3203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.705658912658691,
+      "learning_rate": 2.1752188089926207e-05,
+      "loss": 1.3478,
+      "step": 3204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.247570991516113,
+      "learning_rate": 2.174961386648361e-05,
+      "loss": 0.6832,
+      "step": 3205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.861929893493652,
+      "learning_rate": 2.1747039643041017e-05,
+      "loss": 0.6828,
+      "step": 3206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.63481616973877,
+      "learning_rate": 2.174446541959842e-05,
+      "loss": 1.0992,
+      "step": 3207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.865202903747559,
+      "learning_rate": 2.1741891196155827e-05,
+      "loss": 0.9223,
+      "step": 3208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.262078285217285,
+      "learning_rate": 2.173931697271323e-05,
+      "loss": 0.9563,
+      "step": 3209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.30843734741211,
+      "learning_rate": 2.1736742749270637e-05,
+      "loss": 1.1408,
+      "step": 3210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.31630802154541,
+      "learning_rate": 2.173416852582804e-05,
+      "loss": 0.8361,
+      "step": 3211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.973668575286865,
+      "learning_rate": 2.173159430238545e-05,
+      "loss": 0.8109,
+      "step": 3212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.872391700744629,
+      "learning_rate": 2.1729020078942854e-05,
+      "loss": 0.9446,
+      "step": 3213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.826184272766113,
+      "learning_rate": 2.172644585550026e-05,
+      "loss": 0.73,
+      "step": 3214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.253357887268066,
+      "learning_rate": 2.1723871632057664e-05,
+      "loss": 1.113,
+      "step": 3215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.210947036743164,
+      "learning_rate": 2.1721297408615067e-05,
+      "loss": 1.0653,
+      "step": 3216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.294479370117188,
+      "learning_rate": 2.1718723185172474e-05,
+      "loss": 0.8931,
+      "step": 3217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.98333740234375,
+      "learning_rate": 2.1716148961729877e-05,
+      "loss": 0.8752,
+      "step": 3218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.869829177856445,
+      "learning_rate": 2.1713574738287283e-05,
+      "loss": 1.0788,
+      "step": 3219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.192008018493652,
+      "learning_rate": 2.1711000514844687e-05,
+      "loss": 0.783,
+      "step": 3220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.066743850708008,
+      "learning_rate": 2.1708426291402097e-05,
+      "loss": 1.1949,
+      "step": 3221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.277292251586914,
+      "learning_rate": 2.17058520679595e-05,
+      "loss": 0.9494,
+      "step": 3222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.366868019104004,
+      "learning_rate": 2.1703277844516907e-05,
+      "loss": 1.0176,
+      "step": 3223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.33558464050293,
+      "learning_rate": 2.170070362107431e-05,
+      "loss": 0.7573,
+      "step": 3224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.411799430847168,
+      "learning_rate": 2.1698129397631717e-05,
+      "loss": 0.7129,
+      "step": 3225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.274582862854004,
+      "learning_rate": 2.169555517418912e-05,
+      "loss": 1.0729,
+      "step": 3226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.059366226196289,
+      "learning_rate": 2.1692980950746523e-05,
+      "loss": 0.9163,
+      "step": 3227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.27176570892334,
+      "learning_rate": 2.169040672730393e-05,
+      "loss": 1.4554,
+      "step": 3228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.528203010559082,
+      "learning_rate": 2.1687832503861333e-05,
+      "loss": 0.8814,
+      "step": 3229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.73957347869873,
+      "learning_rate": 2.168525828041874e-05,
+      "loss": 0.9085,
+      "step": 3230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.321489334106445,
+      "learning_rate": 2.1682684056976147e-05,
+      "loss": 0.9523,
+      "step": 3231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.32539176940918,
+      "learning_rate": 2.1680109833533553e-05,
+      "loss": 0.6604,
+      "step": 3232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.537572860717773,
+      "learning_rate": 2.1677535610090957e-05,
+      "loss": 1.0714,
+      "step": 3233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.214726448059082,
+      "learning_rate": 2.1674961386648363e-05,
+      "loss": 0.8779,
+      "step": 3234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.187397003173828,
+      "learning_rate": 2.1672387163205767e-05,
+      "loss": 1.4253,
+      "step": 3235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.91473388671875,
+      "learning_rate": 2.1669812939763173e-05,
+      "loss": 0.8857,
+      "step": 3236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.964056015014648,
+      "learning_rate": 2.1667238716320577e-05,
+      "loss": 0.8446,
+      "step": 3237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.088868141174316,
+      "learning_rate": 2.166466449287798e-05,
+      "loss": 0.9246,
+      "step": 3238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.25182819366455,
+      "learning_rate": 2.1662090269435387e-05,
+      "loss": 0.983,
+      "step": 3239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.780190467834473,
+      "learning_rate": 2.1659516045992793e-05,
+      "loss": 0.9359,
+      "step": 3240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.661764144897461,
+      "learning_rate": 2.16569418225502e-05,
+      "loss": 0.8701,
+      "step": 3241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.114421844482422,
+      "learning_rate": 2.1654367599107603e-05,
+      "loss": 1.2242,
+      "step": 3242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.93411636352539,
+      "learning_rate": 2.165179337566501e-05,
+      "loss": 1.0718,
+      "step": 3243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.305581092834473,
+      "learning_rate": 2.1649219152222413e-05,
+      "loss": 1.0008,
+      "step": 3244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.578404903411865,
+      "learning_rate": 2.164664492877982e-05,
+      "loss": 0.7234,
+      "step": 3245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.716707229614258,
+      "learning_rate": 2.1644070705337223e-05,
+      "loss": 0.7725,
+      "step": 3246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.51779556274414,
+      "learning_rate": 2.1641496481894627e-05,
+      "loss": 1.0431,
+      "step": 3247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.817296028137207,
+      "learning_rate": 2.1638922258452033e-05,
+      "loss": 1.0968,
+      "step": 3248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.080244064331055,
+      "learning_rate": 2.1636348035009437e-05,
+      "loss": 0.8871,
+      "step": 3249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.9187650680542,
+      "learning_rate": 2.1633773811566847e-05,
+      "loss": 1.3985,
+      "step": 3250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.561849594116211,
+      "learning_rate": 2.163119958812425e-05,
+      "loss": 1.1111,
+      "step": 3251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.092209815979004,
+      "learning_rate": 2.1628625364681657e-05,
+      "loss": 0.9586,
+      "step": 3252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.601984977722168,
+      "learning_rate": 2.162605114123906e-05,
+      "loss": 0.9554,
+      "step": 3253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.11715030670166,
+      "learning_rate": 2.1623476917796467e-05,
+      "loss": 0.8711,
+      "step": 3254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.338886260986328,
+      "learning_rate": 2.162090269435387e-05,
+      "loss": 1.079,
+      "step": 3255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.088749885559082,
+      "learning_rate": 2.1618328470911276e-05,
+      "loss": 1.0061,
+      "step": 3256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.813129901885986,
+      "learning_rate": 2.161575424746868e-05,
+      "loss": 0.9419,
+      "step": 3257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.794445037841797,
+      "learning_rate": 2.1613180024026083e-05,
+      "loss": 1.1162,
+      "step": 3258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.691902160644531,
+      "learning_rate": 2.1610605800583493e-05,
+      "loss": 0.7736,
+      "step": 3259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.716875076293945,
+      "learning_rate": 2.1608031577140896e-05,
+      "loss": 0.6622,
+      "step": 3260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.189523696899414,
+      "learning_rate": 2.1605457353698303e-05,
+      "loss": 1.0522,
+      "step": 3261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.608463287353516,
+      "learning_rate": 2.1602883130255706e-05,
+      "loss": 0.7976,
+      "step": 3262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.183985710144043,
+      "learning_rate": 2.1600308906813113e-05,
+      "loss": 1.0986,
+      "step": 3263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.639912605285645,
+      "learning_rate": 2.1597734683370516e-05,
+      "loss": 0.6988,
+      "step": 3264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.809829711914062,
+      "learning_rate": 2.1595160459927923e-05,
+      "loss": 1.5204,
+      "step": 3265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.907243728637695,
+      "learning_rate": 2.1592586236485326e-05,
+      "loss": 0.7407,
+      "step": 3266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.971426963806152,
+      "learning_rate": 2.1590012013042733e-05,
+      "loss": 1.0083,
+      "step": 3267
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.05289363861084,
+      "learning_rate": 2.1587437789600136e-05,
+      "loss": 0.8034,
+      "step": 3268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.950414657592773,
+      "learning_rate": 2.1584863566157543e-05,
+      "loss": 0.8737,
+      "step": 3269
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.235519409179688,
+      "learning_rate": 2.158228934271495e-05,
+      "loss": 0.7837,
+      "step": 3270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.354198455810547,
+      "learning_rate": 2.1579715119272353e-05,
+      "loss": 0.9803,
+      "step": 3271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.11938762664795,
+      "learning_rate": 2.157714089582976e-05,
+      "loss": 0.708,
+      "step": 3272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.488126754760742,
+      "learning_rate": 2.1574566672387163e-05,
+      "loss": 1.0758,
+      "step": 3273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.257621765136719,
+      "learning_rate": 2.157199244894457e-05,
+      "loss": 0.9413,
+      "step": 3274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.837946891784668,
+      "learning_rate": 2.1569418225501973e-05,
+      "loss": 0.8158,
+      "step": 3275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.066544532775879,
+      "learning_rate": 2.156684400205938e-05,
+      "loss": 1.1956,
+      "step": 3276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.441474914550781,
+      "learning_rate": 2.1564269778616783e-05,
+      "loss": 0.8086,
+      "step": 3277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.52379322052002,
+      "learning_rate": 2.156169555517419e-05,
+      "loss": 0.6603,
+      "step": 3278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.316561698913574,
+      "learning_rate": 2.1559121331731596e-05,
+      "loss": 0.8135,
+      "step": 3279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.021976470947266,
+      "learning_rate": 2.1556547108289e-05,
+      "loss": 0.8317,
+      "step": 3280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.18840503692627,
+      "learning_rate": 2.1553972884846406e-05,
+      "loss": 1.0824,
+      "step": 3281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.439772605895996,
+      "learning_rate": 2.155139866140381e-05,
+      "loss": 0.8689,
+      "step": 3282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.884522438049316,
+      "learning_rate": 2.1548824437961216e-05,
+      "loss": 0.9889,
+      "step": 3283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.708690643310547,
+      "learning_rate": 2.154625021451862e-05,
+      "loss": 0.9767,
+      "step": 3284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.476131439208984,
+      "learning_rate": 2.1543675991076026e-05,
+      "loss": 1.0251,
+      "step": 3285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.36382007598877,
+      "learning_rate": 2.154110176763343e-05,
+      "loss": 0.7405,
+      "step": 3286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.07170295715332,
+      "learning_rate": 2.1538527544190836e-05,
+      "loss": 0.9733,
+      "step": 3287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.972197532653809,
+      "learning_rate": 2.1535953320748243e-05,
+      "loss": 0.9596,
+      "step": 3288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.184207916259766,
+      "learning_rate": 2.1533379097305646e-05,
+      "loss": 0.7975,
+      "step": 3289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.137417793273926,
+      "learning_rate": 2.1530804873863053e-05,
+      "loss": 0.7493,
+      "step": 3290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.522597312927246,
+      "learning_rate": 2.1528230650420456e-05,
+      "loss": 1.1567,
+      "step": 3291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.457935333251953,
+      "learning_rate": 2.1525656426977863e-05,
+      "loss": 0.8999,
+      "step": 3292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.231308937072754,
+      "learning_rate": 2.1523082203535266e-05,
+      "loss": 1.1664,
+      "step": 3293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.125234603881836,
+      "learning_rate": 2.1520507980092673e-05,
+      "loss": 0.8378,
+      "step": 3294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.983253479003906,
+      "learning_rate": 2.1517933756650076e-05,
+      "loss": 0.8401,
+      "step": 3295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.115766525268555,
+      "learning_rate": 2.1515359533207483e-05,
+      "loss": 1.0702,
+      "step": 3296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.25951862335205,
+      "learning_rate": 2.151278530976489e-05,
+      "loss": 1.1353,
+      "step": 3297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.079184532165527,
+      "learning_rate": 2.1510211086322296e-05,
+      "loss": 0.9575,
+      "step": 3298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.612562656402588,
+      "learning_rate": 2.15076368628797e-05,
+      "loss": 0.9446,
+      "step": 3299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.672316551208496,
+      "learning_rate": 2.1505062639437103e-05,
+      "loss": 0.9791,
+      "step": 3300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.614047527313232,
+      "learning_rate": 2.150248841599451e-05,
+      "loss": 0.846,
+      "step": 3301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.674379348754883,
+      "learning_rate": 2.1499914192551913e-05,
+      "loss": 1.1468,
+      "step": 3302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.892300605773926,
+      "learning_rate": 2.149733996910932e-05,
+      "loss": 1.0865,
+      "step": 3303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.093138694763184,
+      "learning_rate": 2.1494765745666723e-05,
+      "loss": 1.0762,
+      "step": 3304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.564008712768555,
+      "learning_rate": 2.149219152222413e-05,
+      "loss": 0.9487,
+      "step": 3305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.3695707321167,
+      "learning_rate": 2.1489617298781533e-05,
+      "loss": 0.8785,
+      "step": 3306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.876216888427734,
+      "learning_rate": 2.1487043075338943e-05,
+      "loss": 0.9495,
+      "step": 3307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.334044456481934,
+      "learning_rate": 2.1484468851896346e-05,
+      "loss": 1.2583,
+      "step": 3308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.992374420166016,
+      "learning_rate": 2.148189462845375e-05,
+      "loss": 0.8078,
+      "step": 3309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.051708221435547,
+      "learning_rate": 2.1479320405011156e-05,
+      "loss": 0.8927,
+      "step": 3310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.372456550598145,
+      "learning_rate": 2.147674618156856e-05,
+      "loss": 0.8778,
+      "step": 3311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.7460761070251465,
+      "learning_rate": 2.1474171958125966e-05,
+      "loss": 0.9121,
+      "step": 3312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.750205993652344,
+      "learning_rate": 2.147159773468337e-05,
+      "loss": 0.8885,
+      "step": 3313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.668148040771484,
+      "learning_rate": 2.1469023511240776e-05,
+      "loss": 0.8839,
+      "step": 3314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.011385917663574,
+      "learning_rate": 2.146644928779818e-05,
+      "loss": 1.0066,
+      "step": 3315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.701638221740723,
+      "learning_rate": 2.146387506435559e-05,
+      "loss": 1.3349,
+      "step": 3316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.443819999694824,
+      "learning_rate": 2.1461300840912993e-05,
+      "loss": 1.265,
+      "step": 3317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.117536544799805,
+      "learning_rate": 2.14587266174704e-05,
+      "loss": 0.6747,
+      "step": 3318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.16749382019043,
+      "learning_rate": 2.1456152394027803e-05,
+      "loss": 1.0698,
+      "step": 3319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.960887908935547,
+      "learning_rate": 2.1453578170585206e-05,
+      "loss": 0.928,
+      "step": 3320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.629446983337402,
+      "learning_rate": 2.1451003947142613e-05,
+      "loss": 0.8203,
+      "step": 3321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.186299324035645,
+      "learning_rate": 2.1448429723700016e-05,
+      "loss": 0.8447,
+      "step": 3322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.769874572753906,
+      "learning_rate": 2.1445855500257423e-05,
+      "loss": 0.6914,
+      "step": 3323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.716658592224121,
+      "learning_rate": 2.1443281276814826e-05,
+      "loss": 0.7656,
+      "step": 3324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.508240699768066,
+      "learning_rate": 2.1440707053372236e-05,
+      "loss": 0.9904,
+      "step": 3325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.989377975463867,
+      "learning_rate": 2.143813282992964e-05,
+      "loss": 1.0246,
+      "step": 3326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.998805046081543,
+      "learning_rate": 2.1435558606487046e-05,
+      "loss": 0.8177,
+      "step": 3327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.800924301147461,
+      "learning_rate": 2.143298438304445e-05,
+      "loss": 1.2029,
+      "step": 3328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.99470043182373,
+      "learning_rate": 2.1430410159601856e-05,
+      "loss": 0.8444,
+      "step": 3329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.246833801269531,
+      "learning_rate": 2.142783593615926e-05,
+      "loss": 0.8711,
+      "step": 3330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.696725845336914,
+      "learning_rate": 2.1425261712716662e-05,
+      "loss": 0.8906,
+      "step": 3331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.879152297973633,
+      "learning_rate": 2.142268748927407e-05,
+      "loss": 1.0487,
+      "step": 3332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.030535697937012,
+      "learning_rate": 2.1420113265831472e-05,
+      "loss": 0.6264,
+      "step": 3333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.199321746826172,
+      "learning_rate": 2.141753904238888e-05,
+      "loss": 0.91,
+      "step": 3334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.858607292175293,
+      "learning_rate": 2.1414964818946286e-05,
+      "loss": 1.1586,
+      "step": 3335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.799261093139648,
+      "learning_rate": 2.1412390595503692e-05,
+      "loss": 1.061,
+      "step": 3336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.063508033752441,
+      "learning_rate": 2.1409816372061096e-05,
+      "loss": 0.9201,
+      "step": 3337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.47541618347168,
+      "learning_rate": 2.1407242148618502e-05,
+      "loss": 1.0478,
+      "step": 3338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.947546005249023,
+      "learning_rate": 2.1404667925175906e-05,
+      "loss": 1.1379,
+      "step": 3339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.848004341125488,
+      "learning_rate": 2.1402093701733312e-05,
+      "loss": 0.9509,
+      "step": 3340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.302692413330078,
+      "learning_rate": 2.1399519478290716e-05,
+      "loss": 0.8064,
+      "step": 3341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.407553672790527,
+      "learning_rate": 2.139694525484812e-05,
+      "loss": 0.9112,
+      "step": 3342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.795711517333984,
+      "learning_rate": 2.1394371031405526e-05,
+      "loss": 0.7802,
+      "step": 3343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.757826805114746,
+      "learning_rate": 2.1391796807962932e-05,
+      "loss": 1.0588,
+      "step": 3344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.820745468139648,
+      "learning_rate": 2.138922258452034e-05,
+      "loss": 0.8998,
+      "step": 3345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.344983100891113,
+      "learning_rate": 2.1386648361077742e-05,
+      "loss": 1.1015,
+      "step": 3346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.325364112854004,
+      "learning_rate": 2.138407413763515e-05,
+      "loss": 0.8903,
+      "step": 3347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.416016578674316,
+      "learning_rate": 2.1381499914192552e-05,
+      "loss": 1.2725,
+      "step": 3348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.414058685302734,
+      "learning_rate": 2.137892569074996e-05,
+      "loss": 0.9045,
+      "step": 3349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.113632202148438,
+      "learning_rate": 2.1376351467307362e-05,
+      "loss": 0.8869,
+      "step": 3350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.85736083984375,
+      "learning_rate": 2.1373777243864766e-05,
+      "loss": 0.9481,
+      "step": 3351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.1739501953125,
+      "learning_rate": 2.1371203020422172e-05,
+      "loss": 0.9657,
+      "step": 3352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.000965118408203,
+      "learning_rate": 2.1368628796979576e-05,
+      "loss": 0.771,
+      "step": 3353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.271477699279785,
+      "learning_rate": 2.1366054573536986e-05,
+      "loss": 0.936,
+      "step": 3354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.239330291748047,
+      "learning_rate": 2.136348035009439e-05,
+      "loss": 1.0011,
+      "step": 3355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.274884223937988,
+      "learning_rate": 2.1360906126651796e-05,
+      "loss": 1.1459,
+      "step": 3356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.972211837768555,
+      "learning_rate": 2.13583319032092e-05,
+      "loss": 0.8815,
+      "step": 3357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.910994529724121,
+      "learning_rate": 2.1355757679766606e-05,
+      "loss": 0.9931,
+      "step": 3358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.614021301269531,
+      "learning_rate": 2.135318345632401e-05,
+      "loss": 0.9974,
+      "step": 3359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.335184097290039,
+      "learning_rate": 2.1350609232881416e-05,
+      "loss": 1.1987,
+      "step": 3360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.170494556427002,
+      "learning_rate": 2.134803500943882e-05,
+      "loss": 0.8815,
+      "step": 3361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.56536340713501,
+      "learning_rate": 2.1345460785996222e-05,
+      "loss": 0.7037,
+      "step": 3362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.479547500610352,
+      "learning_rate": 2.1342886562553632e-05,
+      "loss": 0.9098,
+      "step": 3363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.641366004943848,
+      "learning_rate": 2.1340312339111035e-05,
+      "loss": 1.1193,
+      "step": 3364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.93000602722168,
+      "learning_rate": 2.1337738115668442e-05,
+      "loss": 0.8947,
+      "step": 3365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.783208847045898,
+      "learning_rate": 2.1335163892225845e-05,
+      "loss": 0.695,
+      "step": 3366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.242003440856934,
+      "learning_rate": 2.1332589668783252e-05,
+      "loss": 0.8784,
+      "step": 3367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.016057014465332,
+      "learning_rate": 2.1330015445340655e-05,
+      "loss": 0.7613,
+      "step": 3368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.449103355407715,
+      "learning_rate": 2.1327441221898062e-05,
+      "loss": 0.87,
+      "step": 3369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.125410079956055,
+      "learning_rate": 2.1324866998455465e-05,
+      "loss": 0.8774,
+      "step": 3370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.989790916442871,
+      "learning_rate": 2.1322292775012872e-05,
+      "loss": 1.0129,
+      "step": 3371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.345392227172852,
+      "learning_rate": 2.1319718551570275e-05,
+      "loss": 0.6481,
+      "step": 3372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.129225730895996,
+      "learning_rate": 2.1317144328127682e-05,
+      "loss": 0.9379,
+      "step": 3373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.49839973449707,
+      "learning_rate": 2.131457010468509e-05,
+      "loss": 0.805,
+      "step": 3374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.823487281799316,
+      "learning_rate": 2.1311995881242492e-05,
+      "loss": 0.8994,
+      "step": 3375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.875161170959473,
+      "learning_rate": 2.13094216577999e-05,
+      "loss": 1.0175,
+      "step": 3376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.919576644897461,
+      "learning_rate": 2.1306847434357302e-05,
+      "loss": 1.0558,
+      "step": 3377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.043708801269531,
+      "learning_rate": 2.130427321091471e-05,
+      "loss": 0.9635,
+      "step": 3378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.710693359375,
+      "learning_rate": 2.1301698987472112e-05,
+      "loss": 0.8354,
+      "step": 3379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.491646766662598,
+      "learning_rate": 2.129912476402952e-05,
+      "loss": 0.6101,
+      "step": 3380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.665020942687988,
+      "learning_rate": 2.1296550540586922e-05,
+      "loss": 1.1387,
+      "step": 3381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.968727111816406,
+      "learning_rate": 2.129397631714433e-05,
+      "loss": 0.8374,
+      "step": 3382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.36660385131836,
+      "learning_rate": 2.1291402093701735e-05,
+      "loss": 0.7927,
+      "step": 3383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.348801612854004,
+      "learning_rate": 2.128882787025914e-05,
+      "loss": 0.8111,
+      "step": 3384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.692757606506348,
+      "learning_rate": 2.1286253646816545e-05,
+      "loss": 0.7851,
+      "step": 3385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.30654525756836,
+      "learning_rate": 2.128367942337395e-05,
+      "loss": 0.8872,
+      "step": 3386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.270379066467285,
+      "learning_rate": 2.1281105199931355e-05,
+      "loss": 0.8711,
+      "step": 3387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.431641578674316,
+      "learning_rate": 2.127853097648876e-05,
+      "loss": 0.6263,
+      "step": 3388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.345833778381348,
+      "learning_rate": 2.1275956753046165e-05,
+      "loss": 0.8163,
+      "step": 3389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.337143898010254,
+      "learning_rate": 2.127338252960357e-05,
+      "loss": 0.8398,
+      "step": 3390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.857279777526855,
+      "learning_rate": 2.1270808306160975e-05,
+      "loss": 0.867,
+      "step": 3391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.896108627319336,
+      "learning_rate": 2.1268234082718382e-05,
+      "loss": 0.9562,
+      "step": 3392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.186572074890137,
+      "learning_rate": 2.1265659859275785e-05,
+      "loss": 0.8309,
+      "step": 3393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.086546897888184,
+      "learning_rate": 2.1263085635833192e-05,
+      "loss": 0.7872,
+      "step": 3394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.540240287780762,
+      "learning_rate": 2.1260511412390595e-05,
+      "loss": 0.7891,
+      "step": 3395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.109271049499512,
+      "learning_rate": 2.1257937188948002e-05,
+      "loss": 1.0625,
+      "step": 3396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.673296928405762,
+      "learning_rate": 2.1255362965505405e-05,
+      "loss": 1.0933,
+      "step": 3397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.194646835327148,
+      "learning_rate": 2.1252788742062812e-05,
+      "loss": 0.6912,
+      "step": 3398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.768902778625488,
+      "learning_rate": 2.1250214518620215e-05,
+      "loss": 0.8915,
+      "step": 3399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.473485946655273,
+      "learning_rate": 2.1247640295177622e-05,
+      "loss": 0.8354,
+      "step": 3400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.248024940490723,
+      "learning_rate": 2.124506607173503e-05,
+      "loss": 0.721,
+      "step": 3401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.484456062316895,
+      "learning_rate": 2.1242491848292435e-05,
+      "loss": 1.2405,
+      "step": 3402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.60517692565918,
+      "learning_rate": 2.123991762484984e-05,
+      "loss": 0.9914,
+      "step": 3403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.870101928710938,
+      "learning_rate": 2.1237343401407242e-05,
+      "loss": 1.0822,
+      "step": 3404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.580711364746094,
+      "learning_rate": 2.123476917796465e-05,
+      "loss": 0.874,
+      "step": 3405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.969808578491211,
+      "learning_rate": 2.1232194954522052e-05,
+      "loss": 0.9747,
+      "step": 3406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.010531425476074,
+      "learning_rate": 2.122962073107946e-05,
+      "loss": 0.8243,
+      "step": 3407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.558971405029297,
+      "learning_rate": 2.122704650763686e-05,
+      "loss": 0.8482,
+      "step": 3408
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.554405212402344,
+      "learning_rate": 2.122447228419427e-05,
+      "loss": 0.8888,
+      "step": 3409
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.682576179504395,
+      "learning_rate": 2.122189806075167e-05,
+      "loss": 1.2927,
+      "step": 3410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.798102378845215,
+      "learning_rate": 2.1219323837309082e-05,
+      "loss": 0.923,
+      "step": 3411
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.334259986877441,
+      "learning_rate": 2.1216749613866485e-05,
+      "loss": 0.9114,
+      "step": 3412
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.9736008644104,
+      "learning_rate": 2.121417539042389e-05,
+      "loss": 0.7468,
+      "step": 3413
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.202890396118164,
+      "learning_rate": 2.1211601166981295e-05,
+      "loss": 0.9652,
+      "step": 3414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.913182258605957,
+      "learning_rate": 2.1209026943538698e-05,
+      "loss": 0.9204,
+      "step": 3415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.688068389892578,
+      "learning_rate": 2.1206452720096105e-05,
+      "loss": 0.8886,
+      "step": 3416
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.632074356079102,
+      "learning_rate": 2.1203878496653508e-05,
+      "loss": 0.9347,
+      "step": 3417
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.534497261047363,
+      "learning_rate": 2.1201304273210915e-05,
+      "loss": 1.0295,
+      "step": 3418
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.882832527160645,
+      "learning_rate": 2.1198730049768318e-05,
+      "loss": 0.895,
+      "step": 3419
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.662093162536621,
+      "learning_rate": 2.1196155826325728e-05,
+      "loss": 0.8651,
+      "step": 3420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.008186340332031,
+      "learning_rate": 2.119358160288313e-05,
+      "loss": 1.03,
+      "step": 3421
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.827733993530273,
+      "learning_rate": 2.1191007379440538e-05,
+      "loss": 0.9715,
+      "step": 3422
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.083715438842773,
+      "learning_rate": 2.118843315599794e-05,
+      "loss": 0.8469,
+      "step": 3423
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.305743217468262,
+      "learning_rate": 2.1185858932555345e-05,
+      "loss": 0.8906,
+      "step": 3424
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.867901802062988,
+      "learning_rate": 2.118328470911275e-05,
+      "loss": 0.9321,
+      "step": 3425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.041646957397461,
+      "learning_rate": 2.1180710485670155e-05,
+      "loss": 1.1969,
+      "step": 3426
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.933911323547363,
+      "learning_rate": 2.117813626222756e-05,
+      "loss": 0.9315,
+      "step": 3427
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.391996383666992,
+      "learning_rate": 2.1175562038784965e-05,
+      "loss": 0.8534,
+      "step": 3428
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.745831489562988,
+      "learning_rate": 2.117298781534237e-05,
+      "loss": 0.8504,
+      "step": 3429
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.904464721679688,
+      "learning_rate": 2.1170413591899778e-05,
+      "loss": 0.7972,
+      "step": 3430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.746960639953613,
+      "learning_rate": 2.1167839368457185e-05,
+      "loss": 0.9098,
+      "step": 3431
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.606380462646484,
+      "learning_rate": 2.1165265145014588e-05,
+      "loss": 0.8763,
+      "step": 3432
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.09662389755249,
+      "learning_rate": 2.1162690921571995e-05,
+      "loss": 0.8915,
+      "step": 3433
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.510503768920898,
+      "learning_rate": 2.1160116698129398e-05,
+      "loss": 0.6151,
+      "step": 3434
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.570076942443848,
+      "learning_rate": 2.11575424746868e-05,
+      "loss": 0.8435,
+      "step": 3435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.956568717956543,
+      "learning_rate": 2.1154968251244208e-05,
+      "loss": 0.9828,
+      "step": 3436
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.706549644470215,
+      "learning_rate": 2.115239402780161e-05,
+      "loss": 0.912,
+      "step": 3437
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.461691856384277,
+      "learning_rate": 2.1149819804359018e-05,
+      "loss": 0.7623,
+      "step": 3438
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.87739372253418,
+      "learning_rate": 2.1147245580916425e-05,
+      "loss": 1.0354,
+      "step": 3439
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.290151596069336,
+      "learning_rate": 2.114467135747383e-05,
+      "loss": 0.8056,
+      "step": 3440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.44156551361084,
+      "learning_rate": 2.1142097134031235e-05,
+      "loss": 0.7438,
+      "step": 3441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.8095121383667,
+      "learning_rate": 2.113952291058864e-05,
+      "loss": 1.038,
+      "step": 3442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.168571472167969,
+      "learning_rate": 2.1136948687146045e-05,
+      "loss": 0.6805,
+      "step": 3443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.736566543579102,
+      "learning_rate": 2.113437446370345e-05,
+      "loss": 0.7682,
+      "step": 3444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.09931755065918,
+      "learning_rate": 2.1131800240260855e-05,
+      "loss": 0.7133,
+      "step": 3445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.87078857421875,
+      "learning_rate": 2.1129226016818258e-05,
+      "loss": 0.9044,
+      "step": 3446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.737774848937988,
+      "learning_rate": 2.1126651793375665e-05,
+      "loss": 0.9536,
+      "step": 3447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.273669242858887,
+      "learning_rate": 2.112407756993307e-05,
+      "loss": 0.8887,
+      "step": 3448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.952434539794922,
+      "learning_rate": 2.1121503346490478e-05,
+      "loss": 0.9554,
+      "step": 3449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.560873985290527,
+      "learning_rate": 2.111892912304788e-05,
+      "loss": 0.959,
+      "step": 3450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.96397590637207,
+      "learning_rate": 2.1116354899605288e-05,
+      "loss": 0.917,
+      "step": 3451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.227838516235352,
+      "learning_rate": 2.111378067616269e-05,
+      "loss": 0.759,
+      "step": 3452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.357260704040527,
+      "learning_rate": 2.1111206452720098e-05,
+      "loss": 1.0566,
+      "step": 3453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.383551597595215,
+      "learning_rate": 2.11086322292775e-05,
+      "loss": 0.6032,
+      "step": 3454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.857072830200195,
+      "learning_rate": 2.1106058005834905e-05,
+      "loss": 0.8897,
+      "step": 3455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.75000286102295,
+      "learning_rate": 2.110348378239231e-05,
+      "loss": 0.9349,
+      "step": 3456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.371879577636719,
+      "learning_rate": 2.1100909558949715e-05,
+      "loss": 0.7341,
+      "step": 3457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.56074333190918,
+      "learning_rate": 2.1098335335507125e-05,
+      "loss": 0.7258,
+      "step": 3458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.784070014953613,
+      "learning_rate": 2.1095761112064528e-05,
+      "loss": 0.9596,
+      "step": 3459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.739653587341309,
+      "learning_rate": 2.1093186888621935e-05,
+      "loss": 0.7732,
+      "step": 3460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.197305679321289,
+      "learning_rate": 2.1090612665179338e-05,
+      "loss": 0.8965,
+      "step": 3461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.749576568603516,
+      "learning_rate": 2.1088038441736745e-05,
+      "loss": 0.822,
+      "step": 3462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.039815902709961,
+      "learning_rate": 2.1085464218294148e-05,
+      "loss": 1.2785,
+      "step": 3463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.522802352905273,
+      "learning_rate": 2.1082889994851555e-05,
+      "loss": 0.684,
+      "step": 3464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.451534271240234,
+      "learning_rate": 2.1080315771408958e-05,
+      "loss": 1.3313,
+      "step": 3465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.86819076538086,
+      "learning_rate": 2.107774154796636e-05,
+      "loss": 0.9297,
+      "step": 3466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.803569793701172,
+      "learning_rate": 2.107516732452377e-05,
+      "loss": 0.8183,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.220539093017578,
+      "learning_rate": 2.1072593101081174e-05,
+      "loss": 0.7242,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.610380172729492,
+      "learning_rate": 2.107001887763858e-05,
+      "loss": 0.933,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.882061958312988,
+      "learning_rate": 2.1067444654195984e-05,
+      "loss": 0.562,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.083072662353516,
+      "learning_rate": 2.106487043075339e-05,
+      "loss": 1.2563,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.101675033569336,
+      "learning_rate": 2.1062296207310794e-05,
+      "loss": 1.0115,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.753694534301758,
+      "learning_rate": 2.10597219838682e-05,
+      "loss": 0.8939,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.481141090393066,
+      "learning_rate": 2.1057147760425604e-05,
+      "loss": 0.9582,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.974809646606445,
+      "learning_rate": 2.105457353698301e-05,
+      "loss": 0.9938,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.764436721801758,
+      "learning_rate": 2.1051999313540414e-05,
+      "loss": 1.0903,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.915582656860352,
+      "learning_rate": 2.104942509009782e-05,
+      "loss": 0.9686,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.737418174743652,
+      "learning_rate": 2.1046850866655228e-05,
+      "loss": 1.0,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.196995735168457,
+      "learning_rate": 2.104427664321263e-05,
+      "loss": 0.808,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.176024436950684,
+      "learning_rate": 2.1041702419770038e-05,
+      "loss": 0.8546,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.725537300109863,
+      "learning_rate": 2.103912819632744e-05,
+      "loss": 1.122,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.425695419311523,
+      "learning_rate": 2.1036553972884848e-05,
+      "loss": 0.6694,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.48117733001709,
+      "learning_rate": 2.103397974944225e-05,
+      "loss": 0.8846,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.184659004211426,
+      "learning_rate": 2.1031405525999658e-05,
+      "loss": 0.9249,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.160643577575684,
+      "learning_rate": 2.102883130255706e-05,
+      "loss": 0.8291,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.763010025024414,
+      "learning_rate": 2.1026257079114468e-05,
+      "loss": 0.9809,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.668115615844727,
+      "learning_rate": 2.1023682855671874e-05,
+      "loss": 0.8612,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.439923286437988,
+      "learning_rate": 2.1021108632229278e-05,
+      "loss": 0.9016,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.359257698059082,
+      "learning_rate": 2.1018534408786684e-05,
+      "loss": 0.8972,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.74252700805664,
+      "learning_rate": 2.1015960185344088e-05,
+      "loss": 1.0338,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.9962005615234375,
+      "learning_rate": 2.1013385961901494e-05,
+      "loss": 0.6492,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.262776374816895,
+      "learning_rate": 2.1010811738458898e-05,
+      "loss": 0.9953,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.824647903442383,
+      "learning_rate": 2.1008237515016304e-05,
+      "loss": 1.1729,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.100343704223633,
+      "learning_rate": 2.1005663291573708e-05,
+      "loss": 0.8705,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.947909355163574,
+      "learning_rate": 2.1003089068131114e-05,
+      "loss": 0.7091,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.15250015258789,
+      "learning_rate": 2.100051484468852e-05,
+      "loss": 0.8618,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.630385398864746,
+      "learning_rate": 2.0997940621245924e-05,
+      "loss": 0.9365,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.370593070983887,
+      "learning_rate": 2.099536639780333e-05,
+      "loss": 0.8769,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.930132865905762,
+      "learning_rate": 2.0992792174360734e-05,
+      "loss": 1.1851,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.988710403442383,
+      "learning_rate": 2.099021795091814e-05,
+      "loss": 1.0623,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.411948204040527,
+      "learning_rate": 2.0987643727475544e-05,
+      "loss": 0.8566,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.198237419128418,
+      "learning_rate": 2.098506950403295e-05,
+      "loss": 0.8863,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.336523056030273,
+      "learning_rate": 2.0982495280590354e-05,
+      "loss": 0.9533,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.431465148925781,
+      "learning_rate": 2.097992105714776e-05,
+      "loss": 1.0873,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.55638313293457,
+      "learning_rate": 2.0977346833705167e-05,
+      "loss": 1.1952,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.894783020019531,
+      "learning_rate": 2.0974772610262574e-05,
+      "loss": 0.7156,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.867124557495117,
+      "learning_rate": 2.0972198386819977e-05,
+      "loss": 1.0386,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.643655776977539,
+      "learning_rate": 2.096962416337738e-05,
+      "loss": 1.1146,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.800186157226562,
+      "learning_rate": 2.0967049939934787e-05,
+      "loss": 1.0199,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.355324745178223,
+      "learning_rate": 2.096447571649219e-05,
+      "loss": 0.9845,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.809758186340332,
+      "learning_rate": 2.0961901493049597e-05,
+      "loss": 0.863,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.43313217163086,
+      "learning_rate": 2.0959327269607e-05,
+      "loss": 0.8327,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.357420921325684,
+      "learning_rate": 2.0956753046164407e-05,
+      "loss": 0.9441,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.324047088623047,
+      "learning_rate": 2.095417882272181e-05,
+      "loss": 0.867,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.328709602355957,
+      "learning_rate": 2.095160459927922e-05,
+      "loss": 1.003,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.511649131774902,
+      "learning_rate": 2.0949030375836624e-05,
+      "loss": 1.1902,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.947142601013184,
+      "learning_rate": 2.0946456152394027e-05,
+      "loss": 0.8807,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.93046760559082,
+      "learning_rate": 2.0943881928951434e-05,
+      "loss": 1.0436,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.775500297546387,
+      "learning_rate": 2.0941307705508837e-05,
+      "loss": 1.0016,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.531303405761719,
+      "learning_rate": 2.0938733482066244e-05,
+      "loss": 0.966,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.145430564880371,
+      "learning_rate": 2.0936159258623647e-05,
+      "loss": 0.7067,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.787616729736328,
+      "learning_rate": 2.0933585035181054e-05,
+      "loss": 1.0376,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.452864646911621,
+      "learning_rate": 2.0931010811738457e-05,
+      "loss": 1.0156,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.645156860351562,
+      "learning_rate": 2.0928436588295867e-05,
+      "loss": 0.921,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.26245403289795,
+      "learning_rate": 2.092586236485327e-05,
+      "loss": 0.9125,
+      "step": 3525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.762399673461914,
+      "learning_rate": 2.0923288141410677e-05,
+      "loss": 0.8004,
+      "step": 3526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.256109237670898,
+      "learning_rate": 2.092071391796808e-05,
+      "loss": 0.8267,
+      "step": 3527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.301924705505371,
+      "learning_rate": 2.0918139694525484e-05,
+      "loss": 0.9439,
+      "step": 3528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.991838455200195,
+      "learning_rate": 2.091556547108289e-05,
+      "loss": 1.0661,
+      "step": 3529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.099761009216309,
+      "learning_rate": 2.0912991247640294e-05,
+      "loss": 0.8532,
+      "step": 3530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.121550559997559,
+      "learning_rate": 2.09104170241977e-05,
+      "loss": 0.8051,
+      "step": 3531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.28807258605957,
+      "learning_rate": 2.0907842800755104e-05,
+      "loss": 0.7424,
+      "step": 3532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.281378746032715,
+      "learning_rate": 2.090526857731251e-05,
+      "loss": 0.7045,
+      "step": 3533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.212347030639648,
+      "learning_rate": 2.0902694353869917e-05,
+      "loss": 0.8286,
+      "step": 3534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.407116889953613,
+      "learning_rate": 2.0900120130427324e-05,
+      "loss": 1.0528,
+      "step": 3535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.708871841430664,
+      "learning_rate": 2.0897545906984727e-05,
+      "loss": 1.1231,
+      "step": 3536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.085511207580566,
+      "learning_rate": 2.0894971683542134e-05,
+      "loss": 0.7443,
+      "step": 3537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.9285888671875,
+      "learning_rate": 2.0892397460099537e-05,
+      "loss": 1.0158,
+      "step": 3538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.206541061401367,
+      "learning_rate": 2.088982323665694e-05,
+      "loss": 0.7936,
+      "step": 3539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.908160209655762,
+      "learning_rate": 2.0887249013214347e-05,
+      "loss": 0.6885,
+      "step": 3540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.243167877197266,
+      "learning_rate": 2.088467478977175e-05,
+      "loss": 1.0813,
+      "step": 3541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.120624542236328,
+      "learning_rate": 2.0882100566329157e-05,
+      "loss": 1.3071,
+      "step": 3542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.454063415527344,
+      "learning_rate": 2.0879526342886564e-05,
+      "loss": 1.0242,
+      "step": 3543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.565157890319824,
+      "learning_rate": 2.087695211944397e-05,
+      "loss": 1.0155,
+      "step": 3544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.01719856262207,
+      "learning_rate": 2.0874377896001374e-05,
+      "loss": 1.0074,
+      "step": 3545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.954242706298828,
+      "learning_rate": 2.087180367255878e-05,
+      "loss": 0.7199,
+      "step": 3546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.109999656677246,
+      "learning_rate": 2.0869229449116184e-05,
+      "loss": 0.736,
+      "step": 3547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.616226196289062,
+      "learning_rate": 2.086665522567359e-05,
+      "loss": 0.8655,
+      "step": 3548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.14283561706543,
+      "learning_rate": 2.0864081002230994e-05,
+      "loss": 0.7814,
+      "step": 3549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.235616683959961,
+      "learning_rate": 2.0861506778788397e-05,
+      "loss": 0.9135,
+      "step": 3550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.810868263244629,
+      "learning_rate": 2.0858932555345804e-05,
+      "loss": 1.1635,
+      "step": 3551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.180346488952637,
+      "learning_rate": 2.0856358331903207e-05,
+      "loss": 0.9643,
+      "step": 3552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.242701530456543,
+      "learning_rate": 2.0853784108460617e-05,
+      "loss": 0.6851,
+      "step": 3553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.4722261428833,
+      "learning_rate": 2.085120988501802e-05,
+      "loss": 0.9304,
+      "step": 3554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.554118156433105,
+      "learning_rate": 2.0848635661575427e-05,
+      "loss": 1.2391,
+      "step": 3555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.139974594116211,
+      "learning_rate": 2.084606143813283e-05,
+      "loss": 0.7742,
+      "step": 3556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.187524795532227,
+      "learning_rate": 2.0843487214690237e-05,
+      "loss": 0.818,
+      "step": 3557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.278107643127441,
+      "learning_rate": 2.084091299124764e-05,
+      "loss": 0.7832,
+      "step": 3558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.729706764221191,
+      "learning_rate": 2.0838338767805044e-05,
+      "loss": 0.8033,
+      "step": 3559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.261286735534668,
+      "learning_rate": 2.083576454436245e-05,
+      "loss": 1.0819,
+      "step": 3560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.894670486450195,
+      "learning_rate": 2.0833190320919854e-05,
+      "loss": 0.8773,
+      "step": 3561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.732855796813965,
+      "learning_rate": 2.0830616097477264e-05,
+      "loss": 1.083,
+      "step": 3562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.935416221618652,
+      "learning_rate": 2.0828041874034667e-05,
+      "loss": 0.823,
+      "step": 3563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.149697303771973,
+      "learning_rate": 2.0825467650592074e-05,
+      "loss": 1.0436,
+      "step": 3564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.783563613891602,
+      "learning_rate": 2.0822893427149477e-05,
+      "loss": 0.907,
+      "step": 3565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.004741668701172,
+      "learning_rate": 2.0820319203706884e-05,
+      "loss": 0.8219,
+      "step": 3566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.74236011505127,
+      "learning_rate": 2.0817744980264287e-05,
+      "loss": 1.1861,
+      "step": 3567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.484176635742188,
+      "learning_rate": 2.0815170756821694e-05,
+      "loss": 1.0975,
+      "step": 3568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.467572212219238,
+      "learning_rate": 2.0812596533379097e-05,
+      "loss": 0.7448,
+      "step": 3569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.28938102722168,
+      "learning_rate": 2.08100223099365e-05,
+      "loss": 1.1065,
+      "step": 3570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.573892593383789,
+      "learning_rate": 2.080744808649391e-05,
+      "loss": 1.1743,
+      "step": 3571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.535249710083008,
+      "learning_rate": 2.0804873863051313e-05,
+      "loss": 0.8175,
+      "step": 3572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.04575252532959,
+      "learning_rate": 2.080229963960872e-05,
+      "loss": 0.7422,
+      "step": 3573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.565536499023438,
+      "learning_rate": 2.0799725416166123e-05,
+      "loss": 0.9624,
+      "step": 3574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.618603706359863,
+      "learning_rate": 2.079715119272353e-05,
+      "loss": 0.9904,
+      "step": 3575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.872036933898926,
+      "learning_rate": 2.0794576969280933e-05,
+      "loss": 0.9718,
+      "step": 3576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.641082763671875,
+      "learning_rate": 2.079200274583834e-05,
+      "loss": 1.0681,
+      "step": 3577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.628007888793945,
+      "learning_rate": 2.0789428522395743e-05,
+      "loss": 1.004,
+      "step": 3578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.607502937316895,
+      "learning_rate": 2.078685429895315e-05,
+      "loss": 1.1708,
+      "step": 3579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.71158218383789,
+      "learning_rate": 2.0784280075510553e-05,
+      "loss": 0.9265,
+      "step": 3580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.65300464630127,
+      "learning_rate": 2.078170585206796e-05,
+      "loss": 0.6915,
+      "step": 3581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.42484188079834,
+      "learning_rate": 2.0779131628625367e-05,
+      "loss": 0.9448,
+      "step": 3582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.00444221496582,
+      "learning_rate": 2.077655740518277e-05,
+      "loss": 0.9024,
+      "step": 3583
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.023063659667969,
+      "learning_rate": 2.0773983181740177e-05,
+      "loss": 0.8951,
+      "step": 3584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.056589126586914,
+      "learning_rate": 2.077140895829758e-05,
+      "loss": 0.6078,
+      "step": 3585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.09450912475586,
+      "learning_rate": 2.0768834734854987e-05,
+      "loss": 1.0143,
+      "step": 3586
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.711925506591797,
+      "learning_rate": 2.076626051141239e-05,
+      "loss": 1.0569,
+      "step": 3587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.987104415893555,
+      "learning_rate": 2.0763686287969797e-05,
+      "loss": 0.9301,
+      "step": 3588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.281723022460938,
+      "learning_rate": 2.07611120645272e-05,
+      "loss": 0.8745,
+      "step": 3589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.499015808105469,
+      "learning_rate": 2.0758537841084607e-05,
+      "loss": 0.9749,
+      "step": 3590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.553611755371094,
+      "learning_rate": 2.0755963617642013e-05,
+      "loss": 0.8495,
+      "step": 3591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.892396926879883,
+      "learning_rate": 2.0753389394199417e-05,
+      "loss": 0.9487,
+      "step": 3592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.20673942565918,
+      "learning_rate": 2.0750815170756823e-05,
+      "loss": 0.7373,
+      "step": 3593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.146307945251465,
+      "learning_rate": 2.0748240947314227e-05,
+      "loss": 0.637,
+      "step": 3594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.793533325195312,
+      "learning_rate": 2.0745666723871633e-05,
+      "loss": 0.7771,
+      "step": 3595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.788816928863525,
+      "learning_rate": 2.0743092500429037e-05,
+      "loss": 0.7002,
+      "step": 3596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.618098258972168,
+      "learning_rate": 2.0740518276986443e-05,
+      "loss": 0.8879,
+      "step": 3597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.012236595153809,
+      "learning_rate": 2.0737944053543847e-05,
+      "loss": 0.8965,
+      "step": 3598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.48543643951416,
+      "learning_rate": 2.0735369830101253e-05,
+      "loss": 0.9293,
+      "step": 3599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.59442138671875,
+      "learning_rate": 2.073279560665866e-05,
+      "loss": 1.0291,
+      "step": 3600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.945700645446777,
+      "learning_rate": 2.0730221383216063e-05,
+      "loss": 1.0973,
+      "step": 3601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.3099365234375,
+      "learning_rate": 2.072764715977347e-05,
+      "loss": 0.5768,
+      "step": 3602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.856770515441895,
+      "learning_rate": 2.0725072936330873e-05,
+      "loss": 0.8684,
+      "step": 3603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.140678405761719,
+      "learning_rate": 2.072249871288828e-05,
+      "loss": 0.8551,
+      "step": 3604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.962964057922363,
+      "learning_rate": 2.0719924489445683e-05,
+      "loss": 0.8009,
+      "step": 3605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.85222053527832,
+      "learning_rate": 2.071735026600309e-05,
+      "loss": 1.1754,
+      "step": 3606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.708244323730469,
+      "learning_rate": 2.0714776042560493e-05,
+      "loss": 1.0253,
+      "step": 3607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.539422988891602,
+      "learning_rate": 2.07122018191179e-05,
+      "loss": 1.039,
+      "step": 3608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.20456314086914,
+      "learning_rate": 2.0709627595675306e-05,
+      "loss": 1.0313,
+      "step": 3609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.100445747375488,
+      "learning_rate": 2.0707053372232713e-05,
+      "loss": 0.7912,
+      "step": 3610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.814589500427246,
+      "learning_rate": 2.0704479148790116e-05,
+      "loss": 0.9934,
+      "step": 3611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.972722053527832,
+      "learning_rate": 2.070190492534752e-05,
+      "loss": 1.1131,
+      "step": 3612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.813591003417969,
+      "learning_rate": 2.0699330701904926e-05,
+      "loss": 0.8968,
+      "step": 3613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.923067092895508,
+      "learning_rate": 2.069675647846233e-05,
+      "loss": 0.9494,
+      "step": 3614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.696853160858154,
+      "learning_rate": 2.0694182255019736e-05,
+      "loss": 0.6798,
+      "step": 3615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.308184623718262,
+      "learning_rate": 2.069160803157714e-05,
+      "loss": 0.8866,
+      "step": 3616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.513751029968262,
+      "learning_rate": 2.0689033808134546e-05,
+      "loss": 1.0054,
+      "step": 3617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.617974281311035,
+      "learning_rate": 2.068645958469195e-05,
+      "loss": 0.9725,
+      "step": 3618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.4360198974609375,
+      "learning_rate": 2.068388536124936e-05,
+      "loss": 0.8346,
+      "step": 3619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.895474433898926,
+      "learning_rate": 2.0681311137806763e-05,
+      "loss": 0.883,
+      "step": 3620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.972665786743164,
+      "learning_rate": 2.0678736914364166e-05,
+      "loss": 0.9801,
+      "step": 3621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.942977905273438,
+      "learning_rate": 2.0676162690921573e-05,
+      "loss": 0.7673,
+      "step": 3622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.423605918884277,
+      "learning_rate": 2.0673588467478976e-05,
+      "loss": 0.9278,
+      "step": 3623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.923066139221191,
+      "learning_rate": 2.0671014244036383e-05,
+      "loss": 0.825,
+      "step": 3624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.148574829101562,
+      "learning_rate": 2.0668440020593786e-05,
+      "loss": 0.8397,
+      "step": 3625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.27463436126709,
+      "learning_rate": 2.0665865797151193e-05,
+      "loss": 0.7673,
+      "step": 3626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.153227806091309,
+      "learning_rate": 2.0663291573708596e-05,
+      "loss": 1.045,
+      "step": 3627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.231895446777344,
+      "learning_rate": 2.0660717350266006e-05,
+      "loss": 0.9806,
+      "step": 3628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.98068904876709,
+      "learning_rate": 2.065814312682341e-05,
+      "loss": 0.9754,
+      "step": 3629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.418566703796387,
+      "learning_rate": 2.0655568903380816e-05,
+      "loss": 0.8759,
+      "step": 3630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.500130653381348,
+      "learning_rate": 2.065299467993822e-05,
+      "loss": 0.9201,
+      "step": 3631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.282633781433105,
+      "learning_rate": 2.0650420456495623e-05,
+      "loss": 1.0528,
+      "step": 3632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.78018569946289,
+      "learning_rate": 2.064784623305303e-05,
+      "loss": 1.0521,
+      "step": 3633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.125130653381348,
+      "learning_rate": 2.0645272009610433e-05,
+      "loss": 0.7093,
+      "step": 3634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.378721237182617,
+      "learning_rate": 2.064269778616784e-05,
+      "loss": 1.1703,
+      "step": 3635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.150736808776855,
+      "learning_rate": 2.0640123562725243e-05,
+      "loss": 1.0198,
+      "step": 3636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.465408325195312,
+      "learning_rate": 2.063754933928265e-05,
+      "loss": 0.8568,
+      "step": 3637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.055120468139648,
+      "learning_rate": 2.0634975115840056e-05,
+      "loss": 0.7264,
+      "step": 3638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.949440002441406,
+      "learning_rate": 2.0632400892397463e-05,
+      "loss": 0.7122,
+      "step": 3639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.54444694519043,
+      "learning_rate": 2.0629826668954866e-05,
+      "loss": 0.8656,
+      "step": 3640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.26147174835205,
+      "learning_rate": 2.0627252445512273e-05,
+      "loss": 1.0589,
+      "step": 3641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.616227149963379,
+      "learning_rate": 2.0624678222069676e-05,
+      "loss": 1.0893,
+      "step": 3642
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.967999458312988,
+      "learning_rate": 2.062210399862708e-05,
+      "loss": 0.8218,
+      "step": 3643
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.495426177978516,
+      "learning_rate": 2.0619529775184486e-05,
+      "loss": 0.8297,
+      "step": 3644
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.690934181213379,
+      "learning_rate": 2.061695555174189e-05,
+      "loss": 0.8979,
+      "step": 3645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.918691635131836,
+      "learning_rate": 2.0614381328299296e-05,
+      "loss": 1.068,
+      "step": 3646
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.609515190124512,
+      "learning_rate": 2.0611807104856703e-05,
+      "loss": 1.0839,
+      "step": 3647
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.08926010131836,
+      "learning_rate": 2.060923288141411e-05,
+      "loss": 0.6693,
+      "step": 3648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.391222953796387,
+      "learning_rate": 2.0606658657971513e-05,
+      "loss": 0.7525,
+      "step": 3649
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.984003067016602,
+      "learning_rate": 2.060408443452892e-05,
+      "loss": 1.16,
+      "step": 3650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.724226951599121,
+      "learning_rate": 2.0601510211086323e-05,
+      "loss": 1.1775,
+      "step": 3651
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.231865882873535,
+      "learning_rate": 2.059893598764373e-05,
+      "loss": 1.1622,
+      "step": 3652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.671733856201172,
+      "learning_rate": 2.0596361764201133e-05,
+      "loss": 0.8171,
+      "step": 3653
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.980101585388184,
+      "learning_rate": 2.0593787540758536e-05,
+      "loss": 0.9338,
+      "step": 3654
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.749847412109375,
+      "learning_rate": 2.0591213317315943e-05,
+      "loss": 0.8241,
+      "step": 3655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.234421730041504,
+      "learning_rate": 2.0588639093873346e-05,
+      "loss": 0.8507,
+      "step": 3656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.617901802062988,
+      "learning_rate": 2.0586064870430756e-05,
+      "loss": 1.1041,
+      "step": 3657
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.370726585388184,
+      "learning_rate": 2.058349064698816e-05,
+      "loss": 1.0995,
+      "step": 3658
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.447821617126465,
+      "learning_rate": 2.0580916423545566e-05,
+      "loss": 0.6795,
+      "step": 3659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.15230941772461,
+      "learning_rate": 2.057834220010297e-05,
+      "loss": 0.849,
+      "step": 3660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.862015724182129,
+      "learning_rate": 2.0575767976660376e-05,
+      "loss": 0.9088,
+      "step": 3661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.513533592224121,
+      "learning_rate": 2.057319375321778e-05,
+      "loss": 0.8844,
+      "step": 3662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.859278678894043,
+      "learning_rate": 2.0570619529775183e-05,
+      "loss": 1.1208,
+      "step": 3663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.01164722442627,
+      "learning_rate": 2.056804530633259e-05,
+      "loss": 0.8136,
+      "step": 3664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.088756561279297,
+      "learning_rate": 2.0565471082889993e-05,
+      "loss": 0.9166,
+      "step": 3665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.779842376708984,
+      "learning_rate": 2.0562896859447403e-05,
+      "loss": 0.8252,
+      "step": 3666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.122698783874512,
+      "learning_rate": 2.0560322636004806e-05,
+      "loss": 1.1255,
+      "step": 3667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.587461471557617,
+      "learning_rate": 2.0557748412562213e-05,
+      "loss": 0.9872,
+      "step": 3668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.257893562316895,
+      "learning_rate": 2.0555174189119616e-05,
+      "loss": 0.9899,
+      "step": 3669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.810105323791504,
+      "learning_rate": 2.0552599965677023e-05,
+      "loss": 0.8676,
+      "step": 3670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.048532485961914,
+      "learning_rate": 2.0550025742234426e-05,
+      "loss": 0.8357,
+      "step": 3671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.374921798706055,
+      "learning_rate": 2.0547451518791833e-05,
+      "loss": 0.805,
+      "step": 3672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.545283317565918,
+      "learning_rate": 2.0544877295349236e-05,
+      "loss": 0.7126,
+      "step": 3673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.644498825073242,
+      "learning_rate": 2.054230307190664e-05,
+      "loss": 1.0556,
+      "step": 3674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.367122650146484,
+      "learning_rate": 2.053972884846405e-05,
+      "loss": 0.7204,
+      "step": 3675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.233128547668457,
+      "learning_rate": 2.0537154625021452e-05,
+      "loss": 0.9228,
+      "step": 3676
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.225369453430176,
+      "learning_rate": 2.053458040157886e-05,
+      "loss": 1.1144,
+      "step": 3677
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.940107345581055,
+      "learning_rate": 2.0532006178136262e-05,
+      "loss": 1.012,
+      "step": 3678
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.932011604309082,
+      "learning_rate": 2.052943195469367e-05,
+      "loss": 0.7457,
+      "step": 3679
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.4862060546875,
+      "learning_rate": 2.0526857731251072e-05,
+      "loss": 1.4854,
+      "step": 3680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.594715118408203,
+      "learning_rate": 2.052428350780848e-05,
+      "loss": 1.0158,
+      "step": 3681
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.786680698394775,
+      "learning_rate": 2.0521709284365882e-05,
+      "loss": 0.6139,
+      "step": 3682
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.549161911010742,
+      "learning_rate": 2.051913506092329e-05,
+      "loss": 0.7108,
+      "step": 3683
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.391383171081543,
+      "learning_rate": 2.0516560837480692e-05,
+      "loss": 1.3806,
+      "step": 3684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.751867294311523,
+      "learning_rate": 2.05139866140381e-05,
+      "loss": 0.7695,
+      "step": 3685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.59591007232666,
+      "learning_rate": 2.0511412390595506e-05,
+      "loss": 1.2076,
+      "step": 3686
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.624353885650635,
+      "learning_rate": 2.050883816715291e-05,
+      "loss": 0.9371,
+      "step": 3687
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.224845886230469,
+      "learning_rate": 2.0506263943710316e-05,
+      "loss": 0.8086,
+      "step": 3688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.501087188720703,
+      "learning_rate": 2.050368972026772e-05,
+      "loss": 0.9888,
+      "step": 3689
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.994588851928711,
+      "learning_rate": 2.0501115496825126e-05,
+      "loss": 1.1196,
+      "step": 3690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.628186225891113,
+      "learning_rate": 2.049854127338253e-05,
+      "loss": 0.8487,
+      "step": 3691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.290560722351074,
+      "learning_rate": 2.0495967049939936e-05,
+      "loss": 0.8929,
+      "step": 3692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.528678894042969,
+      "learning_rate": 2.049339282649734e-05,
+      "loss": 0.8967,
+      "step": 3693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.490278244018555,
+      "learning_rate": 2.0490818603054746e-05,
+      "loss": 0.8183,
+      "step": 3694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.31004810333252,
+      "learning_rate": 2.0488244379612152e-05,
+      "loss": 0.8496,
+      "step": 3695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.87453842163086,
+      "learning_rate": 2.0485670156169556e-05,
+      "loss": 0.6821,
+      "step": 3696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.041169166564941,
+      "learning_rate": 2.0483095932726962e-05,
+      "loss": 1.1253,
+      "step": 3697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.960317134857178,
+      "learning_rate": 2.0480521709284366e-05,
+      "loss": 0.7579,
+      "step": 3698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.513557434082031,
+      "learning_rate": 2.0477947485841772e-05,
+      "loss": 0.7965,
+      "step": 3699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.026494026184082,
+      "learning_rate": 2.0475373262399176e-05,
+      "loss": 0.6793,
+      "step": 3700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.285836219787598,
+      "learning_rate": 2.0472799038956582e-05,
+      "loss": 0.7486,
+      "step": 3701
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.21915054321289,
+      "learning_rate": 2.0470224815513986e-05,
+      "loss": 0.8206,
+      "step": 3702
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.627345085144043,
+      "learning_rate": 2.0467650592071392e-05,
+      "loss": 1.1697,
+      "step": 3703
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.660696029663086,
+      "learning_rate": 2.04650763686288e-05,
+      "loss": 1.0654,
+      "step": 3704
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.604650497436523,
+      "learning_rate": 2.0462502145186202e-05,
+      "loss": 0.9215,
+      "step": 3705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.867598533630371,
+      "learning_rate": 2.045992792174361e-05,
+      "loss": 0.854,
+      "step": 3706
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.247719764709473,
+      "learning_rate": 2.0457353698301012e-05,
+      "loss": 0.8712,
+      "step": 3707
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.443865776062012,
+      "learning_rate": 2.045477947485842e-05,
+      "loss": 0.9615,
+      "step": 3708
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.16934871673584,
+      "learning_rate": 2.0452205251415822e-05,
+      "loss": 0.9865,
+      "step": 3709
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.650999069213867,
+      "learning_rate": 2.044963102797323e-05,
+      "loss": 0.9038,
+      "step": 3710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.862171173095703,
+      "learning_rate": 2.0447056804530632e-05,
+      "loss": 0.8635,
+      "step": 3711
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.971321105957031,
+      "learning_rate": 2.044448258108804e-05,
+      "loss": 0.7518,
+      "step": 3712
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.484099388122559,
+      "learning_rate": 2.0441908357645445e-05,
+      "loss": 0.888,
+      "step": 3713
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.152271270751953,
+      "learning_rate": 2.0439334134202852e-05,
+      "loss": 0.856,
+      "step": 3714
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.827516555786133,
+      "learning_rate": 2.0436759910760255e-05,
+      "loss": 0.7075,
+      "step": 3715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.254996299743652,
+      "learning_rate": 2.043418568731766e-05,
+      "loss": 0.7263,
+      "step": 3716
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.175752639770508,
+      "learning_rate": 2.0431611463875065e-05,
+      "loss": 0.9317,
+      "step": 3717
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.29155158996582,
+      "learning_rate": 2.042903724043247e-05,
+      "loss": 0.7879,
+      "step": 3718
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.067744255065918,
+      "learning_rate": 2.0426463016989875e-05,
+      "loss": 0.9092,
+      "step": 3719
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.410298347473145,
+      "learning_rate": 2.042388879354728e-05,
+      "loss": 0.803,
+      "step": 3720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.558835983276367,
+      "learning_rate": 2.0421314570104685e-05,
+      "loss": 1.0914,
+      "step": 3721
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.75089168548584,
+      "learning_rate": 2.041874034666209e-05,
+      "loss": 0.7516,
+      "step": 3722
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.924942970275879,
+      "learning_rate": 2.04161661232195e-05,
+      "loss": 1.0185,
+      "step": 3723
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.148600578308105,
+      "learning_rate": 2.0413591899776902e-05,
+      "loss": 0.6848,
+      "step": 3724
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.697454452514648,
+      "learning_rate": 2.041101767633431e-05,
+      "loss": 0.7205,
+      "step": 3725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.940282821655273,
+      "learning_rate": 2.0408443452891712e-05,
+      "loss": 1.0038,
+      "step": 3726
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.05156421661377,
+      "learning_rate": 2.0405869229449115e-05,
+      "loss": 0.7903,
+      "step": 3727
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.532210350036621,
+      "learning_rate": 2.0403295006006522e-05,
+      "loss": 0.9322,
+      "step": 3728
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.266775131225586,
+      "learning_rate": 2.0400720782563925e-05,
+      "loss": 1.1588,
+      "step": 3729
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.365345001220703,
+      "learning_rate": 2.0398146559121332e-05,
+      "loss": 1.3292,
+      "step": 3730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.505094528198242,
+      "learning_rate": 2.0395572335678735e-05,
+      "loss": 0.9166,
+      "step": 3731
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.55389404296875,
+      "learning_rate": 2.0392998112236145e-05,
+      "loss": 0.9683,
+      "step": 3732
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.614384651184082,
+      "learning_rate": 2.039042388879355e-05,
+      "loss": 0.6841,
+      "step": 3733
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.463835716247559,
+      "learning_rate": 2.0387849665350955e-05,
+      "loss": 0.7642,
+      "step": 3734
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.269347190856934,
+      "learning_rate": 2.038527544190836e-05,
+      "loss": 1.3927,
+      "step": 3735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.42772388458252,
+      "learning_rate": 2.0382701218465762e-05,
+      "loss": 0.8991,
+      "step": 3736
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.168792724609375,
+      "learning_rate": 2.038012699502317e-05,
+      "loss": 1.0906,
+      "step": 3737
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.61593246459961,
+      "learning_rate": 2.0377552771580572e-05,
+      "loss": 0.9604,
+      "step": 3738
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.775289535522461,
+      "learning_rate": 2.037497854813798e-05,
+      "loss": 0.7586,
+      "step": 3739
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.232149124145508,
+      "learning_rate": 2.0372404324695382e-05,
+      "loss": 0.9341,
+      "step": 3740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.806316375732422,
+      "learning_rate": 2.036983010125279e-05,
+      "loss": 0.9023,
+      "step": 3741
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.716303825378418,
+      "learning_rate": 2.0367255877810195e-05,
+      "loss": 0.764,
+      "step": 3742
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.610280990600586,
+      "learning_rate": 2.0364681654367602e-05,
+      "loss": 0.7535,
+      "step": 3743
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.894516944885254,
+      "learning_rate": 2.0362107430925005e-05,
+      "loss": 0.8253,
+      "step": 3744
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.50018310546875,
+      "learning_rate": 2.0359533207482412e-05,
+      "loss": 0.7683,
+      "step": 3745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.97812557220459,
+      "learning_rate": 2.0356958984039815e-05,
+      "loss": 0.8469,
+      "step": 3746
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.179431915283203,
+      "learning_rate": 2.035438476059722e-05,
+      "loss": 1.0417,
+      "step": 3747
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.565556526184082,
+      "learning_rate": 2.0351810537154625e-05,
+      "loss": 0.9344,
+      "step": 3748
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.892313957214355,
+      "learning_rate": 2.034923631371203e-05,
+      "loss": 0.8471,
+      "step": 3749
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.469517707824707,
+      "learning_rate": 2.0346662090269435e-05,
+      "loss": 0.6998,
+      "step": 3750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.924263954162598,
+      "learning_rate": 2.0344087866826842e-05,
+      "loss": 0.9014,
+      "step": 3751
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.163795471191406,
+      "learning_rate": 2.034151364338425e-05,
+      "loss": 0.724,
+      "step": 3752
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.019913196563721,
+      "learning_rate": 2.0338939419941652e-05,
+      "loss": 0.7146,
+      "step": 3753
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.92927074432373,
+      "learning_rate": 2.033636519649906e-05,
+      "loss": 0.7491,
+      "step": 3754
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.876102447509766,
+      "learning_rate": 2.0333790973056462e-05,
+      "loss": 0.9295,
+      "step": 3755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.042825698852539,
+      "learning_rate": 2.033121674961387e-05,
+      "loss": 1.0661,
+      "step": 3756
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.4817795753479,
+      "learning_rate": 2.0328642526171272e-05,
+      "loss": 0.7045,
+      "step": 3757
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.642267227172852,
+      "learning_rate": 2.0326068302728675e-05,
+      "loss": 0.8738,
+      "step": 3758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.736170768737793,
+      "learning_rate": 2.032349407928608e-05,
+      "loss": 0.7063,
+      "step": 3759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.759143829345703,
+      "learning_rate": 2.0320919855843485e-05,
+      "loss": 0.884,
+      "step": 3760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.837376594543457,
+      "learning_rate": 2.0318345632400895e-05,
+      "loss": 1.0298,
+      "step": 3761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.604540824890137,
+      "learning_rate": 2.03157714089583e-05,
+      "loss": 0.8982,
+      "step": 3762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.09512710571289,
+      "learning_rate": 2.0313197185515705e-05,
+      "loss": 0.8747,
+      "step": 3763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.5975923538208,
+      "learning_rate": 2.031062296207311e-05,
+      "loss": 0.9215,
+      "step": 3764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.992774963378906,
+      "learning_rate": 2.0308048738630515e-05,
+      "loss": 1.066,
+      "step": 3765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.00985050201416,
+      "learning_rate": 2.0305474515187918e-05,
+      "loss": 1.0393,
+      "step": 3766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.537951469421387,
+      "learning_rate": 2.030290029174532e-05,
+      "loss": 0.8109,
+      "step": 3767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.143230438232422,
+      "learning_rate": 2.0300326068302728e-05,
+      "loss": 0.9042,
+      "step": 3768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.912277221679688,
+      "learning_rate": 2.029775184486013e-05,
+      "loss": 0.89,
+      "step": 3769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.441620826721191,
+      "learning_rate": 2.029517762141754e-05,
+      "loss": 0.7692,
+      "step": 3770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.406990051269531,
+      "learning_rate": 2.0292603397974945e-05,
+      "loss": 0.6933,
+      "step": 3771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.918211936950684,
+      "learning_rate": 2.029002917453235e-05,
+      "loss": 0.8019,
+      "step": 3772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.296481132507324,
+      "learning_rate": 2.0287454951089755e-05,
+      "loss": 0.7531,
+      "step": 3773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.972326278686523,
+      "learning_rate": 2.028488072764716e-05,
+      "loss": 0.9363,
+      "step": 3774
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.636072158813477,
+      "learning_rate": 2.0282306504204565e-05,
+      "loss": 0.7106,
+      "step": 3775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.79617691040039,
+      "learning_rate": 2.027973228076197e-05,
+      "loss": 1.0762,
+      "step": 3776
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.521615982055664,
+      "learning_rate": 2.0277158057319375e-05,
+      "loss": 0.6283,
+      "step": 3777
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.016718864440918,
+      "learning_rate": 2.0274583833876778e-05,
+      "loss": 0.7252,
+      "step": 3778
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.147590637207031,
+      "learning_rate": 2.0272009610434185e-05,
+      "loss": 0.7279,
+      "step": 3779
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.012465476989746,
+      "learning_rate": 2.026943538699159e-05,
+      "loss": 0.9239,
+      "step": 3780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.803175926208496,
+      "learning_rate": 2.0266861163548998e-05,
+      "loss": 0.7739,
+      "step": 3781
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.089409828186035,
+      "learning_rate": 2.02642869401064e-05,
+      "loss": 0.9459,
+      "step": 3782
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.030314445495605,
+      "learning_rate": 2.0261712716663808e-05,
+      "loss": 1.0363,
+      "step": 3783
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.39607048034668,
+      "learning_rate": 2.025913849322121e-05,
+      "loss": 0.7716,
+      "step": 3784
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.431344032287598,
+      "learning_rate": 2.0256564269778618e-05,
+      "loss": 1.1317,
+      "step": 3785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.764358520507812,
+      "learning_rate": 2.025399004633602e-05,
+      "loss": 0.819,
+      "step": 3786
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.26692008972168,
+      "learning_rate": 2.0251415822893428e-05,
+      "loss": 0.8458,
+      "step": 3787
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.75743293762207,
+      "learning_rate": 2.024884159945083e-05,
+      "loss": 0.6654,
+      "step": 3788
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.401911735534668,
+      "learning_rate": 2.0246267376008238e-05,
+      "loss": 0.7295,
+      "step": 3789
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.62918758392334,
+      "learning_rate": 2.0243693152565645e-05,
+      "loss": 0.7644,
+      "step": 3790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.303372383117676,
+      "learning_rate": 2.0241118929123048e-05,
+      "loss": 0.8104,
+      "step": 3791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.16428279876709,
+      "learning_rate": 2.0238544705680455e-05,
+      "loss": 0.6437,
+      "step": 3792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.093573570251465,
+      "learning_rate": 2.0235970482237858e-05,
+      "loss": 0.8171,
+      "step": 3793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.73033618927002,
+      "learning_rate": 2.0233396258795265e-05,
+      "loss": 0.8519,
+      "step": 3794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.583580017089844,
+      "learning_rate": 2.0230822035352668e-05,
+      "loss": 0.8865,
+      "step": 3795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.221098899841309,
+      "learning_rate": 2.0228247811910075e-05,
+      "loss": 0.8021,
+      "step": 3796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.95055866241455,
+      "learning_rate": 2.0225673588467478e-05,
+      "loss": 0.6669,
+      "step": 3797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.826077461242676,
+      "learning_rate": 2.0223099365024885e-05,
+      "loss": 0.8876,
+      "step": 3798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.529410362243652,
+      "learning_rate": 2.022052514158229e-05,
+      "loss": 0.7734,
+      "step": 3799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.645349502563477,
+      "learning_rate": 2.0217950918139695e-05,
+      "loss": 0.9514,
+      "step": 3800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.191333770751953,
+      "learning_rate": 2.02153766946971e-05,
+      "loss": 0.9086,
+      "step": 3801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.085418701171875,
+      "learning_rate": 2.0212802471254505e-05,
+      "loss": 0.7162,
+      "step": 3802
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.082696914672852,
+      "learning_rate": 2.021022824781191e-05,
+      "loss": 0.9013,
+      "step": 3803
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.440038681030273,
+      "learning_rate": 2.0207654024369315e-05,
+      "loss": 0.9378,
+      "step": 3804
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.477656364440918,
+      "learning_rate": 2.020507980092672e-05,
+      "loss": 0.9534,
+      "step": 3805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.915143013000488,
+      "learning_rate": 2.0202505577484125e-05,
+      "loss": 0.8992,
+      "step": 3806
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.547405242919922,
+      "learning_rate": 2.019993135404153e-05,
+      "loss": 0.6859,
+      "step": 3807
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.993768692016602,
+      "learning_rate": 2.0197357130598938e-05,
+      "loss": 0.9447,
+      "step": 3808
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.867547035217285,
+      "learning_rate": 2.019478290715634e-05,
+      "loss": 0.7099,
+      "step": 3809
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.236151695251465,
+      "learning_rate": 2.0192208683713748e-05,
+      "loss": 0.8131,
+      "step": 3810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.449265480041504,
+      "learning_rate": 2.018963446027115e-05,
+      "loss": 1.2849,
+      "step": 3811
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.297179222106934,
+      "learning_rate": 2.0187060236828558e-05,
+      "loss": 0.8853,
+      "step": 3812
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.278464317321777,
+      "learning_rate": 2.018448601338596e-05,
+      "loss": 1.0639,
+      "step": 3813
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.30207633972168,
+      "learning_rate": 2.0181911789943368e-05,
+      "loss": 1.0393,
+      "step": 3814
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.025157928466797,
+      "learning_rate": 2.017933756650077e-05,
+      "loss": 0.8197,
+      "step": 3815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.642430305480957,
+      "learning_rate": 2.0176763343058178e-05,
+      "loss": 0.7759,
+      "step": 3816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.751938819885254,
+      "learning_rate": 2.0174189119615585e-05,
+      "loss": 0.7446,
+      "step": 3817
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.939157485961914,
+      "learning_rate": 2.017161489617299e-05,
+      "loss": 0.9189,
+      "step": 3818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.48426628112793,
+      "learning_rate": 2.0169040672730394e-05,
+      "loss": 1.0649,
+      "step": 3819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.230118751525879,
+      "learning_rate": 2.0166466449287798e-05,
+      "loss": 0.6961,
+      "step": 3820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.982062339782715,
+      "learning_rate": 2.0163892225845204e-05,
+      "loss": 0.7987,
+      "step": 3821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.723980903625488,
+      "learning_rate": 2.0161318002402608e-05,
+      "loss": 0.8645,
+      "step": 3822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.927157402038574,
+      "learning_rate": 2.0158743778960014e-05,
+      "loss": 0.9157,
+      "step": 3823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.691457748413086,
+      "learning_rate": 2.0156169555517418e-05,
+      "loss": 1.1289,
+      "step": 3824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.984818458557129,
+      "learning_rate": 2.0153595332074824e-05,
+      "loss": 0.9488,
+      "step": 3825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.21733570098877,
+      "learning_rate": 2.0151021108632228e-05,
+      "loss": 1.0839,
+      "step": 3826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.890050888061523,
+      "learning_rate": 2.0148446885189638e-05,
+      "loss": 1.0869,
+      "step": 3827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.107467651367188,
+      "learning_rate": 2.014587266174704e-05,
+      "loss": 0.9242,
+      "step": 3828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.55062198638916,
+      "learning_rate": 2.0143298438304448e-05,
+      "loss": 0.8801,
+      "step": 3829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.245901107788086,
+      "learning_rate": 2.014072421486185e-05,
+      "loss": 1.0323,
+      "step": 3830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.906715393066406,
+      "learning_rate": 2.0138149991419254e-05,
+      "loss": 0.8801,
+      "step": 3831
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.886914253234863,
+      "learning_rate": 2.013557576797666e-05,
+      "loss": 1.0558,
+      "step": 3832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.154519081115723,
+      "learning_rate": 2.0133001544534064e-05,
+      "loss": 0.9072,
+      "step": 3833
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.438261032104492,
+      "learning_rate": 2.013042732109147e-05,
+      "loss": 0.7267,
+      "step": 3834
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.867873191833496,
+      "learning_rate": 2.0127853097648874e-05,
+      "loss": 1.0893,
+      "step": 3835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.160661697387695,
+      "learning_rate": 2.0125278874206284e-05,
+      "loss": 1.0172,
+      "step": 3836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.30505084991455,
+      "learning_rate": 2.0122704650763688e-05,
+      "loss": 0.9757,
+      "step": 3837
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.593942642211914,
+      "learning_rate": 2.0120130427321094e-05,
+      "loss": 0.8271,
+      "step": 3838
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.254300117492676,
+      "learning_rate": 2.0117556203878498e-05,
+      "loss": 1.0351,
+      "step": 3839
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.483967781066895,
+      "learning_rate": 2.01149819804359e-05,
+      "loss": 0.8084,
+      "step": 3840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.560583114624023,
+      "learning_rate": 2.0112407756993308e-05,
+      "loss": 1.077,
+      "step": 3841
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.414691925048828,
+      "learning_rate": 2.010983353355071e-05,
+      "loss": 0.9408,
+      "step": 3842
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.46604061126709,
+      "learning_rate": 2.0107259310108118e-05,
+      "loss": 0.8097,
+      "step": 3843
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.858001708984375,
+      "learning_rate": 2.010468508666552e-05,
+      "loss": 0.9786,
+      "step": 3844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.043100357055664,
+      "learning_rate": 2.0102110863222928e-05,
+      "loss": 1.1017,
+      "step": 3845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.875686645507812,
+      "learning_rate": 2.0099536639780334e-05,
+      "loss": 1.1055,
+      "step": 3846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.703948974609375,
+      "learning_rate": 2.009696241633774e-05,
+      "loss": 0.8754,
+      "step": 3847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.572092056274414,
+      "learning_rate": 2.0094388192895144e-05,
+      "loss": 0.7629,
+      "step": 3848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.499730110168457,
+      "learning_rate": 2.009181396945255e-05,
+      "loss": 0.7391,
+      "step": 3849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.371528625488281,
+      "learning_rate": 2.0089239746009954e-05,
+      "loss": 1.2184,
+      "step": 3850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.71910285949707,
+      "learning_rate": 2.0086665522567357e-05,
+      "loss": 0.9444,
+      "step": 3851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.443791389465332,
+      "learning_rate": 2.0084091299124764e-05,
+      "loss": 0.8411,
+      "step": 3852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.28685474395752,
+      "learning_rate": 2.0081517075682167e-05,
+      "loss": 1.0332,
+      "step": 3853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.836456298828125,
+      "learning_rate": 2.0078942852239574e-05,
+      "loss": 0.8048,
+      "step": 3854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.466194152832031,
+      "learning_rate": 2.007636862879698e-05,
+      "loss": 0.9095,
+      "step": 3855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.917708396911621,
+      "learning_rate": 2.0073794405354387e-05,
+      "loss": 1.3415,
+      "step": 3856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.729721069335938,
+      "learning_rate": 2.007122018191179e-05,
+      "loss": 1.0201,
+      "step": 3857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.793363571166992,
+      "learning_rate": 2.0068645958469197e-05,
+      "loss": 1.1044,
+      "step": 3858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.919968605041504,
+      "learning_rate": 2.00660717350266e-05,
+      "loss": 0.706,
+      "step": 3859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.555896759033203,
+      "learning_rate": 2.0063497511584007e-05,
+      "loss": 0.6672,
+      "step": 3860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.051361083984375,
+      "learning_rate": 2.006092328814141e-05,
+      "loss": 0.9201,
+      "step": 3861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.385688781738281,
+      "learning_rate": 2.0058349064698814e-05,
+      "loss": 0.9987,
+      "step": 3862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.015984535217285,
+      "learning_rate": 2.005577484125622e-05,
+      "loss": 0.9494,
+      "step": 3863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.236820220947266,
+      "learning_rate": 2.0053200617813624e-05,
+      "loss": 0.9098,
+      "step": 3864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.320958137512207,
+      "learning_rate": 2.0050626394371034e-05,
+      "loss": 0.8217,
+      "step": 3865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.640387535095215,
+      "learning_rate": 2.0048052170928437e-05,
+      "loss": 0.8337,
+      "step": 3866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.85649299621582,
+      "learning_rate": 2.0045477947485844e-05,
+      "loss": 0.8443,
+      "step": 3867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.217432975769043,
+      "learning_rate": 2.0042903724043247e-05,
+      "loss": 0.9053,
+      "step": 3868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.322803497314453,
+      "learning_rate": 2.0040329500600654e-05,
+      "loss": 0.9291,
+      "step": 3869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.638202667236328,
+      "learning_rate": 2.0037755277158057e-05,
+      "loss": 0.7854,
+      "step": 3870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.8377103805542,
+      "learning_rate": 2.003518105371546e-05,
+      "loss": 1.0823,
+      "step": 3871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.107596397399902,
+      "learning_rate": 2.0032606830272867e-05,
+      "loss": 0.817,
+      "step": 3872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.954251289367676,
+      "learning_rate": 2.003003260683027e-05,
+      "loss": 1.035,
+      "step": 3873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.055277824401855,
+      "learning_rate": 2.002745838338768e-05,
+      "loss": 0.629,
+      "step": 3874
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.547800064086914,
+      "learning_rate": 2.0024884159945084e-05,
+      "loss": 0.864,
+      "step": 3875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.284165382385254,
+      "learning_rate": 2.002230993650249e-05,
+      "loss": 0.9818,
+      "step": 3876
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.371208190917969,
+      "learning_rate": 2.0019735713059894e-05,
+      "loss": 0.883,
+      "step": 3877
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.45896053314209,
+      "learning_rate": 2.00171614896173e-05,
+      "loss": 0.7917,
+      "step": 3878
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.150925636291504,
+      "learning_rate": 2.0014587266174704e-05,
+      "loss": 0.8472,
+      "step": 3879
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.101434707641602,
+      "learning_rate": 2.001201304273211e-05,
+      "loss": 0.6148,
+      "step": 3880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.535784721374512,
+      "learning_rate": 2.0009438819289514e-05,
+      "loss": 0.8225,
+      "step": 3881
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.274863243103027,
+      "learning_rate": 2.0006864595846917e-05,
+      "loss": 0.6738,
+      "step": 3882
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.1603422164917,
+      "learning_rate": 2.0004290372404324e-05,
+      "loss": 0.9445,
+      "step": 3883
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.697280883789062,
+      "learning_rate": 2.000171614896173e-05,
+      "loss": 0.7578,
+      "step": 3884
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.526755332946777,
+      "learning_rate": 1.9999141925519137e-05,
+      "loss": 0.8279,
+      "step": 3885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.127225875854492,
+      "learning_rate": 1.999656770207654e-05,
+      "loss": 1.3112,
+      "step": 3886
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.579535484313965,
+      "learning_rate": 1.9993993478633947e-05,
+      "loss": 0.9747,
+      "step": 3887
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.592826843261719,
+      "learning_rate": 1.999141925519135e-05,
+      "loss": 0.7073,
+      "step": 3888
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.447958946228027,
+      "learning_rate": 1.9988845031748757e-05,
+      "loss": 0.6782,
+      "step": 3889
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.861957550048828,
+      "learning_rate": 1.998627080830616e-05,
+      "loss": 0.908,
+      "step": 3890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.820218086242676,
+      "learning_rate": 1.9983696584863567e-05,
+      "loss": 0.8226,
+      "step": 3891
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.044750213623047,
+      "learning_rate": 1.998112236142097e-05,
+      "loss": 0.7616,
+      "step": 3892
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.752326011657715,
+      "learning_rate": 1.9978548137978377e-05,
+      "loss": 0.6696,
+      "step": 3893
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.505807876586914,
+      "learning_rate": 1.9975973914535784e-05,
+      "loss": 0.916,
+      "step": 3894
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.49383544921875,
+      "learning_rate": 1.9973399691093187e-05,
+      "loss": 0.8205,
+      "step": 3895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.455961227416992,
+      "learning_rate": 1.9970825467650594e-05,
+      "loss": 0.7455,
+      "step": 3896
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.989890098571777,
+      "learning_rate": 1.9968251244207997e-05,
+      "loss": 0.7624,
+      "step": 3897
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.868483543395996,
+      "learning_rate": 1.9965677020765404e-05,
+      "loss": 1.0551,
+      "step": 3898
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.663898468017578,
+      "learning_rate": 1.9963102797322807e-05,
+      "loss": 0.7522,
+      "step": 3899
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.32986068725586,
+      "learning_rate": 1.9960528573880214e-05,
+      "loss": 0.6805,
+      "step": 3900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.079046249389648,
+      "learning_rate": 1.9957954350437617e-05,
+      "loss": 0.9041,
+      "step": 3901
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.203667640686035,
+      "learning_rate": 1.9955380126995024e-05,
+      "loss": 0.8628,
+      "step": 3902
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.46178913116455,
+      "learning_rate": 1.995280590355243e-05,
+      "loss": 1.0608,
+      "step": 3903
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.199135780334473,
+      "learning_rate": 1.9950231680109834e-05,
+      "loss": 0.7819,
+      "step": 3904
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.815032005310059,
+      "learning_rate": 1.994765745666724e-05,
+      "loss": 0.8539,
+      "step": 3905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.437705039978027,
+      "learning_rate": 1.9945083233224644e-05,
+      "loss": 0.8162,
+      "step": 3906
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.97113037109375,
+      "learning_rate": 1.994250900978205e-05,
+      "loss": 0.7135,
+      "step": 3907
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.980389595031738,
+      "learning_rate": 1.9939934786339454e-05,
+      "loss": 0.9366,
+      "step": 3908
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.487383842468262,
+      "learning_rate": 1.993736056289686e-05,
+      "loss": 0.7006,
+      "step": 3909
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.992122650146484,
+      "learning_rate": 1.9934786339454264e-05,
+      "loss": 0.7282,
+      "step": 3910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.490836143493652,
+      "learning_rate": 1.993221211601167e-05,
+      "loss": 0.7352,
+      "step": 3911
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.785042762756348,
+      "learning_rate": 1.9929637892569077e-05,
+      "loss": 1.1043,
+      "step": 3912
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.121511459350586,
+      "learning_rate": 1.992706366912648e-05,
+      "loss": 1.0134,
+      "step": 3913
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.977069854736328,
+      "learning_rate": 1.9924489445683887e-05,
+      "loss": 0.8898,
+      "step": 3914
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.827605247497559,
+      "learning_rate": 1.992191522224129e-05,
+      "loss": 0.7977,
+      "step": 3915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.312994956970215,
+      "learning_rate": 1.9919340998798697e-05,
+      "loss": 0.8871,
+      "step": 3916
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.360579490661621,
+      "learning_rate": 1.99167667753561e-05,
+      "loss": 0.7374,
+      "step": 3917
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.107073783874512,
+      "learning_rate": 1.9914192551913507e-05,
+      "loss": 0.7401,
+      "step": 3918
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.83326530456543,
+      "learning_rate": 1.991161832847091e-05,
+      "loss": 0.8274,
+      "step": 3919
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.00495719909668,
+      "learning_rate": 1.9909044105028317e-05,
+      "loss": 0.8952,
+      "step": 3920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.67477798461914,
+      "learning_rate": 1.9906469881585724e-05,
+      "loss": 0.873,
+      "step": 3921
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.246710777282715,
+      "learning_rate": 1.990389565814313e-05,
+      "loss": 0.6519,
+      "step": 3922
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.416906356811523,
+      "learning_rate": 1.9901321434700533e-05,
+      "loss": 0.8081,
+      "step": 3923
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.19828987121582,
+      "learning_rate": 1.9898747211257937e-05,
+      "loss": 0.9664,
+      "step": 3924
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.778648376464844,
+      "learning_rate": 1.9896172987815343e-05,
+      "loss": 0.9061,
+      "step": 3925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.403685569763184,
+      "learning_rate": 1.9893598764372747e-05,
+      "loss": 0.8616,
+      "step": 3926
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.282853126525879,
+      "learning_rate": 1.9891024540930153e-05,
+      "loss": 1.2102,
+      "step": 3927
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.931404113769531,
+      "learning_rate": 1.9888450317487557e-05,
+      "loss": 0.8525,
+      "step": 3928
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.738006591796875,
+      "learning_rate": 1.9885876094044963e-05,
+      "loss": 0.9559,
+      "step": 3929
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.042181015014648,
+      "learning_rate": 1.9883301870602367e-05,
+      "loss": 0.9463,
+      "step": 3930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.851289749145508,
+      "learning_rate": 1.9880727647159777e-05,
+      "loss": 0.7099,
+      "step": 3931
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.815483093261719,
+      "learning_rate": 1.987815342371718e-05,
+      "loss": 0.5558,
+      "step": 3932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.740373611450195,
+      "learning_rate": 1.9875579200274587e-05,
+      "loss": 1.0212,
+      "step": 3933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.694733619689941,
+      "learning_rate": 1.987300497683199e-05,
+      "loss": 1.0334,
+      "step": 3934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.542588233947754,
+      "learning_rate": 1.9870430753389393e-05,
+      "loss": 0.6258,
+      "step": 3935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.80351448059082,
+      "learning_rate": 1.98678565299468e-05,
+      "loss": 0.7614,
+      "step": 3936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.947404861450195,
+      "learning_rate": 1.9865282306504203e-05,
+      "loss": 0.6481,
+      "step": 3937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.279672622680664,
+      "learning_rate": 1.986270808306161e-05,
+      "loss": 0.9361,
+      "step": 3938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.397329330444336,
+      "learning_rate": 1.9860133859619013e-05,
+      "loss": 0.6903,
+      "step": 3939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.429089546203613,
+      "learning_rate": 1.9857559636176423e-05,
+      "loss": 1.1352,
+      "step": 3940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.662603378295898,
+      "learning_rate": 1.9854985412733827e-05,
+      "loss": 0.9475,
+      "step": 3941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.281474113464355,
+      "learning_rate": 1.9852411189291233e-05,
+      "loss": 1.1532,
+      "step": 3942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.530838966369629,
+      "learning_rate": 1.9849836965848637e-05,
+      "loss": 0.9931,
+      "step": 3943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.199763298034668,
+      "learning_rate": 1.984726274240604e-05,
+      "loss": 0.7225,
+      "step": 3944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.985010147094727,
+      "learning_rate": 1.9844688518963447e-05,
+      "loss": 0.6305,
+      "step": 3945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.202077865600586,
+      "learning_rate": 1.984211429552085e-05,
+      "loss": 1.0916,
+      "step": 3946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.160751342773438,
+      "learning_rate": 1.9839540072078257e-05,
+      "loss": 0.9651,
+      "step": 3947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.579391479492188,
+      "learning_rate": 1.983696584863566e-05,
+      "loss": 0.8875,
+      "step": 3948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.264220237731934,
+      "learning_rate": 1.9834391625193067e-05,
+      "loss": 0.7546,
+      "step": 3949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.203835487365723,
+      "learning_rate": 1.9831817401750473e-05,
+      "loss": 0.8868,
+      "step": 3950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.139389991760254,
+      "learning_rate": 1.982924317830788e-05,
+      "loss": 0.7785,
+      "step": 3951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.022866249084473,
+      "learning_rate": 1.9826668954865283e-05,
+      "loss": 0.7928,
+      "step": 3952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.815255165100098,
+      "learning_rate": 1.982409473142269e-05,
+      "loss": 0.9878,
+      "step": 3953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.767159461975098,
+      "learning_rate": 1.9821520507980093e-05,
+      "loss": 1.0266,
+      "step": 3954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.081512451171875,
+      "learning_rate": 1.9818946284537496e-05,
+      "loss": 0.6789,
+      "step": 3955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.766496658325195,
+      "learning_rate": 1.9816372061094903e-05,
+      "loss": 0.7038,
+      "step": 3956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.34941577911377,
+      "learning_rate": 1.9813797837652306e-05,
+      "loss": 0.6806,
+      "step": 3957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.309305191040039,
+      "learning_rate": 1.9811223614209713e-05,
+      "loss": 0.8272,
+      "step": 3958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.701016426086426,
+      "learning_rate": 1.980864939076712e-05,
+      "loss": 0.9789,
+      "step": 3959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.719133377075195,
+      "learning_rate": 1.9806075167324526e-05,
+      "loss": 0.8249,
+      "step": 3960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.09665298461914,
+      "learning_rate": 1.980350094388193e-05,
+      "loss": 0.6647,
+      "step": 3961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.239025115966797,
+      "learning_rate": 1.9800926720439336e-05,
+      "loss": 1.0294,
+      "step": 3962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.544559478759766,
+      "learning_rate": 1.979835249699674e-05,
+      "loss": 0.7316,
+      "step": 3963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.569293022155762,
+      "learning_rate": 1.9795778273554146e-05,
+      "loss": 0.98,
+      "step": 3964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.094076156616211,
+      "learning_rate": 1.979320405011155e-05,
+      "loss": 0.75,
+      "step": 3965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.303956985473633,
+      "learning_rate": 1.9790629826668953e-05,
+      "loss": 0.7886,
+      "step": 3966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.011849403381348,
+      "learning_rate": 1.978805560322636e-05,
+      "loss": 0.8513,
+      "step": 3967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.344328880310059,
+      "learning_rate": 1.9785481379783763e-05,
+      "loss": 0.8999,
+      "step": 3968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.330906867980957,
+      "learning_rate": 1.9782907156341173e-05,
+      "loss": 1.0393,
+      "step": 3969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.964336395263672,
+      "learning_rate": 1.9780332932898576e-05,
+      "loss": 0.7451,
+      "step": 3970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.336642265319824,
+      "learning_rate": 1.9777758709455983e-05,
+      "loss": 0.9648,
+      "step": 3971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.92069673538208,
+      "learning_rate": 1.9775184486013386e-05,
+      "loss": 0.6064,
+      "step": 3972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.91940689086914,
+      "learning_rate": 1.9772610262570793e-05,
+      "loss": 0.7281,
+      "step": 3973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.428845405578613,
+      "learning_rate": 1.9770036039128196e-05,
+      "loss": 0.8193,
+      "step": 3974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.538307189941406,
+      "learning_rate": 1.97674618156856e-05,
+      "loss": 0.8031,
+      "step": 3975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.502593994140625,
+      "learning_rate": 1.9764887592243006e-05,
+      "loss": 0.8204,
+      "step": 3976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.508402824401855,
+      "learning_rate": 1.976231336880041e-05,
+      "loss": 0.8356,
+      "step": 3977
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.833074569702148,
+      "learning_rate": 1.975973914535782e-05,
+      "loss": 0.675,
+      "step": 3978
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.171976089477539,
+      "learning_rate": 1.9757164921915223e-05,
+      "loss": 1.0401,
+      "step": 3979
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.629979133605957,
+      "learning_rate": 1.975459069847263e-05,
+      "loss": 1.0111,
+      "step": 3980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.351917266845703,
+      "learning_rate": 1.9752016475030033e-05,
+      "loss": 0.8158,
+      "step": 3981
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.653636932373047,
+      "learning_rate": 1.974944225158744e-05,
+      "loss": 0.9946,
+      "step": 3982
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.064470291137695,
+      "learning_rate": 1.9746868028144843e-05,
+      "loss": 0.8905,
+      "step": 3983
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.972502708435059,
+      "learning_rate": 1.974429380470225e-05,
+      "loss": 0.7342,
+      "step": 3984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.094321250915527,
+      "learning_rate": 1.9741719581259653e-05,
+      "loss": 0.8376,
+      "step": 3985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.028883934020996,
+      "learning_rate": 1.9739145357817056e-05,
+      "loss": 0.9002,
+      "step": 3986
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.039002418518066,
+      "learning_rate": 1.9736571134374463e-05,
+      "loss": 0.9004,
+      "step": 3987
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.977614879608154,
+      "learning_rate": 1.973399691093187e-05,
+      "loss": 0.6631,
+      "step": 3988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.446571350097656,
+      "learning_rate": 1.9731422687489276e-05,
+      "loss": 0.7548,
+      "step": 3989
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.09100341796875,
+      "learning_rate": 1.972884846404668e-05,
+      "loss": 1.1521,
+      "step": 3990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.782296180725098,
+      "learning_rate": 1.9726274240604086e-05,
+      "loss": 0.8137,
+      "step": 3991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.478397369384766,
+      "learning_rate": 1.972370001716149e-05,
+      "loss": 0.8823,
+      "step": 3992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.257417678833008,
+      "learning_rate": 1.9721125793718896e-05,
+      "loss": 1.0606,
+      "step": 3993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.692514419555664,
+      "learning_rate": 1.97185515702763e-05,
+      "loss": 1.0351,
+      "step": 3994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.016999244689941,
+      "learning_rate": 1.9715977346833706e-05,
+      "loss": 0.9902,
+      "step": 3995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.181031227111816,
+      "learning_rate": 1.971340312339111e-05,
+      "loss": 0.7479,
+      "step": 3996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.7083740234375,
+      "learning_rate": 1.9710828899948516e-05,
+      "loss": 0.8022,
+      "step": 3997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.884144306182861,
+      "learning_rate": 1.9708254676505923e-05,
+      "loss": 0.7198,
+      "step": 3998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.083291053771973,
+      "learning_rate": 1.9705680453063326e-05,
+      "loss": 1.1176,
+      "step": 3999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.507850646972656,
+      "learning_rate": 1.9703106229620733e-05,
+      "loss": 0.7748,
+      "step": 4000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.2831449508667,
+      "learning_rate": 1.9700532006178136e-05,
+      "loss": 0.9409,
+      "step": 4001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.327726364135742,
+      "learning_rate": 1.9697957782735543e-05,
+      "loss": 1.0776,
+      "step": 4002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.2455472946167,
+      "learning_rate": 1.9695383559292946e-05,
+      "loss": 0.714,
+      "step": 4003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.188271522521973,
+      "learning_rate": 1.9692809335850353e-05,
+      "loss": 0.8372,
+      "step": 4004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.892428398132324,
+      "learning_rate": 1.9690235112407756e-05,
+      "loss": 0.9201,
+      "step": 4005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.546725273132324,
+      "learning_rate": 1.9687660888965163e-05,
+      "loss": 0.8367,
+      "step": 4006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.803289413452148,
+      "learning_rate": 1.968508666552257e-05,
+      "loss": 0.7534,
+      "step": 4007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.834587097167969,
+      "learning_rate": 1.9682512442079973e-05,
+      "loss": 0.8098,
+      "step": 4008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.217619895935059,
+      "learning_rate": 1.967993821863738e-05,
+      "loss": 0.6934,
+      "step": 4009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.22787857055664,
+      "learning_rate": 1.9677363995194783e-05,
+      "loss": 1.0037,
+      "step": 4010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.565980911254883,
+      "learning_rate": 1.967478977175219e-05,
+      "loss": 0.8713,
+      "step": 4011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.52838134765625,
+      "learning_rate": 1.9672215548309593e-05,
+      "loss": 0.8921,
+      "step": 4012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.013540267944336,
+      "learning_rate": 1.9669641324867e-05,
+      "loss": 0.8828,
+      "step": 4013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.014223098754883,
+      "learning_rate": 1.9667067101424403e-05,
+      "loss": 0.8123,
+      "step": 4014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.63780689239502,
+      "learning_rate": 1.966449287798181e-05,
+      "loss": 0.7775,
+      "step": 4015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.96135425567627,
+      "learning_rate": 1.9661918654539216e-05,
+      "loss": 0.8927,
+      "step": 4016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.869874000549316,
+      "learning_rate": 1.965934443109662e-05,
+      "loss": 0.9113,
+      "step": 4017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.326499938964844,
+      "learning_rate": 1.9656770207654026e-05,
+      "loss": 0.8379,
+      "step": 4018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.697678565979004,
+      "learning_rate": 1.965419598421143e-05,
+      "loss": 1.0268,
+      "step": 4019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.77447509765625,
+      "learning_rate": 1.9651621760768836e-05,
+      "loss": 0.958,
+      "step": 4020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.400771141052246,
+      "learning_rate": 1.964904753732624e-05,
+      "loss": 0.7048,
+      "step": 4021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.796031951904297,
+      "learning_rate": 1.9646473313883646e-05,
+      "loss": 0.8138,
+      "step": 4022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.505208969116211,
+      "learning_rate": 1.964389909044105e-05,
+      "loss": 0.9146,
+      "step": 4023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.580912590026855,
+      "learning_rate": 1.9641324866998456e-05,
+      "loss": 0.8553,
+      "step": 4024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.221284866333008,
+      "learning_rate": 1.9638750643555863e-05,
+      "loss": 1.0036,
+      "step": 4025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.90367317199707,
+      "learning_rate": 1.963617642011327e-05,
+      "loss": 0.8058,
+      "step": 4026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.296404838562012,
+      "learning_rate": 1.9633602196670673e-05,
+      "loss": 0.9541,
+      "step": 4027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.00625991821289,
+      "learning_rate": 1.9631027973228076e-05,
+      "loss": 0.7058,
+      "step": 4028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.92353343963623,
+      "learning_rate": 1.9628453749785482e-05,
+      "loss": 0.845,
+      "step": 4029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.148040771484375,
+      "learning_rate": 1.9625879526342886e-05,
+      "loss": 0.9769,
+      "step": 4030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.550756454467773,
+      "learning_rate": 1.9623305302900292e-05,
+      "loss": 1.1777,
+      "step": 4031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.699030876159668,
+      "learning_rate": 1.9620731079457696e-05,
+      "loss": 0.7918,
+      "step": 4032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.76816463470459,
+      "learning_rate": 1.9618156856015102e-05,
+      "loss": 0.5004,
+      "step": 4033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.999948024749756,
+      "learning_rate": 1.9615582632572506e-05,
+      "loss": 0.615,
+      "step": 4034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.726495742797852,
+      "learning_rate": 1.9613008409129916e-05,
+      "loss": 0.9607,
+      "step": 4035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.991649627685547,
+      "learning_rate": 1.961043418568732e-05,
+      "loss": 0.9198,
+      "step": 4036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.714914321899414,
+      "learning_rate": 1.9607859962244726e-05,
+      "loss": 1.0131,
+      "step": 4037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.5370512008667,
+      "learning_rate": 1.960528573880213e-05,
+      "loss": 0.749,
+      "step": 4038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.226852416992188,
+      "learning_rate": 1.9602711515359532e-05,
+      "loss": 0.7837,
+      "step": 4039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.896933555603027,
+      "learning_rate": 1.960013729191694e-05,
+      "loss": 0.6167,
+      "step": 4040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.405925750732422,
+      "learning_rate": 1.9597563068474342e-05,
+      "loss": 0.972,
+      "step": 4041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.742990493774414,
+      "learning_rate": 1.959498884503175e-05,
+      "loss": 0.9503,
+      "step": 4042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.902505874633789,
+      "learning_rate": 1.9592414621589152e-05,
+      "loss": 0.8328,
+      "step": 4043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.653990745544434,
+      "learning_rate": 1.9589840398146562e-05,
+      "loss": 0.6924,
+      "step": 4044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.983641624450684,
+      "learning_rate": 1.9587266174703966e-05,
+      "loss": 0.6876,
+      "step": 4045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.483264923095703,
+      "learning_rate": 1.9584691951261372e-05,
+      "loss": 0.8161,
+      "step": 4046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.088647842407227,
+      "learning_rate": 1.9582117727818776e-05,
+      "loss": 0.9787,
+      "step": 4047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.744452476501465,
+      "learning_rate": 1.957954350437618e-05,
+      "loss": 0.6605,
+      "step": 4048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.046748161315918,
+      "learning_rate": 1.9576969280933586e-05,
+      "loss": 0.7685,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.177292823791504,
+      "learning_rate": 1.957439505749099e-05,
+      "loss": 0.9668,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.781136512756348,
+      "learning_rate": 1.9571820834048396e-05,
+      "loss": 0.9335,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.072296142578125,
+      "learning_rate": 1.95692466106058e-05,
+      "loss": 0.9117,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.207942008972168,
+      "learning_rate": 1.9566672387163206e-05,
+      "loss": 0.9793,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.456552505493164,
+      "learning_rate": 1.9564098163720612e-05,
+      "loss": 0.5707,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.58328914642334,
+      "learning_rate": 1.956152394027802e-05,
+      "loss": 0.8544,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.354524612426758,
+      "learning_rate": 1.9558949716835422e-05,
+      "loss": 0.875,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.670668601989746,
+      "learning_rate": 1.955637549339283e-05,
+      "loss": 0.7347,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.295302391052246,
+      "learning_rate": 1.9553801269950232e-05,
+      "loss": 0.7823,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.055832862854004,
+      "learning_rate": 1.9551227046507635e-05,
+      "loss": 0.8927,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.445508003234863,
+      "learning_rate": 1.9548652823065042e-05,
+      "loss": 0.6966,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.543179512023926,
+      "learning_rate": 1.9546078599622445e-05,
+      "loss": 0.9005,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.495827674865723,
+      "learning_rate": 1.9543504376179852e-05,
+      "loss": 0.7935,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.159778594970703,
+      "learning_rate": 1.954093015273726e-05,
+      "loss": 0.8622,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.753851890563965,
+      "learning_rate": 1.9538355929294666e-05,
+      "loss": 0.7778,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.581764221191406,
+      "learning_rate": 1.953578170585207e-05,
+      "loss": 0.8594,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.900815963745117,
+      "learning_rate": 1.9533207482409475e-05,
+      "loss": 1.0903,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.259923934936523,
+      "learning_rate": 1.953063325896688e-05,
+      "loss": 0.9654,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.771180152893066,
+      "learning_rate": 1.9528059035524285e-05,
+      "loss": 0.6501,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.301082611083984,
+      "learning_rate": 1.952548481208169e-05,
+      "loss": 0.798,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.170222282409668,
+      "learning_rate": 1.9522910588639092e-05,
+      "loss": 0.8041,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.72446346282959,
+      "learning_rate": 1.95203363651965e-05,
+      "loss": 1.018,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.974662780761719,
+      "learning_rate": 1.9517762141753902e-05,
+      "loss": 0.7048,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.767720222473145,
+      "learning_rate": 1.9515187918311312e-05,
+      "loss": 0.76,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.805106163024902,
+      "learning_rate": 1.9512613694868715e-05,
+      "loss": 0.7146,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.177083015441895,
+      "learning_rate": 1.9510039471426122e-05,
+      "loss": 0.6145,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.28121566772461,
+      "learning_rate": 1.9507465247983525e-05,
+      "loss": 0.8947,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.950400352478027,
+      "learning_rate": 1.9504891024540932e-05,
+      "loss": 0.8343,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.08894157409668,
+      "learning_rate": 1.9502316801098335e-05,
+      "loss": 0.7929,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.948592185974121,
+      "learning_rate": 1.949974257765574e-05,
+      "loss": 0.7078,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.790074348449707,
+      "learning_rate": 1.9497168354213145e-05,
+      "loss": 0.9945,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.925834655761719,
+      "learning_rate": 1.949459413077055e-05,
+      "loss": 0.6215,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.762432098388672,
+      "learning_rate": 1.949201990732796e-05,
+      "loss": 1.0637,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.114999771118164,
+      "learning_rate": 1.9489445683885362e-05,
+      "loss": 0.7603,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.416316986083984,
+      "learning_rate": 1.948687146044277e-05,
+      "loss": 0.8125,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.129293441772461,
+      "learning_rate": 1.9484297237000172e-05,
+      "loss": 0.8315,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.308990478515625,
+      "learning_rate": 1.948172301355758e-05,
+      "loss": 0.9291,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.069663047790527,
+      "learning_rate": 1.9479148790114982e-05,
+      "loss": 0.9763,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.242548942565918,
+      "learning_rate": 1.947657456667239e-05,
+      "loss": 0.8067,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.22472095489502,
+      "learning_rate": 1.9474000343229792e-05,
+      "loss": 0.9655,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.0079984664917,
+      "learning_rate": 1.9471426119787195e-05,
+      "loss": 0.8751,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.007758140563965,
+      "learning_rate": 1.9468851896344602e-05,
+      "loss": 1.2319,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.411227226257324,
+      "learning_rate": 1.946627767290201e-05,
+      "loss": 0.8935,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.085384368896484,
+      "learning_rate": 1.9463703449459415e-05,
+      "loss": 0.8864,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.263283729553223,
+      "learning_rate": 1.946112922601682e-05,
+      "loss": 0.7343,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.197152137756348,
+      "learning_rate": 1.9458555002574225e-05,
+      "loss": 0.7345,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.262878894805908,
+      "learning_rate": 1.945598077913163e-05,
+      "loss": 0.7368,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.347179412841797,
+      "learning_rate": 1.9453406555689035e-05,
+      "loss": 0.7142,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.919077396392822,
+      "learning_rate": 1.945083233224644e-05,
+      "loss": 0.5645,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.976900100708008,
+      "learning_rate": 1.9448258108803845e-05,
+      "loss": 1.0602,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.500653266906738,
+      "learning_rate": 1.944568388536125e-05,
+      "loss": 0.6233,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.730000495910645,
+      "learning_rate": 1.9443109661918655e-05,
+      "loss": 0.8537,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.802042007446289,
+      "learning_rate": 1.9440535438476062e-05,
+      "loss": 0.8679,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.940086364746094,
+      "learning_rate": 1.9437961215033465e-05,
+      "loss": 0.9542,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.198892593383789,
+      "learning_rate": 1.9435386991590872e-05,
+      "loss": 0.9823,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.480631828308105,
+      "learning_rate": 1.9432812768148275e-05,
+      "loss": 0.9504,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.300588607788086,
+      "learning_rate": 1.9430238544705682e-05,
+      "loss": 0.9653,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.719565391540527,
+      "learning_rate": 1.9427664321263085e-05,
+      "loss": 0.7701,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.093180656433105,
+      "learning_rate": 1.9425090097820492e-05,
+      "loss": 0.7741,
+      "step": 4108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.70360279083252,
+      "learning_rate": 1.9422515874377895e-05,
+      "loss": 0.789,
+      "step": 4109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.274117469787598,
+      "learning_rate": 1.9419941650935302e-05,
+      "loss": 0.7288,
+      "step": 4110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.471673965454102,
+      "learning_rate": 1.941736742749271e-05,
+      "loss": 0.9696,
+      "step": 4111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.78393840789795,
+      "learning_rate": 1.941479320405011e-05,
+      "loss": 0.8418,
+      "step": 4112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.887661933898926,
+      "learning_rate": 1.941221898060752e-05,
+      "loss": 0.7292,
+      "step": 4113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.562491416931152,
+      "learning_rate": 1.940964475716492e-05,
+      "loss": 0.9941,
+      "step": 4114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.537960052490234,
+      "learning_rate": 1.940707053372233e-05,
+      "loss": 0.72,
+      "step": 4115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.944845199584961,
+      "learning_rate": 1.940449631027973e-05,
+      "loss": 0.9124,
+      "step": 4116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.464266777038574,
+      "learning_rate": 1.9401922086837138e-05,
+      "loss": 1.0593,
+      "step": 4117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.064270973205566,
+      "learning_rate": 1.939934786339454e-05,
+      "loss": 0.8479,
+      "step": 4118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.163776397705078,
+      "learning_rate": 1.9396773639951948e-05,
+      "loss": 0.9357,
+      "step": 4119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.305214881896973,
+      "learning_rate": 1.9394199416509355e-05,
+      "loss": 0.8636,
+      "step": 4120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.623992919921875,
+      "learning_rate": 1.9391625193066758e-05,
+      "loss": 0.9591,
+      "step": 4121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.971527099609375,
+      "learning_rate": 1.9389050969624165e-05,
+      "loss": 1.1617,
+      "step": 4122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.063480377197266,
+      "learning_rate": 1.9386476746181568e-05,
+      "loss": 0.7387,
+      "step": 4123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.922619342803955,
+      "learning_rate": 1.9383902522738975e-05,
+      "loss": 0.647,
+      "step": 4124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.8080472946167,
+      "learning_rate": 1.9381328299296378e-05,
+      "loss": 0.729,
+      "step": 4125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.750015258789062,
+      "learning_rate": 1.9378754075853785e-05,
+      "loss": 0.9147,
+      "step": 4126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.464468002319336,
+      "learning_rate": 1.9376179852411188e-05,
+      "loss": 0.7463,
+      "step": 4127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.865391731262207,
+      "learning_rate": 1.9373605628968595e-05,
+      "loss": 1.2584,
+      "step": 4128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.900446891784668,
+      "learning_rate": 1.9371031405525998e-05,
+      "loss": 0.8376,
+      "step": 4129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.419876098632812,
+      "learning_rate": 1.9368457182083408e-05,
+      "loss": 0.9709,
+      "step": 4130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.008069038391113,
+      "learning_rate": 1.936588295864081e-05,
+      "loss": 0.7046,
+      "step": 4131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.859042167663574,
+      "learning_rate": 1.9363308735198215e-05,
+      "loss": 0.838,
+      "step": 4132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.988000392913818,
+      "learning_rate": 1.936073451175562e-05,
+      "loss": 0.7154,
+      "step": 4133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.62507963180542,
+      "learning_rate": 1.9358160288313025e-05,
+      "loss": 0.7137,
+      "step": 4134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.302682399749756,
+      "learning_rate": 1.935558606487043e-05,
+      "loss": 0.7076,
+      "step": 4135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.77242374420166,
+      "learning_rate": 1.9353011841427835e-05,
+      "loss": 0.7872,
+      "step": 4136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.081562995910645,
+      "learning_rate": 1.935043761798524e-05,
+      "loss": 0.5662,
+      "step": 4137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.44813060760498,
+      "learning_rate": 1.9347863394542645e-05,
+      "loss": 1.0015,
+      "step": 4138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.001461029052734,
+      "learning_rate": 1.9345289171100055e-05,
+      "loss": 0.9098,
+      "step": 4139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.881959915161133,
+      "learning_rate": 1.9342714947657458e-05,
+      "loss": 1.1068,
+      "step": 4140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.648541450500488,
+      "learning_rate": 1.9340140724214865e-05,
+      "loss": 0.9119,
+      "step": 4141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.838765144348145,
+      "learning_rate": 1.9337566500772268e-05,
+      "loss": 0.9047,
+      "step": 4142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.513163566589355,
+      "learning_rate": 1.933499227732967e-05,
+      "loss": 0.7928,
+      "step": 4143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.546553611755371,
+      "learning_rate": 1.9332418053887078e-05,
+      "loss": 0.7728,
+      "step": 4144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.782511711120605,
+      "learning_rate": 1.932984383044448e-05,
+      "loss": 0.8833,
+      "step": 4145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.285393714904785,
+      "learning_rate": 1.9327269607001888e-05,
+      "loss": 0.9278,
+      "step": 4146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.105173110961914,
+      "learning_rate": 1.932469538355929e-05,
+      "loss": 0.9307,
+      "step": 4147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.945642471313477,
+      "learning_rate": 1.93221211601167e-05,
+      "loss": 0.7086,
+      "step": 4148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.571893692016602,
+      "learning_rate": 1.9319546936674105e-05,
+      "loss": 1.0421,
+      "step": 4149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.30229663848877,
+      "learning_rate": 1.931697271323151e-05,
+      "loss": 0.9622,
+      "step": 4150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.79883861541748,
+      "learning_rate": 1.9314398489788915e-05,
+      "loss": 0.8826,
+      "step": 4151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.391386985778809,
+      "learning_rate": 1.9311824266346318e-05,
+      "loss": 1.1381,
+      "step": 4152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.61949348449707,
+      "learning_rate": 1.9309250042903725e-05,
+      "loss": 0.7949,
+      "step": 4153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.430593490600586,
+      "learning_rate": 1.9306675819461128e-05,
+      "loss": 0.7028,
+      "step": 4154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.597196578979492,
+      "learning_rate": 1.9304101596018535e-05,
+      "loss": 0.8895,
+      "step": 4155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.876840591430664,
+      "learning_rate": 1.9301527372575938e-05,
+      "loss": 0.9208,
+      "step": 4156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.680956840515137,
+      "learning_rate": 1.9298953149133345e-05,
+      "loss": 0.6098,
+      "step": 4157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.585216522216797,
+      "learning_rate": 1.929637892569075e-05,
+      "loss": 0.7811,
+      "step": 4158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.264379501342773,
+      "learning_rate": 1.9293804702248158e-05,
+      "loss": 0.7338,
+      "step": 4159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.872125625610352,
+      "learning_rate": 1.929123047880556e-05,
+      "loss": 0.7388,
+      "step": 4160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.399907112121582,
+      "learning_rate": 1.9288656255362968e-05,
+      "loss": 0.9994,
+      "step": 4161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.19143295288086,
+      "learning_rate": 1.928608203192037e-05,
+      "loss": 0.7663,
+      "step": 4162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.973332405090332,
+      "learning_rate": 1.9283507808477774e-05,
+      "loss": 0.7019,
+      "step": 4163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.952611923217773,
+      "learning_rate": 1.928093358503518e-05,
+      "loss": 0.8369,
+      "step": 4164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.02942943572998,
+      "learning_rate": 1.9278359361592584e-05,
+      "loss": 0.6435,
+      "step": 4165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.73558521270752,
+      "learning_rate": 1.927578513814999e-05,
+      "loss": 0.7812,
+      "step": 4166
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.565032958984375,
+      "learning_rate": 1.9273210914707398e-05,
+      "loss": 0.8437,
+      "step": 4167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.904428482055664,
+      "learning_rate": 1.9270636691264805e-05,
+      "loss": 1.0478,
+      "step": 4168
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.629289627075195,
+      "learning_rate": 1.9268062467822208e-05,
+      "loss": 1.1011,
+      "step": 4169
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.416136741638184,
+      "learning_rate": 1.9265488244379614e-05,
+      "loss": 1.1382,
+      "step": 4170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.459004402160645,
+      "learning_rate": 1.9262914020937018e-05,
+      "loss": 0.7723,
+      "step": 4171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.841681480407715,
+      "learning_rate": 1.9260339797494424e-05,
+      "loss": 0.857,
+      "step": 4172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.328619003295898,
+      "learning_rate": 1.9257765574051828e-05,
+      "loss": 0.5077,
+      "step": 4173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.433027267456055,
+      "learning_rate": 1.925519135060923e-05,
+      "loss": 0.5644,
+      "step": 4174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.906169891357422,
+      "learning_rate": 1.9252617127166638e-05,
+      "loss": 0.8372,
+      "step": 4175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.936905860900879,
+      "learning_rate": 1.925004290372404e-05,
+      "loss": 0.7456,
+      "step": 4176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.933682441711426,
+      "learning_rate": 1.924746868028145e-05,
+      "loss": 0.6806,
+      "step": 4177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.381811141967773,
+      "learning_rate": 1.9244894456838854e-05,
+      "loss": 0.7936,
+      "step": 4178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.274429321289062,
+      "learning_rate": 1.924232023339626e-05,
+      "loss": 0.9193,
+      "step": 4179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.535015106201172,
+      "learning_rate": 1.9239746009953664e-05,
+      "loss": 1.034,
+      "step": 4180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.031618118286133,
+      "learning_rate": 1.923717178651107e-05,
+      "loss": 1.0178,
+      "step": 4181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.387957572937012,
+      "learning_rate": 1.9234597563068474e-05,
+      "loss": 0.9654,
+      "step": 4182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.000457763671875,
+      "learning_rate": 1.923202333962588e-05,
+      "loss": 0.9824,
+      "step": 4183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.215188980102539,
+      "learning_rate": 1.9229449116183284e-05,
+      "loss": 0.5607,
+      "step": 4184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.841573715209961,
+      "learning_rate": 1.9226874892740688e-05,
+      "loss": 0.9767,
+      "step": 4185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.959955215454102,
+      "learning_rate": 1.9224300669298098e-05,
+      "loss": 0.8393,
+      "step": 4186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.130094528198242,
+      "learning_rate": 1.92217264458555e-05,
+      "loss": 0.9968,
+      "step": 4187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.533580780029297,
+      "learning_rate": 1.9219152222412908e-05,
+      "loss": 0.7348,
+      "step": 4188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.317439079284668,
+      "learning_rate": 1.921657799897031e-05,
+      "loss": 0.8343,
+      "step": 4189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.607730865478516,
+      "learning_rate": 1.9214003775527718e-05,
+      "loss": 0.8719,
+      "step": 4190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.656519889831543,
+      "learning_rate": 1.921142955208512e-05,
+      "loss": 0.7524,
+      "step": 4191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.75776195526123,
+      "learning_rate": 1.9208855328642528e-05,
+      "loss": 0.7508,
+      "step": 4192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.010034561157227,
+      "learning_rate": 1.920628110519993e-05,
+      "loss": 0.7258,
+      "step": 4193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.178448677062988,
+      "learning_rate": 1.9203706881757334e-05,
+      "loss": 0.8304,
+      "step": 4194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.157059669494629,
+      "learning_rate": 1.920113265831474e-05,
+      "loss": 0.9199,
+      "step": 4195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.935074806213379,
+      "learning_rate": 1.9198558434872148e-05,
+      "loss": 0.9531,
+      "step": 4196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.679344177246094,
+      "learning_rate": 1.9195984211429554e-05,
+      "loss": 0.6097,
+      "step": 4197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.078520774841309,
+      "learning_rate": 1.9193409987986958e-05,
+      "loss": 0.7061,
+      "step": 4198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.979924201965332,
+      "learning_rate": 1.9190835764544364e-05,
+      "loss": 0.8073,
+      "step": 4199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.267154693603516,
+      "learning_rate": 1.9188261541101767e-05,
+      "loss": 0.975,
+      "step": 4200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.881691932678223,
+      "learning_rate": 1.9185687317659174e-05,
+      "loss": 0.7223,
+      "step": 4201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.524392127990723,
+      "learning_rate": 1.9183113094216577e-05,
+      "loss": 0.7263,
+      "step": 4202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.83057975769043,
+      "learning_rate": 1.9180538870773984e-05,
+      "loss": 0.7493,
+      "step": 4203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.009383201599121,
+      "learning_rate": 1.9177964647331387e-05,
+      "loss": 0.5886,
+      "step": 4204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.207352638244629,
+      "learning_rate": 1.9175390423888794e-05,
+      "loss": 1.0468,
+      "step": 4205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.457111358642578,
+      "learning_rate": 1.91728162004462e-05,
+      "loss": 0.9551,
+      "step": 4206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.947735786437988,
+      "learning_rate": 1.9170241977003604e-05,
+      "loss": 0.8575,
+      "step": 4207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.02142333984375,
+      "learning_rate": 1.916766775356101e-05,
+      "loss": 0.8909,
+      "step": 4208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.354259490966797,
+      "learning_rate": 1.9165093530118414e-05,
+      "loss": 0.8369,
+      "step": 4209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.476853370666504,
+      "learning_rate": 1.916251930667582e-05,
+      "loss": 0.7134,
+      "step": 4210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.338850021362305,
+      "learning_rate": 1.9159945083233224e-05,
+      "loss": 0.8601,
+      "step": 4211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.50589370727539,
+      "learning_rate": 1.915737085979063e-05,
+      "loss": 0.834,
+      "step": 4212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.211137771606445,
+      "learning_rate": 1.9154796636348034e-05,
+      "loss": 0.9652,
+      "step": 4213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.985767841339111,
+      "learning_rate": 1.915222241290544e-05,
+      "loss": 0.8474,
+      "step": 4214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.327977180480957,
+      "learning_rate": 1.9149648189462847e-05,
+      "loss": 0.8398,
+      "step": 4215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.896589279174805,
+      "learning_rate": 1.914707396602025e-05,
+      "loss": 0.8843,
+      "step": 4216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.053586959838867,
+      "learning_rate": 1.9144499742577657e-05,
+      "loss": 0.8559,
+      "step": 4217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.480340957641602,
+      "learning_rate": 1.914192551913506e-05,
+      "loss": 0.9843,
+      "step": 4218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.425511360168457,
+      "learning_rate": 1.9139351295692467e-05,
+      "loss": 0.8868,
+      "step": 4219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.773504257202148,
+      "learning_rate": 1.913677707224987e-05,
+      "loss": 0.7274,
+      "step": 4220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.101371765136719,
+      "learning_rate": 1.9134202848807277e-05,
+      "loss": 0.829,
+      "step": 4221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.200393676757812,
+      "learning_rate": 1.913162862536468e-05,
+      "loss": 0.8919,
+      "step": 4222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.389777183532715,
+      "learning_rate": 1.9129054401922087e-05,
+      "loss": 0.7035,
+      "step": 4223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.115687370300293,
+      "learning_rate": 1.9126480178479494e-05,
+      "loss": 0.817,
+      "step": 4224
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.318549156188965,
+      "learning_rate": 1.9123905955036897e-05,
+      "loss": 0.5087,
+      "step": 4225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.522407531738281,
+      "learning_rate": 1.9121331731594304e-05,
+      "loss": 0.8161,
+      "step": 4226
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.159699440002441,
+      "learning_rate": 1.9118757508151707e-05,
+      "loss": 0.9951,
+      "step": 4227
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.890766143798828,
+      "learning_rate": 1.9116183284709114e-05,
+      "loss": 0.6243,
+      "step": 4228
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.272420883178711,
+      "learning_rate": 1.9113609061266517e-05,
+      "loss": 0.9083,
+      "step": 4229
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.610692024230957,
+      "learning_rate": 1.9111034837823924e-05,
+      "loss": 1.009,
+      "step": 4230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.850475311279297,
+      "learning_rate": 1.9108460614381327e-05,
+      "loss": 0.7239,
+      "step": 4231
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.294280052185059,
+      "learning_rate": 1.9105886390938734e-05,
+      "loss": 0.8397,
+      "step": 4232
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.980622291564941,
+      "learning_rate": 1.9103312167496137e-05,
+      "loss": 0.9628,
+      "step": 4233
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.041665077209473,
+      "learning_rate": 1.9100737944053547e-05,
+      "loss": 0.7433,
+      "step": 4234
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.734867095947266,
+      "learning_rate": 1.909816372061095e-05,
+      "loss": 0.7507,
+      "step": 4235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.993006706237793,
+      "learning_rate": 1.9095589497168354e-05,
+      "loss": 0.9328,
+      "step": 4236
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.371488571166992,
+      "learning_rate": 1.909301527372576e-05,
+      "loss": 1.0895,
+      "step": 4237
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.333434104919434,
+      "learning_rate": 1.9090441050283164e-05,
+      "loss": 0.8659,
+      "step": 4238
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.16541862487793,
+      "learning_rate": 1.908786682684057e-05,
+      "loss": 0.8717,
+      "step": 4239
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.871678352355957,
+      "learning_rate": 1.9085292603397974e-05,
+      "loss": 0.646,
+      "step": 4240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.546360969543457,
+      "learning_rate": 1.908271837995538e-05,
+      "loss": 0.7511,
+      "step": 4241
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.895899772644043,
+      "learning_rate": 1.9080144156512784e-05,
+      "loss": 0.7912,
+      "step": 4242
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.175511360168457,
+      "learning_rate": 1.9077569933070194e-05,
+      "loss": 0.9611,
+      "step": 4243
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.024049758911133,
+      "learning_rate": 1.9074995709627597e-05,
+      "loss": 0.9508,
+      "step": 4244
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.078110694885254,
+      "learning_rate": 1.9072421486185004e-05,
+      "loss": 0.9578,
+      "step": 4245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.51116943359375,
+      "learning_rate": 1.9069847262742407e-05,
+      "loss": 1.1497,
+      "step": 4246
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.76651668548584,
+      "learning_rate": 1.906727303929981e-05,
+      "loss": 0.6663,
+      "step": 4247
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.659747123718262,
+      "learning_rate": 1.9064698815857217e-05,
+      "loss": 0.8842,
+      "step": 4248
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.561182022094727,
+      "learning_rate": 1.906212459241462e-05,
+      "loss": 0.8444,
+      "step": 4249
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.045940399169922,
+      "learning_rate": 1.9059550368972027e-05,
+      "loss": 0.9408,
+      "step": 4250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.490055084228516,
+      "learning_rate": 1.905697614552943e-05,
+      "loss": 0.9343,
+      "step": 4251
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.027691841125488,
+      "learning_rate": 1.9054401922086837e-05,
+      "loss": 0.6872,
+      "step": 4252
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.355417251586914,
+      "learning_rate": 1.9051827698644244e-05,
+      "loss": 0.5568,
+      "step": 4253
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.123159408569336,
+      "learning_rate": 1.904925347520165e-05,
+      "loss": 0.7632,
+      "step": 4254
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.372822761535645,
+      "learning_rate": 1.9046679251759054e-05,
+      "loss": 0.8914,
+      "step": 4255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.501077651977539,
+      "learning_rate": 1.9044105028316457e-05,
+      "loss": 0.824,
+      "step": 4256
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.116118431091309,
+      "learning_rate": 1.9041530804873864e-05,
+      "loss": 1.2319,
+      "step": 4257
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.635515213012695,
+      "learning_rate": 1.9038956581431267e-05,
+      "loss": 0.7783,
+      "step": 4258
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.615957260131836,
+      "learning_rate": 1.9036382357988674e-05,
+      "loss": 0.5975,
+      "step": 4259
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.458331108093262,
+      "learning_rate": 1.9033808134546077e-05,
+      "loss": 0.7164,
+      "step": 4260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.402206420898438,
+      "learning_rate": 1.9031233911103484e-05,
+      "loss": 0.7424,
+      "step": 4261
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.506240844726562,
+      "learning_rate": 1.902865968766089e-05,
+      "loss": 1.1194,
+      "step": 4262
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.461847305297852,
+      "learning_rate": 1.9026085464218297e-05,
+      "loss": 0.7483,
+      "step": 4263
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.207564353942871,
+      "learning_rate": 1.90235112407757e-05,
+      "loss": 0.6526,
+      "step": 4264
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.552288055419922,
+      "learning_rate": 1.9020937017333107e-05,
+      "loss": 1.1085,
+      "step": 4265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.87618637084961,
+      "learning_rate": 1.901836279389051e-05,
+      "loss": 0.7648,
+      "step": 4266
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.346138954162598,
+      "learning_rate": 1.9015788570447914e-05,
+      "loss": 0.7646,
+      "step": 4267
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.180513381958008,
+      "learning_rate": 1.901321434700532e-05,
+      "loss": 0.7532,
+      "step": 4268
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.566898345947266,
+      "learning_rate": 1.9010640123562723e-05,
+      "loss": 0.8056,
+      "step": 4269
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.16606616973877,
+      "learning_rate": 1.900806590012013e-05,
+      "loss": 0.7936,
+      "step": 4270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.111207962036133,
+      "learning_rate": 1.9005491676677537e-05,
+      "loss": 0.614,
+      "step": 4271
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.534924507141113,
+      "learning_rate": 1.9002917453234944e-05,
+      "loss": 1.2625,
+      "step": 4272
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.890825271606445,
+      "learning_rate": 1.9000343229792347e-05,
+      "loss": 0.8264,
+      "step": 4273
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.985905647277832,
+      "learning_rate": 1.8997769006349754e-05,
+      "loss": 0.7684,
+      "step": 4274
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.069395065307617,
+      "learning_rate": 1.8995194782907157e-05,
+      "loss": 1.0358,
+      "step": 4275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.809677124023438,
+      "learning_rate": 1.8992620559464563e-05,
+      "loss": 0.8179,
+      "step": 4276
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.157917022705078,
+      "learning_rate": 1.8990046336021967e-05,
+      "loss": 0.6746,
+      "step": 4277
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.686842918395996,
+      "learning_rate": 1.898747211257937e-05,
+      "loss": 0.8305,
+      "step": 4278
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.763129234313965,
+      "learning_rate": 1.8984897889136777e-05,
+      "loss": 1.1049,
+      "step": 4279
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.007186889648438,
+      "learning_rate": 1.898232366569418e-05,
+      "loss": 0.85,
+      "step": 4280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.028447151184082,
+      "learning_rate": 1.897974944225159e-05,
+      "loss": 0.6868,
+      "step": 4281
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.202744483947754,
+      "learning_rate": 1.8977175218808993e-05,
+      "loss": 0.9465,
+      "step": 4282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.139639854431152,
+      "learning_rate": 1.89746009953664e-05,
+      "loss": 1.0126,
+      "step": 4283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.211018562316895,
+      "learning_rate": 1.8972026771923803e-05,
+      "loss": 1.0667,
+      "step": 4284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.292472839355469,
+      "learning_rate": 1.896945254848121e-05,
+      "loss": 0.6016,
+      "step": 4285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.162647247314453,
+      "learning_rate": 1.8966878325038613e-05,
+      "loss": 0.934,
+      "step": 4286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.490167617797852,
+      "learning_rate": 1.896430410159602e-05,
+      "loss": 0.7284,
+      "step": 4287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.79624080657959,
+      "learning_rate": 1.8961729878153423e-05,
+      "loss": 0.8251,
+      "step": 4288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.193233489990234,
+      "learning_rate": 1.8959155654710827e-05,
+      "loss": 0.7528,
+      "step": 4289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.678086280822754,
+      "learning_rate": 1.8956581431268237e-05,
+      "loss": 0.8643,
+      "step": 4290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.563193321228027,
+      "learning_rate": 1.895400720782564e-05,
+      "loss": 0.9623,
+      "step": 4291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.784810066223145,
+      "learning_rate": 1.8951432984383047e-05,
+      "loss": 0.775,
+      "step": 4292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.589600563049316,
+      "learning_rate": 1.894885876094045e-05,
+      "loss": 0.7281,
+      "step": 4293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.031818389892578,
+      "learning_rate": 1.8946284537497857e-05,
+      "loss": 0.5981,
+      "step": 4294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.341252326965332,
+      "learning_rate": 1.894371031405526e-05,
+      "loss": 0.9253,
+      "step": 4295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.869950294494629,
+      "learning_rate": 1.8941136090612667e-05,
+      "loss": 0.8761,
+      "step": 4296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.53954029083252,
+      "learning_rate": 1.893856186717007e-05,
+      "loss": 0.5906,
+      "step": 4297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.133573532104492,
+      "learning_rate": 1.8935987643727473e-05,
+      "loss": 0.6317,
+      "step": 4298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.367504119873047,
+      "learning_rate": 1.893341342028488e-05,
+      "loss": 0.9038,
+      "step": 4299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.127020835876465,
+      "learning_rate": 1.8930839196842287e-05,
+      "loss": 0.9277,
+      "step": 4300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.476219177246094,
+      "learning_rate": 1.8928264973399693e-05,
+      "loss": 0.8453,
+      "step": 4301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.979869842529297,
+      "learning_rate": 1.8925690749957097e-05,
+      "loss": 0.8248,
+      "step": 4302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.847350120544434,
+      "learning_rate": 1.8923116526514503e-05,
+      "loss": 0.7173,
+      "step": 4303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.573976516723633,
+      "learning_rate": 1.8920542303071907e-05,
+      "loss": 1.047,
+      "step": 4304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.462332725524902,
+      "learning_rate": 1.8917968079629313e-05,
+      "loss": 0.6791,
+      "step": 4305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.317166328430176,
+      "learning_rate": 1.8915393856186716e-05,
+      "loss": 0.7915,
+      "step": 4306
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.177624702453613,
+      "learning_rate": 1.8912819632744123e-05,
+      "loss": 0.528,
+      "step": 4307
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.47716236114502,
+      "learning_rate": 1.8910245409301526e-05,
+      "loss": 0.6932,
+      "step": 4308
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.621345520019531,
+      "learning_rate": 1.8907671185858933e-05,
+      "loss": 0.5768,
+      "step": 4309
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.76716423034668,
+      "learning_rate": 1.890509696241634e-05,
+      "loss": 0.7745,
+      "step": 4310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.033775329589844,
+      "learning_rate": 1.8902522738973743e-05,
+      "loss": 0.9711,
+      "step": 4311
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.942069053649902,
+      "learning_rate": 1.889994851553115e-05,
+      "loss": 0.6979,
+      "step": 4312
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.966338157653809,
+      "learning_rate": 1.8897374292088553e-05,
+      "loss": 0.8136,
+      "step": 4313
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.818129539489746,
+      "learning_rate": 1.889480006864596e-05,
+      "loss": 0.8066,
+      "step": 4314
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.387496948242188,
+      "learning_rate": 1.8892225845203363e-05,
+      "loss": 0.6352,
+      "step": 4315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.305728912353516,
+      "learning_rate": 1.888965162176077e-05,
+      "loss": 0.6251,
+      "step": 4316
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.720781326293945,
+      "learning_rate": 1.8887077398318173e-05,
+      "loss": 0.837,
+      "step": 4317
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.153433799743652,
+      "learning_rate": 1.888450317487558e-05,
+      "loss": 1.0185,
+      "step": 4318
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.407583236694336,
+      "learning_rate": 1.8881928951432986e-05,
+      "loss": 0.6867,
+      "step": 4319
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.725436210632324,
+      "learning_rate": 1.887935472799039e-05,
+      "loss": 0.6625,
+      "step": 4320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.523588180541992,
+      "learning_rate": 1.8876780504547796e-05,
+      "loss": 0.7459,
+      "step": 4321
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.071678161621094,
+      "learning_rate": 1.88742062811052e-05,
+      "loss": 0.6611,
+      "step": 4322
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.482538223266602,
+      "learning_rate": 1.8871632057662606e-05,
+      "loss": 0.8529,
+      "step": 4323
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.277457237243652,
+      "learning_rate": 1.886905783422001e-05,
+      "loss": 0.636,
+      "step": 4324
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.441004753112793,
+      "learning_rate": 1.8866483610777416e-05,
+      "loss": 0.7508,
+      "step": 4325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.06314468383789,
+      "learning_rate": 1.886390938733482e-05,
+      "loss": 0.6786,
+      "step": 4326
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.114742279052734,
+      "learning_rate": 1.8861335163892226e-05,
+      "loss": 0.9106,
+      "step": 4327
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.636996269226074,
+      "learning_rate": 1.8858760940449633e-05,
+      "loss": 1.137,
+      "step": 4328
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.75655460357666,
+      "learning_rate": 1.8856186717007036e-05,
+      "loss": 1.2952,
+      "step": 4329
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.261377334594727,
+      "learning_rate": 1.8853612493564443e-05,
+      "loss": 0.6218,
+      "step": 4330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.316879272460938,
+      "learning_rate": 1.8851038270121846e-05,
+      "loss": 0.7588,
+      "step": 4331
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.593915939331055,
+      "learning_rate": 1.8848464046679253e-05,
+      "loss": 0.7051,
+      "step": 4332
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.073686599731445,
+      "learning_rate": 1.8845889823236656e-05,
+      "loss": 0.8409,
+      "step": 4333
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.310689926147461,
+      "learning_rate": 1.8843315599794063e-05,
+      "loss": 0.927,
+      "step": 4334
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.633893013000488,
+      "learning_rate": 1.8840741376351466e-05,
+      "loss": 0.8132,
+      "step": 4335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.84637451171875,
+      "learning_rate": 1.8838167152908873e-05,
+      "loss": 0.7943,
+      "step": 4336
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.861120223999023,
+      "learning_rate": 1.8835592929466276e-05,
+      "loss": 0.7672,
+      "step": 4337
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.433868408203125,
+      "learning_rate": 1.8833018706023686e-05,
+      "loss": 0.6974,
+      "step": 4338
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.834355354309082,
+      "learning_rate": 1.883044448258109e-05,
+      "loss": 0.72,
+      "step": 4339
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.720258712768555,
+      "learning_rate": 1.8827870259138493e-05,
+      "loss": 0.9226,
+      "step": 4340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.472895622253418,
+      "learning_rate": 1.88252960356959e-05,
+      "loss": 0.8724,
+      "step": 4341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.831478118896484,
+      "learning_rate": 1.8822721812253303e-05,
+      "loss": 0.8691,
+      "step": 4342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.863024711608887,
+      "learning_rate": 1.882014758881071e-05,
+      "loss": 0.8022,
+      "step": 4343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.083821296691895,
+      "learning_rate": 1.8817573365368113e-05,
+      "loss": 0.758,
+      "step": 4344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.028779983520508,
+      "learning_rate": 1.881499914192552e-05,
+      "loss": 0.7373,
+      "step": 4345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.553868293762207,
+      "learning_rate": 1.8812424918482923e-05,
+      "loss": 0.654,
+      "step": 4346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.712279319763184,
+      "learning_rate": 1.8809850695040333e-05,
+      "loss": 0.7303,
+      "step": 4347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.865372657775879,
+      "learning_rate": 1.8807276471597736e-05,
+      "loss": 1.0221,
+      "step": 4348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.506478309631348,
+      "learning_rate": 1.8804702248155143e-05,
+      "loss": 0.7201,
+      "step": 4349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.890604019165039,
+      "learning_rate": 1.8802128024712546e-05,
+      "loss": 0.7847,
+      "step": 4350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.948525428771973,
+      "learning_rate": 1.879955380126995e-05,
+      "loss": 0.5751,
+      "step": 4351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.099681854248047,
+      "learning_rate": 1.8796979577827356e-05,
+      "loss": 0.8288,
+      "step": 4352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.870171546936035,
+      "learning_rate": 1.879440535438476e-05,
+      "loss": 1.0622,
+      "step": 4353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.078024864196777,
+      "learning_rate": 1.8791831130942166e-05,
+      "loss": 0.9769,
+      "step": 4354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.907835960388184,
+      "learning_rate": 1.878925690749957e-05,
+      "loss": 0.9277,
+      "step": 4355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.870816230773926,
+      "learning_rate": 1.8786682684056976e-05,
+      "loss": 0.7616,
+      "step": 4356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.830236434936523,
+      "learning_rate": 1.8784108460614383e-05,
+      "loss": 0.7201,
+      "step": 4357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.235135078430176,
+      "learning_rate": 1.878153423717179e-05,
+      "loss": 0.8442,
+      "step": 4358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.954062461853027,
+      "learning_rate": 1.8778960013729193e-05,
+      "loss": 0.6129,
+      "step": 4359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.778276443481445,
+      "learning_rate": 1.8776385790286596e-05,
+      "loss": 1.0534,
+      "step": 4360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.113384246826172,
+      "learning_rate": 1.8773811566844003e-05,
+      "loss": 0.7507,
+      "step": 4361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.1980390548706055,
+      "learning_rate": 1.8771237343401406e-05,
+      "loss": 0.4489,
+      "step": 4362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.564810752868652,
+      "learning_rate": 1.8768663119958813e-05,
+      "loss": 0.8186,
+      "step": 4363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.596735954284668,
+      "learning_rate": 1.8766088896516216e-05,
+      "loss": 0.5342,
+      "step": 4364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.781926155090332,
+      "learning_rate": 1.8763514673073623e-05,
+      "loss": 1.0391,
+      "step": 4365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.596892356872559,
+      "learning_rate": 1.876094044963103e-05,
+      "loss": 0.8483,
+      "step": 4366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.084346771240234,
+      "learning_rate": 1.8758366226188436e-05,
+      "loss": 0.7646,
+      "step": 4367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.50231647491455,
+      "learning_rate": 1.875579200274584e-05,
+      "loss": 0.8787,
+      "step": 4368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.484762191772461,
+      "learning_rate": 1.8753217779303246e-05,
+      "loss": 0.7781,
+      "step": 4369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.743453979492188,
+      "learning_rate": 1.875064355586065e-05,
+      "loss": 0.9012,
+      "step": 4370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.534058570861816,
+      "learning_rate": 1.8748069332418053e-05,
+      "loss": 0.9213,
+      "step": 4371
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.27253246307373,
+      "learning_rate": 1.874549510897546e-05,
+      "loss": 0.7629,
+      "step": 4372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.572549819946289,
+      "learning_rate": 1.8742920885532862e-05,
+      "loss": 0.7689,
+      "step": 4373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.57839298248291,
+      "learning_rate": 1.874034666209027e-05,
+      "loss": 0.9119,
+      "step": 4374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.892982482910156,
+      "learning_rate": 1.8737772438647676e-05,
+      "loss": 0.9118,
+      "step": 4375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.607345581054688,
+      "learning_rate": 1.8735198215205083e-05,
+      "loss": 0.6165,
+      "step": 4376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.54564094543457,
+      "learning_rate": 1.8732623991762486e-05,
+      "loss": 0.9474,
+      "step": 4377
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.732884407043457,
+      "learning_rate": 1.8730049768319893e-05,
+      "loss": 0.8056,
+      "step": 4378
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.062182426452637,
+      "learning_rate": 1.8727475544877296e-05,
+      "loss": 0.9305,
+      "step": 4379
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.44251537322998,
+      "learning_rate": 1.8724901321434702e-05,
+      "loss": 0.9487,
+      "step": 4380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.254076957702637,
+      "learning_rate": 1.8722327097992106e-05,
+      "loss": 0.8822,
+      "step": 4381
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.709226608276367,
+      "learning_rate": 1.871975287454951e-05,
+      "loss": 0.7476,
+      "step": 4382
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.419149398803711,
+      "learning_rate": 1.8717178651106916e-05,
+      "loss": 0.7381,
+      "step": 4383
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.5823392868042,
+      "learning_rate": 1.871460442766432e-05,
+      "loss": 0.6735,
+      "step": 4384
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.25571060180664,
+      "learning_rate": 1.871203020422173e-05,
+      "loss": 0.78,
+      "step": 4385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.075605392456055,
+      "learning_rate": 1.8709455980779132e-05,
+      "loss": 0.9736,
+      "step": 4386
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.6416754722595215,
+      "learning_rate": 1.870688175733654e-05,
+      "loss": 0.6801,
+      "step": 4387
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.600601196289062,
+      "learning_rate": 1.8704307533893942e-05,
+      "loss": 0.8792,
+      "step": 4388
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.597159385681152,
+      "learning_rate": 1.870173331045135e-05,
+      "loss": 0.723,
+      "step": 4389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.653776168823242,
+      "learning_rate": 1.8699159087008752e-05,
+      "loss": 1.0082,
+      "step": 4390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.68677806854248,
+      "learning_rate": 1.869658486356616e-05,
+      "loss": 1.106,
+      "step": 4391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.043035507202148,
+      "learning_rate": 1.8694010640123562e-05,
+      "loss": 0.7793,
+      "step": 4392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.986228942871094,
+      "learning_rate": 1.8691436416680966e-05,
+      "loss": 0.8221,
+      "step": 4393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.062551498413086,
+      "learning_rate": 1.8688862193238376e-05,
+      "loss": 0.8659,
+      "step": 4394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.912751197814941,
+      "learning_rate": 1.868628796979578e-05,
+      "loss": 0.756,
+      "step": 4395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.395503044128418,
+      "learning_rate": 1.8683713746353186e-05,
+      "loss": 0.9355,
+      "step": 4396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.25494384765625,
+      "learning_rate": 1.868113952291059e-05,
+      "loss": 0.7836,
+      "step": 4397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.707947731018066,
+      "learning_rate": 1.8678565299467996e-05,
+      "loss": 0.9053,
+      "step": 4398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.001806259155273,
+      "learning_rate": 1.86759910760254e-05,
+      "loss": 0.7759,
+      "step": 4399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.141298294067383,
+      "learning_rate": 1.8673416852582806e-05,
+      "loss": 0.5919,
+      "step": 4400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.65458869934082,
+      "learning_rate": 1.867084262914021e-05,
+      "loss": 1.0048,
+      "step": 4401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.322049140930176,
+      "learning_rate": 1.8668268405697612e-05,
+      "loss": 0.8787,
+      "step": 4402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.761757850646973,
+      "learning_rate": 1.866569418225502e-05,
+      "loss": 1.0129,
+      "step": 4403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.715259552001953,
+      "learning_rate": 1.8663119958812426e-05,
+      "loss": 0.7009,
+      "step": 4404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.733769416809082,
+      "learning_rate": 1.8660545735369832e-05,
+      "loss": 0.9702,
+      "step": 4405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.306440353393555,
+      "learning_rate": 1.8657971511927236e-05,
+      "loss": 0.9039,
+      "step": 4406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.9626145362854,
+      "learning_rate": 1.8655397288484642e-05,
+      "loss": 0.7413,
+      "step": 4407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.926734924316406,
+      "learning_rate": 1.8652823065042046e-05,
+      "loss": 0.8488,
+      "step": 4408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.632416725158691,
+      "learning_rate": 1.8650248841599452e-05,
+      "loss": 0.6175,
+      "step": 4409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.221813201904297,
+      "learning_rate": 1.8647674618156855e-05,
+      "loss": 0.8773,
+      "step": 4410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.929887771606445,
+      "learning_rate": 1.8645100394714262e-05,
+      "loss": 0.638,
+      "step": 4411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.351088523864746,
+      "learning_rate": 1.8642526171271665e-05,
+      "loss": 0.756,
+      "step": 4412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.052444458007812,
+      "learning_rate": 1.8639951947829072e-05,
+      "loss": 1.0514,
+      "step": 4413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.6525297164917,
+      "learning_rate": 1.863737772438648e-05,
+      "loss": 0.7541,
+      "step": 4414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.138541221618652,
+      "learning_rate": 1.8634803500943882e-05,
+      "loss": 0.7851,
+      "step": 4415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.111345291137695,
+      "learning_rate": 1.863222927750129e-05,
+      "loss": 0.884,
+      "step": 4416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.87191104888916,
+      "learning_rate": 1.8629655054058692e-05,
+      "loss": 1.0476,
+      "step": 4417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.426712036132812,
+      "learning_rate": 1.86270808306161e-05,
+      "loss": 0.8644,
+      "step": 4418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.35315990447998,
+      "learning_rate": 1.8624506607173502e-05,
+      "loss": 0.9767,
+      "step": 4419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.673909187316895,
+      "learning_rate": 1.862193238373091e-05,
+      "loss": 0.9786,
+      "step": 4420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.761059761047363,
+      "learning_rate": 1.8619358160288312e-05,
+      "loss": 0.7912,
+      "step": 4421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.249832153320312,
+      "learning_rate": 1.861678393684572e-05,
+      "loss": 0.8706,
+      "step": 4422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.88196849822998,
+      "learning_rate": 1.8614209713403125e-05,
+      "loss": 0.7316,
+      "step": 4423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.350408554077148,
+      "learning_rate": 1.861163548996053e-05,
+      "loss": 0.7068,
+      "step": 4424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.889779090881348,
+      "learning_rate": 1.8609061266517935e-05,
+      "loss": 0.9263,
+      "step": 4425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.010964393615723,
+      "learning_rate": 1.860648704307534e-05,
+      "loss": 1.3115,
+      "step": 4426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.4732666015625,
+      "learning_rate": 1.8603912819632745e-05,
+      "loss": 0.7906,
+      "step": 4427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.549907684326172,
+      "learning_rate": 1.860133859619015e-05,
+      "loss": 0.781,
+      "step": 4428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.509839057922363,
+      "learning_rate": 1.8598764372747555e-05,
+      "loss": 0.647,
+      "step": 4429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.75,
+      "learning_rate": 1.859619014930496e-05,
+      "loss": 0.7365,
+      "step": 4430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.616283416748047,
+      "learning_rate": 1.8593615925862365e-05,
+      "loss": 0.9994,
+      "step": 4431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.0025634765625,
+      "learning_rate": 1.8591041702419772e-05,
+      "loss": 0.7246,
+      "step": 4432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.842690467834473,
+      "learning_rate": 1.8588467478977175e-05,
+      "loss": 0.8649,
+      "step": 4433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.467083930969238,
+      "learning_rate": 1.8585893255534582e-05,
+      "loss": 0.6283,
+      "step": 4434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.707691192626953,
+      "learning_rate": 1.8583319032091985e-05,
+      "loss": 0.8441,
+      "step": 4435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.234848976135254,
+      "learning_rate": 1.8580744808649392e-05,
+      "loss": 1.0229,
+      "step": 4436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.026167869567871,
+      "learning_rate": 1.8578170585206795e-05,
+      "loss": 0.95,
+      "step": 4437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.936605453491211,
+      "learning_rate": 1.8575596361764202e-05,
+      "loss": 0.9674,
+      "step": 4438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.654797554016113,
+      "learning_rate": 1.8573022138321605e-05,
+      "loss": 0.5985,
+      "step": 4439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.77736759185791,
+      "learning_rate": 1.8570447914879012e-05,
+      "loss": 0.8257,
+      "step": 4440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.64514446258545,
+      "learning_rate": 1.8567873691436415e-05,
+      "loss": 0.9767,
+      "step": 4441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.730393409729004,
+      "learning_rate": 1.8565299467993825e-05,
+      "loss": 0.8074,
+      "step": 4442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.7951078414917,
+      "learning_rate": 1.856272524455123e-05,
+      "loss": 0.5972,
+      "step": 4443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.156877517700195,
+      "learning_rate": 1.8560151021108632e-05,
+      "loss": 0.6072,
+      "step": 4444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.159260749816895,
+      "learning_rate": 1.855757679766604e-05,
+      "loss": 0.8721,
+      "step": 4445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.914385795593262,
+      "learning_rate": 1.8555002574223442e-05,
+      "loss": 0.8044,
+      "step": 4446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.912651062011719,
+      "learning_rate": 1.855242835078085e-05,
+      "loss": 0.7537,
+      "step": 4447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.371186256408691,
+      "learning_rate": 1.8549854127338252e-05,
+      "loss": 0.6967,
+      "step": 4448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.682121276855469,
+      "learning_rate": 1.854727990389566e-05,
+      "loss": 1.0528,
+      "step": 4449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.716703414916992,
+      "learning_rate": 1.8544705680453062e-05,
+      "loss": 0.8986,
+      "step": 4450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.13463020324707,
+      "learning_rate": 1.8542131457010472e-05,
+      "loss": 0.7043,
+      "step": 4451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.702315330505371,
+      "learning_rate": 1.8539557233567875e-05,
+      "loss": 0.8191,
+      "step": 4452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.548389434814453,
+      "learning_rate": 1.8536983010125282e-05,
+      "loss": 0.5203,
+      "step": 4453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.226734161376953,
+      "learning_rate": 1.8534408786682685e-05,
+      "loss": 0.6779,
+      "step": 4454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.5157470703125,
+      "learning_rate": 1.853183456324009e-05,
+      "loss": 0.695,
+      "step": 4455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.56718635559082,
+      "learning_rate": 1.8529260339797495e-05,
+      "loss": 0.7932,
+      "step": 4456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.645631790161133,
+      "learning_rate": 1.85266861163549e-05,
+      "loss": 0.8119,
+      "step": 4457
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.036974906921387,
+      "learning_rate": 1.8524111892912305e-05,
+      "loss": 0.7608,
+      "step": 4458
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.421669960021973,
+      "learning_rate": 1.852153766946971e-05,
+      "loss": 0.7267,
+      "step": 4459
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.169737815856934,
+      "learning_rate": 1.8518963446027115e-05,
+      "loss": 0.5726,
+      "step": 4460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.590944290161133,
+      "learning_rate": 1.8516389222584522e-05,
+      "loss": 0.8535,
+      "step": 4461
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.078580856323242,
+      "learning_rate": 1.851381499914193e-05,
+      "loss": 0.7788,
+      "step": 4462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.609288215637207,
+      "learning_rate": 1.851124077569933e-05,
+      "loss": 0.9815,
+      "step": 4463
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.61540412902832,
+      "learning_rate": 1.8508666552256735e-05,
+      "loss": 0.817,
+      "step": 4464
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.372817039489746,
+      "learning_rate": 1.850609232881414e-05,
+      "loss": 0.9686,
+      "step": 4465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.059974670410156,
+      "learning_rate": 1.8503518105371545e-05,
+      "loss": 0.6221,
+      "step": 4466
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.431178092956543,
+      "learning_rate": 1.850094388192895e-05,
+      "loss": 0.624,
+      "step": 4467
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.192831039428711,
+      "learning_rate": 1.8498369658486355e-05,
+      "loss": 0.7813,
+      "step": 4468
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.22218132019043,
+      "learning_rate": 1.849579543504376e-05,
+      "loss": 0.9372,
+      "step": 4469
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.294771194458008,
+      "learning_rate": 1.8493221211601168e-05,
+      "loss": 0.6635,
+      "step": 4470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.357868194580078,
+      "learning_rate": 1.8490646988158575e-05,
+      "loss": 0.8066,
+      "step": 4471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.630175590515137,
+      "learning_rate": 1.8488072764715978e-05,
+      "loss": 0.8054,
+      "step": 4472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.369306564331055,
+      "learning_rate": 1.8485498541273385e-05,
+      "loss": 0.9648,
+      "step": 4473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.858946800231934,
+      "learning_rate": 1.8482924317830788e-05,
+      "loss": 0.9286,
+      "step": 4474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.09499740600586,
+      "learning_rate": 1.848035009438819e-05,
+      "loss": 0.7685,
+      "step": 4475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.864041328430176,
+      "learning_rate": 1.8477775870945598e-05,
+      "loss": 0.834,
+      "step": 4476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.916614532470703,
+      "learning_rate": 1.8475201647503e-05,
+      "loss": 0.7069,
+      "step": 4477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.010899543762207,
+      "learning_rate": 1.8472627424060408e-05,
+      "loss": 0.8302,
+      "step": 4478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.282487869262695,
+      "learning_rate": 1.847005320061781e-05,
+      "loss": 0.9146,
+      "step": 4479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.03085994720459,
+      "learning_rate": 1.846747897717522e-05,
+      "loss": 0.8117,
+      "step": 4480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.702635765075684,
+      "learning_rate": 1.8464904753732625e-05,
+      "loss": 0.5839,
+      "step": 4481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.291382789611816,
+      "learning_rate": 1.846233053029003e-05,
+      "loss": 0.7454,
+      "step": 4482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.000566482543945,
+      "learning_rate": 1.8459756306847435e-05,
+      "loss": 0.879,
+      "step": 4483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.346214294433594,
+      "learning_rate": 1.845718208340484e-05,
+      "loss": 1.0057,
+      "step": 4484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.30116081237793,
+      "learning_rate": 1.8454607859962245e-05,
+      "loss": 0.5507,
+      "step": 4485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.069503784179688,
+      "learning_rate": 1.8452033636519648e-05,
+      "loss": 0.6537,
+      "step": 4486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.520319938659668,
+      "learning_rate": 1.8449459413077055e-05,
+      "loss": 0.6122,
+      "step": 4487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.078132629394531,
+      "learning_rate": 1.8446885189634458e-05,
+      "loss": 0.6766,
+      "step": 4488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.916189193725586,
+      "learning_rate": 1.8444310966191868e-05,
+      "loss": 0.8757,
+      "step": 4489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.934427261352539,
+      "learning_rate": 1.844173674274927e-05,
+      "loss": 0.7369,
+      "step": 4490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.735384941101074,
+      "learning_rate": 1.8439162519306678e-05,
+      "loss": 0.9513,
+      "step": 4491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.546665191650391,
+      "learning_rate": 1.843658829586408e-05,
+      "loss": 0.514,
+      "step": 4492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.621626853942871,
+      "learning_rate": 1.8434014072421488e-05,
+      "loss": 0.8021,
+      "step": 4493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.1311616897583,
+      "learning_rate": 1.843143984897889e-05,
+      "loss": 0.8665,
+      "step": 4494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.149614334106445,
+      "learning_rate": 1.8428865625536298e-05,
+      "loss": 0.8601,
+      "step": 4495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.288810729980469,
+      "learning_rate": 1.84262914020937e-05,
+      "loss": 0.5983,
+      "step": 4496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.577433586120605,
+      "learning_rate": 1.8423717178651105e-05,
+      "loss": 0.9949,
+      "step": 4497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.912052154541016,
+      "learning_rate": 1.8421142955208515e-05,
+      "loss": 0.8605,
+      "step": 4498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.540184020996094,
+      "learning_rate": 1.8418568731765918e-05,
+      "loss": 0.6795,
+      "step": 4499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.33013916015625,
+      "learning_rate": 1.8415994508323325e-05,
+      "loss": 0.741,
+      "step": 4500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.999069213867188,
+      "learning_rate": 1.8413420284880728e-05,
+      "loss": 0.7492,
+      "step": 4501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.603501319885254,
+      "learning_rate": 1.8410846061438135e-05,
+      "loss": 0.6832,
+      "step": 4502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.333259582519531,
+      "learning_rate": 1.8408271837995538e-05,
+      "loss": 0.7259,
+      "step": 4503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.627447128295898,
+      "learning_rate": 1.8405697614552945e-05,
+      "loss": 0.8405,
+      "step": 4504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.332725524902344,
+      "learning_rate": 1.8403123391110348e-05,
+      "loss": 1.0176,
+      "step": 4505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.903879165649414,
+      "learning_rate": 1.840054916766775e-05,
+      "loss": 0.7921,
+      "step": 4506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.524657249450684,
+      "learning_rate": 1.8397974944225158e-05,
+      "loss": 0.833,
+      "step": 4507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.497197151184082,
+      "learning_rate": 1.8395400720782565e-05,
+      "loss": 0.9854,
+      "step": 4508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.108022689819336,
+      "learning_rate": 1.839282649733997e-05,
+      "loss": 0.8133,
+      "step": 4509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.229659080505371,
+      "learning_rate": 1.8390252273897375e-05,
+      "loss": 0.8802,
+      "step": 4510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.211379051208496,
+      "learning_rate": 1.838767805045478e-05,
+      "loss": 0.8888,
+      "step": 4511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.35076904296875,
+      "learning_rate": 1.8385103827012185e-05,
+      "loss": 0.9631,
+      "step": 4512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.228515625,
+      "learning_rate": 1.838252960356959e-05,
+      "loss": 0.8236,
+      "step": 4513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.220845222473145,
+      "learning_rate": 1.8379955380126995e-05,
+      "loss": 0.9212,
+      "step": 4514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.140191078186035,
+      "learning_rate": 1.83773811566844e-05,
+      "loss": 0.8759,
+      "step": 4515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.232515335083008,
+      "learning_rate": 1.8374806933241804e-05,
+      "loss": 0.9269,
+      "step": 4516
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.290019989013672,
+      "learning_rate": 1.837223270979921e-05,
+      "loss": 0.575,
+      "step": 4517
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.017510414123535,
+      "learning_rate": 1.8369658486356618e-05,
+      "loss": 0.9091,
+      "step": 4518
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.72248649597168,
+      "learning_rate": 1.836708426291402e-05,
+      "loss": 1.0444,
+      "step": 4519
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.022782325744629,
+      "learning_rate": 1.8364510039471428e-05,
+      "loss": 0.7715,
+      "step": 4520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.991741180419922,
+      "learning_rate": 1.836193581602883e-05,
+      "loss": 0.8239,
+      "step": 4521
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.724997520446777,
+      "learning_rate": 1.8359361592586238e-05,
+      "loss": 0.7705,
+      "step": 4522
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.342066764831543,
+      "learning_rate": 1.835678736914364e-05,
+      "loss": 0.7131,
+      "step": 4523
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.394556522369385,
+      "learning_rate": 1.8354213145701048e-05,
+      "loss": 0.4355,
+      "step": 4524
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.0487699508667,
+      "learning_rate": 1.835163892225845e-05,
+      "loss": 0.6849,
+      "step": 4525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.3297758102417,
+      "learning_rate": 1.8349064698815858e-05,
+      "loss": 0.8275,
+      "step": 4526
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.003467559814453,
+      "learning_rate": 1.8346490475373264e-05,
+      "loss": 0.7469,
+      "step": 4527
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.577159881591797,
+      "learning_rate": 1.8343916251930668e-05,
+      "loss": 0.7427,
+      "step": 4528
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.81598949432373,
+      "learning_rate": 1.8341342028488074e-05,
+      "loss": 0.9844,
+      "step": 4529
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.05923843383789,
+      "learning_rate": 1.8338767805045478e-05,
+      "loss": 0.964,
+      "step": 4530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.914913177490234,
+      "learning_rate": 1.8336193581602884e-05,
+      "loss": 0.7655,
+      "step": 4531
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.596419334411621,
+      "learning_rate": 1.8333619358160288e-05,
+      "loss": 0.8081,
+      "step": 4532
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.512883186340332,
+      "learning_rate": 1.8331045134717694e-05,
+      "loss": 0.7835,
+      "step": 4533
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.740601539611816,
+      "learning_rate": 1.8328470911275098e-05,
+      "loss": 1.0698,
+      "step": 4534
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.815478801727295,
+      "learning_rate": 1.8325896687832504e-05,
+      "loss": 0.6298,
+      "step": 4535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.781991004943848,
+      "learning_rate": 1.832332246438991e-05,
+      "loss": 0.8833,
+      "step": 4536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.7704691886901855,
+      "learning_rate": 1.8320748240947314e-05,
+      "loss": 0.6173,
+      "step": 4537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.388161659240723,
+      "learning_rate": 1.831817401750472e-05,
+      "loss": 0.6387,
+      "step": 4538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.602012634277344,
+      "learning_rate": 1.8315599794062124e-05,
+      "loss": 0.8828,
+      "step": 4539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.607589721679688,
+      "learning_rate": 1.831302557061953e-05,
+      "loss": 0.7411,
+      "step": 4540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.596658706665039,
+      "learning_rate": 1.8310451347176934e-05,
+      "loss": 0.665,
+      "step": 4541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.29475212097168,
+      "learning_rate": 1.830787712373434e-05,
+      "loss": 0.9465,
+      "step": 4542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.66930103302002,
+      "learning_rate": 1.8305302900291744e-05,
+      "loss": 0.6908,
+      "step": 4543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.661069869995117,
+      "learning_rate": 1.830272867684915e-05,
+      "loss": 0.5897,
+      "step": 4544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.405319213867188,
+      "learning_rate": 1.8300154453406554e-05,
+      "loss": 0.5914,
+      "step": 4545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.724068641662598,
+      "learning_rate": 1.8297580229963964e-05,
+      "loss": 0.7239,
+      "step": 4546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.002670288085938,
+      "learning_rate": 1.8295006006521368e-05,
+      "loss": 0.7828,
+      "step": 4547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.000532150268555,
+      "learning_rate": 1.829243178307877e-05,
+      "loss": 0.7399,
+      "step": 4548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.797389030456543,
+      "learning_rate": 1.8289857559636178e-05,
+      "loss": 0.8916,
+      "step": 4549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.00051212310791,
+      "learning_rate": 1.828728333619358e-05,
+      "loss": 0.6231,
+      "step": 4550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.578989028930664,
+      "learning_rate": 1.8284709112750988e-05,
+      "loss": 0.7963,
+      "step": 4551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.214009284973145,
+      "learning_rate": 1.828213488930839e-05,
+      "loss": 1.0731,
+      "step": 4552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.950447082519531,
+      "learning_rate": 1.8279560665865797e-05,
+      "loss": 0.8213,
+      "step": 4553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.635833740234375,
+      "learning_rate": 1.82769864424232e-05,
+      "loss": 0.9224,
+      "step": 4554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.554736137390137,
+      "learning_rate": 1.827441221898061e-05,
+      "loss": 0.7679,
+      "step": 4555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.232850074768066,
+      "learning_rate": 1.8271837995538014e-05,
+      "loss": 0.6687,
+      "step": 4556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.844307899475098,
+      "learning_rate": 1.826926377209542e-05,
+      "loss": 0.9273,
+      "step": 4557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.278690338134766,
+      "learning_rate": 1.8266689548652824e-05,
+      "loss": 0.7023,
+      "step": 4558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.582830429077148,
+      "learning_rate": 1.8264115325210227e-05,
+      "loss": 0.8023,
+      "step": 4559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.576903343200684,
+      "learning_rate": 1.8261541101767634e-05,
+      "loss": 0.7627,
+      "step": 4560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.200854301452637,
+      "learning_rate": 1.8258966878325037e-05,
+      "loss": 0.9509,
+      "step": 4561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.573077201843262,
+      "learning_rate": 1.8256392654882444e-05,
+      "loss": 0.911,
+      "step": 4562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.515393257141113,
+      "learning_rate": 1.8253818431439847e-05,
+      "loss": 0.8064,
+      "step": 4563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.103652000427246,
+      "learning_rate": 1.8251244207997254e-05,
+      "loss": 1.0753,
+      "step": 4564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.32536506652832,
+      "learning_rate": 1.824866998455466e-05,
+      "loss": 0.6403,
+      "step": 4565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.977213859558105,
+      "learning_rate": 1.8246095761112067e-05,
+      "loss": 0.7228,
+      "step": 4566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.798955917358398,
+      "learning_rate": 1.824352153766947e-05,
+      "loss": 0.8165,
+      "step": 4567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.52734375,
+      "learning_rate": 1.8240947314226874e-05,
+      "loss": 0.6456,
+      "step": 4568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.67576789855957,
+      "learning_rate": 1.823837309078428e-05,
+      "loss": 0.6377,
+      "step": 4569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.465245246887207,
+      "learning_rate": 1.8235798867341684e-05,
+      "loss": 0.7506,
+      "step": 4570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.725875854492188,
+      "learning_rate": 1.823322464389909e-05,
+      "loss": 0.9004,
+      "step": 4571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.636159896850586,
+      "learning_rate": 1.8230650420456494e-05,
+      "loss": 0.6241,
+      "step": 4572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.581345558166504,
+      "learning_rate": 1.82280761970139e-05,
+      "loss": 0.8353,
+      "step": 4573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.311539649963379,
+      "learning_rate": 1.8225501973571307e-05,
+      "loss": 0.9389,
+      "step": 4574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.504876136779785,
+      "learning_rate": 1.8222927750128714e-05,
+      "loss": 0.7312,
+      "step": 4575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.0369234085083,
+      "learning_rate": 1.8220353526686117e-05,
+      "loss": 0.8633,
+      "step": 4576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.956113815307617,
+      "learning_rate": 1.8217779303243524e-05,
+      "loss": 0.731,
+      "step": 4577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.403385162353516,
+      "learning_rate": 1.8215205079800927e-05,
+      "loss": 0.7374,
+      "step": 4578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.958566188812256,
+      "learning_rate": 1.821263085635833e-05,
+      "loss": 0.3727,
+      "step": 4579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.500529289245605,
+      "learning_rate": 1.8210056632915737e-05,
+      "loss": 1.0132,
+      "step": 4580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.410884857177734,
+      "learning_rate": 1.820748240947314e-05,
+      "loss": 0.7981,
+      "step": 4581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.440316200256348,
+      "learning_rate": 1.8204908186030547e-05,
+      "loss": 0.7446,
+      "step": 4582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.056214332580566,
+      "learning_rate": 1.820233396258795e-05,
+      "loss": 0.7824,
+      "step": 4583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.531798362731934,
+      "learning_rate": 1.819975973914536e-05,
+      "loss": 0.7659,
+      "step": 4584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.205221176147461,
+      "learning_rate": 1.8197185515702764e-05,
+      "loss": 0.8347,
+      "step": 4585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.71817684173584,
+      "learning_rate": 1.819461129226017e-05,
+      "loss": 0.7105,
+      "step": 4586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.199410438537598,
+      "learning_rate": 1.8192037068817574e-05,
+      "loss": 0.8129,
+      "step": 4587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.65517520904541,
+      "learning_rate": 1.818946284537498e-05,
+      "loss": 0.7054,
+      "step": 4588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.244235038757324,
+      "learning_rate": 1.8186888621932384e-05,
+      "loss": 0.8342,
+      "step": 4589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.863982200622559,
+      "learning_rate": 1.8184314398489787e-05,
+      "loss": 0.6068,
+      "step": 4590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.03433895111084,
+      "learning_rate": 1.8181740175047194e-05,
+      "loss": 0.8217,
+      "step": 4591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.143538475036621,
+      "learning_rate": 1.8179165951604597e-05,
+      "loss": 0.7609,
+      "step": 4592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.755047798156738,
+      "learning_rate": 1.8176591728162007e-05,
+      "loss": 0.6745,
+      "step": 4593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.403046607971191,
+      "learning_rate": 1.817401750471941e-05,
+      "loss": 0.8431,
+      "step": 4594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.723485946655273,
+      "learning_rate": 1.8171443281276817e-05,
+      "loss": 0.7758,
+      "step": 4595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.263933181762695,
+      "learning_rate": 1.816886905783422e-05,
+      "loss": 0.8909,
+      "step": 4596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.518206596374512,
+      "learning_rate": 1.8166294834391627e-05,
+      "loss": 0.8204,
+      "step": 4597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.208568572998047,
+      "learning_rate": 1.816372061094903e-05,
+      "loss": 0.9405,
+      "step": 4598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.45848274230957,
+      "learning_rate": 1.8161146387506437e-05,
+      "loss": 1.0782,
+      "step": 4599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.41175365447998,
+      "learning_rate": 1.815857216406384e-05,
+      "loss": 0.9453,
+      "step": 4600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.178896903991699,
+      "learning_rate": 1.8155997940621244e-05,
+      "loss": 0.4872,
+      "step": 4601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.46392822265625,
+      "learning_rate": 1.815342371717865e-05,
+      "loss": 0.8833,
+      "step": 4602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.135536193847656,
+      "learning_rate": 1.8150849493736057e-05,
+      "loss": 0.6072,
+      "step": 4603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.613280296325684,
+      "learning_rate": 1.8148275270293464e-05,
+      "loss": 0.6517,
+      "step": 4604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.646885871887207,
+      "learning_rate": 1.8145701046850867e-05,
+      "loss": 0.8452,
+      "step": 4605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.22172737121582,
+      "learning_rate": 1.8143126823408274e-05,
+      "loss": 0.8655,
+      "step": 4606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.80384349822998,
+      "learning_rate": 1.8140552599965677e-05,
+      "loss": 0.7573,
+      "step": 4607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.838275909423828,
+      "learning_rate": 1.8137978376523084e-05,
+      "loss": 0.5721,
+      "step": 4608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.503539085388184,
+      "learning_rate": 1.8135404153080487e-05,
+      "loss": 0.6233,
+      "step": 4609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.924944877624512,
+      "learning_rate": 1.813282992963789e-05,
+      "loss": 0.6523,
+      "step": 4610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.088214874267578,
+      "learning_rate": 1.8130255706195297e-05,
+      "loss": 0.8185,
+      "step": 4611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.18722152709961,
+      "learning_rate": 1.8127681482752704e-05,
+      "loss": 0.7974,
+      "step": 4612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.173393249511719,
+      "learning_rate": 1.812510725931011e-05,
+      "loss": 0.7328,
+      "step": 4613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.351269721984863,
+      "learning_rate": 1.8122533035867514e-05,
+      "loss": 0.6004,
+      "step": 4614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.80124568939209,
+      "learning_rate": 1.811995881242492e-05,
+      "loss": 0.812,
+      "step": 4615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.095589637756348,
+      "learning_rate": 1.8117384588982324e-05,
+      "loss": 0.8113,
+      "step": 4616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.35456371307373,
+      "learning_rate": 1.811481036553973e-05,
+      "loss": 0.9161,
+      "step": 4617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.764464378356934,
+      "learning_rate": 1.8112236142097134e-05,
+      "loss": 0.8103,
+      "step": 4618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.464253425598145,
+      "learning_rate": 1.810966191865454e-05,
+      "loss": 0.903,
+      "step": 4619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.548282623291016,
+      "learning_rate": 1.8107087695211943e-05,
+      "loss": 0.773,
+      "step": 4620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.999635696411133,
+      "learning_rate": 1.810451347176935e-05,
+      "loss": 0.4953,
+      "step": 4621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.703523635864258,
+      "learning_rate": 1.8101939248326757e-05,
+      "loss": 0.9012,
+      "step": 4622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.518583297729492,
+      "learning_rate": 1.809936502488416e-05,
+      "loss": 0.9153,
+      "step": 4623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.757223129272461,
+      "learning_rate": 1.8096790801441567e-05,
+      "loss": 0.8886,
+      "step": 4624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.543490409851074,
+      "learning_rate": 1.809421657799897e-05,
+      "loss": 0.8808,
+      "step": 4625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.924624919891357,
+      "learning_rate": 1.8091642354556377e-05,
+      "loss": 0.6956,
+      "step": 4626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.829492568969727,
+      "learning_rate": 1.808906813111378e-05,
+      "loss": 0.7511,
+      "step": 4627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.045244216918945,
+      "learning_rate": 1.8086493907671187e-05,
+      "loss": 0.8252,
+      "step": 4628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.998333930969238,
+      "learning_rate": 1.808391968422859e-05,
+      "loss": 0.8851,
+      "step": 4629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.977302551269531,
+      "learning_rate": 1.8081345460785997e-05,
+      "loss": 1.103,
+      "step": 4630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.581474304199219,
+      "learning_rate": 1.8078771237343403e-05,
+      "loss": 0.6281,
+      "step": 4631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.6852445602417,
+      "learning_rate": 1.8076197013900807e-05,
+      "loss": 0.8449,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.24959659576416,
+      "learning_rate": 1.8073622790458213e-05,
+      "loss": 0.7105,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.15105152130127,
+      "learning_rate": 1.8071048567015617e-05,
+      "loss": 0.7902,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.813551902770996,
+      "learning_rate": 1.8068474343573023e-05,
+      "loss": 0.5855,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.484671592712402,
+      "learning_rate": 1.8065900120130427e-05,
+      "loss": 0.7479,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.329991340637207,
+      "learning_rate": 1.8063325896687833e-05,
+      "loss": 0.8843,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.500802993774414,
+      "learning_rate": 1.8060751673245237e-05,
+      "loss": 0.721,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.564055442810059,
+      "learning_rate": 1.8058177449802643e-05,
+      "loss": 0.6692,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.584227561950684,
+      "learning_rate": 1.805560322636005e-05,
+      "loss": 0.6948,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.984607696533203,
+      "learning_rate": 1.8053029002917453e-05,
+      "loss": 0.9441,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.908760070800781,
+      "learning_rate": 1.805045477947486e-05,
+      "loss": 0.8061,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.603954315185547,
+      "learning_rate": 1.8047880556032263e-05,
+      "loss": 0.8863,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.7311429977417,
+      "learning_rate": 1.804530633258967e-05,
+      "loss": 0.8869,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.57198715209961,
+      "learning_rate": 1.8042732109147073e-05,
+      "loss": 0.5754,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.147098541259766,
+      "learning_rate": 1.804015788570448e-05,
+      "loss": 0.6755,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.718560218811035,
+      "learning_rate": 1.8037583662261883e-05,
+      "loss": 0.575,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.119189262390137,
+      "learning_rate": 1.803500943881929e-05,
+      "loss": 0.5091,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.507854461669922,
+      "learning_rate": 1.8032435215376693e-05,
+      "loss": 0.9684,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.607927322387695,
+      "learning_rate": 1.8029860991934103e-05,
+      "loss": 0.7739,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.133891105651855,
+      "learning_rate": 1.8027286768491507e-05,
+      "loss": 0.8511,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.290109634399414,
+      "learning_rate": 1.802471254504891e-05,
+      "loss": 0.7475,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.49523639678955,
+      "learning_rate": 1.8022138321606317e-05,
+      "loss": 0.7153,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.8330254554748535,
+      "learning_rate": 1.801956409816372e-05,
+      "loss": 0.6563,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.321264266967773,
+      "learning_rate": 1.8016989874721127e-05,
+      "loss": 0.6686,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.530916213989258,
+      "learning_rate": 1.801441565127853e-05,
+      "loss": 0.6511,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.36806869506836,
+      "learning_rate": 1.8011841427835936e-05,
+      "loss": 0.8904,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.371545791625977,
+      "learning_rate": 1.800926720439334e-05,
+      "loss": 0.7104,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.19075870513916,
+      "learning_rate": 1.800669298095075e-05,
+      "loss": 0.6715,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.404905319213867,
+      "learning_rate": 1.8004118757508153e-05,
+      "loss": 0.8377,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.673076629638672,
+      "learning_rate": 1.800154453406556e-05,
+      "loss": 0.5113,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.194426536560059,
+      "learning_rate": 1.7998970310622963e-05,
+      "loss": 0.8991,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.455160140991211,
+      "learning_rate": 1.7996396087180366e-05,
+      "loss": 1.0607,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.950098037719727,
+      "learning_rate": 1.7993821863737773e-05,
+      "loss": 0.7847,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.758997917175293,
+      "learning_rate": 1.7991247640295176e-05,
+      "loss": 0.7718,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.240198135375977,
+      "learning_rate": 1.7988673416852583e-05,
+      "loss": 0.7921,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.519044876098633,
+      "learning_rate": 1.7986099193409986e-05,
+      "loss": 0.8884,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.16071891784668,
+      "learning_rate": 1.7983524969967393e-05,
+      "loss": 0.6255,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.046331405639648,
+      "learning_rate": 1.79809507465248e-05,
+      "loss": 0.7597,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.17223072052002,
+      "learning_rate": 1.7978376523082206e-05,
+      "loss": 0.6832,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.298058032989502,
+      "learning_rate": 1.797580229963961e-05,
+      "loss": 0.5662,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.127120018005371,
+      "learning_rate": 1.7973228076197013e-05,
+      "loss": 0.8818,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.152034759521484,
+      "learning_rate": 1.797065385275442e-05,
+      "loss": 0.7842,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.81672477722168,
+      "learning_rate": 1.7968079629311823e-05,
+      "loss": 1.113,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.437200546264648,
+      "learning_rate": 1.796550540586923e-05,
+      "loss": 0.7425,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.65206241607666,
+      "learning_rate": 1.7962931182426633e-05,
+      "loss": 0.7048,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.43521785736084,
+      "learning_rate": 1.796035695898404e-05,
+      "loss": 0.8755,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.043573379516602,
+      "learning_rate": 1.7957782735541446e-05,
+      "loss": 0.8106,
+      "step": 4678
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.109203338623047,
+      "learning_rate": 1.7955208512098853e-05,
+      "loss": 0.9277,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.440686225891113,
+      "learning_rate": 1.7952634288656256e-05,
+      "loss": 0.8093,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.53641128540039,
+      "learning_rate": 1.7950060065213663e-05,
+      "loss": 0.5236,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.874175071716309,
+      "learning_rate": 1.7947485841771066e-05,
+      "loss": 0.878,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.545369148254395,
+      "learning_rate": 1.794491161832847e-05,
+      "loss": 1.1202,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.253897666931152,
+      "learning_rate": 1.7942337394885876e-05,
+      "loss": 0.7012,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.787257194519043,
+      "learning_rate": 1.793976317144328e-05,
+      "loss": 0.6988,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.269976615905762,
+      "learning_rate": 1.7937188948000686e-05,
+      "loss": 0.7213,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.339044570922852,
+      "learning_rate": 1.793461472455809e-05,
+      "loss": 0.6812,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.309968948364258,
+      "learning_rate": 1.79320405011155e-05,
+      "loss": 0.7178,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.741803169250488,
+      "learning_rate": 1.7929466277672903e-05,
+      "loss": 0.9432,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.693902969360352,
+      "learning_rate": 1.792689205423031e-05,
+      "loss": 0.7564,
+      "step": 4690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.328643798828125,
+      "learning_rate": 1.7924317830787713e-05,
+      "loss": 0.6594,
+      "step": 4691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.708168029785156,
+      "learning_rate": 1.792174360734512e-05,
+      "loss": 1.0262,
+      "step": 4692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.268716812133789,
+      "learning_rate": 1.7919169383902523e-05,
+      "loss": 0.6609,
+      "step": 4693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.470682144165039,
+      "learning_rate": 1.7916595160459926e-05,
+      "loss": 0.9439,
+      "step": 4694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.904067039489746,
+      "learning_rate": 1.7914020937017333e-05,
+      "loss": 0.8827,
+      "step": 4695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.216064453125,
+      "learning_rate": 1.7911446713574736e-05,
+      "loss": 0.8426,
+      "step": 4696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.793227195739746,
+      "learning_rate": 1.7908872490132146e-05,
+      "loss": 0.7214,
+      "step": 4697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.365936279296875,
+      "learning_rate": 1.790629826668955e-05,
+      "loss": 0.9337,
+      "step": 4698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.970952987670898,
+      "learning_rate": 1.7903724043246956e-05,
+      "loss": 0.544,
+      "step": 4699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.585055351257324,
+      "learning_rate": 1.790114981980436e-05,
+      "loss": 0.7607,
+      "step": 4700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.284903526306152,
+      "learning_rate": 1.7898575596361766e-05,
+      "loss": 0.7597,
+      "step": 4701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.15507698059082,
+      "learning_rate": 1.789600137291917e-05,
+      "loss": 0.5263,
+      "step": 4702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.795269966125488,
+      "learning_rate": 1.7893427149476576e-05,
+      "loss": 1.1584,
+      "step": 4703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.412137985229492,
+      "learning_rate": 1.789085292603398e-05,
+      "loss": 0.7047,
+      "step": 4704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.058917999267578,
+      "learning_rate": 1.7888278702591383e-05,
+      "loss": 0.8618,
+      "step": 4705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.027851104736328,
+      "learning_rate": 1.788570447914879e-05,
+      "loss": 0.6486,
+      "step": 4706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.5999116897583,
+      "learning_rate": 1.7883130255706196e-05,
+      "loss": 0.679,
+      "step": 4707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.76766300201416,
+      "learning_rate": 1.7880556032263603e-05,
+      "loss": 0.8544,
+      "step": 4708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.434048652648926,
+      "learning_rate": 1.7877981808821006e-05,
+      "loss": 0.5485,
+      "step": 4709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.41886043548584,
+      "learning_rate": 1.7875407585378413e-05,
+      "loss": 1.045,
+      "step": 4710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.896388053894043,
+      "learning_rate": 1.7872833361935816e-05,
+      "loss": 1.0275,
+      "step": 4711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.855283737182617,
+      "learning_rate": 1.7870259138493223e-05,
+      "loss": 0.8951,
+      "step": 4712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.453601837158203,
+      "learning_rate": 1.7867684915050626e-05,
+      "loss": 0.5739,
+      "step": 4713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.944012641906738,
+      "learning_rate": 1.786511069160803e-05,
+      "loss": 0.7602,
+      "step": 4714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.393771171569824,
+      "learning_rate": 1.7862536468165436e-05,
+      "loss": 0.6744,
+      "step": 4715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.566618919372559,
+      "learning_rate": 1.7859962244722843e-05,
+      "loss": 0.7501,
+      "step": 4716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.667789459228516,
+      "learning_rate": 1.785738802128025e-05,
+      "loss": 0.6859,
+      "step": 4717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.746609687805176,
+      "learning_rate": 1.7854813797837653e-05,
+      "loss": 1.0023,
+      "step": 4718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.340370178222656,
+      "learning_rate": 1.785223957439506e-05,
+      "loss": 0.7859,
+      "step": 4719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.385442733764648,
+      "learning_rate": 1.7849665350952463e-05,
+      "loss": 0.9263,
+      "step": 4720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.542134284973145,
+      "learning_rate": 1.784709112750987e-05,
+      "loss": 0.8354,
+      "step": 4721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.154189109802246,
+      "learning_rate": 1.7844516904067273e-05,
+      "loss": 0.6012,
+      "step": 4722
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.087700366973877,
+      "learning_rate": 1.784194268062468e-05,
+      "loss": 0.4578,
+      "step": 4723
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.976358413696289,
+      "learning_rate": 1.7839368457182083e-05,
+      "loss": 0.6712,
+      "step": 4724
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.208373069763184,
+      "learning_rate": 1.783679423373949e-05,
+      "loss": 0.9655,
+      "step": 4725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.266886711120605,
+      "learning_rate": 1.7834220010296896e-05,
+      "loss": 0.8008,
+      "step": 4726
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.366615295410156,
+      "learning_rate": 1.78316457868543e-05,
+      "loss": 0.7056,
+      "step": 4727
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.03531265258789,
+      "learning_rate": 1.7829071563411706e-05,
+      "loss": 0.9338,
+      "step": 4728
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.879033088684082,
+      "learning_rate": 1.782649733996911e-05,
+      "loss": 0.8495,
+      "step": 4729
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.972820281982422,
+      "learning_rate": 1.7823923116526516e-05,
+      "loss": 0.9394,
+      "step": 4730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.600964546203613,
+      "learning_rate": 1.782134889308392e-05,
+      "loss": 0.8586,
+      "step": 4731
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.581879615783691,
+      "learning_rate": 1.7818774669641326e-05,
+      "loss": 0.7712,
+      "step": 4732
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.14083480834961,
+      "learning_rate": 1.781620044619873e-05,
+      "loss": 0.7987,
+      "step": 4733
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.475602149963379,
+      "learning_rate": 1.7813626222756136e-05,
+      "loss": 0.7109,
+      "step": 4734
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.210009574890137,
+      "learning_rate": 1.7811051999313542e-05,
+      "loss": 0.7231,
+      "step": 4735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.758543014526367,
+      "learning_rate": 1.7808477775870946e-05,
+      "loss": 0.5857,
+      "step": 4736
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.707775115966797,
+      "learning_rate": 1.7805903552428352e-05,
+      "loss": 0.9456,
+      "step": 4737
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.69056510925293,
+      "learning_rate": 1.7803329328985756e-05,
+      "loss": 0.7248,
+      "step": 4738
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.393020629882812,
+      "learning_rate": 1.7800755105543162e-05,
+      "loss": 0.6604,
+      "step": 4739
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.653887748718262,
+      "learning_rate": 1.7798180882100566e-05,
+      "loss": 0.6858,
+      "step": 4740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.445959091186523,
+      "learning_rate": 1.7795606658657972e-05,
+      "loss": 0.8981,
+      "step": 4741
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.136672973632812,
+      "learning_rate": 1.7793032435215376e-05,
+      "loss": 0.6346,
+      "step": 4742
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.509740829467773,
+      "learning_rate": 1.7790458211772782e-05,
+      "loss": 0.6559,
+      "step": 4743
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.922419548034668,
+      "learning_rate": 1.778788398833019e-05,
+      "loss": 0.6987,
+      "step": 4744
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.939035415649414,
+      "learning_rate": 1.7785309764887592e-05,
+      "loss": 0.7952,
+      "step": 4745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.739534378051758,
+      "learning_rate": 1.7782735541445e-05,
+      "loss": 0.6478,
+      "step": 4746
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.027047157287598,
+      "learning_rate": 1.7780161318002402e-05,
+      "loss": 0.9522,
+      "step": 4747
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.606758117675781,
+      "learning_rate": 1.777758709455981e-05,
+      "loss": 0.5663,
+      "step": 4748
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.612149238586426,
+      "learning_rate": 1.7775012871117212e-05,
+      "loss": 0.7487,
+      "step": 4749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.411164283752441,
+      "learning_rate": 1.777243864767462e-05,
+      "loss": 0.8331,
+      "step": 4750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.100812911987305,
+      "learning_rate": 1.7769864424232022e-05,
+      "loss": 0.6652,
+      "step": 4751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.680567741394043,
+      "learning_rate": 1.776729020078943e-05,
+      "loss": 0.8855,
+      "step": 4752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.902464866638184,
+      "learning_rate": 1.7764715977346832e-05,
+      "loss": 0.6956,
+      "step": 4753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.097615242004395,
+      "learning_rate": 1.7762141753904242e-05,
+      "loss": 0.5542,
+      "step": 4754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.215203285217285,
+      "learning_rate": 1.7759567530461646e-05,
+      "loss": 0.7032,
+      "step": 4755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.553426742553711,
+      "learning_rate": 1.775699330701905e-05,
+      "loss": 0.872,
+      "step": 4756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.280723571777344,
+      "learning_rate": 1.7754419083576456e-05,
+      "loss": 0.7258,
+      "step": 4757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.920151710510254,
+      "learning_rate": 1.775184486013386e-05,
+      "loss": 0.7815,
+      "step": 4758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.299314498901367,
+      "learning_rate": 1.7749270636691266e-05,
+      "loss": 0.8043,
+      "step": 4759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.703652381896973,
+      "learning_rate": 1.774669641324867e-05,
+      "loss": 0.9688,
+      "step": 4760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.8981351852417,
+      "learning_rate": 1.7744122189806076e-05,
+      "loss": 0.8533,
+      "step": 4761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.716887474060059,
+      "learning_rate": 1.774154796636348e-05,
+      "loss": 0.7105,
+      "step": 4762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.862297058105469,
+      "learning_rate": 1.773897374292089e-05,
+      "loss": 0.9016,
+      "step": 4763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.814416885375977,
+      "learning_rate": 1.7736399519478292e-05,
+      "loss": 0.7525,
+      "step": 4764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.40992259979248,
+      "learning_rate": 1.77338252960357e-05,
+      "loss": 0.7877,
+      "step": 4765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.749227523803711,
+      "learning_rate": 1.7731251072593102e-05,
+      "loss": 0.6553,
+      "step": 4766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.865870475769043,
+      "learning_rate": 1.7728676849150505e-05,
+      "loss": 0.7681,
+      "step": 4767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.250228881835938,
+      "learning_rate": 1.7726102625707912e-05,
+      "loss": 0.6776,
+      "step": 4768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.269866943359375,
+      "learning_rate": 1.7723528402265315e-05,
+      "loss": 0.6895,
+      "step": 4769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.050362586975098,
+      "learning_rate": 1.7720954178822722e-05,
+      "loss": 0.7331,
+      "step": 4770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.744763374328613,
+      "learning_rate": 1.7718379955380125e-05,
+      "loss": 0.867,
+      "step": 4771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.055499076843262,
+      "learning_rate": 1.7715805731937532e-05,
+      "loss": 0.739,
+      "step": 4772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.691717624664307,
+      "learning_rate": 1.771323150849494e-05,
+      "loss": 0.5485,
+      "step": 4773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.498370170593262,
+      "learning_rate": 1.7710657285052345e-05,
+      "loss": 0.91,
+      "step": 4774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.988570213317871,
+      "learning_rate": 1.770808306160975e-05,
+      "loss": 0.6445,
+      "step": 4775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.056323051452637,
+      "learning_rate": 1.7705508838167155e-05,
+      "loss": 0.4321,
+      "step": 4776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.658989906311035,
+      "learning_rate": 1.770293461472456e-05,
+      "loss": 0.7111,
+      "step": 4777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.714153289794922,
+      "learning_rate": 1.7700360391281962e-05,
+      "loss": 0.7901,
+      "step": 4778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.393281936645508,
+      "learning_rate": 1.769778616783937e-05,
+      "loss": 0.6817,
+      "step": 4779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.453137397766113,
+      "learning_rate": 1.7695211944396772e-05,
+      "loss": 0.9336,
+      "step": 4780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.852784156799316,
+      "learning_rate": 1.769263772095418e-05,
+      "loss": 0.7724,
+      "step": 4781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.578731536865234,
+      "learning_rate": 1.7690063497511585e-05,
+      "loss": 0.7911,
+      "step": 4782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.003982543945312,
+      "learning_rate": 1.7687489274068992e-05,
+      "loss": 0.8182,
+      "step": 4783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.337196350097656,
+      "learning_rate": 1.7684915050626395e-05,
+      "loss": 0.7574,
+      "step": 4784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.3565673828125,
+      "learning_rate": 1.7682340827183802e-05,
+      "loss": 0.6727,
+      "step": 4785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.473671913146973,
+      "learning_rate": 1.7679766603741205e-05,
+      "loss": 0.6664,
+      "step": 4786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.382794380187988,
+      "learning_rate": 1.767719238029861e-05,
+      "loss": 0.6981,
+      "step": 4787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.083308219909668,
+      "learning_rate": 1.7674618156856015e-05,
+      "loss": 0.7013,
+      "step": 4788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.678533554077148,
+      "learning_rate": 1.767204393341342e-05,
+      "loss": 0.8264,
+      "step": 4789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.436578750610352,
+      "learning_rate": 1.7669469709970825e-05,
+      "loss": 0.8023,
+      "step": 4790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.489977836608887,
+      "learning_rate": 1.766689548652823e-05,
+      "loss": 0.7519,
+      "step": 4791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.302467346191406,
+      "learning_rate": 1.766432126308564e-05,
+      "loss": 0.5428,
+      "step": 4792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.264116287231445,
+      "learning_rate": 1.7661747039643042e-05,
+      "loss": 0.7665,
+      "step": 4793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.950275421142578,
+      "learning_rate": 1.765917281620045e-05,
+      "loss": 0.8895,
+      "step": 4794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.896724700927734,
+      "learning_rate": 1.7656598592757852e-05,
+      "loss": 0.6034,
+      "step": 4795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.13314151763916,
+      "learning_rate": 1.765402436931526e-05,
+      "loss": 0.8166,
+      "step": 4796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.60966968536377,
+      "learning_rate": 1.7651450145872662e-05,
+      "loss": 0.8992,
+      "step": 4797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.669915199279785,
+      "learning_rate": 1.7648875922430065e-05,
+      "loss": 0.6287,
+      "step": 4798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.607988357543945,
+      "learning_rate": 1.7646301698987472e-05,
+      "loss": 0.9002,
+      "step": 4799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.235288619995117,
+      "learning_rate": 1.7643727475544875e-05,
+      "loss": 0.7838,
+      "step": 4800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.377300262451172,
+      "learning_rate": 1.7641153252102285e-05,
+      "loss": 0.7719,
+      "step": 4801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.074933052062988,
+      "learning_rate": 1.763857902865969e-05,
+      "loss": 0.6962,
+      "step": 4802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.036238670349121,
+      "learning_rate": 1.7636004805217095e-05,
+      "loss": 0.6683,
+      "step": 4803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.239750862121582,
+      "learning_rate": 1.76334305817745e-05,
+      "loss": 0.7341,
+      "step": 4804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.698104858398438,
+      "learning_rate": 1.7630856358331905e-05,
+      "loss": 0.7769,
+      "step": 4805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.367387771606445,
+      "learning_rate": 1.762828213488931e-05,
+      "loss": 0.6348,
+      "step": 4806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.038697242736816,
+      "learning_rate": 1.7625707911446715e-05,
+      "loss": 0.6603,
+      "step": 4807
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.054304122924805,
+      "learning_rate": 1.762313368800412e-05,
+      "loss": 0.8838,
+      "step": 4808
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.05933666229248,
+      "learning_rate": 1.762055946456152e-05,
+      "loss": 0.4696,
+      "step": 4809
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.898219108581543,
+      "learning_rate": 1.761798524111893e-05,
+      "loss": 0.7013,
+      "step": 4810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.645747184753418,
+      "learning_rate": 1.7615411017676335e-05,
+      "loss": 0.9902,
+      "step": 4811
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.932388305664062,
+      "learning_rate": 1.7612836794233742e-05,
+      "loss": 0.8642,
+      "step": 4812
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.371105194091797,
+      "learning_rate": 1.7610262570791145e-05,
+      "loss": 1.0138,
+      "step": 4813
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.310482025146484,
+      "learning_rate": 1.7607688347348552e-05,
+      "loss": 0.7208,
+      "step": 4814
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.735977172851562,
+      "learning_rate": 1.7605114123905955e-05,
+      "loss": 0.7119,
+      "step": 4815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.394113540649414,
+      "learning_rate": 1.760253990046336e-05,
+      "loss": 0.8329,
+      "step": 4816
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.97292423248291,
+      "learning_rate": 1.7599965677020765e-05,
+      "loss": 0.8301,
+      "step": 4817
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.534422874450684,
+      "learning_rate": 1.7597391453578168e-05,
+      "loss": 1.0778,
+      "step": 4818
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.125338554382324,
+      "learning_rate": 1.7594817230135575e-05,
+      "loss": 0.9699,
+      "step": 4819
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.227917671203613,
+      "learning_rate": 1.759224300669298e-05,
+      "loss": 0.7768,
+      "step": 4820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.644247055053711,
+      "learning_rate": 1.7589668783250388e-05,
+      "loss": 0.7802,
+      "step": 4821
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.752542495727539,
+      "learning_rate": 1.758709455980779e-05,
+      "loss": 0.9837,
+      "step": 4822
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.625603675842285,
+      "learning_rate": 1.7584520336365198e-05,
+      "loss": 0.8077,
+      "step": 4823
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.706161499023438,
+      "learning_rate": 1.75819461129226e-05,
+      "loss": 0.7319,
+      "step": 4824
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.121040344238281,
+      "learning_rate": 1.7579371889480008e-05,
+      "loss": 0.8439,
+      "step": 4825
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.253231048583984,
+      "learning_rate": 1.757679766603741e-05,
+      "loss": 0.7152,
+      "step": 4826
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.22000503540039,
+      "learning_rate": 1.7574223442594818e-05,
+      "loss": 0.461,
+      "step": 4827
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.102100372314453,
+      "learning_rate": 1.757164921915222e-05,
+      "loss": 0.8478,
+      "step": 4828
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.431357383728027,
+      "learning_rate": 1.7569074995709625e-05,
+      "loss": 0.8168,
+      "step": 4829
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.592718124389648,
+      "learning_rate": 1.7566500772267035e-05,
+      "loss": 0.8607,
+      "step": 4830
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.677876472473145,
+      "learning_rate": 1.7563926548824438e-05,
+      "loss": 0.9186,
+      "step": 4831
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.456489562988281,
+      "learning_rate": 1.7561352325381845e-05,
+      "loss": 0.6835,
+      "step": 4832
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.43439769744873,
+      "learning_rate": 1.7558778101939248e-05,
+      "loss": 0.9202,
+      "step": 4833
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.976131439208984,
+      "learning_rate": 1.7556203878496655e-05,
+      "loss": 0.6594,
+      "step": 4834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.379373550415039,
+      "learning_rate": 1.7553629655054058e-05,
+      "loss": 0.7108,
+      "step": 4835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.393901824951172,
+      "learning_rate": 1.7551055431611465e-05,
+      "loss": 0.6935,
+      "step": 4836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.976284980773926,
+      "learning_rate": 1.7548481208168868e-05,
+      "loss": 0.409,
+      "step": 4837
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.669224739074707,
+      "learning_rate": 1.7545906984726275e-05,
+      "loss": 0.7548,
+      "step": 4838
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.577635765075684,
+      "learning_rate": 1.754333276128368e-05,
+      "loss": 0.7728,
+      "step": 4839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.180777549743652,
+      "learning_rate": 1.7540758537841085e-05,
+      "loss": 0.9255,
+      "step": 4840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.16286849975586,
+      "learning_rate": 1.753818431439849e-05,
+      "loss": 0.8874,
+      "step": 4841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.272332191467285,
+      "learning_rate": 1.7535610090955895e-05,
+      "loss": 0.8475,
+      "step": 4842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.789423942565918,
+      "learning_rate": 1.75330358675133e-05,
+      "loss": 0.5312,
+      "step": 4843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.371023178100586,
+      "learning_rate": 1.7530461644070705e-05,
+      "loss": 0.5665,
+      "step": 4844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.805109977722168,
+      "learning_rate": 1.752788742062811e-05,
+      "loss": 0.6869,
+      "step": 4845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.336012840270996,
+      "learning_rate": 1.7525313197185515e-05,
+      "loss": 0.8833,
+      "step": 4846
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.783849716186523,
+      "learning_rate": 1.752273897374292e-05,
+      "loss": 0.9346,
+      "step": 4847
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.484649658203125,
+      "learning_rate": 1.7520164750300328e-05,
+      "loss": 0.6735,
+      "step": 4848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.093966484069824,
+      "learning_rate": 1.751759052685773e-05,
+      "loss": 0.6745,
+      "step": 4849
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.905133247375488,
+      "learning_rate": 1.7515016303415138e-05,
+      "loss": 0.9312,
+      "step": 4850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.303994178771973,
+      "learning_rate": 1.751244207997254e-05,
+      "loss": 0.714,
+      "step": 4851
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.163971900939941,
+      "learning_rate": 1.7509867856529948e-05,
+      "loss": 0.8427,
+      "step": 4852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.35700798034668,
+      "learning_rate": 1.750729363308735e-05,
+      "loss": 0.6725,
+      "step": 4853
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.923184394836426,
+      "learning_rate": 1.7504719409644758e-05,
+      "loss": 0.8309,
+      "step": 4854
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.923500061035156,
+      "learning_rate": 1.750214518620216e-05,
+      "loss": 0.7293,
+      "step": 4855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.36686897277832,
+      "learning_rate": 1.7499570962759568e-05,
+      "loss": 0.7542,
+      "step": 4856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.174764633178711,
+      "learning_rate": 1.749699673931697e-05,
+      "loss": 0.7159,
+      "step": 4857
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.03627872467041,
+      "learning_rate": 1.749442251587438e-05,
+      "loss": 0.8438,
+      "step": 4858
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.963062286376953,
+      "learning_rate": 1.7491848292431785e-05,
+      "loss": 0.844,
+      "step": 4859
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.55281925201416,
+      "learning_rate": 1.7489274068989188e-05,
+      "loss": 1.0125,
+      "step": 4860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.869220733642578,
+      "learning_rate": 1.7486699845546595e-05,
+      "loss": 0.6718,
+      "step": 4861
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.24552059173584,
+      "learning_rate": 1.7484125622103998e-05,
+      "loss": 0.5701,
+      "step": 4862
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.85763168334961,
+      "learning_rate": 1.7481551398661405e-05,
+      "loss": 0.6519,
+      "step": 4863
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.538481712341309,
+      "learning_rate": 1.7478977175218808e-05,
+      "loss": 0.6596,
+      "step": 4864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.993450164794922,
+      "learning_rate": 1.7476402951776215e-05,
+      "loss": 0.8185,
+      "step": 4865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.547932624816895,
+      "learning_rate": 1.7473828728333618e-05,
+      "loss": 0.8077,
+      "step": 4866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.072369575500488,
+      "learning_rate": 1.7471254504891028e-05,
+      "loss": 0.5572,
+      "step": 4867
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.876172065734863,
+      "learning_rate": 1.746868028144843e-05,
+      "loss": 0.7,
+      "step": 4868
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.314702987670898,
+      "learning_rate": 1.7466106058005838e-05,
+      "loss": 0.6645,
+      "step": 4869
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.259981155395508,
+      "learning_rate": 1.746353183456324e-05,
+      "loss": 0.7082,
+      "step": 4870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.255072593688965,
+      "learning_rate": 1.7460957611120644e-05,
+      "loss": 0.8642,
+      "step": 4871
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.429109573364258,
+      "learning_rate": 1.745838338767805e-05,
+      "loss": 0.7689,
+      "step": 4872
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.946885108947754,
+      "learning_rate": 1.7455809164235454e-05,
+      "loss": 1.0431,
+      "step": 4873
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.412161827087402,
+      "learning_rate": 1.745323494079286e-05,
+      "loss": 0.6067,
+      "step": 4874
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.824974060058594,
+      "learning_rate": 1.7450660717350264e-05,
+      "loss": 0.712,
+      "step": 4875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.531627655029297,
+      "learning_rate": 1.744808649390767e-05,
+      "loss": 1.3283,
+      "step": 4876
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.385008811950684,
+      "learning_rate": 1.7445512270465078e-05,
+      "loss": 0.9784,
+      "step": 4877
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.504364013671875,
+      "learning_rate": 1.7442938047022484e-05,
+      "loss": 0.7913,
+      "step": 4878
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.046695709228516,
+      "learning_rate": 1.7440363823579888e-05,
+      "loss": 0.9239,
+      "step": 4879
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.872415542602539,
+      "learning_rate": 1.7437789600137294e-05,
+      "loss": 0.4716,
+      "step": 4880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.566971778869629,
+      "learning_rate": 1.7435215376694698e-05,
+      "loss": 0.4739,
+      "step": 4881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.889461517333984,
+      "learning_rate": 1.74326411532521e-05,
+      "loss": 0.6172,
+      "step": 4882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.186433792114258,
+      "learning_rate": 1.7430066929809508e-05,
+      "loss": 0.5798,
+      "step": 4883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.673079490661621,
+      "learning_rate": 1.742749270636691e-05,
+      "loss": 0.7373,
+      "step": 4884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.75408935546875,
+      "learning_rate": 1.7424918482924318e-05,
+      "loss": 0.8041,
+      "step": 4885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.072247505187988,
+      "learning_rate": 1.7422344259481724e-05,
+      "loss": 0.6379,
+      "step": 4886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.713211059570312,
+      "learning_rate": 1.741977003603913e-05,
+      "loss": 0.7199,
+      "step": 4887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.176946640014648,
+      "learning_rate": 1.7417195812596534e-05,
+      "loss": 0.674,
+      "step": 4888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.093537330627441,
+      "learning_rate": 1.741462158915394e-05,
+      "loss": 0.7224,
+      "step": 4889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.163803100585938,
+      "learning_rate": 1.7412047365711344e-05,
+      "loss": 0.722,
+      "step": 4890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.443512916564941,
+      "learning_rate": 1.7409473142268748e-05,
+      "loss": 0.5932,
+      "step": 4891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.385797500610352,
+      "learning_rate": 1.7406898918826154e-05,
+      "loss": 0.6597,
+      "step": 4892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.46291732788086,
+      "learning_rate": 1.7404324695383558e-05,
+      "loss": 0.7738,
+      "step": 4893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.094919204711914,
+      "learning_rate": 1.7401750471940964e-05,
+      "loss": 0.694,
+      "step": 4894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.094781875610352,
+      "learning_rate": 1.7399176248498368e-05,
+      "loss": 0.6353,
+      "step": 4895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.404618263244629,
+      "learning_rate": 1.7396602025055778e-05,
+      "loss": 0.6247,
+      "step": 4896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.333617210388184,
+      "learning_rate": 1.739402780161318e-05,
+      "loss": 0.9858,
+      "step": 4897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.884780883789062,
+      "learning_rate": 1.7391453578170588e-05,
+      "loss": 0.7308,
+      "step": 4898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.797029495239258,
+      "learning_rate": 1.738887935472799e-05,
+      "loss": 0.7795,
+      "step": 4899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.812067031860352,
+      "learning_rate": 1.7386305131285398e-05,
+      "loss": 0.9258,
+      "step": 4900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.568648338317871,
+      "learning_rate": 1.73837309078428e-05,
+      "loss": 0.6762,
+      "step": 4901
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.883368492126465,
+      "learning_rate": 1.7381156684400204e-05,
+      "loss": 0.7522,
+      "step": 4902
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.427814483642578,
+      "learning_rate": 1.737858246095761e-05,
+      "loss": 0.7882,
+      "step": 4903
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.573358535766602,
+      "learning_rate": 1.7376008237515014e-05,
+      "loss": 0.7208,
+      "step": 4904
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.028850555419922,
+      "learning_rate": 1.7373434014072424e-05,
+      "loss": 0.9841,
+      "step": 4905
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.733213424682617,
+      "learning_rate": 1.7370859790629827e-05,
+      "loss": 0.7934,
+      "step": 4906
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.079960823059082,
+      "learning_rate": 1.7368285567187234e-05,
+      "loss": 0.7694,
+      "step": 4907
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.276900291442871,
+      "learning_rate": 1.7365711343744637e-05,
+      "loss": 0.7879,
+      "step": 4908
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.477254867553711,
+      "learning_rate": 1.7363137120302044e-05,
+      "loss": 0.7624,
+      "step": 4909
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.625418663024902,
+      "learning_rate": 1.7360562896859447e-05,
+      "loss": 0.9777,
+      "step": 4910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.324193954467773,
+      "learning_rate": 1.7357988673416854e-05,
+      "loss": 0.7598,
+      "step": 4911
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.814605712890625,
+      "learning_rate": 1.7355414449974257e-05,
+      "loss": 0.7478,
+      "step": 4912
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.550097465515137,
+      "learning_rate": 1.735284022653166e-05,
+      "loss": 0.6007,
+      "step": 4913
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.572664260864258,
+      "learning_rate": 1.7350266003089067e-05,
+      "loss": 0.8189,
+      "step": 4914
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.535879135131836,
+      "learning_rate": 1.7347691779646474e-05,
+      "loss": 0.7733,
+      "step": 4915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.643726348876953,
+      "learning_rate": 1.734511755620388e-05,
+      "loss": 0.8174,
+      "step": 4916
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.396097183227539,
+      "learning_rate": 1.7342543332761284e-05,
+      "loss": 0.914,
+      "step": 4917
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.25436019897461,
+      "learning_rate": 1.733996910931869e-05,
+      "loss": 0.8079,
+      "step": 4918
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.029252052307129,
+      "learning_rate": 1.7337394885876094e-05,
+      "loss": 0.7383,
+      "step": 4919
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.062467575073242,
+      "learning_rate": 1.73348206624335e-05,
+      "loss": 0.5442,
+      "step": 4920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.284063339233398,
+      "learning_rate": 1.7332246438990904e-05,
+      "loss": 0.6925,
+      "step": 4921
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.368736267089844,
+      "learning_rate": 1.7329672215548307e-05,
+      "loss": 0.786,
+      "step": 4922
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.50294017791748,
+      "learning_rate": 1.7327097992105714e-05,
+      "loss": 1.0086,
+      "step": 4923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.022612571716309,
+      "learning_rate": 1.732452376866312e-05,
+      "loss": 0.6656,
+      "step": 4924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.516199111938477,
+      "learning_rate": 1.7321949545220527e-05,
+      "loss": 0.6494,
+      "step": 4925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.147618293762207,
+      "learning_rate": 1.731937532177793e-05,
+      "loss": 0.8569,
+      "step": 4926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.019527435302734,
+      "learning_rate": 1.7316801098335337e-05,
+      "loss": 1.1321,
+      "step": 4927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.74247932434082,
+      "learning_rate": 1.731422687489274e-05,
+      "loss": 0.9128,
+      "step": 4928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.714862823486328,
+      "learning_rate": 1.7311652651450147e-05,
+      "loss": 1.023,
+      "step": 4929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.445389747619629,
+      "learning_rate": 1.730907842800755e-05,
+      "loss": 0.7577,
+      "step": 4930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.68558406829834,
+      "learning_rate": 1.7306504204564957e-05,
+      "loss": 0.7211,
+      "step": 4931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.790580749511719,
+      "learning_rate": 1.730392998112236e-05,
+      "loss": 0.8245,
+      "step": 4932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.841532707214355,
+      "learning_rate": 1.7301355757679764e-05,
+      "loss": 0.5434,
+      "step": 4933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.869279384613037,
+      "learning_rate": 1.7298781534237174e-05,
+      "loss": 0.5786,
+      "step": 4934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.581611633300781,
+      "learning_rate": 1.7296207310794577e-05,
+      "loss": 0.9899,
+      "step": 4935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.266263961791992,
+      "learning_rate": 1.7293633087351984e-05,
+      "loss": 0.8071,
+      "step": 4936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.568883895874023,
+      "learning_rate": 1.7291058863909387e-05,
+      "loss": 0.9715,
+      "step": 4937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.261860847473145,
+      "learning_rate": 1.7288484640466794e-05,
+      "loss": 0.765,
+      "step": 4938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.793584823608398,
+      "learning_rate": 1.7285910417024197e-05,
+      "loss": 0.6247,
+      "step": 4939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.055368423461914,
+      "learning_rate": 1.7283336193581604e-05,
+      "loss": 0.8427,
+      "step": 4940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.205330848693848,
+      "learning_rate": 1.7280761970139007e-05,
+      "loss": 0.6852,
+      "step": 4941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.361076354980469,
+      "learning_rate": 1.7278187746696414e-05,
+      "loss": 0.7573,
+      "step": 4942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.780275344848633,
+      "learning_rate": 1.727561352325382e-05,
+      "loss": 0.6961,
+      "step": 4943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.784602165222168,
+      "learning_rate": 1.7273039299811224e-05,
+      "loss": 0.8277,
+      "step": 4944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.712226867675781,
+      "learning_rate": 1.727046507636863e-05,
+      "loss": 0.5497,
+      "step": 4945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.916309356689453,
+      "learning_rate": 1.7267890852926034e-05,
+      "loss": 0.6639,
+      "step": 4946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.422940254211426,
+      "learning_rate": 1.726531662948344e-05,
+      "loss": 0.682,
+      "step": 4947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.396326065063477,
+      "learning_rate": 1.7262742406040844e-05,
+      "loss": 0.682,
+      "step": 4948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.037121772766113,
+      "learning_rate": 1.726016818259825e-05,
+      "loss": 0.8204,
+      "step": 4949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.892542839050293,
+      "learning_rate": 1.7257593959155654e-05,
+      "loss": 0.7595,
+      "step": 4950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.303763389587402,
+      "learning_rate": 1.725501973571306e-05,
+      "loss": 0.7321,
+      "step": 4951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.016682624816895,
+      "learning_rate": 1.7252445512270464e-05,
+      "loss": 0.8576,
+      "step": 4952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.216480255126953,
+      "learning_rate": 1.724987128882787e-05,
+      "loss": 0.5196,
+      "step": 4953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.688359260559082,
+      "learning_rate": 1.7247297065385277e-05,
+      "loss": 0.6283,
+      "step": 4954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.238533973693848,
+      "learning_rate": 1.724472284194268e-05,
+      "loss": 0.6993,
+      "step": 4955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.846015930175781,
+      "learning_rate": 1.7242148618500087e-05,
+      "loss": 0.6817,
+      "step": 4956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.082934379577637,
+      "learning_rate": 1.723957439505749e-05,
+      "loss": 0.774,
+      "step": 4957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.129729270935059,
+      "learning_rate": 1.7237000171614897e-05,
+      "loss": 0.7532,
+      "step": 4958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.17959976196289,
+      "learning_rate": 1.72344259481723e-05,
+      "loss": 0.7147,
+      "step": 4959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.005559921264648,
+      "learning_rate": 1.7231851724729707e-05,
+      "loss": 0.6481,
+      "step": 4960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.59207534790039,
+      "learning_rate": 1.722927750128711e-05,
+      "loss": 1.1346,
+      "step": 4961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.42135238647461,
+      "learning_rate": 1.722670327784452e-05,
+      "loss": 0.9172,
+      "step": 4962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.474270820617676,
+      "learning_rate": 1.7224129054401924e-05,
+      "loss": 0.6278,
+      "step": 4963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.794137954711914,
+      "learning_rate": 1.7221554830959327e-05,
+      "loss": 0.6219,
+      "step": 4964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.153670310974121,
+      "learning_rate": 1.7218980607516734e-05,
+      "loss": 0.6652,
+      "step": 4965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.129912376403809,
+      "learning_rate": 1.7216406384074137e-05,
+      "loss": 0.7677,
+      "step": 4966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.544198036193848,
+      "learning_rate": 1.7213832160631544e-05,
+      "loss": 1.0513,
+      "step": 4967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.749582290649414,
+      "learning_rate": 1.7211257937188947e-05,
+      "loss": 0.827,
+      "step": 4968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.035421371459961,
+      "learning_rate": 1.7208683713746354e-05,
+      "loss": 0.7915,
+      "step": 4969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.850958824157715,
+      "learning_rate": 1.7206109490303757e-05,
+      "loss": 0.6275,
+      "step": 4970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.161952018737793,
+      "learning_rate": 1.7203535266861167e-05,
+      "loss": 0.8849,
+      "step": 4971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.796048164367676,
+      "learning_rate": 1.720096104341857e-05,
+      "loss": 0.7508,
+      "step": 4972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.304971694946289,
+      "learning_rate": 1.7198386819975977e-05,
+      "loss": 0.6869,
+      "step": 4973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.204222679138184,
+      "learning_rate": 1.719581259653338e-05,
+      "loss": 0.87,
+      "step": 4974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.814250946044922,
+      "learning_rate": 1.7193238373090783e-05,
+      "loss": 0.4595,
+      "step": 4975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.42802906036377,
+      "learning_rate": 1.719066414964819e-05,
+      "loss": 0.6612,
+      "step": 4976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.03331184387207,
+      "learning_rate": 1.7188089926205593e-05,
+      "loss": 0.5956,
+      "step": 4977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.797418594360352,
+      "learning_rate": 1.7185515702763e-05,
+      "loss": 0.7991,
+      "step": 4978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.321005821228027,
+      "learning_rate": 1.7182941479320403e-05,
+      "loss": 1.0181,
+      "step": 4979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.11483383178711,
+      "learning_rate": 1.718036725587781e-05,
+      "loss": 0.6289,
+      "step": 4980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.9276123046875,
+      "learning_rate": 1.7177793032435217e-05,
+      "loss": 0.7497,
+      "step": 4981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.66787052154541,
+      "learning_rate": 1.7175218808992623e-05,
+      "loss": 0.7535,
+      "step": 4982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.72194766998291,
+      "learning_rate": 1.7172644585550027e-05,
+      "loss": 0.753,
+      "step": 4983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.778253555297852,
+      "learning_rate": 1.7170070362107433e-05,
+      "loss": 0.6282,
+      "step": 4984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.868002891540527,
+      "learning_rate": 1.7167496138664837e-05,
+      "loss": 0.7105,
+      "step": 4985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.45827865600586,
+      "learning_rate": 1.716492191522224e-05,
+      "loss": 0.636,
+      "step": 4986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.1046781539917,
+      "learning_rate": 1.7162347691779647e-05,
+      "loss": 0.8897,
+      "step": 4987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.193692207336426,
+      "learning_rate": 1.715977346833705e-05,
+      "loss": 0.7267,
+      "step": 4988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.283628463745117,
+      "learning_rate": 1.7157199244894457e-05,
+      "loss": 0.75,
+      "step": 4989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.282515525817871,
+      "learning_rate": 1.7154625021451863e-05,
+      "loss": 0.7044,
+      "step": 4990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.588621139526367,
+      "learning_rate": 1.715205079800927e-05,
+      "loss": 0.6408,
+      "step": 4991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.896646499633789,
+      "learning_rate": 1.7149476574566673e-05,
+      "loss": 0.7279,
+      "step": 4992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.332626342773438,
+      "learning_rate": 1.714690235112408e-05,
+      "loss": 0.6586,
+      "step": 4993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.131858825683594,
+      "learning_rate": 1.7144328127681483e-05,
+      "loss": 0.883,
+      "step": 4994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.531006813049316,
+      "learning_rate": 1.7141753904238887e-05,
+      "loss": 0.8161,
+      "step": 4995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.387847900390625,
+      "learning_rate": 1.7139179680796293e-05,
+      "loss": 0.605,
+      "step": 4996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.156927108764648,
+      "learning_rate": 1.7136605457353697e-05,
+      "loss": 0.9032,
+      "step": 4997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.074671745300293,
+      "learning_rate": 1.7134031233911103e-05,
+      "loss": 0.737,
+      "step": 4998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.010988235473633,
+      "learning_rate": 1.7131457010468507e-05,
+      "loss": 0.4996,
+      "step": 4999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.418106079101562,
+      "learning_rate": 1.7128882787025917e-05,
+      "loss": 0.8099,
+      "step": 5000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.001021385192871,
+      "learning_rate": 1.712630856358332e-05,
+      "loss": 0.8199,
+      "step": 5001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.040778160095215,
+      "learning_rate": 1.7123734340140727e-05,
+      "loss": 0.5868,
+      "step": 5002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.183379173278809,
+      "learning_rate": 1.712116011669813e-05,
+      "loss": 0.9051,
+      "step": 5003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.451287269592285,
+      "learning_rate": 1.7118585893255537e-05,
+      "loss": 0.961,
+      "step": 5004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.420075416564941,
+      "learning_rate": 1.711601166981294e-05,
+      "loss": 0.6247,
+      "step": 5005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.370603561401367,
+      "learning_rate": 1.7113437446370343e-05,
+      "loss": 0.8457,
+      "step": 5006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.350581169128418,
+      "learning_rate": 1.711086322292775e-05,
+      "loss": 0.5281,
+      "step": 5007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.304046630859375,
+      "learning_rate": 1.7108288999485153e-05,
+      "loss": 0.8896,
+      "step": 5008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.822914123535156,
+      "learning_rate": 1.7105714776042563e-05,
+      "loss": 0.6603,
+      "step": 5009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.38222885131836,
+      "learning_rate": 1.7103140552599966e-05,
+      "loss": 0.7601,
+      "step": 5010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.930339813232422,
+      "learning_rate": 1.7100566329157373e-05,
+      "loss": 0.7744,
+      "step": 5011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.399715423583984,
+      "learning_rate": 1.7097992105714776e-05,
+      "loss": 0.4404,
+      "step": 5012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.872371673583984,
+      "learning_rate": 1.7095417882272183e-05,
+      "loss": 0.5168,
+      "step": 5013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.332186698913574,
+      "learning_rate": 1.7092843658829586e-05,
+      "loss": 0.8214,
+      "step": 5014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.342545509338379,
+      "learning_rate": 1.7090269435386993e-05,
+      "loss": 0.8542,
+      "step": 5015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.146199226379395,
+      "learning_rate": 1.7087695211944396e-05,
+      "loss": 0.5202,
+      "step": 5016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.357397079467773,
+      "learning_rate": 1.70851209885018e-05,
+      "loss": 0.6796,
+      "step": 5017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.00622844696045,
+      "learning_rate": 1.7082546765059206e-05,
+      "loss": 0.8675,
+      "step": 5018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.33556079864502,
+      "learning_rate": 1.7079972541616613e-05,
+      "loss": 0.6171,
+      "step": 5019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.016263008117676,
+      "learning_rate": 1.707739831817402e-05,
+      "loss": 0.6974,
+      "step": 5020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.643027305603027,
+      "learning_rate": 1.7074824094731423e-05,
+      "loss": 1.0009,
+      "step": 5021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.957036018371582,
+      "learning_rate": 1.707224987128883e-05,
+      "loss": 1.0078,
+      "step": 5022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.904595375061035,
+      "learning_rate": 1.7069675647846233e-05,
+      "loss": 0.6725,
+      "step": 5023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.033756256103516,
+      "learning_rate": 1.706710142440364e-05,
+      "loss": 0.5452,
+      "step": 5024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.128388404846191,
+      "learning_rate": 1.7064527200961043e-05,
+      "loss": 0.796,
+      "step": 5025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.647265434265137,
+      "learning_rate": 1.7061952977518446e-05,
+      "loss": 0.7683,
+      "step": 5026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.176582336425781,
+      "learning_rate": 1.7059378754075853e-05,
+      "loss": 0.6503,
+      "step": 5027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.657790184020996,
+      "learning_rate": 1.705680453063326e-05,
+      "loss": 0.9437,
+      "step": 5028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.837785720825195,
+      "learning_rate": 1.7054230307190666e-05,
+      "loss": 0.6766,
+      "step": 5029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.116557121276855,
+      "learning_rate": 1.705165608374807e-05,
+      "loss": 0.8057,
+      "step": 5030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.438505172729492,
+      "learning_rate": 1.7049081860305476e-05,
+      "loss": 0.6892,
+      "step": 5031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.117741584777832,
+      "learning_rate": 1.704650763686288e-05,
+      "loss": 0.7923,
+      "step": 5032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.207502365112305,
+      "learning_rate": 1.7043933413420286e-05,
+      "loss": 0.7875,
+      "step": 5033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.58642578125,
+      "learning_rate": 1.704135918997769e-05,
+      "loss": 0.6887,
+      "step": 5034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.861223220825195,
+      "learning_rate": 1.7038784966535096e-05,
+      "loss": 0.7291,
+      "step": 5035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.419525146484375,
+      "learning_rate": 1.70362107430925e-05,
+      "loss": 0.7209,
+      "step": 5036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.348834037780762,
+      "learning_rate": 1.7033636519649903e-05,
+      "loss": 0.8206,
+      "step": 5037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.353058815002441,
+      "learning_rate": 1.7031062296207313e-05,
+      "loss": 0.91,
+      "step": 5038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.92310905456543,
+      "learning_rate": 1.7028488072764716e-05,
+      "loss": 0.9756,
+      "step": 5039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.245994567871094,
+      "learning_rate": 1.7025913849322123e-05,
+      "loss": 0.7373,
+      "step": 5040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.201173782348633,
+      "learning_rate": 1.7023339625879526e-05,
+      "loss": 1.0927,
+      "step": 5041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.186721801757812,
+      "learning_rate": 1.7020765402436933e-05,
+      "loss": 0.9485,
+      "step": 5042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.173857688903809,
+      "learning_rate": 1.7018191178994336e-05,
+      "loss": 0.7514,
+      "step": 5043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.36408519744873,
+      "learning_rate": 1.7015616955551743e-05,
+      "loss": 0.7456,
+      "step": 5044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.549567222595215,
+      "learning_rate": 1.7013042732109146e-05,
+      "loss": 0.656,
+      "step": 5045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.760339736938477,
+      "learning_rate": 1.7010468508666553e-05,
+      "loss": 0.7951,
+      "step": 5046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.459860801696777,
+      "learning_rate": 1.700789428522396e-05,
+      "loss": 0.678,
+      "step": 5047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.143653869628906,
+      "learning_rate": 1.7005320061781363e-05,
+      "loss": 0.5365,
+      "step": 5048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.60336971282959,
+      "learning_rate": 1.700274583833877e-05,
+      "loss": 0.719,
+      "step": 5049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.393378257751465,
+      "learning_rate": 1.7000171614896173e-05,
+      "loss": 0.7732,
+      "step": 5050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.6954345703125,
+      "learning_rate": 1.699759739145358e-05,
+      "loss": 0.6829,
+      "step": 5051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.231731414794922,
+      "learning_rate": 1.6995023168010983e-05,
+      "loss": 0.7362,
+      "step": 5052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.381374359130859,
+      "learning_rate": 1.699244894456839e-05,
+      "loss": 0.4342,
+      "step": 5053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.06167221069336,
+      "learning_rate": 1.6989874721125793e-05,
+      "loss": 0.7154,
+      "step": 5054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.840649604797363,
+      "learning_rate": 1.69873004976832e-05,
+      "loss": 0.6862,
+      "step": 5055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.552257537841797,
+      "learning_rate": 1.6984726274240603e-05,
+      "loss": 0.9419,
+      "step": 5056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.155996322631836,
+      "learning_rate": 1.698215205079801e-05,
+      "loss": 0.9167,
+      "step": 5057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.930998802185059,
+      "learning_rate": 1.6979577827355416e-05,
+      "loss": 0.8857,
+      "step": 5058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.22996997833252,
+      "learning_rate": 1.697700360391282e-05,
+      "loss": 0.5786,
+      "step": 5059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.228675842285156,
+      "learning_rate": 1.6974429380470226e-05,
+      "loss": 0.7936,
+      "step": 5060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.768718719482422,
+      "learning_rate": 1.697185515702763e-05,
+      "loss": 0.8322,
+      "step": 5061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.743788719177246,
+      "learning_rate": 1.6969280933585036e-05,
+      "loss": 0.7182,
+      "step": 5062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.028605461120605,
+      "learning_rate": 1.696670671014244e-05,
+      "loss": 0.9524,
+      "step": 5063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.297452926635742,
+      "learning_rate": 1.6964132486699846e-05,
+      "loss": 0.9374,
+      "step": 5064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.017545700073242,
+      "learning_rate": 1.696155826325725e-05,
+      "loss": 0.751,
+      "step": 5065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.493666648864746,
+      "learning_rate": 1.695898403981466e-05,
+      "loss": 0.6565,
+      "step": 5066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.744612693786621,
+      "learning_rate": 1.6956409816372063e-05,
+      "loss": 0.7618,
+      "step": 5067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.950468063354492,
+      "learning_rate": 1.6953835592929466e-05,
+      "loss": 0.8004,
+      "step": 5068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.73985767364502,
+      "learning_rate": 1.6951261369486873e-05,
+      "loss": 0.7173,
+      "step": 5069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.84887981414795,
+      "learning_rate": 1.6948687146044276e-05,
+      "loss": 0.7528,
+      "step": 5070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.326107025146484,
+      "learning_rate": 1.6946112922601683e-05,
+      "loss": 0.6849,
+      "step": 5071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.734704971313477,
+      "learning_rate": 1.6943538699159086e-05,
+      "loss": 0.5548,
+      "step": 5072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.152992248535156,
+      "learning_rate": 1.6940964475716493e-05,
+      "loss": 0.9213,
+      "step": 5073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.008866310119629,
+      "learning_rate": 1.6938390252273896e-05,
+      "loss": 0.7313,
+      "step": 5074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.599018096923828,
+      "learning_rate": 1.6935816028831306e-05,
+      "loss": 0.7501,
+      "step": 5075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.486008644104004,
+      "learning_rate": 1.693324180538871e-05,
+      "loss": 0.7916,
+      "step": 5076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.657783508300781,
+      "learning_rate": 1.6930667581946116e-05,
+      "loss": 0.6116,
+      "step": 5077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.07896614074707,
+      "learning_rate": 1.692809335850352e-05,
+      "loss": 0.6275,
+      "step": 5078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.535330772399902,
+      "learning_rate": 1.6925519135060922e-05,
+      "loss": 0.8092,
+      "step": 5079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.815303802490234,
+      "learning_rate": 1.692294491161833e-05,
+      "loss": 0.9915,
+      "step": 5080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.145001411437988,
+      "learning_rate": 1.6920370688175732e-05,
+      "loss": 0.6142,
+      "step": 5081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.141441345214844,
+      "learning_rate": 1.691779646473314e-05,
+      "loss": 0.9283,
+      "step": 5082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.250448226928711,
+      "learning_rate": 1.6915222241290542e-05,
+      "loss": 0.9554,
+      "step": 5083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.658493995666504,
+      "learning_rate": 1.691264801784795e-05,
+      "loss": 0.5825,
+      "step": 5084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.949767112731934,
+      "learning_rate": 1.6910073794405356e-05,
+      "loss": 0.6711,
+      "step": 5085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.661419868469238,
+      "learning_rate": 1.6907499570962762e-05,
+      "loss": 0.6141,
+      "step": 5086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.766616821289062,
+      "learning_rate": 1.6904925347520166e-05,
+      "loss": 0.8972,
+      "step": 5087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.832223892211914,
+      "learning_rate": 1.6902351124077572e-05,
+      "loss": 0.5916,
+      "step": 5088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.83588695526123,
+      "learning_rate": 1.6899776900634976e-05,
+      "loss": 0.4926,
+      "step": 5089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.589808464050293,
+      "learning_rate": 1.689720267719238e-05,
+      "loss": 0.9012,
+      "step": 5090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.086411476135254,
+      "learning_rate": 1.6894628453749786e-05,
+      "loss": 0.8289,
+      "step": 5091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.400313377380371,
+      "learning_rate": 1.689205423030719e-05,
+      "loss": 0.7144,
+      "step": 5092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.67238712310791,
+      "learning_rate": 1.6889480006864596e-05,
+      "loss": 0.6807,
+      "step": 5093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.829276084899902,
+      "learning_rate": 1.6886905783422002e-05,
+      "loss": 0.5496,
+      "step": 5094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.994205474853516,
+      "learning_rate": 1.688433155997941e-05,
+      "loss": 0.5911,
+      "step": 5095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.82376480102539,
+      "learning_rate": 1.6881757336536812e-05,
+      "loss": 0.9673,
+      "step": 5096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.390604019165039,
+      "learning_rate": 1.687918311309422e-05,
+      "loss": 0.7123,
+      "step": 5097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.539375305175781,
+      "learning_rate": 1.6876608889651622e-05,
+      "loss": 0.6047,
+      "step": 5098
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.492993354797363,
+      "learning_rate": 1.6874034666209026e-05,
+      "loss": 0.8519,
+      "step": 5099
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.150948524475098,
+      "learning_rate": 1.6871460442766432e-05,
+      "loss": 0.8469,
+      "step": 5100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.869233131408691,
+      "learning_rate": 1.6868886219323836e-05,
+      "loss": 0.5796,
+      "step": 5101
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.846354484558105,
+      "learning_rate": 1.6866311995881242e-05,
+      "loss": 0.8305,
+      "step": 5102
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.064055442810059,
+      "learning_rate": 1.6863737772438646e-05,
+      "loss": 0.8497,
+      "step": 5103
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.882604598999023,
+      "learning_rate": 1.6861163548996056e-05,
+      "loss": 0.4962,
+      "step": 5104
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.37267017364502,
+      "learning_rate": 1.685858932555346e-05,
+      "loss": 0.8019,
+      "step": 5105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.571423530578613,
+      "learning_rate": 1.6856015102110866e-05,
+      "loss": 0.6557,
+      "step": 5106
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.922707557678223,
+      "learning_rate": 1.685344087866827e-05,
+      "loss": 0.6606,
+      "step": 5107
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.683191299438477,
+      "learning_rate": 1.6850866655225676e-05,
+      "loss": 0.7353,
+      "step": 5108
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.57392406463623,
+      "learning_rate": 1.684829243178308e-05,
+      "loss": 0.8121,
+      "step": 5109
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.889533996582031,
+      "learning_rate": 1.6845718208340482e-05,
+      "loss": 0.6311,
+      "step": 5110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.42176342010498,
+      "learning_rate": 1.684314398489789e-05,
+      "loss": 0.6659,
+      "step": 5111
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.499669075012207,
+      "learning_rate": 1.6840569761455292e-05,
+      "loss": 0.59,
+      "step": 5112
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.916189193725586,
+      "learning_rate": 1.6837995538012702e-05,
+      "loss": 0.7427,
+      "step": 5113
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.4607572555542,
+      "learning_rate": 1.6835421314570105e-05,
+      "loss": 0.7307,
+      "step": 5114
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.490965843200684,
+      "learning_rate": 1.6832847091127512e-05,
+      "loss": 0.9674,
+      "step": 5115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.16115951538086,
+      "learning_rate": 1.6830272867684915e-05,
+      "loss": 0.678,
+      "step": 5116
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.330124855041504,
+      "learning_rate": 1.6827698644242322e-05,
+      "loss": 0.4714,
+      "step": 5117
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.619162559509277,
+      "learning_rate": 1.6825124420799725e-05,
+      "loss": 0.7703,
+      "step": 5118
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.129340171813965,
+      "learning_rate": 1.6822550197357132e-05,
+      "loss": 0.7886,
+      "step": 5119
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.748601913452148,
+      "learning_rate": 1.6819975973914535e-05,
+      "loss": 0.8321,
+      "step": 5120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.726327896118164,
+      "learning_rate": 1.681740175047194e-05,
+      "loss": 0.744,
+      "step": 5121
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.06992244720459,
+      "learning_rate": 1.6814827527029345e-05,
+      "loss": 0.8202,
+      "step": 5122
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.089544296264648,
+      "learning_rate": 1.6812253303586752e-05,
+      "loss": 0.6877,
+      "step": 5123
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.7350435256958,
+      "learning_rate": 1.680967908014416e-05,
+      "loss": 0.5948,
+      "step": 5124
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.103888511657715,
+      "learning_rate": 1.6807104856701562e-05,
+      "loss": 0.5181,
+      "step": 5125
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.345788955688477,
+      "learning_rate": 1.680453063325897e-05,
+      "loss": 0.5736,
+      "step": 5126
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.09771728515625,
+      "learning_rate": 1.6801956409816372e-05,
+      "loss": 0.4244,
+      "step": 5127
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.414392471313477,
+      "learning_rate": 1.679938218637378e-05,
+      "loss": 0.6431,
+      "step": 5128
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.769451141357422,
+      "learning_rate": 1.6796807962931182e-05,
+      "loss": 0.7522,
+      "step": 5129
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.722218990325928,
+      "learning_rate": 1.6794233739488585e-05,
+      "loss": 0.4948,
+      "step": 5130
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.459582328796387,
+      "learning_rate": 1.6791659516045992e-05,
+      "loss": 0.788,
+      "step": 5131
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.663519859313965,
+      "learning_rate": 1.67890852926034e-05,
+      "loss": 0.7317,
+      "step": 5132
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.588348388671875,
+      "learning_rate": 1.6786511069160805e-05,
+      "loss": 0.7427,
+      "step": 5133
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.5723876953125,
+      "learning_rate": 1.678393684571821e-05,
+      "loss": 0.8075,
+      "step": 5134
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.216257095336914,
+      "learning_rate": 1.6781362622275615e-05,
+      "loss": 0.6408,
+      "step": 5135
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.186063766479492,
+      "learning_rate": 1.677878839883302e-05,
+      "loss": 0.499,
+      "step": 5136
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.45901107788086,
+      "learning_rate": 1.6776214175390425e-05,
+      "loss": 0.6864,
+      "step": 5137
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.202424049377441,
+      "learning_rate": 1.677363995194783e-05,
+      "loss": 0.6882,
+      "step": 5138
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.158854484558105,
+      "learning_rate": 1.6771065728505235e-05,
+      "loss": 0.8581,
+      "step": 5139
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.483059883117676,
+      "learning_rate": 1.676849150506264e-05,
+      "loss": 0.6036,
+      "step": 5140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.016020774841309,
+      "learning_rate": 1.6765917281620042e-05,
+      "loss": 0.8377,
+      "step": 5141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.223566055297852,
+      "learning_rate": 1.6763343058177452e-05,
+      "loss": 0.7415,
+      "step": 5142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.837448120117188,
+      "learning_rate": 1.6760768834734855e-05,
+      "loss": 0.459,
+      "step": 5143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.283288955688477,
+      "learning_rate": 1.6758194611292262e-05,
+      "loss": 1.038,
+      "step": 5144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.15881633758545,
+      "learning_rate": 1.6755620387849665e-05,
+      "loss": 0.6349,
+      "step": 5145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.535287857055664,
+      "learning_rate": 1.6753046164407072e-05,
+      "loss": 0.688,
+      "step": 5146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.299918174743652,
+      "learning_rate": 1.6750471940964475e-05,
+      "loss": 1.0001,
+      "step": 5147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.1094970703125,
+      "learning_rate": 1.6747897717521882e-05,
+      "loss": 0.7172,
+      "step": 5148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.257475852966309,
+      "learning_rate": 1.6745323494079285e-05,
+      "loss": 0.544,
+      "step": 5149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.434012413024902,
+      "learning_rate": 1.6742749270636692e-05,
+      "loss": 0.5604,
+      "step": 5150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.975931167602539,
+      "learning_rate": 1.67401750471941e-05,
+      "loss": 0.8215,
+      "step": 5151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.961003303527832,
+      "learning_rate": 1.6737600823751502e-05,
+      "loss": 0.9019,
+      "step": 5152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.639623641967773,
+      "learning_rate": 1.673502660030891e-05,
+      "loss": 0.7083,
+      "step": 5153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.64686107635498,
+      "learning_rate": 1.6732452376866312e-05,
+      "loss": 0.6626,
+      "step": 5154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.169412612915039,
+      "learning_rate": 1.672987815342372e-05,
+      "loss": 0.7328,
+      "step": 5155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.250401496887207,
+      "learning_rate": 1.6727303929981122e-05,
+      "loss": 0.5575,
+      "step": 5156
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.876055717468262,
+      "learning_rate": 1.672472970653853e-05,
+      "loss": 0.7575,
+      "step": 5157
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.425901412963867,
+      "learning_rate": 1.6722155483095932e-05,
+      "loss": 0.6735,
+      "step": 5158
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.946558952331543,
+      "learning_rate": 1.671958125965334e-05,
+      "loss": 0.4543,
+      "step": 5159
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.222345352172852,
+      "learning_rate": 1.671700703621074e-05,
+      "loss": 0.8022,
+      "step": 5160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.4634428024292,
+      "learning_rate": 1.671443281276815e-05,
+      "loss": 0.7809,
+      "step": 5161
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.227447509765625,
+      "learning_rate": 1.6711858589325555e-05,
+      "loss": 0.8119,
+      "step": 5162
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.55435848236084,
+      "learning_rate": 1.670928436588296e-05,
+      "loss": 0.8565,
+      "step": 5163
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.997054100036621,
+      "learning_rate": 1.6706710142440365e-05,
+      "loss": 1.0876,
+      "step": 5164
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.246514320373535,
+      "learning_rate": 1.670413591899777e-05,
+      "loss": 0.7111,
+      "step": 5165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.147298812866211,
+      "learning_rate": 1.6701561695555175e-05,
+      "loss": 0.667,
+      "step": 5166
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.040410041809082,
+      "learning_rate": 1.6698987472112578e-05,
+      "loss": 0.8043,
+      "step": 5167
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.274188041687012,
+      "learning_rate": 1.6696413248669985e-05,
+      "loss": 0.5311,
+      "step": 5168
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.591151237487793,
+      "learning_rate": 1.6693839025227388e-05,
+      "loss": 0.8394,
+      "step": 5169
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.305787086486816,
+      "learning_rate": 1.66912648017848e-05,
+      "loss": 0.8099,
+      "step": 5170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.254986763000488,
+      "learning_rate": 1.66886905783422e-05,
+      "loss": 0.7154,
+      "step": 5171
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.92106819152832,
+      "learning_rate": 1.6686116354899605e-05,
+      "loss": 0.8145,
+      "step": 5172
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.297516822814941,
+      "learning_rate": 1.668354213145701e-05,
+      "loss": 0.7931,
+      "step": 5173
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.657331466674805,
+      "learning_rate": 1.6680967908014415e-05,
+      "loss": 0.7878,
+      "step": 5174
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.934943199157715,
+      "learning_rate": 1.667839368457182e-05,
+      "loss": 0.7537,
+      "step": 5175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.230857849121094,
+      "learning_rate": 1.6675819461129225e-05,
+      "loss": 0.7391,
+      "step": 5176
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.595080375671387,
+      "learning_rate": 1.667324523768663e-05,
+      "loss": 0.62,
+      "step": 5177
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.526241302490234,
+      "learning_rate": 1.6670671014244035e-05,
+      "loss": 0.9329,
+      "step": 5178
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.58484935760498,
+      "learning_rate": 1.666809679080144e-05,
+      "loss": 0.5699,
+      "step": 5179
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.413646697998047,
+      "learning_rate": 1.6665522567358848e-05,
+      "loss": 0.7794,
+      "step": 5180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.08458423614502,
+      "learning_rate": 1.6662948343916255e-05,
+      "loss": 0.7778,
+      "step": 5181
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.542162895202637,
+      "learning_rate": 1.6660374120473658e-05,
+      "loss": 0.6902,
+      "step": 5182
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.577092170715332,
+      "learning_rate": 1.665779989703106e-05,
+      "loss": 0.6979,
+      "step": 5183
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.598038673400879,
+      "learning_rate": 1.6655225673588468e-05,
+      "loss": 0.8171,
+      "step": 5184
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.036426544189453,
+      "learning_rate": 1.665265145014587e-05,
+      "loss": 1.0667,
+      "step": 5185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.628640174865723,
+      "learning_rate": 1.6650077226703278e-05,
+      "loss": 0.9543,
+      "step": 5186
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.614828109741211,
+      "learning_rate": 1.664750300326068e-05,
+      "loss": 0.6174,
+      "step": 5187
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.248889923095703,
+      "learning_rate": 1.6644928779818088e-05,
+      "loss": 0.8864,
+      "step": 5188
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.222512245178223,
+      "learning_rate": 1.6642354556375495e-05,
+      "loss": 0.7374,
+      "step": 5189
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.618529319763184,
+      "learning_rate": 1.66397803329329e-05,
+      "loss": 0.7824,
+      "step": 5190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.544220924377441,
+      "learning_rate": 1.6637206109490305e-05,
+      "loss": 0.7273,
+      "step": 5191
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.569472312927246,
+      "learning_rate": 1.663463188604771e-05,
+      "loss": 0.7878,
+      "step": 5192
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.192878723144531,
+      "learning_rate": 1.6632057662605115e-05,
+      "loss": 0.7754,
+      "step": 5193
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.241219520568848,
+      "learning_rate": 1.6629483439162518e-05,
+      "loss": 0.7687,
+      "step": 5194
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.386351585388184,
+      "learning_rate": 1.6626909215719925e-05,
+      "loss": 0.9269,
+      "step": 5195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.048517227172852,
+      "learning_rate": 1.6624334992277328e-05,
+      "loss": 0.9212,
+      "step": 5196
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.806606769561768,
+      "learning_rate": 1.6621760768834735e-05,
+      "loss": 0.3998,
+      "step": 5197
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.608366966247559,
+      "learning_rate": 1.661918654539214e-05,
+      "loss": 0.7697,
+      "step": 5198
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.722428321838379,
+      "learning_rate": 1.6616612321949548e-05,
+      "loss": 0.7826,
+      "step": 5199
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.099624633789062,
+      "learning_rate": 1.661403809850695e-05,
+      "loss": 0.7108,
+      "step": 5200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.431052207946777,
+      "learning_rate": 1.6611463875064358e-05,
+      "loss": 0.6995,
+      "step": 5201
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.23111343383789,
+      "learning_rate": 1.660888965162176e-05,
+      "loss": 0.8535,
+      "step": 5202
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.609075546264648,
+      "learning_rate": 1.6606315428179165e-05,
+      "loss": 0.6788,
+      "step": 5203
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.281352996826172,
+      "learning_rate": 1.660374120473657e-05,
+      "loss": 0.8985,
+      "step": 5204
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.516215324401855,
+      "learning_rate": 1.6601166981293975e-05,
+      "loss": 0.7634,
+      "step": 5205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.74077033996582,
+      "learning_rate": 1.659859275785138e-05,
+      "loss": 0.8806,
+      "step": 5206
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.645482063293457,
+      "learning_rate": 1.6596018534408785e-05,
+      "loss": 0.5675,
+      "step": 5207
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.549633979797363,
+      "learning_rate": 1.6593444310966195e-05,
+      "loss": 0.7807,
+      "step": 5208
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.692028999328613,
+      "learning_rate": 1.6590870087523598e-05,
+      "loss": 0.8201,
+      "step": 5209
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.625863075256348,
+      "learning_rate": 1.6588295864081005e-05,
+      "loss": 0.7049,
+      "step": 5210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.592692375183105,
+      "learning_rate": 1.6585721640638408e-05,
+      "loss": 0.7609,
+      "step": 5211
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.667631149291992,
+      "learning_rate": 1.6583147417195815e-05,
+      "loss": 0.9495,
+      "step": 5212
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.62667179107666,
+      "learning_rate": 1.6580573193753218e-05,
+      "loss": 0.5782,
+      "step": 5213
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.771811485290527,
+      "learning_rate": 1.657799897031062e-05,
+      "loss": 0.6552,
+      "step": 5214
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.704256057739258,
+      "learning_rate": 1.6575424746868028e-05,
+      "loss": 0.7867,
+      "step": 5215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.04047966003418,
+      "learning_rate": 1.657285052342543e-05,
+      "loss": 0.6639,
+      "step": 5216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.500970840454102,
+      "learning_rate": 1.657027629998284e-05,
+      "loss": 0.6785,
+      "step": 5217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.975128173828125,
+      "learning_rate": 1.6567702076540245e-05,
+      "loss": 0.6367,
+      "step": 5218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.489596366882324,
+      "learning_rate": 1.656512785309765e-05,
+      "loss": 0.7099,
+      "step": 5219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.5138521194458,
+      "learning_rate": 1.6562553629655054e-05,
+      "loss": 0.6627,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.358445167541504,
+      "learning_rate": 1.655997940621246e-05,
+      "loss": 0.5574,
+      "step": 5221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.792037963867188,
+      "learning_rate": 1.6557405182769864e-05,
+      "loss": 0.7517,
+      "step": 5222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.092707633972168,
+      "learning_rate": 1.655483095932727e-05,
+      "loss": 0.6959,
+      "step": 5223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.377098083496094,
+      "learning_rate": 1.6552256735884674e-05,
+      "loss": 0.5483,
+      "step": 5224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.945281982421875,
+      "learning_rate": 1.6549682512442078e-05,
+      "loss": 0.8506,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.849797248840332,
+      "learning_rate": 1.6547108288999484e-05,
+      "loss": 0.7787,
+      "step": 5226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.040619850158691,
+      "learning_rate": 1.654453406555689e-05,
+      "loss": 0.7812,
+      "step": 5227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.434410095214844,
+      "learning_rate": 1.6541959842114298e-05,
+      "loss": 0.742,
+      "step": 5228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.027389526367188,
+      "learning_rate": 1.65393856186717e-05,
+      "loss": 0.6401,
+      "step": 5229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.295379638671875,
+      "learning_rate": 1.6536811395229108e-05,
+      "loss": 0.6541,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.087778091430664,
+      "learning_rate": 1.653423717178651e-05,
+      "loss": 0.4963,
+      "step": 5231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.502301216125488,
+      "learning_rate": 1.6531662948343918e-05,
+      "loss": 0.8584,
+      "step": 5232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.659680366516113,
+      "learning_rate": 1.652908872490132e-05,
+      "loss": 0.7465,
+      "step": 5233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.165035247802734,
+      "learning_rate": 1.6526514501458728e-05,
+      "loss": 0.6075,
+      "step": 5234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.327047348022461,
+      "learning_rate": 1.652394027801613e-05,
+      "loss": 0.8386,
+      "step": 5235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.365690231323242,
+      "learning_rate": 1.6521366054573538e-05,
+      "loss": 0.6087,
+      "step": 5236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.67734146118164,
+      "learning_rate": 1.6518791831130944e-05,
+      "loss": 0.6336,
+      "step": 5237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.418550491333008,
+      "learning_rate": 1.6516217607688348e-05,
+      "loss": 0.76,
+      "step": 5238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.90666389465332,
+      "learning_rate": 1.6513643384245754e-05,
+      "loss": 0.8328,
+      "step": 5239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.265645027160645,
+      "learning_rate": 1.6511069160803158e-05,
+      "loss": 0.5108,
+      "step": 5240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.731019020080566,
+      "learning_rate": 1.6508494937360564e-05,
+      "loss": 0.7937,
+      "step": 5241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.847784042358398,
+      "learning_rate": 1.6505920713917968e-05,
+      "loss": 0.606,
+      "step": 5242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.213117599487305,
+      "learning_rate": 1.6503346490475374e-05,
+      "loss": 0.7917,
+      "step": 5243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.141951560974121,
+      "learning_rate": 1.6500772267032778e-05,
+      "loss": 0.6665,
+      "step": 5244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.041945457458496,
+      "learning_rate": 1.649819804359018e-05,
+      "loss": 0.8224,
+      "step": 5245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.495555877685547,
+      "learning_rate": 1.649562382014759e-05,
+      "loss": 0.4823,
+      "step": 5246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.829453468322754,
+      "learning_rate": 1.6493049596704994e-05,
+      "loss": 0.6583,
+      "step": 5247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.408878326416016,
+      "learning_rate": 1.64904753732624e-05,
+      "loss": 0.492,
+      "step": 5248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.759248733520508,
+      "learning_rate": 1.6487901149819804e-05,
+      "loss": 0.722,
+      "step": 5249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.546459197998047,
+      "learning_rate": 1.648532692637721e-05,
+      "loss": 0.8778,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.193469047546387,
+      "learning_rate": 1.6482752702934614e-05,
+      "loss": 0.8,
+      "step": 5251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.022163391113281,
+      "learning_rate": 1.648017847949202e-05,
+      "loss": 0.4692,
+      "step": 5252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.208653450012207,
+      "learning_rate": 1.6477604256049424e-05,
+      "loss": 0.5389,
+      "step": 5253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.21313762664795,
+      "learning_rate": 1.647503003260683e-05,
+      "loss": 0.7251,
+      "step": 5254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.25330638885498,
+      "learning_rate": 1.6472455809164238e-05,
+      "loss": 0.6814,
+      "step": 5255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.095317840576172,
+      "learning_rate": 1.646988158572164e-05,
+      "loss": 0.7626,
+      "step": 5256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.305283546447754,
+      "learning_rate": 1.6467307362279047e-05,
+      "loss": 0.7507,
+      "step": 5257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.431222915649414,
+      "learning_rate": 1.646473313883645e-05,
+      "loss": 0.5146,
+      "step": 5258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.402660369873047,
+      "learning_rate": 1.6462158915393857e-05,
+      "loss": 0.7854,
+      "step": 5259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.578250885009766,
+      "learning_rate": 1.645958469195126e-05,
+      "loss": 0.8119,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.49059772491455,
+      "learning_rate": 1.6457010468508667e-05,
+      "loss": 0.8049,
+      "step": 5261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.794490814208984,
+      "learning_rate": 1.645443624506607e-05,
+      "loss": 0.7455,
+      "step": 5262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.36746597290039,
+      "learning_rate": 1.6451862021623477e-05,
+      "loss": 0.6569,
+      "step": 5263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.957120418548584,
+      "learning_rate": 1.644928779818088e-05,
+      "loss": 0.5979,
+      "step": 5264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.1671781539917,
+      "learning_rate": 1.6446713574738287e-05,
+      "loss": 0.7981,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.471668243408203,
+      "learning_rate": 1.6444139351295694e-05,
+      "loss": 0.6581,
+      "step": 5266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.272623062133789,
+      "learning_rate": 1.6441565127853097e-05,
+      "loss": 0.9694,
+      "step": 5267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.479127883911133,
+      "learning_rate": 1.6438990904410504e-05,
+      "loss": 0.6734,
+      "step": 5268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.329911231994629,
+      "learning_rate": 1.6436416680967907e-05,
+      "loss": 0.916,
+      "step": 5269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.38661003112793,
+      "learning_rate": 1.6433842457525314e-05,
+      "loss": 0.8645,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.047325134277344,
+      "learning_rate": 1.6431268234082717e-05,
+      "loss": 0.6926,
+      "step": 5271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.207578659057617,
+      "learning_rate": 1.6428694010640124e-05,
+      "loss": 0.6558,
+      "step": 5272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.521034240722656,
+      "learning_rate": 1.6426119787197527e-05,
+      "loss": 0.6857,
+      "step": 5273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.073044776916504,
+      "learning_rate": 1.6423545563754937e-05,
+      "loss": 0.5497,
+      "step": 5274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.431992530822754,
+      "learning_rate": 1.642097134031234e-05,
+      "loss": 0.629,
+      "step": 5275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.577980041503906,
+      "learning_rate": 1.6418397116869744e-05,
+      "loss": 0.5072,
+      "step": 5276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.020289421081543,
+      "learning_rate": 1.641582289342715e-05,
+      "loss": 0.8599,
+      "step": 5277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.187503814697266,
+      "learning_rate": 1.6413248669984554e-05,
+      "loss": 0.7593,
+      "step": 5278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.601643562316895,
+      "learning_rate": 1.641067444654196e-05,
+      "loss": 0.7217,
+      "step": 5279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.692886352539062,
+      "learning_rate": 1.6408100223099364e-05,
+      "loss": 0.7866,
+      "step": 5280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.177303314208984,
+      "learning_rate": 1.640552599965677e-05,
+      "loss": 0.5475,
+      "step": 5281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.756418228149414,
+      "learning_rate": 1.6402951776214174e-05,
+      "loss": 0.5752,
+      "step": 5282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.627341270446777,
+      "learning_rate": 1.640037755277158e-05,
+      "loss": 0.9314,
+      "step": 5283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.185128211975098,
+      "learning_rate": 1.6397803329328987e-05,
+      "loss": 0.7916,
+      "step": 5284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.435806274414062,
+      "learning_rate": 1.6395229105886394e-05,
+      "loss": 0.8947,
+      "step": 5285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.506839752197266,
+      "learning_rate": 1.6392654882443797e-05,
+      "loss": 0.6167,
+      "step": 5286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.203943252563477,
+      "learning_rate": 1.63900806590012e-05,
+      "loss": 0.7711,
+      "step": 5287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.451401710510254,
+      "learning_rate": 1.6387506435558607e-05,
+      "loss": 0.6407,
+      "step": 5288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.373774528503418,
+      "learning_rate": 1.638493221211601e-05,
+      "loss": 0.8961,
+      "step": 5289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.490797996520996,
+      "learning_rate": 1.6382357988673417e-05,
+      "loss": 0.7242,
+      "step": 5290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.302254676818848,
+      "learning_rate": 1.637978376523082e-05,
+      "loss": 0.6484,
+      "step": 5291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.7662935256958,
+      "learning_rate": 1.6377209541788227e-05,
+      "loss": 0.8167,
+      "step": 5292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.014126777648926,
+      "learning_rate": 1.6374635318345634e-05,
+      "loss": 0.6313,
+      "step": 5293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.957836151123047,
+      "learning_rate": 1.637206109490304e-05,
+      "loss": 0.4871,
+      "step": 5294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.586565971374512,
+      "learning_rate": 1.6369486871460444e-05,
+      "loss": 0.645,
+      "step": 5295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.761870384216309,
+      "learning_rate": 1.636691264801785e-05,
+      "loss": 1.0254,
+      "step": 5296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.778366088867188,
+      "learning_rate": 1.6364338424575254e-05,
+      "loss": 0.5141,
+      "step": 5297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.954097747802734,
+      "learning_rate": 1.6361764201132657e-05,
+      "loss": 0.6822,
+      "step": 5298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.593156814575195,
+      "learning_rate": 1.6359189977690064e-05,
+      "loss": 0.7444,
+      "step": 5299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.708962440490723,
+      "learning_rate": 1.6356615754247467e-05,
+      "loss": 0.7988,
+      "step": 5300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.688947677612305,
+      "learning_rate": 1.6354041530804874e-05,
+      "loss": 0.7307,
+      "step": 5301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.107463836669922,
+      "learning_rate": 1.6351467307362277e-05,
+      "loss": 0.7361,
+      "step": 5302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.472657203674316,
+      "learning_rate": 1.6348893083919687e-05,
+      "loss": 0.7446,
+      "step": 5303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.349526405334473,
+      "learning_rate": 1.634631886047709e-05,
+      "loss": 0.9223,
+      "step": 5304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.807222366333008,
+      "learning_rate": 1.6343744637034497e-05,
+      "loss": 0.9837,
+      "step": 5305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.888278007507324,
+      "learning_rate": 1.63411704135919e-05,
+      "loss": 0.6046,
+      "step": 5306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.625167846679688,
+      "learning_rate": 1.6338596190149304e-05,
+      "loss": 0.7998,
+      "step": 5307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.900516510009766,
+      "learning_rate": 1.633602196670671e-05,
+      "loss": 0.6392,
+      "step": 5308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.312675476074219,
+      "learning_rate": 1.6333447743264114e-05,
+      "loss": 0.4499,
+      "step": 5309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.923124313354492,
+      "learning_rate": 1.633087351982152e-05,
+      "loss": 0.5753,
+      "step": 5310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.0786771774292,
+      "learning_rate": 1.6328299296378924e-05,
+      "loss": 0.7728,
+      "step": 5311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.278932571411133,
+      "learning_rate": 1.6325725072936334e-05,
+      "loss": 0.8013,
+      "step": 5312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.094017028808594,
+      "learning_rate": 1.6323150849493737e-05,
+      "loss": 0.6387,
+      "step": 5313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.221474647521973,
+      "learning_rate": 1.6320576626051144e-05,
+      "loss": 0.582,
+      "step": 5314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.438040733337402,
+      "learning_rate": 1.6318002402608547e-05,
+      "loss": 0.5847,
+      "step": 5315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.305757522583008,
+      "learning_rate": 1.6315428179165954e-05,
+      "loss": 0.5777,
+      "step": 5316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.588105201721191,
+      "learning_rate": 1.6312853955723357e-05,
+      "loss": 0.8716,
+      "step": 5317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.558932304382324,
+      "learning_rate": 1.631027973228076e-05,
+      "loss": 0.9259,
+      "step": 5318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.885242462158203,
+      "learning_rate": 1.6307705508838167e-05,
+      "loss": 0.7004,
+      "step": 5319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.342612266540527,
+      "learning_rate": 1.630513128539557e-05,
+      "loss": 0.6183,
+      "step": 5320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.590620994567871,
+      "learning_rate": 1.630255706195298e-05,
+      "loss": 0.5081,
+      "step": 5321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.505029678344727,
+      "learning_rate": 1.6299982838510384e-05,
+      "loss": 0.609,
+      "step": 5322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.353503227233887,
+      "learning_rate": 1.629740861506779e-05,
+      "loss": 0.7164,
+      "step": 5323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.86381721496582,
+      "learning_rate": 1.6294834391625193e-05,
+      "loss": 0.7149,
+      "step": 5324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.773503303527832,
+      "learning_rate": 1.62922601681826e-05,
+      "loss": 0.8802,
+      "step": 5325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.397554397583008,
+      "learning_rate": 1.6289685944740003e-05,
+      "loss": 1.0327,
+      "step": 5326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.209512710571289,
+      "learning_rate": 1.628711172129741e-05,
+      "loss": 1.0638,
+      "step": 5327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.346537590026855,
+      "learning_rate": 1.6284537497854813e-05,
+      "loss": 0.7663,
+      "step": 5328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.009007453918457,
+      "learning_rate": 1.6281963274412217e-05,
+      "loss": 0.6647,
+      "step": 5329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.355412483215332,
+      "learning_rate": 1.6279389050969623e-05,
+      "loss": 0.8087,
+      "step": 5330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.055541038513184,
+      "learning_rate": 1.627681482752703e-05,
+      "loss": 0.5819,
+      "step": 5331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.069061279296875,
+      "learning_rate": 1.6274240604084437e-05,
+      "loss": 0.6965,
+      "step": 5332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.634626388549805,
+      "learning_rate": 1.627166638064184e-05,
+      "loss": 0.7589,
+      "step": 5333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.606729507446289,
+      "learning_rate": 1.6269092157199247e-05,
+      "loss": 0.7905,
+      "step": 5334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.825538635253906,
+      "learning_rate": 1.626651793375665e-05,
+      "loss": 0.6053,
+      "step": 5335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.846492767333984,
+      "learning_rate": 1.6263943710314057e-05,
+      "loss": 0.6376,
+      "step": 5336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.11724853515625,
+      "learning_rate": 1.626136948687146e-05,
+      "loss": 0.9662,
+      "step": 5337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.473100662231445,
+      "learning_rate": 1.6258795263428867e-05,
+      "loss": 0.611,
+      "step": 5338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.624481201171875,
+      "learning_rate": 1.625622103998627e-05,
+      "loss": 0.6607,
+      "step": 5339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.945741653442383,
+      "learning_rate": 1.6253646816543677e-05,
+      "loss": 0.6932,
+      "step": 5340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.372787475585938,
+      "learning_rate": 1.6251072593101083e-05,
+      "loss": 0.6036,
+      "step": 5341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.175030708312988,
+      "learning_rate": 1.6248498369658487e-05,
+      "loss": 0.7683,
+      "step": 5342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.389568328857422,
+      "learning_rate": 1.6245924146215893e-05,
+      "loss": 0.7174,
+      "step": 5343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.045255661010742,
+      "learning_rate": 1.6243349922773297e-05,
+      "loss": 0.7243,
+      "step": 5344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.945423126220703,
+      "learning_rate": 1.6240775699330703e-05,
+      "loss": 0.5235,
+      "step": 5345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.613673210144043,
+      "learning_rate": 1.6238201475888107e-05,
+      "loss": 0.5549,
+      "step": 5346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.48994255065918,
+      "learning_rate": 1.6235627252445513e-05,
+      "loss": 0.7129,
+      "step": 5347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.308940410614014,
+      "learning_rate": 1.6233053029002917e-05,
+      "loss": 0.4572,
+      "step": 5348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.74059772491455,
+      "learning_rate": 1.623047880556032e-05,
+      "loss": 0.5891,
+      "step": 5349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.355896949768066,
+      "learning_rate": 1.622790458211773e-05,
+      "loss": 0.6838,
+      "step": 5350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.994861602783203,
+      "learning_rate": 1.6225330358675133e-05,
+      "loss": 0.7444,
+      "step": 5351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.859659194946289,
+      "learning_rate": 1.622275613523254e-05,
+      "loss": 0.3659,
+      "step": 5352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.659871101379395,
+      "learning_rate": 1.6220181911789943e-05,
+      "loss": 0.5816,
+      "step": 5353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.827496528625488,
+      "learning_rate": 1.621760768834735e-05,
+      "loss": 0.5036,
+      "step": 5354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.024316787719727,
+      "learning_rate": 1.6215033464904753e-05,
+      "loss": 0.7096,
+      "step": 5355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.871668815612793,
+      "learning_rate": 1.621245924146216e-05,
+      "loss": 0.6767,
+      "step": 5356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.786467552185059,
+      "learning_rate": 1.6209885018019563e-05,
+      "loss": 0.7224,
+      "step": 5357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.874835968017578,
+      "learning_rate": 1.620731079457697e-05,
+      "loss": 0.8037,
+      "step": 5358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.968464851379395,
+      "learning_rate": 1.6204736571134377e-05,
+      "loss": 0.6626,
+      "step": 5359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.908061981201172,
+      "learning_rate": 1.620216234769178e-05,
+      "loss": 0.8516,
+      "step": 5360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.781600952148438,
+      "learning_rate": 1.6199588124249186e-05,
+      "loss": 0.9234,
+      "step": 5361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.504817962646484,
+      "learning_rate": 1.619701390080659e-05,
+      "loss": 0.7915,
+      "step": 5362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.987812042236328,
+      "learning_rate": 1.6194439677363996e-05,
+      "loss": 0.6629,
+      "step": 5363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.693304061889648,
+      "learning_rate": 1.61918654539214e-05,
+      "loss": 0.7122,
+      "step": 5364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.195155143737793,
+      "learning_rate": 1.6189291230478806e-05,
+      "loss": 0.7479,
+      "step": 5365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.161410331726074,
+      "learning_rate": 1.618671700703621e-05,
+      "loss": 0.5539,
+      "step": 5366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.984532356262207,
+      "learning_rate": 1.6184142783593616e-05,
+      "loss": 0.7711,
+      "step": 5367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.00101375579834,
+      "learning_rate": 1.618156856015102e-05,
+      "loss": 0.819,
+      "step": 5368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.316864967346191,
+      "learning_rate": 1.617899433670843e-05,
+      "loss": 0.7121,
+      "step": 5369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.20595932006836,
+      "learning_rate": 1.6176420113265833e-05,
+      "loss": 0.5368,
+      "step": 5370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.708369255065918,
+      "learning_rate": 1.6173845889823236e-05,
+      "loss": 1.2529,
+      "step": 5371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.034847259521484,
+      "learning_rate": 1.6171271666380643e-05,
+      "loss": 0.7555,
+      "step": 5372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.021966934204102,
+      "learning_rate": 1.6168697442938046e-05,
+      "loss": 0.604,
+      "step": 5373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.402461051940918,
+      "learning_rate": 1.6166123219495453e-05,
+      "loss": 0.6444,
+      "step": 5374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.433053016662598,
+      "learning_rate": 1.6163548996052856e-05,
+      "loss": 0.8579,
+      "step": 5375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.826925277709961,
+      "learning_rate": 1.6160974772610263e-05,
+      "loss": 0.6832,
+      "step": 5376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.661369323730469,
+      "learning_rate": 1.6158400549167666e-05,
+      "loss": 0.7607,
+      "step": 5377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.432463645935059,
+      "learning_rate": 1.6155826325725076e-05,
+      "loss": 0.6476,
+      "step": 5378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.132686614990234,
+      "learning_rate": 1.615325210228248e-05,
+      "loss": 0.624,
+      "step": 5379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.702945709228516,
+      "learning_rate": 1.6150677878839883e-05,
+      "loss": 0.811,
+      "step": 5380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.254922866821289,
+      "learning_rate": 1.614810365539729e-05,
+      "loss": 0.9329,
+      "step": 5381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.407812118530273,
+      "learning_rate": 1.6145529431954693e-05,
+      "loss": 0.592,
+      "step": 5382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.516083717346191,
+      "learning_rate": 1.61429552085121e-05,
+      "loss": 0.9406,
+      "step": 5383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.30846881866455,
+      "learning_rate": 1.6140380985069503e-05,
+      "loss": 0.5874,
+      "step": 5384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.725110054016113,
+      "learning_rate": 1.613780676162691e-05,
+      "loss": 0.5758,
+      "step": 5385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.639054298400879,
+      "learning_rate": 1.6135232538184313e-05,
+      "loss": 0.9237,
+      "step": 5386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.927342414855957,
+      "learning_rate": 1.613265831474172e-05,
+      "loss": 0.8197,
+      "step": 5387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.463661193847656,
+      "learning_rate": 1.6130084091299126e-05,
+      "loss": 0.5633,
+      "step": 5388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.141190528869629,
+      "learning_rate": 1.6127509867856533e-05,
+      "loss": 0.7412,
+      "step": 5389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.962833404541016,
+      "learning_rate": 1.6124935644413936e-05,
+      "loss": 0.7952,
+      "step": 5390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.48841381072998,
+      "learning_rate": 1.612236142097134e-05,
+      "loss": 0.787,
+      "step": 5391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.667954444885254,
+      "learning_rate": 1.6119787197528746e-05,
+      "loss": 0.6521,
+      "step": 5392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.683670997619629,
+      "learning_rate": 1.611721297408615e-05,
+      "loss": 0.8079,
+      "step": 5393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.248801231384277,
+      "learning_rate": 1.6114638750643556e-05,
+      "loss": 0.6947,
+      "step": 5394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.234082221984863,
+      "learning_rate": 1.611206452720096e-05,
+      "loss": 0.5561,
+      "step": 5395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.410922050476074,
+      "learning_rate": 1.6109490303758366e-05,
+      "loss": 0.8626,
+      "step": 5396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.7655229568481445,
+      "learning_rate": 1.6106916080315773e-05,
+      "loss": 0.7134,
+      "step": 5397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.13786792755127,
+      "learning_rate": 1.610434185687318e-05,
+      "loss": 0.6613,
+      "step": 5398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.836605072021484,
+      "learning_rate": 1.6101767633430583e-05,
+      "loss": 0.7106,
+      "step": 5399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.276037216186523,
+      "learning_rate": 1.609919340998799e-05,
+      "loss": 0.7859,
+      "step": 5400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.147769927978516,
+      "learning_rate": 1.6096619186545393e-05,
+      "loss": 0.6519,
+      "step": 5401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.965616226196289,
+      "learning_rate": 1.6094044963102796e-05,
+      "loss": 0.9663,
+      "step": 5402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.371088027954102,
+      "learning_rate": 1.6091470739660203e-05,
+      "loss": 0.7387,
+      "step": 5403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.353103637695312,
+      "learning_rate": 1.6088896516217606e-05,
+      "loss": 0.9717,
+      "step": 5404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.7879056930542,
+      "learning_rate": 1.6086322292775013e-05,
+      "loss": 0.4814,
+      "step": 5405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.066062927246094,
+      "learning_rate": 1.6083748069332416e-05,
+      "loss": 0.6042,
+      "step": 5406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.037806510925293,
+      "learning_rate": 1.6081173845889826e-05,
+      "loss": 0.5713,
+      "step": 5407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.090384483337402,
+      "learning_rate": 1.607859962244723e-05,
+      "loss": 0.6953,
+      "step": 5408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.951784133911133,
+      "learning_rate": 1.6076025399004636e-05,
+      "loss": 0.6865,
+      "step": 5409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.99825382232666,
+      "learning_rate": 1.607345117556204e-05,
+      "loss": 0.8309,
+      "step": 5410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.917034149169922,
+      "learning_rate": 1.6070876952119443e-05,
+      "loss": 0.7286,
+      "step": 5411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.623841285705566,
+      "learning_rate": 1.606830272867685e-05,
+      "loss": 0.8353,
+      "step": 5412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.034842491149902,
+      "learning_rate": 1.6065728505234253e-05,
+      "loss": 0.7367,
+      "step": 5413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.12311840057373,
+      "learning_rate": 1.606315428179166e-05,
+      "loss": 0.7642,
+      "step": 5414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.247233390808105,
+      "learning_rate": 1.6060580058349063e-05,
+      "loss": 0.858,
+      "step": 5415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.87278938293457,
+      "learning_rate": 1.6058005834906473e-05,
+      "loss": 0.7625,
+      "step": 5416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.773738861083984,
+      "learning_rate": 1.6055431611463876e-05,
+      "loss": 0.6346,
+      "step": 5417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.190245628356934,
+      "learning_rate": 1.6052857388021283e-05,
+      "loss": 0.7055,
+      "step": 5418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.070527076721191,
+      "learning_rate": 1.6050283164578686e-05,
+      "loss": 0.6107,
+      "step": 5419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.259936332702637,
+      "learning_rate": 1.6047708941136093e-05,
+      "loss": 0.6917,
+      "step": 5420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.312941551208496,
+      "learning_rate": 1.6045134717693496e-05,
+      "loss": 0.7604,
+      "step": 5421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.834226608276367,
+      "learning_rate": 1.60425604942509e-05,
+      "loss": 0.7127,
+      "step": 5422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.275663375854492,
+      "learning_rate": 1.6039986270808306e-05,
+      "loss": 0.7497,
+      "step": 5423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.038375854492188,
+      "learning_rate": 1.603741204736571e-05,
+      "loss": 0.7263,
+      "step": 5424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.30618953704834,
+      "learning_rate": 1.603483782392312e-05,
+      "loss": 0.6782,
+      "step": 5425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.826001167297363,
+      "learning_rate": 1.6032263600480523e-05,
+      "loss": 0.6668,
+      "step": 5426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.249032974243164,
+      "learning_rate": 1.602968937703793e-05,
+      "loss": 0.6276,
+      "step": 5427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.06242847442627,
+      "learning_rate": 1.6027115153595333e-05,
+      "loss": 0.6218,
+      "step": 5428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.804352760314941,
+      "learning_rate": 1.602454093015274e-05,
+      "loss": 0.7112,
+      "step": 5429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.694586753845215,
+      "learning_rate": 1.6021966706710142e-05,
+      "loss": 0.9347,
+      "step": 5430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.349326133728027,
+      "learning_rate": 1.601939248326755e-05,
+      "loss": 0.689,
+      "step": 5431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.02199649810791,
+      "learning_rate": 1.6016818259824952e-05,
+      "loss": 0.4271,
+      "step": 5432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.841898918151855,
+      "learning_rate": 1.6014244036382356e-05,
+      "loss": 0.9099,
+      "step": 5433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.227035522460938,
+      "learning_rate": 1.6011669812939762e-05,
+      "loss": 0.6111,
+      "step": 5434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.281566619873047,
+      "learning_rate": 1.600909558949717e-05,
+      "loss": 0.8266,
+      "step": 5435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.28485107421875,
+      "learning_rate": 1.6006521366054576e-05,
+      "loss": 0.701,
+      "step": 5436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.885891914367676,
+      "learning_rate": 1.600394714261198e-05,
+      "loss": 0.6103,
+      "step": 5437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.44804573059082,
+      "learning_rate": 1.6001372919169386e-05,
+      "loss": 0.9201,
+      "step": 5438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.522759437561035,
+      "learning_rate": 1.599879869572679e-05,
+      "loss": 0.8267,
+      "step": 5439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.83130931854248,
+      "learning_rate": 1.5996224472284196e-05,
+      "loss": 0.5862,
+      "step": 5440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.161563873291016,
+      "learning_rate": 1.59936502488416e-05,
+      "loss": 0.8455,
+      "step": 5441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.917141914367676,
+      "learning_rate": 1.5991076025399006e-05,
+      "loss": 0.8669,
+      "step": 5442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.429743766784668,
+      "learning_rate": 1.598850180195641e-05,
+      "loss": 0.7464,
+      "step": 5443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.11600399017334,
+      "learning_rate": 1.5985927578513816e-05,
+      "loss": 0.6064,
+      "step": 5444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.265796661376953,
+      "learning_rate": 1.5983353355071222e-05,
+      "loss": 0.4947,
+      "step": 5445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.254405975341797,
+      "learning_rate": 1.5980779131628626e-05,
+      "loss": 0.6366,
+      "step": 5446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.471699714660645,
+      "learning_rate": 1.5978204908186032e-05,
+      "loss": 0.7729,
+      "step": 5447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.652485847473145,
+      "learning_rate": 1.5975630684743436e-05,
+      "loss": 0.6428,
+      "step": 5448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.507447242736816,
+      "learning_rate": 1.5973056461300842e-05,
+      "loss": 0.7779,
+      "step": 5449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.162154197692871,
+      "learning_rate": 1.5970482237858246e-05,
+      "loss": 0.7809,
+      "step": 5450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.86877727508545,
+      "learning_rate": 1.5967908014415652e-05,
+      "loss": 0.5767,
+      "step": 5451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.88869571685791,
+      "learning_rate": 1.5965333790973056e-05,
+      "loss": 0.601,
+      "step": 5452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.477707862854004,
+      "learning_rate": 1.596275956753046e-05,
+      "loss": 0.4985,
+      "step": 5453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.238130569458008,
+      "learning_rate": 1.596018534408787e-05,
+      "loss": 0.566,
+      "step": 5454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.624138355255127,
+      "learning_rate": 1.5957611120645272e-05,
+      "loss": 0.4094,
+      "step": 5455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.222182273864746,
+      "learning_rate": 1.595503689720268e-05,
+      "loss": 0.6509,
+      "step": 5456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.988697052001953,
+      "learning_rate": 1.5952462673760082e-05,
+      "loss": 1.0603,
+      "step": 5457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.02749252319336,
+      "learning_rate": 1.594988845031749e-05,
+      "loss": 0.6549,
+      "step": 5458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.617202758789062,
+      "learning_rate": 1.5947314226874892e-05,
+      "loss": 0.7358,
+      "step": 5459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.516440391540527,
+      "learning_rate": 1.59447400034323e-05,
+      "loss": 0.9274,
+      "step": 5460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.804156303405762,
+      "learning_rate": 1.5942165779989702e-05,
+      "loss": 0.7639,
+      "step": 5461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.205520629882812,
+      "learning_rate": 1.593959155654711e-05,
+      "loss": 0.7911,
+      "step": 5462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.2595796585083,
+      "learning_rate": 1.5937017333104516e-05,
+      "loss": 0.608,
+      "step": 5463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.216423034667969,
+      "learning_rate": 1.593444310966192e-05,
+      "loss": 0.922,
+      "step": 5464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.078286170959473,
+      "learning_rate": 1.5931868886219326e-05,
+      "loss": 0.618,
+      "step": 5465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.788003921508789,
+      "learning_rate": 1.592929466277673e-05,
+      "loss": 1.0134,
+      "step": 5466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.516185760498047,
+      "learning_rate": 1.5926720439334135e-05,
+      "loss": 0.6751,
+      "step": 5467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.466375350952148,
+      "learning_rate": 1.592414621589154e-05,
+      "loss": 0.5006,
+      "step": 5468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.340472221374512,
+      "learning_rate": 1.5921571992448945e-05,
+      "loss": 0.8121,
+      "step": 5469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.013021469116211,
+      "learning_rate": 1.591899776900635e-05,
+      "loss": 0.8057,
+      "step": 5470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.822932243347168,
+      "learning_rate": 1.5916423545563755e-05,
+      "loss": 0.521,
+      "step": 5471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.534416198730469,
+      "learning_rate": 1.591384932212116e-05,
+      "loss": 0.7171,
+      "step": 5472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.124041557312012,
+      "learning_rate": 1.591127509867857e-05,
+      "loss": 0.5739,
+      "step": 5473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.582082748413086,
+      "learning_rate": 1.5908700875235972e-05,
+      "loss": 0.9217,
+      "step": 5474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.603957176208496,
+      "learning_rate": 1.5906126651793375e-05,
+      "loss": 0.6886,
+      "step": 5475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.811408042907715,
+      "learning_rate": 1.5903552428350782e-05,
+      "loss": 0.7656,
+      "step": 5476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.944864273071289,
+      "learning_rate": 1.5900978204908185e-05,
+      "loss": 0.7287,
+      "step": 5477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.971102714538574,
+      "learning_rate": 1.5898403981465592e-05,
+      "loss": 0.7978,
+      "step": 5478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.381170272827148,
+      "learning_rate": 1.5895829758022995e-05,
+      "loss": 0.8835,
+      "step": 5479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.820755004882812,
+      "learning_rate": 1.5893255534580402e-05,
+      "loss": 0.6181,
+      "step": 5480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.809287071228027,
+      "learning_rate": 1.5890681311137805e-05,
+      "loss": 0.6795,
+      "step": 5481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.557952880859375,
+      "learning_rate": 1.5888107087695215e-05,
+      "loss": 0.6398,
+      "step": 5482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.693344116210938,
+      "learning_rate": 1.588553286425262e-05,
+      "loss": 0.8036,
+      "step": 5483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.018784523010254,
+      "learning_rate": 1.5882958640810022e-05,
+      "loss": 0.6238,
+      "step": 5484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.494216918945312,
+      "learning_rate": 1.588038441736743e-05,
+      "loss": 0.7607,
+      "step": 5485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.84225082397461,
+      "learning_rate": 1.5877810193924832e-05,
+      "loss": 0.7017,
+      "step": 5486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.3541259765625,
+      "learning_rate": 1.587523597048224e-05,
+      "loss": 0.8281,
+      "step": 5487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.750877857208252,
+      "learning_rate": 1.5872661747039642e-05,
+      "loss": 0.5429,
+      "step": 5488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.787055015563965,
+      "learning_rate": 1.587008752359705e-05,
+      "loss": 0.7507,
+      "step": 5489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.039846420288086,
+      "learning_rate": 1.5867513300154452e-05,
+      "loss": 0.5399,
+      "step": 5490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.538870811462402,
+      "learning_rate": 1.586493907671186e-05,
+      "loss": 0.6185,
+      "step": 5491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.526559829711914,
+      "learning_rate": 1.5862364853269265e-05,
+      "loss": 0.7405,
+      "step": 5492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.64937686920166,
+      "learning_rate": 1.5859790629826672e-05,
+      "loss": 1.0004,
+      "step": 5493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.196798324584961,
+      "learning_rate": 1.5857216406384075e-05,
+      "loss": 0.7292,
+      "step": 5494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.425310134887695,
+      "learning_rate": 1.585464218294148e-05,
+      "loss": 0.6555,
+      "step": 5495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.53381633758545,
+      "learning_rate": 1.5852067959498885e-05,
+      "loss": 0.8289,
+      "step": 5496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.954731941223145,
+      "learning_rate": 1.584949373605629e-05,
+      "loss": 0.709,
+      "step": 5497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.777397155761719,
+      "learning_rate": 1.5846919512613695e-05,
+      "loss": 0.8233,
+      "step": 5498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.766757011413574,
+      "learning_rate": 1.58443452891711e-05,
+      "loss": 0.8391,
+      "step": 5499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.392250061035156,
+      "learning_rate": 1.5841771065728505e-05,
+      "loss": 0.8215,
+      "step": 5500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.856593608856201,
+      "learning_rate": 1.5839196842285912e-05,
+      "loss": 0.6673,
+      "step": 5501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.711043357849121,
+      "learning_rate": 1.583662261884332e-05,
+      "loss": 0.8859,
+      "step": 5502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.095938682556152,
+      "learning_rate": 1.5834048395400722e-05,
+      "loss": 0.7732,
+      "step": 5503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.732726097106934,
+      "learning_rate": 1.583147417195813e-05,
+      "loss": 0.841,
+      "step": 5504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.311925888061523,
+      "learning_rate": 1.5828899948515532e-05,
+      "loss": 0.893,
+      "step": 5505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.491415023803711,
+      "learning_rate": 1.5826325725072935e-05,
+      "loss": 0.7547,
+      "step": 5506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.706657409667969,
+      "learning_rate": 1.5823751501630342e-05,
+      "loss": 0.8287,
+      "step": 5507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.36621379852295,
+      "learning_rate": 1.5821177278187745e-05,
+      "loss": 0.6054,
+      "step": 5508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.965426445007324,
+      "learning_rate": 1.5818603054745152e-05,
+      "loss": 0.737,
+      "step": 5509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.062065124511719,
+      "learning_rate": 1.5816028831302555e-05,
+      "loss": 0.9597,
+      "step": 5510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.279169082641602,
+      "learning_rate": 1.5813454607859965e-05,
+      "loss": 0.8107,
+      "step": 5511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.843371391296387,
+      "learning_rate": 1.581088038441737e-05,
+      "loss": 0.4532,
+      "step": 5512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.130149841308594,
+      "learning_rate": 1.5808306160974775e-05,
+      "loss": 0.6675,
+      "step": 5513
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.880720138549805,
+      "learning_rate": 1.580573193753218e-05,
+      "loss": 0.5462,
+      "step": 5514
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.71376895904541,
+      "learning_rate": 1.580315771408958e-05,
+      "loss": 0.6443,
+      "step": 5515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.150823593139648,
+      "learning_rate": 1.580058349064699e-05,
+      "loss": 0.8697,
+      "step": 5516
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.813486099243164,
+      "learning_rate": 1.579800926720439e-05,
+      "loss": 0.7977,
+      "step": 5517
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.239295959472656,
+      "learning_rate": 1.57954350437618e-05,
+      "loss": 0.6793,
+      "step": 5518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.66236686706543,
+      "learning_rate": 1.57928608203192e-05,
+      "loss": 0.7512,
+      "step": 5519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.604142189025879,
+      "learning_rate": 1.579028659687661e-05,
+      "loss": 0.6044,
+      "step": 5520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.841404914855957,
+      "learning_rate": 1.5787712373434015e-05,
+      "loss": 0.579,
+      "step": 5521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.501837730407715,
+      "learning_rate": 1.578513814999142e-05,
+      "loss": 0.7205,
+      "step": 5522
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.540206909179688,
+      "learning_rate": 1.5782563926548825e-05,
+      "loss": 0.7782,
+      "step": 5523
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.388627052307129,
+      "learning_rate": 1.577998970310623e-05,
+      "loss": 0.7765,
+      "step": 5524
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.31390380859375,
+      "learning_rate": 1.5777415479663635e-05,
+      "loss": 0.5311,
+      "step": 5525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.276957511901855,
+      "learning_rate": 1.5774841256221038e-05,
+      "loss": 0.9633,
+      "step": 5526
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.880576133728027,
+      "learning_rate": 1.5772267032778445e-05,
+      "loss": 0.5084,
+      "step": 5527
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.902566909790039,
+      "learning_rate": 1.5769692809335848e-05,
+      "loss": 0.4669,
+      "step": 5528
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.811795234680176,
+      "learning_rate": 1.5767118585893255e-05,
+      "loss": 0.6889,
+      "step": 5529
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.534520149230957,
+      "learning_rate": 1.576454436245066e-05,
+      "loss": 1.0246,
+      "step": 5530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.52299690246582,
+      "learning_rate": 1.5761970139008068e-05,
+      "loss": 0.8349,
+      "step": 5531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.835138320922852,
+      "learning_rate": 1.575939591556547e-05,
+      "loss": 0.8441,
+      "step": 5532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.813920021057129,
+      "learning_rate": 1.5756821692122878e-05,
+      "loss": 0.7039,
+      "step": 5533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.295403480529785,
+      "learning_rate": 1.575424746868028e-05,
+      "loss": 0.8619,
+      "step": 5534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.677909851074219,
+      "learning_rate": 1.5751673245237688e-05,
+      "loss": 0.6487,
+      "step": 5535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.992595672607422,
+      "learning_rate": 1.574909902179509e-05,
+      "loss": 0.7828,
+      "step": 5536
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.744140625,
+      "learning_rate": 1.5746524798352495e-05,
+      "loss": 0.6547,
+      "step": 5537
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.94292163848877,
+      "learning_rate": 1.57439505749099e-05,
+      "loss": 0.8616,
+      "step": 5538
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.255457878112793,
+      "learning_rate": 1.5741376351467308e-05,
+      "loss": 0.7145,
+      "step": 5539
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.098308563232422,
+      "learning_rate": 1.5738802128024715e-05,
+      "loss": 0.8017,
+      "step": 5540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.505480766296387,
+      "learning_rate": 1.5736227904582118e-05,
+      "loss": 0.9495,
+      "step": 5541
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.07521915435791,
+      "learning_rate": 1.5733653681139525e-05,
+      "loss": 0.8291,
+      "step": 5542
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.103720664978027,
+      "learning_rate": 1.5731079457696928e-05,
+      "loss": 0.9458,
+      "step": 5543
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.460010528564453,
+      "learning_rate": 1.5728505234254335e-05,
+      "loss": 0.5839,
+      "step": 5544
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.207123756408691,
+      "learning_rate": 1.5725931010811738e-05,
+      "loss": 0.7652,
+      "step": 5545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.013753890991211,
+      "learning_rate": 1.5723356787369145e-05,
+      "loss": 0.5342,
+      "step": 5546
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.64785385131836,
+      "learning_rate": 1.5720782563926548e-05,
+      "loss": 0.8422,
+      "step": 5547
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.991358757019043,
+      "learning_rate": 1.5718208340483955e-05,
+      "loss": 0.8929,
+      "step": 5548
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.931777954101562,
+      "learning_rate": 1.571563411704136e-05,
+      "loss": 0.6247,
+      "step": 5549
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.044845581054688,
+      "learning_rate": 1.5713059893598765e-05,
+      "loss": 0.685,
+      "step": 5550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.44978666305542,
+      "learning_rate": 1.571048567015617e-05,
+      "loss": 0.4404,
+      "step": 5551
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.396489143371582,
+      "learning_rate": 1.5707911446713575e-05,
+      "loss": 0.6322,
+      "step": 5552
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.507052421569824,
+      "learning_rate": 1.570533722327098e-05,
+      "loss": 0.6494,
+      "step": 5553
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.790435791015625,
+      "learning_rate": 1.5702762999828385e-05,
+      "loss": 0.5111,
+      "step": 5554
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.560773849487305,
+      "learning_rate": 1.570018877638579e-05,
+      "loss": 0.5038,
+      "step": 5555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.470216751098633,
+      "learning_rate": 1.5697614552943195e-05,
+      "loss": 0.7754,
+      "step": 5556
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.905423164367676,
+      "learning_rate": 1.5695040329500598e-05,
+      "loss": 0.7474,
+      "step": 5557
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.42026138305664,
+      "learning_rate": 1.5692466106058008e-05,
+      "loss": 0.936,
+      "step": 5558
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.158493041992188,
+      "learning_rate": 1.568989188261541e-05,
+      "loss": 0.556,
+      "step": 5559
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.384407997131348,
+      "learning_rate": 1.5687317659172818e-05,
+      "loss": 0.6649,
+      "step": 5560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.067281723022461,
+      "learning_rate": 1.568474343573022e-05,
+      "loss": 0.7918,
+      "step": 5561
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.902259826660156,
+      "learning_rate": 1.5682169212287628e-05,
+      "loss": 0.8508,
+      "step": 5562
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.968539237976074,
+      "learning_rate": 1.567959498884503e-05,
+      "loss": 0.6452,
+      "step": 5563
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.663578033447266,
+      "learning_rate": 1.5677020765402438e-05,
+      "loss": 0.6886,
+      "step": 5564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.051219940185547,
+      "learning_rate": 1.567444654195984e-05,
+      "loss": 0.9093,
+      "step": 5565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.78207015991211,
+      "learning_rate": 1.5671872318517248e-05,
+      "loss": 0.6888,
+      "step": 5566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.442506790161133,
+      "learning_rate": 1.5669298095074655e-05,
+      "loss": 0.7745,
+      "step": 5567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.913450241088867,
+      "learning_rate": 1.5666723871632058e-05,
+      "loss": 0.6244,
+      "step": 5568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.408548355102539,
+      "learning_rate": 1.5664149648189465e-05,
+      "loss": 0.875,
+      "step": 5569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.3099365234375,
+      "learning_rate": 1.5661575424746868e-05,
+      "loss": 0.7039,
+      "step": 5570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.898088455200195,
+      "learning_rate": 1.5659001201304274e-05,
+      "loss": 0.5219,
+      "step": 5571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.329687118530273,
+      "learning_rate": 1.5656426977861678e-05,
+      "loss": 0.5974,
+      "step": 5572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.845309257507324,
+      "learning_rate": 1.5653852754419084e-05,
+      "loss": 0.6365,
+      "step": 5573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.738419532775879,
+      "learning_rate": 1.5651278530976488e-05,
+      "loss": 0.6588,
+      "step": 5574
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.223743438720703,
+      "learning_rate": 1.5648704307533894e-05,
+      "loss": 0.608,
+      "step": 5575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.70917797088623,
+      "learning_rate": 1.5646130084091298e-05,
+      "loss": 0.6659,
+      "step": 5576
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.924893379211426,
+      "learning_rate": 1.5643555860648708e-05,
+      "loss": 0.879,
+      "step": 5577
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.618356704711914,
+      "learning_rate": 1.564098163720611e-05,
+      "loss": 0.6395,
+      "step": 5578
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.092124938964844,
+      "learning_rate": 1.5638407413763514e-05,
+      "loss": 0.4761,
+      "step": 5579
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.847647666931152,
+      "learning_rate": 1.563583319032092e-05,
+      "loss": 0.6705,
+      "step": 5580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.82312297821045,
+      "learning_rate": 1.5633258966878324e-05,
+      "loss": 0.998,
+      "step": 5581
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.251140594482422,
+      "learning_rate": 1.563068474343573e-05,
+      "loss": 0.6906,
+      "step": 5582
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.709558486938477,
+      "learning_rate": 1.5628110519993134e-05,
+      "loss": 0.822,
+      "step": 5583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.888821601867676,
+      "learning_rate": 1.562553629655054e-05,
+      "loss": 0.7191,
+      "step": 5584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.68293571472168,
+      "learning_rate": 1.5622962073107944e-05,
+      "loss": 0.5517,
+      "step": 5585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.949898719787598,
+      "learning_rate": 1.5620387849665354e-05,
+      "loss": 0.9547,
+      "step": 5586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.29300308227539,
+      "learning_rate": 1.5617813626222758e-05,
+      "loss": 0.739,
+      "step": 5587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.012667179107666,
+      "learning_rate": 1.561523940278016e-05,
+      "loss": 0.5322,
+      "step": 5588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.041412353515625,
+      "learning_rate": 1.5612665179337568e-05,
+      "loss": 0.7658,
+      "step": 5589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.061575889587402,
+      "learning_rate": 1.561009095589497e-05,
+      "loss": 0.7991,
+      "step": 5590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.600547790527344,
+      "learning_rate": 1.5607516732452378e-05,
+      "loss": 0.6509,
+      "step": 5591
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.921608924865723,
+      "learning_rate": 1.560494250900978e-05,
+      "loss": 0.7163,
+      "step": 5592
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.13800048828125,
+      "learning_rate": 1.5602368285567188e-05,
+      "loss": 0.6234,
+      "step": 5593
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.188948631286621,
+      "learning_rate": 1.559979406212459e-05,
+      "loss": 0.7124,
+      "step": 5594
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.671902656555176,
+      "learning_rate": 1.5597219838681998e-05,
+      "loss": 1.0355,
+      "step": 5595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.45449447631836,
+      "learning_rate": 1.5594645615239404e-05,
+      "loss": 0.8463,
+      "step": 5596
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.132134437561035,
+      "learning_rate": 1.559207139179681e-05,
+      "loss": 0.6501,
+      "step": 5597
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.015098571777344,
+      "learning_rate": 1.5589497168354214e-05,
+      "loss": 0.5043,
+      "step": 5598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.141470909118652,
+      "learning_rate": 1.5586922944911618e-05,
+      "loss": 0.3979,
+      "step": 5599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.37172794342041,
+      "learning_rate": 1.5584348721469024e-05,
+      "loss": 0.4986,
+      "step": 5600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.976631164550781,
+      "learning_rate": 1.5581774498026428e-05,
+      "loss": 0.6687,
+      "step": 5601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.840478897094727,
+      "learning_rate": 1.5579200274583834e-05,
+      "loss": 0.7312,
+      "step": 5602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.035346984863281,
+      "learning_rate": 1.5576626051141237e-05,
+      "loss": 0.9072,
+      "step": 5603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.700462341308594,
+      "learning_rate": 1.5574051827698644e-05,
+      "loss": 0.7465,
+      "step": 5604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.324307441711426,
+      "learning_rate": 1.557147760425605e-05,
+      "loss": 0.7753,
+      "step": 5605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.882134437561035,
+      "learning_rate": 1.5568903380813458e-05,
+      "loss": 0.5816,
+      "step": 5606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.551189422607422,
+      "learning_rate": 1.556632915737086e-05,
+      "loss": 0.6115,
+      "step": 5607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.379423141479492,
+      "learning_rate": 1.5563754933928267e-05,
+      "loss": 0.5545,
+      "step": 5608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.434854507446289,
+      "learning_rate": 1.556118071048567e-05,
+      "loss": 0.5058,
+      "step": 5609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.850112915039062,
+      "learning_rate": 1.5558606487043074e-05,
+      "loss": 0.6237,
+      "step": 5610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.739999771118164,
+      "learning_rate": 1.555603226360048e-05,
+      "loss": 0.9191,
+      "step": 5611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.242504119873047,
+      "learning_rate": 1.5553458040157884e-05,
+      "loss": 0.5725,
+      "step": 5612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.402344703674316,
+      "learning_rate": 1.555088381671529e-05,
+      "loss": 0.8992,
+      "step": 5613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.253782272338867,
+      "learning_rate": 1.5548309593272694e-05,
+      "loss": 0.4061,
+      "step": 5614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.940531730651855,
+      "learning_rate": 1.5545735369830104e-05,
+      "loss": 0.6459,
+      "step": 5615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.35522174835205,
+      "learning_rate": 1.5543161146387507e-05,
+      "loss": 0.7855,
+      "step": 5616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.31906509399414,
+      "learning_rate": 1.5540586922944914e-05,
+      "loss": 0.6374,
+      "step": 5617
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.768364906311035,
+      "learning_rate": 1.5538012699502317e-05,
+      "loss": 0.8213,
+      "step": 5618
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.215407371520996,
+      "learning_rate": 1.553543847605972e-05,
+      "loss": 0.7401,
+      "step": 5619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.41373348236084,
+      "learning_rate": 1.5532864252617127e-05,
+      "loss": 0.7124,
+      "step": 5620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.58165168762207,
+      "learning_rate": 1.553029002917453e-05,
+      "loss": 0.5673,
+      "step": 5621
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.350859642028809,
+      "learning_rate": 1.5527715805731937e-05,
+      "loss": 0.6906,
+      "step": 5622
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.630638122558594,
+      "learning_rate": 1.552514158228934e-05,
+      "loss": 1.0032,
+      "step": 5623
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.555750846862793,
+      "learning_rate": 1.552256735884675e-05,
+      "loss": 0.9854,
+      "step": 5624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.529454231262207,
+      "learning_rate": 1.5519993135404154e-05,
+      "loss": 0.5593,
+      "step": 5625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.921797752380371,
+      "learning_rate": 1.551741891196156e-05,
+      "loss": 0.7022,
+      "step": 5626
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.80700969696045,
+      "learning_rate": 1.5514844688518964e-05,
+      "loss": 0.7121,
+      "step": 5627
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.550198554992676,
+      "learning_rate": 1.551227046507637e-05,
+      "loss": 0.451,
+      "step": 5628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.028861999511719,
+      "learning_rate": 1.5509696241633774e-05,
+      "loss": 0.7682,
+      "step": 5629
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.928542137145996,
+      "learning_rate": 1.5507122018191177e-05,
+      "loss": 0.7565,
+      "step": 5630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.143327713012695,
+      "learning_rate": 1.5504547794748584e-05,
+      "loss": 0.5115,
+      "step": 5631
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.377349853515625,
+      "learning_rate": 1.5501973571305987e-05,
+      "loss": 0.6014,
+      "step": 5632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.393840789794922,
+      "learning_rate": 1.5499399347863394e-05,
+      "loss": 0.9157,
+      "step": 5633
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.588883399963379,
+      "learning_rate": 1.54968251244208e-05,
+      "loss": 0.7787,
+      "step": 5634
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.148269176483154,
+      "learning_rate": 1.5494250900978207e-05,
+      "loss": 0.5022,
+      "step": 5635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.02895450592041,
+      "learning_rate": 1.549167667753561e-05,
+      "loss": 0.7828,
+      "step": 5636
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.645138740539551,
+      "learning_rate": 1.5489102454093017e-05,
+      "loss": 0.688,
+      "step": 5637
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.871220588684082,
+      "learning_rate": 1.548652823065042e-05,
+      "loss": 0.6839,
+      "step": 5638
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.227852821350098,
+      "learning_rate": 1.5483954007207827e-05,
+      "loss": 0.6695,
+      "step": 5639
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.953507423400879,
+      "learning_rate": 1.548137978376523e-05,
+      "loss": 0.6187,
+      "step": 5640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.557047843933105,
+      "learning_rate": 1.5478805560322634e-05,
+      "loss": 0.6582,
+      "step": 5641
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.249770164489746,
+      "learning_rate": 1.547623133688004e-05,
+      "loss": 0.6913,
+      "step": 5642
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.216269493103027,
+      "learning_rate": 1.5473657113437447e-05,
+      "loss": 0.7298,
+      "step": 5643
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.961535453796387,
+      "learning_rate": 1.5471082889994854e-05,
+      "loss": 0.7452,
+      "step": 5644
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.155769348144531,
+      "learning_rate": 1.5468508666552257e-05,
+      "loss": 0.8345,
+      "step": 5645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.022253036499023,
+      "learning_rate": 1.5465934443109664e-05,
+      "loss": 0.4066,
+      "step": 5646
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.56843090057373,
+      "learning_rate": 1.5463360219667067e-05,
+      "loss": 0.9937,
+      "step": 5647
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.05449104309082,
+      "learning_rate": 1.5460785996224474e-05,
+      "loss": 0.7719,
+      "step": 5648
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.686256408691406,
+      "learning_rate": 1.5458211772781877e-05,
+      "loss": 0.6502,
+      "step": 5649
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.032071113586426,
+      "learning_rate": 1.5455637549339284e-05,
+      "loss": 0.7303,
+      "step": 5650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.978096008300781,
+      "learning_rate": 1.5453063325896687e-05,
+      "loss": 0.7298,
+      "step": 5651
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.85512638092041,
+      "learning_rate": 1.545048910245409e-05,
+      "loss": 0.806,
+      "step": 5652
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.632328033447266,
+      "learning_rate": 1.54479148790115e-05,
+      "loss": 0.688,
+      "step": 5653
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.15112018585205,
+      "learning_rate": 1.5445340655568904e-05,
+      "loss": 0.6182,
+      "step": 5654
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.540310859680176,
+      "learning_rate": 1.544276643212631e-05,
+      "loss": 0.5802,
+      "step": 5655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.291088104248047,
+      "learning_rate": 1.5440192208683714e-05,
+      "loss": 0.9598,
+      "step": 5656
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.799928188323975,
+      "learning_rate": 1.543761798524112e-05,
+      "loss": 0.5829,
+      "step": 5657
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.369114875793457,
+      "learning_rate": 1.5435043761798524e-05,
+      "loss": 0.81,
+      "step": 5658
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.463690757751465,
+      "learning_rate": 1.543246953835593e-05,
+      "loss": 0.7768,
+      "step": 5659
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.777408599853516,
+      "learning_rate": 1.5429895314913334e-05,
+      "loss": 0.8269,
+      "step": 5660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.610390663146973,
+      "learning_rate": 1.5427321091470737e-05,
+      "loss": 0.8049,
+      "step": 5661
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.596389770507812,
+      "learning_rate": 1.5424746868028147e-05,
+      "loss": 0.6828,
+      "step": 5662
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.544905662536621,
+      "learning_rate": 1.542217264458555e-05,
+      "loss": 0.6366,
+      "step": 5663
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.65538215637207,
+      "learning_rate": 1.5419598421142957e-05,
+      "loss": 0.5219,
+      "step": 5664
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.16901969909668,
+      "learning_rate": 1.541702419770036e-05,
+      "loss": 0.6199,
+      "step": 5665
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.136821746826172,
+      "learning_rate": 1.5414449974257767e-05,
+      "loss": 0.7471,
+      "step": 5666
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.340278625488281,
+      "learning_rate": 1.541187575081517e-05,
+      "loss": 0.82,
+      "step": 5667
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.651012420654297,
+      "learning_rate": 1.5409301527372577e-05,
+      "loss": 0.9482,
+      "step": 5668
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.645423889160156,
+      "learning_rate": 1.540672730392998e-05,
+      "loss": 0.6914,
+      "step": 5669
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.610738754272461,
+      "learning_rate": 1.5404153080487387e-05,
+      "loss": 0.6715,
+      "step": 5670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.92861270904541,
+      "learning_rate": 1.5401578857044794e-05,
+      "loss": 0.8211,
+      "step": 5671
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.625335693359375,
+      "learning_rate": 1.5399004633602197e-05,
+      "loss": 0.7101,
+      "step": 5672
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.360447883605957,
+      "learning_rate": 1.5396430410159604e-05,
+      "loss": 0.7596,
+      "step": 5673
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.988791465759277,
+      "learning_rate": 1.5393856186717007e-05,
+      "loss": 0.6096,
+      "step": 5674
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.392794609069824,
+      "learning_rate": 1.5391281963274414e-05,
+      "loss": 0.9062,
+      "step": 5675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.133210182189941,
+      "learning_rate": 1.5388707739831817e-05,
+      "loss": 0.956,
+      "step": 5676
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.403499603271484,
+      "learning_rate": 1.5386133516389223e-05,
+      "loss": 0.5713,
+      "step": 5677
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.655791282653809,
+      "learning_rate": 1.5383559292946627e-05,
+      "loss": 0.7032,
+      "step": 5678
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.98996639251709,
+      "learning_rate": 1.5380985069504033e-05,
+      "loss": 0.4332,
+      "step": 5679
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.498340606689453,
+      "learning_rate": 1.5378410846061437e-05,
+      "loss": 0.6589,
+      "step": 5680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.076794624328613,
+      "learning_rate": 1.5375836622618847e-05,
+      "loss": 0.5753,
+      "step": 5681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.586895942687988,
+      "learning_rate": 1.537326239917625e-05,
+      "loss": 0.6369,
+      "step": 5682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.331412315368652,
+      "learning_rate": 1.5370688175733653e-05,
+      "loss": 0.8093,
+      "step": 5683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.331785202026367,
+      "learning_rate": 1.536811395229106e-05,
+      "loss": 0.5642,
+      "step": 5684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.352324485778809,
+      "learning_rate": 1.5365539728848463e-05,
+      "loss": 1.1086,
+      "step": 5685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.84002685546875,
+      "learning_rate": 1.536296550540587e-05,
+      "loss": 0.7655,
+      "step": 5686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.961621284484863,
+      "learning_rate": 1.5360391281963273e-05,
+      "loss": 0.5642,
+      "step": 5687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.137855529785156,
+      "learning_rate": 1.535781705852068e-05,
+      "loss": 0.7061,
+      "step": 5688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.495466232299805,
+      "learning_rate": 1.5355242835078083e-05,
+      "loss": 0.8364,
+      "step": 5689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.012463569641113,
+      "learning_rate": 1.5352668611635493e-05,
+      "loss": 0.8006,
+      "step": 5690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.657920837402344,
+      "learning_rate": 1.5350094388192897e-05,
+      "loss": 0.6529,
+      "step": 5691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.993449211120605,
+      "learning_rate": 1.53475201647503e-05,
+      "loss": 0.6604,
+      "step": 5692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.653358459472656,
+      "learning_rate": 1.5344945941307707e-05,
+      "loss": 0.788,
+      "step": 5693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.433140754699707,
+      "learning_rate": 1.534237171786511e-05,
+      "loss": 0.6557,
+      "step": 5694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.148329734802246,
+      "learning_rate": 1.5339797494422517e-05,
+      "loss": 0.7572,
+      "step": 5695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.149693489074707,
+      "learning_rate": 1.533722327097992e-05,
+      "loss": 0.5877,
+      "step": 5696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.582533836364746,
+      "learning_rate": 1.5334649047537327e-05,
+      "loss": 0.8213,
+      "step": 5697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.772625923156738,
+      "learning_rate": 1.533207482409473e-05,
+      "loss": 0.6146,
+      "step": 5698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.03817367553711,
+      "learning_rate": 1.5329500600652137e-05,
+      "loss": 0.7681,
+      "step": 5699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.559398651123047,
+      "learning_rate": 1.5326926377209543e-05,
+      "loss": 0.6298,
+      "step": 5700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.982650756835938,
+      "learning_rate": 1.532435215376695e-05,
+      "loss": 0.6202,
+      "step": 5701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.992537498474121,
+      "learning_rate": 1.5321777930324353e-05,
+      "loss": 0.8079,
+      "step": 5702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.939094543457031,
+      "learning_rate": 1.5319203706881757e-05,
+      "loss": 0.5668,
+      "step": 5703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.615065574645996,
+      "learning_rate": 1.5316629483439163e-05,
+      "loss": 0.6235,
+      "step": 5704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.674843788146973,
+      "learning_rate": 1.5314055259996567e-05,
+      "loss": 0.7264,
+      "step": 5705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.875312805175781,
+      "learning_rate": 1.5311481036553973e-05,
+      "loss": 0.7015,
+      "step": 5706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.64791202545166,
+      "learning_rate": 1.5308906813111376e-05,
+      "loss": 0.8735,
+      "step": 5707
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.883646965026855,
+      "learning_rate": 1.5306332589668783e-05,
+      "loss": 0.7349,
+      "step": 5708
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.677424430847168,
+      "learning_rate": 1.530375836622619e-05,
+      "loss": 0.5359,
+      "step": 5709
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.181770324707031,
+      "learning_rate": 1.5301184142783597e-05,
+      "loss": 0.5543,
+      "step": 5710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.04793930053711,
+      "learning_rate": 1.5298609919341e-05,
+      "loss": 0.9618,
+      "step": 5711
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.564706802368164,
+      "learning_rate": 1.5296035695898407e-05,
+      "loss": 0.4316,
+      "step": 5712
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.771110534667969,
+      "learning_rate": 1.529346147245581e-05,
+      "loss": 0.7018,
+      "step": 5713
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.063695907592773,
+      "learning_rate": 1.5290887249013213e-05,
+      "loss": 0.7138,
+      "step": 5714
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.364028930664062,
+      "learning_rate": 1.528831302557062e-05,
+      "loss": 0.6103,
+      "step": 5715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.407870292663574,
+      "learning_rate": 1.5285738802128023e-05,
+      "loss": 0.6776,
+      "step": 5716
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.684124946594238,
+      "learning_rate": 1.528316457868543e-05,
+      "loss": 0.8618,
+      "step": 5717
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.859552383422852,
+      "learning_rate": 1.5280590355242833e-05,
+      "loss": 0.7771,
+      "step": 5718
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.708332061767578,
+      "learning_rate": 1.5278016131800243e-05,
+      "loss": 0.9494,
+      "step": 5719
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.849374771118164,
+      "learning_rate": 1.5275441908357646e-05,
+      "loss": 0.6346,
+      "step": 5720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.793769836425781,
+      "learning_rate": 1.5272867684915053e-05,
+      "loss": 0.6131,
+      "step": 5721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.740447998046875,
+      "learning_rate": 1.5270293461472456e-05,
+      "loss": 0.9266,
+      "step": 5722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.526744842529297,
+      "learning_rate": 1.526771923802986e-05,
+      "loss": 0.9554,
+      "step": 5723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.141620635986328,
+      "learning_rate": 1.5265145014587266e-05,
+      "loss": 0.7873,
+      "step": 5724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.561409950256348,
+      "learning_rate": 1.526257079114467e-05,
+      "loss": 0.5033,
+      "step": 5725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.583409309387207,
+      "learning_rate": 1.5259996567702076e-05,
+      "loss": 0.7237,
+      "step": 5726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.0056791305542,
+      "learning_rate": 1.5257422344259481e-05,
+      "loss": 0.5162,
+      "step": 5727
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.778443336486816,
+      "learning_rate": 1.5254848120816888e-05,
+      "loss": 0.6362,
+      "step": 5728
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.222827911376953,
+      "learning_rate": 1.5252273897374293e-05,
+      "loss": 0.9317,
+      "step": 5729
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.880953311920166,
+      "learning_rate": 1.5249699673931698e-05,
+      "loss": 0.4786,
+      "step": 5730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.761146545410156,
+      "learning_rate": 1.5247125450489103e-05,
+      "loss": 0.4865,
+      "step": 5731
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.01679515838623,
+      "learning_rate": 1.5244551227046508e-05,
+      "loss": 0.7607,
+      "step": 5732
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.966072082519531,
+      "learning_rate": 1.5241977003603913e-05,
+      "loss": 0.6358,
+      "step": 5733
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.958866119384766,
+      "learning_rate": 1.5239402780161318e-05,
+      "loss": 0.6514,
+      "step": 5734
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.584814071655273,
+      "learning_rate": 1.5236828556718723e-05,
+      "loss": 0.6512,
+      "step": 5735
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.573479652404785,
+      "learning_rate": 1.5234254333276128e-05,
+      "loss": 0.5867,
+      "step": 5736
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.151217460632324,
+      "learning_rate": 1.5231680109833533e-05,
+      "loss": 0.5512,
+      "step": 5737
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.794495582580566,
+      "learning_rate": 1.5229105886390941e-05,
+      "loss": 0.6392,
+      "step": 5738
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.396523475646973,
+      "learning_rate": 1.5226531662948345e-05,
+      "loss": 0.6959,
+      "step": 5739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.361831665039062,
+      "learning_rate": 1.522395743950575e-05,
+      "loss": 0.6456,
+      "step": 5740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.718506813049316,
+      "learning_rate": 1.5221383216063155e-05,
+      "loss": 0.837,
+      "step": 5741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.24177074432373,
+      "learning_rate": 1.521880899262056e-05,
+      "loss": 0.7099,
+      "step": 5742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.529495239257812,
+      "learning_rate": 1.5216234769177965e-05,
+      "loss": 0.7896,
+      "step": 5743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.81344223022461,
+      "learning_rate": 1.521366054573537e-05,
+      "loss": 0.6909,
+      "step": 5744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.75560474395752,
+      "learning_rate": 1.5211086322292774e-05,
+      "loss": 0.52,
+      "step": 5745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.679831504821777,
+      "learning_rate": 1.520851209885018e-05,
+      "loss": 0.5626,
+      "step": 5746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.186136245727539,
+      "learning_rate": 1.5205937875407588e-05,
+      "loss": 0.6164,
+      "step": 5747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.998634338378906,
+      "learning_rate": 1.5203363651964993e-05,
+      "loss": 0.776,
+      "step": 5748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.41545295715332,
+      "learning_rate": 1.5200789428522396e-05,
+      "loss": 0.4787,
+      "step": 5749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.25280475616455,
+      "learning_rate": 1.5198215205079801e-05,
+      "loss": 0.5787,
+      "step": 5750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.7398681640625,
+      "learning_rate": 1.5195640981637206e-05,
+      "loss": 0.7759,
+      "step": 5751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.124595642089844,
+      "learning_rate": 1.5193066758194611e-05,
+      "loss": 0.6163,
+      "step": 5752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.229990005493164,
+      "learning_rate": 1.5190492534752016e-05,
+      "loss": 0.6962,
+      "step": 5753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.292335510253906,
+      "learning_rate": 1.5187918311309421e-05,
+      "loss": 0.7061,
+      "step": 5754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.922220230102539,
+      "learning_rate": 1.5185344087866826e-05,
+      "loss": 0.6132,
+      "step": 5755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.665719032287598,
+      "learning_rate": 1.5182769864424231e-05,
+      "loss": 0.6725,
+      "step": 5756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.034870147705078,
+      "learning_rate": 1.518019564098164e-05,
+      "loss": 0.6868,
+      "step": 5757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.273383140563965,
+      "learning_rate": 1.5177621417539044e-05,
+      "loss": 0.54,
+      "step": 5758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.658744812011719,
+      "learning_rate": 1.517504719409645e-05,
+      "loss": 0.5812,
+      "step": 5759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.452827453613281,
+      "learning_rate": 1.5172472970653853e-05,
+      "loss": 0.5922,
+      "step": 5760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.487955093383789,
+      "learning_rate": 1.5169898747211258e-05,
+      "loss": 0.7667,
+      "step": 5761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.797428131103516,
+      "learning_rate": 1.5167324523768663e-05,
+      "loss": 0.6671,
+      "step": 5762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.021202087402344,
+      "learning_rate": 1.5164750300326068e-05,
+      "loss": 0.4791,
+      "step": 5763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.655767440795898,
+      "learning_rate": 1.5162176076883473e-05,
+      "loss": 1.0293,
+      "step": 5764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.35323429107666,
+      "learning_rate": 1.5159601853440878e-05,
+      "loss": 0.6349,
+      "step": 5765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.621145248413086,
+      "learning_rate": 1.5157027629998286e-05,
+      "loss": 0.4933,
+      "step": 5766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.433503150939941,
+      "learning_rate": 1.5154453406555691e-05,
+      "loss": 0.7231,
+      "step": 5767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.64327621459961,
+      "learning_rate": 1.5151879183113096e-05,
+      "loss": 0.8214,
+      "step": 5768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.654711723327637,
+      "learning_rate": 1.5149304959670501e-05,
+      "loss": 0.5217,
+      "step": 5769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.769917488098145,
+      "learning_rate": 1.5146730736227904e-05,
+      "loss": 0.7467,
+      "step": 5770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.103089332580566,
+      "learning_rate": 1.514415651278531e-05,
+      "loss": 0.4932,
+      "step": 5771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.62012004852295,
+      "learning_rate": 1.5141582289342714e-05,
+      "loss": 0.5993,
+      "step": 5772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.610448837280273,
+      "learning_rate": 1.513900806590012e-05,
+      "loss": 0.6998,
+      "step": 5773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.021026611328125,
+      "learning_rate": 1.5136433842457524e-05,
+      "loss": 0.5324,
+      "step": 5774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.560358047485352,
+      "learning_rate": 1.5133859619014933e-05,
+      "loss": 0.4598,
+      "step": 5775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.713452339172363,
+      "learning_rate": 1.5131285395572338e-05,
+      "loss": 0.8595,
+      "step": 5776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.286099433898926,
+      "learning_rate": 1.5128711172129743e-05,
+      "loss": 0.7166,
+      "step": 5777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.132092475891113,
+      "learning_rate": 1.5126136948687148e-05,
+      "loss": 0.5737,
+      "step": 5778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.688316345214844,
+      "learning_rate": 1.5123562725244553e-05,
+      "loss": 0.7009,
+      "step": 5779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.423620223999023,
+      "learning_rate": 1.5120988501801958e-05,
+      "loss": 0.6277,
+      "step": 5780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.948112487792969,
+      "learning_rate": 1.511841427835936e-05,
+      "loss": 0.8181,
+      "step": 5781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.351773262023926,
+      "learning_rate": 1.5115840054916766e-05,
+      "loss": 0.5594,
+      "step": 5782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.852547645568848,
+      "learning_rate": 1.511326583147417e-05,
+      "loss": 0.9714,
+      "step": 5783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.424971580505371,
+      "learning_rate": 1.5110691608031576e-05,
+      "loss": 0.6949,
+      "step": 5784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.850028038024902,
+      "learning_rate": 1.5108117384588984e-05,
+      "loss": 0.8682,
+      "step": 5785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.023741722106934,
+      "learning_rate": 1.5105543161146389e-05,
+      "loss": 0.6199,
+      "step": 5786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.838698387145996,
+      "learning_rate": 1.5102968937703794e-05,
+      "loss": 0.5625,
+      "step": 5787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.173796653747559,
+      "learning_rate": 1.5100394714261199e-05,
+      "loss": 0.6288,
+      "step": 5788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.8538818359375,
+      "learning_rate": 1.5097820490818604e-05,
+      "loss": 0.4745,
+      "step": 5789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.6320319175720215,
+      "learning_rate": 1.5095246267376009e-05,
+      "loss": 0.527,
+      "step": 5790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.076807975769043,
+      "learning_rate": 1.5092672043933412e-05,
+      "loss": 0.5604,
+      "step": 5791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.503802299499512,
+      "learning_rate": 1.5090097820490817e-05,
+      "loss": 0.5523,
+      "step": 5792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.966728210449219,
+      "learning_rate": 1.5087523597048222e-05,
+      "loss": 0.8937,
+      "step": 5793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.7310209274292,
+      "learning_rate": 1.508494937360563e-05,
+      "loss": 0.8369,
+      "step": 5794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.883963584899902,
+      "learning_rate": 1.5082375150163036e-05,
+      "loss": 0.7915,
+      "step": 5795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.055442810058594,
+      "learning_rate": 1.507980092672044e-05,
+      "loss": 0.5225,
+      "step": 5796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.309677124023438,
+      "learning_rate": 1.5077226703277846e-05,
+      "loss": 0.5614,
+      "step": 5797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.3325834274292,
+      "learning_rate": 1.507465247983525e-05,
+      "loss": 0.8401,
+      "step": 5798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.112789154052734,
+      "learning_rate": 1.5072078256392656e-05,
+      "loss": 0.7446,
+      "step": 5799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.968636512756348,
+      "learning_rate": 1.506950403295006e-05,
+      "loss": 0.5288,
+      "step": 5800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.94079875946045,
+      "learning_rate": 1.5066929809507466e-05,
+      "loss": 0.9001,
+      "step": 5801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.881861686706543,
+      "learning_rate": 1.5064355586064869e-05,
+      "loss": 0.591,
+      "step": 5802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.56390380859375,
+      "learning_rate": 1.5061781362622274e-05,
+      "loss": 0.4244,
+      "step": 5803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.640572547912598,
+      "learning_rate": 1.5059207139179682e-05,
+      "loss": 0.6966,
+      "step": 5804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.12524700164795,
+      "learning_rate": 1.5056632915737087e-05,
+      "loss": 0.7201,
+      "step": 5805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.12606430053711,
+      "learning_rate": 1.5054058692294492e-05,
+      "loss": 0.9234,
+      "step": 5806
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.305171489715576,
+      "learning_rate": 1.5051484468851897e-05,
+      "loss": 0.4994,
+      "step": 5807
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.895528793334961,
+      "learning_rate": 1.5048910245409302e-05,
+      "loss": 0.5146,
+      "step": 5808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.512129783630371,
+      "learning_rate": 1.5046336021966707e-05,
+      "loss": 0.674,
+      "step": 5809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.729283332824707,
+      "learning_rate": 1.5043761798524112e-05,
+      "loss": 0.5777,
+      "step": 5810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.76346206665039,
+      "learning_rate": 1.5041187575081517e-05,
+      "loss": 0.7952,
+      "step": 5811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.924771308898926,
+      "learning_rate": 1.503861335163892e-05,
+      "loss": 0.6634,
+      "step": 5812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.528681755065918,
+      "learning_rate": 1.5036039128196329e-05,
+      "loss": 0.7472,
+      "step": 5813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.787907600402832,
+      "learning_rate": 1.5033464904753734e-05,
+      "loss": 0.501,
+      "step": 5814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.359318733215332,
+      "learning_rate": 1.5030890681311139e-05,
+      "loss": 0.567,
+      "step": 5815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.569374084472656,
+      "learning_rate": 1.5028316457868544e-05,
+      "loss": 0.796,
+      "step": 5816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.525404930114746,
+      "learning_rate": 1.5025742234425949e-05,
+      "loss": 0.8468,
+      "step": 5817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.533405303955078,
+      "learning_rate": 1.5023168010983354e-05,
+      "loss": 0.7123,
+      "step": 5818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.92939567565918,
+      "learning_rate": 1.5020593787540759e-05,
+      "loss": 0.6604,
+      "step": 5819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.282258033752441,
+      "learning_rate": 1.5018019564098164e-05,
+      "loss": 0.6263,
+      "step": 5820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.508992195129395,
+      "learning_rate": 1.5015445340655569e-05,
+      "loss": 0.7314,
+      "step": 5821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.85505485534668,
+      "learning_rate": 1.5012871117212972e-05,
+      "loss": 0.5178,
+      "step": 5822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.519545555114746,
+      "learning_rate": 1.501029689377038e-05,
+      "loss": 1.1049,
+      "step": 5823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.095693588256836,
+      "learning_rate": 1.5007722670327785e-05,
+      "loss": 0.4994,
+      "step": 5824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.617183685302734,
+      "learning_rate": 1.500514844688519e-05,
+      "loss": 0.733,
+      "step": 5825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.229683876037598,
+      "learning_rate": 1.5002574223442595e-05,
+      "loss": 0.6827,
+      "step": 5826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.437691688537598,
+      "learning_rate": 1.5e-05,
+      "loss": 0.5757,
+      "step": 5827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.996328353881836,
+      "learning_rate": 1.4997425776557405e-05,
+      "loss": 0.5054,
+      "step": 5828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.449949264526367,
+      "learning_rate": 1.499485155311481e-05,
+      "loss": 0.7316,
+      "step": 5829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.471872329711914,
+      "learning_rate": 1.4992277329672217e-05,
+      "loss": 0.6715,
+      "step": 5830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.9858980178833,
+      "learning_rate": 1.4989703106229622e-05,
+      "loss": 0.4177,
+      "step": 5831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.0646390914917,
+      "learning_rate": 1.4987128882787027e-05,
+      "loss": 0.5055,
+      "step": 5832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.068058013916016,
+      "learning_rate": 1.498455465934443e-05,
+      "loss": 0.5957,
+      "step": 5833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.851325035095215,
+      "learning_rate": 1.4981980435901835e-05,
+      "loss": 0.5739,
+      "step": 5834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.4882230758667,
+      "learning_rate": 1.4979406212459242e-05,
+      "loss": 0.4547,
+      "step": 5835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.539527893066406,
+      "learning_rate": 1.4976831989016647e-05,
+      "loss": 0.5826,
+      "step": 5836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.06041145324707,
+      "learning_rate": 1.4974257765574052e-05,
+      "loss": 0.6335,
+      "step": 5837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.827814102172852,
+      "learning_rate": 1.4971683542131457e-05,
+      "loss": 0.6449,
+      "step": 5838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.952459335327148,
+      "learning_rate": 1.4969109318688864e-05,
+      "loss": 0.6177,
+      "step": 5839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.339852333068848,
+      "learning_rate": 1.4966535095246269e-05,
+      "loss": 0.7179,
+      "step": 5840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.543264389038086,
+      "learning_rate": 1.4963960871803674e-05,
+      "loss": 0.6508,
+      "step": 5841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.495478630065918,
+      "learning_rate": 1.4961386648361079e-05,
+      "loss": 0.6763,
+      "step": 5842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.267716407775879,
+      "learning_rate": 1.4958812424918482e-05,
+      "loss": 0.7016,
+      "step": 5843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.899517059326172,
+      "learning_rate": 1.4956238201475889e-05,
+      "loss": 0.7314,
+      "step": 5844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.8445658683776855,
+      "learning_rate": 1.4953663978033294e-05,
+      "loss": 0.5011,
+      "step": 5845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.901155471801758,
+      "learning_rate": 1.4951089754590699e-05,
+      "loss": 0.813,
+      "step": 5846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.040955543518066,
+      "learning_rate": 1.4948515531148104e-05,
+      "loss": 0.6457,
+      "step": 5847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.827336311340332,
+      "learning_rate": 1.4945941307705509e-05,
+      "loss": 0.5509,
+      "step": 5848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.569506645202637,
+      "learning_rate": 1.4943367084262915e-05,
+      "loss": 0.5852,
+      "step": 5849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.296262741088867,
+      "learning_rate": 1.494079286082032e-05,
+      "loss": 0.4235,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.892028331756592,
+      "learning_rate": 1.4938218637377725e-05,
+      "loss": 0.4394,
+      "step": 5851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.314630508422852,
+      "learning_rate": 1.493564441393513e-05,
+      "loss": 0.6473,
+      "step": 5852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.771956443786621,
+      "learning_rate": 1.4933070190492535e-05,
+      "loss": 0.5424,
+      "step": 5853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.939587593078613,
+      "learning_rate": 1.493049596704994e-05,
+      "loss": 0.8219,
+      "step": 5854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.545926570892334,
+      "learning_rate": 1.4927921743607345e-05,
+      "loss": 0.4095,
+      "step": 5855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.807062149047852,
+      "learning_rate": 1.492534752016475e-05,
+      "loss": 0.5949,
+      "step": 5856
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.42801570892334,
+      "learning_rate": 1.4922773296722155e-05,
+      "loss": 0.4882,
+      "step": 5857
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.682374954223633,
+      "learning_rate": 1.4920199073279562e-05,
+      "loss": 0.6485,
+      "step": 5858
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.165955543518066,
+      "learning_rate": 1.4917624849836967e-05,
+      "loss": 0.6441,
+      "step": 5859
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.182374954223633,
+      "learning_rate": 1.4915050626394372e-05,
+      "loss": 0.5382,
+      "step": 5860
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.19049072265625,
+      "learning_rate": 1.4912476402951777e-05,
+      "loss": 0.5167,
+      "step": 5861
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.343765258789062,
+      "learning_rate": 1.4909902179509182e-05,
+      "loss": 0.5953,
+      "step": 5862
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.150381088256836,
+      "learning_rate": 1.4907327956066588e-05,
+      "loss": 0.8045,
+      "step": 5863
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.685317039489746,
+      "learning_rate": 1.4904753732623992e-05,
+      "loss": 0.5688,
+      "step": 5864
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.326414108276367,
+      "learning_rate": 1.4902179509181397e-05,
+      "loss": 0.8766,
+      "step": 5865
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.847957134246826,
+      "learning_rate": 1.4899605285738802e-05,
+      "loss": 0.4393,
+      "step": 5866
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.0089693069458,
+      "learning_rate": 1.4897031062296207e-05,
+      "loss": 0.5271,
+      "step": 5867
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.443350315093994,
+      "learning_rate": 1.4894456838853613e-05,
+      "loss": 0.4036,
+      "step": 5868
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.425400733947754,
+      "learning_rate": 1.4891882615411018e-05,
+      "loss": 0.8541,
+      "step": 5869
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.673444747924805,
+      "learning_rate": 1.4889308391968423e-05,
+      "loss": 0.6116,
+      "step": 5870
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.117676734924316,
+      "learning_rate": 1.4886734168525828e-05,
+      "loss": 0.4866,
+      "step": 5871
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.535839080810547,
+      "learning_rate": 1.4884159945083235e-05,
+      "loss": 0.496,
+      "step": 5872
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.214252471923828,
+      "learning_rate": 1.488158572164064e-05,
+      "loss": 0.6112,
+      "step": 5873
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.156107902526855,
+      "learning_rate": 1.4879011498198043e-05,
+      "loss": 0.7178,
+      "step": 5874
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.114778518676758,
+      "learning_rate": 1.4876437274755448e-05,
+      "loss": 0.716,
+      "step": 5875
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.132325172424316,
+      "learning_rate": 1.4873863051312853e-05,
+      "loss": 0.8676,
+      "step": 5876
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.38235092163086,
+      "learning_rate": 1.487128882787026e-05,
+      "loss": 0.5939,
+      "step": 5877
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.981067657470703,
+      "learning_rate": 1.4868714604427665e-05,
+      "loss": 0.595,
+      "step": 5878
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.459383010864258,
+      "learning_rate": 1.486614038098507e-05,
+      "loss": 0.7795,
+      "step": 5879
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.010276794433594,
+      "learning_rate": 1.4863566157542475e-05,
+      "loss": 0.7097,
+      "step": 5880
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.208502769470215,
+      "learning_rate": 1.486099193409988e-05,
+      "loss": 0.3053,
+      "step": 5881
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.99191665649414,
+      "learning_rate": 1.4858417710657287e-05,
+      "loss": 0.4777,
+      "step": 5882
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.625293731689453,
+      "learning_rate": 1.4855843487214692e-05,
+      "loss": 0.6028,
+      "step": 5883
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.562214851379395,
+      "learning_rate": 1.4853269263772097e-05,
+      "loss": 0.5358,
+      "step": 5884
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.344232559204102,
+      "learning_rate": 1.48506950403295e-05,
+      "loss": 0.5031,
+      "step": 5885
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.200315475463867,
+      "learning_rate": 1.4848120816886905e-05,
+      "loss": 0.6646,
+      "step": 5886
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.999994277954102,
+      "learning_rate": 1.4845546593444311e-05,
+      "loss": 0.5921,
+      "step": 5887
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.149828910827637,
+      "learning_rate": 1.4842972370001716e-05,
+      "loss": 0.4568,
+      "step": 5888
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.828720092773438,
+      "learning_rate": 1.4840398146559121e-05,
+      "loss": 0.5932,
+      "step": 5889
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.309082984924316,
+      "learning_rate": 1.4837823923116526e-05,
+      "loss": 0.4213,
+      "step": 5890
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.649279594421387,
+      "learning_rate": 1.4835249699673933e-05,
+      "loss": 0.4919,
+      "step": 5891
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.159218788146973,
+      "learning_rate": 1.4832675476231338e-05,
+      "loss": 0.6065,
+      "step": 5892
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.690546035766602,
+      "learning_rate": 1.4830101252788743e-05,
+      "loss": 0.5822,
+      "step": 5893
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.738336563110352,
+      "learning_rate": 1.4827527029346148e-05,
+      "loss": 0.5516,
+      "step": 5894
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.259581565856934,
+      "learning_rate": 1.4824952805903551e-05,
+      "loss": 0.5583,
+      "step": 5895
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.709681510925293,
+      "learning_rate": 1.4822378582460958e-05,
+      "loss": 0.4218,
+      "step": 5896
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.941258430480957,
+      "learning_rate": 1.4819804359018363e-05,
+      "loss": 0.3773,
+      "step": 5897
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.74229621887207,
+      "learning_rate": 1.4817230135575768e-05,
+      "loss": 0.6493,
+      "step": 5898
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.19922161102295,
+      "learning_rate": 1.4814655912133173e-05,
+      "loss": 0.5747,
+      "step": 5899
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.202072143554688,
+      "learning_rate": 1.4812081688690578e-05,
+      "loss": 0.4071,
+      "step": 5900
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.328014373779297,
+      "learning_rate": 1.4809507465247985e-05,
+      "loss": 0.7073,
+      "step": 5901
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.359185218811035,
+      "learning_rate": 1.480693324180539e-05,
+      "loss": 0.6825,
+      "step": 5902
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.029674530029297,
+      "learning_rate": 1.4804359018362795e-05,
+      "loss": 0.4647,
+      "step": 5903
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.793015480041504,
+      "learning_rate": 1.48017847949202e-05,
+      "loss": 0.7146,
+      "step": 5904
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.733183860778809,
+      "learning_rate": 1.4799210571477605e-05,
+      "loss": 0.3918,
+      "step": 5905
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.152375221252441,
+      "learning_rate": 1.479663634803501e-05,
+      "loss": 0.4684,
+      "step": 5906
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.329442024230957,
+      "learning_rate": 1.4794062124592415e-05,
+      "loss": 0.5085,
+      "step": 5907
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.1718111038208,
+      "learning_rate": 1.479148790114982e-05,
+      "loss": 0.5066,
+      "step": 5908
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.822720527648926,
+      "learning_rate": 1.4788913677707225e-05,
+      "loss": 0.6753,
+      "step": 5909
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.792845726013184,
+      "learning_rate": 1.4786339454264631e-05,
+      "loss": 0.3951,
+      "step": 5910
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.74872875213623,
+      "learning_rate": 1.4783765230822036e-05,
+      "loss": 0.3757,
+      "step": 5911
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.287698745727539,
+      "learning_rate": 1.4781191007379441e-05,
+      "loss": 0.6864,
+      "step": 5912
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.916751861572266,
+      "learning_rate": 1.4778616783936846e-05,
+      "loss": 0.4277,
+      "step": 5913
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.870539665222168,
+      "learning_rate": 1.4776042560494251e-05,
+      "loss": 0.592,
+      "step": 5914
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.008909225463867,
+      "learning_rate": 1.4773468337051658e-05,
+      "loss": 0.7179,
+      "step": 5915
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.516189575195312,
+      "learning_rate": 1.4770894113609061e-05,
+      "loss": 0.5969,
+      "step": 5916
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.788212776184082,
+      "learning_rate": 1.4768319890166466e-05,
+      "loss": 0.518,
+      "step": 5917
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.915926933288574,
+      "learning_rate": 1.4765745666723871e-05,
+      "loss": 0.623,
+      "step": 5918
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.850130081176758,
+      "learning_rate": 1.4763171443281276e-05,
+      "loss": 0.5302,
+      "step": 5919
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.732810020446777,
+      "learning_rate": 1.4760597219838683e-05,
+      "loss": 0.7025,
+      "step": 5920
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.842598915100098,
+      "learning_rate": 1.4758022996396088e-05,
+      "loss": 0.7036,
+      "step": 5921
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.236542701721191,
+      "learning_rate": 1.4755448772953493e-05,
+      "loss": 0.4438,
+      "step": 5922
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.911584854125977,
+      "learning_rate": 1.4752874549510898e-05,
+      "loss": 0.4967,
+      "step": 5923
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.001578330993652,
+      "learning_rate": 1.4750300326068304e-05,
+      "loss": 0.5498,
+      "step": 5924
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.843045234680176,
+      "learning_rate": 1.474772610262571e-05,
+      "loss": 0.6339,
+      "step": 5925
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.488044738769531,
+      "learning_rate": 1.4745151879183113e-05,
+      "loss": 0.7011,
+      "step": 5926
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.051386833190918,
+      "learning_rate": 1.4742577655740518e-05,
+      "loss": 0.6415,
+      "step": 5927
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.3502779006958,
+      "learning_rate": 1.4740003432297923e-05,
+      "loss": 0.5564,
+      "step": 5928
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.254146575927734,
+      "learning_rate": 1.473742920885533e-05,
+      "loss": 0.4641,
+      "step": 5929
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.9171781539917,
+      "learning_rate": 1.4734854985412734e-05,
+      "loss": 0.6273,
+      "step": 5930
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.963153839111328,
+      "learning_rate": 1.473228076197014e-05,
+      "loss": 0.5815,
+      "step": 5931
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.150697708129883,
+      "learning_rate": 1.4729706538527544e-05,
+      "loss": 0.6865,
+      "step": 5932
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.16169548034668,
+      "learning_rate": 1.472713231508495e-05,
+      "loss": 0.6854,
+      "step": 5933
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.712112426757812,
+      "learning_rate": 1.4724558091642356e-05,
+      "loss": 0.5281,
+      "step": 5934
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.701800346374512,
+      "learning_rate": 1.4721983868199761e-05,
+      "loss": 0.6936,
+      "step": 5935
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.796243667602539,
+      "learning_rate": 1.4719409644757166e-05,
+      "loss": 0.3823,
+      "step": 5936
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.319583892822266,
+      "learning_rate": 1.471683542131457e-05,
+      "loss": 0.4861,
+      "step": 5937
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.393757820129395,
+      "learning_rate": 1.4714261197871974e-05,
+      "loss": 0.7571,
+      "step": 5938
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.345955848693848,
+      "learning_rate": 1.4711686974429381e-05,
+      "loss": 0.497,
+      "step": 5939
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.249866485595703,
+      "learning_rate": 1.4709112750986786e-05,
+      "loss": 0.5631,
+      "step": 5940
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.252386093139648,
+      "learning_rate": 1.4706538527544191e-05,
+      "loss": 0.5427,
+      "step": 5941
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.183730125427246,
+      "learning_rate": 1.4703964304101596e-05,
+      "loss": 0.7649,
+      "step": 5942
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.999211311340332,
+      "learning_rate": 1.4701390080659003e-05,
+      "loss": 0.7697,
+      "step": 5943
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.877355575561523,
+      "learning_rate": 1.4698815857216408e-05,
+      "loss": 0.5126,
+      "step": 5944
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.60456657409668,
+      "learning_rate": 1.4696241633773813e-05,
+      "loss": 0.5625,
+      "step": 5945
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.823182106018066,
+      "learning_rate": 1.4693667410331218e-05,
+      "loss": 0.707,
+      "step": 5946
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.35696029663086,
+      "learning_rate": 1.4691093186888621e-05,
+      "loss": 0.6407,
+      "step": 5947
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.381437301635742,
+      "learning_rate": 1.4688518963446028e-05,
+      "loss": 0.6908,
+      "step": 5948
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.455245971679688,
+      "learning_rate": 1.4685944740003433e-05,
+      "loss": 0.5932,
+      "step": 5949
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.504873275756836,
+      "learning_rate": 1.4683370516560838e-05,
+      "loss": 0.6447,
+      "step": 5950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.666354179382324,
+      "learning_rate": 1.4680796293118243e-05,
+      "loss": 0.6577,
+      "step": 5951
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.356197357177734,
+      "learning_rate": 1.4678222069675648e-05,
+      "loss": 0.4181,
+      "step": 5952
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.843424797058105,
+      "learning_rate": 1.4675647846233054e-05,
+      "loss": 0.6225,
+      "step": 5953
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.660552978515625,
+      "learning_rate": 1.467307362279046e-05,
+      "loss": 0.5203,
+      "step": 5954
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.64244556427002,
+      "learning_rate": 1.4670499399347864e-05,
+      "loss": 0.5729,
+      "step": 5955
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.673271179199219,
+      "learning_rate": 1.466792517590527e-05,
+      "loss": 0.4832,
+      "step": 5956
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.190770149230957,
+      "learning_rate": 1.4665350952462674e-05,
+      "loss": 0.5042,
+      "step": 5957
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.698134422302246,
+      "learning_rate": 1.4662776729020079e-05,
+      "loss": 0.4508,
+      "step": 5958
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.254560470581055,
+      "learning_rate": 1.4660202505577484e-05,
+      "loss": 0.6634,
+      "step": 5959
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.10484790802002,
+      "learning_rate": 1.4657628282134889e-05,
+      "loss": 0.3792,
+      "step": 5960
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.497947692871094,
+      "learning_rate": 1.4655054058692294e-05,
+      "loss": 0.511,
+      "step": 5961
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.054791450500488,
+      "learning_rate": 1.46524798352497e-05,
+      "loss": 0.5077,
+      "step": 5962
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.14282512664795,
+      "learning_rate": 1.4649905611807106e-05,
+      "loss": 0.4494,
+      "step": 5963
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.792989730834961,
+      "learning_rate": 1.464733138836451e-05,
+      "loss": 0.9415,
+      "step": 5964
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.341668128967285,
+      "learning_rate": 1.4644757164921916e-05,
+      "loss": 0.4478,
+      "step": 5965
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.086487770080566,
+      "learning_rate": 1.464218294147932e-05,
+      "loss": 0.6163,
+      "step": 5966
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.91039752960205,
+      "learning_rate": 1.4639608718036727e-05,
+      "loss": 0.637,
+      "step": 5967
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.866921424865723,
+      "learning_rate": 1.463703449459413e-05,
+      "loss": 0.409,
+      "step": 5968
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.187800407409668,
+      "learning_rate": 1.4634460271151536e-05,
+      "loss": 0.5432,
+      "step": 5969
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.714569091796875,
+      "learning_rate": 1.463188604770894e-05,
+      "loss": 0.5639,
+      "step": 5970
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.751258373260498,
+      "learning_rate": 1.4629311824266346e-05,
+      "loss": 0.3676,
+      "step": 5971
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.975625038146973,
+      "learning_rate": 1.4626737600823752e-05,
+      "loss": 0.5023,
+      "step": 5972
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.23196792602539,
+      "learning_rate": 1.4624163377381157e-05,
+      "loss": 0.5034,
+      "step": 5973
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.886031150817871,
+      "learning_rate": 1.4621589153938562e-05,
+      "loss": 0.4551,
+      "step": 5974
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.783336639404297,
+      "learning_rate": 1.4619014930495967e-05,
+      "loss": 0.6406,
+      "step": 5975
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.055586814880371,
+      "learning_rate": 1.4616440707053374e-05,
+      "loss": 0.548,
+      "step": 5976
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.34635066986084,
+      "learning_rate": 1.4613866483610779e-05,
+      "loss": 0.5204,
+      "step": 5977
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.790834426879883,
+      "learning_rate": 1.4611292260168182e-05,
+      "loss": 0.4637,
+      "step": 5978
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.629433631896973,
+      "learning_rate": 1.4608718036725587e-05,
+      "loss": 0.3886,
+      "step": 5979
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.253458976745605,
+      "learning_rate": 1.4606143813282992e-05,
+      "loss": 0.5302,
+      "step": 5980
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.117012977600098,
+      "learning_rate": 1.4603569589840399e-05,
+      "loss": 0.5157,
+      "step": 5981
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.922342300415039,
+      "learning_rate": 1.4600995366397804e-05,
+      "loss": 0.5678,
+      "step": 5982
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.276809692382812,
+      "learning_rate": 1.4598421142955209e-05,
+      "loss": 0.4387,
+      "step": 5983
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.134427070617676,
+      "learning_rate": 1.4595846919512614e-05,
+      "loss": 0.5558,
+      "step": 5984
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.616376876831055,
+      "learning_rate": 1.4593272696070019e-05,
+      "loss": 0.6387,
+      "step": 5985
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.63627815246582,
+      "learning_rate": 1.4590698472627426e-05,
+      "loss": 0.6726,
+      "step": 5986
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.850713729858398,
+      "learning_rate": 1.458812424918483e-05,
+      "loss": 0.6863,
+      "step": 5987
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.737812995910645,
+      "learning_rate": 1.4585550025742236e-05,
+      "loss": 0.3511,
+      "step": 5988
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.919577598571777,
+      "learning_rate": 1.4582975802299639e-05,
+      "loss": 0.4189,
+      "step": 5989
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.109214782714844,
+      "learning_rate": 1.4580401578857044e-05,
+      "loss": 0.5978,
+      "step": 5990
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.94290542602539,
+      "learning_rate": 1.457782735541445e-05,
+      "loss": 0.5975,
+      "step": 5991
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.736477851867676,
+      "learning_rate": 1.4575253131971855e-05,
+      "loss": 0.5522,
+      "step": 5992
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.462236404418945,
+      "learning_rate": 1.457267890852926e-05,
+      "loss": 0.7905,
+      "step": 5993
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.172720909118652,
+      "learning_rate": 1.4570104685086665e-05,
+      "loss": 0.6194,
+      "step": 5994
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.092374801635742,
+      "learning_rate": 1.4567530461644072e-05,
+      "loss": 0.5071,
+      "step": 5995
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.301154136657715,
+      "learning_rate": 1.4564956238201477e-05,
+      "loss": 0.5913,
+      "step": 5996
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.991229057312012,
+      "learning_rate": 1.4562382014758882e-05,
+      "loss": 0.4657,
+      "step": 5997
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.99020004272461,
+      "learning_rate": 1.4559807791316287e-05,
+      "loss": 0.4323,
+      "step": 5998
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.797056198120117,
+      "learning_rate": 1.455723356787369e-05,
+      "loss": 0.6752,
+      "step": 5999
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.071431159973145,
+      "learning_rate": 1.4554659344431097e-05,
+      "loss": 0.3541,
+      "step": 6000
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.672924995422363,
+      "learning_rate": 1.4552085120988502e-05,
+      "loss": 0.631,
+      "step": 6001
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.877246856689453,
+      "learning_rate": 1.4549510897545907e-05,
+      "loss": 0.485,
+      "step": 6002
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.03546142578125,
+      "learning_rate": 1.4546936674103312e-05,
+      "loss": 0.7234,
+      "step": 6003
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.23678970336914,
+      "learning_rate": 1.4544362450660717e-05,
+      "loss": 0.5195,
+      "step": 6004
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.419454574584961,
+      "learning_rate": 1.4541788227218124e-05,
+      "loss": 0.5899,
+      "step": 6005
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.912901878356934,
+      "learning_rate": 1.4539214003775529e-05,
+      "loss": 0.5016,
+      "step": 6006
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.926216125488281,
+      "learning_rate": 1.4536639780332934e-05,
+      "loss": 0.5323,
+      "step": 6007
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.15507984161377,
+      "learning_rate": 1.4534065556890339e-05,
+      "loss": 0.5297,
+      "step": 6008
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.702319145202637,
+      "learning_rate": 1.4531491333447744e-05,
+      "loss": 0.5739,
+      "step": 6009
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.169788360595703,
+      "learning_rate": 1.4528917110005149e-05,
+      "loss": 0.675,
+      "step": 6010
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.871467590332031,
+      "learning_rate": 1.4526342886562554e-05,
+      "loss": 0.4943,
+      "step": 6011
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.006232261657715,
+      "learning_rate": 1.4523768663119959e-05,
+      "loss": 0.4178,
+      "step": 6012
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.051650047302246,
+      "learning_rate": 1.4521194439677364e-05,
+      "loss": 0.7047,
+      "step": 6013
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.217045783996582,
+      "learning_rate": 1.451862021623477e-05,
+      "loss": 0.8165,
+      "step": 6014
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.112601280212402,
+      "learning_rate": 1.4516045992792175e-05,
+      "loss": 0.6061,
+      "step": 6015
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.458157539367676,
+      "learning_rate": 1.451347176934958e-05,
+      "loss": 0.6302,
+      "step": 6016
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.235051155090332,
+      "learning_rate": 1.4510897545906985e-05,
+      "loss": 0.4292,
+      "step": 6017
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.034002304077148,
+      "learning_rate": 1.450832332246439e-05,
+      "loss": 0.4661,
+      "step": 6018
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.086384773254395,
+      "learning_rate": 1.4505749099021797e-05,
+      "loss": 0.6491,
+      "step": 6019
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.544063568115234,
+      "learning_rate": 1.45031748755792e-05,
+      "loss": 0.6637,
+      "step": 6020
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.971297264099121,
+      "learning_rate": 1.4500600652136605e-05,
+      "loss": 0.4539,
+      "step": 6021
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.303975105285645,
+      "learning_rate": 1.449802642869401e-05,
+      "loss": 0.6362,
+      "step": 6022
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.662263870239258,
+      "learning_rate": 1.4495452205251415e-05,
+      "loss": 0.8009,
+      "step": 6023
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.112136840820312,
+      "learning_rate": 1.4492877981808822e-05,
+      "loss": 0.738,
+      "step": 6024
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.446643829345703,
+      "learning_rate": 1.4490303758366227e-05,
+      "loss": 0.7352,
+      "step": 6025
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.696676254272461,
+      "learning_rate": 1.4487729534923632e-05,
+      "loss": 0.5278,
+      "step": 6026
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.39923095703125,
+      "learning_rate": 1.4485155311481037e-05,
+      "loss": 0.5024,
+      "step": 6027
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.176897048950195,
+      "learning_rate": 1.4482581088038442e-05,
+      "loss": 0.573,
+      "step": 6028
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.741162300109863,
+      "learning_rate": 1.4480006864595848e-05,
+      "loss": 0.6059,
+      "step": 6029
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.605361938476562,
+      "learning_rate": 1.4477432641153252e-05,
+      "loss": 0.4713,
+      "step": 6030
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.7548956871032715,
+      "learning_rate": 1.4474858417710657e-05,
+      "loss": 0.3838,
+      "step": 6031
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.093345642089844,
+      "learning_rate": 1.4472284194268062e-05,
+      "loss": 0.3958,
+      "step": 6032
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.763676643371582,
+      "learning_rate": 1.4469709970825468e-05,
+      "loss": 0.4929,
+      "step": 6033
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.30533504486084,
+      "learning_rate": 1.4467135747382873e-05,
+      "loss": 0.5723,
+      "step": 6034
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.449240684509277,
+      "learning_rate": 1.4464561523940278e-05,
+      "loss": 0.5791,
+      "step": 6035
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.579524040222168,
+      "learning_rate": 1.4461987300497683e-05,
+      "loss": 0.4807,
+      "step": 6036
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.388040542602539,
+      "learning_rate": 1.4459413077055088e-05,
+      "loss": 0.4591,
+      "step": 6037
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.479193687438965,
+      "learning_rate": 1.4456838853612495e-05,
+      "loss": 0.5965,
+      "step": 6038
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.420098304748535,
+      "learning_rate": 1.44542646301699e-05,
+      "loss": 0.6916,
+      "step": 6039
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.806793212890625,
+      "learning_rate": 1.4451690406727305e-05,
+      "loss": 0.854,
+      "step": 6040
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.206016540527344,
+      "learning_rate": 1.4449116183284708e-05,
+      "loss": 0.4748,
+      "step": 6041
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.494976043701172,
+      "learning_rate": 1.4446541959842113e-05,
+      "loss": 0.4301,
+      "step": 6042
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.136125564575195,
+      "learning_rate": 1.444396773639952e-05,
+      "loss": 0.9049,
+      "step": 6043
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.781466484069824,
+      "learning_rate": 1.4441393512956925e-05,
+      "loss": 0.6523,
+      "step": 6044
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.593643188476562,
+      "learning_rate": 1.443881928951433e-05,
+      "loss": 0.4749,
+      "step": 6045
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.517271995544434,
+      "learning_rate": 1.4436245066071735e-05,
+      "loss": 0.7917,
+      "step": 6046
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.094876289367676,
+      "learning_rate": 1.4433670842629142e-05,
+      "loss": 0.4401,
+      "step": 6047
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.035802841186523,
+      "learning_rate": 1.4431096619186547e-05,
+      "loss": 0.6043,
+      "step": 6048
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.315948486328125,
+      "learning_rate": 1.4428522395743952e-05,
+      "loss": 0.6069,
+      "step": 6049
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.189122200012207,
+      "learning_rate": 1.4425948172301357e-05,
+      "loss": 0.6013,
+      "step": 6050
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.863465309143066,
+      "learning_rate": 1.442337394885876e-05,
+      "loss": 0.7592,
+      "step": 6051
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.783334732055664,
+      "learning_rate": 1.4420799725416167e-05,
+      "loss": 0.512,
+      "step": 6052
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.720061302185059,
+      "learning_rate": 1.4418225501973572e-05,
+      "loss": 0.6549,
+      "step": 6053
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.834259033203125,
+      "learning_rate": 1.4415651278530977e-05,
+      "loss": 0.5887,
+      "step": 6054
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.825617790222168,
+      "learning_rate": 1.4413077055088382e-05,
+      "loss": 0.5317,
+      "step": 6055
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.527396202087402,
+      "learning_rate": 1.4410502831645787e-05,
+      "loss": 0.5497,
+      "step": 6056
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.533768653869629,
+      "learning_rate": 1.4407928608203193e-05,
+      "loss": 0.3781,
+      "step": 6057
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.703971862792969,
+      "learning_rate": 1.4405354384760598e-05,
+      "loss": 0.5077,
+      "step": 6058
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.134206771850586,
+      "learning_rate": 1.4402780161318003e-05,
+      "loss": 0.608,
+      "step": 6059
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.357513427734375,
+      "learning_rate": 1.4400205937875408e-05,
+      "loss": 0.6031,
+      "step": 6060
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.609781265258789,
+      "learning_rate": 1.4397631714432813e-05,
+      "loss": 0.3727,
+      "step": 6061
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.689179420471191,
+      "learning_rate": 1.4395057490990218e-05,
+      "loss": 0.5953,
+      "step": 6062
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.445324897766113,
+      "learning_rate": 1.4392483267547623e-05,
+      "loss": 0.5379,
+      "step": 6063
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.921080589294434,
+      "learning_rate": 1.4389909044105028e-05,
+      "loss": 0.5976,
+      "step": 6064
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.612894058227539,
+      "learning_rate": 1.4387334820662433e-05,
+      "loss": 0.5477,
+      "step": 6065
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.109848976135254,
+      "learning_rate": 1.438476059721984e-05,
+      "loss": 0.5091,
+      "step": 6066
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.760492324829102,
+      "learning_rate": 1.4382186373777245e-05,
+      "loss": 0.5526,
+      "step": 6067
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.640450954437256,
+      "learning_rate": 1.437961215033465e-05,
+      "loss": 0.3258,
+      "step": 6068
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.686336517333984,
+      "learning_rate": 1.4377037926892055e-05,
+      "loss": 0.7677,
+      "step": 6069
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.44484806060791,
+      "learning_rate": 1.437446370344946e-05,
+      "loss": 0.4966,
+      "step": 6070
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.284594535827637,
+      "learning_rate": 1.4371889480006866e-05,
+      "loss": 0.501,
+      "step": 6071
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.47825813293457,
+      "learning_rate": 1.436931525656427e-05,
+      "loss": 0.6583,
+      "step": 6072
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.232020378112793,
+      "learning_rate": 1.4366741033121675e-05,
+      "loss": 0.5751,
+      "step": 6073
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.311654090881348,
+      "learning_rate": 1.436416680967908e-05,
+      "loss": 0.6769,
+      "step": 6074
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.717187881469727,
+      "learning_rate": 1.4361592586236485e-05,
+      "loss": 0.5025,
+      "step": 6075
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.880813598632812,
+      "learning_rate": 1.4359018362793891e-05,
+      "loss": 0.6143,
+      "step": 6076
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.355393409729004,
+      "learning_rate": 1.4356444139351296e-05,
+      "loss": 0.6914,
+      "step": 6077
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.87582778930664,
+      "learning_rate": 1.4353869915908701e-05,
+      "loss": 0.5311,
+      "step": 6078
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.29086971282959,
+      "learning_rate": 1.4351295692466106e-05,
+      "loss": 0.5469,
+      "step": 6079
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.28413200378418,
+      "learning_rate": 1.4348721469023511e-05,
+      "loss": 0.4143,
+      "step": 6080
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.597415924072266,
+      "learning_rate": 1.4346147245580918e-05,
+      "loss": 0.6202,
+      "step": 6081
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.337462425231934,
+      "learning_rate": 1.4343573022138321e-05,
+      "loss": 0.4907,
+      "step": 6082
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.425405502319336,
+      "learning_rate": 1.4340998798695726e-05,
+      "loss": 0.343,
+      "step": 6083
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.436531066894531,
+      "learning_rate": 1.4338424575253131e-05,
+      "loss": 0.6264,
+      "step": 6084
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.24310302734375,
+      "learning_rate": 1.4335850351810538e-05,
+      "loss": 0.6385,
+      "step": 6085
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.942517280578613,
+      "learning_rate": 1.4333276128367943e-05,
+      "loss": 0.473,
+      "step": 6086
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.65208911895752,
+      "learning_rate": 1.4330701904925348e-05,
+      "loss": 0.5932,
+      "step": 6087
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.827574729919434,
+      "learning_rate": 1.4328127681482753e-05,
+      "loss": 0.4508,
+      "step": 6088
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.024276733398438,
+      "learning_rate": 1.4325553458040158e-05,
+      "loss": 0.7384,
+      "step": 6089
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.000886917114258,
+      "learning_rate": 1.4322979234597565e-05,
+      "loss": 0.543,
+      "step": 6090
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.903984069824219,
+      "learning_rate": 1.432040501115497e-05,
+      "loss": 0.703,
+      "step": 6091
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.271925926208496,
+      "learning_rate": 1.4317830787712375e-05,
+      "loss": 0.5366,
+      "step": 6092
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.144278526306152,
+      "learning_rate": 1.4315256564269778e-05,
+      "loss": 0.5025,
+      "step": 6093
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.94912338256836,
+      "learning_rate": 1.4312682340827183e-05,
+      "loss": 0.5062,
+      "step": 6094
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.116493225097656,
+      "learning_rate": 1.431010811738459e-05,
+      "loss": 0.4661,
+      "step": 6095
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.658554077148438,
+      "learning_rate": 1.4307533893941994e-05,
+      "loss": 0.5392,
+      "step": 6096
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.043598175048828,
+      "learning_rate": 1.43049596704994e-05,
+      "loss": 0.6206,
+      "step": 6097
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.47064208984375,
+      "learning_rate": 1.4302385447056804e-05,
+      "loss": 0.7444,
+      "step": 6098
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.273091316223145,
+      "learning_rate": 1.4299811223614211e-05,
+      "loss": 0.5602,
+      "step": 6099
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.650313377380371,
+      "learning_rate": 1.4297237000171616e-05,
+      "loss": 0.6733,
+      "step": 6100
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.835964679718018,
+      "learning_rate": 1.4294662776729021e-05,
+      "loss": 0.4004,
+      "step": 6101
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.33560562133789,
+      "learning_rate": 1.4292088553286426e-05,
+      "loss": 0.6252,
+      "step": 6102
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.087122917175293,
+      "learning_rate": 1.428951432984383e-05,
+      "loss": 0.4333,
+      "step": 6103
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.257527351379395,
+      "learning_rate": 1.4286940106401236e-05,
+      "loss": 0.4685,
+      "step": 6104
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.58927059173584,
+      "learning_rate": 1.4284365882958641e-05,
+      "loss": 0.4967,
+      "step": 6105
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.252796173095703,
+      "learning_rate": 1.4281791659516046e-05,
+      "loss": 0.5997,
+      "step": 6106
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.413355827331543,
+      "learning_rate": 1.4279217436073451e-05,
+      "loss": 0.3748,
+      "step": 6107
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.290119171142578,
+      "learning_rate": 1.4276643212630856e-05,
+      "loss": 0.6509,
+      "step": 6108
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.814628601074219,
+      "learning_rate": 1.4274068989188263e-05,
+      "loss": 0.4193,
+      "step": 6109
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.923113822937012,
+      "learning_rate": 1.4271494765745668e-05,
+      "loss": 0.6374,
+      "step": 6110
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.146993637084961,
+      "learning_rate": 1.4268920542303073e-05,
+      "loss": 0.3913,
+      "step": 6111
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.459632873535156,
+      "learning_rate": 1.4266346318860478e-05,
+      "loss": 0.7129,
+      "step": 6112
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.424100875854492,
+      "learning_rate": 1.4263772095417883e-05,
+      "loss": 0.4261,
+      "step": 6113
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.101913452148438,
+      "learning_rate": 1.4261197871975288e-05,
+      "loss": 0.6028,
+      "step": 6114
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.185338973999023,
+      "learning_rate": 1.4258623648532693e-05,
+      "loss": 0.5468,
+      "step": 6115
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.860051155090332,
+      "learning_rate": 1.4256049425090098e-05,
+      "loss": 0.5122,
+      "step": 6116
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.243034362792969,
+      "learning_rate": 1.4253475201647503e-05,
+      "loss": 0.5197,
+      "step": 6117
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.012742042541504,
+      "learning_rate": 1.425090097820491e-05,
+      "loss": 0.5302,
+      "step": 6118
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.1295804977417,
+      "learning_rate": 1.4248326754762314e-05,
+      "loss": 0.4654,
+      "step": 6119
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.8541388511657715,
+      "learning_rate": 1.424575253131972e-05,
+      "loss": 0.3855,
+      "step": 6120
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.729656219482422,
+      "learning_rate": 1.4243178307877124e-05,
+      "loss": 0.5408,
+      "step": 6121
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.455890655517578,
+      "learning_rate": 1.424060408443453e-05,
+      "loss": 0.5054,
+      "step": 6122
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.174029350280762,
+      "learning_rate": 1.4238029860991936e-05,
+      "loss": 0.5078,
+      "step": 6123
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.40572738647461,
+      "learning_rate": 1.423545563754934e-05,
+      "loss": 0.7425,
+      "step": 6124
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.948651313781738,
+      "learning_rate": 1.4232881414106744e-05,
+      "loss": 0.7963,
+      "step": 6125
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.797237396240234,
+      "learning_rate": 1.423030719066415e-05,
+      "loss": 0.5415,
+      "step": 6126
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.037702560424805,
+      "learning_rate": 1.4227732967221554e-05,
+      "loss": 0.4369,
+      "step": 6127
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.486723899841309,
+      "learning_rate": 1.4225158743778961e-05,
+      "loss": 0.5017,
+      "step": 6128
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.453106880187988,
+      "learning_rate": 1.4222584520336366e-05,
+      "loss": 0.5797,
+      "step": 6129
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.221585273742676,
+      "learning_rate": 1.422001029689377e-05,
+      "loss": 0.4769,
+      "step": 6130
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.47728443145752,
+      "learning_rate": 1.4217436073451176e-05,
+      "loss": 0.6353,
+      "step": 6131
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.76475715637207,
+      "learning_rate": 1.421486185000858e-05,
+      "loss": 0.5109,
+      "step": 6132
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.694698333740234,
+      "learning_rate": 1.4212287626565987e-05,
+      "loss": 0.6959,
+      "step": 6133
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.754035949707031,
+      "learning_rate": 1.420971340312339e-05,
+      "loss": 0.855,
+      "step": 6134
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.095348358154297,
+      "learning_rate": 1.4207139179680796e-05,
+      "loss": 0.1835,
+      "step": 6135
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.529303550720215,
+      "learning_rate": 1.42045649562382e-05,
+      "loss": 0.4942,
+      "step": 6136
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.631980895996094,
+      "learning_rate": 1.4201990732795607e-05,
+      "loss": 0.7518,
+      "step": 6137
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.752338409423828,
+      "learning_rate": 1.4199416509353012e-05,
+      "loss": 0.7107,
+      "step": 6138
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.647884368896484,
+      "learning_rate": 1.4196842285910417e-05,
+      "loss": 0.6206,
+      "step": 6139
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.637974739074707,
+      "learning_rate": 1.4194268062467822e-05,
+      "loss": 0.5882,
+      "step": 6140
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.008574485778809,
+      "learning_rate": 1.4191693839025227e-05,
+      "loss": 0.6725,
+      "step": 6141
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.038464546203613,
+      "learning_rate": 1.4189119615582634e-05,
+      "loss": 0.5212,
+      "step": 6142
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.971981048583984,
+      "learning_rate": 1.4186545392140039e-05,
+      "loss": 0.6147,
+      "step": 6143
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.331040382385254,
+      "learning_rate": 1.4183971168697444e-05,
+      "loss": 0.5533,
+      "step": 6144
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.561538696289062,
+      "learning_rate": 1.4181396945254847e-05,
+      "loss": 0.4664,
+      "step": 6145
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.650005340576172,
+      "learning_rate": 1.4178822721812252e-05,
+      "loss": 0.6292,
+      "step": 6146
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.890765190124512,
+      "learning_rate": 1.4176248498369659e-05,
+      "loss": 0.5113,
+      "step": 6147
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.604499816894531,
+      "learning_rate": 1.4173674274927064e-05,
+      "loss": 0.5598,
+      "step": 6148
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.501017570495605,
+      "learning_rate": 1.4171100051484469e-05,
+      "loss": 0.7128,
+      "step": 6149
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.226709365844727,
+      "learning_rate": 1.4168525828041874e-05,
+      "loss": 0.6862,
+      "step": 6150
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.615044593811035,
+      "learning_rate": 1.416595160459928e-05,
+      "loss": 0.4897,
+      "step": 6151
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.787710189819336,
+      "learning_rate": 1.4163377381156686e-05,
+      "loss": 0.5188,
+      "step": 6152
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.707430839538574,
+      "learning_rate": 1.416080315771409e-05,
+      "loss": 0.6976,
+      "step": 6153
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.319376945495605,
+      "learning_rate": 1.4158228934271496e-05,
+      "loss": 0.7427,
+      "step": 6154
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.069814682006836,
+      "learning_rate": 1.4155654710828899e-05,
+      "loss": 0.6196,
+      "step": 6155
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.175179481506348,
+      "learning_rate": 1.4153080487386306e-05,
+      "loss": 0.5862,
+      "step": 6156
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.734851837158203,
+      "learning_rate": 1.415050626394371e-05,
+      "loss": 0.8946,
+      "step": 6157
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.9085516929626465,
+      "learning_rate": 1.4147932040501116e-05,
+      "loss": 0.2874,
+      "step": 6158
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.46718692779541,
+      "learning_rate": 1.414535781705852e-05,
+      "loss": 0.6351,
+      "step": 6159
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.667304992675781,
+      "learning_rate": 1.4142783593615926e-05,
+      "loss": 0.6053,
+      "step": 6160
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.37648868560791,
+      "learning_rate": 1.4140209370173332e-05,
+      "loss": 0.6371,
+      "step": 6161
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.953534126281738,
+      "learning_rate": 1.4137635146730737e-05,
+      "loss": 0.5669,
+      "step": 6162
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.759281158447266,
+      "learning_rate": 1.4135060923288142e-05,
+      "loss": 0.5766,
+      "step": 6163
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.923698425292969,
+      "learning_rate": 1.4132486699845547e-05,
+      "loss": 0.6697,
+      "step": 6164
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.941294193267822,
+      "learning_rate": 1.4129912476402952e-05,
+      "loss": 0.6458,
+      "step": 6165
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.728187561035156,
+      "learning_rate": 1.4127338252960357e-05,
+      "loss": 0.4643,
+      "step": 6166
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.354240417480469,
+      "learning_rate": 1.4124764029517762e-05,
+      "loss": 0.5838,
+      "step": 6167
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.264355659484863,
+      "learning_rate": 1.4122189806075167e-05,
+      "loss": 0.8504,
+      "step": 6168
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.871974468231201,
+      "learning_rate": 1.4119615582632572e-05,
+      "loss": 0.4049,
+      "step": 6169
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.0407075881958,
+      "learning_rate": 1.4117041359189979e-05,
+      "loss": 0.6596,
+      "step": 6170
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.660916328430176,
+      "learning_rate": 1.4114467135747384e-05,
+      "loss": 0.458,
+      "step": 6171
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.8273344039917,
+      "learning_rate": 1.4111892912304789e-05,
+      "loss": 0.3837,
+      "step": 6172
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.746609687805176,
+      "learning_rate": 1.4109318688862194e-05,
+      "loss": 0.5208,
+      "step": 6173
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.71379280090332,
+      "learning_rate": 1.4106744465419599e-05,
+      "loss": 0.6932,
+      "step": 6174
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.956168174743652,
+      "learning_rate": 1.4104170241977005e-05,
+      "loss": 0.8068,
+      "step": 6175
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.844505310058594,
+      "learning_rate": 1.4101596018534409e-05,
+      "loss": 0.7495,
+      "step": 6176
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.569572925567627,
+      "learning_rate": 1.4099021795091814e-05,
+      "loss": 0.414,
+      "step": 6177
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.275727272033691,
+      "learning_rate": 1.4096447571649219e-05,
+      "loss": 0.5681,
+      "step": 6178
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.412195205688477,
+      "learning_rate": 1.4093873348206624e-05,
+      "loss": 0.5247,
+      "step": 6179
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.642062187194824,
+      "learning_rate": 1.409129912476403e-05,
+      "loss": 0.5671,
+      "step": 6180
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.485599517822266,
+      "learning_rate": 1.4088724901321435e-05,
+      "loss": 0.5321,
+      "step": 6181
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.10834789276123,
+      "learning_rate": 1.408615067787884e-05,
+      "loss": 0.6384,
+      "step": 6182
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.994007110595703,
+      "learning_rate": 1.4083576454436245e-05,
+      "loss": 0.7008,
+      "step": 6183
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.241264343261719,
+      "learning_rate": 1.408100223099365e-05,
+      "loss": 0.552,
+      "step": 6184
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.289605140686035,
+      "learning_rate": 1.4078428007551057e-05,
+      "loss": 0.7744,
+      "step": 6185
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.377355575561523,
+      "learning_rate": 1.407585378410846e-05,
+      "loss": 0.5369,
+      "step": 6186
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.560219764709473,
+      "learning_rate": 1.4073279560665865e-05,
+      "loss": 0.5053,
+      "step": 6187
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.4827241897583,
+      "learning_rate": 1.407070533722327e-05,
+      "loss": 0.5675,
+      "step": 6188
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.145089149475098,
+      "learning_rate": 1.4068131113780677e-05,
+      "loss": 0.4695,
+      "step": 6189
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.306835174560547,
+      "learning_rate": 1.4065556890338082e-05,
+      "loss": 0.4757,
+      "step": 6190
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.71056842803955,
+      "learning_rate": 1.4062982666895487e-05,
+      "loss": 0.6799,
+      "step": 6191
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.993293762207031,
+      "learning_rate": 1.4060408443452892e-05,
+      "loss": 0.6636,
+      "step": 6192
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.135930061340332,
+      "learning_rate": 1.4057834220010297e-05,
+      "loss": 0.4951,
+      "step": 6193
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.61836051940918,
+      "learning_rate": 1.4055259996567704e-05,
+      "loss": 0.6253,
+      "step": 6194
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.176191329956055,
+      "learning_rate": 1.4052685773125109e-05,
+      "loss": 0.6785,
+      "step": 6195
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.15185546875,
+      "learning_rate": 1.4050111549682514e-05,
+      "loss": 0.6408,
+      "step": 6196
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.607077598571777,
+      "learning_rate": 1.4047537326239917e-05,
+      "loss": 0.6836,
+      "step": 6197
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.048965454101562,
+      "learning_rate": 1.4044963102797322e-05,
+      "loss": 0.5954,
+      "step": 6198
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.19587230682373,
+      "learning_rate": 1.4042388879354729e-05,
+      "loss": 0.4546,
+      "step": 6199
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.464704513549805,
+      "learning_rate": 1.4039814655912134e-05,
+      "loss": 0.441,
+      "step": 6200
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.671178817749023,
+      "learning_rate": 1.4037240432469538e-05,
+      "loss": 0.4576,
+      "step": 6201
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.676521301269531,
+      "learning_rate": 1.4034666209026943e-05,
+      "loss": 0.4821,
+      "step": 6202
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.109301567077637,
+      "learning_rate": 1.4032091985584348e-05,
+      "loss": 0.5541,
+      "step": 6203
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.648665428161621,
+      "learning_rate": 1.4029517762141755e-05,
+      "loss": 0.4771,
+      "step": 6204
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.39564323425293,
+      "learning_rate": 1.402694353869916e-05,
+      "loss": 0.8375,
+      "step": 6205
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.830465316772461,
+      "learning_rate": 1.4024369315256565e-05,
+      "loss": 0.444,
+      "step": 6206
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.146523475646973,
+      "learning_rate": 1.4021795091813968e-05,
+      "loss": 0.709,
+      "step": 6207
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.60805606842041,
+      "learning_rate": 1.4019220868371375e-05,
+      "loss": 0.6636,
+      "step": 6208
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.50390911102295,
+      "learning_rate": 1.401664664492878e-05,
+      "loss": 0.3645,
+      "step": 6209
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.364044189453125,
+      "learning_rate": 1.4014072421486185e-05,
+      "loss": 0.6259,
+      "step": 6210
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.997735977172852,
+      "learning_rate": 1.401149819804359e-05,
+      "loss": 0.8093,
+      "step": 6211
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.80439567565918,
+      "learning_rate": 1.4008923974600995e-05,
+      "loss": 0.6535,
+      "step": 6212
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.367583274841309,
+      "learning_rate": 1.4006349751158402e-05,
+      "loss": 0.4333,
+      "step": 6213
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.364331245422363,
+      "learning_rate": 1.4003775527715807e-05,
+      "loss": 0.6987,
+      "step": 6214
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.478676795959473,
+      "learning_rate": 1.4001201304273212e-05,
+      "loss": 0.4178,
+      "step": 6215
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.883241653442383,
+      "learning_rate": 1.3998627080830617e-05,
+      "loss": 0.8292,
+      "step": 6216
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.789076328277588,
+      "learning_rate": 1.3996052857388022e-05,
+      "loss": 0.459,
+      "step": 6217
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.890912055969238,
+      "learning_rate": 1.3993478633945427e-05,
+      "loss": 0.5262,
+      "step": 6218
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.356431007385254,
+      "learning_rate": 1.3990904410502832e-05,
+      "loss": 0.6706,
+      "step": 6219
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.556934356689453,
+      "learning_rate": 1.3988330187060237e-05,
+      "loss": 0.7136,
+      "step": 6220
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.441965103149414,
+      "learning_rate": 1.3985755963617642e-05,
+      "loss": 0.5053,
+      "step": 6221
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.535511016845703,
+      "learning_rate": 1.3983181740175048e-05,
+      "loss": 0.5035,
+      "step": 6222
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.299717903137207,
+      "learning_rate": 1.3980607516732453e-05,
+      "loss": 0.6058,
+      "step": 6223
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.213362693786621,
+      "learning_rate": 1.3978033293289858e-05,
+      "loss": 0.5974,
+      "step": 6224
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.583738327026367,
+      "learning_rate": 1.3975459069847263e-05,
+      "loss": 0.5529,
+      "step": 6225
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.918071746826172,
+      "learning_rate": 1.3972884846404668e-05,
+      "loss": 0.7726,
+      "step": 6226
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.898406028747559,
+      "learning_rate": 1.3970310622962075e-05,
+      "loss": 0.4654,
+      "step": 6227
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.304642677307129,
+      "learning_rate": 1.3967736399519478e-05,
+      "loss": 0.4431,
+      "step": 6228
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.091839790344238,
+      "learning_rate": 1.3965162176076883e-05,
+      "loss": 0.5877,
+      "step": 6229
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.9913911819458,
+      "learning_rate": 1.3962587952634288e-05,
+      "loss": 0.4119,
+      "step": 6230
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.420130729675293,
+      "learning_rate": 1.3960013729191693e-05,
+      "loss": 0.6257,
+      "step": 6231
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.855796813964844,
+      "learning_rate": 1.39574395057491e-05,
+      "loss": 0.515,
+      "step": 6232
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.3886079788208,
+      "learning_rate": 1.3954865282306505e-05,
+      "loss": 0.4054,
+      "step": 6233
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.868733406066895,
+      "learning_rate": 1.395229105886391e-05,
+      "loss": 0.5337,
+      "step": 6234
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.671782493591309,
+      "learning_rate": 1.3949716835421315e-05,
+      "loss": 0.422,
+      "step": 6235
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.012228965759277,
+      "learning_rate": 1.394714261197872e-05,
+      "loss": 0.4726,
+      "step": 6236
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.76025676727295,
+      "learning_rate": 1.3944568388536127e-05,
+      "loss": 0.3327,
+      "step": 6237
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.357230186462402,
+      "learning_rate": 1.394199416509353e-05,
+      "loss": 0.4904,
+      "step": 6238
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.895170211791992,
+      "learning_rate": 1.3939419941650935e-05,
+      "loss": 0.6192,
+      "step": 6239
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.209478378295898,
+      "learning_rate": 1.393684571820834e-05,
+      "loss": 0.5817,
+      "step": 6240
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.764213562011719,
+      "learning_rate": 1.3934271494765746e-05,
+      "loss": 0.6675,
+      "step": 6241
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.004423141479492,
+      "learning_rate": 1.3931697271323151e-05,
+      "loss": 0.517,
+      "step": 6242
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.49935245513916,
+      "learning_rate": 1.3929123047880556e-05,
+      "loss": 0.6067,
+      "step": 6243
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.47079086303711,
+      "learning_rate": 1.3926548824437961e-05,
+      "loss": 0.4759,
+      "step": 6244
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.959939002990723,
+      "learning_rate": 1.3923974600995366e-05,
+      "loss": 0.5148,
+      "step": 6245
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.511268615722656,
+      "learning_rate": 1.3921400377552773e-05,
+      "loss": 0.7078,
+      "step": 6246
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.832539558410645,
+      "learning_rate": 1.3918826154110178e-05,
+      "loss": 0.421,
+      "step": 6247
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.258830070495605,
+      "learning_rate": 1.3916251930667583e-05,
+      "loss": 0.6067,
+      "step": 6248
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.234213829040527,
+      "learning_rate": 1.3913677707224986e-05,
+      "loss": 0.5588,
+      "step": 6249
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.037192344665527,
+      "learning_rate": 1.3911103483782391e-05,
+      "loss": 0.5106,
+      "step": 6250
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.15066909790039,
+      "learning_rate": 1.3908529260339798e-05,
+      "loss": 0.561,
+      "step": 6251
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.844749450683594,
+      "learning_rate": 1.3905955036897203e-05,
+      "loss": 0.7623,
+      "step": 6252
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.8617582321167,
+      "learning_rate": 1.3903380813454608e-05,
+      "loss": 0.7299,
+      "step": 6253
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.919418334960938,
+      "learning_rate": 1.3900806590012013e-05,
+      "loss": 0.8348,
+      "step": 6254
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.06890869140625,
+      "learning_rate": 1.3898232366569418e-05,
+      "loss": 0.5835,
+      "step": 6255
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.210168838500977,
+      "learning_rate": 1.3895658143126825e-05,
+      "loss": 0.6658,
+      "step": 6256
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.673344612121582,
+      "learning_rate": 1.389308391968423e-05,
+      "loss": 0.5187,
+      "step": 6257
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.394471168518066,
+      "learning_rate": 1.3890509696241635e-05,
+      "loss": 0.6847,
+      "step": 6258
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.45179557800293,
+      "learning_rate": 1.3887935472799038e-05,
+      "loss": 0.565,
+      "step": 6259
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.54520034790039,
+      "learning_rate": 1.3885361249356445e-05,
+      "loss": 0.4987,
+      "step": 6260
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.653621673583984,
+      "learning_rate": 1.388278702591385e-05,
+      "loss": 0.5881,
+      "step": 6261
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.62857437133789,
+      "learning_rate": 1.3880212802471255e-05,
+      "loss": 0.479,
+      "step": 6262
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.425786018371582,
+      "learning_rate": 1.387763857902866e-05,
+      "loss": 0.5928,
+      "step": 6263
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.90549087524414,
+      "learning_rate": 1.3875064355586065e-05,
+      "loss": 0.474,
+      "step": 6264
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.314797401428223,
+      "learning_rate": 1.3872490132143471e-05,
+      "loss": 0.5046,
+      "step": 6265
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.694711208343506,
+      "learning_rate": 1.3869915908700876e-05,
+      "loss": 0.5418,
+      "step": 6266
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.304850578308105,
+      "learning_rate": 1.3867341685258281e-05,
+      "loss": 0.5628,
+      "step": 6267
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.094827651977539,
+      "learning_rate": 1.3864767461815686e-05,
+      "loss": 0.5786,
+      "step": 6268
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.304694175720215,
+      "learning_rate": 1.3862193238373091e-05,
+      "loss": 0.6707,
+      "step": 6269
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.792288780212402,
+      "learning_rate": 1.3859619014930496e-05,
+      "loss": 0.5119,
+      "step": 6270
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.740971565246582,
+      "learning_rate": 1.3857044791487901e-05,
+      "loss": 0.5572,
+      "step": 6271
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.444781303405762,
+      "learning_rate": 1.3854470568045306e-05,
+      "loss": 0.6394,
+      "step": 6272
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.442985534667969,
+      "learning_rate": 1.3851896344602711e-05,
+      "loss": 0.4735,
+      "step": 6273
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.584697723388672,
+      "learning_rate": 1.3849322121160118e-05,
+      "loss": 0.5799,
+      "step": 6274
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.112217903137207,
+      "learning_rate": 1.3846747897717523e-05,
+      "loss": 0.4647,
+      "step": 6275
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.636457443237305,
+      "learning_rate": 1.3844173674274928e-05,
+      "loss": 0.4893,
+      "step": 6276
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.850175857543945,
+      "learning_rate": 1.3841599450832333e-05,
+      "loss": 0.3756,
+      "step": 6277
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.449297904968262,
+      "learning_rate": 1.3839025227389738e-05,
+      "loss": 0.4627,
+      "step": 6278
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.737735748291016,
+      "learning_rate": 1.3836451003947144e-05,
+      "loss": 0.7851,
+      "step": 6279
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.187867164611816,
+      "learning_rate": 1.3833876780504548e-05,
+      "loss": 0.4972,
+      "step": 6280
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.601746559143066,
+      "learning_rate": 1.3831302557061953e-05,
+      "loss": 0.5256,
+      "step": 6281
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.227254867553711,
+      "learning_rate": 1.3828728333619358e-05,
+      "loss": 0.4618,
+      "step": 6282
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.60275411605835,
+      "learning_rate": 1.3826154110176763e-05,
+      "loss": 0.4363,
+      "step": 6283
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.972223281860352,
+      "learning_rate": 1.382357988673417e-05,
+      "loss": 0.4141,
+      "step": 6284
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.946643829345703,
+      "learning_rate": 1.3821005663291574e-05,
+      "loss": 0.5362,
+      "step": 6285
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.773284912109375,
+      "learning_rate": 1.381843143984898e-05,
+      "loss": 0.5777,
+      "step": 6286
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.536645889282227,
+      "learning_rate": 1.3815857216406384e-05,
+      "loss": 0.541,
+      "step": 6287
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.107416152954102,
+      "learning_rate": 1.381328299296379e-05,
+      "loss": 0.5172,
+      "step": 6288
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.692770004272461,
+      "learning_rate": 1.3810708769521196e-05,
+      "loss": 0.6203,
+      "step": 6289
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.433701515197754,
+      "learning_rate": 1.3808134546078601e-05,
+      "loss": 0.5653,
+      "step": 6290
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.546643257141113,
+      "learning_rate": 1.3805560322636004e-05,
+      "loss": 0.5105,
+      "step": 6291
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.81982135772705,
+      "learning_rate": 1.380298609919341e-05,
+      "loss": 0.6135,
+      "step": 6292
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.829192161560059,
+      "learning_rate": 1.3800411875750816e-05,
+      "loss": 0.4661,
+      "step": 6293
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.260122299194336,
+      "learning_rate": 1.3797837652308221e-05,
+      "loss": 0.6475,
+      "step": 6294
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.663241386413574,
+      "learning_rate": 1.3795263428865626e-05,
+      "loss": 0.4079,
+      "step": 6295
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.937297821044922,
+      "learning_rate": 1.3792689205423031e-05,
+      "loss": 0.6343,
+      "step": 6296
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.142282485961914,
+      "learning_rate": 1.3790114981980436e-05,
+      "loss": 0.4985,
+      "step": 6297
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.680667400360107,
+      "learning_rate": 1.3787540758537843e-05,
+      "loss": 0.2924,
+      "step": 6298
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.194073677062988,
+      "learning_rate": 1.3784966535095248e-05,
+      "loss": 0.4541,
+      "step": 6299
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.01657485961914,
+      "learning_rate": 1.3782392311652653e-05,
+      "loss": 0.6727,
+      "step": 6300
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.428760528564453,
+      "learning_rate": 1.3779818088210056e-05,
+      "loss": 0.6225,
+      "step": 6301
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.794927597045898,
+      "learning_rate": 1.377724386476746e-05,
+      "loss": 0.585,
+      "step": 6302
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.727364540100098,
+      "learning_rate": 1.3774669641324868e-05,
+      "loss": 0.5741,
+      "step": 6303
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.955038070678711,
+      "learning_rate": 1.3772095417882273e-05,
+      "loss": 0.5454,
+      "step": 6304
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.521261215209961,
+      "learning_rate": 1.3769521194439678e-05,
+      "loss": 0.5624,
+      "step": 6305
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.289969444274902,
+      "learning_rate": 1.3766946970997082e-05,
+      "loss": 0.7043,
+      "step": 6306
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.969691276550293,
+      "learning_rate": 1.3764372747554487e-05,
+      "loss": 0.5271,
+      "step": 6307
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.186813354492188,
+      "learning_rate": 1.3761798524111894e-05,
+      "loss": 0.4161,
+      "step": 6308
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.47191047668457,
+      "learning_rate": 1.3759224300669299e-05,
+      "loss": 0.6287,
+      "step": 6309
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.426264762878418,
+      "learning_rate": 1.3756650077226704e-05,
+      "loss": 0.5705,
+      "step": 6310
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.465685844421387,
+      "learning_rate": 1.3754075853784107e-05,
+      "loss": 0.6473,
+      "step": 6311
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.629545211791992,
+      "learning_rate": 1.3751501630341514e-05,
+      "loss": 0.5713,
+      "step": 6312
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.682764053344727,
+      "learning_rate": 1.3748927406898919e-05,
+      "loss": 0.6149,
+      "step": 6313
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.982463836669922,
+      "learning_rate": 1.3746353183456324e-05,
+      "loss": 0.7517,
+      "step": 6314
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.972879409790039,
+      "learning_rate": 1.3743778960013729e-05,
+      "loss": 0.7649,
+      "step": 6315
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.674260139465332,
+      "learning_rate": 1.3741204736571134e-05,
+      "loss": 0.412,
+      "step": 6316
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.33468246459961,
+      "learning_rate": 1.373863051312854e-05,
+      "loss": 0.4099,
+      "step": 6317
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.371768951416016,
+      "learning_rate": 1.3736056289685946e-05,
+      "loss": 0.9034,
+      "step": 6318
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.01196575164795,
+      "learning_rate": 1.373348206624335e-05,
+      "loss": 0.5516,
+      "step": 6319
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.978898525238037,
+      "learning_rate": 1.3730907842800756e-05,
+      "loss": 0.4562,
+      "step": 6320
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.977396488189697,
+      "learning_rate": 1.372833361935816e-05,
+      "loss": 0.5025,
+      "step": 6321
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.403948783874512,
+      "learning_rate": 1.3725759395915566e-05,
+      "loss": 0.5199,
+      "step": 6322
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.480254173278809,
+      "learning_rate": 1.372318517247297e-05,
+      "loss": 0.6569,
+      "step": 6323
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.804694652557373,
+      "learning_rate": 1.3720610949030376e-05,
+      "loss": 0.4094,
+      "step": 6324
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.974738597869873,
+      "learning_rate": 1.371803672558778e-05,
+      "loss": 0.577,
+      "step": 6325
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.148778915405273,
+      "learning_rate": 1.3715462502145187e-05,
+      "loss": 0.6078,
+      "step": 6326
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.429791450500488,
+      "learning_rate": 1.3712888278702592e-05,
+      "loss": 0.423,
+      "step": 6327
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.341646194458008,
+      "learning_rate": 1.3710314055259997e-05,
+      "loss": 0.7242,
+      "step": 6328
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.757133483886719,
+      "learning_rate": 1.3707739831817402e-05,
+      "loss": 0.3231,
+      "step": 6329
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.56751823425293,
+      "learning_rate": 1.3705165608374807e-05,
+      "loss": 0.716,
+      "step": 6330
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.949222564697266,
+      "learning_rate": 1.3702591384932214e-05,
+      "loss": 0.5034,
+      "step": 6331
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.675853729248047,
+      "learning_rate": 1.3700017161489617e-05,
+      "loss": 0.5055,
+      "step": 6332
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.437301635742188,
+      "learning_rate": 1.3697442938047022e-05,
+      "loss": 0.6229,
+      "step": 6333
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.74101448059082,
+      "learning_rate": 1.3694868714604427e-05,
+      "loss": 0.367,
+      "step": 6334
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.52475643157959,
+      "learning_rate": 1.3692294491161832e-05,
+      "loss": 0.4028,
+      "step": 6335
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.165909767150879,
+      "learning_rate": 1.3689720267719239e-05,
+      "loss": 0.4518,
+      "step": 6336
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.376535415649414,
+      "learning_rate": 1.3687146044276644e-05,
+      "loss": 0.7996,
+      "step": 6337
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.022297859191895,
+      "learning_rate": 1.3684571820834049e-05,
+      "loss": 0.5165,
+      "step": 6338
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.504289627075195,
+      "learning_rate": 1.3681997597391454e-05,
+      "loss": 0.6616,
+      "step": 6339
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.874673843383789,
+      "learning_rate": 1.3679423373948859e-05,
+      "loss": 0.5026,
+      "step": 6340
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.333306312561035,
+      "learning_rate": 1.3676849150506266e-05,
+      "loss": 0.4877,
+      "step": 6341
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.488816261291504,
+      "learning_rate": 1.367427492706367e-05,
+      "loss": 0.44,
+      "step": 6342
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.746158599853516,
+      "learning_rate": 1.3671700703621074e-05,
+      "loss": 0.5407,
+      "step": 6343
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.122489929199219,
+      "learning_rate": 1.3669126480178479e-05,
+      "loss": 0.5318,
+      "step": 6344
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.632287979125977,
+      "learning_rate": 1.3666552256735885e-05,
+      "loss": 0.4774,
+      "step": 6345
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.978248596191406,
+      "learning_rate": 1.366397803329329e-05,
+      "loss": 0.5511,
+      "step": 6346
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.998766899108887,
+      "learning_rate": 1.3661403809850695e-05,
+      "loss": 0.4407,
+      "step": 6347
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.10408878326416,
+      "learning_rate": 1.36588295864081e-05,
+      "loss": 0.4636,
+      "step": 6348
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.009697914123535,
+      "learning_rate": 1.3656255362965505e-05,
+      "loss": 0.3376,
+      "step": 6349
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.989333152770996,
+      "learning_rate": 1.3653681139522912e-05,
+      "loss": 0.4242,
+      "step": 6350
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.495859146118164,
+      "learning_rate": 1.3651106916080317e-05,
+      "loss": 0.4429,
+      "step": 6351
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.282435417175293,
+      "learning_rate": 1.3648532692637722e-05,
+      "loss": 0.4659,
+      "step": 6352
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.34176254272461,
+      "learning_rate": 1.3645958469195125e-05,
+      "loss": 0.4175,
+      "step": 6353
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.509920120239258,
+      "learning_rate": 1.364338424575253e-05,
+      "loss": 0.5349,
+      "step": 6354
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.715743064880371,
+      "learning_rate": 1.3640810022309937e-05,
+      "loss": 0.499,
+      "step": 6355
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.126643180847168,
+      "learning_rate": 1.3638235798867342e-05,
+      "loss": 0.6813,
+      "step": 6356
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.739013671875,
+      "learning_rate": 1.3635661575424747e-05,
+      "loss": 0.5275,
+      "step": 6357
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.351797103881836,
+      "learning_rate": 1.3633087351982152e-05,
+      "loss": 0.4292,
+      "step": 6358
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.987312316894531,
+      "learning_rate": 1.3630513128539557e-05,
+      "loss": 0.6248,
+      "step": 6359
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.425827026367188,
+      "learning_rate": 1.3627938905096964e-05,
+      "loss": 0.4033,
+      "step": 6360
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.433403015136719,
+      "learning_rate": 1.3625364681654369e-05,
+      "loss": 0.547,
+      "step": 6361
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.584619522094727,
+      "learning_rate": 1.3622790458211774e-05,
+      "loss": 0.6028,
+      "step": 6362
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.98801040649414,
+      "learning_rate": 1.3620216234769177e-05,
+      "loss": 0.6658,
+      "step": 6363
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.8985595703125,
+      "learning_rate": 1.3617642011326584e-05,
+      "loss": 0.4547,
+      "step": 6364
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.857511520385742,
+      "learning_rate": 1.3615067787883989e-05,
+      "loss": 0.518,
+      "step": 6365
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.525308609008789,
+      "learning_rate": 1.3612493564441394e-05,
+      "loss": 0.4712,
+      "step": 6366
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.426017761230469,
+      "learning_rate": 1.3609919340998799e-05,
+      "loss": 0.6713,
+      "step": 6367
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.638419151306152,
+      "learning_rate": 1.3607345117556204e-05,
+      "loss": 0.5222,
+      "step": 6368
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.876932144165039,
+      "learning_rate": 1.360477089411361e-05,
+      "loss": 0.5994,
+      "step": 6369
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.331613540649414,
+      "learning_rate": 1.3602196670671015e-05,
+      "loss": 0.83,
+      "step": 6370
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.791149139404297,
+      "learning_rate": 1.359962244722842e-05,
+      "loss": 0.5903,
+      "step": 6371
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.320500373840332,
+      "learning_rate": 1.3597048223785825e-05,
+      "loss": 0.7425,
+      "step": 6372
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.020678520202637,
+      "learning_rate": 1.359447400034323e-05,
+      "loss": 0.6261,
+      "step": 6373
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.463134288787842,
+      "learning_rate": 1.3591899776900635e-05,
+      "loss": 0.3939,
+      "step": 6374
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.661871910095215,
+      "learning_rate": 1.358932555345804e-05,
+      "loss": 0.4429,
+      "step": 6375
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.510074615478516,
+      "learning_rate": 1.3586751330015445e-05,
+      "loss": 0.5624,
+      "step": 6376
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.635680198669434,
+      "learning_rate": 1.358417710657285e-05,
+      "loss": 0.5875,
+      "step": 6377
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.059264183044434,
+      "learning_rate": 1.3581602883130255e-05,
+      "loss": 0.4513,
+      "step": 6378
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.257065773010254,
+      "learning_rate": 1.3579028659687662e-05,
+      "loss": 0.4845,
+      "step": 6379
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.997725486755371,
+      "learning_rate": 1.3576454436245067e-05,
+      "loss": 0.5778,
+      "step": 6380
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.622173309326172,
+      "learning_rate": 1.3573880212802472e-05,
+      "loss": 0.641,
+      "step": 6381
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.398406982421875,
+      "learning_rate": 1.3571305989359877e-05,
+      "loss": 0.4863,
+      "step": 6382
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.99822998046875,
+      "learning_rate": 1.3568731765917283e-05,
+      "loss": 0.91,
+      "step": 6383
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.757206916809082,
+      "learning_rate": 1.3566157542474687e-05,
+      "loss": 0.5074,
+      "step": 6384
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.954026222229004,
+      "learning_rate": 1.3563583319032092e-05,
+      "loss": 0.4112,
+      "step": 6385
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.116982460021973,
+      "learning_rate": 1.3561009095589497e-05,
+      "loss": 0.3211,
+      "step": 6386
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.540571212768555,
+      "learning_rate": 1.3558434872146902e-05,
+      "loss": 0.5083,
+      "step": 6387
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.422941207885742,
+      "learning_rate": 1.3555860648704308e-05,
+      "loss": 0.6311,
+      "step": 6388
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.01567554473877,
+      "learning_rate": 1.3553286425261713e-05,
+      "loss": 0.6047,
+      "step": 6389
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.086312294006348,
+      "learning_rate": 1.3550712201819118e-05,
+      "loss": 0.5753,
+      "step": 6390
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.17856216430664,
+      "learning_rate": 1.3548137978376523e-05,
+      "loss": 0.696,
+      "step": 6391
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.220876693725586,
+      "learning_rate": 1.3545563754933928e-05,
+      "loss": 0.4285,
+      "step": 6392
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.963923454284668,
+      "learning_rate": 1.3542989531491335e-05,
+      "loss": 0.5472,
+      "step": 6393
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.718011856079102,
+      "learning_rate": 1.354041530804874e-05,
+      "loss": 0.5005,
+      "step": 6394
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.599753379821777,
+      "learning_rate": 1.3537841084606143e-05,
+      "loss": 0.551,
+      "step": 6395
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.1625337600708,
+      "learning_rate": 1.3535266861163548e-05,
+      "loss": 0.5666,
+      "step": 6396
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.343599319458008,
+      "learning_rate": 1.3532692637720955e-05,
+      "loss": 0.4813,
+      "step": 6397
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.8057279586792,
+      "learning_rate": 1.353011841427836e-05,
+      "loss": 0.5189,
+      "step": 6398
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.72898006439209,
+      "learning_rate": 1.3527544190835765e-05,
+      "loss": 0.7332,
+      "step": 6399
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.966325759887695,
+      "learning_rate": 1.352496996739317e-05,
+      "loss": 0.3484,
+      "step": 6400
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.623270988464355,
+      "learning_rate": 1.3522395743950575e-05,
+      "loss": 0.4777,
+      "step": 6401
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.14462947845459,
+      "learning_rate": 1.3519821520507982e-05,
+      "loss": 0.5484,
+      "step": 6402
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.376391410827637,
+      "learning_rate": 1.3517247297065387e-05,
+      "loss": 0.7525,
+      "step": 6403
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.11588191986084,
+      "learning_rate": 1.3514673073622792e-05,
+      "loss": 0.6512,
+      "step": 6404
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.11837387084961,
+      "learning_rate": 1.3512098850180195e-05,
+      "loss": 0.4214,
+      "step": 6405
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.572527885437012,
+      "learning_rate": 1.35095246267376e-05,
+      "loss": 0.4418,
+      "step": 6406
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.243064880371094,
+      "learning_rate": 1.3506950403295007e-05,
+      "loss": 0.3604,
+      "step": 6407
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.808772563934326,
+      "learning_rate": 1.3504376179852412e-05,
+      "loss": 0.4495,
+      "step": 6408
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.752381324768066,
+      "learning_rate": 1.3501801956409817e-05,
+      "loss": 0.6866,
+      "step": 6409
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.337677955627441,
+      "learning_rate": 1.3499227732967221e-05,
+      "loss": 0.5827,
+      "step": 6410
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.918295860290527,
+      "learning_rate": 1.3496653509524626e-05,
+      "loss": 0.427,
+      "step": 6411
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.244314193725586,
+      "learning_rate": 1.3494079286082033e-05,
+      "loss": 0.5604,
+      "step": 6412
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.718318939208984,
+      "learning_rate": 1.3491505062639438e-05,
+      "loss": 0.5631,
+      "step": 6413
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.136080741882324,
+      "learning_rate": 1.3488930839196843e-05,
+      "loss": 0.7477,
+      "step": 6414
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.44372844696045,
+      "learning_rate": 1.3486356615754246e-05,
+      "loss": 0.6166,
+      "step": 6415
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.982657432556152,
+      "learning_rate": 1.3483782392311653e-05,
+      "loss": 0.6096,
+      "step": 6416
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.544780731201172,
+      "learning_rate": 1.3481208168869058e-05,
+      "loss": 0.694,
+      "step": 6417
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.923746109008789,
+      "learning_rate": 1.3478633945426463e-05,
+      "loss": 0.7195,
+      "step": 6418
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.48173713684082,
+      "learning_rate": 1.3476059721983868e-05,
+      "loss": 0.5827,
+      "step": 6419
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.413176536560059,
+      "learning_rate": 1.3473485498541273e-05,
+      "loss": 0.6107,
+      "step": 6420
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.797290802001953,
+      "learning_rate": 1.347091127509868e-05,
+      "loss": 0.7206,
+      "step": 6421
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.028078079223633,
+      "learning_rate": 1.3468337051656085e-05,
+      "loss": 0.3351,
+      "step": 6422
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.243917465209961,
+      "learning_rate": 1.346576282821349e-05,
+      "loss": 0.5047,
+      "step": 6423
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.005094528198242,
+      "learning_rate": 1.3463188604770895e-05,
+      "loss": 0.4533,
+      "step": 6424
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.739798545837402,
+      "learning_rate": 1.34606143813283e-05,
+      "loss": 0.3979,
+      "step": 6425
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.81917667388916,
+      "learning_rate": 1.3458040157885705e-05,
+      "loss": 0.8191,
+      "step": 6426
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.305505752563477,
+      "learning_rate": 1.345546593444311e-05,
+      "loss": 0.5539,
+      "step": 6427
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.53671932220459,
+      "learning_rate": 1.3452891711000515e-05,
+      "loss": 0.3946,
+      "step": 6428
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.575438499450684,
+      "learning_rate": 1.345031748755792e-05,
+      "loss": 0.4827,
+      "step": 6429
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.569454193115234,
+      "learning_rate": 1.3447743264115325e-05,
+      "loss": 0.5711,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.72888469696045,
+      "learning_rate": 1.3445169040672731e-05,
+      "loss": 0.6241,
+      "step": 6431
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.193585395812988,
+      "learning_rate": 1.3442594817230136e-05,
+      "loss": 0.6339,
+      "step": 6432
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.300786972045898,
+      "learning_rate": 1.3440020593787541e-05,
+      "loss": 0.6643,
+      "step": 6433
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.59980583190918,
+      "learning_rate": 1.3437446370344946e-05,
+      "loss": 0.6161,
+      "step": 6434
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.819564819335938,
+      "learning_rate": 1.3434872146902353e-05,
+      "loss": 0.6186,
+      "step": 6435
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.4440279006958,
+      "learning_rate": 1.3432297923459756e-05,
+      "loss": 0.484,
+      "step": 6436
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.75587272644043,
+      "learning_rate": 1.3429723700017161e-05,
+      "loss": 0.5328,
+      "step": 6437
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.196776390075684,
+      "learning_rate": 1.3427149476574566e-05,
+      "loss": 0.5457,
+      "step": 6438
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.927118301391602,
+      "learning_rate": 1.3424575253131971e-05,
+      "loss": 0.5412,
+      "step": 6439
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.193121910095215,
+      "learning_rate": 1.3422001029689378e-05,
+      "loss": 0.4445,
+      "step": 6440
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.619534492492676,
+      "learning_rate": 1.3419426806246783e-05,
+      "loss": 0.4927,
+      "step": 6441
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.151309967041016,
+      "learning_rate": 1.3416852582804188e-05,
+      "loss": 0.6449,
+      "step": 6442
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.470060348510742,
+      "learning_rate": 1.3414278359361593e-05,
+      "loss": 0.3805,
+      "step": 6443
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.84939956665039,
+      "learning_rate": 1.3411704135918998e-05,
+      "loss": 0.5345,
+      "step": 6444
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.695511817932129,
+      "learning_rate": 1.3409129912476405e-05,
+      "loss": 0.4607,
+      "step": 6445
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.321296691894531,
+      "learning_rate": 1.340655568903381e-05,
+      "loss": 0.5508,
+      "step": 6446
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.543728828430176,
+      "learning_rate": 1.3403981465591213e-05,
+      "loss": 0.6933,
+      "step": 6447
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.99250602722168,
+      "learning_rate": 1.3401407242148618e-05,
+      "loss": 0.5453,
+      "step": 6448
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.549531936645508,
+      "learning_rate": 1.3398833018706024e-05,
+      "loss": 0.7422,
+      "step": 6449
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.516470909118652,
+      "learning_rate": 1.339625879526343e-05,
+      "loss": 0.6039,
+      "step": 6450
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.269268035888672,
+      "learning_rate": 1.3393684571820834e-05,
+      "loss": 0.5617,
+      "step": 6451
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.108945846557617,
+      "learning_rate": 1.339111034837824e-05,
+      "loss": 0.4165,
+      "step": 6452
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.033666610717773,
+      "learning_rate": 1.3388536124935644e-05,
+      "loss": 0.6895,
+      "step": 6453
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.472251892089844,
+      "learning_rate": 1.3385961901493051e-05,
+      "loss": 0.6872,
+      "step": 6454
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.815091133117676,
+      "learning_rate": 1.3383387678050456e-05,
+      "loss": 0.4326,
+      "step": 6455
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.05883502960205,
+      "learning_rate": 1.3380813454607861e-05,
+      "loss": 0.5259,
+      "step": 6456
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.620383262634277,
+      "learning_rate": 1.3378239231165264e-05,
+      "loss": 0.6472,
+      "step": 6457
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.336908340454102,
+      "learning_rate": 1.337566500772267e-05,
+      "loss": 0.4718,
+      "step": 6458
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.147042274475098,
+      "learning_rate": 1.3373090784280076e-05,
+      "loss": 0.4651,
+      "step": 6459
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.0889253616333,
+      "learning_rate": 1.3370516560837481e-05,
+      "loss": 0.6705,
+      "step": 6460
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.031773567199707,
+      "learning_rate": 1.3367942337394886e-05,
+      "loss": 0.6216,
+      "step": 6461
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.55211067199707,
+      "learning_rate": 1.3365368113952291e-05,
+      "loss": 0.693,
+      "step": 6462
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.207207679748535,
+      "learning_rate": 1.3362793890509696e-05,
+      "loss": 0.4096,
+      "step": 6463
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.5676679611206055,
+      "learning_rate": 1.3360219667067103e-05,
+      "loss": 0.4043,
+      "step": 6464
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.421072959899902,
+      "learning_rate": 1.3357645443624508e-05,
+      "loss": 0.6345,
+      "step": 6465
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.281852722167969,
+      "learning_rate": 1.3355071220181913e-05,
+      "loss": 0.7356,
+      "step": 6466
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.934675216674805,
+      "learning_rate": 1.3352496996739316e-05,
+      "loss": 0.6215,
+      "step": 6467
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.282097816467285,
+      "learning_rate": 1.3349922773296723e-05,
+      "loss": 0.8024,
+      "step": 6468
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.563453674316406,
+      "learning_rate": 1.3347348549854128e-05,
+      "loss": 0.553,
+      "step": 6469
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.30909538269043,
+      "learning_rate": 1.3344774326411533e-05,
+      "loss": 0.7434,
+      "step": 6470
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.07658576965332,
+      "learning_rate": 1.3342200102968938e-05,
+      "loss": 0.5206,
+      "step": 6471
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.457747459411621,
+      "learning_rate": 1.3339625879526343e-05,
+      "loss": 0.4478,
+      "step": 6472
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.680563926696777,
+      "learning_rate": 1.333705165608375e-05,
+      "loss": 0.6205,
+      "step": 6473
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.597335815429688,
+      "learning_rate": 1.3334477432641154e-05,
+      "loss": 0.6658,
+      "step": 6474
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.61847972869873,
+      "learning_rate": 1.333190320919856e-05,
+      "loss": 0.4784,
+      "step": 6475
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.983537673950195,
+      "learning_rate": 1.3329328985755964e-05,
+      "loss": 0.5591,
+      "step": 6476
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.116236686706543,
+      "learning_rate": 1.332675476231337e-05,
+      "loss": 0.4129,
+      "step": 6477
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.38467025756836,
+      "learning_rate": 1.3324180538870774e-05,
+      "loss": 0.6361,
+      "step": 6478
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.936291694641113,
+      "learning_rate": 1.332160631542818e-05,
+      "loss": 0.4836,
+      "step": 6479
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.047820091247559,
+      "learning_rate": 1.3319032091985584e-05,
+      "loss": 0.6939,
+      "step": 6480
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.784460067749023,
+      "learning_rate": 1.3316457868542989e-05,
+      "loss": 0.6411,
+      "step": 6481
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.867308616638184,
+      "learning_rate": 1.3313883645100394e-05,
+      "loss": 0.5686,
+      "step": 6482
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.41030502319336,
+      "learning_rate": 1.33113094216578e-05,
+      "loss": 0.4649,
+      "step": 6483
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.422324180603027,
+      "learning_rate": 1.3308735198215206e-05,
+      "loss": 0.4035,
+      "step": 6484
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.874316215515137,
+      "learning_rate": 1.330616097477261e-05,
+      "loss": 0.4512,
+      "step": 6485
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.03233814239502,
+      "learning_rate": 1.3303586751330016e-05,
+      "loss": 0.4717,
+      "step": 6486
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.250544548034668,
+      "learning_rate": 1.3301012527887422e-05,
+      "loss": 0.5532,
+      "step": 6487
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.79434585571289,
+      "learning_rate": 1.3298438304444826e-05,
+      "loss": 0.4571,
+      "step": 6488
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.573814392089844,
+      "learning_rate": 1.329586408100223e-05,
+      "loss": 0.4652,
+      "step": 6489
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.933924674987793,
+      "learning_rate": 1.3293289857559636e-05,
+      "loss": 0.6203,
+      "step": 6490
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.825202941894531,
+      "learning_rate": 1.329071563411704e-05,
+      "loss": 0.5241,
+      "step": 6491
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.141525268554688,
+      "learning_rate": 1.3288141410674447e-05,
+      "loss": 0.5837,
+      "step": 6492
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.80006217956543,
+      "learning_rate": 1.3285567187231852e-05,
+      "loss": 0.6059,
+      "step": 6493
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.984138488769531,
+      "learning_rate": 1.3282992963789257e-05,
+      "loss": 0.5903,
+      "step": 6494
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.175521850585938,
+      "learning_rate": 1.3280418740346662e-05,
+      "loss": 0.4014,
+      "step": 6495
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.794022560119629,
+      "learning_rate": 1.3277844516904067e-05,
+      "loss": 0.4755,
+      "step": 6496
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.497796058654785,
+      "learning_rate": 1.3275270293461474e-05,
+      "loss": 0.538,
+      "step": 6497
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.756322860717773,
+      "learning_rate": 1.3272696070018879e-05,
+      "loss": 0.4544,
+      "step": 6498
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.190783500671387,
+      "learning_rate": 1.3270121846576282e-05,
+      "loss": 0.5956,
+      "step": 6499
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.597745895385742,
+      "learning_rate": 1.3267547623133687e-05,
+      "loss": 0.5907,
+      "step": 6500
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.086770057678223,
+      "learning_rate": 1.3264973399691094e-05,
+      "loss": 0.6616,
+      "step": 6501
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.597195625305176,
+      "learning_rate": 1.3262399176248499e-05,
+      "loss": 0.6504,
+      "step": 6502
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.084816932678223,
+      "learning_rate": 1.3259824952805904e-05,
+      "loss": 0.3417,
+      "step": 6503
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.245378494262695,
+      "learning_rate": 1.3257250729363309e-05,
+      "loss": 0.6522,
+      "step": 6504
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.221365928649902,
+      "learning_rate": 1.3254676505920714e-05,
+      "loss": 0.5071,
+      "step": 6505
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.660913467407227,
+      "learning_rate": 1.325210228247812e-05,
+      "loss": 0.7962,
+      "step": 6506
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.997649192810059,
+      "learning_rate": 1.3249528059035526e-05,
+      "loss": 0.5702,
+      "step": 6507
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.492313385009766,
+      "learning_rate": 1.324695383559293e-05,
+      "loss": 0.4827,
+      "step": 6508
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.572259902954102,
+      "learning_rate": 1.3244379612150334e-05,
+      "loss": 0.5088,
+      "step": 6509
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.79743766784668,
+      "learning_rate": 1.3241805388707739e-05,
+      "loss": 0.5494,
+      "step": 6510
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.583270072937012,
+      "learning_rate": 1.3239231165265146e-05,
+      "loss": 0.714,
+      "step": 6511
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.860138893127441,
+      "learning_rate": 1.323665694182255e-05,
+      "loss": 0.4833,
+      "step": 6512
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.43123722076416,
+      "learning_rate": 1.3234082718379956e-05,
+      "loss": 0.739,
+      "step": 6513
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.040709495544434,
+      "learning_rate": 1.323150849493736e-05,
+      "loss": 0.7545,
+      "step": 6514
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.854092121124268,
+      "learning_rate": 1.3228934271494765e-05,
+      "loss": 0.3384,
+      "step": 6515
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.156378746032715,
+      "learning_rate": 1.3226360048052172e-05,
+      "loss": 0.5428,
+      "step": 6516
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.444669723510742,
+      "learning_rate": 1.3223785824609577e-05,
+      "loss": 0.5682,
+      "step": 6517
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.886656761169434,
+      "learning_rate": 1.3221211601166982e-05,
+      "loss": 0.5117,
+      "step": 6518
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.365802764892578,
+      "learning_rate": 1.3218637377724387e-05,
+      "loss": 0.5611,
+      "step": 6519
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.611737251281738,
+      "learning_rate": 1.3216063154281792e-05,
+      "loss": 0.5034,
+      "step": 6520
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.446642875671387,
+      "learning_rate": 1.3213488930839197e-05,
+      "loss": 0.7348,
+      "step": 6521
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.14039134979248,
+      "learning_rate": 1.3210914707396602e-05,
+      "loss": 0.3526,
+      "step": 6522
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.861857414245605,
+      "learning_rate": 1.3208340483954007e-05,
+      "loss": 0.5412,
+      "step": 6523
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.091312408447266,
+      "learning_rate": 1.3205766260511412e-05,
+      "loss": 0.5061,
+      "step": 6524
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.006696701049805,
+      "learning_rate": 1.3203192037068819e-05,
+      "loss": 0.5319,
+      "step": 6525
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.957270622253418,
+      "learning_rate": 1.3200617813626224e-05,
+      "loss": 0.3701,
+      "step": 6526
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.448317527770996,
+      "learning_rate": 1.3198043590183629e-05,
+      "loss": 0.4563,
+      "step": 6527
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.945698738098145,
+      "learning_rate": 1.3195469366741034e-05,
+      "loss": 0.4663,
+      "step": 6528
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.893011093139648,
+      "learning_rate": 1.3192895143298439e-05,
+      "loss": 0.5749,
+      "step": 6529
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.641251564025879,
+      "learning_rate": 1.3190320919855844e-05,
+      "loss": 0.527,
+      "step": 6530
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.86562728881836,
+      "learning_rate": 1.3187746696413249e-05,
+      "loss": 0.4909,
+      "step": 6531
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.261011123657227,
+      "learning_rate": 1.3185172472970654e-05,
+      "loss": 0.6065,
+      "step": 6532
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.138240814208984,
+      "learning_rate": 1.3182598249528059e-05,
+      "loss": 0.6694,
+      "step": 6533
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.378952026367188,
+      "learning_rate": 1.3180024026085464e-05,
+      "loss": 0.4776,
+      "step": 6534
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.386820793151855,
+      "learning_rate": 1.317744980264287e-05,
+      "loss": 0.5835,
+      "step": 6535
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.621814727783203,
+      "learning_rate": 1.3174875579200275e-05,
+      "loss": 0.6197,
+      "step": 6536
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.80805778503418,
+      "learning_rate": 1.317230135575768e-05,
+      "loss": 0.4563,
+      "step": 6537
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.223711967468262,
+      "learning_rate": 1.3169727132315085e-05,
+      "loss": 0.7816,
+      "step": 6538
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.302451133728027,
+      "learning_rate": 1.3167152908872492e-05,
+      "loss": 0.4981,
+      "step": 6539
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.550187110900879,
+      "learning_rate": 1.3164578685429895e-05,
+      "loss": 0.612,
+      "step": 6540
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.217123031616211,
+      "learning_rate": 1.31620044619873e-05,
+      "loss": 0.5187,
+      "step": 6541
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.528560638427734,
+      "learning_rate": 1.3159430238544705e-05,
+      "loss": 0.6893,
+      "step": 6542
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.158309936523438,
+      "learning_rate": 1.315685601510211e-05,
+      "loss": 0.5037,
+      "step": 6543
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.893864631652832,
+      "learning_rate": 1.3154281791659517e-05,
+      "loss": 0.6542,
+      "step": 6544
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.234877586364746,
+      "learning_rate": 1.3151707568216922e-05,
+      "loss": 0.5203,
+      "step": 6545
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.685635566711426,
+      "learning_rate": 1.3149133344774327e-05,
+      "loss": 0.6837,
+      "step": 6546
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.813433647155762,
+      "learning_rate": 1.3146559121331732e-05,
+      "loss": 0.4363,
+      "step": 6547
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.92897367477417,
+      "learning_rate": 1.3143984897889137e-05,
+      "loss": 0.4507,
+      "step": 6548
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.177204132080078,
+      "learning_rate": 1.3141410674446544e-05,
+      "loss": 0.4679,
+      "step": 6549
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.392449378967285,
+      "learning_rate": 1.3138836451003949e-05,
+      "loss": 0.5465,
+      "step": 6550
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.450992584228516,
+      "learning_rate": 1.3136262227561352e-05,
+      "loss": 0.5741,
+      "step": 6551
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.513679504394531,
+      "learning_rate": 1.3133688004118757e-05,
+      "loss": 0.4398,
+      "step": 6552
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.889396667480469,
+      "learning_rate": 1.3131113780676162e-05,
+      "loss": 0.8657,
+      "step": 6553
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.439090728759766,
+      "learning_rate": 1.3128539557233568e-05,
+      "loss": 0.5031,
+      "step": 6554
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.649551391601562,
+      "learning_rate": 1.3125965333790973e-05,
+      "loss": 0.4117,
+      "step": 6555
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.325549125671387,
+      "learning_rate": 1.3123391110348378e-05,
+      "loss": 0.4438,
+      "step": 6556
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.233033180236816,
+      "learning_rate": 1.3120816886905783e-05,
+      "loss": 0.7199,
+      "step": 6557
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.1182222366333,
+      "learning_rate": 1.311824266346319e-05,
+      "loss": 0.569,
+      "step": 6558
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.452754974365234,
+      "learning_rate": 1.3115668440020595e-05,
+      "loss": 0.5249,
+      "step": 6559
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.976300239562988,
+      "learning_rate": 1.3113094216578e-05,
+      "loss": 0.7084,
+      "step": 6560
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.58817195892334,
+      "learning_rate": 1.3110519993135403e-05,
+      "loss": 0.4723,
+      "step": 6561
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.377490043640137,
+      "learning_rate": 1.3107945769692808e-05,
+      "loss": 0.4459,
+      "step": 6562
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.727197647094727,
+      "learning_rate": 1.3105371546250215e-05,
+      "loss": 0.4807,
+      "step": 6563
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.826410293579102,
+      "learning_rate": 1.310279732280762e-05,
+      "loss": 0.5837,
+      "step": 6564
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.527913093566895,
+      "learning_rate": 1.3100223099365025e-05,
+      "loss": 0.3558,
+      "step": 6565
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.869670867919922,
+      "learning_rate": 1.309764887592243e-05,
+      "loss": 0.6012,
+      "step": 6566
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.175188064575195,
+      "learning_rate": 1.3095074652479835e-05,
+      "loss": 0.6005,
+      "step": 6567
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.989154815673828,
+      "learning_rate": 1.3092500429037242e-05,
+      "loss": 0.6041,
+      "step": 6568
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.257312774658203,
+      "learning_rate": 1.3089926205594647e-05,
+      "loss": 0.6224,
+      "step": 6569
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.00068473815918,
+      "learning_rate": 1.3087351982152052e-05,
+      "loss": 0.5205,
+      "step": 6570
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.617300987243652,
+      "learning_rate": 1.3084777758709457e-05,
+      "loss": 0.5443,
+      "step": 6571
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.372429847717285,
+      "learning_rate": 1.3082203535266862e-05,
+      "loss": 0.4517,
+      "step": 6572
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.8631591796875,
+      "learning_rate": 1.3079629311824267e-05,
+      "loss": 0.4206,
+      "step": 6573
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.8099541664123535,
+      "learning_rate": 1.3077055088381672e-05,
+      "loss": 0.3376,
+      "step": 6574
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.668180465698242,
+      "learning_rate": 1.3074480864939077e-05,
+      "loss": 0.6727,
+      "step": 6575
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.970752716064453,
+      "learning_rate": 1.3071906641496482e-05,
+      "loss": 0.3809,
+      "step": 6576
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.777921676635742,
+      "learning_rate": 1.3069332418053888e-05,
+      "loss": 0.4781,
+      "step": 6577
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.43701171875,
+      "learning_rate": 1.3066758194611293e-05,
+      "loss": 0.5624,
+      "step": 6578
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.084794998168945,
+      "learning_rate": 1.3064183971168698e-05,
+      "loss": 0.5276,
+      "step": 6579
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.189397811889648,
+      "learning_rate": 1.3061609747726103e-05,
+      "loss": 0.4693,
+      "step": 6580
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.391263961791992,
+      "learning_rate": 1.3059035524283508e-05,
+      "loss": 0.5776,
+      "step": 6581
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.65726089477539,
+      "learning_rate": 1.3056461300840913e-05,
+      "loss": 0.4856,
+      "step": 6582
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.06060791015625,
+      "learning_rate": 1.3053887077398318e-05,
+      "loss": 0.6063,
+      "step": 6583
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 26.931459426879883,
+      "learning_rate": 1.3051312853955723e-05,
+      "loss": 0.5167,
+      "step": 6584
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.135416030883789,
+      "learning_rate": 1.3048738630513128e-05,
+      "loss": 0.6144,
+      "step": 6585
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.115387916564941,
+      "learning_rate": 1.3046164407070533e-05,
+      "loss": 0.5541,
+      "step": 6586
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.395634651184082,
+      "learning_rate": 1.304359018362794e-05,
+      "loss": 0.3439,
+      "step": 6587
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.804046630859375,
+      "learning_rate": 1.3041015960185345e-05,
+      "loss": 0.5401,
+      "step": 6588
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.473113059997559,
+      "learning_rate": 1.303844173674275e-05,
+      "loss": 0.4357,
+      "step": 6589
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.7532854080200195,
+      "learning_rate": 1.3035867513300155e-05,
+      "loss": 0.3525,
+      "step": 6590
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.283815383911133,
+      "learning_rate": 1.3033293289857561e-05,
+      "loss": 0.4762,
+      "step": 6591
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.335599899291992,
+      "learning_rate": 1.3030719066414965e-05,
+      "loss": 0.3369,
+      "step": 6592
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.679243087768555,
+      "learning_rate": 1.302814484297237e-05,
+      "loss": 0.5246,
+      "step": 6593
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.233473777770996,
+      "learning_rate": 1.3025570619529775e-05,
+      "loss": 0.5099,
+      "step": 6594
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.13751220703125,
+      "learning_rate": 1.302299639608718e-05,
+      "loss": 0.3833,
+      "step": 6595
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.868047714233398,
+      "learning_rate": 1.3020422172644586e-05,
+      "loss": 0.3488,
+      "step": 6596
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.519608497619629,
+      "learning_rate": 1.3017847949201991e-05,
+      "loss": 0.4467,
+      "step": 6597
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.756606101989746,
+      "learning_rate": 1.3015273725759396e-05,
+      "loss": 0.5031,
+      "step": 6598
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.033324241638184,
+      "learning_rate": 1.3012699502316801e-05,
+      "loss": 0.5938,
+      "step": 6599
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.303865432739258,
+      "learning_rate": 1.3010125278874206e-05,
+      "loss": 0.4744,
+      "step": 6600
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 22.110292434692383,
+      "learning_rate": 1.3007551055431613e-05,
+      "loss": 0.4931,
+      "step": 6601
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.843348503112793,
+      "learning_rate": 1.3004976831989018e-05,
+      "loss": 0.5608,
+      "step": 6602
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.494935989379883,
+      "learning_rate": 1.3002402608546421e-05,
+      "loss": 0.5409,
+      "step": 6603
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.201446533203125,
+      "learning_rate": 1.2999828385103826e-05,
+      "loss": 0.4608,
+      "step": 6604
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.345254898071289,
+      "learning_rate": 1.2997254161661231e-05,
+      "loss": 0.6298,
+      "step": 6605
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.122580528259277,
+      "learning_rate": 1.2994679938218638e-05,
+      "loss": 0.4921,
+      "step": 6606
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.820239067077637,
+      "learning_rate": 1.2992105714776043e-05,
+      "loss": 0.5318,
+      "step": 6607
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.083476066589355,
+      "learning_rate": 1.2989531491333448e-05,
+      "loss": 0.6027,
+      "step": 6608
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.082131385803223,
+      "learning_rate": 1.2986957267890853e-05,
+      "loss": 0.4911,
+      "step": 6609
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.85802173614502,
+      "learning_rate": 1.298438304444826e-05,
+      "loss": 0.706,
+      "step": 6610
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.752325057983398,
+      "learning_rate": 1.2981808821005665e-05,
+      "loss": 0.7447,
+      "step": 6611
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.42308521270752,
+      "learning_rate": 1.297923459756307e-05,
+      "loss": 0.6757,
+      "step": 6612
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.254287719726562,
+      "learning_rate": 1.2976660374120473e-05,
+      "loss": 0.5193,
+      "step": 6613
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.064400672912598,
+      "learning_rate": 1.2974086150677878e-05,
+      "loss": 0.5508,
+      "step": 6614
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.173871994018555,
+      "learning_rate": 1.2971511927235285e-05,
+      "loss": 0.5603,
+      "step": 6615
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.750433921813965,
+      "learning_rate": 1.296893770379269e-05,
+      "loss": 0.616,
+      "step": 6616
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.16136646270752,
+      "learning_rate": 1.2966363480350095e-05,
+      "loss": 0.7753,
+      "step": 6617
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.198493957519531,
+      "learning_rate": 1.29637892569075e-05,
+      "loss": 0.5215,
+      "step": 6618
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.20034122467041,
+      "learning_rate": 1.2961215033464905e-05,
+      "loss": 0.5867,
+      "step": 6619
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.259267807006836,
+      "learning_rate": 1.2958640810022311e-05,
+      "loss": 0.5525,
+      "step": 6620
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.16513442993164,
+      "learning_rate": 1.2956066586579716e-05,
+      "loss": 0.4784,
+      "step": 6621
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.509882926940918,
+      "learning_rate": 1.2953492363137121e-05,
+      "loss": 0.7367,
+      "step": 6622
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.642499923706055,
+      "learning_rate": 1.2950918139694526e-05,
+      "loss": 0.835,
+      "step": 6623
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.391217231750488,
+      "learning_rate": 1.2948343916251931e-05,
+      "loss": 0.7651,
+      "step": 6624
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.915867328643799,
+      "learning_rate": 1.2945769692809336e-05,
+      "loss": 0.4071,
+      "step": 6625
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.243711471557617,
+      "learning_rate": 1.2943195469366741e-05,
+      "loss": 0.6216,
+      "step": 6626
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.474575996398926,
+      "learning_rate": 1.2940621245924146e-05,
+      "loss": 0.4674,
+      "step": 6627
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.771490097045898,
+      "learning_rate": 1.2938047022481551e-05,
+      "loss": 0.3771,
+      "step": 6628
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.810847282409668,
+      "learning_rate": 1.2935472799038958e-05,
+      "loss": 0.6311,
+      "step": 6629
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.244412422180176,
+      "learning_rate": 1.2932898575596363e-05,
+      "loss": 0.4131,
+      "step": 6630
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.575753211975098,
+      "learning_rate": 1.2930324352153768e-05,
+      "loss": 0.5737,
+      "step": 6631
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.175783157348633,
+      "learning_rate": 1.2927750128711173e-05,
+      "loss": 0.5061,
+      "step": 6632
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.517440795898438,
+      "learning_rate": 1.2925175905268578e-05,
+      "loss": 0.5074,
+      "step": 6633
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.021347999572754,
+      "learning_rate": 1.2922601681825983e-05,
+      "loss": 0.5154,
+      "step": 6634
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.485918045043945,
+      "learning_rate": 1.2920027458383388e-05,
+      "loss": 0.4024,
+      "step": 6635
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.599707126617432,
+      "learning_rate": 1.2917453234940793e-05,
+      "loss": 0.5677,
+      "step": 6636
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.97475528717041,
+      "learning_rate": 1.2914879011498198e-05,
+      "loss": 0.5792,
+      "step": 6637
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.74142074584961,
+      "learning_rate": 1.2912304788055603e-05,
+      "loss": 0.564,
+      "step": 6638
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.491780757904053,
+      "learning_rate": 1.290973056461301e-05,
+      "loss": 0.4438,
+      "step": 6639
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.42711353302002,
+      "learning_rate": 1.2907156341170414e-05,
+      "loss": 0.4575,
+      "step": 6640
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.107260704040527,
+      "learning_rate": 1.290458211772782e-05,
+      "loss": 0.4211,
+      "step": 6641
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.416579246520996,
+      "learning_rate": 1.2902007894285224e-05,
+      "loss": 0.4333,
+      "step": 6642
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.01895523071289,
+      "learning_rate": 1.2899433670842631e-05,
+      "loss": 0.5369,
+      "step": 6643
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.207151412963867,
+      "learning_rate": 1.2896859447400034e-05,
+      "loss": 0.6145,
+      "step": 6644
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.951322555541992,
+      "learning_rate": 1.289428522395744e-05,
+      "loss": 0.6184,
+      "step": 6645
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.277907371520996,
+      "learning_rate": 1.2891711000514844e-05,
+      "loss": 0.3816,
+      "step": 6646
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.779108047485352,
+      "learning_rate": 1.288913677707225e-05,
+      "loss": 0.5212,
+      "step": 6647
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.678054809570312,
+      "learning_rate": 1.2886562553629656e-05,
+      "loss": 0.6432,
+      "step": 6648
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.1563081741333,
+      "learning_rate": 1.2883988330187061e-05,
+      "loss": 0.3937,
+      "step": 6649
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.920218467712402,
+      "learning_rate": 1.2881414106744466e-05,
+      "loss": 0.4412,
+      "step": 6650
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.496282577514648,
+      "learning_rate": 1.2878839883301871e-05,
+      "loss": 0.6261,
+      "step": 6651
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.1200590133667,
+      "learning_rate": 1.2876265659859276e-05,
+      "loss": 0.5965,
+      "step": 6652
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.59368896484375,
+      "learning_rate": 1.2873691436416683e-05,
+      "loss": 0.4906,
+      "step": 6653
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.63156509399414,
+      "learning_rate": 1.2871117212974088e-05,
+      "loss": 0.5116,
+      "step": 6654
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.79605484008789,
+      "learning_rate": 1.286854298953149e-05,
+      "loss": 0.592,
+      "step": 6655
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.004490852355957,
+      "learning_rate": 1.2865968766088896e-05,
+      "loss": 0.4054,
+      "step": 6656
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.764505386352539,
+      "learning_rate": 1.28633945426463e-05,
+      "loss": 0.5286,
+      "step": 6657
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.3403902053833,
+      "learning_rate": 1.2860820319203707e-05,
+      "loss": 0.7439,
+      "step": 6658
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.892584800720215,
+      "learning_rate": 1.2858246095761112e-05,
+      "loss": 0.7384,
+      "step": 6659
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.002338409423828,
+      "learning_rate": 1.2855671872318517e-05,
+      "loss": 0.6576,
+      "step": 6660
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.832758903503418,
+      "learning_rate": 1.2853097648875922e-05,
+      "loss": 0.6233,
+      "step": 6661
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.900686264038086,
+      "learning_rate": 1.2850523425433329e-05,
+      "loss": 0.5388,
+      "step": 6662
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.024088859558105,
+      "learning_rate": 1.2847949201990734e-05,
+      "loss": 0.476,
+      "step": 6663
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.43239688873291,
+      "learning_rate": 1.2845374978548139e-05,
+      "loss": 0.4752,
+      "step": 6664
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.08057975769043,
+      "learning_rate": 1.2842800755105542e-05,
+      "loss": 0.5479,
+      "step": 6665
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.494189262390137,
+      "learning_rate": 1.2840226531662947e-05,
+      "loss": 0.5553,
+      "step": 6666
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.32431697845459,
+      "learning_rate": 1.2837652308220354e-05,
+      "loss": 0.4402,
+      "step": 6667
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.670583724975586,
+      "learning_rate": 1.2835078084777759e-05,
+      "loss": 0.5737,
+      "step": 6668
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.428166389465332,
+      "learning_rate": 1.2832503861335164e-05,
+      "loss": 0.3968,
+      "step": 6669
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.724882125854492,
+      "learning_rate": 1.2829929637892569e-05,
+      "loss": 0.4201,
+      "step": 6670
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.7952656745910645,
+      "learning_rate": 1.2827355414449974e-05,
+      "loss": 0.4525,
+      "step": 6671
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.33945083618164,
+      "learning_rate": 1.282478119100738e-05,
+      "loss": 0.5353,
+      "step": 6672
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.126082420349121,
+      "learning_rate": 1.2822206967564786e-05,
+      "loss": 0.5401,
+      "step": 6673
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.593513488769531,
+      "learning_rate": 1.281963274412219e-05,
+      "loss": 0.7047,
+      "step": 6674
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.098726272583008,
+      "learning_rate": 1.2817058520679596e-05,
+      "loss": 0.6034,
+      "step": 6675
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.617947578430176,
+      "learning_rate": 1.2814484297237e-05,
+      "loss": 0.3692,
+      "step": 6676
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.953615188598633,
+      "learning_rate": 1.2811910073794406e-05,
+      "loss": 0.5437,
+      "step": 6677
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.29536247253418,
+      "learning_rate": 1.280933585035181e-05,
+      "loss": 0.53,
+      "step": 6678
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.399231910705566,
+      "learning_rate": 1.2806761626909216e-05,
+      "loss": 0.6281,
+      "step": 6679
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.361248970031738,
+      "learning_rate": 1.280418740346662e-05,
+      "loss": 0.3911,
+      "step": 6680
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.777186393737793,
+      "learning_rate": 1.2801613180024027e-05,
+      "loss": 0.4876,
+      "step": 6681
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.768627166748047,
+      "learning_rate": 1.2799038956581432e-05,
+      "loss": 0.7891,
+      "step": 6682
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.145493507385254,
+      "learning_rate": 1.2796464733138837e-05,
+      "loss": 0.8949,
+      "step": 6683
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.15176773071289,
+      "learning_rate": 1.2793890509696242e-05,
+      "loss": 0.5289,
+      "step": 6684
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.483684539794922,
+      "learning_rate": 1.2791316286253647e-05,
+      "loss": 0.6661,
+      "step": 6685
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.809720039367676,
+      "learning_rate": 1.2788742062811052e-05,
+      "loss": 0.4736,
+      "step": 6686
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.834911346435547,
+      "learning_rate": 1.2786167839368457e-05,
+      "loss": 0.6128,
+      "step": 6687
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.260591506958008,
+      "learning_rate": 1.2783593615925862e-05,
+      "loss": 0.7091,
+      "step": 6688
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.857799530029297,
+      "learning_rate": 1.2781019392483267e-05,
+      "loss": 0.5103,
+      "step": 6689
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.976391792297363,
+      "learning_rate": 1.2778445169040672e-05,
+      "loss": 0.693,
+      "step": 6690
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.264769554138184,
+      "learning_rate": 1.2775870945598079e-05,
+      "loss": 0.4276,
+      "step": 6691
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.126923084259033,
+      "learning_rate": 1.2773296722155484e-05,
+      "loss": 0.309,
+      "step": 6692
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.771899223327637,
+      "learning_rate": 1.2770722498712889e-05,
+      "loss": 0.548,
+      "step": 6693
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.559141159057617,
+      "learning_rate": 1.2768148275270294e-05,
+      "loss": 0.6396,
+      "step": 6694
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.344493865966797,
+      "learning_rate": 1.27655740518277e-05,
+      "loss": 0.6079,
+      "step": 6695
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.279424667358398,
+      "learning_rate": 1.2762999828385104e-05,
+      "loss": 0.9092,
+      "step": 6696
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.471848487854004,
+      "learning_rate": 1.2760425604942509e-05,
+      "loss": 0.7587,
+      "step": 6697
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.34504222869873,
+      "learning_rate": 1.2757851381499914e-05,
+      "loss": 0.5484,
+      "step": 6698
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.52061653137207,
+      "learning_rate": 1.2755277158057319e-05,
+      "loss": 0.5139,
+      "step": 6699
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.919501304626465,
+      "learning_rate": 1.2752702934614725e-05,
+      "loss": 0.4915,
+      "step": 6700
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.039825439453125,
+      "learning_rate": 1.275012871117213e-05,
+      "loss": 0.3715,
+      "step": 6701
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.143620491027832,
+      "learning_rate": 1.2747554487729535e-05,
+      "loss": 0.4217,
+      "step": 6702
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.147356986999512,
+      "learning_rate": 1.274498026428694e-05,
+      "loss": 0.5138,
+      "step": 6703
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.371033668518066,
+      "learning_rate": 1.2742406040844345e-05,
+      "loss": 0.5299,
+      "step": 6704
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.041821479797363,
+      "learning_rate": 1.2739831817401752e-05,
+      "loss": 0.4868,
+      "step": 6705
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.828451156616211,
+      "learning_rate": 1.2737257593959157e-05,
+      "loss": 0.5587,
+      "step": 6706
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.729795455932617,
+      "learning_rate": 1.273468337051656e-05,
+      "loss": 0.3908,
+      "step": 6707
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.917746543884277,
+      "learning_rate": 1.2732109147073965e-05,
+      "loss": 0.3145,
+      "step": 6708
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.760326385498047,
+      "learning_rate": 1.272953492363137e-05,
+      "loss": 0.3523,
+      "step": 6709
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.388890266418457,
+      "learning_rate": 1.2726960700188777e-05,
+      "loss": 0.5523,
+      "step": 6710
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.174628257751465,
+      "learning_rate": 1.2724386476746182e-05,
+      "loss": 0.5867,
+      "step": 6711
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.072940826416016,
+      "learning_rate": 1.2721812253303587e-05,
+      "loss": 0.5981,
+      "step": 6712
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.517169952392578,
+      "learning_rate": 1.2719238029860992e-05,
+      "loss": 0.4229,
+      "step": 6713
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.435815811157227,
+      "learning_rate": 1.2716663806418399e-05,
+      "loss": 0.8389,
+      "step": 6714
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.92951774597168,
+      "learning_rate": 1.2714089582975804e-05,
+      "loss": 0.5887,
+      "step": 6715
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.015429496765137,
+      "learning_rate": 1.2711515359533209e-05,
+      "loss": 0.5204,
+      "step": 6716
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.950034141540527,
+      "learning_rate": 1.2708941136090612e-05,
+      "loss": 0.5696,
+      "step": 6717
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.45562744140625,
+      "learning_rate": 1.2706366912648017e-05,
+      "loss": 0.7551,
+      "step": 6718
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.538154602050781,
+      "learning_rate": 1.2703792689205424e-05,
+      "loss": 0.5767,
+      "step": 6719
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.079790115356445,
+      "learning_rate": 1.2701218465762829e-05,
+      "loss": 0.5863,
+      "step": 6720
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.61666488647461,
+      "learning_rate": 1.2698644242320234e-05,
+      "loss": 0.691,
+      "step": 6721
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.60115909576416,
+      "learning_rate": 1.2696070018877639e-05,
+      "loss": 0.4377,
+      "step": 6722
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.699920654296875,
+      "learning_rate": 1.2693495795435044e-05,
+      "loss": 0.6973,
+      "step": 6723
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.728428840637207,
+      "learning_rate": 1.269092157199245e-05,
+      "loss": 0.4249,
+      "step": 6724
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.760560989379883,
+      "learning_rate": 1.2688347348549855e-05,
+      "loss": 0.9823,
+      "step": 6725
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.318336486816406,
+      "learning_rate": 1.268577312510726e-05,
+      "loss": 0.4204,
+      "step": 6726
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.182144165039062,
+      "learning_rate": 1.2683198901664665e-05,
+      "loss": 0.4436,
+      "step": 6727
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.014725685119629,
+      "learning_rate": 1.2680624678222068e-05,
+      "loss": 0.7163,
+      "step": 6728
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.626301765441895,
+      "learning_rate": 1.2678050454779475e-05,
+      "loss": 0.6035,
+      "step": 6729
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.62788200378418,
+      "learning_rate": 1.267547623133688e-05,
+      "loss": 0.4954,
+      "step": 6730
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.553505897521973,
+      "learning_rate": 1.2672902007894285e-05,
+      "loss": 0.4281,
+      "step": 6731
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.797175407409668,
+      "learning_rate": 1.267032778445169e-05,
+      "loss": 0.6781,
+      "step": 6732
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.341609954833984,
+      "learning_rate": 1.2667753561009097e-05,
+      "loss": 0.6553,
+      "step": 6733
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.617149353027344,
+      "learning_rate": 1.2665179337566502e-05,
+      "loss": 0.5201,
+      "step": 6734
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.680614471435547,
+      "learning_rate": 1.2662605114123907e-05,
+      "loss": 0.6146,
+      "step": 6735
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.916034698486328,
+      "learning_rate": 1.2660030890681312e-05,
+      "loss": 0.8438,
+      "step": 6736
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.099270820617676,
+      "learning_rate": 1.2657456667238717e-05,
+      "loss": 0.5887,
+      "step": 6737
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.548237800598145,
+      "learning_rate": 1.2654882443796122e-05,
+      "loss": 0.6987,
+      "step": 6738
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.586852073669434,
+      "learning_rate": 1.2652308220353527e-05,
+      "loss": 0.4909,
+      "step": 6739
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.159618377685547,
+      "learning_rate": 1.2649733996910932e-05,
+      "loss": 0.3954,
+      "step": 6740
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.554677963256836,
+      "learning_rate": 1.2647159773468337e-05,
+      "loss": 0.5544,
+      "step": 6741
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.785699844360352,
+      "learning_rate": 1.2644585550025742e-05,
+      "loss": 0.5357,
+      "step": 6742
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.965754508972168,
+      "learning_rate": 1.2642011326583148e-05,
+      "loss": 0.5779,
+      "step": 6743
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.249988555908203,
+      "learning_rate": 1.2639437103140553e-05,
+      "loss": 0.5882,
+      "step": 6744
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.296032905578613,
+      "learning_rate": 1.2636862879697958e-05,
+      "loss": 0.3963,
+      "step": 6745
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.62956428527832,
+      "learning_rate": 1.2634288656255363e-05,
+      "loss": 0.4947,
+      "step": 6746
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.288333892822266,
+      "learning_rate": 1.263171443281277e-05,
+      "loss": 0.4609,
+      "step": 6747
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.94849681854248,
+      "learning_rate": 1.2629140209370173e-05,
+      "loss": 0.6603,
+      "step": 6748
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.329885482788086,
+      "learning_rate": 1.2626565985927578e-05,
+      "loss": 0.569,
+      "step": 6749
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.659663200378418,
+      "learning_rate": 1.2623991762484983e-05,
+      "loss": 0.4487,
+      "step": 6750
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.303452491760254,
+      "learning_rate": 1.2621417539042388e-05,
+      "loss": 0.5362,
+      "step": 6751
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.65714168548584,
+      "learning_rate": 1.2618843315599795e-05,
+      "loss": 0.7225,
+      "step": 6752
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.553597450256348,
+      "learning_rate": 1.26162690921572e-05,
+      "loss": 0.3538,
+      "step": 6753
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.719615936279297,
+      "learning_rate": 1.2613694868714605e-05,
+      "loss": 0.5659,
+      "step": 6754
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.221165657043457,
+      "learning_rate": 1.261112064527201e-05,
+      "loss": 0.4145,
+      "step": 6755
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.493268013000488,
+      "learning_rate": 1.2608546421829415e-05,
+      "loss": 0.5086,
+      "step": 6756
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.034260749816895,
+      "learning_rate": 1.2605972198386822e-05,
+      "loss": 0.6157,
+      "step": 6757
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.103609085083008,
+      "learning_rate": 1.2603397974944227e-05,
+      "loss": 0.6273,
+      "step": 6758
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.861727714538574,
+      "learning_rate": 1.260082375150163e-05,
+      "loss": 0.4413,
+      "step": 6759
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.260859489440918,
+      "learning_rate": 1.2598249528059035e-05,
+      "loss": 0.5558,
+      "step": 6760
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.298737525939941,
+      "learning_rate": 1.259567530461644e-05,
+      "loss": 0.3999,
+      "step": 6761
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.542593002319336,
+      "learning_rate": 1.2593101081173846e-05,
+      "loss": 0.4122,
+      "step": 6762
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.299402236938477,
+      "learning_rate": 1.2590526857731251e-05,
+      "loss": 0.4636,
+      "step": 6763
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.075308799743652,
+      "learning_rate": 1.2587952634288656e-05,
+      "loss": 0.6894,
+      "step": 6764
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.396122932434082,
+      "learning_rate": 1.2585378410846061e-05,
+      "loss": 0.5546,
+      "step": 6765
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.977927207946777,
+      "learning_rate": 1.2582804187403468e-05,
+      "loss": 0.5561,
+      "step": 6766
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.994551658630371,
+      "learning_rate": 1.2580229963960873e-05,
+      "loss": 0.3995,
+      "step": 6767
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.106865882873535,
+      "learning_rate": 1.2577655740518278e-05,
+      "loss": 0.7406,
+      "step": 6768
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.34880542755127,
+      "learning_rate": 1.2575081517075681e-05,
+      "loss": 0.427,
+      "step": 6769
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.80229663848877,
+      "learning_rate": 1.2572507293633086e-05,
+      "loss": 0.6487,
+      "step": 6770
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.099350929260254,
+      "learning_rate": 1.2569933070190493e-05,
+      "loss": 0.4811,
+      "step": 6771
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.811274528503418,
+      "learning_rate": 1.2567358846747898e-05,
+      "loss": 0.6307,
+      "step": 6772
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.495442390441895,
+      "learning_rate": 1.2564784623305303e-05,
+      "loss": 0.6225,
+      "step": 6773
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.954984188079834,
+      "learning_rate": 1.2562210399862708e-05,
+      "loss": 0.3954,
+      "step": 6774
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.646415710449219,
+      "learning_rate": 1.2559636176420113e-05,
+      "loss": 0.5928,
+      "step": 6775
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.01123046875,
+      "learning_rate": 1.255706195297752e-05,
+      "loss": 0.4359,
+      "step": 6776
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.666677474975586,
+      "learning_rate": 1.2554487729534925e-05,
+      "loss": 0.4723,
+      "step": 6777
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.862643241882324,
+      "learning_rate": 1.255191350609233e-05,
+      "loss": 0.684,
+      "step": 6778
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.741910934448242,
+      "learning_rate": 1.2549339282649735e-05,
+      "loss": 0.5769,
+      "step": 6779
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.529659271240234,
+      "learning_rate": 1.2546765059207138e-05,
+      "loss": 0.5424,
+      "step": 6780
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.7211012840271,
+      "learning_rate": 1.2544190835764545e-05,
+      "loss": 0.3708,
+      "step": 6781
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.521566390991211,
+      "learning_rate": 1.254161661232195e-05,
+      "loss": 0.4304,
+      "step": 6782
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.458416938781738,
+      "learning_rate": 1.2539042388879355e-05,
+      "loss": 0.5796,
+      "step": 6783
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.682881355285645,
+      "learning_rate": 1.253646816543676e-05,
+      "loss": 0.7297,
+      "step": 6784
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.332749366760254,
+      "learning_rate": 1.2533893941994166e-05,
+      "loss": 0.4912,
+      "step": 6785
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.674663543701172,
+      "learning_rate": 1.2531319718551571e-05,
+      "loss": 0.4537,
+      "step": 6786
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.876639366149902,
+      "learning_rate": 1.2528745495108976e-05,
+      "loss": 0.6138,
+      "step": 6787
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.832884788513184,
+      "learning_rate": 1.2526171271666381e-05,
+      "loss": 0.6242,
+      "step": 6788
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.507376670837402,
+      "learning_rate": 1.2523597048223786e-05,
+      "loss": 0.3926,
+      "step": 6789
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.872946739196777,
+      "learning_rate": 1.2521022824781191e-05,
+      "loss": 0.4891,
+      "step": 6790
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.619639873504639,
+      "learning_rate": 1.2518448601338596e-05,
+      "loss": 0.5368,
+      "step": 6791
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.240330696105957,
+      "learning_rate": 1.2515874377896001e-05,
+      "loss": 0.4658,
+      "step": 6792
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.234431266784668,
+      "learning_rate": 1.2513300154453406e-05,
+      "loss": 0.4802,
+      "step": 6793
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.64299201965332,
+      "learning_rate": 1.2510725931010811e-05,
+      "loss": 0.5694,
+      "step": 6794
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.616625785827637,
+      "learning_rate": 1.2508151707568218e-05,
+      "loss": 0.4409,
+      "step": 6795
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.579665184020996,
+      "learning_rate": 1.2505577484125623e-05,
+      "loss": 0.4268,
+      "step": 6796
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.091870307922363,
+      "learning_rate": 1.2503003260683028e-05,
+      "loss": 0.4956,
+      "step": 6797
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.065946578979492,
+      "learning_rate": 1.2500429037240433e-05,
+      "loss": 0.7655,
+      "step": 6798
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.62783145904541,
+      "learning_rate": 1.249785481379784e-05,
+      "loss": 0.5838,
+      "step": 6799
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.778759002685547,
+      "learning_rate": 1.2495280590355243e-05,
+      "loss": 0.4989,
+      "step": 6800
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.538947105407715,
+      "learning_rate": 1.2492706366912648e-05,
+      "loss": 0.6402,
+      "step": 6801
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.49754810333252,
+      "learning_rate": 1.2490132143470053e-05,
+      "loss": 0.3548,
+      "step": 6802
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.568778991699219,
+      "learning_rate": 1.2487557920027458e-05,
+      "loss": 0.3727,
+      "step": 6803
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.141242027282715,
+      "learning_rate": 1.2484983696584864e-05,
+      "loss": 0.4706,
+      "step": 6804
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.662147521972656,
+      "learning_rate": 1.248240947314227e-05,
+      "loss": 0.6372,
+      "step": 6805
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.579142570495605,
+      "learning_rate": 1.2479835249699674e-05,
+      "loss": 0.513,
+      "step": 6806
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.628792762756348,
+      "learning_rate": 1.247726102625708e-05,
+      "loss": 0.6056,
+      "step": 6807
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.767404556274414,
+      "learning_rate": 1.2474686802814484e-05,
+      "loss": 0.6001,
+      "step": 6808
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.569428443908691,
+      "learning_rate": 1.2472112579371891e-05,
+      "loss": 0.5376,
+      "step": 6809
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.378584861755371,
+      "learning_rate": 1.2469538355929296e-05,
+      "loss": 0.3777,
+      "step": 6810
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.086263656616211,
+      "learning_rate": 1.24669641324867e-05,
+      "loss": 0.6188,
+      "step": 6811
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.007267951965332,
+      "learning_rate": 1.2464389909044104e-05,
+      "loss": 0.5115,
+      "step": 6812
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.275233745574951,
+      "learning_rate": 1.246181568560151e-05,
+      "loss": 0.3426,
+      "step": 6813
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.040640830993652,
+      "learning_rate": 1.2459241462158916e-05,
+      "loss": 0.6234,
+      "step": 6814
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.325801849365234,
+      "learning_rate": 1.2456667238716321e-05,
+      "loss": 0.4771,
+      "step": 6815
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.979382514953613,
+      "learning_rate": 1.2454093015273726e-05,
+      "loss": 0.3407,
+      "step": 6816
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.458011627197266,
+      "learning_rate": 1.2451518791831131e-05,
+      "loss": 0.567,
+      "step": 6817
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.544333457946777,
+      "learning_rate": 1.2448944568388538e-05,
+      "loss": 0.615,
+      "step": 6818
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.83591079711914,
+      "learning_rate": 1.2446370344945943e-05,
+      "loss": 0.4769,
+      "step": 6819
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.802139282226562,
+      "learning_rate": 1.2443796121503348e-05,
+      "loss": 0.5772,
+      "step": 6820
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.135982513427734,
+      "learning_rate": 1.2441221898060751e-05,
+      "loss": 0.5482,
+      "step": 6821
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.574701309204102,
+      "learning_rate": 1.2438647674618156e-05,
+      "loss": 0.3975,
+      "step": 6822
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.294801712036133,
+      "learning_rate": 1.2436073451175563e-05,
+      "loss": 0.4745,
+      "step": 6823
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.599040031433105,
+      "learning_rate": 1.2433499227732968e-05,
+      "loss": 0.5905,
+      "step": 6824
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.782919406890869,
+      "learning_rate": 1.2430925004290373e-05,
+      "loss": 0.3649,
+      "step": 6825
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.188220977783203,
+      "learning_rate": 1.2428350780847778e-05,
+      "loss": 0.4365,
+      "step": 6826
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.269412994384766,
+      "learning_rate": 1.2425776557405183e-05,
+      "loss": 0.5385,
+      "step": 6827
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.031576156616211,
+      "learning_rate": 1.242320233396259e-05,
+      "loss": 0.6859,
+      "step": 6828
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.004095077514648,
+      "learning_rate": 1.2420628110519994e-05,
+      "loss": 0.5292,
+      "step": 6829
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.24494457244873,
+      "learning_rate": 1.24180538870774e-05,
+      "loss": 0.424,
+      "step": 6830
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.175368309020996,
+      "learning_rate": 1.2415479663634804e-05,
+      "loss": 0.6736,
+      "step": 6831
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.513937950134277,
+      "learning_rate": 1.2412905440192207e-05,
+      "loss": 0.551,
+      "step": 6832
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.278603553771973,
+      "learning_rate": 1.2410331216749614e-05,
+      "loss": 0.8424,
+      "step": 6833
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.5023832321167,
+      "learning_rate": 1.2407756993307019e-05,
+      "loss": 0.5901,
+      "step": 6834
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.795166969299316,
+      "learning_rate": 1.2405182769864424e-05,
+      "loss": 0.5726,
+      "step": 6835
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.320119857788086,
+      "learning_rate": 1.2402608546421829e-05,
+      "loss": 0.5802,
+      "step": 6836
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.436756134033203,
+      "learning_rate": 1.2400034322979236e-05,
+      "loss": 0.5802,
+      "step": 6837
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.720342636108398,
+      "learning_rate": 1.239746009953664e-05,
+      "loss": 0.6414,
+      "step": 6838
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.858325481414795,
+      "learning_rate": 1.2394885876094046e-05,
+      "loss": 0.4137,
+      "step": 6839
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.454837799072266,
+      "learning_rate": 1.239231165265145e-05,
+      "loss": 0.3474,
+      "step": 6840
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.06265926361084,
+      "learning_rate": 1.2389737429208856e-05,
+      "loss": 0.7606,
+      "step": 6841
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.420174598693848,
+      "learning_rate": 1.238716320576626e-05,
+      "loss": 0.7873,
+      "step": 6842
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.158156394958496,
+      "learning_rate": 1.2384588982323666e-05,
+      "loss": 0.643,
+      "step": 6843
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.007019996643066,
+      "learning_rate": 1.238201475888107e-05,
+      "loss": 0.4884,
+      "step": 6844
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.749552726745605,
+      "learning_rate": 1.2379440535438476e-05,
+      "loss": 0.3755,
+      "step": 6845
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.634915351867676,
+      "learning_rate": 1.237686631199588e-05,
+      "loss": 0.6722,
+      "step": 6846
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.04141902923584,
+      "learning_rate": 1.2374292088553287e-05,
+      "loss": 0.489,
+      "step": 6847
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.348457336425781,
+      "learning_rate": 1.2371717865110692e-05,
+      "loss": 0.5139,
+      "step": 6848
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.05846881866455,
+      "learning_rate": 1.2369143641668097e-05,
+      "loss": 0.4805,
+      "step": 6849
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.247013092041016,
+      "learning_rate": 1.2366569418225502e-05,
+      "loss": 0.8583,
+      "step": 6850
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.411354064941406,
+      "learning_rate": 1.2363995194782909e-05,
+      "loss": 0.5649,
+      "step": 6851
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.20920181274414,
+      "learning_rate": 1.2361420971340312e-05,
+      "loss": 0.7022,
+      "step": 6852
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.595610618591309,
+      "learning_rate": 1.2358846747897717e-05,
+      "loss": 0.4616,
+      "step": 6853
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.999131202697754,
+      "learning_rate": 1.2356272524455122e-05,
+      "loss": 0.6998,
+      "step": 6854
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.027018547058105,
+      "learning_rate": 1.2353698301012527e-05,
+      "loss": 0.7248,
+      "step": 6855
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.674300193786621,
+      "learning_rate": 1.2351124077569934e-05,
+      "loss": 0.5586,
+      "step": 6856
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.837050437927246,
+      "learning_rate": 1.2348549854127339e-05,
+      "loss": 0.4872,
+      "step": 6857
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.066168785095215,
+      "learning_rate": 1.2345975630684744e-05,
+      "loss": 0.6862,
+      "step": 6858
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.072893142700195,
+      "learning_rate": 1.2343401407242149e-05,
+      "loss": 0.4017,
+      "step": 6859
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.817351341247559,
+      "learning_rate": 1.2340827183799554e-05,
+      "loss": 0.545,
+      "step": 6860
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.184842109680176,
+      "learning_rate": 1.233825296035696e-05,
+      "loss": 0.3974,
+      "step": 6861
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.7413859367370605,
+      "learning_rate": 1.2335678736914366e-05,
+      "loss": 0.4536,
+      "step": 6862
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.22570514678955,
+      "learning_rate": 1.2333104513471769e-05,
+      "loss": 0.5915,
+      "step": 6863
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.475497245788574,
+      "learning_rate": 1.2330530290029174e-05,
+      "loss": 0.4488,
+      "step": 6864
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.001102447509766,
+      "learning_rate": 1.2327956066586579e-05,
+      "loss": 0.4683,
+      "step": 6865
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.96658706665039,
+      "learning_rate": 1.2325381843143986e-05,
+      "loss": 0.494,
+      "step": 6866
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.885152816772461,
+      "learning_rate": 1.232280761970139e-05,
+      "loss": 0.3838,
+      "step": 6867
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.549115180969238,
+      "learning_rate": 1.2320233396258795e-05,
+      "loss": 0.4423,
+      "step": 6868
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.89554214477539,
+      "learning_rate": 1.23176591728162e-05,
+      "loss": 0.5626,
+      "step": 6869
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.456453323364258,
+      "learning_rate": 1.2315084949373607e-05,
+      "loss": 0.4202,
+      "step": 6870
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.577530860900879,
+      "learning_rate": 1.2312510725931012e-05,
+      "loss": 0.6112,
+      "step": 6871
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.862924575805664,
+      "learning_rate": 1.2309936502488417e-05,
+      "loss": 0.478,
+      "step": 6872
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.98381519317627,
+      "learning_rate": 1.230736227904582e-05,
+      "loss": 0.4468,
+      "step": 6873
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.726097106933594,
+      "learning_rate": 1.2304788055603225e-05,
+      "loss": 0.608,
+      "step": 6874
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.779783248901367,
+      "learning_rate": 1.2302213832160632e-05,
+      "loss": 0.3484,
+      "step": 6875
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.76326847076416,
+      "learning_rate": 1.2299639608718037e-05,
+      "loss": 0.4865,
+      "step": 6876
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.6524658203125,
+      "learning_rate": 1.2297065385275442e-05,
+      "loss": 0.537,
+      "step": 6877
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.634624481201172,
+      "learning_rate": 1.2294491161832847e-05,
+      "loss": 0.4657,
+      "step": 6878
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.680264472961426,
+      "learning_rate": 1.2291916938390252e-05,
+      "loss": 0.4945,
+      "step": 6879
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.019296646118164,
+      "learning_rate": 1.2289342714947659e-05,
+      "loss": 0.4204,
+      "step": 6880
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.379463195800781,
+      "learning_rate": 1.2286768491505064e-05,
+      "loss": 0.5197,
+      "step": 6881
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.925402641296387,
+      "learning_rate": 1.2284194268062469e-05,
+      "loss": 0.6828,
+      "step": 6882
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.060747146606445,
+      "learning_rate": 1.2281620044619874e-05,
+      "loss": 0.3344,
+      "step": 6883
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.3075590133667,
+      "learning_rate": 1.2279045821177277e-05,
+      "loss": 0.7703,
+      "step": 6884
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.060023307800293,
+      "learning_rate": 1.2276471597734684e-05,
+      "loss": 0.65,
+      "step": 6885
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.080811500549316,
+      "learning_rate": 1.2273897374292089e-05,
+      "loss": 0.5368,
+      "step": 6886
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.84549617767334,
+      "learning_rate": 1.2271323150849494e-05,
+      "loss": 0.381,
+      "step": 6887
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.63058853149414,
+      "learning_rate": 1.2268748927406899e-05,
+      "loss": 0.4809,
+      "step": 6888
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.03759765625,
+      "learning_rate": 1.2266174703964305e-05,
+      "loss": 0.5285,
+      "step": 6889
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.594520568847656,
+      "learning_rate": 1.226360048052171e-05,
+      "loss": 0.48,
+      "step": 6890
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.152198791503906,
+      "learning_rate": 1.2261026257079115e-05,
+      "loss": 0.3318,
+      "step": 6891
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.063015937805176,
+      "learning_rate": 1.225845203363652e-05,
+      "loss": 0.5612,
+      "step": 6892
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.247658729553223,
+      "learning_rate": 1.2255877810193925e-05,
+      "loss": 0.6018,
+      "step": 6893
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.392963409423828,
+      "learning_rate": 1.225330358675133e-05,
+      "loss": 0.394,
+      "step": 6894
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.330645561218262,
+      "learning_rate": 1.2250729363308735e-05,
+      "loss": 0.7683,
+      "step": 6895
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.127904415130615,
+      "learning_rate": 1.224815513986614e-05,
+      "loss": 0.4794,
+      "step": 6896
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.829099655151367,
+      "learning_rate": 1.2245580916423545e-05,
+      "loss": 0.4362,
+      "step": 6897
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.440721035003662,
+      "learning_rate": 1.224300669298095e-05,
+      "loss": 0.4115,
+      "step": 6898
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.553647994995117,
+      "learning_rate": 1.2240432469538357e-05,
+      "loss": 0.431,
+      "step": 6899
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.637636184692383,
+      "learning_rate": 1.2237858246095762e-05,
+      "loss": 0.5577,
+      "step": 6900
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.499363899230957,
+      "learning_rate": 1.2235284022653167e-05,
+      "loss": 0.4076,
+      "step": 6901
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.196478843688965,
+      "learning_rate": 1.2232709799210572e-05,
+      "loss": 0.677,
+      "step": 6902
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.579410552978516,
+      "learning_rate": 1.2230135575767977e-05,
+      "loss": 0.6375,
+      "step": 6903
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.992631912231445,
+      "learning_rate": 1.2227561352325382e-05,
+      "loss": 0.5329,
+      "step": 6904
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.567315101623535,
+      "learning_rate": 1.2224987128882787e-05,
+      "loss": 0.7124,
+      "step": 6905
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.849156379699707,
+      "learning_rate": 1.2222412905440192e-05,
+      "loss": 0.439,
+      "step": 6906
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.673786163330078,
+      "learning_rate": 1.2219838681997597e-05,
+      "loss": 0.3328,
+      "step": 6907
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.686080932617188,
+      "learning_rate": 1.2217264458555003e-05,
+      "loss": 0.5408,
+      "step": 6908
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.886615753173828,
+      "learning_rate": 1.2214690235112408e-05,
+      "loss": 0.4232,
+      "step": 6909
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.084321022033691,
+      "learning_rate": 1.2212116011669813e-05,
+      "loss": 0.5372,
+      "step": 6910
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.213665008544922,
+      "learning_rate": 1.2209541788227218e-05,
+      "loss": 0.3082,
+      "step": 6911
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.59633731842041,
+      "learning_rate": 1.2206967564784623e-05,
+      "loss": 0.5706,
+      "step": 6912
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.798014640808105,
+      "learning_rate": 1.220439334134203e-05,
+      "loss": 0.5154,
+      "step": 6913
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.033978462219238,
+      "learning_rate": 1.2201819117899435e-05,
+      "loss": 0.6533,
+      "step": 6914
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.254484176635742,
+      "learning_rate": 1.2199244894456838e-05,
+      "loss": 0.4753,
+      "step": 6915
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.755403518676758,
+      "learning_rate": 1.2196670671014243e-05,
+      "loss": 0.6712,
+      "step": 6916
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.937376976013184,
+      "learning_rate": 1.2194096447571648e-05,
+      "loss": 0.5788,
+      "step": 6917
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.414369583129883,
+      "learning_rate": 1.2191522224129055e-05,
+      "loss": 0.5647,
+      "step": 6918
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.309571266174316,
+      "learning_rate": 1.218894800068646e-05,
+      "loss": 0.4972,
+      "step": 6919
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.03283405303955,
+      "learning_rate": 1.2186373777243865e-05,
+      "loss": 0.3565,
+      "step": 6920
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.124144554138184,
+      "learning_rate": 1.218379955380127e-05,
+      "loss": 0.6071,
+      "step": 6921
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.392751693725586,
+      "learning_rate": 1.2181225330358677e-05,
+      "loss": 0.5957,
+      "step": 6922
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.871675491333008,
+      "learning_rate": 1.2178651106916082e-05,
+      "loss": 0.4522,
+      "step": 6923
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.38727855682373,
+      "learning_rate": 1.2176076883473487e-05,
+      "loss": 0.3741,
+      "step": 6924
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.851292610168457,
+      "learning_rate": 1.217350266003089e-05,
+      "loss": 0.4121,
+      "step": 6925
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.155853271484375,
+      "learning_rate": 1.2170928436588295e-05,
+      "loss": 0.5876,
+      "step": 6926
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.711142539978027,
+      "learning_rate": 1.2168354213145702e-05,
+      "loss": 0.5739,
+      "step": 6927
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.95273208618164,
+      "learning_rate": 1.2165779989703107e-05,
+      "loss": 0.5317,
+      "step": 6928
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.308259010314941,
+      "learning_rate": 1.2163205766260512e-05,
+      "loss": 0.422,
+      "step": 6929
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.667654037475586,
+      "learning_rate": 1.2160631542817917e-05,
+      "loss": 0.4162,
+      "step": 6930
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.427236557006836,
+      "learning_rate": 1.2158057319375322e-05,
+      "loss": 0.4395,
+      "step": 6931
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.244489669799805,
+      "learning_rate": 1.2155483095932728e-05,
+      "loss": 0.4093,
+      "step": 6932
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.465315818786621,
+      "learning_rate": 1.2152908872490133e-05,
+      "loss": 0.4847,
+      "step": 6933
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.299603462219238,
+      "learning_rate": 1.2150334649047538e-05,
+      "loss": 0.627,
+      "step": 6934
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.604196548461914,
+      "learning_rate": 1.2147760425604943e-05,
+      "loss": 0.5017,
+      "step": 6935
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.398388862609863,
+      "learning_rate": 1.2145186202162346e-05,
+      "loss": 0.4372,
+      "step": 6936
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.70260238647461,
+      "learning_rate": 1.2142611978719753e-05,
+      "loss": 0.671,
+      "step": 6937
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.968806266784668,
+      "learning_rate": 1.2140037755277158e-05,
+      "loss": 0.3225,
+      "step": 6938
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.71995735168457,
+      "learning_rate": 1.2137463531834563e-05,
+      "loss": 0.5295,
+      "step": 6939
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.185986518859863,
+      "learning_rate": 1.2134889308391968e-05,
+      "loss": 0.4386,
+      "step": 6940
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.755692481994629,
+      "learning_rate": 1.2132315084949375e-05,
+      "loss": 0.8041,
+      "step": 6941
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.157242774963379,
+      "learning_rate": 1.212974086150678e-05,
+      "loss": 0.4863,
+      "step": 6942
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.33084487915039,
+      "learning_rate": 1.2127166638064185e-05,
+      "loss": 0.4258,
+      "step": 6943
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.067972183227539,
+      "learning_rate": 1.212459241462159e-05,
+      "loss": 0.6102,
+      "step": 6944
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.8798246383667,
+      "learning_rate": 1.2122018191178995e-05,
+      "loss": 0.5402,
+      "step": 6945
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.439143180847168,
+      "learning_rate": 1.21194439677364e-05,
+      "loss": 0.458,
+      "step": 6946
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.634234428405762,
+      "learning_rate": 1.2116869744293805e-05,
+      "loss": 0.415,
+      "step": 6947
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.790219306945801,
+      "learning_rate": 1.211429552085121e-05,
+      "loss": 0.4349,
+      "step": 6948
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.400364875793457,
+      "learning_rate": 1.2111721297408615e-05,
+      "loss": 0.5006,
+      "step": 6949
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.733245849609375,
+      "learning_rate": 1.210914707396602e-05,
+      "loss": 0.5136,
+      "step": 6950
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.066165924072266,
+      "learning_rate": 1.2106572850523426e-05,
+      "loss": 0.4303,
+      "step": 6951
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.95359992980957,
+      "learning_rate": 1.2103998627080831e-05,
+      "loss": 0.5602,
+      "step": 6952
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.627713203430176,
+      "learning_rate": 1.2101424403638236e-05,
+      "loss": 0.495,
+      "step": 6953
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.105578422546387,
+      "learning_rate": 1.2098850180195641e-05,
+      "loss": 0.5282,
+      "step": 6954
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.413445472717285,
+      "learning_rate": 1.2096275956753046e-05,
+      "loss": 0.3457,
+      "step": 6955
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.681183815002441,
+      "learning_rate": 1.2093701733310451e-05,
+      "loss": 0.4098,
+      "step": 6956
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.14177417755127,
+      "learning_rate": 1.2091127509867856e-05,
+      "loss": 0.6411,
+      "step": 6957
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.76486587524414,
+      "learning_rate": 1.2088553286425261e-05,
+      "loss": 0.6086,
+      "step": 6958
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.164863586425781,
+      "learning_rate": 1.2085979062982666e-05,
+      "loss": 0.5373,
+      "step": 6959
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.067230224609375,
+      "learning_rate": 1.2083404839540073e-05,
+      "loss": 0.5701,
+      "step": 6960
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.132375717163086,
+      "learning_rate": 1.2080830616097478e-05,
+      "loss": 0.5944,
+      "step": 6961
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.549464225769043,
+      "learning_rate": 1.2078256392654883e-05,
+      "loss": 0.5234,
+      "step": 6962
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.204045295715332,
+      "learning_rate": 1.2075682169212288e-05,
+      "loss": 0.673,
+      "step": 6963
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.438495635986328,
+      "learning_rate": 1.2073107945769693e-05,
+      "loss": 0.5943,
+      "step": 6964
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.335518836975098,
+      "learning_rate": 1.20705337223271e-05,
+      "loss": 0.4016,
+      "step": 6965
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.443399429321289,
+      "learning_rate": 1.2067959498884505e-05,
+      "loss": 0.4555,
+      "step": 6966
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.275962829589844,
+      "learning_rate": 1.2065385275441908e-05,
+      "loss": 0.6056,
+      "step": 6967
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.003582954406738,
+      "learning_rate": 1.2062811051999313e-05,
+      "loss": 0.4402,
+      "step": 6968
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.950246810913086,
+      "learning_rate": 1.2060236828556718e-05,
+      "loss": 0.5363,
+      "step": 6969
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.460467338562012,
+      "learning_rate": 1.2057662605114125e-05,
+      "loss": 0.5114,
+      "step": 6970
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.345368385314941,
+      "learning_rate": 1.205508838167153e-05,
+      "loss": 0.5689,
+      "step": 6971
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.838040351867676,
+      "learning_rate": 1.2052514158228934e-05,
+      "loss": 0.4341,
+      "step": 6972
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.360549926757812,
+      "learning_rate": 1.204993993478634e-05,
+      "loss": 0.4572,
+      "step": 6973
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.673266410827637,
+      "learning_rate": 1.2047365711343746e-05,
+      "loss": 0.8732,
+      "step": 6974
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.888762474060059,
+      "learning_rate": 1.2044791487901151e-05,
+      "loss": 0.6281,
+      "step": 6975
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.244769096374512,
+      "learning_rate": 1.2042217264458556e-05,
+      "loss": 0.7336,
+      "step": 6976
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.339118957519531,
+      "learning_rate": 1.203964304101596e-05,
+      "loss": 0.5565,
+      "step": 6977
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.494194030761719,
+      "learning_rate": 1.2037068817573364e-05,
+      "loss": 0.5927,
+      "step": 6978
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.338953018188477,
+      "learning_rate": 1.2034494594130771e-05,
+      "loss": 0.3203,
+      "step": 6979
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.837724685668945,
+      "learning_rate": 1.2031920370688176e-05,
+      "loss": 0.5107,
+      "step": 6980
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.4344482421875,
+      "learning_rate": 1.2029346147245581e-05,
+      "loss": 0.4333,
+      "step": 6981
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.859993934631348,
+      "learning_rate": 1.2026771923802986e-05,
+      "loss": 0.6073,
+      "step": 6982
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.880913734436035,
+      "learning_rate": 1.2024197700360391e-05,
+      "loss": 0.7665,
+      "step": 6983
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.9858880043029785,
+      "learning_rate": 1.2021623476917798e-05,
+      "loss": 0.3886,
+      "step": 6984
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.160512924194336,
+      "learning_rate": 1.2019049253475203e-05,
+      "loss": 0.3861,
+      "step": 6985
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.845983505249023,
+      "learning_rate": 1.2016475030032608e-05,
+      "loss": 0.712,
+      "step": 6986
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.091763496398926,
+      "learning_rate": 1.2013900806590013e-05,
+      "loss": 0.4011,
+      "step": 6987
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.557726860046387,
+      "learning_rate": 1.2011326583147416e-05,
+      "loss": 0.471,
+      "step": 6988
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.745680809020996,
+      "learning_rate": 1.2008752359704823e-05,
+      "loss": 0.4615,
+      "step": 6989
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.776237487792969,
+      "learning_rate": 1.2006178136262228e-05,
+      "loss": 0.8105,
+      "step": 6990
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.352907180786133,
+      "learning_rate": 1.2003603912819633e-05,
+      "loss": 0.5183,
+      "step": 6991
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.638228416442871,
+      "learning_rate": 1.2001029689377038e-05,
+      "loss": 0.5895,
+      "step": 6992
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.045228004455566,
+      "learning_rate": 1.1998455465934444e-05,
+      "loss": 0.6126,
+      "step": 6993
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.426182746887207,
+      "learning_rate": 1.199588124249185e-05,
+      "loss": 0.574,
+      "step": 6994
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.595043659210205,
+      "learning_rate": 1.1993307019049254e-05,
+      "loss": 0.4957,
+      "step": 6995
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.870427131652832,
+      "learning_rate": 1.199073279560666e-05,
+      "loss": 0.4357,
+      "step": 6996
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.806071758270264,
+      "learning_rate": 1.1988158572164064e-05,
+      "loss": 0.3712,
+      "step": 6997
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.191230773925781,
+      "learning_rate": 1.198558434872147e-05,
+      "loss": 0.5265,
+      "step": 6998
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.952571868896484,
+      "learning_rate": 1.1983010125278874e-05,
+      "loss": 0.6739,
+      "step": 6999
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.765110969543457,
+      "learning_rate": 1.198043590183628e-05,
+      "loss": 0.5615,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.294563293457031,
+      "learning_rate": 1.1977861678393684e-05,
+      "loss": 0.2494,
+      "step": 7001
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.038796424865723,
+      "learning_rate": 1.197528745495109e-05,
+      "loss": 0.7038,
+      "step": 7002
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.618143081665039,
+      "learning_rate": 1.1972713231508496e-05,
+      "loss": 0.4741,
+      "step": 7003
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.299175262451172,
+      "learning_rate": 1.1970139008065901e-05,
+      "loss": 0.5161,
+      "step": 7004
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.32515811920166,
+      "learning_rate": 1.1967564784623306e-05,
+      "loss": 0.6179,
+      "step": 7005
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.920764923095703,
+      "learning_rate": 1.196499056118071e-05,
+      "loss": 0.542,
+      "step": 7006
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.704802513122559,
+      "learning_rate": 1.1962416337738116e-05,
+      "loss": 0.5491,
+      "step": 7007
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.108997344970703,
+      "learning_rate": 1.195984211429552e-05,
+      "loss": 0.4458,
+      "step": 7008
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.183677673339844,
+      "learning_rate": 1.1957267890852926e-05,
+      "loss": 0.3906,
+      "step": 7009
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.915328979492188,
+      "learning_rate": 1.195469366741033e-05,
+      "loss": 0.4644,
+      "step": 7010
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.178936958312988,
+      "learning_rate": 1.1952119443967736e-05,
+      "loss": 0.3103,
+      "step": 7011
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.293974876403809,
+      "learning_rate": 1.1949545220525142e-05,
+      "loss": 0.41,
+      "step": 7012
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.273841857910156,
+      "learning_rate": 1.1946970997082547e-05,
+      "loss": 0.8585,
+      "step": 7013
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.48833465576172,
+      "learning_rate": 1.1944396773639952e-05,
+      "loss": 0.3737,
+      "step": 7014
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.675256729125977,
+      "learning_rate": 1.1941822550197357e-05,
+      "loss": 0.4264,
+      "step": 7015
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.285032272338867,
+      "learning_rate": 1.1939248326754762e-05,
+      "loss": 0.6321,
+      "step": 7016
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.52556037902832,
+      "learning_rate": 1.1936674103312169e-05,
+      "loss": 0.6268,
+      "step": 7017
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.654465675354004,
+      "learning_rate": 1.1934099879869574e-05,
+      "loss": 0.5037,
+      "step": 7018
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.316587448120117,
+      "learning_rate": 1.1931525656426977e-05,
+      "loss": 0.4142,
+      "step": 7019
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.695708274841309,
+      "learning_rate": 1.1928951432984382e-05,
+      "loss": 0.7036,
+      "step": 7020
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.179616928100586,
+      "learning_rate": 1.1926377209541787e-05,
+      "loss": 0.3866,
+      "step": 7021
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.021726608276367,
+      "learning_rate": 1.1923802986099194e-05,
+      "loss": 0.4903,
+      "step": 7022
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.754122734069824,
+      "learning_rate": 1.1921228762656599e-05,
+      "loss": 0.4308,
+      "step": 7023
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.344573020935059,
+      "learning_rate": 1.1918654539214004e-05,
+      "loss": 0.5596,
+      "step": 7024
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.760412693023682,
+      "learning_rate": 1.1916080315771409e-05,
+      "loss": 0.4381,
+      "step": 7025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.74301815032959,
+      "learning_rate": 1.1913506092328816e-05,
+      "loss": 0.3956,
+      "step": 7026
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.595786094665527,
+      "learning_rate": 1.191093186888622e-05,
+      "loss": 0.8368,
+      "step": 7027
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.126429557800293,
+      "learning_rate": 1.1908357645443626e-05,
+      "loss": 0.5201,
+      "step": 7028
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.102508544921875,
+      "learning_rate": 1.1905783422001029e-05,
+      "loss": 0.5917,
+      "step": 7029
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.898194313049316,
+      "learning_rate": 1.1903209198558434e-05,
+      "loss": 0.5386,
+      "step": 7030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.192670822143555,
+      "learning_rate": 1.190063497511584e-05,
+      "loss": 0.4813,
+      "step": 7031
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.713973999023438,
+      "learning_rate": 1.1898060751673246e-05,
+      "loss": 0.5942,
+      "step": 7032
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.153640747070312,
+      "learning_rate": 1.189548652823065e-05,
+      "loss": 0.5826,
+      "step": 7033
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.870328903198242,
+      "learning_rate": 1.1892912304788056e-05,
+      "loss": 0.4872,
+      "step": 7034
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.00422191619873,
+      "learning_rate": 1.189033808134546e-05,
+      "loss": 0.4963,
+      "step": 7035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.160073280334473,
+      "learning_rate": 1.1887763857902867e-05,
+      "loss": 0.3616,
+      "step": 7036
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.520591735839844,
+      "learning_rate": 1.1885189634460272e-05,
+      "loss": 0.6143,
+      "step": 7037
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.063573837280273,
+      "learning_rate": 1.1882615411017677e-05,
+      "loss": 0.4386,
+      "step": 7038
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.25976848602295,
+      "learning_rate": 1.1880041187575082e-05,
+      "loss": 0.4612,
+      "step": 7039
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.411417961120605,
+      "learning_rate": 1.1877466964132485e-05,
+      "loss": 0.5498,
+      "step": 7040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.323909759521484,
+      "learning_rate": 1.1874892740689892e-05,
+      "loss": 0.7059,
+      "step": 7041
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.794978141784668,
+      "learning_rate": 1.1872318517247297e-05,
+      "loss": 0.5785,
+      "step": 7042
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.585016250610352,
+      "learning_rate": 1.1869744293804702e-05,
+      "loss": 0.5028,
+      "step": 7043
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.982545852661133,
+      "learning_rate": 1.1867170070362107e-05,
+      "loss": 0.6212,
+      "step": 7044
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.801260948181152,
+      "learning_rate": 1.1864595846919514e-05,
+      "loss": 0.371,
+      "step": 7045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.951095581054688,
+      "learning_rate": 1.1862021623476919e-05,
+      "loss": 0.6263,
+      "step": 7046
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.839865684509277,
+      "learning_rate": 1.1859447400034324e-05,
+      "loss": 0.5179,
+      "step": 7047
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.950497627258301,
+      "learning_rate": 1.1856873176591729e-05,
+      "loss": 0.5286,
+      "step": 7048
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.106574058532715,
+      "learning_rate": 1.1854298953149134e-05,
+      "loss": 0.4785,
+      "step": 7049
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.89876937866211,
+      "learning_rate": 1.1851724729706539e-05,
+      "loss": 0.8721,
+      "step": 7050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.699302673339844,
+      "learning_rate": 1.1849150506263944e-05,
+      "loss": 0.578,
+      "step": 7051
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.155644416809082,
+      "learning_rate": 1.1846576282821349e-05,
+      "loss": 0.5321,
+      "step": 7052
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.794529914855957,
+      "learning_rate": 1.1844002059378754e-05,
+      "loss": 0.549,
+      "step": 7053
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.91904354095459,
+      "learning_rate": 1.1841427835936159e-05,
+      "loss": 0.5655,
+      "step": 7054
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.819161415100098,
+      "learning_rate": 1.1838853612493565e-05,
+      "loss": 0.4241,
+      "step": 7055
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.905237197875977,
+      "learning_rate": 1.183627938905097e-05,
+      "loss": 0.4594,
+      "step": 7056
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.802270889282227,
+      "learning_rate": 1.1833705165608375e-05,
+      "loss": 0.5523,
+      "step": 7057
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.213642120361328,
+      "learning_rate": 1.183113094216578e-05,
+      "loss": 0.7524,
+      "step": 7058
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.1716227531433105,
+      "learning_rate": 1.1828556718723185e-05,
+      "loss": 0.3411,
+      "step": 7059
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.224350929260254,
+      "learning_rate": 1.182598249528059e-05,
+      "loss": 0.3936,
+      "step": 7060
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.481537818908691,
+      "learning_rate": 1.1823408271837995e-05,
+      "loss": 0.5573,
+      "step": 7061
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.607392311096191,
+      "learning_rate": 1.18208340483954e-05,
+      "loss": 0.4878,
+      "step": 7062
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.242002487182617,
+      "learning_rate": 1.1818259824952805e-05,
+      "loss": 0.565,
+      "step": 7063
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.567395210266113,
+      "learning_rate": 1.1815685601510212e-05,
+      "loss": 0.5141,
+      "step": 7064
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.485620498657227,
+      "learning_rate": 1.1813111378067617e-05,
+      "loss": 0.4656,
+      "step": 7065
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.69052505493164,
+      "learning_rate": 1.1810537154625022e-05,
+      "loss": 0.483,
+      "step": 7066
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.210465431213379,
+      "learning_rate": 1.1807962931182427e-05,
+      "loss": 0.3763,
+      "step": 7067
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.8494234085083,
+      "learning_rate": 1.1805388707739832e-05,
+      "loss": 0.5078,
+      "step": 7068
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.431105613708496,
+      "learning_rate": 1.1802814484297239e-05,
+      "loss": 0.627,
+      "step": 7069
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.741544723510742,
+      "learning_rate": 1.1800240260854644e-05,
+      "loss": 0.7926,
+      "step": 7070
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.597745895385742,
+      "learning_rate": 1.1797666037412047e-05,
+      "loss": 0.4042,
+      "step": 7071
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.433899879455566,
+      "learning_rate": 1.1795091813969452e-05,
+      "loss": 0.5409,
+      "step": 7072
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.573753356933594,
+      "learning_rate": 1.1792517590526857e-05,
+      "loss": 0.5791,
+      "step": 7073
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.764334678649902,
+      "learning_rate": 1.1789943367084264e-05,
+      "loss": 0.7008,
+      "step": 7074
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.19047737121582,
+      "learning_rate": 1.1787369143641669e-05,
+      "loss": 0.7471,
+      "step": 7075
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.98177719116211,
+      "learning_rate": 1.1784794920199074e-05,
+      "loss": 0.4211,
+      "step": 7076
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.5682053565979,
+      "learning_rate": 1.1782220696756478e-05,
+      "loss": 0.3395,
+      "step": 7077
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.34294319152832,
+      "learning_rate": 1.1779646473313883e-05,
+      "loss": 0.4304,
+      "step": 7078
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.188375473022461,
+      "learning_rate": 1.177707224987129e-05,
+      "loss": 0.3686,
+      "step": 7079
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.635316848754883,
+      "learning_rate": 1.1774498026428695e-05,
+      "loss": 0.4264,
+      "step": 7080
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.043983459472656,
+      "learning_rate": 1.1771923802986098e-05,
+      "loss": 0.5094,
+      "step": 7081
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.92259693145752,
+      "learning_rate": 1.1769349579543503e-05,
+      "loss": 0.4642,
+      "step": 7082
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.077810287475586,
+      "learning_rate": 1.176677535610091e-05,
+      "loss": 0.6337,
+      "step": 7083
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.101829528808594,
+      "learning_rate": 1.1764201132658315e-05,
+      "loss": 0.6638,
+      "step": 7084
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.365436553955078,
+      "learning_rate": 1.176162690921572e-05,
+      "loss": 0.5677,
+      "step": 7085
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.324374198913574,
+      "learning_rate": 1.1759052685773125e-05,
+      "loss": 0.5725,
+      "step": 7086
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.412105560302734,
+      "learning_rate": 1.175647846233053e-05,
+      "loss": 0.5346,
+      "step": 7087
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.5609769821167,
+      "learning_rate": 1.1753904238887937e-05,
+      "loss": 0.7969,
+      "step": 7088
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.079243659973145,
+      "learning_rate": 1.1751330015445342e-05,
+      "loss": 0.524,
+      "step": 7089
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.235267639160156,
+      "learning_rate": 1.1748755792002747e-05,
+      "loss": 0.6835,
+      "step": 7090
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.892448425292969,
+      "learning_rate": 1.1746181568560152e-05,
+      "loss": 0.5749,
+      "step": 7091
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.017757415771484,
+      "learning_rate": 1.1743607345117555e-05,
+      "loss": 0.5958,
+      "step": 7092
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.532471656799316,
+      "learning_rate": 1.1741033121674962e-05,
+      "loss": 0.5163,
+      "step": 7093
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.933429718017578,
+      "learning_rate": 1.1738458898232367e-05,
+      "loss": 0.4198,
+      "step": 7094
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.935221672058105,
+      "learning_rate": 1.1735884674789772e-05,
+      "loss": 0.4781,
+      "step": 7095
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.14262580871582,
+      "learning_rate": 1.1733310451347177e-05,
+      "loss": 0.4835,
+      "step": 7096
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.401575088500977,
+      "learning_rate": 1.1730736227904583e-05,
+      "loss": 0.6201,
+      "step": 7097
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.443571090698242,
+      "learning_rate": 1.1728162004461988e-05,
+      "loss": 0.4969,
+      "step": 7098
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.579925537109375,
+      "learning_rate": 1.1725587781019393e-05,
+      "loss": 0.4178,
+      "step": 7099
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.502985954284668,
+      "learning_rate": 1.1723013557576798e-05,
+      "loss": 0.5585,
+      "step": 7100
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.74077844619751,
+      "learning_rate": 1.1720439334134203e-05,
+      "loss": 0.5048,
+      "step": 7101
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.656698703765869,
+      "learning_rate": 1.1717865110691608e-05,
+      "loss": 0.3651,
+      "step": 7102
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.099870681762695,
+      "learning_rate": 1.1715290887249013e-05,
+      "loss": 0.4139,
+      "step": 7103
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.223795890808105,
+      "learning_rate": 1.1712716663806418e-05,
+      "loss": 0.4273,
+      "step": 7104
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.487719535827637,
+      "learning_rate": 1.1710142440363823e-05,
+      "loss": 0.8251,
+      "step": 7105
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.5261359214782715,
+      "learning_rate": 1.1707568216921228e-05,
+      "loss": 0.3658,
+      "step": 7106
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.580143451690674,
+      "learning_rate": 1.1704993993478635e-05,
+      "loss": 0.3119,
+      "step": 7107
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.47298526763916,
+      "learning_rate": 1.170241977003604e-05,
+      "loss": 0.6249,
+      "step": 7108
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.980270385742188,
+      "learning_rate": 1.1699845546593445e-05,
+      "loss": 0.5136,
+      "step": 7109
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.487540245056152,
+      "learning_rate": 1.169727132315085e-05,
+      "loss": 0.4227,
+      "step": 7110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.91495418548584,
+      "learning_rate": 1.1694697099708255e-05,
+      "loss": 0.528,
+      "step": 7111
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.30009651184082,
+      "learning_rate": 1.1692122876265662e-05,
+      "loss": 0.5401,
+      "step": 7112
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.5504231452941895,
+      "learning_rate": 1.1689548652823065e-05,
+      "loss": 0.5919,
+      "step": 7113
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.687311172485352,
+      "learning_rate": 1.168697442938047e-05,
+      "loss": 0.4475,
+      "step": 7114
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.62994384765625,
+      "learning_rate": 1.1684400205937875e-05,
+      "loss": 0.3877,
+      "step": 7115
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.701693534851074,
+      "learning_rate": 1.1681825982495281e-05,
+      "loss": 0.4388,
+      "step": 7116
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.731943130493164,
+      "learning_rate": 1.1679251759052686e-05,
+      "loss": 0.5003,
+      "step": 7117
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.448636054992676,
+      "learning_rate": 1.1676677535610091e-05,
+      "loss": 0.6091,
+      "step": 7118
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.859493255615234,
+      "learning_rate": 1.1674103312167496e-05,
+      "loss": 0.4222,
+      "step": 7119
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.540654182434082,
+      "learning_rate": 1.1671529088724901e-05,
+      "loss": 0.4202,
+      "step": 7120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.63926887512207,
+      "learning_rate": 1.1668954865282308e-05,
+      "loss": 0.4924,
+      "step": 7121
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.28764533996582,
+      "learning_rate": 1.1666380641839713e-05,
+      "loss": 0.3627,
+      "step": 7122
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.108075141906738,
+      "learning_rate": 1.1663806418397116e-05,
+      "loss": 0.4352,
+      "step": 7123
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.017898559570312,
+      "learning_rate": 1.1661232194954521e-05,
+      "loss": 0.404,
+      "step": 7124
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.9288740158081055,
+      "learning_rate": 1.1658657971511926e-05,
+      "loss": 0.3302,
+      "step": 7125
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.391546249389648,
+      "learning_rate": 1.1656083748069333e-05,
+      "loss": 0.6349,
+      "step": 7126
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.188262462615967,
+      "learning_rate": 1.1653509524626738e-05,
+      "loss": 0.2733,
+      "step": 7127
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.3939266204834,
+      "learning_rate": 1.1650935301184143e-05,
+      "loss": 0.6341,
+      "step": 7128
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.814071655273438,
+      "learning_rate": 1.1648361077741548e-05,
+      "loss": 0.6084,
+      "step": 7129
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.20941162109375,
+      "learning_rate": 1.1645786854298953e-05,
+      "loss": 0.3172,
+      "step": 7130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.053966522216797,
+      "learning_rate": 1.164321263085636e-05,
+      "loss": 0.562,
+      "step": 7131
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.195056915283203,
+      "learning_rate": 1.1640638407413765e-05,
+      "loss": 0.4252,
+      "step": 7132
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.162524223327637,
+      "learning_rate": 1.1638064183971168e-05,
+      "loss": 0.4871,
+      "step": 7133
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.25905704498291,
+      "learning_rate": 1.1635489960528573e-05,
+      "loss": 0.5897,
+      "step": 7134
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.6315279006958,
+      "learning_rate": 1.163291573708598e-05,
+      "loss": 0.3175,
+      "step": 7135
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.307784080505371,
+      "learning_rate": 1.1630341513643385e-05,
+      "loss": 0.53,
+      "step": 7136
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.675056457519531,
+      "learning_rate": 1.162776729020079e-05,
+      "loss": 0.7125,
+      "step": 7137
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.297932624816895,
+      "learning_rate": 1.1625193066758195e-05,
+      "loss": 0.6564,
+      "step": 7138
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.99775505065918,
+      "learning_rate": 1.16226188433156e-05,
+      "loss": 0.5808,
+      "step": 7139
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.949821949005127,
+      "learning_rate": 1.1620044619873006e-05,
+      "loss": 0.3286,
+      "step": 7140
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.827123641967773,
+      "learning_rate": 1.1617470396430411e-05,
+      "loss": 0.4466,
+      "step": 7141
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.465693473815918,
+      "learning_rate": 1.1614896172987816e-05,
+      "loss": 0.4905,
+      "step": 7142
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.595473289489746,
+      "learning_rate": 1.1612321949545221e-05,
+      "loss": 0.3843,
+      "step": 7143
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.307246208190918,
+      "learning_rate": 1.1609747726102625e-05,
+      "loss": 0.6517,
+      "step": 7144
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.332794189453125,
+      "learning_rate": 1.1607173502660031e-05,
+      "loss": 0.3841,
+      "step": 7145
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.55783748626709,
+      "learning_rate": 1.1604599279217436e-05,
+      "loss": 0.6285,
+      "step": 7146
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.59277057647705,
+      "learning_rate": 1.1602025055774841e-05,
+      "loss": 0.4413,
+      "step": 7147
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.730266571044922,
+      "learning_rate": 1.1599450832332246e-05,
+      "loss": 0.5179,
+      "step": 7148
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.36196231842041,
+      "learning_rate": 1.1596876608889653e-05,
+      "loss": 0.5347,
+      "step": 7149
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.165251731872559,
+      "learning_rate": 1.1594302385447058e-05,
+      "loss": 0.4246,
+      "step": 7150
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.816166877746582,
+      "learning_rate": 1.1591728162004463e-05,
+      "loss": 0.7186,
+      "step": 7151
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.423781394958496,
+      "learning_rate": 1.1589153938561868e-05,
+      "loss": 0.4496,
+      "step": 7152
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.16259479522705,
+      "learning_rate": 1.1586579715119273e-05,
+      "loss": 0.4833,
+      "step": 7153
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.660272598266602,
+      "learning_rate": 1.1584005491676678e-05,
+      "loss": 0.422,
+      "step": 7154
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.854595184326172,
+      "learning_rate": 1.1581431268234083e-05,
+      "loss": 0.4919,
+      "step": 7155
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.222434997558594,
+      "learning_rate": 1.1578857044791488e-05,
+      "loss": 0.5104,
+      "step": 7156
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.050089836120605,
+      "learning_rate": 1.1576282821348893e-05,
+      "loss": 0.4817,
+      "step": 7157
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.66711711883545,
+      "learning_rate": 1.1573708597906298e-05,
+      "loss": 0.4955,
+      "step": 7158
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.91462230682373,
+      "learning_rate": 1.1571134374463704e-05,
+      "loss": 0.481,
+      "step": 7159
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.339996337890625,
+      "learning_rate": 1.156856015102111e-05,
+      "loss": 0.4842,
+      "step": 7160
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.144424438476562,
+      "learning_rate": 1.1565985927578514e-05,
+      "loss": 0.5748,
+      "step": 7161
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.7717866897583,
+      "learning_rate": 1.156341170413592e-05,
+      "loss": 0.4264,
+      "step": 7162
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.315986633300781,
+      "learning_rate": 1.1560837480693324e-05,
+      "loss": 0.5858,
+      "step": 7163
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.716745376586914,
+      "learning_rate": 1.1558263257250731e-05,
+      "loss": 0.3456,
+      "step": 7164
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.27481746673584,
+      "learning_rate": 1.1555689033808134e-05,
+      "loss": 0.616,
+      "step": 7165
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.431517601013184,
+      "learning_rate": 1.155311481036554e-05,
+      "loss": 0.5691,
+      "step": 7166
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.07856559753418,
+      "learning_rate": 1.1550540586922944e-05,
+      "loss": 0.4524,
+      "step": 7167
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.139068603515625,
+      "learning_rate": 1.1547966363480351e-05,
+      "loss": 0.6126,
+      "step": 7168
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.351325035095215,
+      "learning_rate": 1.1545392140037756e-05,
+      "loss": 0.6218,
+      "step": 7169
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.678258895874023,
+      "learning_rate": 1.1542817916595161e-05,
+      "loss": 0.5069,
+      "step": 7170
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.375031471252441,
+      "learning_rate": 1.1540243693152566e-05,
+      "loss": 0.3749,
+      "step": 7171
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.02367115020752,
+      "learning_rate": 1.1537669469709971e-05,
+      "loss": 0.5074,
+      "step": 7172
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.382370948791504,
+      "learning_rate": 1.1535095246267378e-05,
+      "loss": 0.5201,
+      "step": 7173
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.980612754821777,
+      "learning_rate": 1.1532521022824783e-05,
+      "loss": 0.4341,
+      "step": 7174
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.573546409606934,
+      "learning_rate": 1.1529946799382186e-05,
+      "loss": 0.4113,
+      "step": 7175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.847718238830566,
+      "learning_rate": 1.1527372575939591e-05,
+      "loss": 0.4756,
+      "step": 7176
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.996875762939453,
+      "learning_rate": 1.1524798352496996e-05,
+      "loss": 0.5359,
+      "step": 7177
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.408705711364746,
+      "learning_rate": 1.1522224129054403e-05,
+      "loss": 0.455,
+      "step": 7178
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.923229217529297,
+      "learning_rate": 1.1519649905611808e-05,
+      "loss": 0.5556,
+      "step": 7179
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.602019309997559,
+      "learning_rate": 1.1517075682169213e-05,
+      "loss": 0.75,
+      "step": 7180
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.265124320983887,
+      "learning_rate": 1.1514501458726618e-05,
+      "loss": 0.5565,
+      "step": 7181
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.500020980834961,
+      "learning_rate": 1.1511927235284022e-05,
+      "loss": 0.5501,
+      "step": 7182
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.173694610595703,
+      "learning_rate": 1.150935301184143e-05,
+      "loss": 0.4323,
+      "step": 7183
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.560260772705078,
+      "learning_rate": 1.1506778788398834e-05,
+      "loss": 0.4102,
+      "step": 7184
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.73781967163086,
+      "learning_rate": 1.1504204564956237e-05,
+      "loss": 0.5474,
+      "step": 7185
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.796578407287598,
+      "learning_rate": 1.1501630341513642e-05,
+      "loss": 0.5905,
+      "step": 7186
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.797844886779785,
+      "learning_rate": 1.1499056118071049e-05,
+      "loss": 0.521,
+      "step": 7187
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.736172676086426,
+      "learning_rate": 1.1496481894628454e-05,
+      "loss": 0.5932,
+      "step": 7188
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.909501075744629,
+      "learning_rate": 1.1493907671185859e-05,
+      "loss": 0.5309,
+      "step": 7189
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.898690223693848,
+      "learning_rate": 1.1491333447743264e-05,
+      "loss": 0.4736,
+      "step": 7190
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.287887573242188,
+      "learning_rate": 1.1488759224300669e-05,
+      "loss": 0.5086,
+      "step": 7191
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.551624298095703,
+      "learning_rate": 1.1486185000858076e-05,
+      "loss": 0.4678,
+      "step": 7192
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.073768615722656,
+      "learning_rate": 1.148361077741548e-05,
+      "loss": 0.5858,
+      "step": 7193
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.796259880065918,
+      "learning_rate": 1.1481036553972886e-05,
+      "loss": 0.6086,
+      "step": 7194
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.341981887817383,
+      "learning_rate": 1.147846233053029e-05,
+      "loss": 0.349,
+      "step": 7195
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.76344633102417,
+      "learning_rate": 1.1475888107087694e-05,
+      "loss": 0.3582,
+      "step": 7196
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.240875244140625,
+      "learning_rate": 1.14733138836451e-05,
+      "loss": 0.4038,
+      "step": 7197
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.505542755126953,
+      "learning_rate": 1.1470739660202506e-05,
+      "loss": 0.4493,
+      "step": 7198
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.862126350402832,
+      "learning_rate": 1.146816543675991e-05,
+      "loss": 0.3836,
+      "step": 7199
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.115520477294922,
+      "learning_rate": 1.1465591213317316e-05,
+      "loss": 0.4824,
+      "step": 7200
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.578757286071777,
+      "learning_rate": 1.1463016989874722e-05,
+      "loss": 0.5646,
+      "step": 7201
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.81511116027832,
+      "learning_rate": 1.1460442766432127e-05,
+      "loss": 0.5767,
+      "step": 7202
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.9080810546875,
+      "learning_rate": 1.1457868542989532e-05,
+      "loss": 0.589,
+      "step": 7203
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.988289833068848,
+      "learning_rate": 1.1455294319546937e-05,
+      "loss": 0.3225,
+      "step": 7204
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.365975379943848,
+      "learning_rate": 1.1452720096104342e-05,
+      "loss": 0.6056,
+      "step": 7205
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.194970607757568,
+      "learning_rate": 1.1450145872661747e-05,
+      "loss": 0.3617,
+      "step": 7206
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.30859661102295,
+      "learning_rate": 1.1447571649219152e-05,
+      "loss": 0.5695,
+      "step": 7207
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.60885238647461,
+      "learning_rate": 1.1444997425776557e-05,
+      "loss": 0.5869,
+      "step": 7208
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.791413307189941,
+      "learning_rate": 1.1442423202333962e-05,
+      "loss": 0.6055,
+      "step": 7209
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.433341026306152,
+      "learning_rate": 1.1439848978891367e-05,
+      "loss": 0.6103,
+      "step": 7210
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.403416633605957,
+      "learning_rate": 1.1437274755448774e-05,
+      "loss": 0.395,
+      "step": 7211
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.142620086669922,
+      "learning_rate": 1.1434700532006179e-05,
+      "loss": 0.5486,
+      "step": 7212
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.775880813598633,
+      "learning_rate": 1.1432126308563584e-05,
+      "loss": 0.5146,
+      "step": 7213
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.822942733764648,
+      "learning_rate": 1.1429552085120989e-05,
+      "loss": 0.5121,
+      "step": 7214
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.373087882995605,
+      "learning_rate": 1.1426977861678394e-05,
+      "loss": 0.3958,
+      "step": 7215
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.169090270996094,
+      "learning_rate": 1.14244036382358e-05,
+      "loss": 0.3961,
+      "step": 7216
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.584774017333984,
+      "learning_rate": 1.1421829414793204e-05,
+      "loss": 0.441,
+      "step": 7217
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.406041145324707,
+      "learning_rate": 1.1419255191350609e-05,
+      "loss": 0.7111,
+      "step": 7218
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.626038551330566,
+      "learning_rate": 1.1416680967908014e-05,
+      "loss": 0.6714,
+      "step": 7219
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.445282936096191,
+      "learning_rate": 1.141410674446542e-05,
+      "loss": 0.4856,
+      "step": 7220
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.321600914001465,
+      "learning_rate": 1.1411532521022825e-05,
+      "loss": 0.356,
+      "step": 7221
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.86656379699707,
+      "learning_rate": 1.140895829758023e-05,
+      "loss": 0.3288,
+      "step": 7222
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.213930130004883,
+      "learning_rate": 1.1406384074137635e-05,
+      "loss": 0.3596,
+      "step": 7223
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.23892879486084,
+      "learning_rate": 1.140380985069504e-05,
+      "loss": 0.4785,
+      "step": 7224
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.818760871887207,
+      "learning_rate": 1.1401235627252447e-05,
+      "loss": 0.5261,
+      "step": 7225
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.000699043273926,
+      "learning_rate": 1.1398661403809852e-05,
+      "loss": 0.4484,
+      "step": 7226
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.705554962158203,
+      "learning_rate": 1.1396087180367255e-05,
+      "loss": 0.4674,
+      "step": 7227
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.233939170837402,
+      "learning_rate": 1.139351295692466e-05,
+      "loss": 0.4736,
+      "step": 7228
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.031034469604492,
+      "learning_rate": 1.1390938733482065e-05,
+      "loss": 0.4833,
+      "step": 7229
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.996298789978027,
+      "learning_rate": 1.1388364510039472e-05,
+      "loss": 0.5249,
+      "step": 7230
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.602757453918457,
+      "learning_rate": 1.1385790286596877e-05,
+      "loss": 0.6922,
+      "step": 7231
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.56598949432373,
+      "learning_rate": 1.1383216063154282e-05,
+      "loss": 0.4054,
+      "step": 7232
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.37939453125,
+      "learning_rate": 1.1380641839711687e-05,
+      "loss": 0.5647,
+      "step": 7233
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.546008110046387,
+      "learning_rate": 1.1378067616269092e-05,
+      "loss": 0.35,
+      "step": 7234
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.051502227783203,
+      "learning_rate": 1.1375493392826499e-05,
+      "loss": 0.5013,
+      "step": 7235
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.882331848144531,
+      "learning_rate": 1.1372919169383904e-05,
+      "loss": 0.4023,
+      "step": 7236
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.841806411743164,
+      "learning_rate": 1.1370344945941307e-05,
+      "loss": 0.5238,
+      "step": 7237
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.044391632080078,
+      "learning_rate": 1.1367770722498712e-05,
+      "loss": 0.4652,
+      "step": 7238
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.846569061279297,
+      "learning_rate": 1.1365196499056119e-05,
+      "loss": 0.606,
+      "step": 7239
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.960699081420898,
+      "learning_rate": 1.1362622275613524e-05,
+      "loss": 0.4853,
+      "step": 7240
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.793228149414062,
+      "learning_rate": 1.1360048052170929e-05,
+      "loss": 0.4019,
+      "step": 7241
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.172778129577637,
+      "learning_rate": 1.1357473828728334e-05,
+      "loss": 0.5707,
+      "step": 7242
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.804939270019531,
+      "learning_rate": 1.1354899605285739e-05,
+      "loss": 0.3863,
+      "step": 7243
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.950481414794922,
+      "learning_rate": 1.1352325381843145e-05,
+      "loss": 0.4967,
+      "step": 7244
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.23788833618164,
+      "learning_rate": 1.134975115840055e-05,
+      "loss": 0.4025,
+      "step": 7245
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.630895614624023,
+      "learning_rate": 1.1347176934957955e-05,
+      "loss": 0.3437,
+      "step": 7246
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.612194061279297,
+      "learning_rate": 1.134460271151536e-05,
+      "loss": 0.553,
+      "step": 7247
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.624996185302734,
+      "learning_rate": 1.1342028488072764e-05,
+      "loss": 0.5355,
+      "step": 7248
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.33304500579834,
+      "learning_rate": 1.133945426463017e-05,
+      "loss": 0.5072,
+      "step": 7249
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.467058181762695,
+      "learning_rate": 1.1336880041187575e-05,
+      "loss": 0.3252,
+      "step": 7250
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.657370090484619,
+      "learning_rate": 1.133430581774498e-05,
+      "loss": 0.499,
+      "step": 7251
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.748489379882812,
+      "learning_rate": 1.1331731594302385e-05,
+      "loss": 0.4994,
+      "step": 7252
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.288755416870117,
+      "learning_rate": 1.132915737085979e-05,
+      "loss": 0.4601,
+      "step": 7253
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.3094482421875,
+      "learning_rate": 1.1326583147417197e-05,
+      "loss": 0.5718,
+      "step": 7254
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.617788314819336,
+      "learning_rate": 1.1324008923974602e-05,
+      "loss": 0.4283,
+      "step": 7255
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.05827808380127,
+      "learning_rate": 1.1321434700532007e-05,
+      "loss": 0.5677,
+      "step": 7256
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.439905166625977,
+      "learning_rate": 1.1318860477089412e-05,
+      "loss": 0.441,
+      "step": 7257
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.854427337646484,
+      "learning_rate": 1.1316286253646817e-05,
+      "loss": 0.419,
+      "step": 7258
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.072861671447754,
+      "learning_rate": 1.1313712030204222e-05,
+      "loss": 0.4099,
+      "step": 7259
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.507295608520508,
+      "learning_rate": 1.1311137806761627e-05,
+      "loss": 0.5666,
+      "step": 7260
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.564096450805664,
+      "learning_rate": 1.1308563583319032e-05,
+      "loss": 0.4455,
+      "step": 7261
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.835600852966309,
+      "learning_rate": 1.1305989359876437e-05,
+      "loss": 0.5052,
+      "step": 7262
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.7724809646606445,
+      "learning_rate": 1.1303415136433843e-05,
+      "loss": 0.514,
+      "step": 7263
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.510541915893555,
+      "learning_rate": 1.1300840912991248e-05,
+      "loss": 0.5399,
+      "step": 7264
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.288419723510742,
+      "learning_rate": 1.1298266689548653e-05,
+      "loss": 0.5107,
+      "step": 7265
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.084488868713379,
+      "learning_rate": 1.1295692466106058e-05,
+      "loss": 0.4922,
+      "step": 7266
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.851661682128906,
+      "learning_rate": 1.1293118242663463e-05,
+      "loss": 0.5858,
+      "step": 7267
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.195343017578125,
+      "learning_rate": 1.129054401922087e-05,
+      "loss": 0.6452,
+      "step": 7268
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.215805053710938,
+      "learning_rate": 1.1287969795778273e-05,
+      "loss": 0.4136,
+      "step": 7269
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.631438255310059,
+      "learning_rate": 1.1285395572335678e-05,
+      "loss": 0.4646,
+      "step": 7270
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.711803436279297,
+      "learning_rate": 1.1282821348893083e-05,
+      "loss": 0.3928,
+      "step": 7271
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.507469177246094,
+      "learning_rate": 1.128024712545049e-05,
+      "loss": 0.5241,
+      "step": 7272
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.972332000732422,
+      "learning_rate": 1.1277672902007895e-05,
+      "loss": 0.4817,
+      "step": 7273
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.054027557373047,
+      "learning_rate": 1.12750986785653e-05,
+      "loss": 0.7787,
+      "step": 7274
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.904644012451172,
+      "learning_rate": 1.1272524455122705e-05,
+      "loss": 0.5479,
+      "step": 7275
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.885504722595215,
+      "learning_rate": 1.126995023168011e-05,
+      "loss": 0.4343,
+      "step": 7276
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.406169891357422,
+      "learning_rate": 1.1267376008237517e-05,
+      "loss": 0.5582,
+      "step": 7277
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.902565002441406,
+      "learning_rate": 1.1264801784794922e-05,
+      "loss": 0.5416,
+      "step": 7278
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.308808326721191,
+      "learning_rate": 1.1262227561352325e-05,
+      "loss": 0.5342,
+      "step": 7279
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.628055572509766,
+      "learning_rate": 1.125965333790973e-05,
+      "loss": 0.701,
+      "step": 7280
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.303853034973145,
+      "learning_rate": 1.1257079114467135e-05,
+      "loss": 0.6888,
+      "step": 7281
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.253119468688965,
+      "learning_rate": 1.1254504891024542e-05,
+      "loss": 0.4342,
+      "step": 7282
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.1643648147583,
+      "learning_rate": 1.1251930667581947e-05,
+      "loss": 0.3325,
+      "step": 7283
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.911738395690918,
+      "learning_rate": 1.1249356444139352e-05,
+      "loss": 0.5293,
+      "step": 7284
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.523740768432617,
+      "learning_rate": 1.1246782220696757e-05,
+      "loss": 0.4431,
+      "step": 7285
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.141708374023438,
+      "learning_rate": 1.1244207997254162e-05,
+      "loss": 0.604,
+      "step": 7286
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.976603507995605,
+      "learning_rate": 1.1241633773811568e-05,
+      "loss": 0.4651,
+      "step": 7287
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.72829818725586,
+      "learning_rate": 1.1239059550368973e-05,
+      "loss": 0.6487,
+      "step": 7288
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.384855270385742,
+      "learning_rate": 1.1236485326926376e-05,
+      "loss": 0.5895,
+      "step": 7289
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.792872428894043,
+      "learning_rate": 1.1233911103483781e-05,
+      "loss": 0.4809,
+      "step": 7290
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.717025756835938,
+      "learning_rate": 1.1231336880041188e-05,
+      "loss": 0.4717,
+      "step": 7291
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.081117630004883,
+      "learning_rate": 1.1228762656598593e-05,
+      "loss": 0.4903,
+      "step": 7292
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.190896034240723,
+      "learning_rate": 1.1226188433155998e-05,
+      "loss": 0.4885,
+      "step": 7293
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.908560752868652,
+      "learning_rate": 1.1223614209713403e-05,
+      "loss": 0.4172,
+      "step": 7294
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.686976909637451,
+      "learning_rate": 1.1221039986270808e-05,
+      "loss": 0.3909,
+      "step": 7295
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.062314987182617,
+      "learning_rate": 1.1218465762828215e-05,
+      "loss": 0.5018,
+      "step": 7296
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.092988014221191,
+      "learning_rate": 1.121589153938562e-05,
+      "loss": 0.4151,
+      "step": 7297
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.299968719482422,
+      "learning_rate": 1.1213317315943025e-05,
+      "loss": 0.6049,
+      "step": 7298
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.324273109436035,
+      "learning_rate": 1.121074309250043e-05,
+      "loss": 0.5683,
+      "step": 7299
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.269415855407715,
+      "learning_rate": 1.1208168869057833e-05,
+      "loss": 0.4494,
+      "step": 7300
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.625930786132812,
+      "learning_rate": 1.120559464561524e-05,
+      "loss": 0.651,
+      "step": 7301
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.52757740020752,
+      "learning_rate": 1.1203020422172645e-05,
+      "loss": 0.4696,
+      "step": 7302
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.25619125366211,
+      "learning_rate": 1.120044619873005e-05,
+      "loss": 0.578,
+      "step": 7303
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.537073135375977,
+      "learning_rate": 1.1197871975287455e-05,
+      "loss": 0.4884,
+      "step": 7304
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.265997886657715,
+      "learning_rate": 1.119529775184486e-05,
+      "loss": 0.3923,
+      "step": 7305
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.996153831481934,
+      "learning_rate": 1.1192723528402266e-05,
+      "loss": 0.4961,
+      "step": 7306
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.89450740814209,
+      "learning_rate": 1.1190149304959671e-05,
+      "loss": 0.6458,
+      "step": 7307
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.06220817565918,
+      "learning_rate": 1.1187575081517076e-05,
+      "loss": 0.6208,
+      "step": 7308
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.924788475036621,
+      "learning_rate": 1.1185000858074481e-05,
+      "loss": 0.412,
+      "step": 7309
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.411330223083496,
+      "learning_rate": 1.1182426634631886e-05,
+      "loss": 0.629,
+      "step": 7310
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.121463775634766,
+      "learning_rate": 1.1179852411189291e-05,
+      "loss": 0.4451,
+      "step": 7311
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.201952934265137,
+      "learning_rate": 1.1177278187746696e-05,
+      "loss": 0.4098,
+      "step": 7312
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.90984058380127,
+      "learning_rate": 1.1174703964304101e-05,
+      "loss": 0.4821,
+      "step": 7313
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.660990715026855,
+      "learning_rate": 1.1172129740861506e-05,
+      "loss": 0.3648,
+      "step": 7314
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.568854331970215,
+      "learning_rate": 1.1169555517418913e-05,
+      "loss": 0.4303,
+      "step": 7315
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.792206764221191,
+      "learning_rate": 1.1166981293976318e-05,
+      "loss": 0.5768,
+      "step": 7316
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.611165523529053,
+      "learning_rate": 1.1164407070533723e-05,
+      "loss": 0.3432,
+      "step": 7317
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.740774154663086,
+      "learning_rate": 1.1161832847091128e-05,
+      "loss": 0.2775,
+      "step": 7318
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.653419494628906,
+      "learning_rate": 1.1159258623648533e-05,
+      "loss": 0.3901,
+      "step": 7319
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.659016609191895,
+      "learning_rate": 1.115668440020594e-05,
+      "loss": 0.4967,
+      "step": 7320
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.037394523620605,
+      "learning_rate": 1.1154110176763343e-05,
+      "loss": 0.59,
+      "step": 7321
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.336116790771484,
+      "learning_rate": 1.1151535953320748e-05,
+      "loss": 0.539,
+      "step": 7322
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.488786697387695,
+      "learning_rate": 1.1148961729878153e-05,
+      "loss": 0.3024,
+      "step": 7323
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.387650489807129,
+      "learning_rate": 1.114638750643556e-05,
+      "loss": 0.4917,
+      "step": 7324
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.015610694885254,
+      "learning_rate": 1.1143813282992964e-05,
+      "loss": 0.3866,
+      "step": 7325
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.84521484375,
+      "learning_rate": 1.114123905955037e-05,
+      "loss": 0.3983,
+      "step": 7326
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.11248779296875,
+      "learning_rate": 1.1138664836107774e-05,
+      "loss": 0.5833,
+      "step": 7327
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.666871070861816,
+      "learning_rate": 1.113609061266518e-05,
+      "loss": 0.5804,
+      "step": 7328
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.14678955078125,
+      "learning_rate": 1.1133516389222586e-05,
+      "loss": 0.3713,
+      "step": 7329
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.727951049804688,
+      "learning_rate": 1.1130942165779991e-05,
+      "loss": 0.4718,
+      "step": 7330
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.37954044342041,
+      "learning_rate": 1.1128367942337394e-05,
+      "loss": 0.4232,
+      "step": 7331
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.583354949951172,
+      "learning_rate": 1.11257937188948e-05,
+      "loss": 0.486,
+      "step": 7332
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.33794116973877,
+      "learning_rate": 1.1123219495452204e-05,
+      "loss": 0.3675,
+      "step": 7333
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.987041473388672,
+      "learning_rate": 1.1120645272009611e-05,
+      "loss": 0.5623,
+      "step": 7334
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.565496444702148,
+      "learning_rate": 1.1118071048567016e-05,
+      "loss": 0.6642,
+      "step": 7335
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.611883163452148,
+      "learning_rate": 1.1115496825124421e-05,
+      "loss": 0.4221,
+      "step": 7336
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.722855567932129,
+      "learning_rate": 1.1112922601681826e-05,
+      "loss": 0.5212,
+      "step": 7337
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.42210578918457,
+      "learning_rate": 1.1110348378239231e-05,
+      "loss": 0.5113,
+      "step": 7338
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.983506202697754,
+      "learning_rate": 1.1107774154796638e-05,
+      "loss": 0.3533,
+      "step": 7339
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.018218994140625,
+      "learning_rate": 1.1105199931354043e-05,
+      "loss": 0.5285,
+      "step": 7340
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.314803123474121,
+      "learning_rate": 1.1102625707911446e-05,
+      "loss": 0.382,
+      "step": 7341
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.689424514770508,
+      "learning_rate": 1.1100051484468851e-05,
+      "loss": 0.4992,
+      "step": 7342
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.650017738342285,
+      "learning_rate": 1.1097477261026258e-05,
+      "loss": 0.2609,
+      "step": 7343
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.291447639465332,
+      "learning_rate": 1.1094903037583663e-05,
+      "loss": 0.4485,
+      "step": 7344
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.257018089294434,
+      "learning_rate": 1.1092328814141068e-05,
+      "loss": 0.5109,
+      "step": 7345
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.063050270080566,
+      "learning_rate": 1.1089754590698473e-05,
+      "loss": 0.6356,
+      "step": 7346
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.568032264709473,
+      "learning_rate": 1.1087180367255878e-05,
+      "loss": 0.3379,
+      "step": 7347
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.283342361450195,
+      "learning_rate": 1.1084606143813284e-05,
+      "loss": 0.585,
+      "step": 7348
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.898557662963867,
+      "learning_rate": 1.108203192037069e-05,
+      "loss": 0.4101,
+      "step": 7349
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.641866207122803,
+      "learning_rate": 1.1079457696928094e-05,
+      "loss": 0.41,
+      "step": 7350
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.478236198425293,
+      "learning_rate": 1.10768834734855e-05,
+      "loss": 0.426,
+      "step": 7351
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.841486930847168,
+      "learning_rate": 1.1074309250042903e-05,
+      "loss": 0.4654,
+      "step": 7352
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.286558151245117,
+      "learning_rate": 1.107173502660031e-05,
+      "loss": 0.3845,
+      "step": 7353
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.059571266174316,
+      "learning_rate": 1.1069160803157714e-05,
+      "loss": 0.4252,
+      "step": 7354
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.960970878601074,
+      "learning_rate": 1.106658657971512e-05,
+      "loss": 0.5646,
+      "step": 7355
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.519693374633789,
+      "learning_rate": 1.1064012356272524e-05,
+      "loss": 0.5789,
+      "step": 7356
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.587873458862305,
+      "learning_rate": 1.106143813282993e-05,
+      "loss": 0.5044,
+      "step": 7357
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.576017379760742,
+      "learning_rate": 1.1058863909387336e-05,
+      "loss": 0.4697,
+      "step": 7358
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.494189262390137,
+      "learning_rate": 1.105628968594474e-05,
+      "loss": 0.3495,
+      "step": 7359
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.299254417419434,
+      "learning_rate": 1.1053715462502146e-05,
+      "loss": 0.5613,
+      "step": 7360
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.70302677154541,
+      "learning_rate": 1.105114123905955e-05,
+      "loss": 0.4824,
+      "step": 7361
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.210065841674805,
+      "learning_rate": 1.1048567015616956e-05,
+      "loss": 0.353,
+      "step": 7362
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.171737670898438,
+      "learning_rate": 1.104599279217436e-05,
+      "loss": 0.673,
+      "step": 7363
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.690330505371094,
+      "learning_rate": 1.1043418568731766e-05,
+      "loss": 0.5517,
+      "step": 7364
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.84161376953125,
+      "learning_rate": 1.104084434528917e-05,
+      "loss": 0.5083,
+      "step": 7365
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.02966594696045,
+      "learning_rate": 1.1038270121846576e-05,
+      "loss": 0.5422,
+      "step": 7366
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.594504356384277,
+      "learning_rate": 1.1035695898403982e-05,
+      "loss": 0.4964,
+      "step": 7367
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.446212768554688,
+      "learning_rate": 1.1033121674961387e-05,
+      "loss": 0.7646,
+      "step": 7368
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.387414932250977,
+      "learning_rate": 1.1030547451518792e-05,
+      "loss": 0.3946,
+      "step": 7369
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.785445213317871,
+      "learning_rate": 1.1027973228076197e-05,
+      "loss": 0.3826,
+      "step": 7370
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.900127410888672,
+      "learning_rate": 1.1025399004633602e-05,
+      "loss": 0.5517,
+      "step": 7371
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.443618774414062,
+      "learning_rate": 1.1022824781191009e-05,
+      "loss": 0.3948,
+      "step": 7372
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.678475379943848,
+      "learning_rate": 1.1020250557748412e-05,
+      "loss": 0.5166,
+      "step": 7373
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.408212661743164,
+      "learning_rate": 1.1017676334305817e-05,
+      "loss": 0.7953,
+      "step": 7374
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.494850158691406,
+      "learning_rate": 1.1015102110863222e-05,
+      "loss": 0.4099,
+      "step": 7375
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.82183837890625,
+      "learning_rate": 1.1012527887420629e-05,
+      "loss": 0.5599,
+      "step": 7376
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.489359855651855,
+      "learning_rate": 1.1009953663978034e-05,
+      "loss": 0.5565,
+      "step": 7377
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.595519065856934,
+      "learning_rate": 1.1007379440535439e-05,
+      "loss": 0.5726,
+      "step": 7378
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.113448143005371,
+      "learning_rate": 1.1004805217092844e-05,
+      "loss": 0.5839,
+      "step": 7379
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.382354736328125,
+      "learning_rate": 1.1002230993650249e-05,
+      "loss": 0.4741,
+      "step": 7380
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.775054931640625,
+      "learning_rate": 1.0999656770207656e-05,
+      "loss": 0.5934,
+      "step": 7381
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.917232513427734,
+      "learning_rate": 1.099708254676506e-05,
+      "loss": 0.4876,
+      "step": 7382
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.447864532470703,
+      "learning_rate": 1.0994508323322464e-05,
+      "loss": 0.4363,
+      "step": 7383
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.548230171203613,
+      "learning_rate": 1.0991934099879869e-05,
+      "loss": 0.5875,
+      "step": 7384
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.100251197814941,
+      "learning_rate": 1.0989359876437274e-05,
+      "loss": 0.4899,
+      "step": 7385
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.407061576843262,
+      "learning_rate": 1.098678565299468e-05,
+      "loss": 0.6766,
+      "step": 7386
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.539560317993164,
+      "learning_rate": 1.0984211429552086e-05,
+      "loss": 0.4533,
+      "step": 7387
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.022130966186523,
+      "learning_rate": 1.098163720610949e-05,
+      "loss": 0.6325,
+      "step": 7388
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.235760688781738,
+      "learning_rate": 1.0979062982666896e-05,
+      "loss": 0.4195,
+      "step": 7389
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.352241516113281,
+      "learning_rate": 1.09764887592243e-05,
+      "loss": 0.4234,
+      "step": 7390
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.590688705444336,
+      "learning_rate": 1.0973914535781707e-05,
+      "loss": 0.4648,
+      "step": 7391
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.004841804504395,
+      "learning_rate": 1.0971340312339112e-05,
+      "loss": 0.558,
+      "step": 7392
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.055572509765625,
+      "learning_rate": 1.0968766088896517e-05,
+      "loss": 0.5331,
+      "step": 7393
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.71291732788086,
+      "learning_rate": 1.096619186545392e-05,
+      "loss": 0.6749,
+      "step": 7394
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.808960914611816,
+      "learning_rate": 1.0963617642011327e-05,
+      "loss": 0.5916,
+      "step": 7395
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.335718154907227,
+      "learning_rate": 1.0961043418568732e-05,
+      "loss": 0.5091,
+      "step": 7396
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.797740936279297,
+      "learning_rate": 1.0958469195126137e-05,
+      "loss": 0.3448,
+      "step": 7397
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.89153003692627,
+      "learning_rate": 1.0955894971683542e-05,
+      "loss": 0.4089,
+      "step": 7398
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.435091018676758,
+      "learning_rate": 1.0953320748240947e-05,
+      "loss": 0.5835,
+      "step": 7399
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.908112525939941,
+      "learning_rate": 1.0950746524798354e-05,
+      "loss": 0.4312,
+      "step": 7400
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.195268630981445,
+      "learning_rate": 1.0948172301355759e-05,
+      "loss": 0.42,
+      "step": 7401
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.425474166870117,
+      "learning_rate": 1.0945598077913164e-05,
+      "loss": 0.3167,
+      "step": 7402
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.138252258300781,
+      "learning_rate": 1.0943023854470569e-05,
+      "loss": 0.4717,
+      "step": 7403
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.880084037780762,
+      "learning_rate": 1.0940449631027972e-05,
+      "loss": 0.4504,
+      "step": 7404
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.932443618774414,
+      "learning_rate": 1.0937875407585379e-05,
+      "loss": 0.4649,
+      "step": 7405
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.748109817504883,
+      "learning_rate": 1.0935301184142784e-05,
+      "loss": 0.5998,
+      "step": 7406
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.836468696594238,
+      "learning_rate": 1.0932726960700189e-05,
+      "loss": 0.4931,
+      "step": 7407
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.112396240234375,
+      "learning_rate": 1.0930152737257594e-05,
+      "loss": 0.4949,
+      "step": 7408
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.53786563873291,
+      "learning_rate": 1.0927578513814999e-05,
+      "loss": 0.4807,
+      "step": 7409
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.181134223937988,
+      "learning_rate": 1.0925004290372405e-05,
+      "loss": 0.539,
+      "step": 7410
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.775850296020508,
+      "learning_rate": 1.092243006692981e-05,
+      "loss": 0.4724,
+      "step": 7411
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.097833633422852,
+      "learning_rate": 1.0919855843487215e-05,
+      "loss": 0.8134,
+      "step": 7412
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.104572296142578,
+      "learning_rate": 1.091728162004462e-05,
+      "loss": 0.46,
+      "step": 7413
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.703893661499023,
+      "learning_rate": 1.0914707396602025e-05,
+      "loss": 0.842,
+      "step": 7414
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.625191688537598,
+      "learning_rate": 1.091213317315943e-05,
+      "loss": 0.4612,
+      "step": 7415
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.64117956161499,
+      "learning_rate": 1.0909558949716835e-05,
+      "loss": 0.2308,
+      "step": 7416
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.883071899414062,
+      "learning_rate": 1.090698472627424e-05,
+      "loss": 0.7752,
+      "step": 7417
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.58494758605957,
+      "learning_rate": 1.0904410502831645e-05,
+      "loss": 0.3848,
+      "step": 7418
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.34266471862793,
+      "learning_rate": 1.0901836279389052e-05,
+      "loss": 0.557,
+      "step": 7419
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.875296592712402,
+      "learning_rate": 1.0899262055946457e-05,
+      "loss": 0.6447,
+      "step": 7420
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.082805633544922,
+      "learning_rate": 1.0896687832503862e-05,
+      "loss": 0.3538,
+      "step": 7421
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.51256275177002,
+      "learning_rate": 1.0894113609061267e-05,
+      "loss": 0.6898,
+      "step": 7422
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.663081169128418,
+      "learning_rate": 1.0891539385618672e-05,
+      "loss": 0.5729,
+      "step": 7423
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.39887809753418,
+      "learning_rate": 1.0888965162176079e-05,
+      "loss": 0.4438,
+      "step": 7424
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.030675888061523,
+      "learning_rate": 1.0886390938733482e-05,
+      "loss": 0.3858,
+      "step": 7425
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.428672790527344,
+      "learning_rate": 1.0883816715290887e-05,
+      "loss": 0.7735,
+      "step": 7426
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.508746147155762,
+      "learning_rate": 1.0881242491848292e-05,
+      "loss": 0.5378,
+      "step": 7427
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.198284149169922,
+      "learning_rate": 1.0878668268405697e-05,
+      "loss": 0.3171,
+      "step": 7428
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.191280364990234,
+      "learning_rate": 1.0876094044963103e-05,
+      "loss": 0.3814,
+      "step": 7429
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.388025283813477,
+      "learning_rate": 1.0873519821520508e-05,
+      "loss": 0.6304,
+      "step": 7430
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.87787914276123,
+      "learning_rate": 1.0870945598077913e-05,
+      "loss": 0.5966,
+      "step": 7431
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.583935260772705,
+      "learning_rate": 1.0868371374635318e-05,
+      "loss": 0.3429,
+      "step": 7432
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.591317176818848,
+      "learning_rate": 1.0865797151192725e-05,
+      "loss": 0.3968,
+      "step": 7433
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.970666885375977,
+      "learning_rate": 1.086322292775013e-05,
+      "loss": 0.4822,
+      "step": 7434
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.406126976013184,
+      "learning_rate": 1.0860648704307533e-05,
+      "loss": 0.512,
+      "step": 7435
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.578034400939941,
+      "learning_rate": 1.0858074480864938e-05,
+      "loss": 0.411,
+      "step": 7436
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.091978073120117,
+      "learning_rate": 1.0855500257422343e-05,
+      "loss": 0.2314,
+      "step": 7437
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.776302337646484,
+      "learning_rate": 1.085292603397975e-05,
+      "loss": 0.3227,
+      "step": 7438
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.795964241027832,
+      "learning_rate": 1.0850351810537155e-05,
+      "loss": 0.5227,
+      "step": 7439
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.46214485168457,
+      "learning_rate": 1.084777758709456e-05,
+      "loss": 0.4356,
+      "step": 7440
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.215465545654297,
+      "learning_rate": 1.0845203363651965e-05,
+      "loss": 0.6005,
+      "step": 7441
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.202252388000488,
+      "learning_rate": 1.084262914020937e-05,
+      "loss": 0.4234,
+      "step": 7442
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.662891387939453,
+      "learning_rate": 1.0840054916766777e-05,
+      "loss": 0.4677,
+      "step": 7443
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.376526832580566,
+      "learning_rate": 1.0837480693324182e-05,
+      "loss": 0.5627,
+      "step": 7444
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.635258674621582,
+      "learning_rate": 1.0834906469881587e-05,
+      "loss": 0.4817,
+      "step": 7445
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.524569511413574,
+      "learning_rate": 1.083233224643899e-05,
+      "loss": 0.8378,
+      "step": 7446
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.190790176391602,
+      "learning_rate": 1.0829758022996397e-05,
+      "loss": 0.4123,
+      "step": 7447
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.145949363708496,
+      "learning_rate": 1.0827183799553802e-05,
+      "loss": 0.4973,
+      "step": 7448
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.877215385437012,
+      "learning_rate": 1.0824609576111207e-05,
+      "loss": 0.4719,
+      "step": 7449
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.523372650146484,
+      "learning_rate": 1.0822035352668612e-05,
+      "loss": 0.6474,
+      "step": 7450
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.625123023986816,
+      "learning_rate": 1.0819461129226017e-05,
+      "loss": 0.5119,
+      "step": 7451
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.660490989685059,
+      "learning_rate": 1.0816886905783423e-05,
+      "loss": 0.6141,
+      "step": 7452
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.527603149414062,
+      "learning_rate": 1.0814312682340828e-05,
+      "loss": 0.3769,
+      "step": 7453
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.848456382751465,
+      "learning_rate": 1.0811738458898233e-05,
+      "loss": 0.4286,
+      "step": 7454
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.680733680725098,
+      "learning_rate": 1.0809164235455638e-05,
+      "loss": 0.5844,
+      "step": 7455
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.627582550048828,
+      "learning_rate": 1.0806590012013042e-05,
+      "loss": 0.4866,
+      "step": 7456
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.343541145324707,
+      "learning_rate": 1.0804015788570448e-05,
+      "loss": 0.5801,
+      "step": 7457
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.711421012878418,
+      "learning_rate": 1.0801441565127853e-05,
+      "loss": 0.6033,
+      "step": 7458
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.719820976257324,
+      "learning_rate": 1.0798867341685258e-05,
+      "loss": 0.4498,
+      "step": 7459
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.25599193572998,
+      "learning_rate": 1.0796293118242663e-05,
+      "loss": 0.5434,
+      "step": 7460
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.425467491149902,
+      "learning_rate": 1.0793718894800068e-05,
+      "loss": 0.4365,
+      "step": 7461
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.971858024597168,
+      "learning_rate": 1.0791144671357475e-05,
+      "loss": 0.4072,
+      "step": 7462
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.246363639831543,
+      "learning_rate": 1.078857044791488e-05,
+      "loss": 0.5188,
+      "step": 7463
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.017495155334473,
+      "learning_rate": 1.0785996224472285e-05,
+      "loss": 0.3781,
+      "step": 7464
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.844362258911133,
+      "learning_rate": 1.078342200102969e-05,
+      "loss": 0.5011,
+      "step": 7465
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.767553329467773,
+      "learning_rate": 1.0780847777587095e-05,
+      "loss": 0.5336,
+      "step": 7466
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.728456497192383,
+      "learning_rate": 1.07782735541445e-05,
+      "loss": 0.5207,
+      "step": 7467
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.234371185302734,
+      "learning_rate": 1.0775699330701905e-05,
+      "loss": 0.6167,
+      "step": 7468
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.180463790893555,
+      "learning_rate": 1.077312510725931e-05,
+      "loss": 0.5041,
+      "step": 7469
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.594744682312012,
+      "learning_rate": 1.0770550883816715e-05,
+      "loss": 0.3767,
+      "step": 7470
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.505976676940918,
+      "learning_rate": 1.0767976660374121e-05,
+      "loss": 0.5005,
+      "step": 7471
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.433834075927734,
+      "learning_rate": 1.0765402436931526e-05,
+      "loss": 0.3806,
+      "step": 7472
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.749080657958984,
+      "learning_rate": 1.0762828213488931e-05,
+      "loss": 0.5238,
+      "step": 7473
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.561195373535156,
+      "learning_rate": 1.0760253990046336e-05,
+      "loss": 0.5212,
+      "step": 7474
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.059864044189453,
+      "learning_rate": 1.0757679766603741e-05,
+      "loss": 0.5987,
+      "step": 7475
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.482023239135742,
+      "learning_rate": 1.0755105543161148e-05,
+      "loss": 0.5982,
+      "step": 7476
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.836862564086914,
+      "learning_rate": 1.0752531319718551e-05,
+      "loss": 0.3857,
+      "step": 7477
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.901195526123047,
+      "learning_rate": 1.0749957096275956e-05,
+      "loss": 0.6019,
+      "step": 7478
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.829532623291016,
+      "learning_rate": 1.0747382872833361e-05,
+      "loss": 0.5685,
+      "step": 7479
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.477641105651855,
+      "learning_rate": 1.0744808649390766e-05,
+      "loss": 0.6177,
+      "step": 7480
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.40280532836914,
+      "learning_rate": 1.0742234425948173e-05,
+      "loss": 0.5157,
+      "step": 7481
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.046311378479004,
+      "learning_rate": 1.0739660202505578e-05,
+      "loss": 0.6029,
+      "step": 7482
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.58298110961914,
+      "learning_rate": 1.0737085979062983e-05,
+      "loss": 0.421,
+      "step": 7483
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.21367359161377,
+      "learning_rate": 1.0734511755620388e-05,
+      "loss": 0.4925,
+      "step": 7484
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.266897201538086,
+      "learning_rate": 1.0731937532177795e-05,
+      "loss": 0.3255,
+      "step": 7485
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.406457901000977,
+      "learning_rate": 1.07293633087352e-05,
+      "loss": 0.3816,
+      "step": 7486
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.383731842041016,
+      "learning_rate": 1.0726789085292603e-05,
+      "loss": 0.5075,
+      "step": 7487
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.13294792175293,
+      "learning_rate": 1.0724214861850008e-05,
+      "loss": 0.5304,
+      "step": 7488
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.950424194335938,
+      "learning_rate": 1.0721640638407413e-05,
+      "loss": 0.4476,
+      "step": 7489
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.416092872619629,
+      "learning_rate": 1.071906641496482e-05,
+      "loss": 0.4161,
+      "step": 7490
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.153877258300781,
+      "learning_rate": 1.0716492191522225e-05,
+      "loss": 0.5224,
+      "step": 7491
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.353551864624023,
+      "learning_rate": 1.071391796807963e-05,
+      "loss": 0.4199,
+      "step": 7492
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.87924861907959,
+      "learning_rate": 1.0711343744637035e-05,
+      "loss": 0.2971,
+      "step": 7493
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.318202018737793,
+      "learning_rate": 1.070876952119444e-05,
+      "loss": 0.3374,
+      "step": 7494
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.892108917236328,
+      "learning_rate": 1.0706195297751846e-05,
+      "loss": 0.6273,
+      "step": 7495
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.575667381286621,
+      "learning_rate": 1.0703621074309251e-05,
+      "loss": 0.462,
+      "step": 7496
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.327213287353516,
+      "learning_rate": 1.0701046850866656e-05,
+      "loss": 0.6198,
+      "step": 7497
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.397997856140137,
+      "learning_rate": 1.069847262742406e-05,
+      "loss": 0.3286,
+      "step": 7498
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.168190956115723,
+      "learning_rate": 1.0695898403981466e-05,
+      "loss": 0.5473,
+      "step": 7499
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.656102180480957,
+      "learning_rate": 1.0693324180538871e-05,
+      "loss": 0.6185,
+      "step": 7500
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.660324096679688,
+      "learning_rate": 1.0690749957096276e-05,
+      "loss": 0.4035,
+      "step": 7501
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.037286758422852,
+      "learning_rate": 1.0688175733653681e-05,
+      "loss": 0.4422,
+      "step": 7502
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.997214317321777,
+      "learning_rate": 1.0685601510211086e-05,
+      "loss": 0.4572,
+      "step": 7503
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.286606788635254,
+      "learning_rate": 1.0683027286768493e-05,
+      "loss": 0.4222,
+      "step": 7504
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.196907043457031,
+      "learning_rate": 1.0680453063325898e-05,
+      "loss": 0.5055,
+      "step": 7505
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.247546195983887,
+      "learning_rate": 1.0677878839883303e-05,
+      "loss": 0.5824,
+      "step": 7506
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.00897979736328,
+      "learning_rate": 1.0675304616440708e-05,
+      "loss": 0.5158,
+      "step": 7507
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.000876426696777,
+      "learning_rate": 1.0672730392998111e-05,
+      "loss": 0.902,
+      "step": 7508
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.657307624816895,
+      "learning_rate": 1.0670156169555518e-05,
+      "loss": 0.2476,
+      "step": 7509
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.91662883758545,
+      "learning_rate": 1.0667581946112923e-05,
+      "loss": 0.5422,
+      "step": 7510
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.896361351013184,
+      "learning_rate": 1.0665007722670328e-05,
+      "loss": 0.6262,
+      "step": 7511
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.007389068603516,
+      "learning_rate": 1.0662433499227733e-05,
+      "loss": 0.5553,
+      "step": 7512
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.739877700805664,
+      "learning_rate": 1.0659859275785138e-05,
+      "loss": 0.3934,
+      "step": 7513
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.095998764038086,
+      "learning_rate": 1.0657285052342544e-05,
+      "loss": 0.3749,
+      "step": 7514
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.805936813354492,
+      "learning_rate": 1.065471082889995e-05,
+      "loss": 0.4699,
+      "step": 7515
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.338393211364746,
+      "learning_rate": 1.0652136605457354e-05,
+      "loss": 0.5155,
+      "step": 7516
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.030548572540283,
+      "learning_rate": 1.064956238201476e-05,
+      "loss": 0.2224,
+      "step": 7517
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.219165802001953,
+      "learning_rate": 1.0646988158572164e-05,
+      "loss": 0.447,
+      "step": 7518
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.641654014587402,
+      "learning_rate": 1.064441393512957e-05,
+      "loss": 0.4007,
+      "step": 7519
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.736027717590332,
+      "learning_rate": 1.0641839711686974e-05,
+      "loss": 0.4508,
+      "step": 7520
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.524602890014648,
+      "learning_rate": 1.063926548824438e-05,
+      "loss": 0.3543,
+      "step": 7521
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.188543319702148,
+      "learning_rate": 1.0636691264801784e-05,
+      "loss": 0.4827,
+      "step": 7522
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.921015739440918,
+      "learning_rate": 1.0634117041359191e-05,
+      "loss": 0.4687,
+      "step": 7523
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.920520782470703,
+      "learning_rate": 1.0631542817916596e-05,
+      "loss": 0.496,
+      "step": 7524
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.996758460998535,
+      "learning_rate": 1.0628968594474001e-05,
+      "loss": 0.5602,
+      "step": 7525
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.540448188781738,
+      "learning_rate": 1.0626394371031406e-05,
+      "loss": 0.5454,
+      "step": 7526
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.803266525268555,
+      "learning_rate": 1.0623820147588811e-05,
+      "loss": 0.4147,
+      "step": 7527
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.970677375793457,
+      "learning_rate": 1.0621245924146218e-05,
+      "loss": 0.3776,
+      "step": 7528
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.074735641479492,
+      "learning_rate": 1.0618671700703621e-05,
+      "loss": 0.4929,
+      "step": 7529
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.672462463378906,
+      "learning_rate": 1.0616097477261026e-05,
+      "loss": 0.4758,
+      "step": 7530
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.037508964538574,
+      "learning_rate": 1.061352325381843e-05,
+      "loss": 0.6194,
+      "step": 7531
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.949732780456543,
+      "learning_rate": 1.0610949030375836e-05,
+      "loss": 0.8195,
+      "step": 7532
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.576162338256836,
+      "learning_rate": 1.0608374806933243e-05,
+      "loss": 0.6324,
+      "step": 7533
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.948009967803955,
+      "learning_rate": 1.0605800583490647e-05,
+      "loss": 0.3754,
+      "step": 7534
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.722002029418945,
+      "learning_rate": 1.0603226360048052e-05,
+      "loss": 0.4788,
+      "step": 7535
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.692331314086914,
+      "learning_rate": 1.0600652136605457e-05,
+      "loss": 0.4046,
+      "step": 7536
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.092635154724121,
+      "learning_rate": 1.0598077913162864e-05,
+      "loss": 0.3155,
+      "step": 7537
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.974721908569336,
+      "learning_rate": 1.0595503689720269e-05,
+      "loss": 0.4394,
+      "step": 7538
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.534664154052734,
+      "learning_rate": 1.0592929466277672e-05,
+      "loss": 0.3509,
+      "step": 7539
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.622676849365234,
+      "learning_rate": 1.0590355242835077e-05,
+      "loss": 0.4098,
+      "step": 7540
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.259799003601074,
+      "learning_rate": 1.0587781019392482e-05,
+      "loss": 0.3832,
+      "step": 7541
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.262960433959961,
+      "learning_rate": 1.0585206795949889e-05,
+      "loss": 0.2735,
+      "step": 7542
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.705814361572266,
+      "learning_rate": 1.0582632572507294e-05,
+      "loss": 0.4398,
+      "step": 7543
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.44537353515625,
+      "learning_rate": 1.0580058349064699e-05,
+      "loss": 0.4076,
+      "step": 7544
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.455634117126465,
+      "learning_rate": 1.0577484125622104e-05,
+      "loss": 0.5374,
+      "step": 7545
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.263714790344238,
+      "learning_rate": 1.0574909902179509e-05,
+      "loss": 0.5075,
+      "step": 7546
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.596541404724121,
+      "learning_rate": 1.0572335678736916e-05,
+      "loss": 0.5074,
+      "step": 7547
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.088547706604004,
+      "learning_rate": 1.056976145529432e-05,
+      "loss": 0.6784,
+      "step": 7548
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.327363967895508,
+      "learning_rate": 1.0567187231851726e-05,
+      "loss": 0.5328,
+      "step": 7549
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.240397453308105,
+      "learning_rate": 1.0564613008409129e-05,
+      "loss": 0.493,
+      "step": 7550
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.778910636901855,
+      "learning_rate": 1.0562038784966536e-05,
+      "loss": 0.5793,
+      "step": 7551
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.707723617553711,
+      "learning_rate": 1.055946456152394e-05,
+      "loss": 0.4145,
+      "step": 7552
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.224451541900635,
+      "learning_rate": 1.0556890338081346e-05,
+      "loss": 0.3317,
+      "step": 7553
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.40455150604248,
+      "learning_rate": 1.055431611463875e-05,
+      "loss": 0.2934,
+      "step": 7554
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.57699203491211,
+      "learning_rate": 1.0551741891196156e-05,
+      "loss": 0.4265,
+      "step": 7555
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.641803741455078,
+      "learning_rate": 1.0549167667753562e-05,
+      "loss": 0.7676,
+      "step": 7556
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.282593727111816,
+      "learning_rate": 1.0546593444310967e-05,
+      "loss": 0.4041,
+      "step": 7557
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.14730453491211,
+      "learning_rate": 1.0544019220868372e-05,
+      "loss": 0.6872,
+      "step": 7558
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.139691352844238,
+      "learning_rate": 1.0541444997425777e-05,
+      "loss": 0.7045,
+      "step": 7559
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.67737102508545,
+      "learning_rate": 1.053887077398318e-05,
+      "loss": 0.5271,
+      "step": 7560
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.156164169311523,
+      "learning_rate": 1.0536296550540587e-05,
+      "loss": 0.4915,
+      "step": 7561
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.528146743774414,
+      "learning_rate": 1.0533722327097992e-05,
+      "loss": 0.5988,
+      "step": 7562
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.946915626525879,
+      "learning_rate": 1.0531148103655397e-05,
+      "loss": 0.4204,
+      "step": 7563
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.51050090789795,
+      "learning_rate": 1.0528573880212802e-05,
+      "loss": 0.5621,
+      "step": 7564
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.574899673461914,
+      "learning_rate": 1.0525999656770207e-05,
+      "loss": 0.4502,
+      "step": 7565
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.043357849121094,
+      "learning_rate": 1.0523425433327614e-05,
+      "loss": 0.4212,
+      "step": 7566
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.919164657592773,
+      "learning_rate": 1.0520851209885019e-05,
+      "loss": 0.3513,
+      "step": 7567
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.325124740600586,
+      "learning_rate": 1.0518276986442424e-05,
+      "loss": 0.4333,
+      "step": 7568
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.761919021606445,
+      "learning_rate": 1.0515702762999829e-05,
+      "loss": 0.4765,
+      "step": 7569
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.399340629577637,
+      "learning_rate": 1.0513128539557234e-05,
+      "loss": 0.4614,
+      "step": 7570
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.49824047088623,
+      "learning_rate": 1.0510554316114639e-05,
+      "loss": 0.7209,
+      "step": 7571
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.914290428161621,
+      "learning_rate": 1.0507980092672044e-05,
+      "loss": 0.4861,
+      "step": 7572
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.58306884765625,
+      "learning_rate": 1.0505405869229449e-05,
+      "loss": 0.3629,
+      "step": 7573
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.455018997192383,
+      "learning_rate": 1.0502831645786854e-05,
+      "loss": 0.3693,
+      "step": 7574
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.3626508712768555,
+      "learning_rate": 1.050025742234426e-05,
+      "loss": 0.3849,
+      "step": 7575
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.481241226196289,
+      "learning_rate": 1.0497683198901665e-05,
+      "loss": 0.6926,
+      "step": 7576
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.063183784484863,
+      "learning_rate": 1.049510897545907e-05,
+      "loss": 0.4847,
+      "step": 7577
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.723499298095703,
+      "learning_rate": 1.0492534752016475e-05,
+      "loss": 0.5744,
+      "step": 7578
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.72335147857666,
+      "learning_rate": 1.048996052857388e-05,
+      "loss": 0.368,
+      "step": 7579
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.503802299499512,
+      "learning_rate": 1.0487386305131287e-05,
+      "loss": 0.6391,
+      "step": 7580
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.515002250671387,
+      "learning_rate": 1.048481208168869e-05,
+      "loss": 0.5714,
+      "step": 7581
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.594352722167969,
+      "learning_rate": 1.0482237858246095e-05,
+      "loss": 0.5666,
+      "step": 7582
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.503571510314941,
+      "learning_rate": 1.04796636348035e-05,
+      "loss": 0.3397,
+      "step": 7583
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.70746898651123,
+      "learning_rate": 1.0477089411360905e-05,
+      "loss": 0.3438,
+      "step": 7584
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.973519325256348,
+      "learning_rate": 1.0474515187918312e-05,
+      "loss": 0.5933,
+      "step": 7585
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.70559310913086,
+      "learning_rate": 1.0471940964475717e-05,
+      "loss": 0.5616,
+      "step": 7586
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.801112174987793,
+      "learning_rate": 1.0469366741033122e-05,
+      "loss": 0.6084,
+      "step": 7587
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.673832893371582,
+      "learning_rate": 1.0466792517590527e-05,
+      "loss": 0.5221,
+      "step": 7588
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.244690895080566,
+      "learning_rate": 1.0464218294147934e-05,
+      "loss": 0.5931,
+      "step": 7589
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.470046043395996,
+      "learning_rate": 1.0461644070705339e-05,
+      "loss": 0.3065,
+      "step": 7590
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.129098892211914,
+      "learning_rate": 1.0459069847262742e-05,
+      "loss": 0.3822,
+      "step": 7591
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.066459655761719,
+      "learning_rate": 1.0456495623820147e-05,
+      "loss": 0.5738,
+      "step": 7592
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.978337287902832,
+      "learning_rate": 1.0453921400377552e-05,
+      "loss": 0.5539,
+      "step": 7593
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.200026512145996,
+      "learning_rate": 1.0451347176934959e-05,
+      "loss": 0.509,
+      "step": 7594
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.632519721984863,
+      "learning_rate": 1.0448772953492364e-05,
+      "loss": 0.5611,
+      "step": 7595
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.198501110076904,
+      "learning_rate": 1.0446198730049769e-05,
+      "loss": 0.338,
+      "step": 7596
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.813353538513184,
+      "learning_rate": 1.0443624506607174e-05,
+      "loss": 0.812,
+      "step": 7597
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.40137004852295,
+      "learning_rate": 1.0441050283164579e-05,
+      "loss": 0.6986,
+      "step": 7598
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.052644729614258,
+      "learning_rate": 1.0438476059721985e-05,
+      "loss": 0.402,
+      "step": 7599
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.621946334838867,
+      "learning_rate": 1.043590183627939e-05,
+      "loss": 0.5718,
+      "step": 7600
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.031299591064453,
+      "learning_rate": 1.0433327612836795e-05,
+      "loss": 0.2937,
+      "step": 7601
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.721993446350098,
+      "learning_rate": 1.0430753389394198e-05,
+      "loss": 0.3963,
+      "step": 7602
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.292656898498535,
+      "learning_rate": 1.0428179165951603e-05,
+      "loss": 0.5122,
+      "step": 7603
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.778241157531738,
+      "learning_rate": 1.042560494250901e-05,
+      "loss": 0.5224,
+      "step": 7604
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.720760345458984,
+      "learning_rate": 1.0423030719066415e-05,
+      "loss": 0.5592,
+      "step": 7605
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.58338737487793,
+      "learning_rate": 1.042045649562382e-05,
+      "loss": 0.5383,
+      "step": 7606
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.038021087646484,
+      "learning_rate": 1.0417882272181225e-05,
+      "loss": 0.6423,
+      "step": 7607
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.134819984436035,
+      "learning_rate": 1.0415308048738632e-05,
+      "loss": 0.6105,
+      "step": 7608
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.333404541015625,
+      "learning_rate": 1.0412733825296037e-05,
+      "loss": 0.6074,
+      "step": 7609
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.601170539855957,
+      "learning_rate": 1.0410159601853442e-05,
+      "loss": 0.4002,
+      "step": 7610
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.287076950073242,
+      "learning_rate": 1.0407585378410847e-05,
+      "loss": 0.5329,
+      "step": 7611
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.492240905761719,
+      "learning_rate": 1.040501115496825e-05,
+      "loss": 0.4515,
+      "step": 7612
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.017926216125488,
+      "learning_rate": 1.0402436931525657e-05,
+      "loss": 0.4445,
+      "step": 7613
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.896016120910645,
+      "learning_rate": 1.0399862708083062e-05,
+      "loss": 0.5659,
+      "step": 7614
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.269497871398926,
+      "learning_rate": 1.0397288484640467e-05,
+      "loss": 0.5041,
+      "step": 7615
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.545845985412598,
+      "learning_rate": 1.0394714261197872e-05,
+      "loss": 0.6307,
+      "step": 7616
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.45733642578125,
+      "learning_rate": 1.0392140037755277e-05,
+      "loss": 0.5388,
+      "step": 7617
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.823169708251953,
+      "learning_rate": 1.0389565814312683e-05,
+      "loss": 0.4665,
+      "step": 7618
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.971879959106445,
+      "learning_rate": 1.0386991590870088e-05,
+      "loss": 0.461,
+      "step": 7619
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.590045928955078,
+      "learning_rate": 1.0384417367427493e-05,
+      "loss": 0.4452,
+      "step": 7620
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.817612648010254,
+      "learning_rate": 1.0381843143984898e-05,
+      "loss": 0.6212,
+      "step": 7621
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.876532554626465,
+      "learning_rate": 1.0379268920542303e-05,
+      "loss": 0.5217,
+      "step": 7622
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.179770469665527,
+      "learning_rate": 1.0376694697099708e-05,
+      "loss": 0.6029,
+      "step": 7623
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.677385330200195,
+      "learning_rate": 1.0374120473657113e-05,
+      "loss": 0.6449,
+      "step": 7624
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.10698413848877,
+      "learning_rate": 1.0371546250214518e-05,
+      "loss": 0.494,
+      "step": 7625
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.941584587097168,
+      "learning_rate": 1.0368972026771923e-05,
+      "loss": 0.5538,
+      "step": 7626
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.517629623413086,
+      "learning_rate": 1.036639780332933e-05,
+      "loss": 0.3846,
+      "step": 7627
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.007301330566406,
+      "learning_rate": 1.0363823579886735e-05,
+      "loss": 0.5212,
+      "step": 7628
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.88581657409668,
+      "learning_rate": 1.036124935644414e-05,
+      "loss": 0.6526,
+      "step": 7629
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.389724731445312,
+      "learning_rate": 1.0358675133001545e-05,
+      "loss": 0.745,
+      "step": 7630
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.649310111999512,
+      "learning_rate": 1.035610090955895e-05,
+      "loss": 0.5522,
+      "step": 7631
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.6701340675354,
+      "learning_rate": 1.0353526686116357e-05,
+      "loss": 0.287,
+      "step": 7632
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.678239822387695,
+      "learning_rate": 1.035095246267376e-05,
+      "loss": 0.5347,
+      "step": 7633
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.223403930664062,
+      "learning_rate": 1.0348378239231165e-05,
+      "loss": 0.4687,
+      "step": 7634
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.723023414611816,
+      "learning_rate": 1.034580401578857e-05,
+      "loss": 0.4222,
+      "step": 7635
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.584243774414062,
+      "learning_rate": 1.0343229792345975e-05,
+      "loss": 0.6159,
+      "step": 7636
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.441088676452637,
+      "learning_rate": 1.0340655568903382e-05,
+      "loss": 0.5008,
+      "step": 7637
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.49659538269043,
+      "learning_rate": 1.0338081345460787e-05,
+      "loss": 0.668,
+      "step": 7638
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.211221694946289,
+      "learning_rate": 1.0335507122018191e-05,
+      "loss": 0.3276,
+      "step": 7639
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.020726203918457,
+      "learning_rate": 1.0332932898575596e-05,
+      "loss": 0.4316,
+      "step": 7640
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.816468238830566,
+      "learning_rate": 1.0330358675133003e-05,
+      "loss": 0.4717,
+      "step": 7641
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.137051582336426,
+      "learning_rate": 1.0327784451690408e-05,
+      "loss": 0.5587,
+      "step": 7642
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.667003631591797,
+      "learning_rate": 1.0325210228247811e-05,
+      "loss": 0.6343,
+      "step": 7643
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.340344429016113,
+      "learning_rate": 1.0322636004805216e-05,
+      "loss": 0.3575,
+      "step": 7644
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.79090404510498,
+      "learning_rate": 1.0320061781362621e-05,
+      "loss": 0.4042,
+      "step": 7645
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.056097984313965,
+      "learning_rate": 1.0317487557920028e-05,
+      "loss": 0.3421,
+      "step": 7646
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.9146146774292,
+      "learning_rate": 1.0314913334477433e-05,
+      "loss": 0.5977,
+      "step": 7647
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.455370903015137,
+      "learning_rate": 1.0312339111034838e-05,
+      "loss": 0.4442,
+      "step": 7648
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.547144889831543,
+      "learning_rate": 1.0309764887592243e-05,
+      "loss": 0.4097,
+      "step": 7649
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.367593765258789,
+      "learning_rate": 1.0307190664149648e-05,
+      "loss": 0.4782,
+      "step": 7650
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.46049976348877,
+      "learning_rate": 1.0304616440707055e-05,
+      "loss": 0.6173,
+      "step": 7651
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.880172729492188,
+      "learning_rate": 1.030204221726446e-05,
+      "loss": 0.4819,
+      "step": 7652
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.555892944335938,
+      "learning_rate": 1.0299467993821865e-05,
+      "loss": 0.5615,
+      "step": 7653
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.010433197021484,
+      "learning_rate": 1.0296893770379268e-05,
+      "loss": 0.3736,
+      "step": 7654
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.676969528198242,
+      "learning_rate": 1.0294319546936673e-05,
+      "loss": 0.4007,
+      "step": 7655
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.463154792785645,
+      "learning_rate": 1.029174532349408e-05,
+      "loss": 0.6578,
+      "step": 7656
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.027267456054688,
+      "learning_rate": 1.0289171100051485e-05,
+      "loss": 0.31,
+      "step": 7657
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.18458080291748,
+      "learning_rate": 1.028659687660889e-05,
+      "loss": 0.3947,
+      "step": 7658
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.778727531433105,
+      "learning_rate": 1.0284022653166295e-05,
+      "loss": 0.4626,
+      "step": 7659
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.715446472167969,
+      "learning_rate": 1.0281448429723701e-05,
+      "loss": 0.5173,
+      "step": 7660
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.753809928894043,
+      "learning_rate": 1.0278874206281106e-05,
+      "loss": 0.3613,
+      "step": 7661
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.290806770324707,
+      "learning_rate": 1.0276299982838511e-05,
+      "loss": 0.4105,
+      "step": 7662
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.830404281616211,
+      "learning_rate": 1.0273725759395916e-05,
+      "loss": 0.4879,
+      "step": 7663
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.855487823486328,
+      "learning_rate": 1.027115153595332e-05,
+      "loss": 0.4888,
+      "step": 7664
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.879138946533203,
+      "learning_rate": 1.0268577312510726e-05,
+      "loss": 0.4953,
+      "step": 7665
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.021116256713867,
+      "learning_rate": 1.0266003089068131e-05,
+      "loss": 0.5683,
+      "step": 7666
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.994811058044434,
+      "learning_rate": 1.0263428865625536e-05,
+      "loss": 0.5198,
+      "step": 7667
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.031010627746582,
+      "learning_rate": 1.0260854642182941e-05,
+      "loss": 0.6681,
+      "step": 7668
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.21917724609375,
+      "learning_rate": 1.0258280418740346e-05,
+      "loss": 0.4113,
+      "step": 7669
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.923477172851562,
+      "learning_rate": 1.0255706195297753e-05,
+      "loss": 0.5717,
+      "step": 7670
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.71715545654297,
+      "learning_rate": 1.0253131971855158e-05,
+      "loss": 0.6811,
+      "step": 7671
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.599434852600098,
+      "learning_rate": 1.0250557748412563e-05,
+      "loss": 0.4114,
+      "step": 7672
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.938738822937012,
+      "learning_rate": 1.0247983524969968e-05,
+      "loss": 0.4104,
+      "step": 7673
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.35966968536377,
+      "learning_rate": 1.0245409301527373e-05,
+      "loss": 0.5351,
+      "step": 7674
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.487553596496582,
+      "learning_rate": 1.0242835078084778e-05,
+      "loss": 0.3632,
+      "step": 7675
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.132365226745605,
+      "learning_rate": 1.0240260854642183e-05,
+      "loss": 0.5049,
+      "step": 7676
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.248384475708008,
+      "learning_rate": 1.0237686631199588e-05,
+      "loss": 0.7382,
+      "step": 7677
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.453958511352539,
+      "learning_rate": 1.0235112407756993e-05,
+      "loss": 0.6043,
+      "step": 7678
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.014633178710938,
+      "learning_rate": 1.02325381843144e-05,
+      "loss": 0.5675,
+      "step": 7679
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.388157844543457,
+      "learning_rate": 1.0229963960871804e-05,
+      "loss": 0.7631,
+      "step": 7680
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.168139457702637,
+      "learning_rate": 1.022738973742921e-05,
+      "loss": 0.3885,
+      "step": 7681
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.904991149902344,
+      "learning_rate": 1.0224815513986614e-05,
+      "loss": 0.6153,
+      "step": 7682
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.918192863464355,
+      "learning_rate": 1.022224129054402e-05,
+      "loss": 0.4525,
+      "step": 7683
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.1475191116333,
+      "learning_rate": 1.0219667067101426e-05,
+      "loss": 0.5083,
+      "step": 7684
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.759369850158691,
+      "learning_rate": 1.021709284365883e-05,
+      "loss": 0.4407,
+      "step": 7685
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.438612937927246,
+      "learning_rate": 1.0214518620216234e-05,
+      "loss": 0.3931,
+      "step": 7686
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.576518058776855,
+      "learning_rate": 1.021194439677364e-05,
+      "loss": 0.5469,
+      "step": 7687
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.180768966674805,
+      "learning_rate": 1.0209370173331044e-05,
+      "loss": 0.5146,
+      "step": 7688
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.942716598510742,
+      "learning_rate": 1.0206795949888451e-05,
+      "loss": 0.5428,
+      "step": 7689
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.733922004699707,
+      "learning_rate": 1.0204221726445856e-05,
+      "loss": 0.6077,
+      "step": 7690
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.93974781036377,
+      "learning_rate": 1.0201647503003261e-05,
+      "loss": 0.4856,
+      "step": 7691
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.876021385192871,
+      "learning_rate": 1.0199073279560666e-05,
+      "loss": 0.5097,
+      "step": 7692
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.196867942810059,
+      "learning_rate": 1.0196499056118073e-05,
+      "loss": 0.5664,
+      "step": 7693
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.308637619018555,
+      "learning_rate": 1.0193924832675478e-05,
+      "loss": 0.6109,
+      "step": 7694
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.110604286193848,
+      "learning_rate": 1.0191350609232881e-05,
+      "loss": 0.478,
+      "step": 7695
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.802572250366211,
+      "learning_rate": 1.0188776385790286e-05,
+      "loss": 0.4881,
+      "step": 7696
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.356598854064941,
+      "learning_rate": 1.0186202162347691e-05,
+      "loss": 0.5587,
+      "step": 7697
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.789701461791992,
+      "learning_rate": 1.0183627938905098e-05,
+      "loss": 0.4952,
+      "step": 7698
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.346013069152832,
+      "learning_rate": 1.0181053715462503e-05,
+      "loss": 0.6237,
+      "step": 7699
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.338967323303223,
+      "learning_rate": 1.0178479492019908e-05,
+      "loss": 0.5249,
+      "step": 7700
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.160362243652344,
+      "learning_rate": 1.0175905268577313e-05,
+      "loss": 0.4278,
+      "step": 7701
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.647008419036865,
+      "learning_rate": 1.0173331045134718e-05,
+      "loss": 0.3249,
+      "step": 7702
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.2664213180542,
+      "learning_rate": 1.0170756821692124e-05,
+      "loss": 0.5051,
+      "step": 7703
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.073074340820312,
+      "learning_rate": 1.016818259824953e-05,
+      "loss": 0.8712,
+      "step": 7704
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.206207275390625,
+      "learning_rate": 1.0165608374806934e-05,
+      "loss": 0.4776,
+      "step": 7705
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.693732261657715,
+      "learning_rate": 1.0163034151364338e-05,
+      "loss": 0.5567,
+      "step": 7706
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.299224853515625,
+      "learning_rate": 1.0160459927921742e-05,
+      "loss": 0.3744,
+      "step": 7707
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.106921195983887,
+      "learning_rate": 1.015788570447915e-05,
+      "loss": 0.6698,
+      "step": 7708
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.241186141967773,
+      "learning_rate": 1.0155311481036554e-05,
+      "loss": 0.4848,
+      "step": 7709
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.151101112365723,
+      "learning_rate": 1.0152737257593959e-05,
+      "loss": 0.4434,
+      "step": 7710
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.867897033691406,
+      "learning_rate": 1.0150163034151364e-05,
+      "loss": 0.663,
+      "step": 7711
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.96371841430664,
+      "learning_rate": 1.014758881070877e-05,
+      "loss": 0.5116,
+      "step": 7712
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.17064094543457,
+      "learning_rate": 1.0145014587266176e-05,
+      "loss": 0.3887,
+      "step": 7713
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.953055381774902,
+      "learning_rate": 1.014244036382358e-05,
+      "loss": 0.6361,
+      "step": 7714
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.440086364746094,
+      "learning_rate": 1.0139866140380986e-05,
+      "loss": 0.4921,
+      "step": 7715
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.911348342895508,
+      "learning_rate": 1.0137291916938389e-05,
+      "loss": 0.416,
+      "step": 7716
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.459033012390137,
+      "learning_rate": 1.0134717693495796e-05,
+      "loss": 0.3508,
+      "step": 7717
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.743725776672363,
+      "learning_rate": 1.01321434700532e-05,
+      "loss": 0.4247,
+      "step": 7718
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.971477508544922,
+      "learning_rate": 1.0129569246610606e-05,
+      "loss": 0.6408,
+      "step": 7719
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.370979309082031,
+      "learning_rate": 1.012699502316801e-05,
+      "loss": 0.5155,
+      "step": 7720
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.87381649017334,
+      "learning_rate": 1.0124420799725416e-05,
+      "loss": 0.5846,
+      "step": 7721
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.795620918273926,
+      "learning_rate": 1.0121846576282822e-05,
+      "loss": 0.6393,
+      "step": 7722
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.147948265075684,
+      "learning_rate": 1.0119272352840227e-05,
+      "loss": 0.4861,
+      "step": 7723
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.977652549743652,
+      "learning_rate": 1.0116698129397632e-05,
+      "loss": 0.557,
+      "step": 7724
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.888734817504883,
+      "learning_rate": 1.0114123905955037e-05,
+      "loss": 0.5015,
+      "step": 7725
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.407049655914307,
+      "learning_rate": 1.0111549682512442e-05,
+      "loss": 0.2436,
+      "step": 7726
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.32789421081543,
+      "learning_rate": 1.0108975459069847e-05,
+      "loss": 0.3552,
+      "step": 7727
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.634115219116211,
+      "learning_rate": 1.0106401235627252e-05,
+      "loss": 0.4183,
+      "step": 7728
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.24951171875,
+      "learning_rate": 1.0103827012184657e-05,
+      "loss": 0.4233,
+      "step": 7729
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.46374797821045,
+      "learning_rate": 1.0101252788742062e-05,
+      "loss": 0.4374,
+      "step": 7730
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.346308708190918,
+      "learning_rate": 1.0098678565299469e-05,
+      "loss": 0.4001,
+      "step": 7731
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.14107894897461,
+      "learning_rate": 1.0096104341856874e-05,
+      "loss": 0.3783,
+      "step": 7732
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.576476097106934,
+      "learning_rate": 1.0093530118414279e-05,
+      "loss": 0.4913,
+      "step": 7733
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.956659317016602,
+      "learning_rate": 1.0090955894971684e-05,
+      "loss": 0.5134,
+      "step": 7734
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.471206665039062,
+      "learning_rate": 1.0088381671529089e-05,
+      "loss": 0.4451,
+      "step": 7735
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.790603637695312,
+      "learning_rate": 1.0085807448086496e-05,
+      "loss": 0.5092,
+      "step": 7736
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.110894203186035,
+      "learning_rate": 1.0083233224643899e-05,
+      "loss": 0.3445,
+      "step": 7737
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.716071128845215,
+      "learning_rate": 1.0080659001201304e-05,
+      "loss": 0.5015,
+      "step": 7738
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.373269081115723,
+      "learning_rate": 1.0078084777758709e-05,
+      "loss": 0.5107,
+      "step": 7739
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.19774341583252,
+      "learning_rate": 1.0075510554316114e-05,
+      "loss": 0.3752,
+      "step": 7740
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.300834655761719,
+      "learning_rate": 1.007293633087352e-05,
+      "loss": 0.444,
+      "step": 7741
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.19477367401123,
+      "learning_rate": 1.0070362107430926e-05,
+      "loss": 0.4985,
+      "step": 7742
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.454253196716309,
+      "learning_rate": 1.006778788398833e-05,
+      "loss": 0.399,
+      "step": 7743
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.702960968017578,
+      "learning_rate": 1.0065213660545735e-05,
+      "loss": 0.4912,
+      "step": 7744
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.439407348632812,
+      "learning_rate": 1.0062639437103142e-05,
+      "loss": 0.4091,
+      "step": 7745
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.278610229492188,
+      "learning_rate": 1.0060065213660547e-05,
+      "loss": 0.4811,
+      "step": 7746
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.671419143676758,
+      "learning_rate": 1.005749099021795e-05,
+      "loss": 0.5109,
+      "step": 7747
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.953645706176758,
+      "learning_rate": 1.0054916766775355e-05,
+      "loss": 0.4394,
+      "step": 7748
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.745628356933594,
+      "learning_rate": 1.005234254333276e-05,
+      "loss": 0.5516,
+      "step": 7749
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.74183464050293,
+      "learning_rate": 1.0049768319890167e-05,
+      "loss": 0.6131,
+      "step": 7750
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.532737731933594,
+      "learning_rate": 1.0047194096447572e-05,
+      "loss": 0.4789,
+      "step": 7751
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.661445617675781,
+      "learning_rate": 1.0044619873004977e-05,
+      "loss": 0.4724,
+      "step": 7752
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.972615242004395,
+      "learning_rate": 1.0042045649562382e-05,
+      "loss": 0.5184,
+      "step": 7753
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.499295234680176,
+      "learning_rate": 1.0039471426119787e-05,
+      "loss": 0.3339,
+      "step": 7754
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.487306594848633,
+      "learning_rate": 1.0036897202677194e-05,
+      "loss": 0.2722,
+      "step": 7755
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.158554077148438,
+      "learning_rate": 1.0034322979234599e-05,
+      "loss": 0.5955,
+      "step": 7756
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.339704513549805,
+      "learning_rate": 1.0031748755792004e-05,
+      "loss": 0.3881,
+      "step": 7757
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.292365074157715,
+      "learning_rate": 1.0029174532349407e-05,
+      "loss": 0.6178,
+      "step": 7758
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.204378128051758,
+      "learning_rate": 1.0026600308906812e-05,
+      "loss": 0.3679,
+      "step": 7759
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.968137741088867,
+      "learning_rate": 1.0024026085464219e-05,
+      "loss": 0.7739,
+      "step": 7760
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.328302383422852,
+      "learning_rate": 1.0021451862021624e-05,
+      "loss": 0.4007,
+      "step": 7761
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.331490516662598,
+      "learning_rate": 1.0018877638579029e-05,
+      "loss": 0.5925,
+      "step": 7762
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.456241607666016,
+      "learning_rate": 1.0016303415136434e-05,
+      "loss": 0.4873,
+      "step": 7763
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.09859848022461,
+      "learning_rate": 1.001372919169384e-05,
+      "loss": 0.2885,
+      "step": 7764
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.258848190307617,
+      "learning_rate": 1.0011154968251245e-05,
+      "loss": 0.4692,
+      "step": 7765
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.38225269317627,
+      "learning_rate": 1.000858074480865e-05,
+      "loss": 0.7009,
+      "step": 7766
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.7805757522583,
+      "learning_rate": 1.0006006521366055e-05,
+      "loss": 0.7279,
+      "step": 7767
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.964977264404297,
+      "learning_rate": 1.0003432297923459e-05,
+      "loss": 0.4061,
+      "step": 7768
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.108419418334961,
+      "learning_rate": 1.0000858074480865e-05,
+      "loss": 0.4724,
+      "step": 7769
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.253042221069336,
+      "learning_rate": 9.99828385103827e-06,
+      "loss": 0.7647,
+      "step": 7770
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.767439842224121,
+      "learning_rate": 9.995709627595675e-06,
+      "loss": 0.5012,
+      "step": 7771
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.6414794921875,
+      "learning_rate": 9.99313540415308e-06,
+      "loss": 0.4409,
+      "step": 7772
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.81783390045166,
+      "learning_rate": 9.990561180710485e-06,
+      "loss": 0.4807,
+      "step": 7773
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.49046516418457,
+      "learning_rate": 9.987986957267892e-06,
+      "loss": 0.4881,
+      "step": 7774
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.088298797607422,
+      "learning_rate": 9.985412733825297e-06,
+      "loss": 0.3567,
+      "step": 7775
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.536469459533691,
+      "learning_rate": 9.982838510382702e-06,
+      "loss": 0.5164,
+      "step": 7776
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.29196834564209,
+      "learning_rate": 9.980264286940107e-06,
+      "loss": 0.3726,
+      "step": 7777
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.835789680480957,
+      "learning_rate": 9.977690063497512e-06,
+      "loss": 0.724,
+      "step": 7778
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.835392951965332,
+      "learning_rate": 9.975115840054917e-06,
+      "loss": 0.5318,
+      "step": 7779
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.065471649169922,
+      "learning_rate": 9.972541616612322e-06,
+      "loss": 0.2848,
+      "step": 7780
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.572319030761719,
+      "learning_rate": 9.969967393169727e-06,
+      "loss": 0.4961,
+      "step": 7781
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.702765464782715,
+      "learning_rate": 9.967393169727132e-06,
+      "loss": 0.576,
+      "step": 7782
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.804869651794434,
+      "learning_rate": 9.964818946284538e-06,
+      "loss": 0.3448,
+      "step": 7783
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.074567794799805,
+      "learning_rate": 9.962244722841943e-06,
+      "loss": 0.4921,
+      "step": 7784
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.345091819763184,
+      "learning_rate": 9.959670499399348e-06,
+      "loss": 0.5686,
+      "step": 7785
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.821876525878906,
+      "learning_rate": 9.957096275956753e-06,
+      "loss": 0.3736,
+      "step": 7786
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.495122909545898,
+      "learning_rate": 9.954522052514158e-06,
+      "loss": 0.4628,
+      "step": 7787
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.096467018127441,
+      "learning_rate": 9.951947829071565e-06,
+      "loss": 0.5182,
+      "step": 7788
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.715571403503418,
+      "learning_rate": 9.949373605628968e-06,
+      "loss": 0.3708,
+      "step": 7789
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.558807373046875,
+      "learning_rate": 9.946799382186373e-06,
+      "loss": 0.5762,
+      "step": 7790
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.3325777053833,
+      "learning_rate": 9.944225158743778e-06,
+      "loss": 0.5207,
+      "step": 7791
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.696695804595947,
+      "learning_rate": 9.941650935301183e-06,
+      "loss": 0.4032,
+      "step": 7792
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.648896217346191,
+      "learning_rate": 9.93907671185859e-06,
+      "loss": 0.5357,
+      "step": 7793
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.576841354370117,
+      "learning_rate": 9.936502488415995e-06,
+      "loss": 0.5056,
+      "step": 7794
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.248518943786621,
+      "learning_rate": 9.9339282649734e-06,
+      "loss": 0.403,
+      "step": 7795
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.282525062561035,
+      "learning_rate": 9.931354041530805e-06,
+      "loss": 0.4931,
+      "step": 7796
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.213623046875,
+      "learning_rate": 9.928779818088212e-06,
+      "loss": 0.4309,
+      "step": 7797
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.758950233459473,
+      "learning_rate": 9.926205594645617e-06,
+      "loss": 0.4596,
+      "step": 7798
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.991473197937012,
+      "learning_rate": 9.92363137120302e-06,
+      "loss": 0.361,
+      "step": 7799
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.212671279907227,
+      "learning_rate": 9.921057147760425e-06,
+      "loss": 0.3761,
+      "step": 7800
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.036935806274414,
+      "learning_rate": 9.91848292431783e-06,
+      "loss": 0.5292,
+      "step": 7801
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.830004692077637,
+      "learning_rate": 9.915908700875237e-06,
+      "loss": 0.4981,
+      "step": 7802
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.893874168395996,
+      "learning_rate": 9.913334477432642e-06,
+      "loss": 0.6167,
+      "step": 7803
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.476625442504883,
+      "learning_rate": 9.910760253990047e-06,
+      "loss": 0.2795,
+      "step": 7804
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.794510841369629,
+      "learning_rate": 9.908186030547452e-06,
+      "loss": 0.3916,
+      "step": 7805
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.44105339050293,
+      "learning_rate": 9.905611807104857e-06,
+      "loss": 0.3235,
+      "step": 7806
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.826579093933105,
+      "learning_rate": 9.903037583662263e-06,
+      "loss": 0.5203,
+      "step": 7807
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.6405029296875,
+      "learning_rate": 9.900463360219668e-06,
+      "loss": 0.4138,
+      "step": 7808
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.390388488769531,
+      "learning_rate": 9.897889136777073e-06,
+      "loss": 0.6336,
+      "step": 7809
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.235594749450684,
+      "learning_rate": 9.895314913334477e-06,
+      "loss": 0.4451,
+      "step": 7810
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.92341423034668,
+      "learning_rate": 9.892740689891882e-06,
+      "loss": 0.6219,
+      "step": 7811
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 20.03653907775879,
+      "learning_rate": 9.890166466449288e-06,
+      "loss": 0.5885,
+      "step": 7812
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.7859978675842285,
+      "learning_rate": 9.887592243006693e-06,
+      "loss": 0.2507,
+      "step": 7813
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.49488353729248,
+      "learning_rate": 9.885018019564098e-06,
+      "loss": 0.2834,
+      "step": 7814
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.413704872131348,
+      "learning_rate": 9.882443796121503e-06,
+      "loss": 0.432,
+      "step": 7815
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.287511825561523,
+      "learning_rate": 9.87986957267891e-06,
+      "loss": 0.5256,
+      "step": 7816
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.119768142700195,
+      "learning_rate": 9.877295349236315e-06,
+      "loss": 0.3605,
+      "step": 7817
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.733137130737305,
+      "learning_rate": 9.87472112579372e-06,
+      "loss": 0.4332,
+      "step": 7818
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.899484634399414,
+      "learning_rate": 9.872146902351125e-06,
+      "loss": 0.338,
+      "step": 7819
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.59583854675293,
+      "learning_rate": 9.869572678908528e-06,
+      "loss": 0.6533,
+      "step": 7820
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.460033416748047,
+      "learning_rate": 9.866998455465935e-06,
+      "loss": 0.6124,
+      "step": 7821
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.69294548034668,
+      "learning_rate": 9.86442423202334e-06,
+      "loss": 0.5059,
+      "step": 7822
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.422442436218262,
+      "learning_rate": 9.861850008580745e-06,
+      "loss": 0.5904,
+      "step": 7823
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.567588806152344,
+      "learning_rate": 9.85927578513815e-06,
+      "loss": 0.6012,
+      "step": 7824
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.430335998535156,
+      "learning_rate": 9.856701561695555e-06,
+      "loss": 0.3941,
+      "step": 7825
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.163117408752441,
+      "learning_rate": 9.854127338252961e-06,
+      "loss": 0.7988,
+      "step": 7826
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.376513481140137,
+      "learning_rate": 9.851553114810366e-06,
+      "loss": 0.4682,
+      "step": 7827
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.4733304977417,
+      "learning_rate": 9.848978891367771e-06,
+      "loss": 0.4803,
+      "step": 7828
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.11894702911377,
+      "learning_rate": 9.846404667925176e-06,
+      "loss": 0.5094,
+      "step": 7829
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.281590461730957,
+      "learning_rate": 9.843830444482581e-06,
+      "loss": 0.5888,
+      "step": 7830
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.856283187866211,
+      "learning_rate": 9.841256221039986e-06,
+      "loss": 0.6041,
+      "step": 7831
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.023073196411133,
+      "learning_rate": 9.838681997597391e-06,
+      "loss": 0.336,
+      "step": 7832
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.655364990234375,
+      "learning_rate": 9.836107774154796e-06,
+      "loss": 0.4137,
+      "step": 7833
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.6759033203125,
+      "learning_rate": 9.833533550712201e-06,
+      "loss": 0.5383,
+      "step": 7834
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.055542945861816,
+      "learning_rate": 9.830959327269608e-06,
+      "loss": 0.4429,
+      "step": 7835
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.497017860412598,
+      "learning_rate": 9.828385103827013e-06,
+      "loss": 0.6196,
+      "step": 7836
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.390766143798828,
+      "learning_rate": 9.825810880384418e-06,
+      "loss": 0.3786,
+      "step": 7837
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.510671615600586,
+      "learning_rate": 9.823236656941823e-06,
+      "loss": 0.5683,
+      "step": 7838
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.21231460571289,
+      "learning_rate": 9.820662433499228e-06,
+      "loss": 0.3756,
+      "step": 7839
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.280329704284668,
+      "learning_rate": 9.818088210056635e-06,
+      "loss": 0.3279,
+      "step": 7840
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.856691360473633,
+      "learning_rate": 9.815513986614038e-06,
+      "loss": 0.5336,
+      "step": 7841
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.319304466247559,
+      "learning_rate": 9.812939763171443e-06,
+      "loss": 0.3987,
+      "step": 7842
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.5609712600708,
+      "learning_rate": 9.810365539728848e-06,
+      "loss": 0.5165,
+      "step": 7843
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.934987545013428,
+      "learning_rate": 9.807791316286253e-06,
+      "loss": 0.3615,
+      "step": 7844
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.290690422058105,
+      "learning_rate": 9.80521709284366e-06,
+      "loss": 0.3594,
+      "step": 7845
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.969504356384277,
+      "learning_rate": 9.802642869401065e-06,
+      "loss": 0.6047,
+      "step": 7846
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.453792572021484,
+      "learning_rate": 9.80006864595847e-06,
+      "loss": 0.5547,
+      "step": 7847
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.964442253112793,
+      "learning_rate": 9.797494422515875e-06,
+      "loss": 0.512,
+      "step": 7848
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.762725830078125,
+      "learning_rate": 9.794920199073281e-06,
+      "loss": 0.4341,
+      "step": 7849
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.61126708984375,
+      "learning_rate": 9.792345975630686e-06,
+      "loss": 0.7721,
+      "step": 7850
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.368814468383789,
+      "learning_rate": 9.78977175218809e-06,
+      "loss": 0.5659,
+      "step": 7851
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.620594024658203,
+      "learning_rate": 9.787197528745494e-06,
+      "loss": 0.4628,
+      "step": 7852
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.326383590698242,
+      "learning_rate": 9.7846233053029e-06,
+      "loss": 0.5452,
+      "step": 7853
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.69447135925293,
+      "learning_rate": 9.782049081860306e-06,
+      "loss": 0.6278,
+      "step": 7854
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.626816749572754,
+      "learning_rate": 9.779474858417711e-06,
+      "loss": 0.5241,
+      "step": 7855
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.496217727661133,
+      "learning_rate": 9.776900634975116e-06,
+      "loss": 0.3537,
+      "step": 7856
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.960626602172852,
+      "learning_rate": 9.774326411532521e-06,
+      "loss": 0.5719,
+      "step": 7857
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.040411949157715,
+      "learning_rate": 9.771752188089926e-06,
+      "loss": 0.3659,
+      "step": 7858
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.668703079223633,
+      "learning_rate": 9.769177964647333e-06,
+      "loss": 0.3318,
+      "step": 7859
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.654769897460938,
+      "learning_rate": 9.766603741204738e-06,
+      "loss": 0.5931,
+      "step": 7860
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.421550750732422,
+      "learning_rate": 9.764029517762143e-06,
+      "loss": 0.3518,
+      "step": 7861
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.402252197265625,
+      "learning_rate": 9.761455294319546e-06,
+      "loss": 0.4765,
+      "step": 7862
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.763633728027344,
+      "learning_rate": 9.758881070876951e-06,
+      "loss": 0.249,
+      "step": 7863
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.121054649353027,
+      "learning_rate": 9.756306847434358e-06,
+      "loss": 0.4757,
+      "step": 7864
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.769430160522461,
+      "learning_rate": 9.753732623991763e-06,
+      "loss": 0.5496,
+      "step": 7865
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.768556594848633,
+      "learning_rate": 9.751158400549168e-06,
+      "loss": 0.6156,
+      "step": 7866
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.153548240661621,
+      "learning_rate": 9.748584177106573e-06,
+      "loss": 0.6403,
+      "step": 7867
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.739583969116211,
+      "learning_rate": 9.74600995366398e-06,
+      "loss": 0.5516,
+      "step": 7868
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.17764949798584,
+      "learning_rate": 9.743435730221384e-06,
+      "loss": 0.4273,
+      "step": 7869
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.148964881896973,
+      "learning_rate": 9.74086150677879e-06,
+      "loss": 0.4761,
+      "step": 7870
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.738555908203125,
+      "learning_rate": 9.738287283336194e-06,
+      "loss": 0.561,
+      "step": 7871
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.561820030212402,
+      "learning_rate": 9.735713059893598e-06,
+      "loss": 0.4002,
+      "step": 7872
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.05392837524414,
+      "learning_rate": 9.733138836451004e-06,
+      "loss": 0.5664,
+      "step": 7873
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.910451889038086,
+      "learning_rate": 9.73056461300841e-06,
+      "loss": 0.4485,
+      "step": 7874
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.862271308898926,
+      "learning_rate": 9.727990389565814e-06,
+      "loss": 0.5186,
+      "step": 7875
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.9047212600708,
+      "learning_rate": 9.72541616612322e-06,
+      "loss": 0.399,
+      "step": 7876
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.42728042602539,
+      "learning_rate": 9.722841942680624e-06,
+      "loss": 0.4852,
+      "step": 7877
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.16113567352295,
+      "learning_rate": 9.720267719238031e-06,
+      "loss": 0.4994,
+      "step": 7878
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.269696235656738,
+      "learning_rate": 9.717693495795436e-06,
+      "loss": 0.4599,
+      "step": 7879
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.118162155151367,
+      "learning_rate": 9.715119272352841e-06,
+      "loss": 0.494,
+      "step": 7880
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.150639533996582,
+      "learning_rate": 9.712545048910246e-06,
+      "loss": 0.4689,
+      "step": 7881
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.063127517700195,
+      "learning_rate": 9.709970825467651e-06,
+      "loss": 0.485,
+      "step": 7882
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.748597145080566,
+      "learning_rate": 9.707396602025056e-06,
+      "loss": 0.4731,
+      "step": 7883
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.388176918029785,
+      "learning_rate": 9.70482237858246e-06,
+      "loss": 0.5719,
+      "step": 7884
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.090357780456543,
+      "learning_rate": 9.702248155139866e-06,
+      "loss": 0.4772,
+      "step": 7885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.651619911193848,
+      "learning_rate": 9.69967393169727e-06,
+      "loss": 0.3959,
+      "step": 7886
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.236624717712402,
+      "learning_rate": 9.697099708254677e-06,
+      "loss": 0.3725,
+      "step": 7887
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.64133071899414,
+      "learning_rate": 9.694525484812082e-06,
+      "loss": 0.4929,
+      "step": 7888
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.256224632263184,
+      "learning_rate": 9.691951261369487e-06,
+      "loss": 0.4727,
+      "step": 7889
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.375576972961426,
+      "learning_rate": 9.689377037926892e-06,
+      "loss": 0.4567,
+      "step": 7890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.470894813537598,
+      "learning_rate": 9.686802814484297e-06,
+      "loss": 0.5636,
+      "step": 7891
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.24980354309082,
+      "learning_rate": 9.684228591041704e-06,
+      "loss": 0.6526,
+      "step": 7892
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.546586990356445,
+      "learning_rate": 9.681654367599107e-06,
+      "loss": 0.3554,
+      "step": 7893
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.096156120300293,
+      "learning_rate": 9.679080144156512e-06,
+      "loss": 0.3018,
+      "step": 7894
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.843267440795898,
+      "learning_rate": 9.676505920713917e-06,
+      "loss": 0.4242,
+      "step": 7895
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.440269470214844,
+      "learning_rate": 9.673931697271322e-06,
+      "loss": 0.5666,
+      "step": 7896
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.503560066223145,
+      "learning_rate": 9.671357473828729e-06,
+      "loss": 0.3638,
+      "step": 7897
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.833135604858398,
+      "learning_rate": 9.668783250386134e-06,
+      "loss": 0.342,
+      "step": 7898
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.183473587036133,
+      "learning_rate": 9.666209026943539e-06,
+      "loss": 0.6704,
+      "step": 7899
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.344062328338623,
+      "learning_rate": 9.663634803500944e-06,
+      "loss": 0.3328,
+      "step": 7900
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.75399398803711,
+      "learning_rate": 9.66106058005835e-06,
+      "loss": 0.716,
+      "step": 7901
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.551011085510254,
+      "learning_rate": 9.658486356615756e-06,
+      "loss": 0.4762,
+      "step": 7902
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.75141716003418,
+      "learning_rate": 9.655912133173159e-06,
+      "loss": 0.5188,
+      "step": 7903
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.248319625854492,
+      "learning_rate": 9.653337909730564e-06,
+      "loss": 0.4587,
+      "step": 7904
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.654538154602051,
+      "learning_rate": 9.650763686287969e-06,
+      "loss": 0.3596,
+      "step": 7905
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.485015869140625,
+      "learning_rate": 9.648189462845376e-06,
+      "loss": 0.7514,
+      "step": 7906
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.851784706115723,
+      "learning_rate": 9.64561523940278e-06,
+      "loss": 0.3519,
+      "step": 7907
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.043368339538574,
+      "learning_rate": 9.643041015960186e-06,
+      "loss": 0.4275,
+      "step": 7908
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.457150459289551,
+      "learning_rate": 9.64046679251759e-06,
+      "loss": 0.2649,
+      "step": 7909
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.928865432739258,
+      "learning_rate": 9.637892569074996e-06,
+      "loss": 0.7033,
+      "step": 7910
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.327780723571777,
+      "learning_rate": 9.635318345632402e-06,
+      "loss": 0.6201,
+      "step": 7911
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.8857502937316895,
+      "learning_rate": 9.632744122189807e-06,
+      "loss": 0.3789,
+      "step": 7912
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.756880760192871,
+      "learning_rate": 9.630169898747212e-06,
+      "loss": 0.3884,
+      "step": 7913
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.089845657348633,
+      "learning_rate": 9.627595675304616e-06,
+      "loss": 0.4721,
+      "step": 7914
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.617827415466309,
+      "learning_rate": 9.62502145186202e-06,
+      "loss": 0.4425,
+      "step": 7915
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.050688743591309,
+      "learning_rate": 9.622447228419427e-06,
+      "loss": 0.4432,
+      "step": 7916
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.98144817352295,
+      "learning_rate": 9.619873004976832e-06,
+      "loss": 0.494,
+      "step": 7917
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.91801643371582,
+      "learning_rate": 9.617298781534237e-06,
+      "loss": 0.3389,
+      "step": 7918
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.546061515808105,
+      "learning_rate": 9.614724558091642e-06,
+      "loss": 0.5805,
+      "step": 7919
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.63851547241211,
+      "learning_rate": 9.612150334649049e-06,
+      "loss": 0.38,
+      "step": 7920
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.369551658630371,
+      "learning_rate": 9.609576111206454e-06,
+      "loss": 0.3167,
+      "step": 7921
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.338345527648926,
+      "learning_rate": 9.607001887763859e-06,
+      "loss": 0.3936,
+      "step": 7922
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.559334754943848,
+      "learning_rate": 9.604427664321264e-06,
+      "loss": 0.5655,
+      "step": 7923
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.715849876403809,
+      "learning_rate": 9.601853440878667e-06,
+      "loss": 0.6027,
+      "step": 7924
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.786272048950195,
+      "learning_rate": 9.599279217436074e-06,
+      "loss": 0.5646,
+      "step": 7925
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.754862785339355,
+      "learning_rate": 9.596704993993479e-06,
+      "loss": 0.5029,
+      "step": 7926
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.600534439086914,
+      "learning_rate": 9.594130770550884e-06,
+      "loss": 0.6923,
+      "step": 7927
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.364373207092285,
+      "learning_rate": 9.591556547108289e-06,
+      "loss": 0.3036,
+      "step": 7928
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.808643341064453,
+      "learning_rate": 9.588982323665694e-06,
+      "loss": 0.6615,
+      "step": 7929
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.058235168457031,
+      "learning_rate": 9.5864081002231e-06,
+      "loss": 0.7172,
+      "step": 7930
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.112159729003906,
+      "learning_rate": 9.583833876780505e-06,
+      "loss": 0.3329,
+      "step": 7931
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.51227855682373,
+      "learning_rate": 9.58125965333791e-06,
+      "loss": 0.887,
+      "step": 7932
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.956052780151367,
+      "learning_rate": 9.578685429895315e-06,
+      "loss": 0.4386,
+      "step": 7933
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.272078514099121,
+      "learning_rate": 9.57611120645272e-06,
+      "loss": 0.451,
+      "step": 7934
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.831804275512695,
+      "learning_rate": 9.573536983010125e-06,
+      "loss": 0.4947,
+      "step": 7935
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.690449714660645,
+      "learning_rate": 9.57096275956753e-06,
+      "loss": 0.5057,
+      "step": 7936
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.02230453491211,
+      "learning_rate": 9.568388536124935e-06,
+      "loss": 0.5295,
+      "step": 7937
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.09279727935791,
+      "learning_rate": 9.56581431268234e-06,
+      "loss": 0.4545,
+      "step": 7938
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.567049026489258,
+      "learning_rate": 9.563240089239747e-06,
+      "loss": 0.4609,
+      "step": 7939
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.060547828674316,
+      "learning_rate": 9.560665865797152e-06,
+      "loss": 0.3896,
+      "step": 7940
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.55778980255127,
+      "learning_rate": 9.558091642354557e-06,
+      "loss": 0.3546,
+      "step": 7941
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.26447868347168,
+      "learning_rate": 9.555517418911962e-06,
+      "loss": 0.4502,
+      "step": 7942
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.228554725646973,
+      "learning_rate": 9.552943195469367e-06,
+      "loss": 0.297,
+      "step": 7943
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.135771751403809,
+      "learning_rate": 9.550368972026774e-06,
+      "loss": 0.5123,
+      "step": 7944
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.186357498168945,
+      "learning_rate": 9.547794748584177e-06,
+      "loss": 0.5235,
+      "step": 7945
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.003829956054688,
+      "learning_rate": 9.545220525141582e-06,
+      "loss": 0.5746,
+      "step": 7946
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.436480522155762,
+      "learning_rate": 9.542646301698987e-06,
+      "loss": 0.3595,
+      "step": 7947
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.297316551208496,
+      "learning_rate": 9.540072078256392e-06,
+      "loss": 0.5374,
+      "step": 7948
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.019418716430664,
+      "learning_rate": 9.537497854813799e-06,
+      "loss": 0.4533,
+      "step": 7949
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.778726577758789,
+      "learning_rate": 9.534923631371204e-06,
+      "loss": 0.2627,
+      "step": 7950
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.137608528137207,
+      "learning_rate": 9.532349407928609e-06,
+      "loss": 0.4314,
+      "step": 7951
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.141313552856445,
+      "learning_rate": 9.529775184486014e-06,
+      "loss": 0.5498,
+      "step": 7952
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.387163162231445,
+      "learning_rate": 9.527200961043419e-06,
+      "loss": 0.3944,
+      "step": 7953
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.75260066986084,
+      "learning_rate": 9.524626737600825e-06,
+      "loss": 0.4656,
+      "step": 7954
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.301797866821289,
+      "learning_rate": 9.522052514158228e-06,
+      "loss": 0.4064,
+      "step": 7955
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.34122085571289,
+      "learning_rate": 9.519478290715633e-06,
+      "loss": 0.5799,
+      "step": 7956
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.564830780029297,
+      "learning_rate": 9.516904067273038e-06,
+      "loss": 0.403,
+      "step": 7957
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.964445114135742,
+      "learning_rate": 9.514329843830445e-06,
+      "loss": 0.3114,
+      "step": 7958
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.449462890625,
+      "learning_rate": 9.51175562038785e-06,
+      "loss": 0.5315,
+      "step": 7959
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.97594451904297,
+      "learning_rate": 9.509181396945255e-06,
+      "loss": 0.595,
+      "step": 7960
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.211369514465332,
+      "learning_rate": 9.50660717350266e-06,
+      "loss": 0.5134,
+      "step": 7961
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.999001502990723,
+      "learning_rate": 9.504032950060065e-06,
+      "loss": 0.4178,
+      "step": 7962
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.591569900512695,
+      "learning_rate": 9.501458726617472e-06,
+      "loss": 0.3171,
+      "step": 7963
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.294441223144531,
+      "learning_rate": 9.498884503174877e-06,
+      "loss": 0.5068,
+      "step": 7964
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.029129028320312,
+      "learning_rate": 9.496310279732282e-06,
+      "loss": 0.5429,
+      "step": 7965
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.49426555633545,
+      "learning_rate": 9.493736056289685e-06,
+      "loss": 0.5535,
+      "step": 7966
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.20008659362793,
+      "learning_rate": 9.49116183284709e-06,
+      "loss": 0.3495,
+      "step": 7967
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.031464576721191,
+      "learning_rate": 9.488587609404497e-06,
+      "loss": 0.3439,
+      "step": 7968
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.443595886230469,
+      "learning_rate": 9.486013385961902e-06,
+      "loss": 0.4986,
+      "step": 7969
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.72297191619873,
+      "learning_rate": 9.483439162519307e-06,
+      "loss": 0.3509,
+      "step": 7970
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.723140716552734,
+      "learning_rate": 9.480864939076712e-06,
+      "loss": 0.6552,
+      "step": 7971
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.96426010131836,
+      "learning_rate": 9.478290715634118e-06,
+      "loss": 0.423,
+      "step": 7972
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.09008502960205,
+      "learning_rate": 9.475716492191523e-06,
+      "loss": 0.4553,
+      "step": 7973
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.585437297821045,
+      "learning_rate": 9.473142268748928e-06,
+      "loss": 0.3397,
+      "step": 7974
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.272289276123047,
+      "learning_rate": 9.470568045306333e-06,
+      "loss": 0.3997,
+      "step": 7975
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.922517776489258,
+      "learning_rate": 9.467993821863737e-06,
+      "loss": 0.3701,
+      "step": 7976
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.318075180053711,
+      "learning_rate": 9.465419598421143e-06,
+      "loss": 0.3566,
+      "step": 7977
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.127506256103516,
+      "learning_rate": 9.462845374978548e-06,
+      "loss": 0.4298,
+      "step": 7978
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.943914413452148,
+      "learning_rate": 9.460271151535953e-06,
+      "loss": 0.5534,
+      "step": 7979
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.463277816772461,
+      "learning_rate": 9.457696928093358e-06,
+      "loss": 0.3953,
+      "step": 7980
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.071959495544434,
+      "learning_rate": 9.455122704650763e-06,
+      "loss": 0.5181,
+      "step": 7981
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.230708122253418,
+      "learning_rate": 9.45254848120817e-06,
+      "loss": 0.4316,
+      "step": 7982
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.167973518371582,
+      "learning_rate": 9.449974257765575e-06,
+      "loss": 0.5694,
+      "step": 7983
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.809586524963379,
+      "learning_rate": 9.44740003432298e-06,
+      "loss": 0.4548,
+      "step": 7984
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.861859321594238,
+      "learning_rate": 9.444825810880385e-06,
+      "loss": 0.4381,
+      "step": 7985
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.695276260375977,
+      "learning_rate": 9.44225158743779e-06,
+      "loss": 0.4095,
+      "step": 7986
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.268562316894531,
+      "learning_rate": 9.439677363995195e-06,
+      "loss": 0.4532,
+      "step": 7987
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.070253372192383,
+      "learning_rate": 9.4371031405526e-06,
+      "loss": 0.4005,
+      "step": 7988
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.023111820220947,
+      "learning_rate": 9.434528917110005e-06,
+      "loss": 0.3017,
+      "step": 7989
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.786094665527344,
+      "learning_rate": 9.43195469366741e-06,
+      "loss": 0.4113,
+      "step": 7990
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.137983322143555,
+      "learning_rate": 9.429380470224816e-06,
+      "loss": 0.3046,
+      "step": 7991
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.733827590942383,
+      "learning_rate": 9.426806246782221e-06,
+      "loss": 0.6012,
+      "step": 7992
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.344902992248535,
+      "learning_rate": 9.424232023339626e-06,
+      "loss": 0.4224,
+      "step": 7993
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.666622161865234,
+      "learning_rate": 9.421657799897031e-06,
+      "loss": 0.5556,
+      "step": 7994
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.04263973236084,
+      "learning_rate": 9.419083576454436e-06,
+      "loss": 0.4013,
+      "step": 7995
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.478981971740723,
+      "learning_rate": 9.416509353011843e-06,
+      "loss": 0.4965,
+      "step": 7996
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.716177940368652,
+      "learning_rate": 9.413935129569246e-06,
+      "loss": 0.5522,
+      "step": 7997
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.81005859375,
+      "learning_rate": 9.411360906126651e-06,
+      "loss": 0.6467,
+      "step": 7998
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.434504508972168,
+      "learning_rate": 9.408786682684056e-06,
+      "loss": 0.3804,
+      "step": 7999
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.765411376953125,
+      "learning_rate": 9.406212459241461e-06,
+      "loss": 0.5495,
+      "step": 8000
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.332803726196289,
+      "learning_rate": 9.403638235798868e-06,
+      "loss": 0.5431,
+      "step": 8001
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.456473350524902,
+      "learning_rate": 9.401064012356273e-06,
+      "loss": 0.5641,
+      "step": 8002
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.626286506652832,
+      "learning_rate": 9.398489788913678e-06,
+      "loss": 0.4688,
+      "step": 8003
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.766304016113281,
+      "learning_rate": 9.395915565471083e-06,
+      "loss": 0.4206,
+      "step": 8004
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.014063835144043,
+      "learning_rate": 9.393341342028488e-06,
+      "loss": 0.5907,
+      "step": 8005
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.282660484313965,
+      "learning_rate": 9.390767118585895e-06,
+      "loss": 0.5397,
+      "step": 8006
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.133003234863281,
+      "learning_rate": 9.388192895143298e-06,
+      "loss": 0.3617,
+      "step": 8007
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.292089462280273,
+      "learning_rate": 9.385618671700703e-06,
+      "loss": 0.4716,
+      "step": 8008
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.585433959960938,
+      "learning_rate": 9.383044448258108e-06,
+      "loss": 0.4841,
+      "step": 8009
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.634963035583496,
+      "learning_rate": 9.380470224815515e-06,
+      "loss": 0.3696,
+      "step": 8010
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.995577812194824,
+      "learning_rate": 9.37789600137292e-06,
+      "loss": 0.5785,
+      "step": 8011
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.64609718322754,
+      "learning_rate": 9.375321777930325e-06,
+      "loss": 0.6242,
+      "step": 8012
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.982150077819824,
+      "learning_rate": 9.37274755448773e-06,
+      "loss": 0.3501,
+      "step": 8013
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.285284996032715,
+      "learning_rate": 9.370173331045135e-06,
+      "loss": 0.6264,
+      "step": 8014
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.575113773345947,
+      "learning_rate": 9.367599107602541e-06,
+      "loss": 0.3797,
+      "step": 8015
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.67357349395752,
+      "learning_rate": 9.365024884159946e-06,
+      "loss": 0.3867,
+      "step": 8016
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.106900215148926,
+      "learning_rate": 9.362450660717351e-06,
+      "loss": 0.3732,
+      "step": 8017
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.497732162475586,
+      "learning_rate": 9.359876437274755e-06,
+      "loss": 0.7343,
+      "step": 8018
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.809659481048584,
+      "learning_rate": 9.35730221383216e-06,
+      "loss": 0.4074,
+      "step": 8019
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.506635665893555,
+      "learning_rate": 9.354727990389566e-06,
+      "loss": 0.406,
+      "step": 8020
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.403267860412598,
+      "learning_rate": 9.352153766946971e-06,
+      "loss": 0.5285,
+      "step": 8021
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.269692420959473,
+      "learning_rate": 9.349579543504376e-06,
+      "loss": 0.5486,
+      "step": 8022
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.832744598388672,
+      "learning_rate": 9.347005320061781e-06,
+      "loss": 0.3635,
+      "step": 8023
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.928349494934082,
+      "learning_rate": 9.344431096619188e-06,
+      "loss": 0.3154,
+      "step": 8024
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.598359107971191,
+      "learning_rate": 9.341856873176593e-06,
+      "loss": 0.4291,
+      "step": 8025
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.557576179504395,
+      "learning_rate": 9.339282649733998e-06,
+      "loss": 0.4514,
+      "step": 8026
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.876992225646973,
+      "learning_rate": 9.336708426291403e-06,
+      "loss": 0.4586,
+      "step": 8027
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.892364501953125,
+      "learning_rate": 9.334134202848806e-06,
+      "loss": 0.5799,
+      "step": 8028
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.67786693572998,
+      "learning_rate": 9.331559979406213e-06,
+      "loss": 0.3627,
+      "step": 8029
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.558042526245117,
+      "learning_rate": 9.328985755963618e-06,
+      "loss": 0.5747,
+      "step": 8030
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.952147483825684,
+      "learning_rate": 9.326411532521023e-06,
+      "loss": 0.7353,
+      "step": 8031
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.418725967407227,
+      "learning_rate": 9.323837309078428e-06,
+      "loss": 0.4161,
+      "step": 8032
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.857487678527832,
+      "learning_rate": 9.321263085635833e-06,
+      "loss": 0.3946,
+      "step": 8033
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.552834510803223,
+      "learning_rate": 9.31868886219324e-06,
+      "loss": 0.6213,
+      "step": 8034
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.699774742126465,
+      "learning_rate": 9.316114638750644e-06,
+      "loss": 0.4156,
+      "step": 8035
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.114253997802734,
+      "learning_rate": 9.31354041530805e-06,
+      "loss": 0.4122,
+      "step": 8036
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.4076509475708,
+      "learning_rate": 9.310966191865454e-06,
+      "loss": 0.6045,
+      "step": 8037
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.927475929260254,
+      "learning_rate": 9.30839196842286e-06,
+      "loss": 0.5449,
+      "step": 8038
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.94198751449585,
+      "learning_rate": 9.305817744980264e-06,
+      "loss": 0.3991,
+      "step": 8039
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.521130561828613,
+      "learning_rate": 9.30324352153767e-06,
+      "loss": 0.7631,
+      "step": 8040
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.8478193283081055,
+      "learning_rate": 9.300669298095074e-06,
+      "loss": 0.2825,
+      "step": 8041
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.764883995056152,
+      "learning_rate": 9.29809507465248e-06,
+      "loss": 0.5785,
+      "step": 8042
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.743704795837402,
+      "learning_rate": 9.295520851209886e-06,
+      "loss": 0.5258,
+      "step": 8043
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.892135620117188,
+      "learning_rate": 9.292946627767291e-06,
+      "loss": 0.476,
+      "step": 8044
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.146523475646973,
+      "learning_rate": 9.290372404324696e-06,
+      "loss": 0.4902,
+      "step": 8045
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.193522930145264,
+      "learning_rate": 9.287798180882101e-06,
+      "loss": 0.2906,
+      "step": 8046
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.721735954284668,
+      "learning_rate": 9.285223957439506e-06,
+      "loss": 0.4723,
+      "step": 8047
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.398427963256836,
+      "learning_rate": 9.282649733996913e-06,
+      "loss": 0.4065,
+      "step": 8048
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.175658226013184,
+      "learning_rate": 9.280075510554316e-06,
+      "loss": 0.4116,
+      "step": 8049
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.530220031738281,
+      "learning_rate": 9.277501287111721e-06,
+      "loss": 0.2988,
+      "step": 8050
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.05105972290039,
+      "learning_rate": 9.274927063669126e-06,
+      "loss": 0.472,
+      "step": 8051
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.80966567993164,
+      "learning_rate": 9.272352840226531e-06,
+      "loss": 0.5229,
+      "step": 8052
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.220248222351074,
+      "learning_rate": 9.269778616783938e-06,
+      "loss": 0.3409,
+      "step": 8053
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.488533020019531,
+      "learning_rate": 9.267204393341343e-06,
+      "loss": 0.3125,
+      "step": 8054
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.574464797973633,
+      "learning_rate": 9.264630169898748e-06,
+      "loss": 0.4688,
+      "step": 8055
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.483405113220215,
+      "learning_rate": 9.262055946456153e-06,
+      "loss": 0.6974,
+      "step": 8056
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.591333389282227,
+      "learning_rate": 9.259481723013558e-06,
+      "loss": 0.5321,
+      "step": 8057
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.4368257522583,
+      "learning_rate": 9.256907499570964e-06,
+      "loss": 0.5061,
+      "step": 8058
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.434736251831055,
+      "learning_rate": 9.254333276128367e-06,
+      "loss": 0.5674,
+      "step": 8059
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.47308349609375,
+      "learning_rate": 9.251759052685772e-06,
+      "loss": 0.5751,
+      "step": 8060
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.3410062789917,
+      "learning_rate": 9.249184829243177e-06,
+      "loss": 0.4704,
+      "step": 8061
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.103771209716797,
+      "learning_rate": 9.246610605800584e-06,
+      "loss": 0.3633,
+      "step": 8062
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.619467735290527,
+      "learning_rate": 9.244036382357989e-06,
+      "loss": 0.6022,
+      "step": 8063
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.229324340820312,
+      "learning_rate": 9.241462158915394e-06,
+      "loss": 0.4712,
+      "step": 8064
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.660847663879395,
+      "learning_rate": 9.238887935472799e-06,
+      "loss": 0.5006,
+      "step": 8065
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.222514152526855,
+      "learning_rate": 9.236313712030204e-06,
+      "loss": 0.5117,
+      "step": 8066
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.187640190124512,
+      "learning_rate": 9.23373948858761e-06,
+      "loss": 0.5412,
+      "step": 8067
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.165215492248535,
+      "learning_rate": 9.231165265145016e-06,
+      "loss": 0.4262,
+      "step": 8068
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.318355560302734,
+      "learning_rate": 9.22859104170242e-06,
+      "loss": 0.4678,
+      "step": 8069
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.313220024108887,
+      "learning_rate": 9.226016818259824e-06,
+      "loss": 0.6658,
+      "step": 8070
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.866724014282227,
+      "learning_rate": 9.223442594817229e-06,
+      "loss": 0.3462,
+      "step": 8071
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.366093635559082,
+      "learning_rate": 9.220868371374636e-06,
+      "loss": 0.5936,
+      "step": 8072
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.871940612792969,
+      "learning_rate": 9.21829414793204e-06,
+      "loss": 0.3365,
+      "step": 8073
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.973756790161133,
+      "learning_rate": 9.215719924489446e-06,
+      "loss": 0.4507,
+      "step": 8074
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.828827857971191,
+      "learning_rate": 9.21314570104685e-06,
+      "loss": 0.5217,
+      "step": 8075
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.579808235168457,
+      "learning_rate": 9.210571477604257e-06,
+      "loss": 0.497,
+      "step": 8076
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.060633659362793,
+      "learning_rate": 9.207997254161662e-06,
+      "loss": 0.424,
+      "step": 8077
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.351252555847168,
+      "learning_rate": 9.205423030719067e-06,
+      "loss": 0.4474,
+      "step": 8078
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.33030891418457,
+      "learning_rate": 9.202848807276472e-06,
+      "loss": 0.3736,
+      "step": 8079
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.173501014709473,
+      "learning_rate": 9.200274583833876e-06,
+      "loss": 0.5181,
+      "step": 8080
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.958808898925781,
+      "learning_rate": 9.197700360391282e-06,
+      "loss": 0.5828,
+      "step": 8081
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.104315757751465,
+      "learning_rate": 9.195126136948687e-06,
+      "loss": 0.3645,
+      "step": 8082
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.270234107971191,
+      "learning_rate": 9.192551913506092e-06,
+      "loss": 0.6803,
+      "step": 8083
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.381769180297852,
+      "learning_rate": 9.189977690063497e-06,
+      "loss": 0.5663,
+      "step": 8084
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.067337036132812,
+      "learning_rate": 9.187403466620902e-06,
+      "loss": 0.3539,
+      "step": 8085
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.98100757598877,
+      "learning_rate": 9.184829243178309e-06,
+      "loss": 0.3303,
+      "step": 8086
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.848414421081543,
+      "learning_rate": 9.182255019735714e-06,
+      "loss": 0.4793,
+      "step": 8087
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.813301086425781,
+      "learning_rate": 9.179680796293119e-06,
+      "loss": 0.3734,
+      "step": 8088
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.797536849975586,
+      "learning_rate": 9.177106572850524e-06,
+      "loss": 0.3712,
+      "step": 8089
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.001542091369629,
+      "learning_rate": 9.174532349407929e-06,
+      "loss": 0.2863,
+      "step": 8090
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.185687065124512,
+      "learning_rate": 9.171958125965334e-06,
+      "loss": 0.6677,
+      "step": 8091
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.785146713256836,
+      "learning_rate": 9.169383902522739e-06,
+      "loss": 0.306,
+      "step": 8092
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.149352073669434,
+      "learning_rate": 9.166809679080144e-06,
+      "loss": 0.4567,
+      "step": 8093
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.00492000579834,
+      "learning_rate": 9.164235455637549e-06,
+      "loss": 0.5592,
+      "step": 8094
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.96116828918457,
+      "learning_rate": 9.161661232194956e-06,
+      "loss": 0.5866,
+      "step": 8095
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.884201049804688,
+      "learning_rate": 9.15908700875236e-06,
+      "loss": 0.6157,
+      "step": 8096
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.086968421936035,
+      "learning_rate": 9.156512785309765e-06,
+      "loss": 0.6028,
+      "step": 8097
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.086346626281738,
+      "learning_rate": 9.15393856186717e-06,
+      "loss": 0.3687,
+      "step": 8098
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.605083465576172,
+      "learning_rate": 9.151364338424575e-06,
+      "loss": 0.5211,
+      "step": 8099
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.9868316650390625,
+      "learning_rate": 9.148790114981982e-06,
+      "loss": 0.4179,
+      "step": 8100
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.09657096862793,
+      "learning_rate": 9.146215891539385e-06,
+      "loss": 0.2719,
+      "step": 8101
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.188369750976562,
+      "learning_rate": 9.14364166809679e-06,
+      "loss": 0.4122,
+      "step": 8102
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.78367805480957,
+      "learning_rate": 9.141067444654195e-06,
+      "loss": 0.6931,
+      "step": 8103
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.463834762573242,
+      "learning_rate": 9.1384932212116e-06,
+      "loss": 0.2721,
+      "step": 8104
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.522351264953613,
+      "learning_rate": 9.135918997769007e-06,
+      "loss": 0.5886,
+      "step": 8105
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.248406410217285,
+      "learning_rate": 9.133344774326412e-06,
+      "loss": 0.5777,
+      "step": 8106
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.383096218109131,
+      "learning_rate": 9.130770550883817e-06,
+      "loss": 0.2832,
+      "step": 8107
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.572043418884277,
+      "learning_rate": 9.128196327441222e-06,
+      "loss": 0.3077,
+      "step": 8108
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.84365177154541,
+      "learning_rate": 9.125622103998627e-06,
+      "loss": 0.3161,
+      "step": 8109
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.437800407409668,
+      "learning_rate": 9.123047880556034e-06,
+      "loss": 0.5003,
+      "step": 8110
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.400375366210938,
+      "learning_rate": 9.120473657113437e-06,
+      "loss": 0.5481,
+      "step": 8111
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.788758277893066,
+      "learning_rate": 9.117899433670842e-06,
+      "loss": 0.51,
+      "step": 8112
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.329127311706543,
+      "learning_rate": 9.115325210228247e-06,
+      "loss": 0.458,
+      "step": 8113
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.465681076049805,
+      "learning_rate": 9.112750986785654e-06,
+      "loss": 0.5648,
+      "step": 8114
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.650321006774902,
+      "learning_rate": 9.110176763343059e-06,
+      "loss": 0.3974,
+      "step": 8115
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.671797752380371,
+      "learning_rate": 9.107602539900464e-06,
+      "loss": 0.5949,
+      "step": 8116
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.014888763427734,
+      "learning_rate": 9.105028316457869e-06,
+      "loss": 0.3711,
+      "step": 8117
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.839879989624023,
+      "learning_rate": 9.102454093015274e-06,
+      "loss": 0.3674,
+      "step": 8118
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.793872833251953,
+      "learning_rate": 9.09987986957268e-06,
+      "loss": 0.5252,
+      "step": 8119
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.970890998840332,
+      "learning_rate": 9.097305646130085e-06,
+      "loss": 0.4629,
+      "step": 8120
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.441503524780273,
+      "learning_rate": 9.09473142268749e-06,
+      "loss": 0.3309,
+      "step": 8121
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.942561149597168,
+      "learning_rate": 9.092157199244894e-06,
+      "loss": 0.4877,
+      "step": 8122
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.946661949157715,
+      "learning_rate": 9.089582975802299e-06,
+      "loss": 0.4963,
+      "step": 8123
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.9072265625,
+      "learning_rate": 9.087008752359705e-06,
+      "loss": 0.6806,
+      "step": 8124
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.876772403717041,
+      "learning_rate": 9.08443452891711e-06,
+      "loss": 0.3393,
+      "step": 8125
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.20521354675293,
+      "learning_rate": 9.081860305474515e-06,
+      "loss": 0.4055,
+      "step": 8126
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.229073524475098,
+      "learning_rate": 9.07928608203192e-06,
+      "loss": 0.4284,
+      "step": 8127
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.250029563903809,
+      "learning_rate": 9.076711858589325e-06,
+      "loss": 0.3294,
+      "step": 8128
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.435440540313721,
+      "learning_rate": 9.074137635146732e-06,
+      "loss": 0.326,
+      "step": 8129
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.752583980560303,
+      "learning_rate": 9.071563411704137e-06,
+      "loss": 0.3474,
+      "step": 8130
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.20149040222168,
+      "learning_rate": 9.068989188261542e-06,
+      "loss": 0.3749,
+      "step": 8131
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.70844841003418,
+      "learning_rate": 9.066414964818945e-06,
+      "loss": 0.8339,
+      "step": 8132
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.071343421936035,
+      "learning_rate": 9.063840741376352e-06,
+      "loss": 0.414,
+      "step": 8133
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.165002822875977,
+      "learning_rate": 9.061266517933757e-06,
+      "loss": 0.5569,
+      "step": 8134
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.210293769836426,
+      "learning_rate": 9.058692294491162e-06,
+      "loss": 0.6359,
+      "step": 8135
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.353513717651367,
+      "learning_rate": 9.056118071048567e-06,
+      "loss": 0.5061,
+      "step": 8136
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.391908645629883,
+      "learning_rate": 9.053543847605972e-06,
+      "loss": 0.5219,
+      "step": 8137
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.386820793151855,
+      "learning_rate": 9.050969624163378e-06,
+      "loss": 0.3787,
+      "step": 8138
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.763087272644043,
+      "learning_rate": 9.048395400720783e-06,
+      "loss": 0.6168,
+      "step": 8139
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.695484161376953,
+      "learning_rate": 9.045821177278188e-06,
+      "loss": 0.3589,
+      "step": 8140
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.074987411499023,
+      "learning_rate": 9.043246953835593e-06,
+      "loss": 0.4396,
+      "step": 8141
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.367197036743164,
+      "learning_rate": 9.040672730392998e-06,
+      "loss": 0.5388,
+      "step": 8142
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.407917976379395,
+      "learning_rate": 9.038098506950403e-06,
+      "loss": 0.591,
+      "step": 8143
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.238104820251465,
+      "learning_rate": 9.035524283507808e-06,
+      "loss": 0.5808,
+      "step": 8144
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.704313278198242,
+      "learning_rate": 9.032950060065213e-06,
+      "loss": 0.4613,
+      "step": 8145
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.697770118713379,
+      "learning_rate": 9.030375836622618e-06,
+      "loss": 0.5465,
+      "step": 8146
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.017826557159424,
+      "learning_rate": 9.027801613180025e-06,
+      "loss": 0.3571,
+      "step": 8147
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.1393351554870605,
+      "learning_rate": 9.02522738973743e-06,
+      "loss": 0.2915,
+      "step": 8148
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.771018028259277,
+      "learning_rate": 9.022653166294835e-06,
+      "loss": 0.5083,
+      "step": 8149
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.499884605407715,
+      "learning_rate": 9.02007894285224e-06,
+      "loss": 0.5661,
+      "step": 8150
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.548622131347656,
+      "learning_rate": 9.017504719409645e-06,
+      "loss": 0.5096,
+      "step": 8151
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.084818840026855,
+      "learning_rate": 9.014930495967052e-06,
+      "loss": 0.6314,
+      "step": 8152
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.663867950439453,
+      "learning_rate": 9.012356272524455e-06,
+      "loss": 0.4214,
+      "step": 8153
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.084020614624023,
+      "learning_rate": 9.00978204908186e-06,
+      "loss": 0.6374,
+      "step": 8154
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.588614463806152,
+      "learning_rate": 9.007207825639265e-06,
+      "loss": 0.4847,
+      "step": 8155
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.392104148864746,
+      "learning_rate": 9.00463360219667e-06,
+      "loss": 0.3692,
+      "step": 8156
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.744665145874023,
+      "learning_rate": 9.002059378754077e-06,
+      "loss": 0.4344,
+      "step": 8157
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.3781585693359375,
+      "learning_rate": 8.999485155311482e-06,
+      "loss": 0.2844,
+      "step": 8158
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.054342269897461,
+      "learning_rate": 8.996910931868887e-06,
+      "loss": 0.4426,
+      "step": 8159
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.579928398132324,
+      "learning_rate": 8.994336708426292e-06,
+      "loss": 0.4809,
+      "step": 8160
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.54077434539795,
+      "learning_rate": 8.991762484983697e-06,
+      "loss": 0.2959,
+      "step": 8161
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.806890487670898,
+      "learning_rate": 8.989188261541103e-06,
+      "loss": 0.354,
+      "step": 8162
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.789708137512207,
+      "learning_rate": 8.986614038098507e-06,
+      "loss": 0.4683,
+      "step": 8163
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.888046741485596,
+      "learning_rate": 8.984039814655911e-06,
+      "loss": 0.2823,
+      "step": 8164
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.385698318481445,
+      "learning_rate": 8.981465591213316e-06,
+      "loss": 0.4918,
+      "step": 8165
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.205957412719727,
+      "learning_rate": 8.978891367770723e-06,
+      "loss": 0.4328,
+      "step": 8166
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.774950981140137,
+      "learning_rate": 8.976317144328128e-06,
+      "loss": 0.6236,
+      "step": 8167
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.305766105651855,
+      "learning_rate": 8.973742920885533e-06,
+      "loss": 0.5363,
+      "step": 8168
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.135571479797363,
+      "learning_rate": 8.971168697442938e-06,
+      "loss": 0.4644,
+      "step": 8169
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.392585754394531,
+      "learning_rate": 8.968594474000343e-06,
+      "loss": 0.3965,
+      "step": 8170
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.347583770751953,
+      "learning_rate": 8.96602025055775e-06,
+      "loss": 0.6629,
+      "step": 8171
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.156412124633789,
+      "learning_rate": 8.963446027115155e-06,
+      "loss": 0.42,
+      "step": 8172
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.219449043273926,
+      "learning_rate": 8.96087180367256e-06,
+      "loss": 0.5156,
+      "step": 8173
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.61083984375,
+      "learning_rate": 8.958297580229963e-06,
+      "loss": 0.6273,
+      "step": 8174
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.187403678894043,
+      "learning_rate": 8.955723356787368e-06,
+      "loss": 0.2932,
+      "step": 8175
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.252684593200684,
+      "learning_rate": 8.953149133344775e-06,
+      "loss": 0.5857,
+      "step": 8176
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.347857475280762,
+      "learning_rate": 8.95057490990218e-06,
+      "loss": 0.7255,
+      "step": 8177
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.671805381774902,
+      "learning_rate": 8.948000686459585e-06,
+      "loss": 0.4495,
+      "step": 8178
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.4444427490234375,
+      "learning_rate": 8.94542646301699e-06,
+      "loss": 0.3414,
+      "step": 8179
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.483716011047363,
+      "learning_rate": 8.942852239574395e-06,
+      "loss": 0.385,
+      "step": 8180
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.974165916442871,
+      "learning_rate": 8.940278016131801e-06,
+      "loss": 0.4548,
+      "step": 8181
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.356922149658203,
+      "learning_rate": 8.937703792689206e-06,
+      "loss": 0.573,
+      "step": 8182
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.437995910644531,
+      "learning_rate": 8.935129569246611e-06,
+      "loss": 0.427,
+      "step": 8183
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.105772018432617,
+      "learning_rate": 8.932555345804015e-06,
+      "loss": 0.5355,
+      "step": 8184
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.922123908996582,
+      "learning_rate": 8.929981122361421e-06,
+      "loss": 0.3058,
+      "step": 8185
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.451415061950684,
+      "learning_rate": 8.927406898918826e-06,
+      "loss": 0.3559,
+      "step": 8186
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.637255668640137,
+      "learning_rate": 8.924832675476231e-06,
+      "loss": 0.4065,
+      "step": 8187
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.531972885131836,
+      "learning_rate": 8.922258452033636e-06,
+      "loss": 0.2564,
+      "step": 8188
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.943924903869629,
+      "learning_rate": 8.919684228591041e-06,
+      "loss": 0.4977,
+      "step": 8189
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.87394905090332,
+      "learning_rate": 8.917110005148448e-06,
+      "loss": 0.5008,
+      "step": 8190
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.284592628479004,
+      "learning_rate": 8.914535781705853e-06,
+      "loss": 0.404,
+      "step": 8191
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.662132263183594,
+      "learning_rate": 8.911961558263258e-06,
+      "loss": 0.3691,
+      "step": 8192
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.03271484375,
+      "learning_rate": 8.909387334820663e-06,
+      "loss": 0.562,
+      "step": 8193
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.703021049499512,
+      "learning_rate": 8.906813111378068e-06,
+      "loss": 0.3693,
+      "step": 8194
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.925243377685547,
+      "learning_rate": 8.904238887935473e-06,
+      "loss": 0.689,
+      "step": 8195
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.75747013092041,
+      "learning_rate": 8.901664664492878e-06,
+      "loss": 0.3488,
+      "step": 8196
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.172895431518555,
+      "learning_rate": 8.899090441050283e-06,
+      "loss": 0.3431,
+      "step": 8197
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.2479248046875,
+      "learning_rate": 8.896516217607688e-06,
+      "loss": 0.3468,
+      "step": 8198
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.617008209228516,
+      "learning_rate": 8.893941994165095e-06,
+      "loss": 0.3558,
+      "step": 8199
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.736842155456543,
+      "learning_rate": 8.8913677707225e-06,
+      "loss": 0.5643,
+      "step": 8200
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.11459732055664,
+      "learning_rate": 8.888793547279904e-06,
+      "loss": 0.434,
+      "step": 8201
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.434243202209473,
+      "learning_rate": 8.88621932383731e-06,
+      "loss": 0.4352,
+      "step": 8202
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.888657569885254,
+      "learning_rate": 8.883645100394714e-06,
+      "loss": 0.537,
+      "step": 8203
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.245015144348145,
+      "learning_rate": 8.881070876952121e-06,
+      "loss": 0.5832,
+      "step": 8204
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.50074577331543,
+      "learning_rate": 8.878496653509524e-06,
+      "loss": 0.4098,
+      "step": 8205
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.001585960388184,
+      "learning_rate": 8.87592243006693e-06,
+      "loss": 0.5697,
+      "step": 8206
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.823759078979492,
+      "learning_rate": 8.873348206624334e-06,
+      "loss": 0.6038,
+      "step": 8207
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.641414642333984,
+      "learning_rate": 8.87077398318174e-06,
+      "loss": 0.5124,
+      "step": 8208
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.899895668029785,
+      "learning_rate": 8.868199759739146e-06,
+      "loss": 0.4119,
+      "step": 8209
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.082362174987793,
+      "learning_rate": 8.865625536296551e-06,
+      "loss": 0.5197,
+      "step": 8210
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.878350257873535,
+      "learning_rate": 8.863051312853956e-06,
+      "loss": 0.5615,
+      "step": 8211
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.261621475219727,
+      "learning_rate": 8.860477089411361e-06,
+      "loss": 0.4499,
+      "step": 8212
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.753847122192383,
+      "learning_rate": 8.857902865968766e-06,
+      "loss": 0.4265,
+      "step": 8213
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.380756378173828,
+      "learning_rate": 8.855328642526173e-06,
+      "loss": 0.5466,
+      "step": 8214
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.7287492752075195,
+      "learning_rate": 8.852754419083578e-06,
+      "loss": 0.4916,
+      "step": 8215
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.159360885620117,
+      "learning_rate": 8.850180195640981e-06,
+      "loss": 0.5015,
+      "step": 8216
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.102422714233398,
+      "learning_rate": 8.847605972198386e-06,
+      "loss": 0.5305,
+      "step": 8217
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.977442741394043,
+      "learning_rate": 8.845031748755793e-06,
+      "loss": 0.3016,
+      "step": 8218
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.829934120178223,
+      "learning_rate": 8.842457525313198e-06,
+      "loss": 0.3551,
+      "step": 8219
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.341826438903809,
+      "learning_rate": 8.839883301870603e-06,
+      "loss": 0.4647,
+      "step": 8220
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.113983154296875,
+      "learning_rate": 8.837309078428008e-06,
+      "loss": 0.544,
+      "step": 8221
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.224457740783691,
+      "learning_rate": 8.834734854985413e-06,
+      "loss": 0.4098,
+      "step": 8222
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.566207885742188,
+      "learning_rate": 8.83216063154282e-06,
+      "loss": 0.5322,
+      "step": 8223
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.017365455627441,
+      "learning_rate": 8.829586408100224e-06,
+      "loss": 0.759,
+      "step": 8224
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.604937553405762,
+      "learning_rate": 8.82701218465763e-06,
+      "loss": 0.3421,
+      "step": 8225
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.038934707641602,
+      "learning_rate": 8.824437961215033e-06,
+      "loss": 0.3996,
+      "step": 8226
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.238123893737793,
+      "learning_rate": 8.821863737772438e-06,
+      "loss": 0.5725,
+      "step": 8227
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.162654876708984,
+      "learning_rate": 8.819289514329844e-06,
+      "loss": 0.5796,
+      "step": 8228
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.382336616516113,
+      "learning_rate": 8.81671529088725e-06,
+      "loss": 0.3975,
+      "step": 8229
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.843670845031738,
+      "learning_rate": 8.814141067444654e-06,
+      "loss": 0.5249,
+      "step": 8230
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.033681869506836,
+      "learning_rate": 8.81156684400206e-06,
+      "loss": 0.6187,
+      "step": 8231
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.541534423828125,
+      "learning_rate": 8.808992620559464e-06,
+      "loss": 0.4985,
+      "step": 8232
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.84037971496582,
+      "learning_rate": 8.806418397116871e-06,
+      "loss": 0.4992,
+      "step": 8233
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.408158302307129,
+      "learning_rate": 8.803844173674276e-06,
+      "loss": 0.3927,
+      "step": 8234
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.84311294555664,
+      "learning_rate": 8.80126995023168e-06,
+      "loss": 0.5302,
+      "step": 8235
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.212716102600098,
+      "learning_rate": 8.798695726789084e-06,
+      "loss": 0.3294,
+      "step": 8236
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.820342063903809,
+      "learning_rate": 8.79612150334649e-06,
+      "loss": 0.4808,
+      "step": 8237
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.624064445495605,
+      "learning_rate": 8.793547279903896e-06,
+      "loss": 0.6788,
+      "step": 8238
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.806413650512695,
+      "learning_rate": 8.7909730564613e-06,
+      "loss": 0.6901,
+      "step": 8239
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.613114356994629,
+      "learning_rate": 8.788398833018706e-06,
+      "loss": 0.4274,
+      "step": 8240
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.828776359558105,
+      "learning_rate": 8.78582460957611e-06,
+      "loss": 0.3861,
+      "step": 8241
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.33912181854248,
+      "learning_rate": 8.783250386133517e-06,
+      "loss": 0.4479,
+      "step": 8242
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.551612854003906,
+      "learning_rate": 8.780676162690922e-06,
+      "loss": 0.424,
+      "step": 8243
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.652555465698242,
+      "learning_rate": 8.778101939248327e-06,
+      "loss": 0.6361,
+      "step": 8244
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.954252243041992,
+      "learning_rate": 8.775527715805732e-06,
+      "loss": 0.4973,
+      "step": 8245
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.67759895324707,
+      "learning_rate": 8.772953492363137e-06,
+      "loss": 0.467,
+      "step": 8246
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.356545925140381,
+      "learning_rate": 8.770379268920542e-06,
+      "loss": 0.251,
+      "step": 8247
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.583957195281982,
+      "learning_rate": 8.767805045477947e-06,
+      "loss": 0.334,
+      "step": 8248
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.201501846313477,
+      "learning_rate": 8.765230822035352e-06,
+      "loss": 0.4561,
+      "step": 8249
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.017091751098633,
+      "learning_rate": 8.762656598592757e-06,
+      "loss": 0.336,
+      "step": 8250
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.172395706176758,
+      "learning_rate": 8.760082375150164e-06,
+      "loss": 0.3834,
+      "step": 8251
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.870978355407715,
+      "learning_rate": 8.757508151707569e-06,
+      "loss": 0.5051,
+      "step": 8252
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.49899959564209,
+      "learning_rate": 8.754933928264974e-06,
+      "loss": 0.5327,
+      "step": 8253
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.872721672058105,
+      "learning_rate": 8.752359704822379e-06,
+      "loss": 0.5365,
+      "step": 8254
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.230222702026367,
+      "learning_rate": 8.749785481379784e-06,
+      "loss": 0.4969,
+      "step": 8255
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.480122566223145,
+      "learning_rate": 8.74721125793719e-06,
+      "loss": 0.4104,
+      "step": 8256
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.623018264770508,
+      "learning_rate": 8.744637034494594e-06,
+      "loss": 0.5869,
+      "step": 8257
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.163650512695312,
+      "learning_rate": 8.742062811051999e-06,
+      "loss": 0.6378,
+      "step": 8258
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.162103652954102,
+      "learning_rate": 8.739488587609404e-06,
+      "loss": 0.4801,
+      "step": 8259
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.066727638244629,
+      "learning_rate": 8.736914364166809e-06,
+      "loss": 0.4924,
+      "step": 8260
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.020031929016113,
+      "learning_rate": 8.734340140724216e-06,
+      "loss": 0.5363,
+      "step": 8261
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.88711929321289,
+      "learning_rate": 8.73176591728162e-06,
+      "loss": 0.3471,
+      "step": 8262
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.993285179138184,
+      "learning_rate": 8.729191693839026e-06,
+      "loss": 0.3973,
+      "step": 8263
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.434762954711914,
+      "learning_rate": 8.72661747039643e-06,
+      "loss": 0.3801,
+      "step": 8264
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.919322967529297,
+      "learning_rate": 8.724043246953836e-06,
+      "loss": 0.5145,
+      "step": 8265
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.72677230834961,
+      "learning_rate": 8.721469023511242e-06,
+      "loss": 0.3841,
+      "step": 8266
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.11995506286621,
+      "learning_rate": 8.718894800068647e-06,
+      "loss": 0.5194,
+      "step": 8267
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.176628112792969,
+      "learning_rate": 8.71632057662605e-06,
+      "loss": 0.5011,
+      "step": 8268
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.962334632873535,
+      "learning_rate": 8.713746353183455e-06,
+      "loss": 0.4482,
+      "step": 8269
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.004849433898926,
+      "learning_rate": 8.711172129740862e-06,
+      "loss": 0.579,
+      "step": 8270
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.271482467651367,
+      "learning_rate": 8.708597906298267e-06,
+      "loss": 0.5262,
+      "step": 8271
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.614509582519531,
+      "learning_rate": 8.706023682855672e-06,
+      "loss": 0.3924,
+      "step": 8272
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.539762496948242,
+      "learning_rate": 8.703449459413077e-06,
+      "loss": 0.6156,
+      "step": 8273
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.934184551239014,
+      "learning_rate": 8.700875235970482e-06,
+      "loss": 0.3409,
+      "step": 8274
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.466604232788086,
+      "learning_rate": 8.698301012527889e-06,
+      "loss": 0.4886,
+      "step": 8275
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.193823337554932,
+      "learning_rate": 8.695726789085294e-06,
+      "loss": 0.3566,
+      "step": 8276
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.440775394439697,
+      "learning_rate": 8.693152565642699e-06,
+      "loss": 0.2732,
+      "step": 8277
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.61000394821167,
+      "learning_rate": 8.690578342200102e-06,
+      "loss": 0.491,
+      "step": 8278
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.64564037322998,
+      "learning_rate": 8.688004118757507e-06,
+      "loss": 0.7108,
+      "step": 8279
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.754938125610352,
+      "learning_rate": 8.685429895314914e-06,
+      "loss": 0.341,
+      "step": 8280
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.427057266235352,
+      "learning_rate": 8.682855671872319e-06,
+      "loss": 0.3672,
+      "step": 8281
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.298309326171875,
+      "learning_rate": 8.680281448429724e-06,
+      "loss": 0.3598,
+      "step": 8282
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.559104919433594,
+      "learning_rate": 8.677707224987129e-06,
+      "loss": 0.3229,
+      "step": 8283
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.896451950073242,
+      "learning_rate": 8.675133001544534e-06,
+      "loss": 0.5372,
+      "step": 8284
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.25345516204834,
+      "learning_rate": 8.67255877810194e-06,
+      "loss": 0.6119,
+      "step": 8285
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.65485954284668,
+      "learning_rate": 8.669984554659345e-06,
+      "loss": 0.4507,
+      "step": 8286
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.328947067260742,
+      "learning_rate": 8.66741033121675e-06,
+      "loss": 0.5,
+      "step": 8287
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.528351783752441,
+      "learning_rate": 8.664836107774154e-06,
+      "loss": 0.3386,
+      "step": 8288
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.252469062805176,
+      "learning_rate": 8.66226188433156e-06,
+      "loss": 0.4749,
+      "step": 8289
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.864716529846191,
+      "learning_rate": 8.659687660888965e-06,
+      "loss": 0.3129,
+      "step": 8290
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.331801414489746,
+      "learning_rate": 8.65711343744637e-06,
+      "loss": 0.4275,
+      "step": 8291
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.158288955688477,
+      "learning_rate": 8.654539214003775e-06,
+      "loss": 0.4124,
+      "step": 8292
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.22913932800293,
+      "learning_rate": 8.65196499056118e-06,
+      "loss": 0.6633,
+      "step": 8293
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.249902725219727,
+      "learning_rate": 8.649390767118587e-06,
+      "loss": 0.6345,
+      "step": 8294
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.750879287719727,
+      "learning_rate": 8.646816543675992e-06,
+      "loss": 0.4824,
+      "step": 8295
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.58333683013916,
+      "learning_rate": 8.644242320233397e-06,
+      "loss": 0.4946,
+      "step": 8296
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.40911865234375,
+      "learning_rate": 8.641668096790802e-06,
+      "loss": 0.4101,
+      "step": 8297
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.808370590209961,
+      "learning_rate": 8.639093873348207e-06,
+      "loss": 0.3125,
+      "step": 8298
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.678741455078125,
+      "learning_rate": 8.636519649905612e-06,
+      "loss": 0.4791,
+      "step": 8299
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.651301860809326,
+      "learning_rate": 8.633945426463017e-06,
+      "loss": 0.3633,
+      "step": 8300
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.924830436706543,
+      "learning_rate": 8.631371203020422e-06,
+      "loss": 0.2811,
+      "step": 8301
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.635172843933105,
+      "learning_rate": 8.628796979577827e-06,
+      "loss": 0.484,
+      "step": 8302
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.726848602294922,
+      "learning_rate": 8.626222756135232e-06,
+      "loss": 0.385,
+      "step": 8303
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.830549240112305,
+      "learning_rate": 8.623648532692639e-06,
+      "loss": 0.424,
+      "step": 8304
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.686710357666016,
+      "learning_rate": 8.621074309250044e-06,
+      "loss": 0.3862,
+      "step": 8305
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.640059471130371,
+      "learning_rate": 8.618500085807448e-06,
+      "loss": 0.5257,
+      "step": 8306
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.689738273620605,
+      "learning_rate": 8.615925862364853e-06,
+      "loss": 0.6524,
+      "step": 8307
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.673746109008789,
+      "learning_rate": 8.61335163892226e-06,
+      "loss": 0.3279,
+      "step": 8308
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.808671951293945,
+      "learning_rate": 8.610777415479663e-06,
+      "loss": 0.682,
+      "step": 8309
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.548581123352051,
+      "learning_rate": 8.608203192037068e-06,
+      "loss": 0.2683,
+      "step": 8310
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.380051612854004,
+      "learning_rate": 8.605628968594473e-06,
+      "loss": 0.366,
+      "step": 8311
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.494132041931152,
+      "learning_rate": 8.603054745151878e-06,
+      "loss": 0.4889,
+      "step": 8312
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.73214864730835,
+      "learning_rate": 8.600480521709285e-06,
+      "loss": 0.3131,
+      "step": 8313
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.153104782104492,
+      "learning_rate": 8.59790629826669e-06,
+      "loss": 0.5418,
+      "step": 8314
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.794357299804688,
+      "learning_rate": 8.595332074824095e-06,
+      "loss": 0.3256,
+      "step": 8315
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.454607963562012,
+      "learning_rate": 8.5927578513815e-06,
+      "loss": 0.356,
+      "step": 8316
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.618805885314941,
+      "learning_rate": 8.590183627938905e-06,
+      "loss": 0.3531,
+      "step": 8317
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.349067687988281,
+      "learning_rate": 8.587609404496312e-06,
+      "loss": 0.633,
+      "step": 8318
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.352596282958984,
+      "learning_rate": 8.585035181053717e-06,
+      "loss": 0.5056,
+      "step": 8319
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.118658065795898,
+      "learning_rate": 8.58246095761112e-06,
+      "loss": 0.7799,
+      "step": 8320
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.962812423706055,
+      "learning_rate": 8.579886734168525e-06,
+      "loss": 0.5314,
+      "step": 8321
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.80367374420166,
+      "learning_rate": 8.577312510725932e-06,
+      "loss": 0.4823,
+      "step": 8322
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.806482315063477,
+      "learning_rate": 8.574738287283337e-06,
+      "loss": 0.3314,
+      "step": 8323
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.926474571228027,
+      "learning_rate": 8.572164063840742e-06,
+      "loss": 0.334,
+      "step": 8324
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.295524597167969,
+      "learning_rate": 8.569589840398147e-06,
+      "loss": 0.3899,
+      "step": 8325
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.297874450683594,
+      "learning_rate": 8.567015616955552e-06,
+      "loss": 0.3585,
+      "step": 8326
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.662091255187988,
+      "learning_rate": 8.564441393512958e-06,
+      "loss": 0.3612,
+      "step": 8327
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.200091361999512,
+      "learning_rate": 8.561867170070363e-06,
+      "loss": 0.4963,
+      "step": 8328
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.467052459716797,
+      "learning_rate": 8.559292946627768e-06,
+      "loss": 0.4722,
+      "step": 8329
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.857604026794434,
+      "learning_rate": 8.556718723185172e-06,
+      "loss": 0.4791,
+      "step": 8330
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.185227394104004,
+      "learning_rate": 8.554144499742577e-06,
+      "loss": 0.4457,
+      "step": 8331
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.965193748474121,
+      "learning_rate": 8.551570276299983e-06,
+      "loss": 0.4176,
+      "step": 8332
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.929383277893066,
+      "learning_rate": 8.548996052857388e-06,
+      "loss": 0.5286,
+      "step": 8333
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.188926696777344,
+      "learning_rate": 8.546421829414793e-06,
+      "loss": 0.5374,
+      "step": 8334
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.042409896850586,
+      "learning_rate": 8.543847605972198e-06,
+      "loss": 0.471,
+      "step": 8335
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.140226364135742,
+      "learning_rate": 8.541273382529603e-06,
+      "loss": 0.2439,
+      "step": 8336
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.757363319396973,
+      "learning_rate": 8.53869915908701e-06,
+      "loss": 0.2812,
+      "step": 8337
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.087406158447266,
+      "learning_rate": 8.536124935644415e-06,
+      "loss": 0.4378,
+      "step": 8338
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.848742485046387,
+      "learning_rate": 8.53355071220182e-06,
+      "loss": 0.52,
+      "step": 8339
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.599913597106934,
+      "learning_rate": 8.530976488759223e-06,
+      "loss": 0.6345,
+      "step": 8340
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.633811950683594,
+      "learning_rate": 8.52840226531663e-06,
+      "loss": 0.4267,
+      "step": 8341
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.455072402954102,
+      "learning_rate": 8.525828041874035e-06,
+      "loss": 0.4001,
+      "step": 8342
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.543682098388672,
+      "learning_rate": 8.52325381843144e-06,
+      "loss": 0.2689,
+      "step": 8343
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.692220687866211,
+      "learning_rate": 8.520679594988845e-06,
+      "loss": 0.488,
+      "step": 8344
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.151768684387207,
+      "learning_rate": 8.51810537154625e-06,
+      "loss": 0.3154,
+      "step": 8345
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.533817291259766,
+      "learning_rate": 8.515531148103656e-06,
+      "loss": 0.6282,
+      "step": 8346
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.896312713623047,
+      "learning_rate": 8.512956924661061e-06,
+      "loss": 0.5397,
+      "step": 8347
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.509431838989258,
+      "learning_rate": 8.510382701218466e-06,
+      "loss": 0.4958,
+      "step": 8348
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.831765174865723,
+      "learning_rate": 8.507808477775871e-06,
+      "loss": 0.2785,
+      "step": 8349
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.94285774230957,
+      "learning_rate": 8.505234254333276e-06,
+      "loss": 0.6302,
+      "step": 8350
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.702482223510742,
+      "learning_rate": 8.502660030890681e-06,
+      "loss": 0.6218,
+      "step": 8351
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.443445205688477,
+      "learning_rate": 8.500085807448086e-06,
+      "loss": 0.3523,
+      "step": 8352
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.827876091003418,
+      "learning_rate": 8.497511584005491e-06,
+      "loss": 0.3185,
+      "step": 8353
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.086119651794434,
+      "learning_rate": 8.494937360562896e-06,
+      "loss": 0.397,
+      "step": 8354
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.287346839904785,
+      "learning_rate": 8.492363137120301e-06,
+      "loss": 0.3258,
+      "step": 8355
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.768567085266113,
+      "learning_rate": 8.489788913677708e-06,
+      "loss": 0.7088,
+      "step": 8356
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.253920555114746,
+      "learning_rate": 8.487214690235113e-06,
+      "loss": 0.4393,
+      "step": 8357
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.461947441101074,
+      "learning_rate": 8.484640466792518e-06,
+      "loss": 0.4945,
+      "step": 8358
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.910832405090332,
+      "learning_rate": 8.482066243349923e-06,
+      "loss": 0.5817,
+      "step": 8359
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.067282676696777,
+      "learning_rate": 8.47949201990733e-06,
+      "loss": 0.5148,
+      "step": 8360
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.792328834533691,
+      "learning_rate": 8.476917796464733e-06,
+      "loss": 0.5585,
+      "step": 8361
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.34477424621582,
+      "learning_rate": 8.474343573022138e-06,
+      "loss": 0.4886,
+      "step": 8362
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.764257431030273,
+      "learning_rate": 8.471769349579543e-06,
+      "loss": 0.4799,
+      "step": 8363
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.07537841796875,
+      "learning_rate": 8.469195126136948e-06,
+      "loss": 0.5224,
+      "step": 8364
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.835885047912598,
+      "learning_rate": 8.466620902694355e-06,
+      "loss": 0.5015,
+      "step": 8365
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.417989730834961,
+      "learning_rate": 8.46404667925176e-06,
+      "loss": 0.4305,
+      "step": 8366
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.092061042785645,
+      "learning_rate": 8.461472455809165e-06,
+      "loss": 0.481,
+      "step": 8367
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.665326118469238,
+      "learning_rate": 8.45889823236657e-06,
+      "loss": 0.8116,
+      "step": 8368
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.353106498718262,
+      "learning_rate": 8.456324008923975e-06,
+      "loss": 0.3589,
+      "step": 8369
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.502452850341797,
+      "learning_rate": 8.453749785481381e-06,
+      "loss": 0.4884,
+      "step": 8370
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.927734375,
+      "learning_rate": 8.451175562038786e-06,
+      "loss": 0.3571,
+      "step": 8371
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.052430152893066,
+      "learning_rate": 8.44860133859619e-06,
+      "loss": 0.5337,
+      "step": 8372
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.651488304138184,
+      "learning_rate": 8.446027115153594e-06,
+      "loss": 0.5971,
+      "step": 8373
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.430017471313477,
+      "learning_rate": 8.443452891711001e-06,
+      "loss": 0.5956,
+      "step": 8374
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.363251686096191,
+      "learning_rate": 8.440878668268406e-06,
+      "loss": 0.4146,
+      "step": 8375
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.345171928405762,
+      "learning_rate": 8.438304444825811e-06,
+      "loss": 0.4958,
+      "step": 8376
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.103143692016602,
+      "learning_rate": 8.435730221383216e-06,
+      "loss": 0.4342,
+      "step": 8377
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.357614517211914,
+      "learning_rate": 8.433155997940621e-06,
+      "loss": 0.4971,
+      "step": 8378
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.737123489379883,
+      "learning_rate": 8.430581774498028e-06,
+      "loss": 0.4342,
+      "step": 8379
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.8394904136657715,
+      "learning_rate": 8.428007551055433e-06,
+      "loss": 0.3054,
+      "step": 8380
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.861248016357422,
+      "learning_rate": 8.425433327612838e-06,
+      "loss": 0.4876,
+      "step": 8381
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.34762191772461,
+      "learning_rate": 8.422859104170241e-06,
+      "loss": 0.4909,
+      "step": 8382
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.439802169799805,
+      "learning_rate": 8.420284880727646e-06,
+      "loss": 0.6265,
+      "step": 8383
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.666215896606445,
+      "learning_rate": 8.417710657285053e-06,
+      "loss": 0.2877,
+      "step": 8384
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.26455307006836,
+      "learning_rate": 8.415136433842458e-06,
+      "loss": 0.59,
+      "step": 8385
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.452399253845215,
+      "learning_rate": 8.412562210399863e-06,
+      "loss": 0.2802,
+      "step": 8386
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.878459930419922,
+      "learning_rate": 8.409987986957268e-06,
+      "loss": 0.5966,
+      "step": 8387
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.647378921508789,
+      "learning_rate": 8.407413763514673e-06,
+      "loss": 0.2942,
+      "step": 8388
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.983903884887695,
+      "learning_rate": 8.40483954007208e-06,
+      "loss": 0.4485,
+      "step": 8389
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.199638366699219,
+      "learning_rate": 8.402265316629484e-06,
+      "loss": 0.5004,
+      "step": 8390
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.529352188110352,
+      "learning_rate": 8.39969109318689e-06,
+      "loss": 0.398,
+      "step": 8391
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.492273330688477,
+      "learning_rate": 8.397116869744293e-06,
+      "loss": 0.4537,
+      "step": 8392
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.803034782409668,
+      "learning_rate": 8.3945426463017e-06,
+      "loss": 0.6738,
+      "step": 8393
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.096636772155762,
+      "learning_rate": 8.391968422859104e-06,
+      "loss": 0.5831,
+      "step": 8394
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.942265510559082,
+      "learning_rate": 8.38939419941651e-06,
+      "loss": 0.3807,
+      "step": 8395
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.381046295166016,
+      "learning_rate": 8.386819975973914e-06,
+      "loss": 0.3847,
+      "step": 8396
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.610161781311035,
+      "learning_rate": 8.38424575253132e-06,
+      "loss": 0.4258,
+      "step": 8397
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.381559371948242,
+      "learning_rate": 8.381671529088726e-06,
+      "loss": 0.6171,
+      "step": 8398
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.574058532714844,
+      "learning_rate": 8.379097305646131e-06,
+      "loss": 0.342,
+      "step": 8399
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.315040588378906,
+      "learning_rate": 8.376523082203536e-06,
+      "loss": 0.5536,
+      "step": 8400
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.571113586425781,
+      "learning_rate": 8.373948858760941e-06,
+      "loss": 0.3625,
+      "step": 8401
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.108417510986328,
+      "learning_rate": 8.371374635318346e-06,
+      "loss": 0.5363,
+      "step": 8402
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.237990379333496,
+      "learning_rate": 8.368800411875751e-06,
+      "loss": 0.5317,
+      "step": 8403
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.077326774597168,
+      "learning_rate": 8.366226188433156e-06,
+      "loss": 0.2537,
+      "step": 8404
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.669939041137695,
+      "learning_rate": 8.363651964990561e-06,
+      "loss": 0.3767,
+      "step": 8405
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.217723846435547,
+      "learning_rate": 8.361077741547966e-06,
+      "loss": 0.4565,
+      "step": 8406
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.763214111328125,
+      "learning_rate": 8.35850351810537e-06,
+      "loss": 0.3536,
+      "step": 8407
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.262199401855469,
+      "learning_rate": 8.355929294662778e-06,
+      "loss": 0.5818,
+      "step": 8408
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.669377326965332,
+      "learning_rate": 8.353355071220183e-06,
+      "loss": 0.4387,
+      "step": 8409
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.528432846069336,
+      "learning_rate": 8.350780847777588e-06,
+      "loss": 0.5647,
+      "step": 8410
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.921568870544434,
+      "learning_rate": 8.348206624334992e-06,
+      "loss": 0.4815,
+      "step": 8411
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.749783515930176,
+      "learning_rate": 8.3456324008924e-06,
+      "loss": 0.5651,
+      "step": 8412
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.831440448760986,
+      "learning_rate": 8.343058177449802e-06,
+      "loss": 0.3119,
+      "step": 8413
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.510527610778809,
+      "learning_rate": 8.340483954007207e-06,
+      "loss": 0.5791,
+      "step": 8414
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.164763450622559,
+      "learning_rate": 8.337909730564612e-06,
+      "loss": 0.4462,
+      "step": 8415
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.396998405456543,
+      "learning_rate": 8.335335507122017e-06,
+      "loss": 0.4084,
+      "step": 8416
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.974717140197754,
+      "learning_rate": 8.332761283679424e-06,
+      "loss": 0.2652,
+      "step": 8417
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.95771312713623,
+      "learning_rate": 8.330187060236829e-06,
+      "loss": 0.5407,
+      "step": 8418
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.97205638885498,
+      "learning_rate": 8.327612836794234e-06,
+      "loss": 0.4735,
+      "step": 8419
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.182731628417969,
+      "learning_rate": 8.325038613351639e-06,
+      "loss": 0.4982,
+      "step": 8420
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.552624225616455,
+      "learning_rate": 8.322464389909044e-06,
+      "loss": 0.3196,
+      "step": 8421
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.71505355834961,
+      "learning_rate": 8.31989016646645e-06,
+      "loss": 0.5178,
+      "step": 8422
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.814096450805664,
+      "learning_rate": 8.317315943023856e-06,
+      "loss": 0.6398,
+      "step": 8423
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.24152946472168,
+      "learning_rate": 8.314741719581259e-06,
+      "loss": 0.5214,
+      "step": 8424
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.424005508422852,
+      "learning_rate": 8.312167496138664e-06,
+      "loss": 0.405,
+      "step": 8425
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.650248527526855,
+      "learning_rate": 8.30959327269607e-06,
+      "loss": 0.4663,
+      "step": 8426
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.279358386993408,
+      "learning_rate": 8.307019049253476e-06,
+      "loss": 0.3011,
+      "step": 8427
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.594501495361328,
+      "learning_rate": 8.30444482581088e-06,
+      "loss": 0.3335,
+      "step": 8428
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.36333179473877,
+      "learning_rate": 8.301870602368286e-06,
+      "loss": 0.4246,
+      "step": 8429
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.192784309387207,
+      "learning_rate": 8.29929637892569e-06,
+      "loss": 0.4719,
+      "step": 8430
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.950685501098633,
+      "learning_rate": 8.296722155483097e-06,
+      "loss": 0.3771,
+      "step": 8431
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.547282218933105,
+      "learning_rate": 8.294147932040502e-06,
+      "loss": 0.6865,
+      "step": 8432
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.424164772033691,
+      "learning_rate": 8.291573708597907e-06,
+      "loss": 0.5652,
+      "step": 8433
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.865653038024902,
+      "learning_rate": 8.28899948515531e-06,
+      "loss": 0.4456,
+      "step": 8434
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.219254493713379,
+      "learning_rate": 8.286425261712716e-06,
+      "loss": 0.552,
+      "step": 8435
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.369195938110352,
+      "learning_rate": 8.283851038270122e-06,
+      "loss": 0.4501,
+      "step": 8436
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.32481575012207,
+      "learning_rate": 8.281276814827527e-06,
+      "loss": 0.4291,
+      "step": 8437
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.981685638427734,
+      "learning_rate": 8.278702591384932e-06,
+      "loss": 0.4604,
+      "step": 8438
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.983774185180664,
+      "learning_rate": 8.276128367942337e-06,
+      "loss": 0.4458,
+      "step": 8439
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.131538391113281,
+      "learning_rate": 8.273554144499742e-06,
+      "loss": 0.3347,
+      "step": 8440
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.760303497314453,
+      "learning_rate": 8.270979921057149e-06,
+      "loss": 0.4634,
+      "step": 8441
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.544533729553223,
+      "learning_rate": 8.268405697614554e-06,
+      "loss": 0.5307,
+      "step": 8442
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.032904624938965,
+      "learning_rate": 8.265831474171959e-06,
+      "loss": 0.3305,
+      "step": 8443
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.01931381225586,
+      "learning_rate": 8.263257250729364e-06,
+      "loss": 0.3629,
+      "step": 8444
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.627394676208496,
+      "learning_rate": 8.260683027286769e-06,
+      "loss": 0.4172,
+      "step": 8445
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.199492454528809,
+      "learning_rate": 8.258108803844174e-06,
+      "loss": 0.3404,
+      "step": 8446
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.086860656738281,
+      "learning_rate": 8.255534580401579e-06,
+      "loss": 0.3776,
+      "step": 8447
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.144417762756348,
+      "learning_rate": 8.252960356958984e-06,
+      "loss": 0.4841,
+      "step": 8448
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.8674898147583,
+      "learning_rate": 8.250386133516389e-06,
+      "loss": 0.4192,
+      "step": 8449
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.37148666381836,
+      "learning_rate": 8.247811910073795e-06,
+      "loss": 0.4222,
+      "step": 8450
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.26028060913086,
+      "learning_rate": 8.2452376866312e-06,
+      "loss": 0.5195,
+      "step": 8451
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.05476188659668,
+      "learning_rate": 8.242663463188605e-06,
+      "loss": 0.4429,
+      "step": 8452
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.395622253417969,
+      "learning_rate": 8.24008923974601e-06,
+      "loss": 0.5244,
+      "step": 8453
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.304303169250488,
+      "learning_rate": 8.237515016303415e-06,
+      "loss": 0.4451,
+      "step": 8454
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.714488983154297,
+      "learning_rate": 8.23494079286082e-06,
+      "loss": 0.6974,
+      "step": 8455
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.059340476989746,
+      "learning_rate": 8.232366569418225e-06,
+      "loss": 0.728,
+      "step": 8456
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.829148292541504,
+      "learning_rate": 8.22979234597563e-06,
+      "loss": 0.4292,
+      "step": 8457
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.803596496582031,
+      "learning_rate": 8.227218122533035e-06,
+      "loss": 0.45,
+      "step": 8458
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.919529914855957,
+      "learning_rate": 8.22464389909044e-06,
+      "loss": 0.4653,
+      "step": 8459
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.814225196838379,
+      "learning_rate": 8.222069675647847e-06,
+      "loss": 0.3956,
+      "step": 8460
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.573873519897461,
+      "learning_rate": 8.219495452205252e-06,
+      "loss": 0.4426,
+      "step": 8461
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.42048168182373,
+      "learning_rate": 8.216921228762657e-06,
+      "loss": 0.518,
+      "step": 8462
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.783432960510254,
+      "learning_rate": 8.214347005320062e-06,
+      "loss": 0.5066,
+      "step": 8463
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.45118236541748,
+      "learning_rate": 8.211772781877469e-06,
+      "loss": 0.5789,
+      "step": 8464
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.957793235778809,
+      "learning_rate": 8.209198558434872e-06,
+      "loss": 0.4351,
+      "step": 8465
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.776083946228027,
+      "learning_rate": 8.206624334992277e-06,
+      "loss": 0.4333,
+      "step": 8466
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.99911880493164,
+      "learning_rate": 8.204050111549682e-06,
+      "loss": 0.4055,
+      "step": 8467
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.94580364227295,
+      "learning_rate": 8.201475888107087e-06,
+      "loss": 0.5057,
+      "step": 8468
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.577740669250488,
+      "learning_rate": 8.198901664664494e-06,
+      "loss": 0.4571,
+      "step": 8469
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.692513942718506,
+      "learning_rate": 8.196327441221899e-06,
+      "loss": 0.2926,
+      "step": 8470
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.46033763885498,
+      "learning_rate": 8.193753217779304e-06,
+      "loss": 0.6169,
+      "step": 8471
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.683175086975098,
+      "learning_rate": 8.191178994336709e-06,
+      "loss": 0.3517,
+      "step": 8472
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.816125869750977,
+      "learning_rate": 8.188604770894114e-06,
+      "loss": 0.466,
+      "step": 8473
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.747037887573242,
+      "learning_rate": 8.18603054745152e-06,
+      "loss": 0.4297,
+      "step": 8474
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.123262405395508,
+      "learning_rate": 8.183456324008925e-06,
+      "loss": 0.3166,
+      "step": 8475
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.995744705200195,
+      "learning_rate": 8.180882100566329e-06,
+      "loss": 0.5092,
+      "step": 8476
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.646529197692871,
+      "learning_rate": 8.178307877123734e-06,
+      "loss": 0.4361,
+      "step": 8477
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.212658882141113,
+      "learning_rate": 8.175733653681138e-06,
+      "loss": 0.352,
+      "step": 8478
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.237067222595215,
+      "learning_rate": 8.173159430238545e-06,
+      "loss": 0.4433,
+      "step": 8479
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.00698184967041,
+      "learning_rate": 8.17058520679595e-06,
+      "loss": 0.4,
+      "step": 8480
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.731039047241211,
+      "learning_rate": 8.168010983353355e-06,
+      "loss": 0.3632,
+      "step": 8481
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.285453796386719,
+      "learning_rate": 8.16543675991076e-06,
+      "loss": 0.2463,
+      "step": 8482
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.291986465454102,
+      "learning_rate": 8.162862536468167e-06,
+      "loss": 0.4433,
+      "step": 8483
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.476229667663574,
+      "learning_rate": 8.160288313025572e-06,
+      "loss": 0.4293,
+      "step": 8484
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.852197647094727,
+      "learning_rate": 8.157714089582977e-06,
+      "loss": 0.4534,
+      "step": 8485
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.264104843139648,
+      "learning_rate": 8.15513986614038e-06,
+      "loss": 0.5055,
+      "step": 8486
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.131600379943848,
+      "learning_rate": 8.152565642697785e-06,
+      "loss": 0.5458,
+      "step": 8487
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.988832473754883,
+      "learning_rate": 8.149991419255192e-06,
+      "loss": 0.6621,
+      "step": 8488
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.021470069885254,
+      "learning_rate": 8.147417195812597e-06,
+      "loss": 0.4055,
+      "step": 8489
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.236593246459961,
+      "learning_rate": 8.144842972370002e-06,
+      "loss": 0.5235,
+      "step": 8490
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.95366096496582,
+      "learning_rate": 8.142268748927407e-06,
+      "loss": 0.48,
+      "step": 8491
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.198348999023438,
+      "learning_rate": 8.139694525484812e-06,
+      "loss": 0.467,
+      "step": 8492
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.833405494689941,
+      "learning_rate": 8.137120302042218e-06,
+      "loss": 0.3358,
+      "step": 8493
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.655987739562988,
+      "learning_rate": 8.134546078599623e-06,
+      "loss": 0.5777,
+      "step": 8494
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.078511238098145,
+      "learning_rate": 8.131971855157028e-06,
+      "loss": 0.4674,
+      "step": 8495
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.794244766235352,
+      "learning_rate": 8.129397631714433e-06,
+      "loss": 0.4774,
+      "step": 8496
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.90333080291748,
+      "learning_rate": 8.126823408271838e-06,
+      "loss": 0.4316,
+      "step": 8497
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.14542007446289,
+      "learning_rate": 8.124249184829243e-06,
+      "loss": 0.4482,
+      "step": 8498
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.563660621643066,
+      "learning_rate": 8.121674961386648e-06,
+      "loss": 0.5368,
+      "step": 8499
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.97414493560791,
+      "learning_rate": 8.119100737944053e-06,
+      "loss": 0.4153,
+      "step": 8500
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.99500560760498,
+      "learning_rate": 8.116526514501458e-06,
+      "loss": 0.4568,
+      "step": 8501
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.135347366333008,
+      "learning_rate": 8.113952291058865e-06,
+      "loss": 0.3883,
+      "step": 8502
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.680998802185059,
+      "learning_rate": 8.11137806761627e-06,
+      "loss": 0.5393,
+      "step": 8503
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.127410888671875,
+      "learning_rate": 8.108803844173675e-06,
+      "loss": 0.5414,
+      "step": 8504
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.314909934997559,
+      "learning_rate": 8.10622962073108e-06,
+      "loss": 0.3377,
+      "step": 8505
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.325355529785156,
+      "learning_rate": 8.103655397288485e-06,
+      "loss": 0.8095,
+      "step": 8506
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.419428825378418,
+      "learning_rate": 8.10108117384589e-06,
+      "loss": 0.4026,
+      "step": 8507
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.552887916564941,
+      "learning_rate": 8.098506950403295e-06,
+      "loss": 0.5147,
+      "step": 8508
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.195188522338867,
+      "learning_rate": 8.0959327269607e-06,
+      "loss": 0.4182,
+      "step": 8509
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.206841468811035,
+      "learning_rate": 8.093358503518105e-06,
+      "loss": 0.5025,
+      "step": 8510
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.249957084655762,
+      "learning_rate": 8.09078428007551e-06,
+      "loss": 0.2775,
+      "step": 8511
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.243851661682129,
+      "learning_rate": 8.088210056632917e-06,
+      "loss": 0.4352,
+      "step": 8512
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.872894287109375,
+      "learning_rate": 8.085635833190322e-06,
+      "loss": 0.331,
+      "step": 8513
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.04076099395752,
+      "learning_rate": 8.083061609747727e-06,
+      "loss": 0.363,
+      "step": 8514
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.39079475402832,
+      "learning_rate": 8.080487386305131e-06,
+      "loss": 0.4156,
+      "step": 8515
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.378559589385986,
+      "learning_rate": 8.077913162862538e-06,
+      "loss": 0.2945,
+      "step": 8516
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.975764274597168,
+      "learning_rate": 8.075338939419941e-06,
+      "loss": 0.5242,
+      "step": 8517
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.688372611999512,
+      "learning_rate": 8.072764715977346e-06,
+      "loss": 0.5725,
+      "step": 8518
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.530060768127441,
+      "learning_rate": 8.070190492534751e-06,
+      "loss": 0.5253,
+      "step": 8519
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.306748390197754,
+      "learning_rate": 8.067616269092156e-06,
+      "loss": 0.5221,
+      "step": 8520
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.011393547058105,
+      "learning_rate": 8.065042045649563e-06,
+      "loss": 0.5562,
+      "step": 8521
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.171546936035156,
+      "learning_rate": 8.062467822206968e-06,
+      "loss": 0.3144,
+      "step": 8522
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.945513725280762,
+      "learning_rate": 8.059893598764373e-06,
+      "loss": 0.5019,
+      "step": 8523
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.490241050720215,
+      "learning_rate": 8.057319375321778e-06,
+      "loss": 0.4945,
+      "step": 8524
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.841147422790527,
+      "learning_rate": 8.054745151879183e-06,
+      "loss": 0.479,
+      "step": 8525
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.656150817871094,
+      "learning_rate": 8.05217092843659e-06,
+      "loss": 0.4177,
+      "step": 8526
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.543272018432617,
+      "learning_rate": 8.049596704993995e-06,
+      "loss": 0.4383,
+      "step": 8527
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.524797439575195,
+      "learning_rate": 8.047022481551398e-06,
+      "loss": 0.5641,
+      "step": 8528
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.433579444885254,
+      "learning_rate": 8.044448258108803e-06,
+      "loss": 0.4055,
+      "step": 8529
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.726025581359863,
+      "learning_rate": 8.041874034666208e-06,
+      "loss": 0.4021,
+      "step": 8530
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.450407028198242,
+      "learning_rate": 8.039299811223615e-06,
+      "loss": 0.3207,
+      "step": 8531
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.087461471557617,
+      "learning_rate": 8.03672558778102e-06,
+      "loss": 0.4748,
+      "step": 8532
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.776268005371094,
+      "learning_rate": 8.034151364338425e-06,
+      "loss": 0.3336,
+      "step": 8533
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.61547565460205,
+      "learning_rate": 8.03157714089583e-06,
+      "loss": 0.3658,
+      "step": 8534
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.384796142578125,
+      "learning_rate": 8.029002917453236e-06,
+      "loss": 0.6244,
+      "step": 8535
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.831313133239746,
+      "learning_rate": 8.026428694010641e-06,
+      "loss": 0.3752,
+      "step": 8536
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.768578052520752,
+      "learning_rate": 8.023854470568046e-06,
+      "loss": 0.3222,
+      "step": 8537
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.5904693603515625,
+      "learning_rate": 8.02128024712545e-06,
+      "loss": 0.324,
+      "step": 8538
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.682817459106445,
+      "learning_rate": 8.018706023682855e-06,
+      "loss": 0.5733,
+      "step": 8539
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.738873481750488,
+      "learning_rate": 8.016131800240261e-06,
+      "loss": 0.7028,
+      "step": 8540
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.896012306213379,
+      "learning_rate": 8.013557576797666e-06,
+      "loss": 0.2906,
+      "step": 8541
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.165061950683594,
+      "learning_rate": 8.010983353355071e-06,
+      "loss": 0.5081,
+      "step": 8542
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.817131042480469,
+      "learning_rate": 8.008409129912476e-06,
+      "loss": 0.4214,
+      "step": 8543
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.760784149169922,
+      "learning_rate": 8.005834906469881e-06,
+      "loss": 0.5577,
+      "step": 8544
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.888113021850586,
+      "learning_rate": 8.003260683027288e-06,
+      "loss": 0.4521,
+      "step": 8545
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.918608665466309,
+      "learning_rate": 8.000686459584693e-06,
+      "loss": 0.3398,
+      "step": 8546
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.417941093444824,
+      "learning_rate": 7.998112236142098e-06,
+      "loss": 0.3342,
+      "step": 8547
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.36760711669922,
+      "learning_rate": 7.995538012699503e-06,
+      "loss": 0.4828,
+      "step": 8548
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.83072280883789,
+      "learning_rate": 7.992963789256908e-06,
+      "loss": 0.5709,
+      "step": 8549
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.945616722106934,
+      "learning_rate": 7.990389565814313e-06,
+      "loss": 0.4415,
+      "step": 8550
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.12635612487793,
+      "learning_rate": 7.987815342371718e-06,
+      "loss": 0.4411,
+      "step": 8551
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.528286933898926,
+      "learning_rate": 7.985241118929123e-06,
+      "loss": 0.2534,
+      "step": 8552
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.1712646484375,
+      "learning_rate": 7.982666895486528e-06,
+      "loss": 0.3184,
+      "step": 8553
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.892218589782715,
+      "learning_rate": 7.980092672043934e-06,
+      "loss": 0.3492,
+      "step": 8554
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.465339660644531,
+      "learning_rate": 7.97751844860134e-06,
+      "loss": 0.5743,
+      "step": 8555
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.969108581542969,
+      "learning_rate": 7.974944225158744e-06,
+      "loss": 0.5014,
+      "step": 8556
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.110034942626953,
+      "learning_rate": 7.97237000171615e-06,
+      "loss": 0.5086,
+      "step": 8557
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.674614906311035,
+      "learning_rate": 7.969795778273554e-06,
+      "loss": 0.3687,
+      "step": 8558
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.170455932617188,
+      "learning_rate": 7.96722155483096e-06,
+      "loss": 0.3053,
+      "step": 8559
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.258048057556152,
+      "learning_rate": 7.964647331388364e-06,
+      "loss": 0.4365,
+      "step": 8560
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.805398941040039,
+      "learning_rate": 7.96207310794577e-06,
+      "loss": 0.4557,
+      "step": 8561
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.54407787322998,
+      "learning_rate": 7.959498884503174e-06,
+      "loss": 0.4291,
+      "step": 8562
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.470309257507324,
+      "learning_rate": 7.95692466106058e-06,
+      "loss": 0.4608,
+      "step": 8563
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.896372318267822,
+      "learning_rate": 7.954350437617986e-06,
+      "loss": 0.2956,
+      "step": 8564
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.786768913269043,
+      "learning_rate": 7.951776214175391e-06,
+      "loss": 0.4783,
+      "step": 8565
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.320167541503906,
+      "learning_rate": 7.949201990732796e-06,
+      "loss": 0.4196,
+      "step": 8566
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.008817672729492,
+      "learning_rate": 7.946627767290201e-06,
+      "loss": 0.329,
+      "step": 8567
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.293103218078613,
+      "learning_rate": 7.944053543847608e-06,
+      "loss": 0.5097,
+      "step": 8568
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.558850288391113,
+      "learning_rate": 7.941479320405011e-06,
+      "loss": 0.4128,
+      "step": 8569
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.128281593322754,
+      "learning_rate": 7.938905096962416e-06,
+      "loss": 0.4256,
+      "step": 8570
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.740893363952637,
+      "learning_rate": 7.936330873519821e-06,
+      "loss": 0.3899,
+      "step": 8571
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.472328186035156,
+      "learning_rate": 7.933756650077226e-06,
+      "loss": 0.5626,
+      "step": 8572
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.7655611038208,
+      "learning_rate": 7.931182426634633e-06,
+      "loss": 0.4221,
+      "step": 8573
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.645916938781738,
+      "learning_rate": 7.928608203192038e-06,
+      "loss": 0.4746,
+      "step": 8574
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.39921760559082,
+      "learning_rate": 7.926033979749443e-06,
+      "loss": 0.3937,
+      "step": 8575
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.620342254638672,
+      "learning_rate": 7.923459756306848e-06,
+      "loss": 0.4633,
+      "step": 8576
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.516475677490234,
+      "learning_rate": 7.920885532864253e-06,
+      "loss": 0.4572,
+      "step": 8577
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.848872184753418,
+      "learning_rate": 7.91831130942166e-06,
+      "loss": 0.6008,
+      "step": 8578
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.542719841003418,
+      "learning_rate": 7.915737085979064e-06,
+      "loss": 0.4618,
+      "step": 8579
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.783872604370117,
+      "learning_rate": 7.913162862536468e-06,
+      "loss": 0.6557,
+      "step": 8580
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.77277946472168,
+      "learning_rate": 7.910588639093873e-06,
+      "loss": 0.7083,
+      "step": 8581
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.899014472961426,
+      "learning_rate": 7.908014415651278e-06,
+      "loss": 0.5602,
+      "step": 8582
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.835158348083496,
+      "learning_rate": 7.905440192208684e-06,
+      "loss": 0.6009,
+      "step": 8583
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.717292785644531,
+      "learning_rate": 7.90286596876609e-06,
+      "loss": 0.4338,
+      "step": 8584
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.928191184997559,
+      "learning_rate": 7.900291745323494e-06,
+      "loss": 0.5056,
+      "step": 8585
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.180452346801758,
+      "learning_rate": 7.8977175218809e-06,
+      "loss": 0.56,
+      "step": 8586
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.932937622070312,
+      "learning_rate": 7.895143298438306e-06,
+      "loss": 0.5892,
+      "step": 8587
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.647933959960938,
+      "learning_rate": 7.89256907499571e-06,
+      "loss": 0.6721,
+      "step": 8588
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.494047164916992,
+      "learning_rate": 7.889994851553116e-06,
+      "loss": 0.3785,
+      "step": 8589
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.854719161987305,
+      "learning_rate": 7.887420628110519e-06,
+      "loss": 0.5584,
+      "step": 8590
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.866348266601562,
+      "learning_rate": 7.884846404667924e-06,
+      "loss": 0.5017,
+      "step": 8591
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.433438301086426,
+      "learning_rate": 7.88227218122533e-06,
+      "loss": 0.5221,
+      "step": 8592
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.125801086425781,
+      "learning_rate": 7.879697957782736e-06,
+      "loss": 0.3946,
+      "step": 8593
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.81082534790039,
+      "learning_rate": 7.87712373434014e-06,
+      "loss": 0.5728,
+      "step": 8594
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.547959327697754,
+      "learning_rate": 7.874549510897546e-06,
+      "loss": 0.369,
+      "step": 8595
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.833579063415527,
+      "learning_rate": 7.87197528745495e-06,
+      "loss": 0.4068,
+      "step": 8596
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.35151195526123,
+      "learning_rate": 7.869401064012357e-06,
+      "loss": 0.3471,
+      "step": 8597
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.833970069885254,
+      "learning_rate": 7.866826840569762e-06,
+      "loss": 0.3091,
+      "step": 8598
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.919172286987305,
+      "learning_rate": 7.864252617127167e-06,
+      "loss": 0.3985,
+      "step": 8599
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.889188766479492,
+      "learning_rate": 7.861678393684572e-06,
+      "loss": 0.4031,
+      "step": 8600
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.633176803588867,
+      "learning_rate": 7.859104170241977e-06,
+      "loss": 0.3812,
+      "step": 8601
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.027216911315918,
+      "learning_rate": 7.856529946799382e-06,
+      "loss": 0.3904,
+      "step": 8602
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.757156372070312,
+      "learning_rate": 7.853955723356787e-06,
+      "loss": 0.49,
+      "step": 8603
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.519168853759766,
+      "learning_rate": 7.851381499914192e-06,
+      "loss": 0.3731,
+      "step": 8604
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.353914260864258,
+      "learning_rate": 7.848807276471597e-06,
+      "loss": 0.4911,
+      "step": 8605
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.230640411376953,
+      "learning_rate": 7.846233053029004e-06,
+      "loss": 0.4357,
+      "step": 8606
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.016966819763184,
+      "learning_rate": 7.843658829586409e-06,
+      "loss": 0.3928,
+      "step": 8607
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.400897979736328,
+      "learning_rate": 7.841084606143814e-06,
+      "loss": 0.4244,
+      "step": 8608
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.500144004821777,
+      "learning_rate": 7.838510382701219e-06,
+      "loss": 0.495,
+      "step": 8609
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.50043773651123,
+      "learning_rate": 7.835936159258624e-06,
+      "loss": 0.3381,
+      "step": 8610
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.781012535095215,
+      "learning_rate": 7.833361935816029e-06,
+      "loss": 0.5251,
+      "step": 8611
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.041146278381348,
+      "learning_rate": 7.830787712373434e-06,
+      "loss": 0.5681,
+      "step": 8612
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.248954772949219,
+      "learning_rate": 7.828213488930839e-06,
+      "loss": 0.6538,
+      "step": 8613
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.323101043701172,
+      "learning_rate": 7.825639265488244e-06,
+      "loss": 0.5628,
+      "step": 8614
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.977293014526367,
+      "learning_rate": 7.823065042045649e-06,
+      "loss": 0.3591,
+      "step": 8615
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.425943374633789,
+      "learning_rate": 7.820490818603056e-06,
+      "loss": 0.2616,
+      "step": 8616
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.769509792327881,
+      "learning_rate": 7.81791659516046e-06,
+      "loss": 0.4714,
+      "step": 8617
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.433910369873047,
+      "learning_rate": 7.815342371717866e-06,
+      "loss": 0.3575,
+      "step": 8618
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.992460250854492,
+      "learning_rate": 7.81276814827527e-06,
+      "loss": 0.3664,
+      "step": 8619
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.948832511901855,
+      "learning_rate": 7.810193924832677e-06,
+      "loss": 0.5124,
+      "step": 8620
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.187300682067871,
+      "learning_rate": 7.80761970139008e-06,
+      "loss": 0.3882,
+      "step": 8621
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.126838684082031,
+      "learning_rate": 7.805045477947485e-06,
+      "loss": 0.4334,
+      "step": 8622
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.636470794677734,
+      "learning_rate": 7.80247125450489e-06,
+      "loss": 0.2723,
+      "step": 8623
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.808019638061523,
+      "learning_rate": 7.799897031062295e-06,
+      "loss": 0.566,
+      "step": 8624
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.529194831848145,
+      "learning_rate": 7.797322807619702e-06,
+      "loss": 0.3412,
+      "step": 8625
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.64117431640625,
+      "learning_rate": 7.794748584177107e-06,
+      "loss": 0.28,
+      "step": 8626
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.318421363830566,
+      "learning_rate": 7.792174360734512e-06,
+      "loss": 0.3969,
+      "step": 8627
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.63068962097168,
+      "learning_rate": 7.789600137291917e-06,
+      "loss": 0.51,
+      "step": 8628
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.368439674377441,
+      "learning_rate": 7.787025913849322e-06,
+      "loss": 0.5441,
+      "step": 8629
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.952978134155273,
+      "learning_rate": 7.784451690406729e-06,
+      "loss": 0.6755,
+      "step": 8630
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.3170747756958,
+      "learning_rate": 7.781877466964134e-06,
+      "loss": 0.4856,
+      "step": 8631
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.046541213989258,
+      "learning_rate": 7.779303243521537e-06,
+      "loss": 0.4159,
+      "step": 8632
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.291326522827148,
+      "learning_rate": 7.776729020078942e-06,
+      "loss": 0.3775,
+      "step": 8633
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.550436019897461,
+      "learning_rate": 7.774154796636347e-06,
+      "loss": 0.4928,
+      "step": 8634
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.879800796508789,
+      "learning_rate": 7.771580573193754e-06,
+      "loss": 0.4117,
+      "step": 8635
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.166871070861816,
+      "learning_rate": 7.769006349751159e-06,
+      "loss": 0.4399,
+      "step": 8636
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.039190292358398,
+      "learning_rate": 7.766432126308564e-06,
+      "loss": 0.3285,
+      "step": 8637
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.225506782531738,
+      "learning_rate": 7.763857902865969e-06,
+      "loss": 0.3806,
+      "step": 8638
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.985786437988281,
+      "learning_rate": 7.761283679423375e-06,
+      "loss": 0.4005,
+      "step": 8639
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.728994369506836,
+      "learning_rate": 7.75870945598078e-06,
+      "loss": 0.7569,
+      "step": 8640
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.25137710571289,
+      "learning_rate": 7.756135232538185e-06,
+      "loss": 0.4928,
+      "step": 8641
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.436782836914062,
+      "learning_rate": 7.753561009095589e-06,
+      "loss": 0.3492,
+      "step": 8642
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.46200180053711,
+      "learning_rate": 7.750986785652994e-06,
+      "loss": 0.3731,
+      "step": 8643
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.370363235473633,
+      "learning_rate": 7.7484125622104e-06,
+      "loss": 0.5277,
+      "step": 8644
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.704399108886719,
+      "learning_rate": 7.745838338767805e-06,
+      "loss": 0.5361,
+      "step": 8645
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.174317359924316,
+      "learning_rate": 7.74326411532521e-06,
+      "loss": 0.5948,
+      "step": 8646
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.262345314025879,
+      "learning_rate": 7.740689891882615e-06,
+      "loss": 0.3897,
+      "step": 8647
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.378447532653809,
+      "learning_rate": 7.73811566844002e-06,
+      "loss": 0.4219,
+      "step": 8648
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.001533508300781,
+      "learning_rate": 7.735541444997427e-06,
+      "loss": 0.3486,
+      "step": 8649
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.994129180908203,
+      "learning_rate": 7.732967221554832e-06,
+      "loss": 0.5329,
+      "step": 8650
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.308484077453613,
+      "learning_rate": 7.730392998112237e-06,
+      "loss": 0.4263,
+      "step": 8651
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.41646957397461,
+      "learning_rate": 7.727818774669642e-06,
+      "loss": 0.4035,
+      "step": 8652
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.040307998657227,
+      "learning_rate": 7.725244551227045e-06,
+      "loss": 0.4165,
+      "step": 8653
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.31472110748291,
+      "learning_rate": 7.722670327784452e-06,
+      "loss": 0.2045,
+      "step": 8654
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.812907218933105,
+      "learning_rate": 7.720096104341857e-06,
+      "loss": 0.449,
+      "step": 8655
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.812792778015137,
+      "learning_rate": 7.717521880899262e-06,
+      "loss": 0.4032,
+      "step": 8656
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.149288177490234,
+      "learning_rate": 7.714947657456667e-06,
+      "loss": 0.5115,
+      "step": 8657
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.846478462219238,
+      "learning_rate": 7.712373434014073e-06,
+      "loss": 0.4938,
+      "step": 8658
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.545721054077148,
+      "learning_rate": 7.709799210571478e-06,
+      "loss": 0.4208,
+      "step": 8659
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.555359840393066,
+      "learning_rate": 7.707224987128883e-06,
+      "loss": 0.4141,
+      "step": 8660
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.194311141967773,
+      "learning_rate": 7.704650763686288e-06,
+      "loss": 0.3589,
+      "step": 8661
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.833417892456055,
+      "learning_rate": 7.702076540243693e-06,
+      "loss": 0.3768,
+      "step": 8662
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.746025085449219,
+      "learning_rate": 7.699502316801098e-06,
+      "loss": 0.5061,
+      "step": 8663
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.398209571838379,
+      "learning_rate": 7.696928093358503e-06,
+      "loss": 0.5899,
+      "step": 8664
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.046601295471191,
+      "learning_rate": 7.694353869915908e-06,
+      "loss": 0.497,
+      "step": 8665
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.254277229309082,
+      "learning_rate": 7.691779646473313e-06,
+      "loss": 0.4254,
+      "step": 8666
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.432557106018066,
+      "learning_rate": 7.689205423030718e-06,
+      "loss": 0.3644,
+      "step": 8667
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.011787414550781,
+      "learning_rate": 7.686631199588125e-06,
+      "loss": 0.5751,
+      "step": 8668
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.807391166687012,
+      "learning_rate": 7.68405697614553e-06,
+      "loss": 0.6072,
+      "step": 8669
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.687145233154297,
+      "learning_rate": 7.681482752702935e-06,
+      "loss": 0.3684,
+      "step": 8670
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.121129035949707,
+      "learning_rate": 7.67890852926034e-06,
+      "loss": 0.3502,
+      "step": 8671
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.828241348266602,
+      "learning_rate": 7.676334305817747e-06,
+      "loss": 0.3774,
+      "step": 8672
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.942272186279297,
+      "learning_rate": 7.67376008237515e-06,
+      "loss": 0.4005,
+      "step": 8673
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.567776679992676,
+      "learning_rate": 7.671185858932555e-06,
+      "loss": 0.4092,
+      "step": 8674
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.88443660736084,
+      "learning_rate": 7.66861163548996e-06,
+      "loss": 0.3564,
+      "step": 8675
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.76310157775879,
+      "learning_rate": 7.666037412047365e-06,
+      "loss": 0.5737,
+      "step": 8676
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.955726623535156,
+      "learning_rate": 7.663463188604772e-06,
+      "loss": 0.4338,
+      "step": 8677
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.855793952941895,
+      "learning_rate": 7.660888965162177e-06,
+      "loss": 0.4027,
+      "step": 8678
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.609536170959473,
+      "learning_rate": 7.658314741719582e-06,
+      "loss": 0.4467,
+      "step": 8679
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.255777359008789,
+      "learning_rate": 7.655740518276987e-06,
+      "loss": 0.3606,
+      "step": 8680
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.6880521774292,
+      "learning_rate": 7.653166294834392e-06,
+      "loss": 0.5056,
+      "step": 8681
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.712709426879883,
+      "learning_rate": 7.650592071391798e-06,
+      "loss": 0.4961,
+      "step": 8682
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.465036392211914,
+      "learning_rate": 7.648017847949203e-06,
+      "loss": 0.5416,
+      "step": 8683
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.517153263092041,
+      "learning_rate": 7.645443624506607e-06,
+      "loss": 0.3217,
+      "step": 8684
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.14057445526123,
+      "learning_rate": 7.642869401064012e-06,
+      "loss": 0.4006,
+      "step": 8685
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.679612159729004,
+      "learning_rate": 7.640295177621417e-06,
+      "loss": 0.3961,
+      "step": 8686
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.151049613952637,
+      "learning_rate": 7.637720954178823e-06,
+      "loss": 0.6611,
+      "step": 8687
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.980060577392578,
+      "learning_rate": 7.635146730736228e-06,
+      "loss": 0.348,
+      "step": 8688
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.323481559753418,
+      "learning_rate": 7.632572507293633e-06,
+      "loss": 0.4482,
+      "step": 8689
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.23275375366211,
+      "learning_rate": 7.629998283851038e-06,
+      "loss": 0.2792,
+      "step": 8690
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.62765121459961,
+      "learning_rate": 7.627424060408444e-06,
+      "loss": 0.4991,
+      "step": 8691
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.456879138946533,
+      "learning_rate": 7.624849836965849e-06,
+      "loss": 0.3401,
+      "step": 8692
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.61656379699707,
+      "learning_rate": 7.622275613523254e-06,
+      "loss": 0.5759,
+      "step": 8693
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.339529037475586,
+      "learning_rate": 7.619701390080659e-06,
+      "loss": 0.3965,
+      "step": 8694
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.059610366821289,
+      "learning_rate": 7.617127166638064e-06,
+      "loss": 0.4692,
+      "step": 8695
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.67989730834961,
+      "learning_rate": 7.614552943195471e-06,
+      "loss": 0.5743,
+      "step": 8696
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.272076606750488,
+      "learning_rate": 7.611978719752875e-06,
+      "loss": 0.3803,
+      "step": 8697
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.38070011138916,
+      "learning_rate": 7.60940449631028e-06,
+      "loss": 0.4559,
+      "step": 8698
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.39708423614502,
+      "learning_rate": 7.606830272867685e-06,
+      "loss": 0.3623,
+      "step": 8699
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.156476974487305,
+      "learning_rate": 7.60425604942509e-06,
+      "loss": 0.5037,
+      "step": 8700
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.293466567993164,
+      "learning_rate": 7.601681825982496e-06,
+      "loss": 0.5061,
+      "step": 8701
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.782387733459473,
+      "learning_rate": 7.5991076025399006e-06,
+      "loss": 0.2963,
+      "step": 8702
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.22900676727295,
+      "learning_rate": 7.5965333790973055e-06,
+      "loss": 0.4414,
+      "step": 8703
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.715869903564453,
+      "learning_rate": 7.5939591556547105e-06,
+      "loss": 0.5806,
+      "step": 8704
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.745909690856934,
+      "learning_rate": 7.5913849322121155e-06,
+      "loss": 0.3812,
+      "step": 8705
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.796847343444824,
+      "learning_rate": 7.588810708769522e-06,
+      "loss": 0.4646,
+      "step": 8706
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.717659950256348,
+      "learning_rate": 7.586236485326926e-06,
+      "loss": 0.4574,
+      "step": 8707
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.462380409240723,
+      "learning_rate": 7.583662261884331e-06,
+      "loss": 0.3608,
+      "step": 8708
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.34654712677002,
+      "learning_rate": 7.581088038441736e-06,
+      "loss": 0.4558,
+      "step": 8709
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.999104499816895,
+      "learning_rate": 7.578513814999143e-06,
+      "loss": 0.4434,
+      "step": 8710
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.085166931152344,
+      "learning_rate": 7.575939591556548e-06,
+      "loss": 0.5302,
+      "step": 8711
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.097311019897461,
+      "learning_rate": 7.573365368113952e-06,
+      "loss": 0.5046,
+      "step": 8712
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.378292083740234,
+      "learning_rate": 7.570791144671357e-06,
+      "loss": 0.5099,
+      "step": 8713
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.888908386230469,
+      "learning_rate": 7.568216921228762e-06,
+      "loss": 0.4807,
+      "step": 8714
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.898755073547363,
+      "learning_rate": 7.565642697786169e-06,
+      "loss": 0.3219,
+      "step": 8715
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.602651596069336,
+      "learning_rate": 7.563068474343574e-06,
+      "loss": 0.6849,
+      "step": 8716
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.659979820251465,
+      "learning_rate": 7.560494250900979e-06,
+      "loss": 0.3166,
+      "step": 8717
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.922460556030273,
+      "learning_rate": 7.557920027458383e-06,
+      "loss": 0.3806,
+      "step": 8718
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.665297508239746,
+      "learning_rate": 7.555345804015788e-06,
+      "loss": 0.3061,
+      "step": 8719
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.128588676452637,
+      "learning_rate": 7.5527715805731946e-06,
+      "loss": 0.3975,
+      "step": 8720
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.131848335266113,
+      "learning_rate": 7.5501973571305995e-06,
+      "loss": 0.5227,
+      "step": 8721
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.696487426757812,
+      "learning_rate": 7.5476231336880045e-06,
+      "loss": 0.4333,
+      "step": 8722
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.940535545349121,
+      "learning_rate": 7.545048910245409e-06,
+      "loss": 0.4935,
+      "step": 8723
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.68030071258545,
+      "learning_rate": 7.542474686802815e-06,
+      "loss": 0.4789,
+      "step": 8724
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.738661766052246,
+      "learning_rate": 7.53990046336022e-06,
+      "loss": 0.3503,
+      "step": 8725
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.187797546386719,
+      "learning_rate": 7.537326239917625e-06,
+      "loss": 0.4757,
+      "step": 8726
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.924798011779785,
+      "learning_rate": 7.53475201647503e-06,
+      "loss": 0.4424,
+      "step": 8727
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.826604843139648,
+      "learning_rate": 7.5321777930324345e-06,
+      "loss": 0.3712,
+      "step": 8728
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.118193626403809,
+      "learning_rate": 7.529603569589841e-06,
+      "loss": 0.5455,
+      "step": 8729
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.980628967285156,
+      "learning_rate": 7.527029346147246e-06,
+      "loss": 0.4883,
+      "step": 8730
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.775498390197754,
+      "learning_rate": 7.524455122704651e-06,
+      "loss": 0.5881,
+      "step": 8731
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.3221435546875,
+      "learning_rate": 7.521880899262056e-06,
+      "loss": 0.3668,
+      "step": 8732
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.927716255187988,
+      "learning_rate": 7.51930667581946e-06,
+      "loss": 0.4116,
+      "step": 8733
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.804611206054688,
+      "learning_rate": 7.516732452376867e-06,
+      "loss": 0.3445,
+      "step": 8734
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.995001792907715,
+      "learning_rate": 7.514158228934272e-06,
+      "loss": 0.4496,
+      "step": 8735
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.850483894348145,
+      "learning_rate": 7.511584005491677e-06,
+      "loss": 0.5389,
+      "step": 8736
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.620684623718262,
+      "learning_rate": 7.509009782049082e-06,
+      "loss": 0.3659,
+      "step": 8737
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.97374439239502,
+      "learning_rate": 7.506435558606486e-06,
+      "loss": 0.4827,
+      "step": 8738
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.738988876342773,
+      "learning_rate": 7.503861335163893e-06,
+      "loss": 0.3773,
+      "step": 8739
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.055754661560059,
+      "learning_rate": 7.501287111721298e-06,
+      "loss": 0.3924,
+      "step": 8740
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.052687644958496,
+      "learning_rate": 7.498712888278703e-06,
+      "loss": 0.4134,
+      "step": 8741
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.5737223625183105,
+      "learning_rate": 7.4961386648361085e-06,
+      "loss": 0.1778,
+      "step": 8742
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.387993812561035,
+      "learning_rate": 7.4935644413935135e-06,
+      "loss": 0.4352,
+      "step": 8743
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.305328369140625,
+      "learning_rate": 7.490990217950918e-06,
+      "loss": 0.4074,
+      "step": 8744
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.951504230499268,
+      "learning_rate": 7.4884159945083235e-06,
+      "loss": 0.3848,
+      "step": 8745
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.153668403625488,
+      "learning_rate": 7.4858417710657285e-06,
+      "loss": 0.4418,
+      "step": 8746
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.094598770141602,
+      "learning_rate": 7.483267547623134e-06,
+      "loss": 0.4926,
+      "step": 8747
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.65021800994873,
+      "learning_rate": 7.480693324180539e-06,
+      "loss": 0.367,
+      "step": 8748
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.264941215515137,
+      "learning_rate": 7.478119100737944e-06,
+      "loss": 0.4585,
+      "step": 8749
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.575928688049316,
+      "learning_rate": 7.475544877295349e-06,
+      "loss": 0.3404,
+      "step": 8750
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.107653617858887,
+      "learning_rate": 7.472970653852754e-06,
+      "loss": 0.62,
+      "step": 8751
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.695974349975586,
+      "learning_rate": 7.47039643041016e-06,
+      "loss": 0.533,
+      "step": 8752
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.48951244354248,
+      "learning_rate": 7.467822206967565e-06,
+      "loss": 0.3823,
+      "step": 8753
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.928346633911133,
+      "learning_rate": 7.46524798352497e-06,
+      "loss": 0.551,
+      "step": 8754
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.034577369689941,
+      "learning_rate": 7.462673760082375e-06,
+      "loss": 0.2812,
+      "step": 8755
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.299479484558105,
+      "learning_rate": 7.460099536639781e-06,
+      "loss": 0.6269,
+      "step": 8756
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.282571792602539,
+      "learning_rate": 7.457525313197186e-06,
+      "loss": 0.4905,
+      "step": 8757
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.033225059509277,
+      "learning_rate": 7.454951089754591e-06,
+      "loss": 0.3846,
+      "step": 8758
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.580182075500488,
+      "learning_rate": 7.452376866311996e-06,
+      "loss": 0.6844,
+      "step": 8759
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.297609329223633,
+      "learning_rate": 7.449802642869401e-06,
+      "loss": 0.5394,
+      "step": 8760
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.419766426086426,
+      "learning_rate": 7.447228419426807e-06,
+      "loss": 0.3593,
+      "step": 8761
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.03036880493164,
+      "learning_rate": 7.444654195984212e-06,
+      "loss": 0.4566,
+      "step": 8762
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.862035751342773,
+      "learning_rate": 7.4420799725416175e-06,
+      "loss": 0.4694,
+      "step": 8763
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.9309163093566895,
+      "learning_rate": 7.439505749099022e-06,
+      "loss": 0.4008,
+      "step": 8764
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.637907028198242,
+      "learning_rate": 7.436931525656427e-06,
+      "loss": 0.5061,
+      "step": 8765
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.582741737365723,
+      "learning_rate": 7.4343573022138324e-06,
+      "loss": 0.3703,
+      "step": 8766
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.502876281738281,
+      "learning_rate": 7.4317830787712374e-06,
+      "loss": 0.4593,
+      "step": 8767
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.501891136169434,
+      "learning_rate": 7.429208855328643e-06,
+      "loss": 0.4167,
+      "step": 8768
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.096879005432129,
+      "learning_rate": 7.426634631886048e-06,
+      "loss": 0.5099,
+      "step": 8769
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.736690521240234,
+      "learning_rate": 7.424060408443452e-06,
+      "loss": 0.3577,
+      "step": 8770
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.838726997375488,
+      "learning_rate": 7.421486185000858e-06,
+      "loss": 0.609,
+      "step": 8771
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.117023468017578,
+      "learning_rate": 7.418911961558263e-06,
+      "loss": 0.3835,
+      "step": 8772
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.22922134399414,
+      "learning_rate": 7.416337738115669e-06,
+      "loss": 0.4301,
+      "step": 8773
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.718122482299805,
+      "learning_rate": 7.413763514673074e-06,
+      "loss": 0.2766,
+      "step": 8774
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.26065731048584,
+      "learning_rate": 7.411189291230479e-06,
+      "loss": 0.4337,
+      "step": 8775
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.731683254241943,
+      "learning_rate": 7.408615067787884e-06,
+      "loss": 0.259,
+      "step": 8776
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.79808235168457,
+      "learning_rate": 7.406040844345289e-06,
+      "loss": 0.3705,
+      "step": 8777
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.64966106414795,
+      "learning_rate": 7.403466620902695e-06,
+      "loss": 0.425,
+      "step": 8778
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.392223358154297,
+      "learning_rate": 7.4008923974601e-06,
+      "loss": 0.4592,
+      "step": 8779
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.400946617126465,
+      "learning_rate": 7.398318174017505e-06,
+      "loss": 0.2814,
+      "step": 8780
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.687527656555176,
+      "learning_rate": 7.39574395057491e-06,
+      "loss": 0.5491,
+      "step": 8781
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.366559028625488,
+      "learning_rate": 7.393169727132316e-06,
+      "loss": 0.3551,
+      "step": 8782
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.964877128601074,
+      "learning_rate": 7.390595503689721e-06,
+      "loss": 0.5435,
+      "step": 8783
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.320745468139648,
+      "learning_rate": 7.388021280247126e-06,
+      "loss": 0.3814,
+      "step": 8784
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.121947288513184,
+      "learning_rate": 7.385447056804531e-06,
+      "loss": 0.4195,
+      "step": 8785
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.1561279296875,
+      "learning_rate": 7.382872833361936e-06,
+      "loss": 0.5622,
+      "step": 8786
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.045967102050781,
+      "learning_rate": 7.380298609919341e-06,
+      "loss": 0.5008,
+      "step": 8787
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.236106872558594,
+      "learning_rate": 7.377724386476746e-06,
+      "loss": 0.445,
+      "step": 8788
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.588759422302246,
+      "learning_rate": 7.375150163034152e-06,
+      "loss": 0.278,
+      "step": 8789
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.644856452941895,
+      "learning_rate": 7.372575939591556e-06,
+      "loss": 0.4051,
+      "step": 8790
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.774545669555664,
+      "learning_rate": 7.370001716148961e-06,
+      "loss": 0.4701,
+      "step": 8791
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.966662406921387,
+      "learning_rate": 7.367427492706367e-06,
+      "loss": 0.4374,
+      "step": 8792
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.633502960205078,
+      "learning_rate": 7.364853269263772e-06,
+      "loss": 0.494,
+      "step": 8793
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.987869262695312,
+      "learning_rate": 7.362279045821178e-06,
+      "loss": 0.3418,
+      "step": 8794
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.426177024841309,
+      "learning_rate": 7.359704822378583e-06,
+      "loss": 0.3568,
+      "step": 8795
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.096763610839844,
+      "learning_rate": 7.357130598935987e-06,
+      "loss": 0.4037,
+      "step": 8796
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.230522155761719,
+      "learning_rate": 7.354556375493393e-06,
+      "loss": 0.423,
+      "step": 8797
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.045573234558105,
+      "learning_rate": 7.351982152050798e-06,
+      "loss": 0.3674,
+      "step": 8798
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.338088989257812,
+      "learning_rate": 7.349407928608204e-06,
+      "loss": 0.3108,
+      "step": 8799
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.044685363769531,
+      "learning_rate": 7.346833705165609e-06,
+      "loss": 0.4526,
+      "step": 8800
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.16036319732666,
+      "learning_rate": 7.344259481723014e-06,
+      "loss": 0.4964,
+      "step": 8801
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.049835205078125,
+      "learning_rate": 7.341685258280419e-06,
+      "loss": 0.3937,
+      "step": 8802
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.86475658416748,
+      "learning_rate": 7.339111034837824e-06,
+      "loss": 0.5843,
+      "step": 8803
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.912973403930664,
+      "learning_rate": 7.33653681139523e-06,
+      "loss": 0.3669,
+      "step": 8804
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.2679443359375,
+      "learning_rate": 7.333962587952635e-06,
+      "loss": 0.3577,
+      "step": 8805
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.166956901550293,
+      "learning_rate": 7.3313883645100396e-06,
+      "loss": 0.3852,
+      "step": 8806
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.623123168945312,
+      "learning_rate": 7.3288141410674446e-06,
+      "loss": 0.3887,
+      "step": 8807
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.616179466247559,
+      "learning_rate": 7.32623991762485e-06,
+      "loss": 0.329,
+      "step": 8808
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.576656341552734,
+      "learning_rate": 7.323665694182255e-06,
+      "loss": 0.4326,
+      "step": 8809
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.386253356933594,
+      "learning_rate": 7.32109147073966e-06,
+      "loss": 0.3872,
+      "step": 8810
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.359831809997559,
+      "learning_rate": 7.318517247297065e-06,
+      "loss": 0.5517,
+      "step": 8811
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.12345027923584,
+      "learning_rate": 7.31594302385447e-06,
+      "loss": 0.4971,
+      "step": 8812
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.607744216918945,
+      "learning_rate": 7.313368800411876e-06,
+      "loss": 0.6341,
+      "step": 8813
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.886641502380371,
+      "learning_rate": 7.310794576969281e-06,
+      "loss": 0.5391,
+      "step": 8814
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.09367847442627,
+      "learning_rate": 7.308220353526687e-06,
+      "loss": 0.4405,
+      "step": 8815
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.571110725402832,
+      "learning_rate": 7.305646130084091e-06,
+      "loss": 0.4618,
+      "step": 8816
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.874504566192627,
+      "learning_rate": 7.303071906641496e-06,
+      "loss": 0.3187,
+      "step": 8817
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.266547203063965,
+      "learning_rate": 7.300497683198902e-06,
+      "loss": 0.5731,
+      "step": 8818
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.223689079284668,
+      "learning_rate": 7.297923459756307e-06,
+      "loss": 0.5282,
+      "step": 8819
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.01658821105957,
+      "learning_rate": 7.295349236313713e-06,
+      "loss": 0.4101,
+      "step": 8820
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.149382591247559,
+      "learning_rate": 7.292775012871118e-06,
+      "loss": 0.4485,
+      "step": 8821
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.120978355407715,
+      "learning_rate": 7.290200789428522e-06,
+      "loss": 0.3263,
+      "step": 8822
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.204527854919434,
+      "learning_rate": 7.287626565985928e-06,
+      "loss": 0.4411,
+      "step": 8823
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.927420616149902,
+      "learning_rate": 7.285052342543333e-06,
+      "loss": 0.2149,
+      "step": 8824
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.783214092254639,
+      "learning_rate": 7.2824781191007386e-06,
+      "loss": 0.4074,
+      "step": 8825
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.478473663330078,
+      "learning_rate": 7.2799038956581435e-06,
+      "loss": 0.3461,
+      "step": 8826
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.341203689575195,
+      "learning_rate": 7.2773296722155485e-06,
+      "loss": 0.4026,
+      "step": 8827
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.5017728805542,
+      "learning_rate": 7.2747554487729535e-06,
+      "loss": 0.5483,
+      "step": 8828
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.103543281555176,
+      "learning_rate": 7.2721812253303585e-06,
+      "loss": 0.4359,
+      "step": 8829
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.962469100952148,
+      "learning_rate": 7.269607001887764e-06,
+      "loss": 0.4857,
+      "step": 8830
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.309229850769043,
+      "learning_rate": 7.267032778445169e-06,
+      "loss": 0.4297,
+      "step": 8831
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.478974342346191,
+      "learning_rate": 7.264458555002574e-06,
+      "loss": 0.5057,
+      "step": 8832
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.26053524017334,
+      "learning_rate": 7.261884331559979e-06,
+      "loss": 0.5801,
+      "step": 8833
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.356572151184082,
+      "learning_rate": 7.259310108117385e-06,
+      "loss": 0.4289,
+      "step": 8834
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.247749328613281,
+      "learning_rate": 7.25673588467479e-06,
+      "loss": 0.5264,
+      "step": 8835
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.72059440612793,
+      "learning_rate": 7.254161661232195e-06,
+      "loss": 0.3276,
+      "step": 8836
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.958900451660156,
+      "learning_rate": 7.2515874377896e-06,
+      "loss": 0.4227,
+      "step": 8837
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.569451332092285,
+      "learning_rate": 7.249013214347005e-06,
+      "loss": 0.4005,
+      "step": 8838
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.54477310180664,
+      "learning_rate": 7.246438990904411e-06,
+      "loss": 0.4857,
+      "step": 8839
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.293161392211914,
+      "learning_rate": 7.243864767461816e-06,
+      "loss": 0.6239,
+      "step": 8840
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.453324317932129,
+      "learning_rate": 7.241290544019221e-06,
+      "loss": 0.3432,
+      "step": 8841
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.279464721679688,
+      "learning_rate": 7.238716320576626e-06,
+      "loss": 0.44,
+      "step": 8842
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.750658988952637,
+      "learning_rate": 7.236142097134031e-06,
+      "loss": 0.4933,
+      "step": 8843
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.113375663757324,
+      "learning_rate": 7.233567873691437e-06,
+      "loss": 0.6137,
+      "step": 8844
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.700080871582031,
+      "learning_rate": 7.230993650248842e-06,
+      "loss": 0.5215,
+      "step": 8845
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.13374137878418,
+      "learning_rate": 7.2284194268062475e-06,
+      "loss": 0.4844,
+      "step": 8846
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.928226470947266,
+      "learning_rate": 7.2258452033636525e-06,
+      "loss": 0.4761,
+      "step": 8847
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.967079162597656,
+      "learning_rate": 7.223270979921057e-06,
+      "loss": 0.4367,
+      "step": 8848
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.112823486328125,
+      "learning_rate": 7.2206967564784625e-06,
+      "loss": 0.6469,
+      "step": 8849
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.774311542510986,
+      "learning_rate": 7.2181225330358675e-06,
+      "loss": 0.2604,
+      "step": 8850
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.273468971252441,
+      "learning_rate": 7.215548309593273e-06,
+      "loss": 0.5598,
+      "step": 8851
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.786847114562988,
+      "learning_rate": 7.212974086150678e-06,
+      "loss": 0.2727,
+      "step": 8852
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.64932632446289,
+      "learning_rate": 7.210399862708083e-06,
+      "loss": 0.4634,
+      "step": 8853
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.51461124420166,
+      "learning_rate": 7.207825639265488e-06,
+      "loss": 0.714,
+      "step": 8854
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.96403980255127,
+      "learning_rate": 7.205251415822893e-06,
+      "loss": 0.5432,
+      "step": 8855
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.334166526794434,
+      "learning_rate": 7.202677192380299e-06,
+      "loss": 0.3426,
+      "step": 8856
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.989980697631836,
+      "learning_rate": 7.200102968937704e-06,
+      "loss": 0.4662,
+      "step": 8857
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.669950485229492,
+      "learning_rate": 7.197528745495109e-06,
+      "loss": 0.397,
+      "step": 8858
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.047621726989746,
+      "learning_rate": 7.194954522052514e-06,
+      "loss": 0.4182,
+      "step": 8859
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.333658218383789,
+      "learning_rate": 7.19238029860992e-06,
+      "loss": 0.3585,
+      "step": 8860
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.259073257446289,
+      "learning_rate": 7.189806075167325e-06,
+      "loss": 0.4596,
+      "step": 8861
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.235369682312012,
+      "learning_rate": 7.18723185172473e-06,
+      "loss": 0.3864,
+      "step": 8862
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.717405319213867,
+      "learning_rate": 7.184657628282135e-06,
+      "loss": 0.4641,
+      "step": 8863
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.15607738494873,
+      "learning_rate": 7.18208340483954e-06,
+      "loss": 0.5655,
+      "step": 8864
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.473398208618164,
+      "learning_rate": 7.179509181396946e-06,
+      "loss": 0.497,
+      "step": 8865
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.998419761657715,
+      "learning_rate": 7.176934957954351e-06,
+      "loss": 0.3287,
+      "step": 8866
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.598323822021484,
+      "learning_rate": 7.174360734511756e-06,
+      "loss": 0.6619,
+      "step": 8867
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.139378547668457,
+      "learning_rate": 7.171786511069161e-06,
+      "loss": 0.4889,
+      "step": 8868
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.548771858215332,
+      "learning_rate": 7.169212287626566e-06,
+      "loss": 0.4528,
+      "step": 8869
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.250882148742676,
+      "learning_rate": 7.1666380641839715e-06,
+      "loss": 0.6002,
+      "step": 8870
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.281623840332031,
+      "learning_rate": 7.1640638407413764e-06,
+      "loss": 0.3651,
+      "step": 8871
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.502840042114258,
+      "learning_rate": 7.161489617298782e-06,
+      "loss": 0.453,
+      "step": 8872
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.626325607299805,
+      "learning_rate": 7.158915393856187e-06,
+      "loss": 0.3073,
+      "step": 8873
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.452362060546875,
+      "learning_rate": 7.156341170413591e-06,
+      "loss": 0.5024,
+      "step": 8874
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.148929595947266,
+      "learning_rate": 7.153766946970997e-06,
+      "loss": 0.4041,
+      "step": 8875
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.789012908935547,
+      "learning_rate": 7.151192723528402e-06,
+      "loss": 0.2853,
+      "step": 8876
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.083465576171875,
+      "learning_rate": 7.148618500085808e-06,
+      "loss": 0.4297,
+      "step": 8877
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.260814666748047,
+      "learning_rate": 7.146044276643213e-06,
+      "loss": 0.3215,
+      "step": 8878
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.21704387664795,
+      "learning_rate": 7.143470053200618e-06,
+      "loss": 0.4656,
+      "step": 8879
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.400568008422852,
+      "learning_rate": 7.140895829758023e-06,
+      "loss": 0.4299,
+      "step": 8880
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.600332260131836,
+      "learning_rate": 7.138321606315428e-06,
+      "loss": 0.4117,
+      "step": 8881
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.670083045959473,
+      "learning_rate": 7.135747382872834e-06,
+      "loss": 0.4319,
+      "step": 8882
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.604001998901367,
+      "learning_rate": 7.133173159430239e-06,
+      "loss": 0.5183,
+      "step": 8883
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.450639724731445,
+      "learning_rate": 7.130598935987644e-06,
+      "loss": 0.4138,
+      "step": 8884
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.210122108459473,
+      "learning_rate": 7.128024712545049e-06,
+      "loss": 0.5035,
+      "step": 8885
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.121723175048828,
+      "learning_rate": 7.125450489102455e-06,
+      "loss": 0.3984,
+      "step": 8886
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.203651428222656,
+      "learning_rate": 7.12287626565986e-06,
+      "loss": 0.432,
+      "step": 8887
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.5178804397583,
+      "learning_rate": 7.120302042217265e-06,
+      "loss": 0.376,
+      "step": 8888
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.63821029663086,
+      "learning_rate": 7.11772781877467e-06,
+      "loss": 0.3921,
+      "step": 8889
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.6951322555542,
+      "learning_rate": 7.115153595332075e-06,
+      "loss": 0.4086,
+      "step": 8890
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.29124641418457,
+      "learning_rate": 7.1125793718894804e-06,
+      "loss": 0.3772,
+      "step": 8891
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.061479568481445,
+      "learning_rate": 7.110005148446885e-06,
+      "loss": 0.5001,
+      "step": 8892
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.922256469726562,
+      "learning_rate": 7.10743092500429e-06,
+      "loss": 0.4717,
+      "step": 8893
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.323261260986328,
+      "learning_rate": 7.104856701561695e-06,
+      "loss": 0.4749,
+      "step": 8894
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.216887474060059,
+      "learning_rate": 7.1022824781191e-06,
+      "loss": 0.3984,
+      "step": 8895
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.934128284454346,
+      "learning_rate": 7.099708254676506e-06,
+      "loss": 0.347,
+      "step": 8896
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.633722305297852,
+      "learning_rate": 7.097134031233911e-06,
+      "loss": 0.2435,
+      "step": 8897
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.857534408569336,
+      "learning_rate": 7.094559807791317e-06,
+      "loss": 0.4294,
+      "step": 8898
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.378945350646973,
+      "learning_rate": 7.091985584348722e-06,
+      "loss": 0.4699,
+      "step": 8899
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.56597900390625,
+      "learning_rate": 7.089411360906126e-06,
+      "loss": 0.4952,
+      "step": 8900
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.527883052825928,
+      "learning_rate": 7.086837137463532e-06,
+      "loss": 0.2861,
+      "step": 8901
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.969795227050781,
+      "learning_rate": 7.084262914020937e-06,
+      "loss": 0.3248,
+      "step": 8902
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.945636749267578,
+      "learning_rate": 7.081688690578343e-06,
+      "loss": 0.5072,
+      "step": 8903
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.161673545837402,
+      "learning_rate": 7.079114467135748e-06,
+      "loss": 0.5091,
+      "step": 8904
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.02252197265625,
+      "learning_rate": 7.076540243693153e-06,
+      "loss": 0.5127,
+      "step": 8905
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.72762680053711,
+      "learning_rate": 7.073966020250558e-06,
+      "loss": 0.4311,
+      "step": 8906
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.358551025390625,
+      "learning_rate": 7.071391796807963e-06,
+      "loss": 0.471,
+      "step": 8907
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.956890106201172,
+      "learning_rate": 7.068817573365369e-06,
+      "loss": 0.5217,
+      "step": 8908
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.607876777648926,
+      "learning_rate": 7.066243349922774e-06,
+      "loss": 0.4731,
+      "step": 8909
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.18957233428955,
+      "learning_rate": 7.063669126480179e-06,
+      "loss": 0.4956,
+      "step": 8910
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.583208084106445,
+      "learning_rate": 7.0610949030375836e-06,
+      "loss": 0.4276,
+      "step": 8911
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.812610149383545,
+      "learning_rate": 7.058520679594989e-06,
+      "loss": 0.3667,
+      "step": 8912
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.129200458526611,
+      "learning_rate": 7.055946456152394e-06,
+      "loss": 0.3663,
+      "step": 8913
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.108561515808105,
+      "learning_rate": 7.053372232709799e-06,
+      "loss": 0.4189,
+      "step": 8914
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.96666145324707,
+      "learning_rate": 7.050798009267204e-06,
+      "loss": 0.4068,
+      "step": 8915
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.690185546875,
+      "learning_rate": 7.048223785824609e-06,
+      "loss": 0.4316,
+      "step": 8916
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.282842636108398,
+      "learning_rate": 7.045649562382015e-06,
+      "loss": 0.3765,
+      "step": 8917
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.994156837463379,
+      "learning_rate": 7.04307533893942e-06,
+      "loss": 0.3158,
+      "step": 8918
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.971477508544922,
+      "learning_rate": 7.040501115496825e-06,
+      "loss": 0.432,
+      "step": 8919
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.713102340698242,
+      "learning_rate": 7.03792689205423e-06,
+      "loss": 0.4613,
+      "step": 8920
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.9313554763793945,
+      "learning_rate": 7.035352668611635e-06,
+      "loss": 0.2449,
+      "step": 8921
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.674796104431152,
+      "learning_rate": 7.032778445169041e-06,
+      "loss": 0.4648,
+      "step": 8922
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.270509719848633,
+      "learning_rate": 7.030204221726446e-06,
+      "loss": 0.5068,
+      "step": 8923
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.4141206741333,
+      "learning_rate": 7.027629998283852e-06,
+      "loss": 0.5071,
+      "step": 8924
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.394927978515625,
+      "learning_rate": 7.025055774841257e-06,
+      "loss": 0.5717,
+      "step": 8925
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.928528785705566,
+      "learning_rate": 7.022481551398661e-06,
+      "loss": 0.5141,
+      "step": 8926
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.652988910675049,
+      "learning_rate": 7.019907327956067e-06,
+      "loss": 0.3254,
+      "step": 8927
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.221549034118652,
+      "learning_rate": 7.017333104513472e-06,
+      "loss": 0.4598,
+      "step": 8928
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.18213176727295,
+      "learning_rate": 7.0147588810708776e-06,
+      "loss": 0.2799,
+      "step": 8929
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.492405891418457,
+      "learning_rate": 7.0121846576282826e-06,
+      "loss": 0.3331,
+      "step": 8930
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.051721572875977,
+      "learning_rate": 7.0096104341856875e-06,
+      "loss": 0.4662,
+      "step": 8931
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.982498168945312,
+      "learning_rate": 7.0070362107430925e-06,
+      "loss": 0.4243,
+      "step": 8932
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.501224517822266,
+      "learning_rate": 7.0044619873004975e-06,
+      "loss": 0.3283,
+      "step": 8933
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.438892364501953,
+      "learning_rate": 7.001887763857903e-06,
+      "loss": 0.5087,
+      "step": 8934
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.742033004760742,
+      "learning_rate": 6.999313540415308e-06,
+      "loss": 0.326,
+      "step": 8935
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.81969165802002,
+      "learning_rate": 6.996739316972713e-06,
+      "loss": 0.5256,
+      "step": 8936
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.44371509552002,
+      "learning_rate": 6.994165093530118e-06,
+      "loss": 0.4008,
+      "step": 8937
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.709961891174316,
+      "learning_rate": 6.991590870087524e-06,
+      "loss": 0.4038,
+      "step": 8938
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.391035079956055,
+      "learning_rate": 6.989016646644929e-06,
+      "loss": 0.5656,
+      "step": 8939
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.154583930969238,
+      "learning_rate": 6.986442423202334e-06,
+      "loss": 0.4149,
+      "step": 8940
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.773641586303711,
+      "learning_rate": 6.983868199759739e-06,
+      "loss": 0.4047,
+      "step": 8941
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.217438697814941,
+      "learning_rate": 6.981293976317144e-06,
+      "loss": 0.444,
+      "step": 8942
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.037546157836914,
+      "learning_rate": 6.97871975287455e-06,
+      "loss": 0.4977,
+      "step": 8943
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.7324857711792,
+      "learning_rate": 6.976145529431955e-06,
+      "loss": 0.3783,
+      "step": 8944
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.169418334960938,
+      "learning_rate": 6.97357130598936e-06,
+      "loss": 0.3564,
+      "step": 8945
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.172502517700195,
+      "learning_rate": 6.970997082546765e-06,
+      "loss": 0.5238,
+      "step": 8946
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.07720947265625,
+      "learning_rate": 6.96842285910417e-06,
+      "loss": 0.3891,
+      "step": 8947
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.466992378234863,
+      "learning_rate": 6.965848635661576e-06,
+      "loss": 0.3416,
+      "step": 8948
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.41238021850586,
+      "learning_rate": 6.963274412218981e-06,
+      "loss": 0.3941,
+      "step": 8949
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.027323722839355,
+      "learning_rate": 6.9607001887763865e-06,
+      "loss": 0.4548,
+      "step": 8950
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.93846321105957,
+      "learning_rate": 6.9581259653337915e-06,
+      "loss": 0.5196,
+      "step": 8951
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.976224899291992,
+      "learning_rate": 6.955551741891196e-06,
+      "loss": 0.4226,
+      "step": 8952
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.822880744934082,
+      "learning_rate": 6.9529775184486015e-06,
+      "loss": 0.4668,
+      "step": 8953
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.665153503417969,
+      "learning_rate": 6.9504032950060065e-06,
+      "loss": 0.3618,
+      "step": 8954
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.610836029052734,
+      "learning_rate": 6.947829071563412e-06,
+      "loss": 0.5179,
+      "step": 8955
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.167706489562988,
+      "learning_rate": 6.945254848120817e-06,
+      "loss": 0.4276,
+      "step": 8956
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.83510684967041,
+      "learning_rate": 6.942680624678222e-06,
+      "loss": 0.6989,
+      "step": 8957
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.535457611083984,
+      "learning_rate": 6.940106401235627e-06,
+      "loss": 0.2847,
+      "step": 8958
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.525236129760742,
+      "learning_rate": 6.937532177793032e-06,
+      "loss": 0.5602,
+      "step": 8959
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.786735534667969,
+      "learning_rate": 6.934957954350438e-06,
+      "loss": 0.4485,
+      "step": 8960
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.261709213256836,
+      "learning_rate": 6.932383730907843e-06,
+      "loss": 0.4715,
+      "step": 8961
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.32877254486084,
+      "learning_rate": 6.929809507465248e-06,
+      "loss": 0.4365,
+      "step": 8962
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.638503551483154,
+      "learning_rate": 6.927235284022653e-06,
+      "loss": 0.358,
+      "step": 8963
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.470928192138672,
+      "learning_rate": 6.924661060580059e-06,
+      "loss": 0.2755,
+      "step": 8964
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.32476806640625,
+      "learning_rate": 6.922086837137464e-06,
+      "loss": 0.3925,
+      "step": 8965
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.3812642097473145,
+      "learning_rate": 6.919512613694869e-06,
+      "loss": 0.3318,
+      "step": 8966
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.82723331451416,
+      "learning_rate": 6.916938390252274e-06,
+      "loss": 0.4636,
+      "step": 8967
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.385515213012695,
+      "learning_rate": 6.914364166809679e-06,
+      "loss": 0.4615,
+      "step": 8968
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.83295726776123,
+      "learning_rate": 6.911789943367085e-06,
+      "loss": 0.3849,
+      "step": 8969
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.831999778747559,
+      "learning_rate": 6.90921571992449e-06,
+      "loss": 0.408,
+      "step": 8970
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.939336776733398,
+      "learning_rate": 6.906641496481895e-06,
+      "loss": 0.4841,
+      "step": 8971
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.491720199584961,
+      "learning_rate": 6.9040672730393005e-06,
+      "loss": 0.3959,
+      "step": 8972
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.336962699890137,
+      "learning_rate": 6.901493049596705e-06,
+      "loss": 0.5526,
+      "step": 8973
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.131930351257324,
+      "learning_rate": 6.8989188261541105e-06,
+      "loss": 0.598,
+      "step": 8974
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.194853782653809,
+      "learning_rate": 6.8963446027115155e-06,
+      "loss": 0.5016,
+      "step": 8975
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.512124061584473,
+      "learning_rate": 6.893770379268921e-06,
+      "loss": 0.4137,
+      "step": 8976
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.344644546508789,
+      "learning_rate": 6.891196155826326e-06,
+      "loss": 0.5206,
+      "step": 8977
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.666017532348633,
+      "learning_rate": 6.88862193238373e-06,
+      "loss": 0.4154,
+      "step": 8978
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.957460403442383,
+      "learning_rate": 6.886047708941136e-06,
+      "loss": 0.3911,
+      "step": 8979
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.622474670410156,
+      "learning_rate": 6.883473485498541e-06,
+      "loss": 0.4925,
+      "step": 8980
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.922347068786621,
+      "learning_rate": 6.880899262055947e-06,
+      "loss": 0.4897,
+      "step": 8981
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.485464096069336,
+      "learning_rate": 6.878325038613352e-06,
+      "loss": 0.3414,
+      "step": 8982
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.115612030029297,
+      "learning_rate": 6.875750815170757e-06,
+      "loss": 0.3409,
+      "step": 8983
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.991114616394043,
+      "learning_rate": 6.873176591728162e-06,
+      "loss": 0.4447,
+      "step": 8984
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.786396026611328,
+      "learning_rate": 6.870602368285567e-06,
+      "loss": 0.6268,
+      "step": 8985
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.771339416503906,
+      "learning_rate": 6.868028144842973e-06,
+      "loss": 0.3625,
+      "step": 8986
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.360448837280273,
+      "learning_rate": 6.865453921400378e-06,
+      "loss": 0.2887,
+      "step": 8987
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.364771366119385,
+      "learning_rate": 6.862879697957783e-06,
+      "loss": 0.3066,
+      "step": 8988
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.924720764160156,
+      "learning_rate": 6.860305474515188e-06,
+      "loss": 0.3879,
+      "step": 8989
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.039410591125488,
+      "learning_rate": 6.857731251072594e-06,
+      "loss": 0.3501,
+      "step": 8990
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.326106071472168,
+      "learning_rate": 6.855157027629999e-06,
+      "loss": 0.4389,
+      "step": 8991
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.962635040283203,
+      "learning_rate": 6.852582804187404e-06,
+      "loss": 0.4185,
+      "step": 8992
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.42863130569458,
+      "learning_rate": 6.850008580744809e-06,
+      "loss": 0.3356,
+      "step": 8993
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.837570190429688,
+      "learning_rate": 6.847434357302214e-06,
+      "loss": 0.465,
+      "step": 8994
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.257991790771484,
+      "learning_rate": 6.8448601338596194e-06,
+      "loss": 0.3991,
+      "step": 8995
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.193087577819824,
+      "learning_rate": 6.8422859104170244e-06,
+      "loss": 0.3424,
+      "step": 8996
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.22000789642334,
+      "learning_rate": 6.839711686974429e-06,
+      "loss": 0.5843,
+      "step": 8997
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.147873878479004,
+      "learning_rate": 6.837137463531835e-06,
+      "loss": 0.5504,
+      "step": 8998
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.118026733398438,
+      "learning_rate": 6.834563240089239e-06,
+      "loss": 0.4202,
+      "step": 8999
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.525010108947754,
+      "learning_rate": 6.831989016646645e-06,
+      "loss": 0.3876,
+      "step": 9000
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.929567337036133,
+      "learning_rate": 6.82941479320405e-06,
+      "loss": 0.2729,
+      "step": 9001
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.239514350891113,
+      "learning_rate": 6.826840569761456e-06,
+      "loss": 0.3721,
+      "step": 9002
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.794919013977051,
+      "learning_rate": 6.824266346318861e-06,
+      "loss": 0.3175,
+      "step": 9003
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.50118350982666,
+      "learning_rate": 6.821692122876265e-06,
+      "loss": 0.293,
+      "step": 9004
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.85807991027832,
+      "learning_rate": 6.819117899433671e-06,
+      "loss": 0.38,
+      "step": 9005
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.937372207641602,
+      "learning_rate": 6.816543675991076e-06,
+      "loss": 0.4364,
+      "step": 9006
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.955855369567871,
+      "learning_rate": 6.813969452548482e-06,
+      "loss": 0.5283,
+      "step": 9007
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.119773864746094,
+      "learning_rate": 6.811395229105887e-06,
+      "loss": 0.4431,
+      "step": 9008
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.668829917907715,
+      "learning_rate": 6.808821005663292e-06,
+      "loss": 0.5373,
+      "step": 9009
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.992827415466309,
+      "learning_rate": 6.806246782220697e-06,
+      "loss": 0.2839,
+      "step": 9010
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.80026626586914,
+      "learning_rate": 6.803672558778102e-06,
+      "loss": 0.344,
+      "step": 9011
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.294907569885254,
+      "learning_rate": 6.801098335335508e-06,
+      "loss": 0.4355,
+      "step": 9012
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.616134643554688,
+      "learning_rate": 6.798524111892913e-06,
+      "loss": 0.3739,
+      "step": 9013
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.69835090637207,
+      "learning_rate": 6.795949888450318e-06,
+      "loss": 0.3783,
+      "step": 9014
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.762347221374512,
+      "learning_rate": 6.793375665007723e-06,
+      "loss": 0.4616,
+      "step": 9015
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.7774019241333,
+      "learning_rate": 6.7908014415651276e-06,
+      "loss": 0.3046,
+      "step": 9016
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.603497505187988,
+      "learning_rate": 6.788227218122533e-06,
+      "loss": 0.4686,
+      "step": 9017
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.491743087768555,
+      "learning_rate": 6.785652994679938e-06,
+      "loss": 0.5825,
+      "step": 9018
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.463603019714355,
+      "learning_rate": 6.783078771237343e-06,
+      "loss": 0.5188,
+      "step": 9019
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.45566177368164,
+      "learning_rate": 6.780504547794748e-06,
+      "loss": 0.4284,
+      "step": 9020
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.350882530212402,
+      "learning_rate": 6.777930324352154e-06,
+      "loss": 0.3459,
+      "step": 9021
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.684252738952637,
+      "learning_rate": 6.775356100909559e-06,
+      "loss": 0.5135,
+      "step": 9022
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.462064743041992,
+      "learning_rate": 6.772781877466964e-06,
+      "loss": 0.4263,
+      "step": 9023
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.511894226074219,
+      "learning_rate": 6.77020765402437e-06,
+      "loss": 0.3327,
+      "step": 9024
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.778379440307617,
+      "learning_rate": 6.767633430581774e-06,
+      "loss": 0.4092,
+      "step": 9025
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.334806442260742,
+      "learning_rate": 6.76505920713918e-06,
+      "loss": 0.298,
+      "step": 9026
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.341312408447266,
+      "learning_rate": 6.762484983696585e-06,
+      "loss": 0.4947,
+      "step": 9027
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.079545974731445,
+      "learning_rate": 6.759910760253991e-06,
+      "loss": 0.4858,
+      "step": 9028
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.366859436035156,
+      "learning_rate": 6.757336536811396e-06,
+      "loss": 0.4849,
+      "step": 9029
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.166098594665527,
+      "learning_rate": 6.7547623133688e-06,
+      "loss": 0.4984,
+      "step": 9030
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.932158470153809,
+      "learning_rate": 6.752188089926206e-06,
+      "loss": 0.508,
+      "step": 9031
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.931735038757324,
+      "learning_rate": 6.749613866483611e-06,
+      "loss": 0.3258,
+      "step": 9032
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.567687034606934,
+      "learning_rate": 6.747039643041017e-06,
+      "loss": 0.4604,
+      "step": 9033
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.96845817565918,
+      "learning_rate": 6.7444654195984216e-06,
+      "loss": 0.5776,
+      "step": 9034
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.238383293151855,
+      "learning_rate": 6.7418911961558266e-06,
+      "loss": 0.3858,
+      "step": 9035
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.755646705627441,
+      "learning_rate": 6.7393169727132315e-06,
+      "loss": 0.2998,
+      "step": 9036
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.348224639892578,
+      "learning_rate": 6.7367427492706365e-06,
+      "loss": 0.3914,
+      "step": 9037
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.106252670288086,
+      "learning_rate": 6.734168525828042e-06,
+      "loss": 0.3123,
+      "step": 9038
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.297879219055176,
+      "learning_rate": 6.731594302385447e-06,
+      "loss": 0.3621,
+      "step": 9039
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.97221565246582,
+      "learning_rate": 6.729020078942852e-06,
+      "loss": 0.4046,
+      "step": 9040
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.938226699829102,
+      "learning_rate": 6.726445855500257e-06,
+      "loss": 0.3809,
+      "step": 9041
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.279706001281738,
+      "learning_rate": 6.723871632057662e-06,
+      "loss": 0.3617,
+      "step": 9042
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.85624885559082,
+      "learning_rate": 6.721297408615068e-06,
+      "loss": 0.4283,
+      "step": 9043
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.55158519744873,
+      "learning_rate": 6.718723185172473e-06,
+      "loss": 0.3592,
+      "step": 9044
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.287115097045898,
+      "learning_rate": 6.716148961729878e-06,
+      "loss": 0.5733,
+      "step": 9045
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.168319702148438,
+      "learning_rate": 6.713574738287283e-06,
+      "loss": 0.4338,
+      "step": 9046
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.264942169189453,
+      "learning_rate": 6.711000514844689e-06,
+      "loss": 0.3034,
+      "step": 9047
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.841805458068848,
+      "learning_rate": 6.708426291402094e-06,
+      "loss": 0.5603,
+      "step": 9048
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.720842361450195,
+      "learning_rate": 6.705852067959499e-06,
+      "loss": 0.6207,
+      "step": 9049
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.516718864440918,
+      "learning_rate": 6.703277844516905e-06,
+      "loss": 0.3713,
+      "step": 9050
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.178365707397461,
+      "learning_rate": 6.700703621074309e-06,
+      "loss": 0.5639,
+      "step": 9051
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.041665077209473,
+      "learning_rate": 6.698129397631715e-06,
+      "loss": 0.4252,
+      "step": 9052
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.102350234985352,
+      "learning_rate": 6.69555517418912e-06,
+      "loss": 0.5042,
+      "step": 9053
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.4866361618042,
+      "learning_rate": 6.6929809507465256e-06,
+      "loss": 0.4346,
+      "step": 9054
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.604265213012695,
+      "learning_rate": 6.6904067273039305e-06,
+      "loss": 0.4202,
+      "step": 9055
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.317880630493164,
+      "learning_rate": 6.687832503861335e-06,
+      "loss": 0.2992,
+      "step": 9056
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.931004524230957,
+      "learning_rate": 6.6852582804187405e-06,
+      "loss": 0.2154,
+      "step": 9057
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.68259048461914,
+      "learning_rate": 6.6826840569761455e-06,
+      "loss": 0.356,
+      "step": 9058
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.403766632080078,
+      "learning_rate": 6.680109833533551e-06,
+      "loss": 0.4383,
+      "step": 9059
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.369772911071777,
+      "learning_rate": 6.677535610090956e-06,
+      "loss": 0.4297,
+      "step": 9060
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.384984016418457,
+      "learning_rate": 6.674961386648361e-06,
+      "loss": 0.3296,
+      "step": 9061
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.11343002319336,
+      "learning_rate": 6.672387163205766e-06,
+      "loss": 0.2974,
+      "step": 9062
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.19269847869873,
+      "learning_rate": 6.669812939763171e-06,
+      "loss": 0.4339,
+      "step": 9063
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.9432954788208,
+      "learning_rate": 6.667238716320577e-06,
+      "loss": 0.4885,
+      "step": 9064
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.548988342285156,
+      "learning_rate": 6.664664492877982e-06,
+      "loss": 0.418,
+      "step": 9065
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.793389320373535,
+      "learning_rate": 6.662090269435387e-06,
+      "loss": 0.4407,
+      "step": 9066
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.234336853027344,
+      "learning_rate": 6.659516045992792e-06,
+      "loss": 0.4532,
+      "step": 9067
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.90408992767334,
+      "learning_rate": 6.656941822550197e-06,
+      "loss": 0.5414,
+      "step": 9068
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.602997779846191,
+      "learning_rate": 6.654367599107603e-06,
+      "loss": 0.4308,
+      "step": 9069
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.916178703308105,
+      "learning_rate": 6.651793375665008e-06,
+      "loss": 0.5975,
+      "step": 9070
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.920973777770996,
+      "learning_rate": 6.649219152222413e-06,
+      "loss": 0.2993,
+      "step": 9071
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.38058853149414,
+      "learning_rate": 6.646644928779818e-06,
+      "loss": 0.3294,
+      "step": 9072
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.297099113464355,
+      "learning_rate": 6.644070705337224e-06,
+      "loss": 0.5424,
+      "step": 9073
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.37918472290039,
+      "learning_rate": 6.641496481894629e-06,
+      "loss": 0.6976,
+      "step": 9074
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.919448852539062,
+      "learning_rate": 6.638922258452034e-06,
+      "loss": 0.3287,
+      "step": 9075
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.900317192077637,
+      "learning_rate": 6.6363480350094395e-06,
+      "loss": 0.5989,
+      "step": 9076
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.163420677185059,
+      "learning_rate": 6.633773811566844e-06,
+      "loss": 0.4301,
+      "step": 9077
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.837841033935547,
+      "learning_rate": 6.6311995881242495e-06,
+      "loss": 0.3274,
+      "step": 9078
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.50175952911377,
+      "learning_rate": 6.6286253646816545e-06,
+      "loss": 0.4362,
+      "step": 9079
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.980319023132324,
+      "learning_rate": 6.62605114123906e-06,
+      "loss": 0.3522,
+      "step": 9080
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.221705436706543,
+      "learning_rate": 6.623476917796465e-06,
+      "loss": 0.5547,
+      "step": 9081
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.201438903808594,
+      "learning_rate": 6.6209026943538694e-06,
+      "loss": 0.3081,
+      "step": 9082
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.045942306518555,
+      "learning_rate": 6.618328470911275e-06,
+      "loss": 0.3885,
+      "step": 9083
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.674612998962402,
+      "learning_rate": 6.61575424746868e-06,
+      "loss": 0.3766,
+      "step": 9084
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.371481895446777,
+      "learning_rate": 6.613180024026086e-06,
+      "loss": 0.2742,
+      "step": 9085
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.235065460205078,
+      "learning_rate": 6.610605800583491e-06,
+      "loss": 0.2868,
+      "step": 9086
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.495343208312988,
+      "learning_rate": 6.608031577140896e-06,
+      "loss": 0.5213,
+      "step": 9087
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.121530532836914,
+      "learning_rate": 6.605457353698301e-06,
+      "loss": 0.4886,
+      "step": 9088
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.821917533874512,
+      "learning_rate": 6.602883130255706e-06,
+      "loss": 0.4306,
+      "step": 9089
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.319422721862793,
+      "learning_rate": 6.600308906813112e-06,
+      "loss": 0.34,
+      "step": 9090
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.17870044708252,
+      "learning_rate": 6.597734683370517e-06,
+      "loss": 0.5238,
+      "step": 9091
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.177310943603516,
+      "learning_rate": 6.595160459927922e-06,
+      "loss": 0.3137,
+      "step": 9092
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.990798473358154,
+      "learning_rate": 6.592586236485327e-06,
+      "loss": 0.3652,
+      "step": 9093
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.063852310180664,
+      "learning_rate": 6.590012013042732e-06,
+      "loss": 0.4725,
+      "step": 9094
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.781620025634766,
+      "learning_rate": 6.587437789600138e-06,
+      "loss": 0.4493,
+      "step": 9095
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.202735900878906,
+      "learning_rate": 6.584863566157543e-06,
+      "loss": 0.3454,
+      "step": 9096
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.74743938446045,
+      "learning_rate": 6.582289342714948e-06,
+      "loss": 0.4237,
+      "step": 9097
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.113112449645996,
+      "learning_rate": 6.579715119272353e-06,
+      "loss": 0.4049,
+      "step": 9098
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.021286964416504,
+      "learning_rate": 6.5771408958297585e-06,
+      "loss": 0.6689,
+      "step": 9099
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.52872085571289,
+      "learning_rate": 6.5745666723871634e-06,
+      "loss": 0.3713,
+      "step": 9100
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.817242622375488,
+      "learning_rate": 6.5719924489445684e-06,
+      "loss": 0.4241,
+      "step": 9101
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.09481143951416,
+      "learning_rate": 6.569418225501974e-06,
+      "loss": 0.5944,
+      "step": 9102
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.745797157287598,
+      "learning_rate": 6.566844002059378e-06,
+      "loss": 0.2441,
+      "step": 9103
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.217633247375488,
+      "learning_rate": 6.564269778616784e-06,
+      "loss": 0.4832,
+      "step": 9104
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.284358978271484,
+      "learning_rate": 6.561695555174189e-06,
+      "loss": 0.3287,
+      "step": 9105
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.830768585205078,
+      "learning_rate": 6.559121331731595e-06,
+      "loss": 0.3138,
+      "step": 9106
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.012214660644531,
+      "learning_rate": 6.556547108289e-06,
+      "loss": 0.387,
+      "step": 9107
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.80935001373291,
+      "learning_rate": 6.553972884846404e-06,
+      "loss": 0.5218,
+      "step": 9108
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.77457857131958,
+      "learning_rate": 6.55139866140381e-06,
+      "loss": 0.3142,
+      "step": 9109
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.121886253356934,
+      "learning_rate": 6.548824437961215e-06,
+      "loss": 0.3844,
+      "step": 9110
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.070313453674316,
+      "learning_rate": 6.546250214518621e-06,
+      "loss": 0.531,
+      "step": 9111
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.64294147491455,
+      "learning_rate": 6.543675991076026e-06,
+      "loss": 0.2965,
+      "step": 9112
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.200862884521484,
+      "learning_rate": 6.541101767633431e-06,
+      "loss": 0.4901,
+      "step": 9113
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.238682746887207,
+      "learning_rate": 6.538527544190836e-06,
+      "loss": 0.3706,
+      "step": 9114
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.326116561889648,
+      "learning_rate": 6.535953320748241e-06,
+      "loss": 0.3243,
+      "step": 9115
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.95153522491455,
+      "learning_rate": 6.533379097305647e-06,
+      "loss": 0.4127,
+      "step": 9116
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.6895751953125,
+      "learning_rate": 6.530804873863052e-06,
+      "loss": 0.4495,
+      "step": 9117
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.63408088684082,
+      "learning_rate": 6.528230650420457e-06,
+      "loss": 0.3325,
+      "step": 9118
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.318384170532227,
+      "learning_rate": 6.525656426977862e-06,
+      "loss": 0.4507,
+      "step": 9119
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.620682716369629,
+      "learning_rate": 6.5230822035352666e-06,
+      "loss": 0.387,
+      "step": 9120
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.932801246643066,
+      "learning_rate": 6.520507980092672e-06,
+      "loss": 0.3956,
+      "step": 9121
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.0022611618042,
+      "learning_rate": 6.517933756650077e-06,
+      "loss": 0.3439,
+      "step": 9122
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.975605010986328,
+      "learning_rate": 6.515359533207482e-06,
+      "loss": 0.3452,
+      "step": 9123
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.023692607879639,
+      "learning_rate": 6.512785309764887e-06,
+      "loss": 0.3325,
+      "step": 9124
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.353510856628418,
+      "learning_rate": 6.510211086322293e-06,
+      "loss": 0.6038,
+      "step": 9125
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.9276123046875,
+      "learning_rate": 6.507636862879698e-06,
+      "loss": 0.4792,
+      "step": 9126
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.834636688232422,
+      "learning_rate": 6.505062639437103e-06,
+      "loss": 0.5706,
+      "step": 9127
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.11631965637207,
+      "learning_rate": 6.502488415994509e-06,
+      "loss": 0.4871,
+      "step": 9128
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.8363213539123535,
+      "learning_rate": 6.499914192551913e-06,
+      "loss": 0.3815,
+      "step": 9129
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.0677266120910645,
+      "learning_rate": 6.497339969109319e-06,
+      "loss": 0.2796,
+      "step": 9130
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.577725410461426,
+      "learning_rate": 6.494765745666724e-06,
+      "loss": 0.344,
+      "step": 9131
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.923539161682129,
+      "learning_rate": 6.49219152222413e-06,
+      "loss": 0.3185,
+      "step": 9132
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.376360893249512,
+      "learning_rate": 6.489617298781535e-06,
+      "loss": 0.3198,
+      "step": 9133
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.27021598815918,
+      "learning_rate": 6.487043075338939e-06,
+      "loss": 0.5073,
+      "step": 9134
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.934722900390625,
+      "learning_rate": 6.484468851896345e-06,
+      "loss": 0.4404,
+      "step": 9135
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.822287559509277,
+      "learning_rate": 6.48189462845375e-06,
+      "loss": 0.5774,
+      "step": 9136
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.939071655273438,
+      "learning_rate": 6.479320405011156e-06,
+      "loss": 0.2943,
+      "step": 9137
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.28674030303955,
+      "learning_rate": 6.476746181568561e-06,
+      "loss": 0.3376,
+      "step": 9138
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.918496131896973,
+      "learning_rate": 6.4741719581259656e-06,
+      "loss": 0.2892,
+      "step": 9139
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.129401206970215,
+      "learning_rate": 6.4715977346833706e-06,
+      "loss": 0.4588,
+      "step": 9140
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.066826820373535,
+      "learning_rate": 6.4690235112407755e-06,
+      "loss": 0.4128,
+      "step": 9141
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.561490058898926,
+      "learning_rate": 6.466449287798181e-06,
+      "loss": 0.4404,
+      "step": 9142
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.176942825317383,
+      "learning_rate": 6.463875064355586e-06,
+      "loss": 0.3879,
+      "step": 9143
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.307493209838867,
+      "learning_rate": 6.461300840912991e-06,
+      "loss": 0.4464,
+      "step": 9144
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.318291664123535,
+      "learning_rate": 6.458726617470396e-06,
+      "loss": 0.4403,
+      "step": 9145
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.240020751953125,
+      "learning_rate": 6.456152394027801e-06,
+      "loss": 0.4397,
+      "step": 9146
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.238934516906738,
+      "learning_rate": 6.453578170585207e-06,
+      "loss": 0.3925,
+      "step": 9147
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.540674209594727,
+      "learning_rate": 6.451003947142612e-06,
+      "loss": 0.2702,
+      "step": 9148
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.108983039855957,
+      "learning_rate": 6.448429723700017e-06,
+      "loss": 0.3493,
+      "step": 9149
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.288019180297852,
+      "learning_rate": 6.445855500257422e-06,
+      "loss": 0.5413,
+      "step": 9150
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.925826072692871,
+      "learning_rate": 6.443281276814828e-06,
+      "loss": 0.3367,
+      "step": 9151
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.938533306121826,
+      "learning_rate": 6.440707053372233e-06,
+      "loss": 0.3785,
+      "step": 9152
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.446700096130371,
+      "learning_rate": 6.438132829929638e-06,
+      "loss": 0.4267,
+      "step": 9153
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.088892936706543,
+      "learning_rate": 6.435558606487044e-06,
+      "loss": 0.477,
+      "step": 9154
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.959019660949707,
+      "learning_rate": 6.432984383044448e-06,
+      "loss": 0.3541,
+      "step": 9155
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.720050811767578,
+      "learning_rate": 6.430410159601854e-06,
+      "loss": 0.4112,
+      "step": 9156
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.589910507202148,
+      "learning_rate": 6.427835936159259e-06,
+      "loss": 0.2835,
+      "step": 9157
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.143238067626953,
+      "learning_rate": 6.4252617127166646e-06,
+      "loss": 0.4491,
+      "step": 9158
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.869386672973633,
+      "learning_rate": 6.4226874892740696e-06,
+      "loss": 0.6889,
+      "step": 9159
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.199326515197754,
+      "learning_rate": 6.420113265831474e-06,
+      "loss": 0.34,
+      "step": 9160
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.466827392578125,
+      "learning_rate": 6.4175390423888795e-06,
+      "loss": 0.4387,
+      "step": 9161
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.478174209594727,
+      "learning_rate": 6.4149648189462845e-06,
+      "loss": 0.4016,
+      "step": 9162
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.79170036315918,
+      "learning_rate": 6.41239059550369e-06,
+      "loss": 0.3401,
+      "step": 9163
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.144485473632812,
+      "learning_rate": 6.409816372061095e-06,
+      "loss": 0.2904,
+      "step": 9164
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.949034214019775,
+      "learning_rate": 6.4072421486185e-06,
+      "loss": 0.3288,
+      "step": 9165
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.652872085571289,
+      "learning_rate": 6.404667925175905e-06,
+      "loss": 0.5183,
+      "step": 9166
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.029740333557129,
+      "learning_rate": 6.40209370173331e-06,
+      "loss": 0.4428,
+      "step": 9167
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.10626220703125,
+      "learning_rate": 6.399519478290716e-06,
+      "loss": 0.4905,
+      "step": 9168
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.636055946350098,
+      "learning_rate": 6.396945254848121e-06,
+      "loss": 0.3283,
+      "step": 9169
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.930211067199707,
+      "learning_rate": 6.394371031405526e-06,
+      "loss": 0.5027,
+      "step": 9170
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.66115665435791,
+      "learning_rate": 6.391796807962931e-06,
+      "loss": 0.5509,
+      "step": 9171
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.8271050453186035,
+      "learning_rate": 6.389222584520336e-06,
+      "loss": 0.3067,
+      "step": 9172
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.62724781036377,
+      "learning_rate": 6.386648361077742e-06,
+      "loss": 0.4757,
+      "step": 9173
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.797287940979004,
+      "learning_rate": 6.384074137635147e-06,
+      "loss": 0.3841,
+      "step": 9174
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.340659141540527,
+      "learning_rate": 6.381499914192552e-06,
+      "loss": 0.3596,
+      "step": 9175
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.375767707824707,
+      "learning_rate": 6.378925690749957e-06,
+      "loss": 0.4161,
+      "step": 9176
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.873882293701172,
+      "learning_rate": 6.376351467307363e-06,
+      "loss": 0.5284,
+      "step": 9177
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.78491497039795,
+      "learning_rate": 6.373777243864768e-06,
+      "loss": 0.2966,
+      "step": 9178
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.241011619567871,
+      "learning_rate": 6.371203020422173e-06,
+      "loss": 0.3889,
+      "step": 9179
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.602612495422363,
+      "learning_rate": 6.3686287969795785e-06,
+      "loss": 0.4154,
+      "step": 9180
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.44935131072998,
+      "learning_rate": 6.366054573536983e-06,
+      "loss": 0.2783,
+      "step": 9181
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.485203742980957,
+      "learning_rate": 6.3634803500943885e-06,
+      "loss": 0.4032,
+      "step": 9182
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.090597152709961,
+      "learning_rate": 6.3609061266517935e-06,
+      "loss": 0.3144,
+      "step": 9183
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.790277004241943,
+      "learning_rate": 6.358331903209199e-06,
+      "loss": 0.3183,
+      "step": 9184
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.959728240966797,
+      "learning_rate": 6.355757679766604e-06,
+      "loss": 0.4813,
+      "step": 9185
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.528314590454102,
+      "learning_rate": 6.3531834563240084e-06,
+      "loss": 0.4418,
+      "step": 9186
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.280416488647461,
+      "learning_rate": 6.350609232881414e-06,
+      "loss": 0.3754,
+      "step": 9187
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.70815658569336,
+      "learning_rate": 6.348035009438819e-06,
+      "loss": 0.4866,
+      "step": 9188
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.094280242919922,
+      "learning_rate": 6.345460785996225e-06,
+      "loss": 0.4987,
+      "step": 9189
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.443658828735352,
+      "learning_rate": 6.34288656255363e-06,
+      "loss": 0.4744,
+      "step": 9190
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.062834739685059,
+      "learning_rate": 6.340312339111034e-06,
+      "loss": 0.4508,
+      "step": 9191
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.706581115722656,
+      "learning_rate": 6.33773811566844e-06,
+      "loss": 0.4386,
+      "step": 9192
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.26957893371582,
+      "learning_rate": 6.335163892225845e-06,
+      "loss": 0.401,
+      "step": 9193
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.478545188903809,
+      "learning_rate": 6.332589668783251e-06,
+      "loss": 0.4432,
+      "step": 9194
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.519157409667969,
+      "learning_rate": 6.330015445340656e-06,
+      "loss": 0.5666,
+      "step": 9195
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.237707138061523,
+      "learning_rate": 6.327441221898061e-06,
+      "loss": 0.464,
+      "step": 9196
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.189074516296387,
+      "learning_rate": 6.324866998455466e-06,
+      "loss": 0.3563,
+      "step": 9197
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.335261344909668,
+      "learning_rate": 6.322292775012871e-06,
+      "loss": 0.4736,
+      "step": 9198
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.836599349975586,
+      "learning_rate": 6.319718551570277e-06,
+      "loss": 0.3335,
+      "step": 9199
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.865789413452148,
+      "learning_rate": 6.317144328127682e-06,
+      "loss": 0.2711,
+      "step": 9200
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.779120445251465,
+      "learning_rate": 6.314570104685087e-06,
+      "loss": 0.4434,
+      "step": 9201
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.948101043701172,
+      "learning_rate": 6.311995881242492e-06,
+      "loss": 0.398,
+      "step": 9202
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.8967084884643555,
+      "learning_rate": 6.3094216577998975e-06,
+      "loss": 0.2803,
+      "step": 9203
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.806225776672363,
+      "learning_rate": 6.3068474343573025e-06,
+      "loss": 0.3499,
+      "step": 9204
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.917614936828613,
+      "learning_rate": 6.3042732109147074e-06,
+      "loss": 0.6543,
+      "step": 9205
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.861566543579102,
+      "learning_rate": 6.301698987472113e-06,
+      "loss": 0.394,
+      "step": 9206
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.328132629394531,
+      "learning_rate": 6.299124764029517e-06,
+      "loss": 0.3762,
+      "step": 9207
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.741944313049316,
+      "learning_rate": 6.296550540586923e-06,
+      "loss": 0.266,
+      "step": 9208
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.820136070251465,
+      "learning_rate": 6.293976317144328e-06,
+      "loss": 0.3392,
+      "step": 9209
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.18803882598877,
+      "learning_rate": 6.291402093701734e-06,
+      "loss": 0.3968,
+      "step": 9210
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.384278297424316,
+      "learning_rate": 6.288827870259139e-06,
+      "loss": 0.3269,
+      "step": 9211
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.6448140144348145,
+      "learning_rate": 6.286253646816543e-06,
+      "loss": 0.242,
+      "step": 9212
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.856287002563477,
+      "learning_rate": 6.283679423373949e-06,
+      "loss": 0.2546,
+      "step": 9213
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.850491523742676,
+      "learning_rate": 6.281105199931354e-06,
+      "loss": 0.3218,
+      "step": 9214
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.62674617767334,
+      "learning_rate": 6.27853097648876e-06,
+      "loss": 0.4675,
+      "step": 9215
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.44118595123291,
+      "learning_rate": 6.275956753046165e-06,
+      "loss": 0.4394,
+      "step": 9216
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.37971305847168,
+      "learning_rate": 6.273382529603569e-06,
+      "loss": 0.5001,
+      "step": 9217
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.927918434143066,
+      "learning_rate": 6.270808306160975e-06,
+      "loss": 0.4025,
+      "step": 9218
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.683401107788086,
+      "learning_rate": 6.26823408271838e-06,
+      "loss": 0.412,
+      "step": 9219
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.079011917114258,
+      "learning_rate": 6.265659859275786e-06,
+      "loss": 0.419,
+      "step": 9220
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.040717124938965,
+      "learning_rate": 6.263085635833191e-06,
+      "loss": 0.2575,
+      "step": 9221
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.044816970825195,
+      "learning_rate": 6.260511412390596e-06,
+      "loss": 0.2736,
+      "step": 9222
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.928059577941895,
+      "learning_rate": 6.257937188948001e-06,
+      "loss": 0.4892,
+      "step": 9223
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.411823272705078,
+      "learning_rate": 6.255362965505406e-06,
+      "loss": 0.6102,
+      "step": 9224
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.721311569213867,
+      "learning_rate": 6.252788742062811e-06,
+      "loss": 0.4807,
+      "step": 9225
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.812295913696289,
+      "learning_rate": 6.250214518620216e-06,
+      "loss": 0.5312,
+      "step": 9226
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.255032539367676,
+      "learning_rate": 6.247640295177621e-06,
+      "loss": 0.6739,
+      "step": 9227
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.508779525756836,
+      "learning_rate": 6.245066071735026e-06,
+      "loss": 0.5966,
+      "step": 9228
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.748486518859863,
+      "learning_rate": 6.242491848292432e-06,
+      "loss": 0.4089,
+      "step": 9229
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.84903335571289,
+      "learning_rate": 6.239917624849837e-06,
+      "loss": 0.5221,
+      "step": 9230
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.814446449279785,
+      "learning_rate": 6.237343401407242e-06,
+      "loss": 0.5019,
+      "step": 9231
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.281661033630371,
+      "learning_rate": 6.234769177964648e-06,
+      "loss": 0.5408,
+      "step": 9232
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.548513412475586,
+      "learning_rate": 6.232194954522052e-06,
+      "loss": 0.4943,
+      "step": 9233
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.791305541992188,
+      "learning_rate": 6.229620731079458e-06,
+      "loss": 0.4264,
+      "step": 9234
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.57929515838623,
+      "learning_rate": 6.227046507636863e-06,
+      "loss": 0.3723,
+      "step": 9235
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.174640655517578,
+      "learning_rate": 6.224472284194269e-06,
+      "loss": 0.4752,
+      "step": 9236
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.794071197509766,
+      "learning_rate": 6.221898060751674e-06,
+      "loss": 0.4902,
+      "step": 9237
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.173076629638672,
+      "learning_rate": 6.219323837309078e-06,
+      "loss": 0.3757,
+      "step": 9238
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.48410415649414,
+      "learning_rate": 6.216749613866484e-06,
+      "loss": 0.2799,
+      "step": 9239
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.611170291900635,
+      "learning_rate": 6.214175390423889e-06,
+      "loss": 0.3328,
+      "step": 9240
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.497977256774902,
+      "learning_rate": 6.211601166981295e-06,
+      "loss": 0.4005,
+      "step": 9241
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.058220863342285,
+      "learning_rate": 6.2090269435387e-06,
+      "loss": 0.3598,
+      "step": 9242
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.694328308105469,
+      "learning_rate": 6.206452720096104e-06,
+      "loss": 0.3787,
+      "step": 9243
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.689227104187012,
+      "learning_rate": 6.2038784966535096e-06,
+      "loss": 0.607,
+      "step": 9244
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.597221374511719,
+      "learning_rate": 6.2013042732109146e-06,
+      "loss": 0.3211,
+      "step": 9245
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.371646881103516,
+      "learning_rate": 6.19873004976832e-06,
+      "loss": 0.3632,
+      "step": 9246
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.772314071655273,
+      "learning_rate": 6.196155826325725e-06,
+      "loss": 0.5425,
+      "step": 9247
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.395363807678223,
+      "learning_rate": 6.19358160288313e-06,
+      "loss": 0.2751,
+      "step": 9248
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.440888404846191,
+      "learning_rate": 6.191007379440535e-06,
+      "loss": 0.622,
+      "step": 9249
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.794041633605957,
+      "learning_rate": 6.18843315599794e-06,
+      "loss": 0.3511,
+      "step": 9250
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.800326347351074,
+      "learning_rate": 6.185858932555346e-06,
+      "loss": 0.5923,
+      "step": 9251
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.731034278869629,
+      "learning_rate": 6.183284709112751e-06,
+      "loss": 0.5992,
+      "step": 9252
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.585034370422363,
+      "learning_rate": 6.180710485670156e-06,
+      "loss": 0.3608,
+      "step": 9253
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.079184532165527,
+      "learning_rate": 6.178136262227561e-06,
+      "loss": 0.5047,
+      "step": 9254
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.937750816345215,
+      "learning_rate": 6.175562038784967e-06,
+      "loss": 0.3699,
+      "step": 9255
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.716346740722656,
+      "learning_rate": 6.172987815342372e-06,
+      "loss": 0.4097,
+      "step": 9256
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.026440620422363,
+      "learning_rate": 6.170413591899777e-06,
+      "loss": 0.401,
+      "step": 9257
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.726183891296387,
+      "learning_rate": 6.167839368457183e-06,
+      "loss": 0.4434,
+      "step": 9258
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.557024955749512,
+      "learning_rate": 6.165265145014587e-06,
+      "loss": 0.5816,
+      "step": 9259
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.54609203338623,
+      "learning_rate": 6.162690921571993e-06,
+      "loss": 0.4758,
+      "step": 9260
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.337329864501953,
+      "learning_rate": 6.160116698129398e-06,
+      "loss": 0.624,
+      "step": 9261
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.918791770935059,
+      "learning_rate": 6.157542474686804e-06,
+      "loss": 0.2915,
+      "step": 9262
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.622333526611328,
+      "learning_rate": 6.1549682512442086e-06,
+      "loss": 0.4079,
+      "step": 9263
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.2561674118042,
+      "learning_rate": 6.152394027801613e-06,
+      "loss": 0.3809,
+      "step": 9264
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.55340576171875,
+      "learning_rate": 6.1498198043590185e-06,
+      "loss": 0.5374,
+      "step": 9265
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.96012020111084,
+      "learning_rate": 6.1472455809164235e-06,
+      "loss": 0.356,
+      "step": 9266
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.459707260131836,
+      "learning_rate": 6.144671357473829e-06,
+      "loss": 0.6339,
+      "step": 9267
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.842348098754883,
+      "learning_rate": 6.142097134031234e-06,
+      "loss": 0.2857,
+      "step": 9268
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.259499549865723,
+      "learning_rate": 6.1395229105886385e-06,
+      "loss": 0.3934,
+      "step": 9269
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.485591888427734,
+      "learning_rate": 6.136948687146044e-06,
+      "loss": 0.3819,
+      "step": 9270
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.002604961395264,
+      "learning_rate": 6.134374463703449e-06,
+      "loss": 0.2309,
+      "step": 9271
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.152010917663574,
+      "learning_rate": 6.131800240260855e-06,
+      "loss": 0.5052,
+      "step": 9272
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.924613952636719,
+      "learning_rate": 6.12922601681826e-06,
+      "loss": 0.5115,
+      "step": 9273
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.100393295288086,
+      "learning_rate": 6.126651793375665e-06,
+      "loss": 0.4541,
+      "step": 9274
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.020744323730469,
+      "learning_rate": 6.12407756993307e-06,
+      "loss": 0.5774,
+      "step": 9275
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.655982971191406,
+      "learning_rate": 6.121503346490475e-06,
+      "loss": 0.3284,
+      "step": 9276
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.302045822143555,
+      "learning_rate": 6.118929123047881e-06,
+      "loss": 0.2654,
+      "step": 9277
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.965720176696777,
+      "learning_rate": 6.116354899605286e-06,
+      "loss": 0.5395,
+      "step": 9278
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.560702323913574,
+      "learning_rate": 6.113780676162691e-06,
+      "loss": 0.218,
+      "step": 9279
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.798632621765137,
+      "learning_rate": 6.111206452720096e-06,
+      "loss": 0.3781,
+      "step": 9280
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.915528297424316,
+      "learning_rate": 6.108632229277502e-06,
+      "loss": 0.2927,
+      "step": 9281
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.718039512634277,
+      "learning_rate": 6.106058005834907e-06,
+      "loss": 0.3748,
+      "step": 9282
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.610981941223145,
+      "learning_rate": 6.103483782392312e-06,
+      "loss": 0.4156,
+      "step": 9283
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.602054595947266,
+      "learning_rate": 6.1009095589497175e-06,
+      "loss": 0.3729,
+      "step": 9284
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.089277267456055,
+      "learning_rate": 6.098335335507122e-06,
+      "loss": 0.3612,
+      "step": 9285
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.430428504943848,
+      "learning_rate": 6.0957611120645275e-06,
+      "loss": 0.5332,
+      "step": 9286
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.575677871704102,
+      "learning_rate": 6.0931868886219325e-06,
+      "loss": 0.3386,
+      "step": 9287
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.176405906677246,
+      "learning_rate": 6.090612665179338e-06,
+      "loss": 0.494,
+      "step": 9288
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.538809776306152,
+      "learning_rate": 6.088038441736743e-06,
+      "loss": 0.5416,
+      "step": 9289
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.619078636169434,
+      "learning_rate": 6.0854642182941475e-06,
+      "loss": 0.4912,
+      "step": 9290
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.25139045715332,
+      "learning_rate": 6.082889994851553e-06,
+      "loss": 0.3834,
+      "step": 9291
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.719689846038818,
+      "learning_rate": 6.080315771408958e-06,
+      "loss": 0.3456,
+      "step": 9292
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.243241310119629,
+      "learning_rate": 6.077741547966364e-06,
+      "loss": 0.5009,
+      "step": 9293
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.630461692810059,
+      "learning_rate": 6.075167324523769e-06,
+      "loss": 0.3576,
+      "step": 9294
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.423848152160645,
+      "learning_rate": 6.072593101081173e-06,
+      "loss": 0.5059,
+      "step": 9295
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.479129791259766,
+      "learning_rate": 6.070018877638579e-06,
+      "loss": 0.5272,
+      "step": 9296
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.272512435913086,
+      "learning_rate": 6.067444654195984e-06,
+      "loss": 0.3023,
+      "step": 9297
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.018547058105469,
+      "learning_rate": 6.06487043075339e-06,
+      "loss": 0.5643,
+      "step": 9298
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.156089782714844,
+      "learning_rate": 6.062296207310795e-06,
+      "loss": 0.3727,
+      "step": 9299
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.55975341796875,
+      "learning_rate": 6.0597219838682e-06,
+      "loss": 0.2611,
+      "step": 9300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.020785331726074,
+      "learning_rate": 6.057147760425605e-06,
+      "loss": 0.3822,
+      "step": 9301
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.141971588134766,
+      "learning_rate": 6.05457353698301e-06,
+      "loss": 0.5198,
+      "step": 9302
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.88416576385498,
+      "learning_rate": 6.051999313540416e-06,
+      "loss": 0.5505,
+      "step": 9303
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.901981353759766,
+      "learning_rate": 6.049425090097821e-06,
+      "loss": 0.2862,
+      "step": 9304
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.45460319519043,
+      "learning_rate": 6.046850866655226e-06,
+      "loss": 0.3459,
+      "step": 9305
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.849007606506348,
+      "learning_rate": 6.044276643212631e-06,
+      "loss": 0.3254,
+      "step": 9306
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.164941787719727,
+      "learning_rate": 6.0417024197700365e-06,
+      "loss": 0.5973,
+      "step": 9307
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.208996772766113,
+      "learning_rate": 6.0391281963274415e-06,
+      "loss": 0.4844,
+      "step": 9308
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.36326789855957,
+      "learning_rate": 6.0365539728848465e-06,
+      "loss": 0.3321,
+      "step": 9309
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.12881088256836,
+      "learning_rate": 6.033979749442252e-06,
+      "loss": 0.4278,
+      "step": 9310
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.57091999053955,
+      "learning_rate": 6.0314055259996564e-06,
+      "loss": 0.3726,
+      "step": 9311
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.636283874511719,
+      "learning_rate": 6.028831302557062e-06,
+      "loss": 0.392,
+      "step": 9312
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.948846817016602,
+      "learning_rate": 6.026257079114467e-06,
+      "loss": 0.2689,
+      "step": 9313
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.179314613342285,
+      "learning_rate": 6.023682855671873e-06,
+      "loss": 0.368,
+      "step": 9314
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.047548294067383,
+      "learning_rate": 6.021108632229278e-06,
+      "loss": 0.3608,
+      "step": 9315
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.82796859741211,
+      "learning_rate": 6.018534408786682e-06,
+      "loss": 0.353,
+      "step": 9316
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.916614532470703,
+      "learning_rate": 6.015960185344088e-06,
+      "loss": 0.4454,
+      "step": 9317
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.81261920928955,
+      "learning_rate": 6.013385961901493e-06,
+      "loss": 0.3534,
+      "step": 9318
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.931442260742188,
+      "learning_rate": 6.010811738458899e-06,
+      "loss": 0.4239,
+      "step": 9319
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.180985450744629,
+      "learning_rate": 6.008237515016304e-06,
+      "loss": 0.4094,
+      "step": 9320
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.366473197937012,
+      "learning_rate": 6.005663291573708e-06,
+      "loss": 0.6373,
+      "step": 9321
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.796895980834961,
+      "learning_rate": 6.003089068131114e-06,
+      "loss": 0.5091,
+      "step": 9322
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.969415664672852,
+      "learning_rate": 6.000514844688519e-06,
+      "loss": 0.444,
+      "step": 9323
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.278087615966797,
+      "learning_rate": 5.997940621245925e-06,
+      "loss": 0.2579,
+      "step": 9324
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.328569412231445,
+      "learning_rate": 5.99536639780333e-06,
+      "loss": 0.4803,
+      "step": 9325
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.177692413330078,
+      "learning_rate": 5.992792174360735e-06,
+      "loss": 0.4947,
+      "step": 9326
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.504905700683594,
+      "learning_rate": 5.99021795091814e-06,
+      "loss": 0.3134,
+      "step": 9327
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.254817008972168,
+      "learning_rate": 5.987643727475545e-06,
+      "loss": 0.4861,
+      "step": 9328
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.166810989379883,
+      "learning_rate": 5.9850695040329504e-06,
+      "loss": 0.363,
+      "step": 9329
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.391650676727295,
+      "learning_rate": 5.982495280590355e-06,
+      "loss": 0.2166,
+      "step": 9330
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.902772903442383,
+      "learning_rate": 5.97992105714776e-06,
+      "loss": 0.5765,
+      "step": 9331
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.624422073364258,
+      "learning_rate": 5.977346833705165e-06,
+      "loss": 0.4744,
+      "step": 9332
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.857745170593262,
+      "learning_rate": 5.974772610262571e-06,
+      "loss": 0.4948,
+      "step": 9333
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.925243377685547,
+      "learning_rate": 5.972198386819976e-06,
+      "loss": 0.3548,
+      "step": 9334
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.138570785522461,
+      "learning_rate": 5.969624163377381e-06,
+      "loss": 0.3301,
+      "step": 9335
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.094758033752441,
+      "learning_rate": 5.967049939934787e-06,
+      "loss": 0.4501,
+      "step": 9336
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.87204360961914,
+      "learning_rate": 5.964475716492191e-06,
+      "loss": 0.6357,
+      "step": 9337
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.850410461425781,
+      "learning_rate": 5.961901493049597e-06,
+      "loss": 0.4706,
+      "step": 9338
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.774003982543945,
+      "learning_rate": 5.959327269607002e-06,
+      "loss": 0.2983,
+      "step": 9339
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.635725021362305,
+      "learning_rate": 5.956753046164408e-06,
+      "loss": 0.4463,
+      "step": 9340
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.130509376525879,
+      "learning_rate": 5.954178822721813e-06,
+      "loss": 0.2974,
+      "step": 9341
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.487428665161133,
+      "learning_rate": 5.951604599279217e-06,
+      "loss": 0.3373,
+      "step": 9342
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.86618423461914,
+      "learning_rate": 5.949030375836623e-06,
+      "loss": 0.4209,
+      "step": 9343
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.71156120300293,
+      "learning_rate": 5.946456152394028e-06,
+      "loss": 0.4609,
+      "step": 9344
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.448272705078125,
+      "learning_rate": 5.943881928951434e-06,
+      "loss": 0.4087,
+      "step": 9345
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.952189445495605,
+      "learning_rate": 5.941307705508839e-06,
+      "loss": 0.5209,
+      "step": 9346
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.552382469177246,
+      "learning_rate": 5.938733482066243e-06,
+      "loss": 0.4547,
+      "step": 9347
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.495433807373047,
+      "learning_rate": 5.936159258623649e-06,
+      "loss": 0.4151,
+      "step": 9348
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.510540008544922,
+      "learning_rate": 5.9335850351810536e-06,
+      "loss": 0.4371,
+      "step": 9349
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.461602210998535,
+      "learning_rate": 5.931010811738459e-06,
+      "loss": 0.3226,
+      "step": 9350
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.195796966552734,
+      "learning_rate": 5.928436588295864e-06,
+      "loss": 0.3189,
+      "step": 9351
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.487024307250977,
+      "learning_rate": 5.925862364853269e-06,
+      "loss": 0.2571,
+      "step": 9352
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.922492027282715,
+      "learning_rate": 5.923288141410674e-06,
+      "loss": 0.5108,
+      "step": 9353
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.246350288391113,
+      "learning_rate": 5.920713917968079e-06,
+      "loss": 0.3148,
+      "step": 9354
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.257803916931152,
+      "learning_rate": 5.918139694525485e-06,
+      "loss": 0.4622,
+      "step": 9355
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.560223579406738,
+      "learning_rate": 5.91556547108289e-06,
+      "loss": 0.4447,
+      "step": 9356
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.579278945922852,
+      "learning_rate": 5.912991247640295e-06,
+      "loss": 0.2871,
+      "step": 9357
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.096439361572266,
+      "learning_rate": 5.9104170241977e-06,
+      "loss": 0.4621,
+      "step": 9358
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.200118064880371,
+      "learning_rate": 5.907842800755106e-06,
+      "loss": 0.5691,
+      "step": 9359
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.151054382324219,
+      "learning_rate": 5.905268577312511e-06,
+      "loss": 0.4102,
+      "step": 9360
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.757876396179199,
+      "learning_rate": 5.902694353869916e-06,
+      "loss": 0.257,
+      "step": 9361
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.452271461486816,
+      "learning_rate": 5.900120130427322e-06,
+      "loss": 0.3395,
+      "step": 9362
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.132145881652832,
+      "learning_rate": 5.897545906984726e-06,
+      "loss": 0.4416,
+      "step": 9363
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.626276969909668,
+      "learning_rate": 5.894971683542132e-06,
+      "loss": 0.3473,
+      "step": 9364
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.145108222961426,
+      "learning_rate": 5.892397460099537e-06,
+      "loss": 0.4639,
+      "step": 9365
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.048547744750977,
+      "learning_rate": 5.889823236656942e-06,
+      "loss": 0.3527,
+      "step": 9366
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.349409103393555,
+      "learning_rate": 5.8872490132143476e-06,
+      "loss": 0.6066,
+      "step": 9367
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.787160396575928,
+      "learning_rate": 5.884674789771752e-06,
+      "loss": 0.2924,
+      "step": 9368
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.691932678222656,
+      "learning_rate": 5.8821005663291576e-06,
+      "loss": 0.4399,
+      "step": 9369
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.421937942504883,
+      "learning_rate": 5.8795263428865625e-06,
+      "loss": 0.4559,
+      "step": 9370
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.742603302001953,
+      "learning_rate": 5.876952119443968e-06,
+      "loss": 0.642,
+      "step": 9371
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.500776290893555,
+      "learning_rate": 5.874377896001373e-06,
+      "loss": 0.3622,
+      "step": 9372
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.78892707824707,
+      "learning_rate": 5.8718036725587775e-06,
+      "loss": 0.5199,
+      "step": 9373
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.555265426635742,
+      "learning_rate": 5.869229449116183e-06,
+      "loss": 0.3516,
+      "step": 9374
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.721681594848633,
+      "learning_rate": 5.866655225673588e-06,
+      "loss": 0.3237,
+      "step": 9375
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.01641845703125,
+      "learning_rate": 5.864081002230994e-06,
+      "loss": 0.4826,
+      "step": 9376
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.306987762451172,
+      "learning_rate": 5.861506778788399e-06,
+      "loss": 0.2503,
+      "step": 9377
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.787321090698242,
+      "learning_rate": 5.858932555345804e-06,
+      "loss": 0.2964,
+      "step": 9378
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.207783699035645,
+      "learning_rate": 5.856358331903209e-06,
+      "loss": 0.4971,
+      "step": 9379
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.970616340637207,
+      "learning_rate": 5.853784108460614e-06,
+      "loss": 0.4826,
+      "step": 9380
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.678253173828125,
+      "learning_rate": 5.85120988501802e-06,
+      "loss": 0.3736,
+      "step": 9381
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.0572509765625,
+      "learning_rate": 5.848635661575425e-06,
+      "loss": 0.3381,
+      "step": 9382
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.030983924865723,
+      "learning_rate": 5.846061438132831e-06,
+      "loss": 0.3966,
+      "step": 9383
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.285091400146484,
+      "learning_rate": 5.843487214690235e-06,
+      "loss": 0.4965,
+      "step": 9384
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.010630130767822,
+      "learning_rate": 5.840912991247641e-06,
+      "loss": 0.2905,
+      "step": 9385
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.928468704223633,
+      "learning_rate": 5.838338767805046e-06,
+      "loss": 0.3296,
+      "step": 9386
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.42886734008789,
+      "learning_rate": 5.835764544362451e-06,
+      "loss": 0.332,
+      "step": 9387
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.908339500427246,
+      "learning_rate": 5.8331903209198565e-06,
+      "loss": 0.4273,
+      "step": 9388
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.1969499588012695,
+      "learning_rate": 5.830616097477261e-06,
+      "loss": 0.223,
+      "step": 9389
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.444570541381836,
+      "learning_rate": 5.8280418740346665e-06,
+      "loss": 0.3362,
+      "step": 9390
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.364608764648438,
+      "learning_rate": 5.8254676505920715e-06,
+      "loss": 0.3234,
+      "step": 9391
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.901269912719727,
+      "learning_rate": 5.8228934271494765e-06,
+      "loss": 0.5691,
+      "step": 9392
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.817156791687012,
+      "learning_rate": 5.820319203706882e-06,
+      "loss": 0.3869,
+      "step": 9393
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.694347858428955,
+      "learning_rate": 5.8177449802642865e-06,
+      "loss": 0.3568,
+      "step": 9394
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.901276588439941,
+      "learning_rate": 5.815170756821692e-06,
+      "loss": 0.3348,
+      "step": 9395
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.192615509033203,
+      "learning_rate": 5.812596533379097e-06,
+      "loss": 0.2995,
+      "step": 9396
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.507491111755371,
+      "learning_rate": 5.810022309936503e-06,
+      "loss": 0.3902,
+      "step": 9397
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.135817527770996,
+      "learning_rate": 5.807448086493908e-06,
+      "loss": 0.3596,
+      "step": 9398
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.679965019226074,
+      "learning_rate": 5.804873863051312e-06,
+      "loss": 0.4209,
+      "step": 9399
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.001195907592773,
+      "learning_rate": 5.802299639608718e-06,
+      "loss": 0.5544,
+      "step": 9400
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.406121253967285,
+      "learning_rate": 5.799725416166123e-06,
+      "loss": 0.4476,
+      "step": 9401
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.545846939086914,
+      "learning_rate": 5.797151192723529e-06,
+      "loss": 0.5148,
+      "step": 9402
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.235424041748047,
+      "learning_rate": 5.794576969280934e-06,
+      "loss": 0.4847,
+      "step": 9403
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.171060562133789,
+      "learning_rate": 5.792002745838339e-06,
+      "loss": 0.4245,
+      "step": 9404
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.361068725585938,
+      "learning_rate": 5.789428522395744e-06,
+      "loss": 0.401,
+      "step": 9405
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.301602363586426,
+      "learning_rate": 5.786854298953149e-06,
+      "loss": 0.4688,
+      "step": 9406
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.1406831741333,
+      "learning_rate": 5.784280075510555e-06,
+      "loss": 0.2977,
+      "step": 9407
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.393590450286865,
+      "learning_rate": 5.78170585206796e-06,
+      "loss": 0.2248,
+      "step": 9408
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.281673431396484,
+      "learning_rate": 5.7791316286253655e-06,
+      "loss": 0.3988,
+      "step": 9409
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.977659702301025,
+      "learning_rate": 5.77655740518277e-06,
+      "loss": 0.3915,
+      "step": 9410
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.32376480102539,
+      "learning_rate": 5.7739831817401755e-06,
+      "loss": 0.3113,
+      "step": 9411
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.357552528381348,
+      "learning_rate": 5.7714089582975805e-06,
+      "loss": 0.441,
+      "step": 9412
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.72784423828125,
+      "learning_rate": 5.7688347348549855e-06,
+      "loss": 0.4453,
+      "step": 9413
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.66362190246582,
+      "learning_rate": 5.766260511412391e-06,
+      "loss": 0.74,
+      "step": 9414
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.012140274047852,
+      "learning_rate": 5.7636862879697954e-06,
+      "loss": 0.5343,
+      "step": 9415
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.609058380126953,
+      "learning_rate": 5.761112064527201e-06,
+      "loss": 0.28,
+      "step": 9416
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.282861709594727,
+      "learning_rate": 5.758537841084606e-06,
+      "loss": 0.3836,
+      "step": 9417
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.726537704467773,
+      "learning_rate": 5.755963617642011e-06,
+      "loss": 0.4459,
+      "step": 9418
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.606436729431152,
+      "learning_rate": 5.753389394199417e-06,
+      "loss": 0.3538,
+      "step": 9419
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.751702308654785,
+      "learning_rate": 5.750815170756821e-06,
+      "loss": 0.4457,
+      "step": 9420
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.104438781738281,
+      "learning_rate": 5.748240947314227e-06,
+      "loss": 0.6976,
+      "step": 9421
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.153799057006836,
+      "learning_rate": 5.745666723871632e-06,
+      "loss": 0.5936,
+      "step": 9422
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.309557914733887,
+      "learning_rate": 5.743092500429038e-06,
+      "loss": 0.3608,
+      "step": 9423
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.956354141235352,
+      "learning_rate": 5.740518276986443e-06,
+      "loss": 0.4579,
+      "step": 9424
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.102630615234375,
+      "learning_rate": 5.737944053543847e-06,
+      "loss": 0.5007,
+      "step": 9425
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.239076614379883,
+      "learning_rate": 5.735369830101253e-06,
+      "loss": 0.5125,
+      "step": 9426
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.7650980949401855,
+      "learning_rate": 5.732795606658658e-06,
+      "loss": 0.36,
+      "step": 9427
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.341635704040527,
+      "learning_rate": 5.730221383216064e-06,
+      "loss": 0.2929,
+      "step": 9428
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.563714027404785,
+      "learning_rate": 5.727647159773469e-06,
+      "loss": 0.3478,
+      "step": 9429
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.534550666809082,
+      "learning_rate": 5.725072936330874e-06,
+      "loss": 0.417,
+      "step": 9430
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.148838996887207,
+      "learning_rate": 5.722498712888279e-06,
+      "loss": 0.3054,
+      "step": 9431
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.314359664916992,
+      "learning_rate": 5.719924489445684e-06,
+      "loss": 0.3541,
+      "step": 9432
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.9740309715271,
+      "learning_rate": 5.7173502660030894e-06,
+      "loss": 0.2978,
+      "step": 9433
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.401704788208008,
+      "learning_rate": 5.7147760425604944e-06,
+      "loss": 0.4636,
+      "step": 9434
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.361881256103516,
+      "learning_rate": 5.7122018191179e-06,
+      "loss": 0.5099,
+      "step": 9435
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.762758731842041,
+      "learning_rate": 5.709627595675304e-06,
+      "loss": 0.2905,
+      "step": 9436
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.460415840148926,
+      "learning_rate": 5.70705337223271e-06,
+      "loss": 0.3647,
+      "step": 9437
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.269363403320312,
+      "learning_rate": 5.704479148790115e-06,
+      "loss": 0.5,
+      "step": 9438
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.00335693359375,
+      "learning_rate": 5.70190492534752e-06,
+      "loss": 0.3988,
+      "step": 9439
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.040388107299805,
+      "learning_rate": 5.699330701904926e-06,
+      "loss": 0.3993,
+      "step": 9440
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.470085144042969,
+      "learning_rate": 5.69675647846233e-06,
+      "loss": 0.2645,
+      "step": 9441
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.192621231079102,
+      "learning_rate": 5.694182255019736e-06,
+      "loss": 0.5402,
+      "step": 9442
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.485733985900879,
+      "learning_rate": 5.691608031577141e-06,
+      "loss": 0.3496,
+      "step": 9443
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.314024925231934,
+      "learning_rate": 5.689033808134546e-06,
+      "loss": 0.4641,
+      "step": 9444
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.139507293701172,
+      "learning_rate": 5.686459584691952e-06,
+      "loss": 0.4792,
+      "step": 9445
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.422876358032227,
+      "learning_rate": 5.683885361249356e-06,
+      "loss": 0.3295,
+      "step": 9446
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.380331039428711,
+      "learning_rate": 5.681311137806762e-06,
+      "loss": 0.4889,
+      "step": 9447
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.798163414001465,
+      "learning_rate": 5.678736914364167e-06,
+      "loss": 0.4406,
+      "step": 9448
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.419535636901855,
+      "learning_rate": 5.676162690921573e-06,
+      "loss": 0.3113,
+      "step": 9449
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.418760299682617,
+      "learning_rate": 5.673588467478978e-06,
+      "loss": 0.4324,
+      "step": 9450
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.205583572387695,
+      "learning_rate": 5.671014244036382e-06,
+      "loss": 0.3619,
+      "step": 9451
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.42657995223999,
+      "learning_rate": 5.668440020593788e-06,
+      "loss": 0.3224,
+      "step": 9452
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.461694717407227,
+      "learning_rate": 5.665865797151193e-06,
+      "loss": 0.4596,
+      "step": 9453
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.489125728607178,
+      "learning_rate": 5.663291573708598e-06,
+      "loss": 0.2557,
+      "step": 9454
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.991674423217773,
+      "learning_rate": 5.660717350266003e-06,
+      "loss": 0.3223,
+      "step": 9455
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.101436614990234,
+      "learning_rate": 5.658143126823408e-06,
+      "loss": 0.2953,
+      "step": 9456
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.007814407348633,
+      "learning_rate": 5.655568903380813e-06,
+      "loss": 0.3529,
+      "step": 9457
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.626026153564453,
+      "learning_rate": 5.652994679938218e-06,
+      "loss": 0.4621,
+      "step": 9458
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.057463645935059,
+      "learning_rate": 5.650420456495624e-06,
+      "loss": 0.326,
+      "step": 9459
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.18494987487793,
+      "learning_rate": 5.647846233053029e-06,
+      "loss": 0.4238,
+      "step": 9460
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.687639236450195,
+      "learning_rate": 5.645272009610435e-06,
+      "loss": 0.4175,
+      "step": 9461
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.35632610321045,
+      "learning_rate": 5.642697786167839e-06,
+      "loss": 0.3192,
+      "step": 9462
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.568696975708008,
+      "learning_rate": 5.640123562725245e-06,
+      "loss": 0.7296,
+      "step": 9463
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.392620086669922,
+      "learning_rate": 5.63754933928265e-06,
+      "loss": 0.2002,
+      "step": 9464
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 24.698711395263672,
+      "learning_rate": 5.634975115840055e-06,
+      "loss": 0.4289,
+      "step": 9465
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.210901260375977,
+      "learning_rate": 5.632400892397461e-06,
+      "loss": 0.3849,
+      "step": 9466
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.081639289855957,
+      "learning_rate": 5.629826668954865e-06,
+      "loss": 0.4971,
+      "step": 9467
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.283191680908203,
+      "learning_rate": 5.627252445512271e-06,
+      "loss": 0.4543,
+      "step": 9468
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.628883361816406,
+      "learning_rate": 5.624678222069676e-06,
+      "loss": 0.3403,
+      "step": 9469
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.887421607971191,
+      "learning_rate": 5.622103998627081e-06,
+      "loss": 0.4559,
+      "step": 9470
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.605327606201172,
+      "learning_rate": 5.619529775184487e-06,
+      "loss": 0.4851,
+      "step": 9471
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.49120569229126,
+      "learning_rate": 5.616955551741891e-06,
+      "loss": 0.2355,
+      "step": 9472
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.835000991821289,
+      "learning_rate": 5.6143813282992966e-06,
+      "loss": 0.5566,
+      "step": 9473
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.666558265686035,
+      "learning_rate": 5.6118071048567015e-06,
+      "loss": 0.2818,
+      "step": 9474
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.961830139160156,
+      "learning_rate": 5.609232881414107e-06,
+      "loss": 0.4205,
+      "step": 9475
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.174569129943848,
+      "learning_rate": 5.606658657971512e-06,
+      "loss": 0.5047,
+      "step": 9476
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.219046592712402,
+      "learning_rate": 5.6040844345289165e-06,
+      "loss": 0.355,
+      "step": 9477
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.74807357788086,
+      "learning_rate": 5.601510211086322e-06,
+      "loss": 0.412,
+      "step": 9478
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.126022338867188,
+      "learning_rate": 5.598935987643727e-06,
+      "loss": 0.3001,
+      "step": 9479
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.293107032775879,
+      "learning_rate": 5.596361764201133e-06,
+      "loss": 0.5309,
+      "step": 9480
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.276771545410156,
+      "learning_rate": 5.593787540758538e-06,
+      "loss": 0.3959,
+      "step": 9481
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.59830379486084,
+      "learning_rate": 5.591213317315943e-06,
+      "loss": 0.3412,
+      "step": 9482
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.927278995513916,
+      "learning_rate": 5.588639093873348e-06,
+      "loss": 0.3523,
+      "step": 9483
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.802326202392578,
+      "learning_rate": 5.586064870430753e-06,
+      "loss": 0.2781,
+      "step": 9484
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.646251678466797,
+      "learning_rate": 5.583490646988159e-06,
+      "loss": 0.3949,
+      "step": 9485
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.38243579864502,
+      "learning_rate": 5.580916423545564e-06,
+      "loss": 0.5674,
+      "step": 9486
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.333597183227539,
+      "learning_rate": 5.57834220010297e-06,
+      "loss": 0.403,
+      "step": 9487
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.630289077758789,
+      "learning_rate": 5.575767976660374e-06,
+      "loss": 0.5945,
+      "step": 9488
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.838348388671875,
+      "learning_rate": 5.57319375321778e-06,
+      "loss": 0.3882,
+      "step": 9489
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.97502613067627,
+      "learning_rate": 5.570619529775185e-06,
+      "loss": 0.4292,
+      "step": 9490
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.213041305541992,
+      "learning_rate": 5.56804530633259e-06,
+      "loss": 0.6812,
+      "step": 9491
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.27556324005127,
+      "learning_rate": 5.5654710828899956e-06,
+      "loss": 0.3498,
+      "step": 9492
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.039551734924316,
+      "learning_rate": 5.5628968594474e-06,
+      "loss": 0.4898,
+      "step": 9493
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.654840469360352,
+      "learning_rate": 5.5603226360048055e-06,
+      "loss": 0.4408,
+      "step": 9494
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.076395034790039,
+      "learning_rate": 5.5577484125622105e-06,
+      "loss": 0.5059,
+      "step": 9495
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.76390552520752,
+      "learning_rate": 5.5551741891196155e-06,
+      "loss": 0.4985,
+      "step": 9496
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.457161903381348,
+      "learning_rate": 5.552599965677021e-06,
+      "loss": 0.4021,
+      "step": 9497
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.285161972045898,
+      "learning_rate": 5.5500257422344255e-06,
+      "loss": 0.2897,
+      "step": 9498
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.88736343383789,
+      "learning_rate": 5.547451518791831e-06,
+      "loss": 0.5343,
+      "step": 9499
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.263001441955566,
+      "learning_rate": 5.544877295349236e-06,
+      "loss": 0.4679,
+      "step": 9500
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.066070556640625,
+      "learning_rate": 5.542303071906642e-06,
+      "loss": 0.4338,
+      "step": 9501
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.497499465942383,
+      "learning_rate": 5.539728848464047e-06,
+      "loss": 0.4305,
+      "step": 9502
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.847160339355469,
+      "learning_rate": 5.537154625021451e-06,
+      "loss": 0.3502,
+      "step": 9503
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.71319055557251,
+      "learning_rate": 5.534580401578857e-06,
+      "loss": 0.3201,
+      "step": 9504
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.295954704284668,
+      "learning_rate": 5.532006178136262e-06,
+      "loss": 0.3941,
+      "step": 9505
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.266483306884766,
+      "learning_rate": 5.529431954693668e-06,
+      "loss": 0.3982,
+      "step": 9506
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.09190845489502,
+      "learning_rate": 5.526857731251073e-06,
+      "loss": 0.4783,
+      "step": 9507
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.584628105163574,
+      "learning_rate": 5.524283507808478e-06,
+      "loss": 0.4744,
+      "step": 9508
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.742584228515625,
+      "learning_rate": 5.521709284365883e-06,
+      "loss": 0.514,
+      "step": 9509
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.745132446289062,
+      "learning_rate": 5.519135060923288e-06,
+      "loss": 0.2851,
+      "step": 9510
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.359766960144043,
+      "learning_rate": 5.516560837480694e-06,
+      "loss": 0.4297,
+      "step": 9511
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.780399322509766,
+      "learning_rate": 5.513986614038099e-06,
+      "loss": 0.6842,
+      "step": 9512
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.671772003173828,
+      "learning_rate": 5.5114123905955045e-06,
+      "loss": 0.3953,
+      "step": 9513
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.546340942382812,
+      "learning_rate": 5.508838167152909e-06,
+      "loss": 0.483,
+      "step": 9514
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.07878589630127,
+      "learning_rate": 5.5062639437103145e-06,
+      "loss": 0.4839,
+      "step": 9515
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.804040908813477,
+      "learning_rate": 5.5036897202677195e-06,
+      "loss": 0.3769,
+      "step": 9516
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.916842937469482,
+      "learning_rate": 5.5011154968251245e-06,
+      "loss": 0.3174,
+      "step": 9517
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.750128746032715,
+      "learning_rate": 5.49854127338253e-06,
+      "loss": 0.3253,
+      "step": 9518
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.293014526367188,
+      "learning_rate": 5.4959670499399345e-06,
+      "loss": 0.2398,
+      "step": 9519
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.516221046447754,
+      "learning_rate": 5.49339282649734e-06,
+      "loss": 0.2682,
+      "step": 9520
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.01073169708252,
+      "learning_rate": 5.490818603054745e-06,
+      "loss": 0.4094,
+      "step": 9521
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.900822639465332,
+      "learning_rate": 5.48824437961215e-06,
+      "loss": 0.3751,
+      "step": 9522
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.678243637084961,
+      "learning_rate": 5.485670156169556e-06,
+      "loss": 0.393,
+      "step": 9523
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.451507568359375,
+      "learning_rate": 5.48309593272696e-06,
+      "loss": 0.5546,
+      "step": 9524
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.416790962219238,
+      "learning_rate": 5.480521709284366e-06,
+      "loss": 0.3119,
+      "step": 9525
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.565122604370117,
+      "learning_rate": 5.477947485841771e-06,
+      "loss": 0.4713,
+      "step": 9526
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.840173721313477,
+      "learning_rate": 5.475373262399177e-06,
+      "loss": 0.4917,
+      "step": 9527
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.61516284942627,
+      "learning_rate": 5.472799038956582e-06,
+      "loss": 0.5948,
+      "step": 9528
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.036795616149902,
+      "learning_rate": 5.470224815513986e-06,
+      "loss": 0.4526,
+      "step": 9529
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.842708587646484,
+      "learning_rate": 5.467650592071392e-06,
+      "loss": 0.4411,
+      "step": 9530
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.228228569030762,
+      "learning_rate": 5.465076368628797e-06,
+      "loss": 0.3512,
+      "step": 9531
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.198923110961914,
+      "learning_rate": 5.462502145186203e-06,
+      "loss": 0.4442,
+      "step": 9532
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.91535472869873,
+      "learning_rate": 5.459927921743608e-06,
+      "loss": 0.3202,
+      "step": 9533
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.445051193237305,
+      "learning_rate": 5.457353698301013e-06,
+      "loss": 0.3955,
+      "step": 9534
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.360581398010254,
+      "learning_rate": 5.454779474858418e-06,
+      "loss": 0.3812,
+      "step": 9535
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.984142303466797,
+      "learning_rate": 5.452205251415823e-06,
+      "loss": 0.4418,
+      "step": 9536
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.357352256774902,
+      "learning_rate": 5.4496310279732285e-06,
+      "loss": 0.4037,
+      "step": 9537
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.403053283691406,
+      "learning_rate": 5.4470568045306334e-06,
+      "loss": 0.476,
+      "step": 9538
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.646989822387695,
+      "learning_rate": 5.444482581088039e-06,
+      "loss": 0.3284,
+      "step": 9539
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.550874710083008,
+      "learning_rate": 5.441908357645443e-06,
+      "loss": 0.6351,
+      "step": 9540
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.214956283569336,
+      "learning_rate": 5.439334134202848e-06,
+      "loss": 0.3966,
+      "step": 9541
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.978095054626465,
+      "learning_rate": 5.436759910760254e-06,
+      "loss": 0.3507,
+      "step": 9542
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.20533275604248,
+      "learning_rate": 5.434185687317659e-06,
+      "loss": 0.2897,
+      "step": 9543
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.751108169555664,
+      "learning_rate": 5.431611463875065e-06,
+      "loss": 0.4487,
+      "step": 9544
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.129634857177734,
+      "learning_rate": 5.429037240432469e-06,
+      "loss": 0.3483,
+      "step": 9545
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.402081489562988,
+      "learning_rate": 5.426463016989875e-06,
+      "loss": 0.3781,
+      "step": 9546
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.624606132507324,
+      "learning_rate": 5.42388879354728e-06,
+      "loss": 0.5113,
+      "step": 9547
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.68389892578125,
+      "learning_rate": 5.421314570104685e-06,
+      "loss": 0.633,
+      "step": 9548
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.133748054504395,
+      "learning_rate": 5.418740346662091e-06,
+      "loss": 0.5521,
+      "step": 9549
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.241744995117188,
+      "learning_rate": 5.416166123219495e-06,
+      "loss": 0.3083,
+      "step": 9550
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.651822090148926,
+      "learning_rate": 5.413591899776901e-06,
+      "loss": 0.5847,
+      "step": 9551
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.413772583007812,
+      "learning_rate": 5.411017676334306e-06,
+      "loss": 0.4548,
+      "step": 9552
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.035812377929688,
+      "learning_rate": 5.408443452891712e-06,
+      "loss": 0.444,
+      "step": 9553
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.18057632446289,
+      "learning_rate": 5.405869229449117e-06,
+      "loss": 0.5344,
+      "step": 9554
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.828645706176758,
+      "learning_rate": 5.403295006006521e-06,
+      "loss": 0.3964,
+      "step": 9555
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.996955871582031,
+      "learning_rate": 5.400720782563927e-06,
+      "loss": 0.3469,
+      "step": 9556
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.058783531188965,
+      "learning_rate": 5.398146559121332e-06,
+      "loss": 0.3878,
+      "step": 9557
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.217575073242188,
+      "learning_rate": 5.3955723356787374e-06,
+      "loss": 0.4707,
+      "step": 9558
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.300097465515137,
+      "learning_rate": 5.392998112236142e-06,
+      "loss": 0.2818,
+      "step": 9559
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.75239372253418,
+      "learning_rate": 5.390423888793547e-06,
+      "loss": 0.3886,
+      "step": 9560
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.548704147338867,
+      "learning_rate": 5.387849665350952e-06,
+      "loss": 0.2526,
+      "step": 9561
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.713324546813965,
+      "learning_rate": 5.385275441908357e-06,
+      "loss": 0.2906,
+      "step": 9562
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.408305168151855,
+      "learning_rate": 5.382701218465763e-06,
+      "loss": 0.3239,
+      "step": 9563
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.87882137298584,
+      "learning_rate": 5.380126995023168e-06,
+      "loss": 0.5022,
+      "step": 9564
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.475255966186523,
+      "learning_rate": 5.377552771580574e-06,
+      "loss": 0.3187,
+      "step": 9565
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.214444160461426,
+      "learning_rate": 5.374978548137978e-06,
+      "loss": 0.4406,
+      "step": 9566
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.222140312194824,
+      "learning_rate": 5.372404324695383e-06,
+      "loss": 0.4956,
+      "step": 9567
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.52293872833252,
+      "learning_rate": 5.369830101252789e-06,
+      "loss": 0.4043,
+      "step": 9568
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.071854591369629,
+      "learning_rate": 5.367255877810194e-06,
+      "loss": 0.4452,
+      "step": 9569
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.6990203857421875,
+      "learning_rate": 5.3646816543676e-06,
+      "loss": 0.2953,
+      "step": 9570
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.847419738769531,
+      "learning_rate": 5.362107430925004e-06,
+      "loss": 0.4816,
+      "step": 9571
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.557892799377441,
+      "learning_rate": 5.35953320748241e-06,
+      "loss": 0.2918,
+      "step": 9572
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.088288307189941,
+      "learning_rate": 5.356958984039815e-06,
+      "loss": 0.5144,
+      "step": 9573
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.070624351501465,
+      "learning_rate": 5.35438476059722e-06,
+      "loss": 0.4365,
+      "step": 9574
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.944831848144531,
+      "learning_rate": 5.351810537154626e-06,
+      "loss": 0.342,
+      "step": 9575
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.573887825012207,
+      "learning_rate": 5.34923631371203e-06,
+      "loss": 0.3637,
+      "step": 9576
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.548113822937012,
+      "learning_rate": 5.3466620902694356e-06,
+      "loss": 0.5225,
+      "step": 9577
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.11074447631836,
+      "learning_rate": 5.3440878668268406e-06,
+      "loss": 0.2917,
+      "step": 9578
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.624791145324707,
+      "learning_rate": 5.341513643384246e-06,
+      "loss": 0.4344,
+      "step": 9579
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.08116340637207,
+      "learning_rate": 5.338939419941651e-06,
+      "loss": 0.3729,
+      "step": 9580
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.875580787658691,
+      "learning_rate": 5.3363651964990555e-06,
+      "loss": 0.3392,
+      "step": 9581
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.49691104888916,
+      "learning_rate": 5.333790973056461e-06,
+      "loss": 0.3364,
+      "step": 9582
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.008417129516602,
+      "learning_rate": 5.331216749613866e-06,
+      "loss": 0.3941,
+      "step": 9583
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.848654747009277,
+      "learning_rate": 5.328642526171272e-06,
+      "loss": 0.4664,
+      "step": 9584
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.821887016296387,
+      "learning_rate": 5.326068302728677e-06,
+      "loss": 0.3384,
+      "step": 9585
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.210831642150879,
+      "learning_rate": 5.323494079286082e-06,
+      "loss": 0.5148,
+      "step": 9586
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.727355003356934,
+      "learning_rate": 5.320919855843487e-06,
+      "loss": 0.3022,
+      "step": 9587
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.036104202270508,
+      "learning_rate": 5.318345632400892e-06,
+      "loss": 0.4292,
+      "step": 9588
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.513479232788086,
+      "learning_rate": 5.315771408958298e-06,
+      "loss": 0.3374,
+      "step": 9589
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.694671630859375,
+      "learning_rate": 5.313197185515703e-06,
+      "loss": 0.5588,
+      "step": 9590
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.200736999511719,
+      "learning_rate": 5.310622962073109e-06,
+      "loss": 0.4175,
+      "step": 9591
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.50564193725586,
+      "learning_rate": 5.308048738630513e-06,
+      "loss": 0.3543,
+      "step": 9592
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.963687896728516,
+      "learning_rate": 5.305474515187918e-06,
+      "loss": 0.4715,
+      "step": 9593
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.498435974121094,
+      "learning_rate": 5.302900291745324e-06,
+      "loss": 0.3656,
+      "step": 9594
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.893860816955566,
+      "learning_rate": 5.300326068302729e-06,
+      "loss": 0.4367,
+      "step": 9595
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.980158805847168,
+      "learning_rate": 5.2977518448601346e-06,
+      "loss": 0.6962,
+      "step": 9596
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.001141548156738,
+      "learning_rate": 5.295177621417539e-06,
+      "loss": 0.3928,
+      "step": 9597
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.483039855957031,
+      "learning_rate": 5.2926033979749445e-06,
+      "loss": 0.2953,
+      "step": 9598
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.969145774841309,
+      "learning_rate": 5.2900291745323495e-06,
+      "loss": 0.3239,
+      "step": 9599
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.261040687561035,
+      "learning_rate": 5.2874549510897545e-06,
+      "loss": 0.3371,
+      "step": 9600
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.073018074035645,
+      "learning_rate": 5.28488072764716e-06,
+      "loss": 0.3655,
+      "step": 9601
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.37199878692627,
+      "learning_rate": 5.2823065042045645e-06,
+      "loss": 0.4765,
+      "step": 9602
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.612890243530273,
+      "learning_rate": 5.27973228076197e-06,
+      "loss": 0.4252,
+      "step": 9603
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.203518867492676,
+      "learning_rate": 5.277158057319375e-06,
+      "loss": 0.2959,
+      "step": 9604
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.584474563598633,
+      "learning_rate": 5.274583833876781e-06,
+      "loss": 0.2578,
+      "step": 9605
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.519556045532227,
+      "learning_rate": 5.272009610434186e-06,
+      "loss": 0.3959,
+      "step": 9606
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.884371757507324,
+      "learning_rate": 5.26943538699159e-06,
+      "loss": 0.4247,
+      "step": 9607
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.644059181213379,
+      "learning_rate": 5.266861163548996e-06,
+      "loss": 0.5127,
+      "step": 9608
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.072440147399902,
+      "learning_rate": 5.264286940106401e-06,
+      "loss": 0.4754,
+      "step": 9609
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.92575740814209,
+      "learning_rate": 5.261712716663807e-06,
+      "loss": 0.4107,
+      "step": 9610
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.279706001281738,
+      "learning_rate": 5.259138493221212e-06,
+      "loss": 0.3781,
+      "step": 9611
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.406563758850098,
+      "learning_rate": 5.256564269778617e-06,
+      "loss": 0.5921,
+      "step": 9612
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.601865768432617,
+      "learning_rate": 5.253990046336022e-06,
+      "loss": 0.334,
+      "step": 9613
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.434653282165527,
+      "learning_rate": 5.251415822893427e-06,
+      "loss": 0.365,
+      "step": 9614
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.051100254058838,
+      "learning_rate": 5.248841599450833e-06,
+      "loss": 0.2711,
+      "step": 9615
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.551592826843262,
+      "learning_rate": 5.246267376008238e-06,
+      "loss": 0.4609,
+      "step": 9616
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.5572452545166,
+      "learning_rate": 5.2436931525656435e-06,
+      "loss": 0.4668,
+      "step": 9617
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.892151832580566,
+      "learning_rate": 5.241118929123048e-06,
+      "loss": 0.3163,
+      "step": 9618
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.698821067810059,
+      "learning_rate": 5.238544705680453e-06,
+      "loss": 0.5468,
+      "step": 9619
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.577296257019043,
+      "learning_rate": 5.2359704822378585e-06,
+      "loss": 0.3295,
+      "step": 9620
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.120394706726074,
+      "learning_rate": 5.2333962587952635e-06,
+      "loss": 0.5822,
+      "step": 9621
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.945082664489746,
+      "learning_rate": 5.230822035352669e-06,
+      "loss": 0.3464,
+      "step": 9622
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.520858764648438,
+      "learning_rate": 5.2282478119100735e-06,
+      "loss": 0.5536,
+      "step": 9623
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.958598136901855,
+      "learning_rate": 5.225673588467479e-06,
+      "loss": 0.4022,
+      "step": 9624
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.269060134887695,
+      "learning_rate": 5.223099365024884e-06,
+      "loss": 0.3177,
+      "step": 9625
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.411349296569824,
+      "learning_rate": 5.220525141582289e-06,
+      "loss": 0.4978,
+      "step": 9626
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.53354263305664,
+      "learning_rate": 5.217950918139695e-06,
+      "loss": 0.5876,
+      "step": 9627
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.472549438476562,
+      "learning_rate": 5.215376694697099e-06,
+      "loss": 0.4486,
+      "step": 9628
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.45311164855957,
+      "learning_rate": 5.212802471254505e-06,
+      "loss": 0.5712,
+      "step": 9629
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.941349983215332,
+      "learning_rate": 5.21022824781191e-06,
+      "loss": 0.4403,
+      "step": 9630
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.188843727111816,
+      "learning_rate": 5.207654024369316e-06,
+      "loss": 0.4363,
+      "step": 9631
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.22233772277832,
+      "learning_rate": 5.205079800926721e-06,
+      "loss": 0.4265,
+      "step": 9632
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.215051651000977,
+      "learning_rate": 5.202505577484125e-06,
+      "loss": 0.2918,
+      "step": 9633
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.18620491027832,
+      "learning_rate": 5.199931354041531e-06,
+      "loss": 0.5269,
+      "step": 9634
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.463600158691406,
+      "learning_rate": 5.197357130598936e-06,
+      "loss": 0.4948,
+      "step": 9635
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.704997062683105,
+      "learning_rate": 5.194782907156342e-06,
+      "loss": 0.4368,
+      "step": 9636
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.63602352142334,
+      "learning_rate": 5.192208683713747e-06,
+      "loss": 0.3814,
+      "step": 9637
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.676228046417236,
+      "learning_rate": 5.189634460271152e-06,
+      "loss": 0.2348,
+      "step": 9638
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.812514305114746,
+      "learning_rate": 5.187060236828557e-06,
+      "loss": 0.3691,
+      "step": 9639
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.405218124389648,
+      "learning_rate": 5.184486013385962e-06,
+      "loss": 0.3215,
+      "step": 9640
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.067056655883789,
+      "learning_rate": 5.1819117899433675e-06,
+      "loss": 0.2929,
+      "step": 9641
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.108912467956543,
+      "learning_rate": 5.1793375665007725e-06,
+      "loss": 0.6292,
+      "step": 9642
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.607660293579102,
+      "learning_rate": 5.176763343058178e-06,
+      "loss": 0.5454,
+      "step": 9643
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.48611831665039,
+      "learning_rate": 5.1741891196155824e-06,
+      "loss": 0.4604,
+      "step": 9644
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.380729675292969,
+      "learning_rate": 5.171614896172987e-06,
+      "loss": 0.4349,
+      "step": 9645
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.142107963562012,
+      "learning_rate": 5.169040672730393e-06,
+      "loss": 0.5217,
+      "step": 9646
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.150938987731934,
+      "learning_rate": 5.166466449287798e-06,
+      "loss": 0.2636,
+      "step": 9647
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.137070655822754,
+      "learning_rate": 5.163892225845204e-06,
+      "loss": 0.3405,
+      "step": 9648
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.729690551757812,
+      "learning_rate": 5.161318002402608e-06,
+      "loss": 0.5361,
+      "step": 9649
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.038084983825684,
+      "learning_rate": 5.158743778960014e-06,
+      "loss": 0.3174,
+      "step": 9650
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.0342378616333,
+      "learning_rate": 5.156169555517419e-06,
+      "loss": 0.3105,
+      "step": 9651
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.702507972717285,
+      "learning_rate": 5.153595332074824e-06,
+      "loss": 0.5836,
+      "step": 9652
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.81276798248291,
+      "learning_rate": 5.15102110863223e-06,
+      "loss": 0.2117,
+      "step": 9653
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.668899536132812,
+      "learning_rate": 5.148446885189634e-06,
+      "loss": 0.6862,
+      "step": 9654
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.84668254852295,
+      "learning_rate": 5.14587266174704e-06,
+      "loss": 0.4756,
+      "step": 9655
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.694463729858398,
+      "learning_rate": 5.143298438304445e-06,
+      "loss": 0.2502,
+      "step": 9656
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.033036231994629,
+      "learning_rate": 5.140724214861851e-06,
+      "loss": 0.3887,
+      "step": 9657
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.068087577819824,
+      "learning_rate": 5.138149991419256e-06,
+      "loss": 0.399,
+      "step": 9658
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.252431869506836,
+      "learning_rate": 5.13557576797666e-06,
+      "loss": 0.4944,
+      "step": 9659
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.663421630859375,
+      "learning_rate": 5.133001544534066e-06,
+      "loss": 0.3324,
+      "step": 9660
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.257155418395996,
+      "learning_rate": 5.130427321091471e-06,
+      "loss": 0.3699,
+      "step": 9661
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.055291175842285,
+      "learning_rate": 5.1278530976488764e-06,
+      "loss": 0.4766,
+      "step": 9662
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.363739967346191,
+      "learning_rate": 5.1252788742062814e-06,
+      "loss": 0.5042,
+      "step": 9663
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.342509269714355,
+      "learning_rate": 5.122704650763686e-06,
+      "loss": 0.3707,
+      "step": 9664
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.173884391784668,
+      "learning_rate": 5.120130427321091e-06,
+      "loss": 0.4334,
+      "step": 9665
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.110319137573242,
+      "learning_rate": 5.117556203878496e-06,
+      "loss": 0.3685,
+      "step": 9666
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.817700386047363,
+      "learning_rate": 5.114981980435902e-06,
+      "loss": 0.3278,
+      "step": 9667
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.422295570373535,
+      "learning_rate": 5.112407756993307e-06,
+      "loss": 0.3149,
+      "step": 9668
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.302181243896484,
+      "learning_rate": 5.109833533550713e-06,
+      "loss": 0.4029,
+      "step": 9669
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.484495162963867,
+      "learning_rate": 5.107259310108117e-06,
+      "loss": 0.4563,
+      "step": 9670
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.984222412109375,
+      "learning_rate": 5.104685086665522e-06,
+      "loss": 0.3982,
+      "step": 9671
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.562557220458984,
+      "learning_rate": 5.102110863222928e-06,
+      "loss": 0.2225,
+      "step": 9672
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.58154296875,
+      "learning_rate": 5.099536639780333e-06,
+      "loss": 0.5037,
+      "step": 9673
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.111811637878418,
+      "learning_rate": 5.096962416337739e-06,
+      "loss": 0.4773,
+      "step": 9674
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.926980972290039,
+      "learning_rate": 5.094388192895143e-06,
+      "loss": 0.5246,
+      "step": 9675
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.595392227172852,
+      "learning_rate": 5.091813969452549e-06,
+      "loss": 0.361,
+      "step": 9676
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.566325187683105,
+      "learning_rate": 5.089239746009954e-06,
+      "loss": 0.4003,
+      "step": 9677
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.54730224609375,
+      "learning_rate": 5.086665522567359e-06,
+      "loss": 0.4316,
+      "step": 9678
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.703022003173828,
+      "learning_rate": 5.084091299124765e-06,
+      "loss": 0.2642,
+      "step": 9679
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.015544891357422,
+      "learning_rate": 5.081517075682169e-06,
+      "loss": 0.4576,
+      "step": 9680
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.491514205932617,
+      "learning_rate": 5.078942852239575e-06,
+      "loss": 0.2702,
+      "step": 9681
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.422765731811523,
+      "learning_rate": 5.0763686287969796e-06,
+      "loss": 0.3372,
+      "step": 9682
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.247221946716309,
+      "learning_rate": 5.073794405354385e-06,
+      "loss": 0.4937,
+      "step": 9683
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.32010269165039,
+      "learning_rate": 5.07122018191179e-06,
+      "loss": 0.5839,
+      "step": 9684
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.346248626708984,
+      "learning_rate": 5.0686459584691945e-06,
+      "loss": 0.4014,
+      "step": 9685
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.298344612121582,
+      "learning_rate": 5.0660717350266e-06,
+      "loss": 0.3451,
+      "step": 9686
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.95216941833496,
+      "learning_rate": 5.063497511584005e-06,
+      "loss": 0.5503,
+      "step": 9687
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.768074035644531,
+      "learning_rate": 5.060923288141411e-06,
+      "loss": 0.3055,
+      "step": 9688
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.51509952545166,
+      "learning_rate": 5.058349064698816e-06,
+      "loss": 0.3556,
+      "step": 9689
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.47543716430664,
+      "learning_rate": 5.055774841256221e-06,
+      "loss": 0.3333,
+      "step": 9690
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.400741577148438,
+      "learning_rate": 5.053200617813626e-06,
+      "loss": 0.4665,
+      "step": 9691
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.153923034667969,
+      "learning_rate": 5.050626394371031e-06,
+      "loss": 0.4699,
+      "step": 9692
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.132309913635254,
+      "learning_rate": 5.048052170928437e-06,
+      "loss": 0.3326,
+      "step": 9693
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.555898666381836,
+      "learning_rate": 5.045477947485842e-06,
+      "loss": 0.5747,
+      "step": 9694
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.92475700378418,
+      "learning_rate": 5.042903724043248e-06,
+      "loss": 0.386,
+      "step": 9695
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.048136711120605,
+      "learning_rate": 5.040329500600652e-06,
+      "loss": 0.4085,
+      "step": 9696
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.683354377746582,
+      "learning_rate": 5.037755277158057e-06,
+      "loss": 0.4891,
+      "step": 9697
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.044388771057129,
+      "learning_rate": 5.035181053715463e-06,
+      "loss": 0.5794,
+      "step": 9698
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.953500747680664,
+      "learning_rate": 5.032606830272868e-06,
+      "loss": 0.3842,
+      "step": 9699
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.28654956817627,
+      "learning_rate": 5.030032606830274e-06,
+      "loss": 0.3827,
+      "step": 9700
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.839948654174805,
+      "learning_rate": 5.027458383387678e-06,
+      "loss": 0.4443,
+      "step": 9701
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.030762672424316,
+      "learning_rate": 5.0248841599450836e-06,
+      "loss": 0.2864,
+      "step": 9702
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.66850471496582,
+      "learning_rate": 5.0223099365024885e-06,
+      "loss": 0.454,
+      "step": 9703
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.376481056213379,
+      "learning_rate": 5.0197357130598935e-06,
+      "loss": 0.4531,
+      "step": 9704
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.105647087097168,
+      "learning_rate": 5.017161489617299e-06,
+      "loss": 0.3653,
+      "step": 9705
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.068383693695068,
+      "learning_rate": 5.0145872661747035e-06,
+      "loss": 0.3109,
+      "step": 9706
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.775964736938477,
+      "learning_rate": 5.012013042732109e-06,
+      "loss": 0.4885,
+      "step": 9707
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.013874053955078,
+      "learning_rate": 5.009438819289514e-06,
+      "loss": 0.4333,
+      "step": 9708
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.835769653320312,
+      "learning_rate": 5.00686459584692e-06,
+      "loss": 0.4808,
+      "step": 9709
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.118885040283203,
+      "learning_rate": 5.004290372404325e-06,
+      "loss": 0.3596,
+      "step": 9710
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.221210479736328,
+      "learning_rate": 5.001716148961729e-06,
+      "loss": 0.4878,
+      "step": 9711
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.434014320373535,
+      "learning_rate": 4.999141925519135e-06,
+      "loss": 0.3823,
+      "step": 9712
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.95885944366455,
+      "learning_rate": 4.99656770207654e-06,
+      "loss": 0.344,
+      "step": 9713
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.053287506103516,
+      "learning_rate": 4.993993478633946e-06,
+      "loss": 0.4008,
+      "step": 9714
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.965164184570312,
+      "learning_rate": 4.991419255191351e-06,
+      "loss": 0.4148,
+      "step": 9715
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.994836807250977,
+      "learning_rate": 4.988845031748756e-06,
+      "loss": 0.2957,
+      "step": 9716
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.769638061523438,
+      "learning_rate": 4.986270808306161e-06,
+      "loss": 0.3409,
+      "step": 9717
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.909829139709473,
+      "learning_rate": 4.983696584863566e-06,
+      "loss": 0.5229,
+      "step": 9718
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.90284252166748,
+      "learning_rate": 4.981122361420972e-06,
+      "loss": 0.3628,
+      "step": 9719
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.65898609161377,
+      "learning_rate": 4.978548137978377e-06,
+      "loss": 0.563,
+      "step": 9720
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.9395856857299805,
+      "learning_rate": 4.9759739145357825e-06,
+      "loss": 0.4056,
+      "step": 9721
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.578889846801758,
+      "learning_rate": 4.973399691093187e-06,
+      "loss": 0.4731,
+      "step": 9722
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.687485694885254,
+      "learning_rate": 4.970825467650592e-06,
+      "loss": 0.5302,
+      "step": 9723
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.489301681518555,
+      "learning_rate": 4.9682512442079975e-06,
+      "loss": 0.5465,
+      "step": 9724
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.20189094543457,
+      "learning_rate": 4.9656770207654025e-06,
+      "loss": 0.3079,
+      "step": 9725
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.46342945098877,
+      "learning_rate": 4.963102797322808e-06,
+      "loss": 0.5348,
+      "step": 9726
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.839994430541992,
+      "learning_rate": 4.9605285738802125e-06,
+      "loss": 0.6541,
+      "step": 9727
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.444540500640869,
+      "learning_rate": 4.957954350437618e-06,
+      "loss": 0.2223,
+      "step": 9728
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.409849166870117,
+      "learning_rate": 4.955380126995023e-06,
+      "loss": 0.3626,
+      "step": 9729
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.897289276123047,
+      "learning_rate": 4.952805903552428e-06,
+      "loss": 0.3453,
+      "step": 9730
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.75826644897461,
+      "learning_rate": 4.950231680109834e-06,
+      "loss": 0.4334,
+      "step": 9731
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.613908767700195,
+      "learning_rate": 4.947657456667238e-06,
+      "loss": 0.4228,
+      "step": 9732
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.719470977783203,
+      "learning_rate": 4.945083233224644e-06,
+      "loss": 0.427,
+      "step": 9733
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.687437057495117,
+      "learning_rate": 4.942509009782049e-06,
+      "loss": 0.3567,
+      "step": 9734
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.149904251098633,
+      "learning_rate": 4.939934786339455e-06,
+      "loss": 0.4149,
+      "step": 9735
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.584178924560547,
+      "learning_rate": 4.93736056289686e-06,
+      "loss": 0.5128,
+      "step": 9736
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.671844482421875,
+      "learning_rate": 4.934786339454264e-06,
+      "loss": 0.5259,
+      "step": 9737
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.277400016784668,
+      "learning_rate": 4.93221211601167e-06,
+      "loss": 0.599,
+      "step": 9738
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.343419075012207,
+      "learning_rate": 4.929637892569075e-06,
+      "loss": 0.3067,
+      "step": 9739
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.915791511535645,
+      "learning_rate": 4.927063669126481e-06,
+      "loss": 0.485,
+      "step": 9740
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.148524284362793,
+      "learning_rate": 4.924489445683886e-06,
+      "loss": 0.4279,
+      "step": 9741
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.541462898254395,
+      "learning_rate": 4.921915222241291e-06,
+      "loss": 0.3187,
+      "step": 9742
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.05405330657959,
+      "learning_rate": 4.919340998798696e-06,
+      "loss": 0.3563,
+      "step": 9743
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.602696418762207,
+      "learning_rate": 4.916766775356101e-06,
+      "loss": 0.3998,
+      "step": 9744
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.657196044921875,
+      "learning_rate": 4.9141925519135065e-06,
+      "loss": 0.3731,
+      "step": 9745
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.555832862854004,
+      "learning_rate": 4.9116183284709115e-06,
+      "loss": 0.3885,
+      "step": 9746
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.744253635406494,
+      "learning_rate": 4.909044105028317e-06,
+      "loss": 0.364,
+      "step": 9747
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.683523178100586,
+      "learning_rate": 4.9064698815857214e-06,
+      "loss": 0.2569,
+      "step": 9748
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.121380805969238,
+      "learning_rate": 4.9038956581431264e-06,
+      "loss": 0.3385,
+      "step": 9749
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.50162410736084,
+      "learning_rate": 4.901321434700532e-06,
+      "loss": 0.365,
+      "step": 9750
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.024792671203613,
+      "learning_rate": 4.898747211257937e-06,
+      "loss": 0.5786,
+      "step": 9751
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.351205825805664,
+      "learning_rate": 4.896172987815343e-06,
+      "loss": 0.4454,
+      "step": 9752
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.305006980895996,
+      "learning_rate": 4.893598764372747e-06,
+      "loss": 0.3918,
+      "step": 9753
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.149331092834473,
+      "learning_rate": 4.891024540930153e-06,
+      "loss": 0.3914,
+      "step": 9754
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.778213500976562,
+      "learning_rate": 4.888450317487558e-06,
+      "loss": 0.3568,
+      "step": 9755
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.530014038085938,
+      "learning_rate": 4.885876094044963e-06,
+      "loss": 0.5542,
+      "step": 9756
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.421553134918213,
+      "learning_rate": 4.883301870602369e-06,
+      "loss": 0.4059,
+      "step": 9757
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.967906951904297,
+      "learning_rate": 4.880727647159773e-06,
+      "loss": 0.4492,
+      "step": 9758
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.925548553466797,
+      "learning_rate": 4.878153423717179e-06,
+      "loss": 0.4873,
+      "step": 9759
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.683004379272461,
+      "learning_rate": 4.875579200274584e-06,
+      "loss": 0.5783,
+      "step": 9760
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.583817481994629,
+      "learning_rate": 4.87300497683199e-06,
+      "loss": 0.4882,
+      "step": 9761
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.585066795349121,
+      "learning_rate": 4.870430753389395e-06,
+      "loss": 0.5634,
+      "step": 9762
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.594355583190918,
+      "learning_rate": 4.867856529946799e-06,
+      "loss": 0.2008,
+      "step": 9763
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.9620943069458,
+      "learning_rate": 4.865282306504205e-06,
+      "loss": 0.3649,
+      "step": 9764
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.050984382629395,
+      "learning_rate": 4.86270808306161e-06,
+      "loss": 0.5158,
+      "step": 9765
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.703680038452148,
+      "learning_rate": 4.8601338596190155e-06,
+      "loss": 0.3964,
+      "step": 9766
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.653826713562012,
+      "learning_rate": 4.8575596361764204e-06,
+      "loss": 0.4932,
+      "step": 9767
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.42916488647461,
+      "learning_rate": 4.8549854127338254e-06,
+      "loss": 0.4299,
+      "step": 9768
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.416485786437988,
+      "learning_rate": 4.85241118929123e-06,
+      "loss": 0.2953,
+      "step": 9769
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.955766677856445,
+      "learning_rate": 4.849836965848635e-06,
+      "loss": 0.1855,
+      "step": 9770
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.285853385925293,
+      "learning_rate": 4.847262742406041e-06,
+      "loss": 0.4909,
+      "step": 9771
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.007976531982422,
+      "learning_rate": 4.844688518963446e-06,
+      "loss": 0.3296,
+      "step": 9772
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.120858192443848,
+      "learning_rate": 4.842114295520852e-06,
+      "loss": 0.4508,
+      "step": 9773
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.275014877319336,
+      "learning_rate": 4.839540072078256e-06,
+      "loss": 0.3586,
+      "step": 9774
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.946040153503418,
+      "learning_rate": 4.836965848635661e-06,
+      "loss": 0.3457,
+      "step": 9775
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.76372241973877,
+      "learning_rate": 4.834391625193067e-06,
+      "loss": 0.3692,
+      "step": 9776
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.383515357971191,
+      "learning_rate": 4.831817401750472e-06,
+      "loss": 0.3021,
+      "step": 9777
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.727526664733887,
+      "learning_rate": 4.829243178307878e-06,
+      "loss": 0.4489,
+      "step": 9778
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.16355323791504,
+      "learning_rate": 4.826668954865282e-06,
+      "loss": 0.5248,
+      "step": 9779
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.519400596618652,
+      "learning_rate": 4.824094731422688e-06,
+      "loss": 0.487,
+      "step": 9780
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.917197227478027,
+      "learning_rate": 4.821520507980093e-06,
+      "loss": 0.5149,
+      "step": 9781
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.641584873199463,
+      "learning_rate": 4.818946284537498e-06,
+      "loss": 0.3047,
+      "step": 9782
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.241837501525879,
+      "learning_rate": 4.816372061094904e-06,
+      "loss": 0.4307,
+      "step": 9783
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.343546867370605,
+      "learning_rate": 4.813797837652308e-06,
+      "loss": 0.365,
+      "step": 9784
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.740839004516602,
+      "learning_rate": 4.811223614209714e-06,
+      "loss": 0.326,
+      "step": 9785
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.622611045837402,
+      "learning_rate": 4.808649390767119e-06,
+      "loss": 0.4491,
+      "step": 9786
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.138028144836426,
+      "learning_rate": 4.806075167324524e-06,
+      "loss": 0.2401,
+      "step": 9787
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.242894172668457,
+      "learning_rate": 4.803500943881929e-06,
+      "loss": 0.5856,
+      "step": 9788
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.222790718078613,
+      "learning_rate": 4.8009267204393335e-06,
+      "loss": 0.3135,
+      "step": 9789
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.781777381896973,
+      "learning_rate": 4.798352496996739e-06,
+      "loss": 0.4274,
+      "step": 9790
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.9878411293029785,
+      "learning_rate": 4.795778273554144e-06,
+      "loss": 0.363,
+      "step": 9791
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.4044771194458,
+      "learning_rate": 4.79320405011155e-06,
+      "loss": 0.4434,
+      "step": 9792
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.178708076477051,
+      "learning_rate": 4.790629826668955e-06,
+      "loss": 0.3797,
+      "step": 9793
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.6813325881958,
+      "learning_rate": 4.78805560322636e-06,
+      "loss": 0.5054,
+      "step": 9794
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.982411861419678,
+      "learning_rate": 4.785481379783765e-06,
+      "loss": 0.2052,
+      "step": 9795
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.61693000793457,
+      "learning_rate": 4.78290715634117e-06,
+      "loss": 0.5132,
+      "step": 9796
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.12075138092041,
+      "learning_rate": 4.780332932898576e-06,
+      "loss": 0.2804,
+      "step": 9797
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.588252067565918,
+      "learning_rate": 4.777758709455981e-06,
+      "loss": 0.3441,
+      "step": 9798
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.717166900634766,
+      "learning_rate": 4.775184486013387e-06,
+      "loss": 0.4569,
+      "step": 9799
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.314568519592285,
+      "learning_rate": 4.772610262570791e-06,
+      "loss": 0.3885,
+      "step": 9800
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.739814758300781,
+      "learning_rate": 4.770036039128196e-06,
+      "loss": 0.3648,
+      "step": 9801
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.777087211608887,
+      "learning_rate": 4.767461815685602e-06,
+      "loss": 0.3712,
+      "step": 9802
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.593404769897461,
+      "learning_rate": 4.764887592243007e-06,
+      "loss": 0.3444,
+      "step": 9803
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.855645179748535,
+      "learning_rate": 4.762313368800413e-06,
+      "loss": 0.5348,
+      "step": 9804
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.688872337341309,
+      "learning_rate": 4.759739145357817e-06,
+      "loss": 0.5544,
+      "step": 9805
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.552645206451416,
+      "learning_rate": 4.7571649219152226e-06,
+      "loss": 0.224,
+      "step": 9806
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.194353103637695,
+      "learning_rate": 4.7545906984726276e-06,
+      "loss": 0.4566,
+      "step": 9807
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.01422643661499,
+      "learning_rate": 4.7520164750300325e-06,
+      "loss": 0.3212,
+      "step": 9808
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.690058708190918,
+      "learning_rate": 4.749442251587438e-06,
+      "loss": 0.4046,
+      "step": 9809
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.333771705627441,
+      "learning_rate": 4.7468680281448425e-06,
+      "loss": 0.3044,
+      "step": 9810
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.825360298156738,
+      "learning_rate": 4.744293804702248e-06,
+      "loss": 0.4733,
+      "step": 9811
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.755243301391602,
+      "learning_rate": 4.741719581259653e-06,
+      "loss": 0.4944,
+      "step": 9812
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.844978332519531,
+      "learning_rate": 4.739145357817059e-06,
+      "loss": 0.2402,
+      "step": 9813
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.713988304138184,
+      "learning_rate": 4.736571134374464e-06,
+      "loss": 0.3748,
+      "step": 9814
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.40971851348877,
+      "learning_rate": 4.733996910931868e-06,
+      "loss": 0.4246,
+      "step": 9815
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.48250961303711,
+      "learning_rate": 4.731422687489274e-06,
+      "loss": 0.4136,
+      "step": 9816
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.333423614501953,
+      "learning_rate": 4.728848464046679e-06,
+      "loss": 0.3594,
+      "step": 9817
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.316245079040527,
+      "learning_rate": 4.726274240604085e-06,
+      "loss": 0.3243,
+      "step": 9818
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.806426048278809,
+      "learning_rate": 4.72370001716149e-06,
+      "loss": 0.367,
+      "step": 9819
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.316032409667969,
+      "learning_rate": 4.721125793718895e-06,
+      "loss": 0.2928,
+      "step": 9820
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.782794952392578,
+      "learning_rate": 4.7185515702763e-06,
+      "loss": 0.4303,
+      "step": 9821
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.911745071411133,
+      "learning_rate": 4.715977346833705e-06,
+      "loss": 0.3829,
+      "step": 9822
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.591330528259277,
+      "learning_rate": 4.713403123391111e-06,
+      "loss": 0.4084,
+      "step": 9823
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.675152778625488,
+      "learning_rate": 4.710828899948516e-06,
+      "loss": 0.3473,
+      "step": 9824
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.265535354614258,
+      "learning_rate": 4.7082546765059216e-06,
+      "loss": 0.477,
+      "step": 9825
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.301252365112305,
+      "learning_rate": 4.705680453063326e-06,
+      "loss": 0.352,
+      "step": 9826
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.527870178222656,
+      "learning_rate": 4.703106229620731e-06,
+      "loss": 0.3314,
+      "step": 9827
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.516618728637695,
+      "learning_rate": 4.7005320061781365e-06,
+      "loss": 0.2983,
+      "step": 9828
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.598899841308594,
+      "learning_rate": 4.6979577827355415e-06,
+      "loss": 0.3594,
+      "step": 9829
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.302850723266602,
+      "learning_rate": 4.695383559292947e-06,
+      "loss": 0.3796,
+      "step": 9830
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.374128341674805,
+      "learning_rate": 4.6928093358503515e-06,
+      "loss": 0.3827,
+      "step": 9831
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.87947940826416,
+      "learning_rate": 4.690235112407757e-06,
+      "loss": 0.4863,
+      "step": 9832
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.927107810974121,
+      "learning_rate": 4.687660888965162e-06,
+      "loss": 0.3641,
+      "step": 9833
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.19281005859375,
+      "learning_rate": 4.685086665522567e-06,
+      "loss": 0.4519,
+      "step": 9834
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.90243911743164,
+      "learning_rate": 4.682512442079973e-06,
+      "loss": 0.3695,
+      "step": 9835
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.137521743774414,
+      "learning_rate": 4.679938218637377e-06,
+      "loss": 0.5407,
+      "step": 9836
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.40953254699707,
+      "learning_rate": 4.677363995194783e-06,
+      "loss": 0.3313,
+      "step": 9837
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.945343017578125,
+      "learning_rate": 4.674789771752188e-06,
+      "loss": 0.4024,
+      "step": 9838
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.178277969360352,
+      "learning_rate": 4.672215548309594e-06,
+      "loss": 0.4536,
+      "step": 9839
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.213882446289062,
+      "learning_rate": 4.669641324866999e-06,
+      "loss": 0.3644,
+      "step": 9840
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.61800765991211,
+      "learning_rate": 4.667067101424403e-06,
+      "loss": 0.5098,
+      "step": 9841
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.695042610168457,
+      "learning_rate": 4.664492877981809e-06,
+      "loss": 0.3005,
+      "step": 9842
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.714210510253906,
+      "learning_rate": 4.661918654539214e-06,
+      "loss": 0.4691,
+      "step": 9843
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.154513359069824,
+      "learning_rate": 4.65934443109662e-06,
+      "loss": 0.3923,
+      "step": 9844
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.203897476196289,
+      "learning_rate": 4.656770207654025e-06,
+      "loss": 0.4449,
+      "step": 9845
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.143186569213867,
+      "learning_rate": 4.65419598421143e-06,
+      "loss": 0.4072,
+      "step": 9846
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.76897144317627,
+      "learning_rate": 4.651621760768835e-06,
+      "loss": 0.2984,
+      "step": 9847
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.279026985168457,
+      "learning_rate": 4.64904753732624e-06,
+      "loss": 0.317,
+      "step": 9848
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.268970489501953,
+      "learning_rate": 4.6464733138836455e-06,
+      "loss": 0.5681,
+      "step": 9849
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.448714256286621,
+      "learning_rate": 4.6438990904410505e-06,
+      "loss": 0.3394,
+      "step": 9850
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.98882007598877,
+      "learning_rate": 4.641324866998456e-06,
+      "loss": 0.4784,
+      "step": 9851
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.86368465423584,
+      "learning_rate": 4.6387506435558605e-06,
+      "loss": 0.4373,
+      "step": 9852
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.203152656555176,
+      "learning_rate": 4.6361764201132654e-06,
+      "loss": 0.3715,
+      "step": 9853
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.811062812805176,
+      "learning_rate": 4.633602196670671e-06,
+      "loss": 0.3231,
+      "step": 9854
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.87142562866211,
+      "learning_rate": 4.631027973228076e-06,
+      "loss": 0.2538,
+      "step": 9855
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.872397422790527,
+      "learning_rate": 4.628453749785482e-06,
+      "loss": 0.3488,
+      "step": 9856
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.994199752807617,
+      "learning_rate": 4.625879526342886e-06,
+      "loss": 0.4921,
+      "step": 9857
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.76842212677002,
+      "learning_rate": 4.623305302900292e-06,
+      "loss": 0.4806,
+      "step": 9858
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.93956470489502,
+      "learning_rate": 4.620731079457697e-06,
+      "loss": 0.5517,
+      "step": 9859
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.051068305969238,
+      "learning_rate": 4.618156856015102e-06,
+      "loss": 0.463,
+      "step": 9860
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.394959449768066,
+      "learning_rate": 4.615582632572508e-06,
+      "loss": 0.3313,
+      "step": 9861
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.845617294311523,
+      "learning_rate": 4.613008409129912e-06,
+      "loss": 0.5067,
+      "step": 9862
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.060649871826172,
+      "learning_rate": 4.610434185687318e-06,
+      "loss": 0.296,
+      "step": 9863
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.202380180358887,
+      "learning_rate": 4.607859962244723e-06,
+      "loss": 0.2721,
+      "step": 9864
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.583398818969727,
+      "learning_rate": 4.605285738802129e-06,
+      "loss": 0.5201,
+      "step": 9865
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.528497695922852,
+      "learning_rate": 4.602711515359534e-06,
+      "loss": 0.3013,
+      "step": 9866
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.059798240661621,
+      "learning_rate": 4.600137291916938e-06,
+      "loss": 0.4016,
+      "step": 9867
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.830427646636963,
+      "learning_rate": 4.597563068474344e-06,
+      "loss": 0.1739,
+      "step": 9868
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.108186721801758,
+      "learning_rate": 4.594988845031749e-06,
+      "loss": 0.3077,
+      "step": 9869
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.28027057647705,
+      "learning_rate": 4.5924146215891545e-06,
+      "loss": 0.3203,
+      "step": 9870
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.209270477294922,
+      "learning_rate": 4.5898403981465594e-06,
+      "loss": 0.3556,
+      "step": 9871
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.611172676086426,
+      "learning_rate": 4.5872661747039644e-06,
+      "loss": 0.3661,
+      "step": 9872
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.970212459564209,
+      "learning_rate": 4.5846919512613694e-06,
+      "loss": 0.2734,
+      "step": 9873
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.525469779968262,
+      "learning_rate": 4.582117727818774e-06,
+      "loss": 0.323,
+      "step": 9874
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.489565372467041,
+      "learning_rate": 4.57954350437618e-06,
+      "loss": 0.2885,
+      "step": 9875
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.656030654907227,
+      "learning_rate": 4.576969280933585e-06,
+      "loss": 0.3498,
+      "step": 9876
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.637785911560059,
+      "learning_rate": 4.574395057490991e-06,
+      "loss": 0.5924,
+      "step": 9877
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.836075782775879,
+      "learning_rate": 4.571820834048395e-06,
+      "loss": 0.366,
+      "step": 9878
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.544533729553223,
+      "learning_rate": 4.5692466106058e-06,
+      "loss": 0.3082,
+      "step": 9879
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.82541275024414,
+      "learning_rate": 4.566672387163206e-06,
+      "loss": 0.3581,
+      "step": 9880
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.754866600036621,
+      "learning_rate": 4.564098163720611e-06,
+      "loss": 0.4529,
+      "step": 9881
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.363245964050293,
+      "learning_rate": 4.561523940278017e-06,
+      "loss": 0.5338,
+      "step": 9882
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.042471885681152,
+      "learning_rate": 4.558949716835421e-06,
+      "loss": 0.589,
+      "step": 9883
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.8993916511535645,
+      "learning_rate": 4.556375493392827e-06,
+      "loss": 0.4759,
+      "step": 9884
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.697238922119141,
+      "learning_rate": 4.553801269950232e-06,
+      "loss": 0.284,
+      "step": 9885
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.843331336975098,
+      "learning_rate": 4.551227046507637e-06,
+      "loss": 0.4114,
+      "step": 9886
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.613215923309326,
+      "learning_rate": 4.548652823065043e-06,
+      "loss": 0.3233,
+      "step": 9887
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.457439422607422,
+      "learning_rate": 4.546078599622447e-06,
+      "loss": 0.672,
+      "step": 9888
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.027551651000977,
+      "learning_rate": 4.543504376179853e-06,
+      "loss": 0.4279,
+      "step": 9889
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.191003799438477,
+      "learning_rate": 4.540930152737258e-06,
+      "loss": 0.4173,
+      "step": 9890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.850259780883789,
+      "learning_rate": 4.538355929294663e-06,
+      "loss": 0.5251,
+      "step": 9891
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.420716285705566,
+      "learning_rate": 4.535781705852068e-06,
+      "loss": 0.4322,
+      "step": 9892
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.755199432373047,
+      "learning_rate": 4.5332074824094726e-06,
+      "loss": 0.3306,
+      "step": 9893
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.51894474029541,
+      "learning_rate": 4.530633258966878e-06,
+      "loss": 0.3436,
+      "step": 9894
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.663911819458008,
+      "learning_rate": 4.528059035524283e-06,
+      "loss": 0.4141,
+      "step": 9895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.478439331054688,
+      "learning_rate": 4.525484812081689e-06,
+      "loss": 0.3057,
+      "step": 9896
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.557174682617188,
+      "learning_rate": 4.522910588639094e-06,
+      "loss": 0.3271,
+      "step": 9897
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.117825984954834,
+      "learning_rate": 4.520336365196499e-06,
+      "loss": 0.3008,
+      "step": 9898
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.67201042175293,
+      "learning_rate": 4.517762141753904e-06,
+      "loss": 0.3839,
+      "step": 9899
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.920360088348389,
+      "learning_rate": 4.515187918311309e-06,
+      "loss": 0.3075,
+      "step": 9900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.682703971862793,
+      "learning_rate": 4.512613694868715e-06,
+      "loss": 0.2329,
+      "step": 9901
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.152132034301758,
+      "learning_rate": 4.51003947142612e-06,
+      "loss": 0.5156,
+      "step": 9902
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.8434672355651855,
+      "learning_rate": 4.507465247983526e-06,
+      "loss": 0.34,
+      "step": 9903
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.7037353515625,
+      "learning_rate": 4.50489102454093e-06,
+      "loss": 0.2548,
+      "step": 9904
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.541150093078613,
+      "learning_rate": 4.502316801098335e-06,
+      "loss": 0.378,
+      "step": 9905
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.245125770568848,
+      "learning_rate": 4.499742577655741e-06,
+      "loss": 0.2657,
+      "step": 9906
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.534658432006836,
+      "learning_rate": 4.497168354213146e-06,
+      "loss": 0.2804,
+      "step": 9907
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.769990921020508,
+      "learning_rate": 4.494594130770552e-06,
+      "loss": 0.2776,
+      "step": 9908
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.299349784851074,
+      "learning_rate": 4.492019907327956e-06,
+      "loss": 0.589,
+      "step": 9909
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.773977279663086,
+      "learning_rate": 4.489445683885362e-06,
+      "loss": 0.3645,
+      "step": 9910
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.670004844665527,
+      "learning_rate": 4.4868714604427666e-06,
+      "loss": 0.3539,
+      "step": 9911
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.599482536315918,
+      "learning_rate": 4.4842972370001716e-06,
+      "loss": 0.3158,
+      "step": 9912
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.113405227661133,
+      "learning_rate": 4.481723013557577e-06,
+      "loss": 0.6185,
+      "step": 9913
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.548078536987305,
+      "learning_rate": 4.4791487901149815e-06,
+      "loss": 0.3143,
+      "step": 9914
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.87138557434082,
+      "learning_rate": 4.476574566672387e-06,
+      "loss": 0.3951,
+      "step": 9915
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.681567192077637,
+      "learning_rate": 4.474000343229792e-06,
+      "loss": 0.5538,
+      "step": 9916
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.560541152954102,
+      "learning_rate": 4.471426119787197e-06,
+      "loss": 0.3245,
+      "step": 9917
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.184925079345703,
+      "learning_rate": 4.468851896344603e-06,
+      "loss": 0.3684,
+      "step": 9918
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.797863006591797,
+      "learning_rate": 4.466277672902007e-06,
+      "loss": 0.3596,
+      "step": 9919
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.702967643737793,
+      "learning_rate": 4.463703449459413e-06,
+      "loss": 0.4481,
+      "step": 9920
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.415507316589355,
+      "learning_rate": 4.461129226016818e-06,
+      "loss": 0.5798,
+      "step": 9921
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.53695297241211,
+      "learning_rate": 4.458555002574224e-06,
+      "loss": 0.4363,
+      "step": 9922
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.94764232635498,
+      "learning_rate": 4.455980779131629e-06,
+      "loss": 0.5268,
+      "step": 9923
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.764609336853027,
+      "learning_rate": 4.453406555689034e-06,
+      "loss": 0.2574,
+      "step": 9924
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.214308738708496,
+      "learning_rate": 4.450832332246439e-06,
+      "loss": 0.4664,
+      "step": 9925
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.65449047088623,
+      "learning_rate": 4.448258108803844e-06,
+      "loss": 0.4357,
+      "step": 9926
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.322643756866455,
+      "learning_rate": 4.44568388536125e-06,
+      "loss": 0.2718,
+      "step": 9927
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.729557991027832,
+      "learning_rate": 4.443109661918655e-06,
+      "loss": 0.3978,
+      "step": 9928
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.054461479187012,
+      "learning_rate": 4.4405354384760606e-06,
+      "loss": 0.3249,
+      "step": 9929
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.421133995056152,
+      "learning_rate": 4.437961215033465e-06,
+      "loss": 0.339,
+      "step": 9930
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.247720718383789,
+      "learning_rate": 4.43538699159087e-06,
+      "loss": 0.3763,
+      "step": 9931
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.756997108459473,
+      "learning_rate": 4.4328127681482755e-06,
+      "loss": 0.442,
+      "step": 9932
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.602229118347168,
+      "learning_rate": 4.4302385447056805e-06,
+      "loss": 0.3772,
+      "step": 9933
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.061059951782227,
+      "learning_rate": 4.427664321263086e-06,
+      "loss": 0.3439,
+      "step": 9934
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.922179222106934,
+      "learning_rate": 4.4250900978204905e-06,
+      "loss": 0.2709,
+      "step": 9935
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.109374046325684,
+      "learning_rate": 4.422515874377896e-06,
+      "loss": 0.5161,
+      "step": 9936
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.100266456604004,
+      "learning_rate": 4.419941650935301e-06,
+      "loss": 0.3431,
+      "step": 9937
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.069289207458496,
+      "learning_rate": 4.417367427492706e-06,
+      "loss": 0.3434,
+      "step": 9938
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.864737510681152,
+      "learning_rate": 4.414793204050112e-06,
+      "loss": 0.4447,
+      "step": 9939
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.718376159667969,
+      "learning_rate": 4.412218980607516e-06,
+      "loss": 0.3581,
+      "step": 9940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.893866539001465,
+      "learning_rate": 4.409644757164922e-06,
+      "loss": 0.3602,
+      "step": 9941
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.459463119506836,
+      "learning_rate": 4.407070533722327e-06,
+      "loss": 0.2941,
+      "step": 9942
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.783381462097168,
+      "learning_rate": 4.404496310279732e-06,
+      "loss": 0.434,
+      "step": 9943
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.828739166259766,
+      "learning_rate": 4.401922086837138e-06,
+      "loss": 0.4314,
+      "step": 9944
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.140368938446045,
+      "learning_rate": 4.399347863394542e-06,
+      "loss": 0.297,
+      "step": 9945
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.443787574768066,
+      "learning_rate": 4.396773639951948e-06,
+      "loss": 0.4304,
+      "step": 9946
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.49197769165039,
+      "learning_rate": 4.394199416509353e-06,
+      "loss": 0.4603,
+      "step": 9947
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.293033599853516,
+      "learning_rate": 4.391625193066759e-06,
+      "loss": 0.3378,
+      "step": 9948
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.394169807434082,
+      "learning_rate": 4.389050969624164e-06,
+      "loss": 0.3738,
+      "step": 9949
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.421829223632812,
+      "learning_rate": 4.386476746181569e-06,
+      "loss": 0.3499,
+      "step": 9950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.849088668823242,
+      "learning_rate": 4.383902522738974e-06,
+      "loss": 0.376,
+      "step": 9951
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.470183372497559,
+      "learning_rate": 4.381328299296379e-06,
+      "loss": 0.4127,
+      "step": 9952
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.784804344177246,
+      "learning_rate": 4.3787540758537845e-06,
+      "loss": 0.458,
+      "step": 9953
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.23931121826172,
+      "learning_rate": 4.3761798524111895e-06,
+      "loss": 0.3807,
+      "step": 9954
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.81545352935791,
+      "learning_rate": 4.373605628968595e-06,
+      "loss": 0.3841,
+      "step": 9955
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.764395713806152,
+      "learning_rate": 4.3710314055259995e-06,
+      "loss": 0.5208,
+      "step": 9956
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.672592163085938,
+      "learning_rate": 4.3684571820834045e-06,
+      "loss": 0.3087,
+      "step": 9957
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.531686782836914,
+      "learning_rate": 4.36588295864081e-06,
+      "loss": 0.4427,
+      "step": 9958
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.387245178222656,
+      "learning_rate": 4.363308735198215e-06,
+      "loss": 0.3499,
+      "step": 9959
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.841809272766113,
+      "learning_rate": 4.360734511755621e-06,
+      "loss": 0.3798,
+      "step": 9960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.419737815856934,
+      "learning_rate": 4.358160288313025e-06,
+      "loss": 0.344,
+      "step": 9961
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.818804740905762,
+      "learning_rate": 4.355586064870431e-06,
+      "loss": 0.4238,
+      "step": 9962
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.162422180175781,
+      "learning_rate": 4.353011841427836e-06,
+      "loss": 0.3439,
+      "step": 9963
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.351640701293945,
+      "learning_rate": 4.350437617985241e-06,
+      "loss": 0.2175,
+      "step": 9964
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.395376205444336,
+      "learning_rate": 4.347863394542647e-06,
+      "loss": 0.3205,
+      "step": 9965
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.97102165222168,
+      "learning_rate": 4.345289171100051e-06,
+      "loss": 0.3443,
+      "step": 9966
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.703822135925293,
+      "learning_rate": 4.342714947657457e-06,
+      "loss": 0.4079,
+      "step": 9967
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.510187149047852,
+      "learning_rate": 4.340140724214862e-06,
+      "loss": 0.3666,
+      "step": 9968
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.33984088897705,
+      "learning_rate": 4.337566500772267e-06,
+      "loss": 0.509,
+      "step": 9969
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.041324615478516,
+      "learning_rate": 4.334992277329673e-06,
+      "loss": 0.4211,
+      "step": 9970
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.166263580322266,
+      "learning_rate": 4.332418053887077e-06,
+      "loss": 0.4017,
+      "step": 9971
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.305109977722168,
+      "learning_rate": 4.329843830444483e-06,
+      "loss": 0.2828,
+      "step": 9972
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.540048599243164,
+      "learning_rate": 4.327269607001888e-06,
+      "loss": 0.6394,
+      "step": 9973
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.86319351196289,
+      "learning_rate": 4.3246953835592935e-06,
+      "loss": 0.5214,
+      "step": 9974
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.562620162963867,
+      "learning_rate": 4.3221211601166985e-06,
+      "loss": 0.4605,
+      "step": 9975
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.087349891662598,
+      "learning_rate": 4.3195469366741034e-06,
+      "loss": 0.4786,
+      "step": 9976
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.751546859741211,
+      "learning_rate": 4.3169727132315084e-06,
+      "loss": 0.5242,
+      "step": 9977
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.395437240600586,
+      "learning_rate": 4.314398489788913e-06,
+      "loss": 0.4307,
+      "step": 9978
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.593765258789062,
+      "learning_rate": 4.311824266346319e-06,
+      "loss": 0.4633,
+      "step": 9979
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.877291679382324,
+      "learning_rate": 4.309250042903724e-06,
+      "loss": 0.5356,
+      "step": 9980
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.459031105041504,
+      "learning_rate": 4.30667581946113e-06,
+      "loss": 0.3569,
+      "step": 9981
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.680886268615723,
+      "learning_rate": 4.304101596018534e-06,
+      "loss": 0.3856,
+      "step": 9982
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.003732681274414,
+      "learning_rate": 4.301527372575939e-06,
+      "loss": 0.3501,
+      "step": 9983
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.571664810180664,
+      "learning_rate": 4.298953149133345e-06,
+      "loss": 0.3511,
+      "step": 9984
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.06940746307373,
+      "learning_rate": 4.29637892569075e-06,
+      "loss": 0.638,
+      "step": 9985
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.53713607788086,
+      "learning_rate": 4.293804702248156e-06,
+      "loss": 0.2733,
+      "step": 9986
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.928442001342773,
+      "learning_rate": 4.29123047880556e-06,
+      "loss": 0.395,
+      "step": 9987
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.022892951965332,
+      "learning_rate": 4.288656255362966e-06,
+      "loss": 0.238,
+      "step": 9988
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.315838813781738,
+      "learning_rate": 4.286082031920371e-06,
+      "loss": 0.4526,
+      "step": 9989
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.385167121887207,
+      "learning_rate": 4.283507808477776e-06,
+      "loss": 0.3602,
+      "step": 9990
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.09528636932373,
+      "learning_rate": 4.280933585035182e-06,
+      "loss": 0.3881,
+      "step": 9991
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.53307056427002,
+      "learning_rate": 4.278359361592586e-06,
+      "loss": 0.5135,
+      "step": 9992
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.816743850708008,
+      "learning_rate": 4.275785138149992e-06,
+      "loss": 0.476,
+      "step": 9993
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.927627563476562,
+      "learning_rate": 4.273210914707397e-06,
+      "loss": 0.3254,
+      "step": 9994
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.05849838256836,
+      "learning_rate": 4.270636691264802e-06,
+      "loss": 0.3206,
+      "step": 9995
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.244296073913574,
+      "learning_rate": 4.2680624678222074e-06,
+      "loss": 0.3888,
+      "step": 9996
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.192716598510742,
+      "learning_rate": 4.2654882443796116e-06,
+      "loss": 0.2747,
+      "step": 9997
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.146156311035156,
+      "learning_rate": 4.262914020937017e-06,
+      "loss": 0.3986,
+      "step": 9998
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.070487022399902,
+      "learning_rate": 4.260339797494422e-06,
+      "loss": 0.4472,
+      "step": 9999
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.683749198913574,
+      "learning_rate": 4.257765574051828e-06,
+      "loss": 0.5275,
+      "step": 10000
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.69218921661377,
+      "learning_rate": 4.255191350609233e-06,
+      "loss": 0.4133,
+      "step": 10001
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.761098861694336,
+      "learning_rate": 4.252617127166638e-06,
+      "loss": 0.4975,
+      "step": 10002
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.708329200744629,
+      "learning_rate": 4.250042903724043e-06,
+      "loss": 0.2557,
+      "step": 10003
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.929306983947754,
+      "learning_rate": 4.247468680281448e-06,
+      "loss": 0.3182,
+      "step": 10004
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.045339584350586,
+      "learning_rate": 4.244894456838854e-06,
+      "loss": 0.3838,
+      "step": 10005
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.587363243103027,
+      "learning_rate": 4.242320233396259e-06,
+      "loss": 0.4017,
+      "step": 10006
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.690743446350098,
+      "learning_rate": 4.239746009953665e-06,
+      "loss": 0.3563,
+      "step": 10007
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.331488609313965,
+      "learning_rate": 4.237171786511069e-06,
+      "loss": 0.6087,
+      "step": 10008
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.600420951843262,
+      "learning_rate": 4.234597563068474e-06,
+      "loss": 0.3599,
+      "step": 10009
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.76400089263916,
+      "learning_rate": 4.23202333962588e-06,
+      "loss": 0.2751,
+      "step": 10010
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.677290916442871,
+      "learning_rate": 4.229449116183285e-06,
+      "loss": 0.4324,
+      "step": 10011
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.26309871673584,
+      "learning_rate": 4.226874892740691e-06,
+      "loss": 0.4589,
+      "step": 10012
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.799513816833496,
+      "learning_rate": 4.224300669298095e-06,
+      "loss": 0.4897,
+      "step": 10013
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.122469902038574,
+      "learning_rate": 4.221726445855501e-06,
+      "loss": 0.2205,
+      "step": 10014
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.203651428222656,
+      "learning_rate": 4.219152222412906e-06,
+      "loss": 0.3603,
+      "step": 10015
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.591609001159668,
+      "learning_rate": 4.2165779989703106e-06,
+      "loss": 0.2476,
+      "step": 10016
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.541642189025879,
+      "learning_rate": 4.214003775527716e-06,
+      "loss": 0.5272,
+      "step": 10017
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.456793785095215,
+      "learning_rate": 4.2114295520851205e-06,
+      "loss": 0.5535,
+      "step": 10018
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.074565887451172,
+      "learning_rate": 4.208855328642526e-06,
+      "loss": 0.4096,
+      "step": 10019
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.155806541442871,
+      "learning_rate": 4.206281105199931e-06,
+      "loss": 0.488,
+      "step": 10020
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.50942611694336,
+      "learning_rate": 4.203706881757336e-06,
+      "loss": 0.4609,
+      "step": 10021
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.536234855651855,
+      "learning_rate": 4.201132658314742e-06,
+      "loss": 0.4203,
+      "step": 10022
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.204566955566406,
+      "learning_rate": 4.198558434872146e-06,
+      "loss": 0.5778,
+      "step": 10023
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.064000129699707,
+      "learning_rate": 4.195984211429552e-06,
+      "loss": 0.3768,
+      "step": 10024
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.241158485412598,
+      "learning_rate": 4.193409987986957e-06,
+      "loss": 0.3996,
+      "step": 10025
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.174125671386719,
+      "learning_rate": 4.190835764544363e-06,
+      "loss": 0.3392,
+      "step": 10026
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.345174789428711,
+      "learning_rate": 4.188261541101768e-06,
+      "loss": 0.3971,
+      "step": 10027
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.275466918945312,
+      "learning_rate": 4.185687317659173e-06,
+      "loss": 0.5066,
+      "step": 10028
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.62310791015625,
+      "learning_rate": 4.183113094216578e-06,
+      "loss": 0.3347,
+      "step": 10029
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.437335014343262,
+      "learning_rate": 4.180538870773983e-06,
+      "loss": 0.3518,
+      "step": 10030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.76176929473877,
+      "learning_rate": 4.177964647331389e-06,
+      "loss": 0.4014,
+      "step": 10031
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.781759262084961,
+      "learning_rate": 4.175390423888794e-06,
+      "loss": 0.45,
+      "step": 10032
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.804431915283203,
+      "learning_rate": 4.1728162004462e-06,
+      "loss": 0.5412,
+      "step": 10033
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.3230619430542,
+      "learning_rate": 4.170241977003604e-06,
+      "loss": 0.3673,
+      "step": 10034
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.02815055847168,
+      "learning_rate": 4.167667753561009e-06,
+      "loss": 0.4242,
+      "step": 10035
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.48589038848877,
+      "learning_rate": 4.1650935301184145e-06,
+      "loss": 0.4381,
+      "step": 10036
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.080061912536621,
+      "learning_rate": 4.1625193066758195e-06,
+      "loss": 0.3745,
+      "step": 10037
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.295418739318848,
+      "learning_rate": 4.159945083233225e-06,
+      "loss": 0.2397,
+      "step": 10038
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.337566375732422,
+      "learning_rate": 4.1573708597906295e-06,
+      "loss": 0.4904,
+      "step": 10039
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.389517784118652,
+      "learning_rate": 4.154796636348035e-06,
+      "loss": 0.362,
+      "step": 10040
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.826855659484863,
+      "learning_rate": 4.15222241290544e-06,
+      "loss": 0.2595,
+      "step": 10041
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.343708992004395,
+      "learning_rate": 4.149648189462845e-06,
+      "loss": 0.3669,
+      "step": 10042
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.76363468170166,
+      "learning_rate": 4.147073966020251e-06,
+      "loss": 0.2913,
+      "step": 10043
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.270110130310059,
+      "learning_rate": 4.144499742577655e-06,
+      "loss": 0.4417,
+      "step": 10044
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.014410018920898,
+      "learning_rate": 4.141925519135061e-06,
+      "loss": 0.4179,
+      "step": 10045
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.9216227531433105,
+      "learning_rate": 4.139351295692466e-06,
+      "loss": 0.2627,
+      "step": 10046
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.388548851013184,
+      "learning_rate": 4.136777072249871e-06,
+      "loss": 0.4104,
+      "step": 10047
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.152256965637207,
+      "learning_rate": 4.134202848807277e-06,
+      "loss": 0.3494,
+      "step": 10048
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.463470458984375,
+      "learning_rate": 4.131628625364682e-06,
+      "loss": 0.2555,
+      "step": 10049
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.243407249450684,
+      "learning_rate": 4.129054401922087e-06,
+      "loss": 0.3652,
+      "step": 10050
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.6102933883667,
+      "learning_rate": 4.126480178479492e-06,
+      "loss": 0.2998,
+      "step": 10051
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.25294017791748,
+      "learning_rate": 4.123905955036898e-06,
+      "loss": 0.512,
+      "step": 10052
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.124979019165039,
+      "learning_rate": 4.121331731594303e-06,
+      "loss": 0.3626,
+      "step": 10053
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.336242198944092,
+      "learning_rate": 4.118757508151708e-06,
+      "loss": 0.215,
+      "step": 10054
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.061466217041016,
+      "learning_rate": 4.116183284709113e-06,
+      "loss": 0.3215,
+      "step": 10055
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.13331413269043,
+      "learning_rate": 4.113609061266518e-06,
+      "loss": 0.4244,
+      "step": 10056
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.146357536315918,
+      "learning_rate": 4.1110348378239235e-06,
+      "loss": 0.3712,
+      "step": 10057
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.410157203674316,
+      "learning_rate": 4.1084606143813285e-06,
+      "loss": 0.3266,
+      "step": 10058
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.439155578613281,
+      "learning_rate": 4.105886390938734e-06,
+      "loss": 0.2689,
+      "step": 10059
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.937128067016602,
+      "learning_rate": 4.1033121674961385e-06,
+      "loss": 0.298,
+      "step": 10060
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.979667663574219,
+      "learning_rate": 4.1007379440535435e-06,
+      "loss": 0.4274,
+      "step": 10061
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.426986694335938,
+      "learning_rate": 4.098163720610949e-06,
+      "loss": 0.4087,
+      "step": 10062
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.391172409057617,
+      "learning_rate": 4.095589497168354e-06,
+      "loss": 0.3698,
+      "step": 10063
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.403900146484375,
+      "learning_rate": 4.09301527372576e-06,
+      "loss": 0.4726,
+      "step": 10064
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.048985481262207,
+      "learning_rate": 4.090441050283164e-06,
+      "loss": 0.1954,
+      "step": 10065
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.735633850097656,
+      "learning_rate": 4.087866826840569e-06,
+      "loss": 0.2995,
+      "step": 10066
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.164621353149414,
+      "learning_rate": 4.085292603397975e-06,
+      "loss": 0.2752,
+      "step": 10067
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.355125427246094,
+      "learning_rate": 4.08271837995538e-06,
+      "loss": 0.4,
+      "step": 10068
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.642196655273438,
+      "learning_rate": 4.080144156512786e-06,
+      "loss": 0.5601,
+      "step": 10069
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.689550399780273,
+      "learning_rate": 4.07756993307019e-06,
+      "loss": 0.325,
+      "step": 10070
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.884004592895508,
+      "learning_rate": 4.074995709627596e-06,
+      "loss": 0.3896,
+      "step": 10071
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.69339370727539,
+      "learning_rate": 4.072421486185001e-06,
+      "loss": 0.386,
+      "step": 10072
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.559497833251953,
+      "learning_rate": 4.069847262742406e-06,
+      "loss": 0.3497,
+      "step": 10073
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.112516403198242,
+      "learning_rate": 4.067273039299812e-06,
+      "loss": 0.2648,
+      "step": 10074
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.096497535705566,
+      "learning_rate": 4.064698815857217e-06,
+      "loss": 0.4449,
+      "step": 10075
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.329072952270508,
+      "learning_rate": 4.062124592414622e-06,
+      "loss": 0.3324,
+      "step": 10076
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.353119850158691,
+      "learning_rate": 4.059550368972027e-06,
+      "loss": 0.369,
+      "step": 10077
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.638492584228516,
+      "learning_rate": 4.0569761455294325e-06,
+      "loss": 0.2372,
+      "step": 10078
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.622969627380371,
+      "learning_rate": 4.0544019220868375e-06,
+      "loss": 0.2513,
+      "step": 10079
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.213968276977539,
+      "learning_rate": 4.0518276986442425e-06,
+      "loss": 0.5631,
+      "step": 10080
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.203161239624023,
+      "learning_rate": 4.0492534752016474e-06,
+      "loss": 0.4392,
+      "step": 10081
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.644662857055664,
+      "learning_rate": 4.0466792517590524e-06,
+      "loss": 0.4,
+      "step": 10082
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.43625545501709,
+      "learning_rate": 4.044105028316458e-06,
+      "loss": 0.3789,
+      "step": 10083
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.148271560668945,
+      "learning_rate": 4.041530804873863e-06,
+      "loss": 0.3693,
+      "step": 10084
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.060578346252441,
+      "learning_rate": 4.038956581431269e-06,
+      "loss": 0.1837,
+      "step": 10085
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.784847259521484,
+      "learning_rate": 4.036382357988673e-06,
+      "loss": 0.4735,
+      "step": 10086
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.989152908325195,
+      "learning_rate": 4.033808134546078e-06,
+      "loss": 0.3457,
+      "step": 10087
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.844850540161133,
+      "learning_rate": 4.031233911103484e-06,
+      "loss": 0.3449,
+      "step": 10088
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.894591331481934,
+      "learning_rate": 4.028659687660889e-06,
+      "loss": 0.4263,
+      "step": 10089
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.945180892944336,
+      "learning_rate": 4.026085464218295e-06,
+      "loss": 0.2866,
+      "step": 10090
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.589032173156738,
+      "learning_rate": 4.023511240775699e-06,
+      "loss": 0.6389,
+      "step": 10091
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.864961624145508,
+      "learning_rate": 4.020937017333104e-06,
+      "loss": 0.6178,
+      "step": 10092
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.489991188049316,
+      "learning_rate": 4.01836279389051e-06,
+      "loss": 0.3439,
+      "step": 10093
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.373434066772461,
+      "learning_rate": 4.015788570447915e-06,
+      "loss": 0.3362,
+      "step": 10094
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.838357448577881,
+      "learning_rate": 4.013214347005321e-06,
+      "loss": 0.2354,
+      "step": 10095
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.028657913208008,
+      "learning_rate": 4.010640123562725e-06,
+      "loss": 0.281,
+      "step": 10096
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.155238151550293,
+      "learning_rate": 4.008065900120131e-06,
+      "loss": 0.2869,
+      "step": 10097
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.65730094909668,
+      "learning_rate": 4.005491676677536e-06,
+      "loss": 0.4494,
+      "step": 10098
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.49343490600586,
+      "learning_rate": 4.002917453234941e-06,
+      "loss": 0.41,
+      "step": 10099
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.031082153320312,
+      "learning_rate": 4.0003432297923464e-06,
+      "loss": 0.3188,
+      "step": 10100
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.496113777160645,
+      "learning_rate": 3.9977690063497514e-06,
+      "loss": 0.3532,
+      "step": 10101
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.07313060760498,
+      "learning_rate": 3.995194782907156e-06,
+      "loss": 0.4757,
+      "step": 10102
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.511941909790039,
+      "learning_rate": 3.992620559464561e-06,
+      "loss": 0.4832,
+      "step": 10103
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.121332168579102,
+      "learning_rate": 3.990046336021967e-06,
+      "loss": 0.2619,
+      "step": 10104
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.800068855285645,
+      "learning_rate": 3.987472112579372e-06,
+      "loss": 0.4312,
+      "step": 10105
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.701355934143066,
+      "learning_rate": 3.984897889136777e-06,
+      "loss": 0.3494,
+      "step": 10106
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.681970596313477,
+      "learning_rate": 3.982323665694182e-06,
+      "loss": 0.3549,
+      "step": 10107
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.654945373535156,
+      "learning_rate": 3.979749442251587e-06,
+      "loss": 0.2984,
+      "step": 10108
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.594856262207031,
+      "learning_rate": 3.977175218808993e-06,
+      "loss": 0.3414,
+      "step": 10109
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.8992338180542,
+      "learning_rate": 3.974600995366398e-06,
+      "loss": 0.5667,
+      "step": 10110
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.656254768371582,
+      "learning_rate": 3.972026771923804e-06,
+      "loss": 0.464,
+      "step": 10111
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.926775932312012,
+      "learning_rate": 3.969452548481208e-06,
+      "loss": 0.3728,
+      "step": 10112
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.054173469543457,
+      "learning_rate": 3.966878325038613e-06,
+      "loss": 0.5664,
+      "step": 10113
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.295503616333008,
+      "learning_rate": 3.964304101596019e-06,
+      "loss": 0.4331,
+      "step": 10114
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.960628509521484,
+      "learning_rate": 3.961729878153424e-06,
+      "loss": 0.3698,
+      "step": 10115
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.869596481323242,
+      "learning_rate": 3.95915565471083e-06,
+      "loss": 0.4341,
+      "step": 10116
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.166833877563477,
+      "learning_rate": 3.956581431268234e-06,
+      "loss": 0.3685,
+      "step": 10117
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.592395782470703,
+      "learning_rate": 3.954007207825639e-06,
+      "loss": 0.4397,
+      "step": 10118
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.791773796081543,
+      "learning_rate": 3.951432984383045e-06,
+      "loss": 0.3871,
+      "step": 10119
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.0762357711792,
+      "learning_rate": 3.94885876094045e-06,
+      "loss": 0.3591,
+      "step": 10120
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.5891695022583,
+      "learning_rate": 3.946284537497855e-06,
+      "loss": 0.5207,
+      "step": 10121
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.399431228637695,
+      "learning_rate": 3.9437103140552596e-06,
+      "loss": 0.352,
+      "step": 10122
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.89193058013916,
+      "learning_rate": 3.941136090612665e-06,
+      "loss": 0.4992,
+      "step": 10123
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.691988468170166,
+      "learning_rate": 3.93856186717007e-06,
+      "loss": 0.159,
+      "step": 10124
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.005675315856934,
+      "learning_rate": 3.935987643727475e-06,
+      "loss": 0.3312,
+      "step": 10125
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.908039569854736,
+      "learning_rate": 3.933413420284881e-06,
+      "loss": 0.1449,
+      "step": 10126
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.309003829956055,
+      "learning_rate": 3.930839196842286e-06,
+      "loss": 0.4602,
+      "step": 10127
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.464003562927246,
+      "learning_rate": 3.928264973399691e-06,
+      "loss": 0.3839,
+      "step": 10128
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.687711715698242,
+      "learning_rate": 3.925690749957096e-06,
+      "loss": 0.3445,
+      "step": 10129
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.281744956970215,
+      "learning_rate": 3.923116526514502e-06,
+      "loss": 0.5241,
+      "step": 10130
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.498844146728516,
+      "learning_rate": 3.920542303071907e-06,
+      "loss": 0.327,
+      "step": 10131
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.091933250427246,
+      "learning_rate": 3.917968079629312e-06,
+      "loss": 0.2824,
+      "step": 10132
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.312525749206543,
+      "learning_rate": 3.915393856186717e-06,
+      "loss": 0.5629,
+      "step": 10133
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.121058464050293,
+      "learning_rate": 3.912819632744122e-06,
+      "loss": 0.4358,
+      "step": 10134
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.324707984924316,
+      "learning_rate": 3.910245409301528e-06,
+      "loss": 0.3947,
+      "step": 10135
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.844686508178711,
+      "learning_rate": 3.907671185858933e-06,
+      "loss": 0.2847,
+      "step": 10136
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.180686950683594,
+      "learning_rate": 3.905096962416339e-06,
+      "loss": 0.3813,
+      "step": 10137
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.596451759338379,
+      "learning_rate": 3.902522738973743e-06,
+      "loss": 0.3386,
+      "step": 10138
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.376317977905273,
+      "learning_rate": 3.899948515531148e-06,
+      "loss": 0.4449,
+      "step": 10139
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.76968002319336,
+      "learning_rate": 3.8973742920885536e-06,
+      "loss": 0.1939,
+      "step": 10140
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.617429733276367,
+      "learning_rate": 3.8948000686459585e-06,
+      "loss": 0.511,
+      "step": 10141
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.86064624786377,
+      "learning_rate": 3.892225845203364e-06,
+      "loss": 0.3824,
+      "step": 10142
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.791238784790039,
+      "learning_rate": 3.8896516217607685e-06,
+      "loss": 0.3047,
+      "step": 10143
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.900047302246094,
+      "learning_rate": 3.8870773983181735e-06,
+      "loss": 0.2892,
+      "step": 10144
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.590638160705566,
+      "learning_rate": 3.884503174875579e-06,
+      "loss": 0.3112,
+      "step": 10145
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.809215545654297,
+      "learning_rate": 3.881928951432984e-06,
+      "loss": 0.3487,
+      "step": 10146
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.319319725036621,
+      "learning_rate": 3.87935472799039e-06,
+      "loss": 0.5492,
+      "step": 10147
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.404623031616211,
+      "learning_rate": 3.876780504547794e-06,
+      "loss": 0.3197,
+      "step": 10148
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.93860149383545,
+      "learning_rate": 3.8742062811052e-06,
+      "loss": 0.2761,
+      "step": 10149
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.618767738342285,
+      "learning_rate": 3.871632057662605e-06,
+      "loss": 0.6052,
+      "step": 10150
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.340433597564697,
+      "learning_rate": 3.86905783422001e-06,
+      "loss": 0.2744,
+      "step": 10151
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.260811805725098,
+      "learning_rate": 3.866483610777416e-06,
+      "loss": 0.606,
+      "step": 10152
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.70333194732666,
+      "learning_rate": 3.863909387334821e-06,
+      "loss": 0.2935,
+      "step": 10153
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.234350204467773,
+      "learning_rate": 3.861335163892226e-06,
+      "loss": 0.3622,
+      "step": 10154
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.689064025878906,
+      "learning_rate": 3.858760940449631e-06,
+      "loss": 0.5567,
+      "step": 10155
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.362519264221191,
+      "learning_rate": 3.856186717007037e-06,
+      "loss": 0.3005,
+      "step": 10156
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.870559692382812,
+      "learning_rate": 3.853612493564442e-06,
+      "loss": 0.3534,
+      "step": 10157
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.29106616973877,
+      "learning_rate": 3.851038270121847e-06,
+      "loss": 0.2802,
+      "step": 10158
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.681184768676758,
+      "learning_rate": 3.848464046679252e-06,
+      "loss": 0.3886,
+      "step": 10159
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.985565185546875,
+      "learning_rate": 3.845889823236657e-06,
+      "loss": 0.4138,
+      "step": 10160
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.857745170593262,
+      "learning_rate": 3.8433155997940625e-06,
+      "loss": 0.3003,
+      "step": 10161
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.65396785736084,
+      "learning_rate": 3.8407413763514675e-06,
+      "loss": 0.4674,
+      "step": 10162
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.1290864944458,
+      "learning_rate": 3.838167152908873e-06,
+      "loss": 0.4457,
+      "step": 10163
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.762251853942871,
+      "learning_rate": 3.8355929294662775e-06,
+      "loss": 0.3151,
+      "step": 10164
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.683486938476562,
+      "learning_rate": 3.8330187060236825e-06,
+      "loss": 0.4095,
+      "step": 10165
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.60052490234375,
+      "learning_rate": 3.830444482581088e-06,
+      "loss": 0.5007,
+      "step": 10166
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.125239372253418,
+      "learning_rate": 3.827870259138493e-06,
+      "loss": 0.3678,
+      "step": 10167
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.572195053100586,
+      "learning_rate": 3.825296035695899e-06,
+      "loss": 0.383,
+      "step": 10168
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.062673568725586,
+      "learning_rate": 3.822721812253303e-06,
+      "loss": 0.3307,
+      "step": 10169
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.542654991149902,
+      "learning_rate": 3.820147588810708e-06,
+      "loss": 0.3823,
+      "step": 10170
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.79784870147705,
+      "learning_rate": 3.817573365368114e-06,
+      "loss": 0.3434,
+      "step": 10171
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.200736045837402,
+      "learning_rate": 3.814999141925519e-06,
+      "loss": 0.2691,
+      "step": 10172
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.891195297241211,
+      "learning_rate": 3.8124249184829245e-06,
+      "loss": 0.3118,
+      "step": 10173
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.681863784790039,
+      "learning_rate": 3.8098506950403295e-06,
+      "loss": 0.3609,
+      "step": 10174
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.811197280883789,
+      "learning_rate": 3.8072764715977353e-06,
+      "loss": 0.425,
+      "step": 10175
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.526089668273926,
+      "learning_rate": 3.80470224815514e-06,
+      "loss": 0.3092,
+      "step": 10176
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.435626029968262,
+      "learning_rate": 3.802128024712545e-06,
+      "loss": 0.4314,
+      "step": 10177
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.051081657409668,
+      "learning_rate": 3.7995538012699503e-06,
+      "loss": 0.3363,
+      "step": 10178
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.169252395629883,
+      "learning_rate": 3.7969795778273553e-06,
+      "loss": 0.3322,
+      "step": 10179
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.432186126708984,
+      "learning_rate": 3.794405354384761e-06,
+      "loss": 0.3272,
+      "step": 10180
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.028060913085938,
+      "learning_rate": 3.7918311309421657e-06,
+      "loss": 0.4354,
+      "step": 10181
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.405692100524902,
+      "learning_rate": 3.7892569074995715e-06,
+      "loss": 0.3692,
+      "step": 10182
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.497088432312012,
+      "learning_rate": 3.786682684056976e-06,
+      "loss": 0.4336,
+      "step": 10183
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.881348609924316,
+      "learning_rate": 3.784108460614381e-06,
+      "loss": 0.4159,
+      "step": 10184
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.188570022583008,
+      "learning_rate": 3.781534237171787e-06,
+      "loss": 0.4665,
+      "step": 10185
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.117717742919922,
+      "learning_rate": 3.7789600137291914e-06,
+      "loss": 0.3599,
+      "step": 10186
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.287288665771484,
+      "learning_rate": 3.7763857902865973e-06,
+      "loss": 0.2787,
+      "step": 10187
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.764989852905273,
+      "learning_rate": 3.7738115668440023e-06,
+      "loss": 0.389,
+      "step": 10188
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.528788566589355,
+      "learning_rate": 3.7712373434014077e-06,
+      "loss": 0.3383,
+      "step": 10189
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.32126235961914,
+      "learning_rate": 3.7686631199588127e-06,
+      "loss": 0.2743,
+      "step": 10190
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.467133522033691,
+      "learning_rate": 3.7660888965162172e-06,
+      "loss": 0.4855,
+      "step": 10191
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.054813385009766,
+      "learning_rate": 3.763514673073623e-06,
+      "loss": 0.3842,
+      "step": 10192
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.813760757446289,
+      "learning_rate": 3.760940449631028e-06,
+      "loss": 0.409,
+      "step": 10193
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.009858131408691,
+      "learning_rate": 3.7583662261884335e-06,
+      "loss": 0.3529,
+      "step": 10194
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.135607719421387,
+      "learning_rate": 3.7557920027458384e-06,
+      "loss": 0.3987,
+      "step": 10195
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 16.250125885009766,
+      "learning_rate": 3.753217779303243e-06,
+      "loss": 0.3077,
+      "step": 10196
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.336618423461914,
+      "learning_rate": 3.750643555860649e-06,
+      "loss": 0.5505,
+      "step": 10197
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.917937278747559,
+      "learning_rate": 3.7480693324180543e-06,
+      "loss": 0.3855,
+      "step": 10198
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.916064262390137,
+      "learning_rate": 3.745495108975459e-06,
+      "loss": 0.2829,
+      "step": 10199
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.031928062438965,
+      "learning_rate": 3.7429208855328642e-06,
+      "loss": 0.5476,
+      "step": 10200
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.900164604187012,
+      "learning_rate": 3.7403466620902696e-06,
+      "loss": 0.5107,
+      "step": 10201
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.412485122680664,
+      "learning_rate": 3.7377724386476746e-06,
+      "loss": 0.3165,
+      "step": 10202
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.321499824523926,
+      "learning_rate": 3.73519821520508e-06,
+      "loss": 0.5741,
+      "step": 10203
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.0417375564575195,
+      "learning_rate": 3.732623991762485e-06,
+      "loss": 0.322,
+      "step": 10204
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.423901557922363,
+      "learning_rate": 3.7300497683198904e-06,
+      "loss": 0.479,
+      "step": 10205
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.220063209533691,
+      "learning_rate": 3.7274755448772954e-06,
+      "loss": 0.3528,
+      "step": 10206
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.971394538879395,
+      "learning_rate": 3.7249013214347004e-06,
+      "loss": 0.379,
+      "step": 10207
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.72043228149414,
+      "learning_rate": 3.722327097992106e-06,
+      "loss": 0.489,
+      "step": 10208
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.088316917419434,
+      "learning_rate": 3.719752874549511e-06,
+      "loss": 0.437,
+      "step": 10209
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.305198669433594,
+      "learning_rate": 3.7171786511069162e-06,
+      "loss": 0.5663,
+      "step": 10210
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.237088203430176,
+      "learning_rate": 3.7146044276643216e-06,
+      "loss": 0.2819,
+      "step": 10211
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.895726203918457,
+      "learning_rate": 3.712030204221726e-06,
+      "loss": 0.4606,
+      "step": 10212
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.748737335205078,
+      "learning_rate": 3.7094559807791316e-06,
+      "loss": 0.3876,
+      "step": 10213
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.259161949157715,
+      "learning_rate": 3.706881757336537e-06,
+      "loss": 0.4186,
+      "step": 10214
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.028119087219238,
+      "learning_rate": 3.704307533893942e-06,
+      "loss": 0.2588,
+      "step": 10215
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.59246826171875,
+      "learning_rate": 3.7017333104513474e-06,
+      "loss": 0.3857,
+      "step": 10216
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.146876335144043,
+      "learning_rate": 3.6991590870087524e-06,
+      "loss": 0.4382,
+      "step": 10217
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.523021697998047,
+      "learning_rate": 3.696584863566158e-06,
+      "loss": 0.3964,
+      "step": 10218
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.233314514160156,
+      "learning_rate": 3.694010640123563e-06,
+      "loss": 0.3836,
+      "step": 10219
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.950338363647461,
+      "learning_rate": 3.691436416680968e-06,
+      "loss": 0.3473,
+      "step": 10220
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.352520942687988,
+      "learning_rate": 3.688862193238373e-06,
+      "loss": 0.4426,
+      "step": 10221
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.945180892944336,
+      "learning_rate": 3.686287969795778e-06,
+      "loss": 0.4705,
+      "step": 10222
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.934719085693359,
+      "learning_rate": 3.6837137463531836e-06,
+      "loss": 0.2289,
+      "step": 10223
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.892194747924805,
+      "learning_rate": 3.681139522910589e-06,
+      "loss": 0.5061,
+      "step": 10224
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.008149147033691,
+      "learning_rate": 3.6785652994679936e-06,
+      "loss": 0.4234,
+      "step": 10225
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.622702598571777,
+      "learning_rate": 3.675991076025399e-06,
+      "loss": 0.5187,
+      "step": 10226
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.735066413879395,
+      "learning_rate": 3.6734168525828044e-06,
+      "loss": 0.3654,
+      "step": 10227
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.49288558959961,
+      "learning_rate": 3.6708426291402094e-06,
+      "loss": 0.3758,
+      "step": 10228
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.763015747070312,
+      "learning_rate": 3.668268405697615e-06,
+      "loss": 0.4298,
+      "step": 10229
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.570923805236816,
+      "learning_rate": 3.6656941822550198e-06,
+      "loss": 0.2124,
+      "step": 10230
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.82702922821045,
+      "learning_rate": 3.663119958812425e-06,
+      "loss": 0.3749,
+      "step": 10231
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.77657699584961,
+      "learning_rate": 3.66054573536983e-06,
+      "loss": 0.2517,
+      "step": 10232
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.193717002868652,
+      "learning_rate": 3.657971511927235e-06,
+      "loss": 0.4446,
+      "step": 10233
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.4744038581848145,
+      "learning_rate": 3.6553972884846406e-06,
+      "loss": 0.1988,
+      "step": 10234
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.41050910949707,
+      "learning_rate": 3.6528230650420456e-06,
+      "loss": 0.3939,
+      "step": 10235
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.848515510559082,
+      "learning_rate": 3.650248841599451e-06,
+      "loss": 0.4534,
+      "step": 10236
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.180014610290527,
+      "learning_rate": 3.6476746181568564e-06,
+      "loss": 0.2749,
+      "step": 10237
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.61145782470703,
+      "learning_rate": 3.645100394714261e-06,
+      "loss": 0.3995,
+      "step": 10238
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.235429763793945,
+      "learning_rate": 3.6425261712716664e-06,
+      "loss": 0.3171,
+      "step": 10239
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.68899154663086,
+      "learning_rate": 3.6399519478290718e-06,
+      "loss": 0.4255,
+      "step": 10240
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.805556297302246,
+      "learning_rate": 3.6373777243864768e-06,
+      "loss": 0.4189,
+      "step": 10241
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.117131233215332,
+      "learning_rate": 3.634803500943882e-06,
+      "loss": 0.5465,
+      "step": 10242
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.267644882202148,
+      "learning_rate": 3.632229277501287e-06,
+      "loss": 0.4355,
+      "step": 10243
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.417571067810059,
+      "learning_rate": 3.6296550540586926e-06,
+      "loss": 0.3252,
+      "step": 10244
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.063173294067383,
+      "learning_rate": 3.6270808306160976e-06,
+      "loss": 0.2651,
+      "step": 10245
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.665863990783691,
+      "learning_rate": 3.6245066071735025e-06,
+      "loss": 0.3893,
+      "step": 10246
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.454741477966309,
+      "learning_rate": 3.621932383730908e-06,
+      "loss": 0.6182,
+      "step": 10247
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.875772476196289,
+      "learning_rate": 3.619358160288313e-06,
+      "loss": 0.4422,
+      "step": 10248
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.632668495178223,
+      "learning_rate": 3.6167839368457184e-06,
+      "loss": 0.3494,
+      "step": 10249
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.414258003234863,
+      "learning_rate": 3.6142097134031238e-06,
+      "loss": 0.1634,
+      "step": 10250
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.760488510131836,
+      "learning_rate": 3.6116354899605283e-06,
+      "loss": 0.3774,
+      "step": 10251
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.845131874084473,
+      "learning_rate": 3.6090612665179337e-06,
+      "loss": 0.3897,
+      "step": 10252
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.781501770019531,
+      "learning_rate": 3.606487043075339e-06,
+      "loss": 0.3187,
+      "step": 10253
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.929583549499512,
+      "learning_rate": 3.603912819632744e-06,
+      "loss": 0.469,
+      "step": 10254
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.2886323928833,
+      "learning_rate": 3.6013385961901495e-06,
+      "loss": 0.2773,
+      "step": 10255
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.797028541564941,
+      "learning_rate": 3.5987643727475545e-06,
+      "loss": 0.2498,
+      "step": 10256
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.603740692138672,
+      "learning_rate": 3.59619014930496e-06,
+      "loss": 0.4764,
+      "step": 10257
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.143132209777832,
+      "learning_rate": 3.593615925862365e-06,
+      "loss": 0.2996,
+      "step": 10258
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.65654182434082,
+      "learning_rate": 3.59104170241977e-06,
+      "loss": 0.2893,
+      "step": 10259
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.181256294250488,
+      "learning_rate": 3.5884674789771753e-06,
+      "loss": 0.3335,
+      "step": 10260
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.574747085571289,
+      "learning_rate": 3.5858932555345803e-06,
+      "loss": 0.3222,
+      "step": 10261
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.166397094726562,
+      "learning_rate": 3.5833190320919857e-06,
+      "loss": 0.3921,
+      "step": 10262
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.948464393615723,
+      "learning_rate": 3.580744808649391e-06,
+      "loss": 0.3042,
+      "step": 10263
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.119213104248047,
+      "learning_rate": 3.5781705852067957e-06,
+      "loss": 0.2746,
+      "step": 10264
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.38940143585205,
+      "learning_rate": 3.575596361764201e-06,
+      "loss": 0.5122,
+      "step": 10265
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.604372024536133,
+      "learning_rate": 3.5730221383216065e-06,
+      "loss": 0.3038,
+      "step": 10266
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.439919948577881,
+      "learning_rate": 3.5704479148790115e-06,
+      "loss": 0.2691,
+      "step": 10267
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.319008827209473,
+      "learning_rate": 3.567873691436417e-06,
+      "loss": 0.4428,
+      "step": 10268
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.040205955505371,
+      "learning_rate": 3.565299467993822e-06,
+      "loss": 0.2964,
+      "step": 10269
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.960899353027344,
+      "learning_rate": 3.5627252445512273e-06,
+      "loss": 0.3532,
+      "step": 10270
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.433367729187012,
+      "learning_rate": 3.5601510211086323e-06,
+      "loss": 0.3188,
+      "step": 10271
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.964232444763184,
+      "learning_rate": 3.5575767976660373e-06,
+      "loss": 0.5936,
+      "step": 10272
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.903236389160156,
+      "learning_rate": 3.5550025742234427e-06,
+      "loss": 0.5645,
+      "step": 10273
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.747243881225586,
+      "learning_rate": 3.5524283507808477e-06,
+      "loss": 0.6095,
+      "step": 10274
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.366724491119385,
+      "learning_rate": 3.549854127338253e-06,
+      "loss": 0.2494,
+      "step": 10275
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.175317764282227,
+      "learning_rate": 3.5472799038956585e-06,
+      "loss": 0.4956,
+      "step": 10276
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.862730979919434,
+      "learning_rate": 3.544705680453063e-06,
+      "loss": 0.3139,
+      "step": 10277
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.148316383361816,
+      "learning_rate": 3.5421314570104685e-06,
+      "loss": 0.3412,
+      "step": 10278
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.67927360534668,
+      "learning_rate": 3.539557233567874e-06,
+      "loss": 0.4003,
+      "step": 10279
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.142306327819824,
+      "learning_rate": 3.536983010125279e-06,
+      "loss": 0.4048,
+      "step": 10280
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.645539283752441,
+      "learning_rate": 3.5344087866826843e-06,
+      "loss": 0.3232,
+      "step": 10281
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.56098461151123,
+      "learning_rate": 3.5318345632400893e-06,
+      "loss": 0.3882,
+      "step": 10282
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.415877342224121,
+      "learning_rate": 3.5292603397974947e-06,
+      "loss": 0.2145,
+      "step": 10283
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.05709171295166,
+      "learning_rate": 3.5266861163548997e-06,
+      "loss": 0.4451,
+      "step": 10284
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.837220191955566,
+      "learning_rate": 3.5241118929123047e-06,
+      "loss": 0.5567,
+      "step": 10285
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.209320068359375,
+      "learning_rate": 3.52153766946971e-06,
+      "loss": 0.513,
+      "step": 10286
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.352028846740723,
+      "learning_rate": 3.518963446027115e-06,
+      "loss": 0.3012,
+      "step": 10287
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.550464630126953,
+      "learning_rate": 3.5163892225845205e-06,
+      "loss": 0.4202,
+      "step": 10288
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.208223342895508,
+      "learning_rate": 3.513814999141926e-06,
+      "loss": 0.4685,
+      "step": 10289
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.388490676879883,
+      "learning_rate": 3.5112407756993305e-06,
+      "loss": 0.2887,
+      "step": 10290
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.14285659790039,
+      "learning_rate": 3.508666552256736e-06,
+      "loss": 0.5275,
+      "step": 10291
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.451848030090332,
+      "learning_rate": 3.5060923288141413e-06,
+      "loss": 0.6252,
+      "step": 10292
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.768813133239746,
+      "learning_rate": 3.5035181053715463e-06,
+      "loss": 0.3542,
+      "step": 10293
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.399510383605957,
+      "learning_rate": 3.5009438819289517e-06,
+      "loss": 0.4854,
+      "step": 10294
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.238245964050293,
+      "learning_rate": 3.4983696584863567e-06,
+      "loss": 0.3979,
+      "step": 10295
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.908125877380371,
+      "learning_rate": 3.495795435043762e-06,
+      "loss": 0.4981,
+      "step": 10296
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.450529098510742,
+      "learning_rate": 3.493221211601167e-06,
+      "loss": 0.3135,
+      "step": 10297
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.324727535247803,
+      "learning_rate": 3.490646988158572e-06,
+      "loss": 0.2525,
+      "step": 10298
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.562472820281982,
+      "learning_rate": 3.4880727647159775e-06,
+      "loss": 0.263,
+      "step": 10299
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.205446243286133,
+      "learning_rate": 3.4854985412733824e-06,
+      "loss": 0.301,
+      "step": 10300
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.021312713623047,
+      "learning_rate": 3.482924317830788e-06,
+      "loss": 0.463,
+      "step": 10301
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.65982437133789,
+      "learning_rate": 3.4803500943881933e-06,
+      "loss": 0.4438,
+      "step": 10302
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.424298286437988,
+      "learning_rate": 3.477775870945598e-06,
+      "loss": 0.2931,
+      "step": 10303
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.964347839355469,
+      "learning_rate": 3.4752016475030032e-06,
+      "loss": 0.3709,
+      "step": 10304
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.874922752380371,
+      "learning_rate": 3.4726274240604087e-06,
+      "loss": 0.3808,
+      "step": 10305
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.361870765686035,
+      "learning_rate": 3.4700532006178136e-06,
+      "loss": 0.4141,
+      "step": 10306
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.834250450134277,
+      "learning_rate": 3.467478977175219e-06,
+      "loss": 0.5122,
+      "step": 10307
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.968622207641602,
+      "learning_rate": 3.464904753732624e-06,
+      "loss": 0.3735,
+      "step": 10308
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.21124267578125,
+      "learning_rate": 3.4623305302900295e-06,
+      "loss": 0.2812,
+      "step": 10309
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.070713996887207,
+      "learning_rate": 3.4597563068474344e-06,
+      "loss": 0.3327,
+      "step": 10310
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.503514289855957,
+      "learning_rate": 3.4571820834048394e-06,
+      "loss": 0.4168,
+      "step": 10311
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.725349426269531,
+      "learning_rate": 3.454607859962245e-06,
+      "loss": 0.33,
+      "step": 10312
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.072580337524414,
+      "learning_rate": 3.4520336365196502e-06,
+      "loss": 0.3008,
+      "step": 10313
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.163908004760742,
+      "learning_rate": 3.4494594130770552e-06,
+      "loss": 0.4394,
+      "step": 10314
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.541789054870605,
+      "learning_rate": 3.4468851896344606e-06,
+      "loss": 0.5652,
+      "step": 10315
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.008191108703613,
+      "learning_rate": 3.444310966191865e-06,
+      "loss": 0.3386,
+      "step": 10316
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.454290390014648,
+      "learning_rate": 3.4417367427492706e-06,
+      "loss": 0.5166,
+      "step": 10317
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.802310943603516,
+      "learning_rate": 3.439162519306676e-06,
+      "loss": 0.3082,
+      "step": 10318
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.244376182556152,
+      "learning_rate": 3.436588295864081e-06,
+      "loss": 0.3266,
+      "step": 10319
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.181434631347656,
+      "learning_rate": 3.4340140724214864e-06,
+      "loss": 0.4626,
+      "step": 10320
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.8023176193237305,
+      "learning_rate": 3.4314398489788914e-06,
+      "loss": 0.3633,
+      "step": 10321
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.527910232543945,
+      "learning_rate": 3.428865625536297e-06,
+      "loss": 0.4064,
+      "step": 10322
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.953227043151855,
+      "learning_rate": 3.426291402093702e-06,
+      "loss": 0.2656,
+      "step": 10323
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.34505558013916,
+      "learning_rate": 3.423717178651107e-06,
+      "loss": 0.4373,
+      "step": 10324
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.77633285522461,
+      "learning_rate": 3.4211429552085122e-06,
+      "loss": 0.3254,
+      "step": 10325
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.321101188659668,
+      "learning_rate": 3.4185687317659176e-06,
+      "loss": 0.4738,
+      "step": 10326
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.482454299926758,
+      "learning_rate": 3.4159945083233226e-06,
+      "loss": 0.3438,
+      "step": 10327
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.79236125946045,
+      "learning_rate": 3.413420284880728e-06,
+      "loss": 0.2604,
+      "step": 10328
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.345510482788086,
+      "learning_rate": 3.4108460614381326e-06,
+      "loss": 0.5073,
+      "step": 10329
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.138515472412109,
+      "learning_rate": 3.408271837995538e-06,
+      "loss": 0.3654,
+      "step": 10330
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.56738805770874,
+      "learning_rate": 3.4056976145529434e-06,
+      "loss": 0.2656,
+      "step": 10331
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.79387378692627,
+      "learning_rate": 3.4031233911103484e-06,
+      "loss": 0.4011,
+      "step": 10332
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.253780364990234,
+      "learning_rate": 3.400549167667754e-06,
+      "loss": 0.3602,
+      "step": 10333
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.453286170959473,
+      "learning_rate": 3.397974944225159e-06,
+      "loss": 0.3837,
+      "step": 10334
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.63301944732666,
+      "learning_rate": 3.3954007207825638e-06,
+      "loss": 0.3533,
+      "step": 10335
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.314065933227539,
+      "learning_rate": 3.392826497339969e-06,
+      "loss": 0.2284,
+      "step": 10336
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.389789581298828,
+      "learning_rate": 3.390252273897374e-06,
+      "loss": 0.2565,
+      "step": 10337
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.987936019897461,
+      "learning_rate": 3.3876780504547796e-06,
+      "loss": 0.197,
+      "step": 10338
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.762155532836914,
+      "learning_rate": 3.385103827012185e-06,
+      "loss": 0.3853,
+      "step": 10339
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.206766128540039,
+      "learning_rate": 3.38252960356959e-06,
+      "loss": 0.3187,
+      "step": 10340
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.465702056884766,
+      "learning_rate": 3.3799553801269954e-06,
+      "loss": 0.527,
+      "step": 10341
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.702652931213379,
+      "learning_rate": 3.3773811566844e-06,
+      "loss": 0.3219,
+      "step": 10342
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.759143829345703,
+      "learning_rate": 3.3748069332418054e-06,
+      "loss": 0.4821,
+      "step": 10343
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.18113899230957,
+      "learning_rate": 3.3722327097992108e-06,
+      "loss": 0.3623,
+      "step": 10344
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.89564323425293,
+      "learning_rate": 3.3696584863566158e-06,
+      "loss": 0.4238,
+      "step": 10345
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.624650955200195,
+      "learning_rate": 3.367084262914021e-06,
+      "loss": 0.3402,
+      "step": 10346
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.608189582824707,
+      "learning_rate": 3.364510039471426e-06,
+      "loss": 0.4109,
+      "step": 10347
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.358651161193848,
+      "learning_rate": 3.361935816028831e-06,
+      "loss": 0.3574,
+      "step": 10348
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.220965385437012,
+      "learning_rate": 3.3593615925862366e-06,
+      "loss": 0.3318,
+      "step": 10349
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.647798538208008,
+      "learning_rate": 3.3567873691436416e-06,
+      "loss": 0.3023,
+      "step": 10350
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.34692668914795,
+      "learning_rate": 3.354213145701047e-06,
+      "loss": 0.3374,
+      "step": 10351
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.064382553100586,
+      "learning_rate": 3.3516389222584524e-06,
+      "loss": 0.2876,
+      "step": 10352
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.457650184631348,
+      "learning_rate": 3.3490646988158574e-06,
+      "loss": 0.3867,
+      "step": 10353
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.781315803527832,
+      "learning_rate": 3.3464904753732628e-06,
+      "loss": 0.4221,
+      "step": 10354
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.486967086791992,
+      "learning_rate": 3.3439162519306673e-06,
+      "loss": 0.3239,
+      "step": 10355
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.00626277923584,
+      "learning_rate": 3.3413420284880728e-06,
+      "loss": 0.3696,
+      "step": 10356
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.245110511779785,
+      "learning_rate": 3.338767805045478e-06,
+      "loss": 0.3123,
+      "step": 10357
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.767106056213379,
+      "learning_rate": 3.336193581602883e-06,
+      "loss": 0.4389,
+      "step": 10358
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.81732177734375,
+      "learning_rate": 3.3336193581602886e-06,
+      "loss": 0.4039,
+      "step": 10359
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.436320304870605,
+      "learning_rate": 3.3310451347176935e-06,
+      "loss": 0.3122,
+      "step": 10360
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.913230895996094,
+      "learning_rate": 3.3284709112750985e-06,
+      "loss": 0.3536,
+      "step": 10361
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.325385093688965,
+      "learning_rate": 3.325896687832504e-06,
+      "loss": 0.3868,
+      "step": 10362
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.295690536499023,
+      "learning_rate": 3.323322464389909e-06,
+      "loss": 0.4019,
+      "step": 10363
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.48462200164795,
+      "learning_rate": 3.3207482409473143e-06,
+      "loss": 0.3442,
+      "step": 10364
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.8226318359375,
+      "learning_rate": 3.3181740175047198e-06,
+      "loss": 0.3471,
+      "step": 10365
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.76678466796875,
+      "learning_rate": 3.3155997940621247e-06,
+      "loss": 0.3863,
+      "step": 10366
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.450419902801514,
+      "learning_rate": 3.31302557061953e-06,
+      "loss": 0.3797,
+      "step": 10367
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.025487899780273,
+      "learning_rate": 3.3104513471769347e-06,
+      "loss": 0.4222,
+      "step": 10368
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.639636039733887,
+      "learning_rate": 3.30787712373434e-06,
+      "loss": 0.3518,
+      "step": 10369
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.625659942626953,
+      "learning_rate": 3.3053029002917455e-06,
+      "loss": 0.3278,
+      "step": 10370
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.003331184387207,
+      "learning_rate": 3.3027286768491505e-06,
+      "loss": 0.3522,
+      "step": 10371
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.612837314605713,
+      "learning_rate": 3.300154453406556e-06,
+      "loss": 0.2022,
+      "step": 10372
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.15909481048584,
+      "learning_rate": 3.297580229963961e-06,
+      "loss": 0.4,
+      "step": 10373
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.829818725585938,
+      "learning_rate": 3.295006006521366e-06,
+      "loss": 0.3181,
+      "step": 10374
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.90761661529541,
+      "learning_rate": 3.2924317830787713e-06,
+      "loss": 0.3898,
+      "step": 10375
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.742633819580078,
+      "learning_rate": 3.2898575596361763e-06,
+      "loss": 0.519,
+      "step": 10376
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.075239181518555,
+      "learning_rate": 3.2872833361935817e-06,
+      "loss": 0.4766,
+      "step": 10377
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.430364608764648,
+      "learning_rate": 3.284709112750987e-06,
+      "loss": 0.4269,
+      "step": 10378
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.02288818359375,
+      "learning_rate": 3.282134889308392e-06,
+      "loss": 0.3027,
+      "step": 10379
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.890666007995605,
+      "learning_rate": 3.2795606658657975e-06,
+      "loss": 0.2773,
+      "step": 10380
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.060396194458008,
+      "learning_rate": 3.276986442423202e-06,
+      "loss": 0.4201,
+      "step": 10381
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.831722259521484,
+      "learning_rate": 3.2744122189806075e-06,
+      "loss": 0.4713,
+      "step": 10382
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.686012268066406,
+      "learning_rate": 3.271837995538013e-06,
+      "loss": 0.2566,
+      "step": 10383
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.444820404052734,
+      "learning_rate": 3.269263772095418e-06,
+      "loss": 0.5268,
+      "step": 10384
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.96072006225586,
+      "learning_rate": 3.2666895486528233e-06,
+      "loss": 0.423,
+      "step": 10385
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.017065048217773,
+      "learning_rate": 3.2641153252102283e-06,
+      "loss": 0.5799,
+      "step": 10386
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.117496013641357,
+      "learning_rate": 3.2615411017676333e-06,
+      "loss": 0.232,
+      "step": 10387
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.22362995147705,
+      "learning_rate": 3.2589668783250387e-06,
+      "loss": 0.3084,
+      "step": 10388
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.4487943649292,
+      "learning_rate": 3.2563926548824437e-06,
+      "loss": 0.2844,
+      "step": 10389
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.250666618347168,
+      "learning_rate": 3.253818431439849e-06,
+      "loss": 0.3653,
+      "step": 10390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.97622299194336,
+      "learning_rate": 3.2512442079972545e-06,
+      "loss": 0.4187,
+      "step": 10391
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.81460952758789,
+      "learning_rate": 3.2486699845546595e-06,
+      "loss": 0.3936,
+      "step": 10392
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.792677879333496,
+      "learning_rate": 3.246095761112065e-06,
+      "loss": 0.5237,
+      "step": 10393
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.203180313110352,
+      "learning_rate": 3.2435215376694695e-06,
+      "loss": 0.2558,
+      "step": 10394
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.556968688964844,
+      "learning_rate": 3.240947314226875e-06,
+      "loss": 0.4362,
+      "step": 10395
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.968944549560547,
+      "learning_rate": 3.2383730907842803e-06,
+      "loss": 0.511,
+      "step": 10396
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.981884002685547,
+      "learning_rate": 3.2357988673416853e-06,
+      "loss": 0.4566,
+      "step": 10397
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.601639747619629,
+      "learning_rate": 3.2332246438990907e-06,
+      "loss": 0.3828,
+      "step": 10398
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.621521949768066,
+      "learning_rate": 3.2306504204564957e-06,
+      "loss": 0.3328,
+      "step": 10399
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.16479778289795,
+      "learning_rate": 3.2280761970139007e-06,
+      "loss": 0.4389,
+      "step": 10400
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.212334632873535,
+      "learning_rate": 3.225501973571306e-06,
+      "loss": 0.3302,
+      "step": 10401
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.324894905090332,
+      "learning_rate": 3.222927750128711e-06,
+      "loss": 0.4328,
+      "step": 10402
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.898311138153076,
+      "learning_rate": 3.2203535266861165e-06,
+      "loss": 0.3092,
+      "step": 10403
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.091353416442871,
+      "learning_rate": 3.217779303243522e-06,
+      "loss": 0.4459,
+      "step": 10404
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.136836528778076,
+      "learning_rate": 3.215205079800927e-06,
+      "loss": 0.277,
+      "step": 10405
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.946653366088867,
+      "learning_rate": 3.2126308563583323e-06,
+      "loss": 0.3875,
+      "step": 10406
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.466854095458984,
+      "learning_rate": 3.210056632915737e-06,
+      "loss": 0.4126,
+      "step": 10407
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.457801818847656,
+      "learning_rate": 3.2074824094731423e-06,
+      "loss": 0.4534,
+      "step": 10408
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.460442543029785,
+      "learning_rate": 3.2049081860305477e-06,
+      "loss": 0.4763,
+      "step": 10409
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.811309814453125,
+      "learning_rate": 3.2023339625879527e-06,
+      "loss": 0.4383,
+      "step": 10410
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.980772018432617,
+      "learning_rate": 3.199759739145358e-06,
+      "loss": 0.4323,
+      "step": 10411
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.63910961151123,
+      "learning_rate": 3.197185515702763e-06,
+      "loss": 0.3878,
+      "step": 10412
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.61581802368164,
+      "learning_rate": 3.194611292260168e-06,
+      "loss": 0.414,
+      "step": 10413
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.028787612915039,
+      "learning_rate": 3.1920370688175735e-06,
+      "loss": 0.3732,
+      "step": 10414
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.4498291015625,
+      "learning_rate": 3.1894628453749784e-06,
+      "loss": 0.4567,
+      "step": 10415
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.622383117675781,
+      "learning_rate": 3.186888621932384e-06,
+      "loss": 0.3014,
+      "step": 10416
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.299242973327637,
+      "learning_rate": 3.1843143984897893e-06,
+      "loss": 0.3553,
+      "step": 10417
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.99066162109375,
+      "learning_rate": 3.1817401750471942e-06,
+      "loss": 0.4837,
+      "step": 10418
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.404476642608643,
+      "learning_rate": 3.1791659516045997e-06,
+      "loss": 0.2378,
+      "step": 10419
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.202580451965332,
+      "learning_rate": 3.1765917281620042e-06,
+      "loss": 0.3031,
+      "step": 10420
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.294025421142578,
+      "learning_rate": 3.1740175047194096e-06,
+      "loss": 0.3783,
+      "step": 10421
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.027019500732422,
+      "learning_rate": 3.171443281276815e-06,
+      "loss": 0.3435,
+      "step": 10422
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.20676326751709,
+      "learning_rate": 3.16886905783422e-06,
+      "loss": 0.3646,
+      "step": 10423
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.397150993347168,
+      "learning_rate": 3.1662948343916254e-06,
+      "loss": 0.3074,
+      "step": 10424
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.634303092956543,
+      "learning_rate": 3.1637206109490304e-06,
+      "loss": 0.4436,
+      "step": 10425
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.696736335754395,
+      "learning_rate": 3.1611463875064354e-06,
+      "loss": 0.4129,
+      "step": 10426
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.378373146057129,
+      "learning_rate": 3.158572164063841e-06,
+      "loss": 0.318,
+      "step": 10427
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.294855117797852,
+      "learning_rate": 3.155997940621246e-06,
+      "loss": 0.4607,
+      "step": 10428
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.040972709655762,
+      "learning_rate": 3.1534237171786512e-06,
+      "loss": 0.3901,
+      "step": 10429
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.023202896118164,
+      "learning_rate": 3.1508494937360566e-06,
+      "loss": 0.4375,
+      "step": 10430
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.713528633117676,
+      "learning_rate": 3.1482752702934616e-06,
+      "loss": 0.5196,
+      "step": 10431
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.742056846618652,
+      "learning_rate": 3.145701046850867e-06,
+      "loss": 0.3954,
+      "step": 10432
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.226908683776855,
+      "learning_rate": 3.1431268234082716e-06,
+      "loss": 0.343,
+      "step": 10433
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.437498092651367,
+      "learning_rate": 3.140552599965677e-06,
+      "loss": 0.2502,
+      "step": 10434
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.046233177185059,
+      "learning_rate": 3.1379783765230824e-06,
+      "loss": 0.3234,
+      "step": 10435
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.71757984161377,
+      "learning_rate": 3.1354041530804874e-06,
+      "loss": 0.3604,
+      "step": 10436
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.548288345336914,
+      "learning_rate": 3.132829929637893e-06,
+      "loss": 0.4339,
+      "step": 10437
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.2647066116333,
+      "learning_rate": 3.130255706195298e-06,
+      "loss": 0.3799,
+      "step": 10438
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.122357368469238,
+      "learning_rate": 3.127681482752703e-06,
+      "loss": 0.2217,
+      "step": 10439
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.940947532653809,
+      "learning_rate": 3.125107259310108e-06,
+      "loss": 0.2518,
+      "step": 10440
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.62997055053711,
+      "learning_rate": 3.122533035867513e-06,
+      "loss": 0.3733,
+      "step": 10441
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.836966037750244,
+      "learning_rate": 3.1199588124249186e-06,
+      "loss": 0.1797,
+      "step": 10442
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.723001480102539,
+      "learning_rate": 3.117384588982324e-06,
+      "loss": 0.4777,
+      "step": 10443
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.959872722625732,
+      "learning_rate": 3.114810365539729e-06,
+      "loss": 0.2769,
+      "step": 10444
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.834935188293457,
+      "learning_rate": 3.1122361420971344e-06,
+      "loss": 0.3563,
+      "step": 10445
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.509300231933594,
+      "learning_rate": 3.109661918654539e-06,
+      "loss": 0.3915,
+      "step": 10446
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.07111930847168,
+      "learning_rate": 3.1070876952119444e-06,
+      "loss": 0.3267,
+      "step": 10447
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.043294906616211,
+      "learning_rate": 3.10451347176935e-06,
+      "loss": 0.4405,
+      "step": 10448
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.18427562713623,
+      "learning_rate": 3.1019392483267548e-06,
+      "loss": 0.5181,
+      "step": 10449
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.779806137084961,
+      "learning_rate": 3.09936502488416e-06,
+      "loss": 0.2407,
+      "step": 10450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.736502647399902,
+      "learning_rate": 3.096790801441565e-06,
+      "loss": 0.3411,
+      "step": 10451
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.774645805358887,
+      "learning_rate": 3.09421657799897e-06,
+      "loss": 0.4571,
+      "step": 10452
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.356518745422363,
+      "learning_rate": 3.0916423545563756e-06,
+      "loss": 0.4493,
+      "step": 10453
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.229302406311035,
+      "learning_rate": 3.0890681311137806e-06,
+      "loss": 0.3212,
+      "step": 10454
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.343244552612305,
+      "learning_rate": 3.086493907671186e-06,
+      "loss": 0.2629,
+      "step": 10455
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.548308372497559,
+      "learning_rate": 3.0839196842285914e-06,
+      "loss": 0.409,
+      "step": 10456
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.149595260620117,
+      "learning_rate": 3.0813454607859964e-06,
+      "loss": 0.3767,
+      "step": 10457
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.739317893981934,
+      "learning_rate": 3.078771237343402e-06,
+      "loss": 0.5334,
+      "step": 10458
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.763712882995605,
+      "learning_rate": 3.0761970139008064e-06,
+      "loss": 0.335,
+      "step": 10459
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.9581499099731445,
+      "learning_rate": 3.0736227904582118e-06,
+      "loss": 0.287,
+      "step": 10460
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.953165054321289,
+      "learning_rate": 3.071048567015617e-06,
+      "loss": 0.3215,
+      "step": 10461
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.7612943649292,
+      "learning_rate": 3.068474343573022e-06,
+      "loss": 0.3962,
+      "step": 10462
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.252572059631348,
+      "learning_rate": 3.0659001201304276e-06,
+      "loss": 0.3712,
+      "step": 10463
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.732776641845703,
+      "learning_rate": 3.0633258966878326e-06,
+      "loss": 0.4333,
+      "step": 10464
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.15046501159668,
+      "learning_rate": 3.0607516732452375e-06,
+      "loss": 0.4144,
+      "step": 10465
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.169456481933594,
+      "learning_rate": 3.058177449802643e-06,
+      "loss": 0.3307,
+      "step": 10466
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.577850341796875,
+      "learning_rate": 3.055603226360048e-06,
+      "loss": 0.539,
+      "step": 10467
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.831425666809082,
+      "learning_rate": 3.0530290029174534e-06,
+      "loss": 0.5813,
+      "step": 10468
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.429529190063477,
+      "learning_rate": 3.0504547794748588e-06,
+      "loss": 0.6707,
+      "step": 10469
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.450451850891113,
+      "learning_rate": 3.0478805560322638e-06,
+      "loss": 0.485,
+      "step": 10470
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.50145149230957,
+      "learning_rate": 3.045306332589669e-06,
+      "loss": 0.3956,
+      "step": 10471
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.277445793151855,
+      "learning_rate": 3.0427321091470737e-06,
+      "loss": 0.2804,
+      "step": 10472
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.106534004211426,
+      "learning_rate": 3.040157885704479e-06,
+      "loss": 0.4831,
+      "step": 10473
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.272245407104492,
+      "learning_rate": 3.0375836622618846e-06,
+      "loss": 0.434,
+      "step": 10474
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.200016021728516,
+      "learning_rate": 3.0350094388192895e-06,
+      "loss": 0.3605,
+      "step": 10475
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.604349136352539,
+      "learning_rate": 3.032435215376695e-06,
+      "loss": 0.319,
+      "step": 10476
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.937857627868652,
+      "learning_rate": 3.0298609919341e-06,
+      "loss": 0.4413,
+      "step": 10477
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.702960968017578,
+      "learning_rate": 3.027286768491505e-06,
+      "loss": 0.3216,
+      "step": 10478
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.42029857635498,
+      "learning_rate": 3.0247125450489103e-06,
+      "loss": 0.3126,
+      "step": 10479
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.8287992477417,
+      "learning_rate": 3.0221383216063153e-06,
+      "loss": 0.4181,
+      "step": 10480
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.196413040161133,
+      "learning_rate": 3.0195640981637207e-06,
+      "loss": 0.3181,
+      "step": 10481
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.274801254272461,
+      "learning_rate": 3.016989874721126e-06,
+      "loss": 0.3049,
+      "step": 10482
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.921868324279785,
+      "learning_rate": 3.014415651278531e-06,
+      "loss": 0.4739,
+      "step": 10483
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.26804256439209,
+      "learning_rate": 3.0118414278359365e-06,
+      "loss": 0.3415,
+      "step": 10484
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.426739692687988,
+      "learning_rate": 3.009267204393341e-06,
+      "loss": 0.4155,
+      "step": 10485
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.302928924560547,
+      "learning_rate": 3.0066929809507465e-06,
+      "loss": 0.4515,
+      "step": 10486
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.730430603027344,
+      "learning_rate": 3.004118757508152e-06,
+      "loss": 0.3515,
+      "step": 10487
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.60925579071045,
+      "learning_rate": 3.001544534065557e-06,
+      "loss": 0.4456,
+      "step": 10488
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.795517921447754,
+      "learning_rate": 2.9989703106229623e-06,
+      "loss": 0.5469,
+      "step": 10489
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.022937774658203,
+      "learning_rate": 2.9963960871803673e-06,
+      "loss": 0.3824,
+      "step": 10490
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.785831451416016,
+      "learning_rate": 2.9938218637377723e-06,
+      "loss": 0.6176,
+      "step": 10491
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.09007740020752,
+      "learning_rate": 2.9912476402951777e-06,
+      "loss": 0.3668,
+      "step": 10492
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.240140914916992,
+      "learning_rate": 2.9886734168525827e-06,
+      "loss": 0.3375,
+      "step": 10493
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.147300720214844,
+      "learning_rate": 2.986099193409988e-06,
+      "loss": 0.2585,
+      "step": 10494
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.183375358581543,
+      "learning_rate": 2.9835249699673935e-06,
+      "loss": 0.324,
+      "step": 10495
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.306490898132324,
+      "learning_rate": 2.9809507465247985e-06,
+      "loss": 0.4229,
+      "step": 10496
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.241902351379395,
+      "learning_rate": 2.978376523082204e-06,
+      "loss": 0.4034,
+      "step": 10497
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.103628158569336,
+      "learning_rate": 2.9758022996396085e-06,
+      "loss": 0.4971,
+      "step": 10498
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.960226058959961,
+      "learning_rate": 2.973228076197014e-06,
+      "loss": 0.4651,
+      "step": 10499
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.878409385681152,
+      "learning_rate": 2.9706538527544193e-06,
+      "loss": 0.4128,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.149812698364258,
+      "learning_rate": 2.9680796293118243e-06,
+      "loss": 0.3229,
+      "step": 10501
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.156515121459961,
+      "learning_rate": 2.9655054058692297e-06,
+      "loss": 0.2426,
+      "step": 10502
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.101290702819824,
+      "learning_rate": 2.9629311824266347e-06,
+      "loss": 0.5479,
+      "step": 10503
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.113824844360352,
+      "learning_rate": 2.9603569589840397e-06,
+      "loss": 0.2417,
+      "step": 10504
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.302640914916992,
+      "learning_rate": 2.957782735541445e-06,
+      "loss": 0.471,
+      "step": 10505
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.096059799194336,
+      "learning_rate": 2.95520851209885e-06,
+      "loss": 0.3605,
+      "step": 10506
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.7495698928833,
+      "learning_rate": 2.9526342886562555e-06,
+      "loss": 0.3854,
+      "step": 10507
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.65858268737793,
+      "learning_rate": 2.950060065213661e-06,
+      "loss": 0.2668,
+      "step": 10508
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.74853515625,
+      "learning_rate": 2.947485841771066e-06,
+      "loss": 0.328,
+      "step": 10509
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.641555786132812,
+      "learning_rate": 2.944911618328471e-06,
+      "loss": 0.4549,
+      "step": 10510
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.706530570983887,
+      "learning_rate": 2.942337394885876e-06,
+      "loss": 0.4255,
+      "step": 10511
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.585078239440918,
+      "learning_rate": 2.9397631714432813e-06,
+      "loss": 0.4299,
+      "step": 10512
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.333610534667969,
+      "learning_rate": 2.9371889480006867e-06,
+      "loss": 0.1562,
+      "step": 10513
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.311840057373047,
+      "learning_rate": 2.9346147245580917e-06,
+      "loss": 0.2095,
+      "step": 10514
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.91805648803711,
+      "learning_rate": 2.932040501115497e-06,
+      "loss": 0.5584,
+      "step": 10515
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.733766555786133,
+      "learning_rate": 2.929466277672902e-06,
+      "loss": 0.3841,
+      "step": 10516
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.359426498413086,
+      "learning_rate": 2.926892054230307e-06,
+      "loss": 0.3687,
+      "step": 10517
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.029149055480957,
+      "learning_rate": 2.9243178307877125e-06,
+      "loss": 0.2796,
+      "step": 10518
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.420923233032227,
+      "learning_rate": 2.9217436073451175e-06,
+      "loss": 0.3452,
+      "step": 10519
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.36894702911377,
+      "learning_rate": 2.919169383902523e-06,
+      "loss": 0.3778,
+      "step": 10520
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.38289737701416,
+      "learning_rate": 2.9165951604599283e-06,
+      "loss": 0.3045,
+      "step": 10521
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.05566692352295,
+      "learning_rate": 2.9140209370173333e-06,
+      "loss": 0.2796,
+      "step": 10522
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.177206039428711,
+      "learning_rate": 2.9114467135747382e-06,
+      "loss": 0.4427,
+      "step": 10523
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.995177268981934,
+      "learning_rate": 2.9088724901321432e-06,
+      "loss": 0.3749,
+      "step": 10524
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.074337005615234,
+      "learning_rate": 2.9062982666895486e-06,
+      "loss": 0.4551,
+      "step": 10525
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.752120971679688,
+      "learning_rate": 2.903724043246954e-06,
+      "loss": 0.3661,
+      "step": 10526
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.213761329650879,
+      "learning_rate": 2.901149819804359e-06,
+      "loss": 0.2887,
+      "step": 10527
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.839393615722656,
+      "learning_rate": 2.8985755963617645e-06,
+      "loss": 0.2288,
+      "step": 10528
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.381450653076172,
+      "learning_rate": 2.8960013729191694e-06,
+      "loss": 0.3318,
+      "step": 10529
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.615515232086182,
+      "learning_rate": 2.8934271494765744e-06,
+      "loss": 0.2597,
+      "step": 10530
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.628311157226562,
+      "learning_rate": 2.89085292603398e-06,
+      "loss": 0.4385,
+      "step": 10531
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.040813446044922,
+      "learning_rate": 2.888278702591385e-06,
+      "loss": 0.3115,
+      "step": 10532
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.033926963806152,
+      "learning_rate": 2.8857044791487902e-06,
+      "loss": 0.4985,
+      "step": 10533
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.547693252563477,
+      "learning_rate": 2.8831302557061956e-06,
+      "loss": 0.3176,
+      "step": 10534
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.174239158630371,
+      "learning_rate": 2.8805560322636006e-06,
+      "loss": 0.4017,
+      "step": 10535
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.291644096374512,
+      "learning_rate": 2.8779818088210056e-06,
+      "loss": 0.3642,
+      "step": 10536
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.902597427368164,
+      "learning_rate": 2.8754075853784106e-06,
+      "loss": 0.3852,
+      "step": 10537
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.805796146392822,
+      "learning_rate": 2.872833361935816e-06,
+      "loss": 0.2278,
+      "step": 10538
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.182024002075195,
+      "learning_rate": 2.8702591384932214e-06,
+      "loss": 0.3623,
+      "step": 10539
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.437385559082031,
+      "learning_rate": 2.8676849150506264e-06,
+      "loss": 0.4775,
+      "step": 10540
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.261222839355469,
+      "learning_rate": 2.865110691608032e-06,
+      "loss": 0.3547,
+      "step": 10541
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.4518404006958,
+      "learning_rate": 2.862536468165437e-06,
+      "loss": 0.3631,
+      "step": 10542
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.53903865814209,
+      "learning_rate": 2.859962244722842e-06,
+      "loss": 0.2195,
+      "step": 10543
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.522038459777832,
+      "learning_rate": 2.8573880212802472e-06,
+      "loss": 0.2779,
+      "step": 10544
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.359233856201172,
+      "learning_rate": 2.854813797837652e-06,
+      "loss": 0.4727,
+      "step": 10545
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.91067886352539,
+      "learning_rate": 2.8522395743950576e-06,
+      "loss": 0.3306,
+      "step": 10546
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.986410140991211,
+      "learning_rate": 2.849665350952463e-06,
+      "loss": 0.3438,
+      "step": 10547
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.411774635314941,
+      "learning_rate": 2.847091127509868e-06,
+      "loss": 0.2605,
+      "step": 10548
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.776211738586426,
+      "learning_rate": 2.844516904067273e-06,
+      "loss": 0.2656,
+      "step": 10549
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.012097358703613,
+      "learning_rate": 2.841942680624678e-06,
+      "loss": 0.419,
+      "step": 10550
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.288351058959961,
+      "learning_rate": 2.8393684571820834e-06,
+      "loss": 0.2623,
+      "step": 10551
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.65538215637207,
+      "learning_rate": 2.836794233739489e-06,
+      "loss": 0.4126,
+      "step": 10552
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.261811256408691,
+      "learning_rate": 2.834220010296894e-06,
+      "loss": 0.346,
+      "step": 10553
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.221818923950195,
+      "learning_rate": 2.831645786854299e-06,
+      "loss": 0.2815,
+      "step": 10554
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.045987129211426,
+      "learning_rate": 2.829071563411704e-06,
+      "loss": 0.4013,
+      "step": 10555
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.827857971191406,
+      "learning_rate": 2.826497339969109e-06,
+      "loss": 0.3692,
+      "step": 10556
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.186592102050781,
+      "learning_rate": 2.8239231165265146e-06,
+      "loss": 0.3897,
+      "step": 10557
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.697921752929688,
+      "learning_rate": 2.8213488930839196e-06,
+      "loss": 0.3147,
+      "step": 10558
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.759190559387207,
+      "learning_rate": 2.818774669641325e-06,
+      "loss": 0.3754,
+      "step": 10559
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.325169563293457,
+      "learning_rate": 2.8162004461987304e-06,
+      "loss": 0.4412,
+      "step": 10560
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.071754455566406,
+      "learning_rate": 2.8136262227561354e-06,
+      "loss": 0.4155,
+      "step": 10561
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.39793872833252,
+      "learning_rate": 2.8110519993135404e-06,
+      "loss": 0.4341,
+      "step": 10562
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.861881256103516,
+      "learning_rate": 2.8084777758709454e-06,
+      "loss": 0.5423,
+      "step": 10563
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.513286590576172,
+      "learning_rate": 2.8059035524283508e-06,
+      "loss": 0.4751,
+      "step": 10564
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.263647079467773,
+      "learning_rate": 2.803329328985756e-06,
+      "loss": 0.3686,
+      "step": 10565
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.746999740600586,
+      "learning_rate": 2.800755105543161e-06,
+      "loss": 0.3899,
+      "step": 10566
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.294825553894043,
+      "learning_rate": 2.7981808821005666e-06,
+      "loss": 0.4866,
+      "step": 10567
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.171945571899414,
+      "learning_rate": 2.7956066586579716e-06,
+      "loss": 0.5816,
+      "step": 10568
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.502220153808594,
+      "learning_rate": 2.7930324352153766e-06,
+      "loss": 0.4326,
+      "step": 10569
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.98552131652832,
+      "learning_rate": 2.790458211772782e-06,
+      "loss": 0.436,
+      "step": 10570
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.707756996154785,
+      "learning_rate": 2.787883988330187e-06,
+      "loss": 0.2927,
+      "step": 10571
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.05388069152832,
+      "learning_rate": 2.7853097648875924e-06,
+      "loss": 0.3858,
+      "step": 10572
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.113991737365723,
+      "learning_rate": 2.7827355414449978e-06,
+      "loss": 0.465,
+      "step": 10573
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.309956550598145,
+      "learning_rate": 2.7801613180024028e-06,
+      "loss": 0.3354,
+      "step": 10574
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.145524024963379,
+      "learning_rate": 2.7775870945598078e-06,
+      "loss": 0.2914,
+      "step": 10575
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.339261054992676,
+      "learning_rate": 2.7750128711172127e-06,
+      "loss": 0.476,
+      "step": 10576
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.763864994049072,
+      "learning_rate": 2.772438647674618e-06,
+      "loss": 0.2999,
+      "step": 10577
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.158241271972656,
+      "learning_rate": 2.7698644242320236e-06,
+      "loss": 0.5204,
+      "step": 10578
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.755532264709473,
+      "learning_rate": 2.7672902007894285e-06,
+      "loss": 0.346,
+      "step": 10579
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.783491134643555,
+      "learning_rate": 2.764715977346834e-06,
+      "loss": 0.421,
+      "step": 10580
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.496931076049805,
+      "learning_rate": 2.762141753904239e-06,
+      "loss": 0.4242,
+      "step": 10581
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.753774642944336,
+      "learning_rate": 2.759567530461644e-06,
+      "loss": 0.2941,
+      "step": 10582
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.52796745300293,
+      "learning_rate": 2.7569933070190493e-06,
+      "loss": 0.3719,
+      "step": 10583
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.800555229187012,
+      "learning_rate": 2.7544190835764543e-06,
+      "loss": 0.2398,
+      "step": 10584
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.637390613555908,
+      "learning_rate": 2.7518448601338597e-06,
+      "loss": 0.2663,
+      "step": 10585
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.037017822265625,
+      "learning_rate": 2.749270636691265e-06,
+      "loss": 0.4499,
+      "step": 10586
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.532233238220215,
+      "learning_rate": 2.74669641324867e-06,
+      "loss": 0.2082,
+      "step": 10587
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.800809860229492,
+      "learning_rate": 2.744122189806075e-06,
+      "loss": 0.3746,
+      "step": 10588
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.036859512329102,
+      "learning_rate": 2.74154796636348e-06,
+      "loss": 0.338,
+      "step": 10589
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.914230346679688,
+      "learning_rate": 2.7389737429208855e-06,
+      "loss": 0.4457,
+      "step": 10590
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.960654258728027,
+      "learning_rate": 2.736399519478291e-06,
+      "loss": 0.243,
+      "step": 10591
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.706157684326172,
+      "learning_rate": 2.733825296035696e-06,
+      "loss": 0.5498,
+      "step": 10592
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.224727630615234,
+      "learning_rate": 2.7312510725931013e-06,
+      "loss": 0.3512,
+      "step": 10593
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.958061218261719,
+      "learning_rate": 2.7286768491505063e-06,
+      "loss": 0.2833,
+      "step": 10594
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.539462089538574,
+      "learning_rate": 2.7261026257079113e-06,
+      "loss": 0.3899,
+      "step": 10595
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.695630073547363,
+      "learning_rate": 2.7235284022653167e-06,
+      "loss": 0.6017,
+      "step": 10596
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.647037506103516,
+      "learning_rate": 2.7209541788227217e-06,
+      "loss": 0.3946,
+      "step": 10597
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.245928764343262,
+      "learning_rate": 2.718379955380127e-06,
+      "loss": 0.439,
+      "step": 10598
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.757686614990234,
+      "learning_rate": 2.7158057319375325e-06,
+      "loss": 0.2704,
+      "step": 10599
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.669482231140137,
+      "learning_rate": 2.7132315084949375e-06,
+      "loss": 0.3674,
+      "step": 10600
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.072402954101562,
+      "learning_rate": 2.7106572850523425e-06,
+      "loss": 0.411,
+      "step": 10601
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.48277473449707,
+      "learning_rate": 2.7080830616097475e-06,
+      "loss": 0.5575,
+      "step": 10602
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.001754760742188,
+      "learning_rate": 2.705508838167153e-06,
+      "loss": 0.4023,
+      "step": 10603
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.322209358215332,
+      "learning_rate": 2.7029346147245583e-06,
+      "loss": 0.3281,
+      "step": 10604
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.982292175292969,
+      "learning_rate": 2.7003603912819633e-06,
+      "loss": 0.3397,
+      "step": 10605
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.836981773376465,
+      "learning_rate": 2.6977861678393687e-06,
+      "loss": 0.2666,
+      "step": 10606
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.477084159851074,
+      "learning_rate": 2.6952119443967737e-06,
+      "loss": 0.338,
+      "step": 10607
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.189910888671875,
+      "learning_rate": 2.6926377209541787e-06,
+      "loss": 0.3507,
+      "step": 10608
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.914125442504883,
+      "learning_rate": 2.690063497511584e-06,
+      "loss": 0.3637,
+      "step": 10609
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.882555961608887,
+      "learning_rate": 2.687489274068989e-06,
+      "loss": 0.4148,
+      "step": 10610
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.456730842590332,
+      "learning_rate": 2.6849150506263945e-06,
+      "loss": 0.2627,
+      "step": 10611
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.048463821411133,
+      "learning_rate": 2.6823408271838e-06,
+      "loss": 0.3487,
+      "step": 10612
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.227310180664062,
+      "learning_rate": 2.679766603741205e-06,
+      "loss": 0.2775,
+      "step": 10613
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.190180778503418,
+      "learning_rate": 2.67719238029861e-06,
+      "loss": 0.3063,
+      "step": 10614
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.883695602416992,
+      "learning_rate": 2.674618156856015e-06,
+      "loss": 0.4502,
+      "step": 10615
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.71688461303711,
+      "learning_rate": 2.6720439334134203e-06,
+      "loss": 0.5068,
+      "step": 10616
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.551656723022461,
+      "learning_rate": 2.6694697099708257e-06,
+      "loss": 0.2955,
+      "step": 10617
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.369060516357422,
+      "learning_rate": 2.6668954865282307e-06,
+      "loss": 0.2576,
+      "step": 10618
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.58068561553955,
+      "learning_rate": 2.664321263085636e-06,
+      "loss": 0.3137,
+      "step": 10619
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.581269264221191,
+      "learning_rate": 2.661747039643041e-06,
+      "loss": 0.3654,
+      "step": 10620
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.775806427001953,
+      "learning_rate": 2.659172816200446e-06,
+      "loss": 0.2647,
+      "step": 10621
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.271175384521484,
+      "learning_rate": 2.6565985927578515e-06,
+      "loss": 0.3589,
+      "step": 10622
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.722084999084473,
+      "learning_rate": 2.6540243693152565e-06,
+      "loss": 0.3768,
+      "step": 10623
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.016916275024414,
+      "learning_rate": 2.651450145872662e-06,
+      "loss": 0.3311,
+      "step": 10624
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.523698806762695,
+      "learning_rate": 2.6488759224300673e-06,
+      "loss": 0.3587,
+      "step": 10625
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.232637405395508,
+      "learning_rate": 2.6463016989874723e-06,
+      "loss": 0.333,
+      "step": 10626
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 33.780067443847656,
+      "learning_rate": 2.6437274755448773e-06,
+      "loss": 0.2737,
+      "step": 10627
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.261059761047363,
+      "learning_rate": 2.6411532521022822e-06,
+      "loss": 0.5378,
+      "step": 10628
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.210847854614258,
+      "learning_rate": 2.6385790286596877e-06,
+      "loss": 0.5803,
+      "step": 10629
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.004446029663086,
+      "learning_rate": 2.636004805217093e-06,
+      "loss": 0.5725,
+      "step": 10630
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.11176586151123,
+      "learning_rate": 2.633430581774498e-06,
+      "loss": 0.252,
+      "step": 10631
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.555098533630371,
+      "learning_rate": 2.6308563583319035e-06,
+      "loss": 0.5343,
+      "step": 10632
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.724139213562012,
+      "learning_rate": 2.6282821348893085e-06,
+      "loss": 0.2942,
+      "step": 10633
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.260353088378906,
+      "learning_rate": 2.6257079114467134e-06,
+      "loss": 0.3384,
+      "step": 10634
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.129500389099121,
+      "learning_rate": 2.623133688004119e-06,
+      "loss": 0.4465,
+      "step": 10635
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.664830207824707,
+      "learning_rate": 2.620559464561524e-06,
+      "loss": 0.3828,
+      "step": 10636
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.48660659790039,
+      "learning_rate": 2.6179852411189292e-06,
+      "loss": 0.243,
+      "step": 10637
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.969393730163574,
+      "learning_rate": 2.6154110176763347e-06,
+      "loss": 0.3194,
+      "step": 10638
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.693145751953125,
+      "learning_rate": 2.6128367942337396e-06,
+      "loss": 0.2953,
+      "step": 10639
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.943690299987793,
+      "learning_rate": 2.6102625707911446e-06,
+      "loss": 0.3923,
+      "step": 10640
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.0937981605529785,
+      "learning_rate": 2.6076883473485496e-06,
+      "loss": 0.1444,
+      "step": 10641
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.566240310668945,
+      "learning_rate": 2.605114123905955e-06,
+      "loss": 0.3607,
+      "step": 10642
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.5882415771484375,
+      "learning_rate": 2.6025399004633604e-06,
+      "loss": 0.2362,
+      "step": 10643
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.648507118225098,
+      "learning_rate": 2.5999656770207654e-06,
+      "loss": 0.3617,
+      "step": 10644
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.666114330291748,
+      "learning_rate": 2.597391453578171e-06,
+      "loss": 0.2689,
+      "step": 10645
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.404210090637207,
+      "learning_rate": 2.594817230135576e-06,
+      "loss": 0.4654,
+      "step": 10646
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.755274772644043,
+      "learning_rate": 2.592243006692981e-06,
+      "loss": 0.3033,
+      "step": 10647
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.978460311889648,
+      "learning_rate": 2.5896687832503862e-06,
+      "loss": 0.3809,
+      "step": 10648
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.506619453430176,
+      "learning_rate": 2.5870945598077912e-06,
+      "loss": 0.474,
+      "step": 10649
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.281198501586914,
+      "learning_rate": 2.5845203363651966e-06,
+      "loss": 0.4336,
+      "step": 10650
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.126219749450684,
+      "learning_rate": 2.581946112922602e-06,
+      "loss": 0.2947,
+      "step": 10651
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.406610488891602,
+      "learning_rate": 2.579371889480007e-06,
+      "loss": 0.3549,
+      "step": 10652
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.784308433532715,
+      "learning_rate": 2.576797666037412e-06,
+      "loss": 0.2985,
+      "step": 10653
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.192948341369629,
+      "learning_rate": 2.574223442594817e-06,
+      "loss": 0.3816,
+      "step": 10654
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.522122383117676,
+      "learning_rate": 2.5716492191522224e-06,
+      "loss": 0.3252,
+      "step": 10655
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.236865997314453,
+      "learning_rate": 2.569074995709628e-06,
+      "loss": 0.31,
+      "step": 10656
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.734746932983398,
+      "learning_rate": 2.566500772267033e-06,
+      "loss": 0.2426,
+      "step": 10657
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.649822235107422,
+      "learning_rate": 2.5639265488244382e-06,
+      "loss": 0.4776,
+      "step": 10658
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.669262886047363,
+      "learning_rate": 2.561352325381843e-06,
+      "loss": 0.3362,
+      "step": 10659
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.61310863494873,
+      "learning_rate": 2.558778101939248e-06,
+      "loss": 0.2083,
+      "step": 10660
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.194008827209473,
+      "learning_rate": 2.5562038784966536e-06,
+      "loss": 0.3197,
+      "step": 10661
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.544124603271484,
+      "learning_rate": 2.5536296550540586e-06,
+      "loss": 0.3584,
+      "step": 10662
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.95773696899414,
+      "learning_rate": 2.551055431611464e-06,
+      "loss": 0.2641,
+      "step": 10663
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.994258880615234,
+      "learning_rate": 2.5484812081688694e-06,
+      "loss": 0.3933,
+      "step": 10664
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.87153434753418,
+      "learning_rate": 2.5459069847262744e-06,
+      "loss": 0.2925,
+      "step": 10665
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.127371788024902,
+      "learning_rate": 2.5433327612836794e-06,
+      "loss": 0.275,
+      "step": 10666
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.941168785095215,
+      "learning_rate": 2.5407585378410844e-06,
+      "loss": 0.3046,
+      "step": 10667
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.67513656616211,
+      "learning_rate": 2.5381843143984898e-06,
+      "loss": 0.3292,
+      "step": 10668
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.871071815490723,
+      "learning_rate": 2.535610090955895e-06,
+      "loss": 0.2954,
+      "step": 10669
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.738022804260254,
+      "learning_rate": 2.5330358675133e-06,
+      "loss": 0.3821,
+      "step": 10670
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.993451118469238,
+      "learning_rate": 2.5304616440707056e-06,
+      "loss": 0.4609,
+      "step": 10671
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.69482421875,
+      "learning_rate": 2.5278874206281106e-06,
+      "loss": 0.4201,
+      "step": 10672
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.218208312988281,
+      "learning_rate": 2.5253131971855156e-06,
+      "loss": 0.3566,
+      "step": 10673
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.55072021484375,
+      "learning_rate": 2.522738973742921e-06,
+      "loss": 0.3988,
+      "step": 10674
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.507923126220703,
+      "learning_rate": 2.520164750300326e-06,
+      "loss": 0.4861,
+      "step": 10675
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.510515213012695,
+      "learning_rate": 2.5175905268577314e-06,
+      "loss": 0.3368,
+      "step": 10676
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.693829536437988,
+      "learning_rate": 2.515016303415137e-06,
+      "loss": 0.2396,
+      "step": 10677
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.560959815979004,
+      "learning_rate": 2.5124420799725418e-06,
+      "loss": 0.4916,
+      "step": 10678
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.765018463134766,
+      "learning_rate": 2.5098678565299468e-06,
+      "loss": 0.6102,
+      "step": 10679
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.950790405273438,
+      "learning_rate": 2.5072936330873518e-06,
+      "loss": 0.4473,
+      "step": 10680
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.45733642578125,
+      "learning_rate": 2.504719409644757e-06,
+      "loss": 0.4512,
+      "step": 10681
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.298639297485352,
+      "learning_rate": 2.5021451862021626e-06,
+      "loss": 0.3228,
+      "step": 10682
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.985605239868164,
+      "learning_rate": 2.4995709627595676e-06,
+      "loss": 0.5333,
+      "step": 10683
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.020630836486816,
+      "learning_rate": 2.496996739316973e-06,
+      "loss": 0.5044,
+      "step": 10684
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.97929859161377,
+      "learning_rate": 2.494422515874378e-06,
+      "loss": 0.3551,
+      "step": 10685
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.300394058227539,
+      "learning_rate": 2.491848292431783e-06,
+      "loss": 0.408,
+      "step": 10686
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.84925365447998,
+      "learning_rate": 2.4892740689891884e-06,
+      "loss": 0.3481,
+      "step": 10687
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.51993179321289,
+      "learning_rate": 2.4866998455465933e-06,
+      "loss": 0.3779,
+      "step": 10688
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.055931091308594,
+      "learning_rate": 2.4841256221039988e-06,
+      "loss": 0.3127,
+      "step": 10689
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.135323524475098,
+      "learning_rate": 2.481551398661404e-06,
+      "loss": 0.3064,
+      "step": 10690
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.34423828125,
+      "learning_rate": 2.478977175218809e-06,
+      "loss": 0.2762,
+      "step": 10691
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.478259086608887,
+      "learning_rate": 2.476402951776214e-06,
+      "loss": 0.4732,
+      "step": 10692
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.674147129058838,
+      "learning_rate": 2.473828728333619e-06,
+      "loss": 0.3008,
+      "step": 10693
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.84994125366211,
+      "learning_rate": 2.4712545048910245e-06,
+      "loss": 0.4315,
+      "step": 10694
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.25035285949707,
+      "learning_rate": 2.46868028144843e-06,
+      "loss": 0.2382,
+      "step": 10695
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.903910636901855,
+      "learning_rate": 2.466106058005835e-06,
+      "loss": 0.3797,
+      "step": 10696
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.617475509643555,
+      "learning_rate": 2.4635318345632403e-06,
+      "loss": 0.3788,
+      "step": 10697
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.769457817077637,
+      "learning_rate": 2.4609576111206453e-06,
+      "loss": 0.3622,
+      "step": 10698
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.13018798828125,
+      "learning_rate": 2.4583833876780503e-06,
+      "loss": 0.2796,
+      "step": 10699
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.063539505004883,
+      "learning_rate": 2.4558091642354557e-06,
+      "loss": 0.3168,
+      "step": 10700
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.092525482177734,
+      "learning_rate": 2.4532349407928607e-06,
+      "loss": 0.4787,
+      "step": 10701
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.005880355834961,
+      "learning_rate": 2.450660717350266e-06,
+      "loss": 0.2767,
+      "step": 10702
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.908269882202148,
+      "learning_rate": 2.4480864939076715e-06,
+      "loss": 0.27,
+      "step": 10703
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.34350299835205,
+      "learning_rate": 2.4455122704650765e-06,
+      "loss": 0.4592,
+      "step": 10704
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.498222827911377,
+      "learning_rate": 2.4429380470224815e-06,
+      "loss": 0.393,
+      "step": 10705
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.66064739227295,
+      "learning_rate": 2.4403638235798865e-06,
+      "loss": 0.4346,
+      "step": 10706
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.21183967590332,
+      "learning_rate": 2.437789600137292e-06,
+      "loss": 0.3454,
+      "step": 10707
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.232462882995605,
+      "learning_rate": 2.4352153766946973e-06,
+      "loss": 0.3814,
+      "step": 10708
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.885994911193848,
+      "learning_rate": 2.4326411532521023e-06,
+      "loss": 0.4438,
+      "step": 10709
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.98500919342041,
+      "learning_rate": 2.4300669298095077e-06,
+      "loss": 0.4261,
+      "step": 10710
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.819098472595215,
+      "learning_rate": 2.4274927063669127e-06,
+      "loss": 0.4565,
+      "step": 10711
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.974862098693848,
+      "learning_rate": 2.4249184829243177e-06,
+      "loss": 0.455,
+      "step": 10712
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.547346115112305,
+      "learning_rate": 2.422344259481723e-06,
+      "loss": 0.4271,
+      "step": 10713
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.870933532714844,
+      "learning_rate": 2.419770036039128e-06,
+      "loss": 0.3525,
+      "step": 10714
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.99450969696045,
+      "learning_rate": 2.4171958125965335e-06,
+      "loss": 0.5769,
+      "step": 10715
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.365693092346191,
+      "learning_rate": 2.414621589153939e-06,
+      "loss": 0.4526,
+      "step": 10716
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.111851692199707,
+      "learning_rate": 2.412047365711344e-06,
+      "loss": 0.364,
+      "step": 10717
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.152235984802246,
+      "learning_rate": 2.409473142268749e-06,
+      "loss": 0.3813,
+      "step": 10718
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.690625190734863,
+      "learning_rate": 2.406898918826154e-06,
+      "loss": 0.394,
+      "step": 10719
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.726323127746582,
+      "learning_rate": 2.4043246953835593e-06,
+      "loss": 0.3937,
+      "step": 10720
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.641265869140625,
+      "learning_rate": 2.4017504719409647e-06,
+      "loss": 0.4748,
+      "step": 10721
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.505012035369873,
+      "learning_rate": 2.3991762484983697e-06,
+      "loss": 0.3154,
+      "step": 10722
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.628533363342285,
+      "learning_rate": 2.396602025055775e-06,
+      "loss": 0.4944,
+      "step": 10723
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.972872734069824,
+      "learning_rate": 2.39402780161318e-06,
+      "loss": 0.3235,
+      "step": 10724
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.067104816436768,
+      "learning_rate": 2.391453578170585e-06,
+      "loss": 0.2797,
+      "step": 10725
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.320572853088379,
+      "learning_rate": 2.3888793547279905e-06,
+      "loss": 0.3946,
+      "step": 10726
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.184109687805176,
+      "learning_rate": 2.3863051312853955e-06,
+      "loss": 0.4283,
+      "step": 10727
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.515572547912598,
+      "learning_rate": 2.383730907842801e-06,
+      "loss": 0.4074,
+      "step": 10728
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.552699089050293,
+      "learning_rate": 2.3811566844002063e-06,
+      "loss": 0.3594,
+      "step": 10729
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.412210464477539,
+      "learning_rate": 2.3785824609576113e-06,
+      "loss": 0.4901,
+      "step": 10730
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.148717880249023,
+      "learning_rate": 2.3760082375150163e-06,
+      "loss": 0.4589,
+      "step": 10731
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.961956977844238,
+      "learning_rate": 2.3734340140724213e-06,
+      "loss": 0.4587,
+      "step": 10732
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.488950729370117,
+      "learning_rate": 2.3708597906298267e-06,
+      "loss": 0.4681,
+      "step": 10733
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.658753395080566,
+      "learning_rate": 2.368285567187232e-06,
+      "loss": 0.42,
+      "step": 10734
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.891936302185059,
+      "learning_rate": 2.365711343744637e-06,
+      "loss": 0.3699,
+      "step": 10735
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.638575553894043,
+      "learning_rate": 2.3631371203020425e-06,
+      "loss": 0.4098,
+      "step": 10736
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.642004013061523,
+      "learning_rate": 2.3605628968594475e-06,
+      "loss": 0.4884,
+      "step": 10737
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.215935707092285,
+      "learning_rate": 2.3579886734168525e-06,
+      "loss": 0.3861,
+      "step": 10738
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.332646369934082,
+      "learning_rate": 2.355414449974258e-06,
+      "loss": 0.3031,
+      "step": 10739
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.552823066711426,
+      "learning_rate": 2.352840226531663e-06,
+      "loss": 0.3673,
+      "step": 10740
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.942591190338135,
+      "learning_rate": 2.3502660030890683e-06,
+      "loss": 0.3187,
+      "step": 10741
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.71848201751709,
+      "learning_rate": 2.3476917796464737e-06,
+      "loss": 0.4059,
+      "step": 10742
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.311237335205078,
+      "learning_rate": 2.3451175562038787e-06,
+      "loss": 0.3469,
+      "step": 10743
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.839985370635986,
+      "learning_rate": 2.3425433327612836e-06,
+      "loss": 0.2815,
+      "step": 10744
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.345728874206543,
+      "learning_rate": 2.3399691093186886e-06,
+      "loss": 0.3124,
+      "step": 10745
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.57292366027832,
+      "learning_rate": 2.337394885876094e-06,
+      "loss": 0.273,
+      "step": 10746
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.953744888305664,
+      "learning_rate": 2.3348206624334995e-06,
+      "loss": 0.3599,
+      "step": 10747
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.740678787231445,
+      "learning_rate": 2.3322464389909044e-06,
+      "loss": 0.3269,
+      "step": 10748
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 18.84109878540039,
+      "learning_rate": 2.32967221554831e-06,
+      "loss": 0.4604,
+      "step": 10749
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.938599586486816,
+      "learning_rate": 2.327097992105715e-06,
+      "loss": 0.4271,
+      "step": 10750
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.845032691955566,
+      "learning_rate": 2.32452376866312e-06,
+      "loss": 0.4105,
+      "step": 10751
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.102261543273926,
+      "learning_rate": 2.3219495452205252e-06,
+      "loss": 0.4717,
+      "step": 10752
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.609228134155273,
+      "learning_rate": 2.3193753217779302e-06,
+      "loss": 0.4263,
+      "step": 10753
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.596081733703613,
+      "learning_rate": 2.3168010983353356e-06,
+      "loss": 0.2193,
+      "step": 10754
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.777579307556152,
+      "learning_rate": 2.314226874892741e-06,
+      "loss": 0.4518,
+      "step": 10755
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.152752876281738,
+      "learning_rate": 2.311652651450146e-06,
+      "loss": 0.3354,
+      "step": 10756
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.82956314086914,
+      "learning_rate": 2.309078428007551e-06,
+      "loss": 0.4699,
+      "step": 10757
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.48739242553711,
+      "learning_rate": 2.306504204564956e-06,
+      "loss": 0.417,
+      "step": 10758
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.494852066040039,
+      "learning_rate": 2.3039299811223614e-06,
+      "loss": 0.4977,
+      "step": 10759
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.750948905944824,
+      "learning_rate": 2.301355757679767e-06,
+      "loss": 0.455,
+      "step": 10760
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.151665687561035,
+      "learning_rate": 2.298781534237172e-06,
+      "loss": 0.4942,
+      "step": 10761
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 32.819061279296875,
+      "learning_rate": 2.2962073107945772e-06,
+      "loss": 0.404,
+      "step": 10762
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.47062873840332,
+      "learning_rate": 2.2936330873519822e-06,
+      "loss": 0.3854,
+      "step": 10763
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.467101097106934,
+      "learning_rate": 2.291058863909387e-06,
+      "loss": 0.3279,
+      "step": 10764
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.407057762145996,
+      "learning_rate": 2.2884846404667926e-06,
+      "loss": 0.3094,
+      "step": 10765
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.819005012512207,
+      "learning_rate": 2.2859104170241976e-06,
+      "loss": 0.3628,
+      "step": 10766
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.826458930969238,
+      "learning_rate": 2.283336193581603e-06,
+      "loss": 0.2743,
+      "step": 10767
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.1718525886535645,
+      "learning_rate": 2.2807619701390084e-06,
+      "loss": 0.3215,
+      "step": 10768
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.493207931518555,
+      "learning_rate": 2.2781877466964134e-06,
+      "loss": 0.3257,
+      "step": 10769
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.476923942565918,
+      "learning_rate": 2.2756135232538184e-06,
+      "loss": 0.3036,
+      "step": 10770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.054359436035156,
+      "learning_rate": 2.2730392998112234e-06,
+      "loss": 0.4715,
+      "step": 10771
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.640109062194824,
+      "learning_rate": 2.270465076368629e-06,
+      "loss": 0.3296,
+      "step": 10772
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.762101173400879,
+      "learning_rate": 2.267890852926034e-06,
+      "loss": 0.3601,
+      "step": 10773
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.953591346740723,
+      "learning_rate": 2.265316629483439e-06,
+      "loss": 0.4362,
+      "step": 10774
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.266918182373047,
+      "learning_rate": 2.2627424060408446e-06,
+      "loss": 0.4505,
+      "step": 10775
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.953780174255371,
+      "learning_rate": 2.2601681825982496e-06,
+      "loss": 0.4516,
+      "step": 10776
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.710752487182617,
+      "learning_rate": 2.2575939591556546e-06,
+      "loss": 0.3218,
+      "step": 10777
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.6055755615234375,
+      "learning_rate": 2.25501973571306e-06,
+      "loss": 0.2625,
+      "step": 10778
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.41867446899414,
+      "learning_rate": 2.252445512270465e-06,
+      "loss": 0.3481,
+      "step": 10779
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.75260066986084,
+      "learning_rate": 2.2498712888278704e-06,
+      "loss": 0.3312,
+      "step": 10780
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.806591033935547,
+      "learning_rate": 2.247297065385276e-06,
+      "loss": 0.4483,
+      "step": 10781
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.854302406311035,
+      "learning_rate": 2.244722841942681e-06,
+      "loss": 0.3797,
+      "step": 10782
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.459409713745117,
+      "learning_rate": 2.2421486185000858e-06,
+      "loss": 0.3149,
+      "step": 10783
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.881391525268555,
+      "learning_rate": 2.2395743950574908e-06,
+      "loss": 0.4413,
+      "step": 10784
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.098734855651855,
+      "learning_rate": 2.237000171614896e-06,
+      "loss": 0.4845,
+      "step": 10785
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.4308443069458,
+      "learning_rate": 2.2344259481723016e-06,
+      "loss": 0.3798,
+      "step": 10786
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.48336410522461,
+      "learning_rate": 2.2318517247297066e-06,
+      "loss": 0.4244,
+      "step": 10787
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.419635772705078,
+      "learning_rate": 2.229277501287112e-06,
+      "loss": 0.4723,
+      "step": 10788
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.306260585784912,
+      "learning_rate": 2.226703277844517e-06,
+      "loss": 0.3261,
+      "step": 10789
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.113286972045898,
+      "learning_rate": 2.224129054401922e-06,
+      "loss": 0.5654,
+      "step": 10790
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.769453048706055,
+      "learning_rate": 2.2215548309593274e-06,
+      "loss": 0.4038,
+      "step": 10791
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.740857124328613,
+      "learning_rate": 2.2189806075167324e-06,
+      "loss": 0.3157,
+      "step": 10792
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.538228034973145,
+      "learning_rate": 2.2164063840741378e-06,
+      "loss": 0.4004,
+      "step": 10793
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.721349716186523,
+      "learning_rate": 2.213832160631543e-06,
+      "loss": 0.2371,
+      "step": 10794
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.922874450683594,
+      "learning_rate": 2.211257937188948e-06,
+      "loss": 0.368,
+      "step": 10795
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.995516777038574,
+      "learning_rate": 2.208683713746353e-06,
+      "loss": 0.4539,
+      "step": 10796
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.98942756652832,
+      "learning_rate": 2.206109490303758e-06,
+      "loss": 0.3035,
+      "step": 10797
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.0841064453125,
+      "learning_rate": 2.2035352668611636e-06,
+      "loss": 0.288,
+      "step": 10798
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.027642250061035,
+      "learning_rate": 2.200961043418569e-06,
+      "loss": 0.3204,
+      "step": 10799
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.990629196166992,
+      "learning_rate": 2.198386819975974e-06,
+      "loss": 0.3326,
+      "step": 10800
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.817361831665039,
+      "learning_rate": 2.1958125965333794e-06,
+      "loss": 0.5591,
+      "step": 10801
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.330004692077637,
+      "learning_rate": 2.1932383730907843e-06,
+      "loss": 0.4329,
+      "step": 10802
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.918294906616211,
+      "learning_rate": 2.1906641496481893e-06,
+      "loss": 0.333,
+      "step": 10803
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.731829643249512,
+      "learning_rate": 2.1880899262055947e-06,
+      "loss": 0.318,
+      "step": 10804
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.897744178771973,
+      "learning_rate": 2.1855157027629997e-06,
+      "loss": 0.3439,
+      "step": 10805
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.820111274719238,
+      "learning_rate": 2.182941479320405e-06,
+      "loss": 0.3696,
+      "step": 10806
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.642640113830566,
+      "learning_rate": 2.1803672558778106e-06,
+      "loss": 0.5032,
+      "step": 10807
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.144488334655762,
+      "learning_rate": 2.1777930324352155e-06,
+      "loss": 0.3735,
+      "step": 10808
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.489876747131348,
+      "learning_rate": 2.1752188089926205e-06,
+      "loss": 0.3798,
+      "step": 10809
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.61406135559082,
+      "learning_rate": 2.1726445855500255e-06,
+      "loss": 0.3387,
+      "step": 10810
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.711216926574707,
+      "learning_rate": 2.170070362107431e-06,
+      "loss": 0.4337,
+      "step": 10811
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.59476375579834,
+      "learning_rate": 2.1674961386648363e-06,
+      "loss": 0.4709,
+      "step": 10812
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.587848663330078,
+      "learning_rate": 2.1649219152222413e-06,
+      "loss": 0.3485,
+      "step": 10813
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.932808876037598,
+      "learning_rate": 2.1623476917796467e-06,
+      "loss": 0.3501,
+      "step": 10814
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.602524757385254,
+      "learning_rate": 2.1597734683370517e-06,
+      "loss": 0.3511,
+      "step": 10815
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.017490863800049,
+      "learning_rate": 2.1571992448944567e-06,
+      "loss": 0.2124,
+      "step": 10816
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.48502254486084,
+      "learning_rate": 2.154625021451862e-06,
+      "loss": 0.2939,
+      "step": 10817
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.827862739562988,
+      "learning_rate": 2.152050798009267e-06,
+      "loss": 0.3149,
+      "step": 10818
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.767644882202148,
+      "learning_rate": 2.1494765745666725e-06,
+      "loss": 0.2953,
+      "step": 10819
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.464248657226562,
+      "learning_rate": 2.146902351124078e-06,
+      "loss": 0.3244,
+      "step": 10820
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.992477416992188,
+      "learning_rate": 2.144328127681483e-06,
+      "loss": 0.3256,
+      "step": 10821
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.161433219909668,
+      "learning_rate": 2.141753904238888e-06,
+      "loss": 0.3607,
+      "step": 10822
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.0374755859375,
+      "learning_rate": 2.139179680796293e-06,
+      "loss": 0.2819,
+      "step": 10823
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.524991035461426,
+      "learning_rate": 2.1366054573536983e-06,
+      "loss": 0.5337,
+      "step": 10824
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.048221588134766,
+      "learning_rate": 2.1340312339111037e-06,
+      "loss": 0.3204,
+      "step": 10825
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.081626892089844,
+      "learning_rate": 2.1314570104685087e-06,
+      "loss": 0.4938,
+      "step": 10826
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.9117021560668945,
+      "learning_rate": 2.128882787025914e-06,
+      "loss": 0.3639,
+      "step": 10827
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.109366416931152,
+      "learning_rate": 2.126308563583319e-06,
+      "loss": 0.2826,
+      "step": 10828
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.635122299194336,
+      "learning_rate": 2.123734340140724e-06,
+      "loss": 0.4808,
+      "step": 10829
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.299768924713135,
+      "learning_rate": 2.1211601166981295e-06,
+      "loss": 0.2482,
+      "step": 10830
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.62306022644043,
+      "learning_rate": 2.1185858932555345e-06,
+      "loss": 0.4508,
+      "step": 10831
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.016575813293457,
+      "learning_rate": 2.11601166981294e-06,
+      "loss": 0.2345,
+      "step": 10832
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.547022819519043,
+      "learning_rate": 2.1134374463703453e-06,
+      "loss": 0.4968,
+      "step": 10833
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.415725708007812,
+      "learning_rate": 2.1108632229277503e-06,
+      "loss": 0.291,
+      "step": 10834
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.098642349243164,
+      "learning_rate": 2.1082889994851553e-06,
+      "loss": 0.3459,
+      "step": 10835
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.48742961883545,
+      "learning_rate": 2.1057147760425603e-06,
+      "loss": 0.4204,
+      "step": 10836
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.175433158874512,
+      "learning_rate": 2.1031405525999657e-06,
+      "loss": 0.2679,
+      "step": 10837
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.38789176940918,
+      "learning_rate": 2.100566329157371e-06,
+      "loss": 0.3646,
+      "step": 10838
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.256763935089111,
+      "learning_rate": 2.097992105714776e-06,
+      "loss": 0.2612,
+      "step": 10839
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.745893478393555,
+      "learning_rate": 2.0954178822721815e-06,
+      "loss": 0.2839,
+      "step": 10840
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.312209129333496,
+      "learning_rate": 2.0928436588295865e-06,
+      "loss": 0.4288,
+      "step": 10841
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.442017555236816,
+      "learning_rate": 2.0902694353869915e-06,
+      "loss": 0.4444,
+      "step": 10842
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.069182395935059,
+      "learning_rate": 2.087695211944397e-06,
+      "loss": 0.4258,
+      "step": 10843
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.49484920501709,
+      "learning_rate": 2.085120988501802e-06,
+      "loss": 0.3,
+      "step": 10844
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.370274066925049,
+      "learning_rate": 2.0825467650592073e-06,
+      "loss": 0.2795,
+      "step": 10845
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.82091999053955,
+      "learning_rate": 2.0799725416166127e-06,
+      "loss": 0.33,
+      "step": 10846
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.931026458740234,
+      "learning_rate": 2.0773983181740177e-06,
+      "loss": 0.3417,
+      "step": 10847
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.414534568786621,
+      "learning_rate": 2.0748240947314227e-06,
+      "loss": 0.3247,
+      "step": 10848
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.010655403137207,
+      "learning_rate": 2.0722498712888276e-06,
+      "loss": 0.5502,
+      "step": 10849
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.234877586364746,
+      "learning_rate": 2.069675647846233e-06,
+      "loss": 0.3277,
+      "step": 10850
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.902608871459961,
+      "learning_rate": 2.0671014244036385e-06,
+      "loss": 0.4241,
+      "step": 10851
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.112380981445312,
+      "learning_rate": 2.0645272009610435e-06,
+      "loss": 0.4556,
+      "step": 10852
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.702977180480957,
+      "learning_rate": 2.061952977518449e-06,
+      "loss": 0.3529,
+      "step": 10853
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.624886512756348,
+      "learning_rate": 2.059378754075854e-06,
+      "loss": 0.2972,
+      "step": 10854
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.750003814697266,
+      "learning_rate": 2.056804530633259e-06,
+      "loss": 0.3776,
+      "step": 10855
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.668874740600586,
+      "learning_rate": 2.0542303071906643e-06,
+      "loss": 0.3964,
+      "step": 10856
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.533783912658691,
+      "learning_rate": 2.0516560837480692e-06,
+      "loss": 0.2578,
+      "step": 10857
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.777493000030518,
+      "learning_rate": 2.0490818603054746e-06,
+      "loss": 0.2758,
+      "step": 10858
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.253680229187012,
+      "learning_rate": 2.04650763686288e-06,
+      "loss": 0.4073,
+      "step": 10859
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.718351364135742,
+      "learning_rate": 2.0439334134202846e-06,
+      "loss": 0.5825,
+      "step": 10860
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.605161666870117,
+      "learning_rate": 2.04135918997769e-06,
+      "loss": 0.3916,
+      "step": 10861
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.781625747680664,
+      "learning_rate": 2.038784966535095e-06,
+      "loss": 0.2946,
+      "step": 10862
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.692581176757812,
+      "learning_rate": 2.0362107430925004e-06,
+      "loss": 0.3037,
+      "step": 10863
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.491222381591797,
+      "learning_rate": 2.033636519649906e-06,
+      "loss": 0.3868,
+      "step": 10864
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.820777893066406,
+      "learning_rate": 2.031062296207311e-06,
+      "loss": 0.3183,
+      "step": 10865
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.390010833740234,
+      "learning_rate": 2.0284880727647162e-06,
+      "loss": 0.2387,
+      "step": 10866
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.670175552368164,
+      "learning_rate": 2.0259138493221212e-06,
+      "loss": 0.4745,
+      "step": 10867
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.558318138122559,
+      "learning_rate": 2.0233396258795262e-06,
+      "loss": 0.3502,
+      "step": 10868
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.584383964538574,
+      "learning_rate": 2.0207654024369316e-06,
+      "loss": 0.2779,
+      "step": 10869
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.472787857055664,
+      "learning_rate": 2.0181911789943366e-06,
+      "loss": 0.3735,
+      "step": 10870
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.116923332214355,
+      "learning_rate": 2.015616955551742e-06,
+      "loss": 0.3172,
+      "step": 10871
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.554486274719238,
+      "learning_rate": 2.0130427321091474e-06,
+      "loss": 0.3193,
+      "step": 10872
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.012552261352539,
+      "learning_rate": 2.010468508666552e-06,
+      "loss": 0.3107,
+      "step": 10873
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.351103782653809,
+      "learning_rate": 2.0078942852239574e-06,
+      "loss": 0.3081,
+      "step": 10874
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.335416793823242,
+      "learning_rate": 2.0053200617813624e-06,
+      "loss": 0.4063,
+      "step": 10875
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.989412307739258,
+      "learning_rate": 2.002745838338768e-06,
+      "loss": 0.5465,
+      "step": 10876
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.931973457336426,
+      "learning_rate": 2.0001716148961732e-06,
+      "loss": 0.2691,
+      "step": 10877
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.633210182189941,
+      "learning_rate": 1.997597391453578e-06,
+      "loss": 0.5472,
+      "step": 10878
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.967374801635742,
+      "learning_rate": 1.9950231680109836e-06,
+      "loss": 0.408,
+      "step": 10879
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.11751651763916,
+      "learning_rate": 1.9924489445683886e-06,
+      "loss": 0.4009,
+      "step": 10880
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.455737113952637,
+      "learning_rate": 1.9898747211257936e-06,
+      "loss": 0.3342,
+      "step": 10881
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.098809242248535,
+      "learning_rate": 1.987300497683199e-06,
+      "loss": 0.2926,
+      "step": 10882
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.390904426574707,
+      "learning_rate": 1.984726274240604e-06,
+      "loss": 0.4593,
+      "step": 10883
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.999994277954102,
+      "learning_rate": 1.9821520507980094e-06,
+      "loss": 0.4293,
+      "step": 10884
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.855880737304688,
+      "learning_rate": 1.979577827355415e-06,
+      "loss": 0.3421,
+      "step": 10885
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.927961349487305,
+      "learning_rate": 1.9770036039128194e-06,
+      "loss": 0.3545,
+      "step": 10886
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.199851036071777,
+      "learning_rate": 1.974429380470225e-06,
+      "loss": 0.3846,
+      "step": 10887
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.086398124694824,
+      "learning_rate": 1.9718551570276298e-06,
+      "loss": 0.3284,
+      "step": 10888
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.92074728012085,
+      "learning_rate": 1.969280933585035e-06,
+      "loss": 0.2864,
+      "step": 10889
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.711901664733887,
+      "learning_rate": 1.9667067101424406e-06,
+      "loss": 0.4471,
+      "step": 10890
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.300307273864746,
+      "learning_rate": 1.9641324866998456e-06,
+      "loss": 0.3492,
+      "step": 10891
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.916529655456543,
+      "learning_rate": 1.961558263257251e-06,
+      "loss": 0.3667,
+      "step": 10892
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.481868743896484,
+      "learning_rate": 1.958984039814656e-06,
+      "loss": 0.3685,
+      "step": 10893
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.224878311157227,
+      "learning_rate": 1.956409816372061e-06,
+      "loss": 0.3121,
+      "step": 10894
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.719337463378906,
+      "learning_rate": 1.9538355929294664e-06,
+      "loss": 0.3187,
+      "step": 10895
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.698870658874512,
+      "learning_rate": 1.9512613694868714e-06,
+      "loss": 0.3979,
+      "step": 10896
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.925223350524902,
+      "learning_rate": 1.9486871460442768e-06,
+      "loss": 0.2924,
+      "step": 10897
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.224217414855957,
+      "learning_rate": 1.946112922601682e-06,
+      "loss": 0.3423,
+      "step": 10898
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.26035213470459,
+      "learning_rate": 1.9435386991590868e-06,
+      "loss": 0.3215,
+      "step": 10899
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.702499389648438,
+      "learning_rate": 1.940964475716492e-06,
+      "loss": 0.4997,
+      "step": 10900
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.4429731369018555,
+      "learning_rate": 1.938390252273897e-06,
+      "loss": 0.2643,
+      "step": 10901
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.223000526428223,
+      "learning_rate": 1.9358160288313026e-06,
+      "loss": 0.3761,
+      "step": 10902
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.797672271728516,
+      "learning_rate": 1.933241805388708e-06,
+      "loss": 0.4087,
+      "step": 10903
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.24771785736084,
+      "learning_rate": 1.930667581946113e-06,
+      "loss": 0.4276,
+      "step": 10904
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.165146827697754,
+      "learning_rate": 1.9280933585035184e-06,
+      "loss": 0.2561,
+      "step": 10905
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.10590934753418,
+      "learning_rate": 1.9255191350609234e-06,
+      "loss": 0.4682,
+      "step": 10906
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.769000053405762,
+      "learning_rate": 1.9229449116183283e-06,
+      "loss": 0.4245,
+      "step": 10907
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.64234447479248,
+      "learning_rate": 1.9203706881757338e-06,
+      "loss": 0.2173,
+      "step": 10908
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.308282852172852,
+      "learning_rate": 1.9177964647331387e-06,
+      "loss": 0.3512,
+      "step": 10909
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.698214530944824,
+      "learning_rate": 1.915222241290544e-06,
+      "loss": 0.4464,
+      "step": 10910
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.663969039916992,
+      "learning_rate": 1.9126480178479496e-06,
+      "loss": 0.3552,
+      "step": 10911
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.192497253417969,
+      "learning_rate": 1.910073794405354e-06,
+      "loss": 0.4219,
+      "step": 10912
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.867043495178223,
+      "learning_rate": 1.9074995709627595e-06,
+      "loss": 0.3177,
+      "step": 10913
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.983952522277832,
+      "learning_rate": 1.9049253475201647e-06,
+      "loss": 0.2394,
+      "step": 10914
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.460039138793945,
+      "learning_rate": 1.90235112407757e-06,
+      "loss": 0.2937,
+      "step": 10915
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.977998733520508,
+      "learning_rate": 1.8997769006349751e-06,
+      "loss": 0.4923,
+      "step": 10916
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.566569328308105,
+      "learning_rate": 1.8972026771923805e-06,
+      "loss": 0.34,
+      "step": 10917
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.486238479614258,
+      "learning_rate": 1.8946284537497857e-06,
+      "loss": 0.3451,
+      "step": 10918
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.05168628692627,
+      "learning_rate": 1.8920542303071905e-06,
+      "loss": 0.1917,
+      "step": 10919
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.568209648132324,
+      "learning_rate": 1.8894800068645957e-06,
+      "loss": 0.3181,
+      "step": 10920
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.249594688415527,
+      "learning_rate": 1.8869057834220011e-06,
+      "loss": 0.2863,
+      "step": 10921
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.555765151977539,
+      "learning_rate": 1.8843315599794063e-06,
+      "loss": 0.3128,
+      "step": 10922
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.785953998565674,
+      "learning_rate": 1.8817573365368115e-06,
+      "loss": 0.2663,
+      "step": 10923
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.51054573059082,
+      "learning_rate": 1.8791831130942167e-06,
+      "loss": 0.3521,
+      "step": 10924
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.163633346557617,
+      "learning_rate": 1.8766088896516215e-06,
+      "loss": 0.3209,
+      "step": 10925
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.81087589263916,
+      "learning_rate": 1.8740346662090271e-06,
+      "loss": 0.3311,
+      "step": 10926
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.272926330566406,
+      "learning_rate": 1.8714604427664321e-06,
+      "loss": 0.4884,
+      "step": 10927
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.91628646850586,
+      "learning_rate": 1.8688862193238373e-06,
+      "loss": 0.3836,
+      "step": 10928
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.258402824401855,
+      "learning_rate": 1.8663119958812425e-06,
+      "loss": 0.2882,
+      "step": 10929
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.744312286376953,
+      "learning_rate": 1.8637377724386477e-06,
+      "loss": 0.3744,
+      "step": 10930
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.320903301239014,
+      "learning_rate": 1.861163548996053e-06,
+      "loss": 0.1566,
+      "step": 10931
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.218724250793457,
+      "learning_rate": 1.8585893255534581e-06,
+      "loss": 0.3718,
+      "step": 10932
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.666526794433594,
+      "learning_rate": 1.856015102110863e-06,
+      "loss": 0.4264,
+      "step": 10933
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.000092506408691,
+      "learning_rate": 1.8534408786682685e-06,
+      "loss": 0.3998,
+      "step": 10934
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.815633773803711,
+      "learning_rate": 1.8508666552256737e-06,
+      "loss": 0.5304,
+      "step": 10935
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.508477210998535,
+      "learning_rate": 1.848292431783079e-06,
+      "loss": 0.2781,
+      "step": 10936
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.919846534729004,
+      "learning_rate": 1.845718208340484e-06,
+      "loss": 0.5268,
+      "step": 10937
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.646406173706055,
+      "learning_rate": 1.843143984897889e-06,
+      "loss": 0.5366,
+      "step": 10938
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.416731834411621,
+      "learning_rate": 1.8405697614552945e-06,
+      "loss": 0.5303,
+      "step": 10939
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.099388122558594,
+      "learning_rate": 1.8379955380126995e-06,
+      "loss": 0.3049,
+      "step": 10940
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.014859199523926,
+      "learning_rate": 1.8354213145701047e-06,
+      "loss": 0.2689,
+      "step": 10941
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.663738250732422,
+      "learning_rate": 1.8328470911275099e-06,
+      "loss": 0.4063,
+      "step": 10942
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.820629119873047,
+      "learning_rate": 1.830272867684915e-06,
+      "loss": 0.3735,
+      "step": 10943
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.905858039855957,
+      "learning_rate": 1.8276986442423203e-06,
+      "loss": 0.4487,
+      "step": 10944
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.693678855895996,
+      "learning_rate": 1.8251244207997255e-06,
+      "loss": 0.3747,
+      "step": 10945
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.894797325134277,
+      "learning_rate": 1.8225501973571305e-06,
+      "loss": 0.3577,
+      "step": 10946
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.730839729309082,
+      "learning_rate": 1.8199759739145359e-06,
+      "loss": 0.2407,
+      "step": 10947
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.638737678527832,
+      "learning_rate": 1.817401750471941e-06,
+      "loss": 0.6203,
+      "step": 10948
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.555971145629883,
+      "learning_rate": 1.8148275270293463e-06,
+      "loss": 0.272,
+      "step": 10949
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.601404190063477,
+      "learning_rate": 1.8122533035867513e-06,
+      "loss": 0.3058,
+      "step": 10950
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.684060096740723,
+      "learning_rate": 1.8096790801441565e-06,
+      "loss": 0.4711,
+      "step": 10951
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.077581405639648,
+      "learning_rate": 1.8071048567015619e-06,
+      "loss": 0.3769,
+      "step": 10952
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.523407936096191,
+      "learning_rate": 1.8045306332589669e-06,
+      "loss": 0.2662,
+      "step": 10953
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.453701972961426,
+      "learning_rate": 1.801956409816372e-06,
+      "loss": 0.3374,
+      "step": 10954
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.322737693786621,
+      "learning_rate": 1.7993821863737773e-06,
+      "loss": 0.409,
+      "step": 10955
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.129615783691406,
+      "learning_rate": 1.7968079629311825e-06,
+      "loss": 0.3592,
+      "step": 10956
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.893600463867188,
+      "learning_rate": 1.7942337394885877e-06,
+      "loss": 0.4282,
+      "step": 10957
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.170722007751465,
+      "learning_rate": 1.7916595160459929e-06,
+      "loss": 0.2453,
+      "step": 10958
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.843855857849121,
+      "learning_rate": 1.7890852926033979e-06,
+      "loss": 0.297,
+      "step": 10959
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.154544830322266,
+      "learning_rate": 1.7865110691608033e-06,
+      "loss": 0.2612,
+      "step": 10960
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.611052513122559,
+      "learning_rate": 1.7839368457182085e-06,
+      "loss": 0.3408,
+      "step": 10961
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.195291519165039,
+      "learning_rate": 1.7813626222756137e-06,
+      "loss": 0.2961,
+      "step": 10962
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.155021667480469,
+      "learning_rate": 1.7787883988330186e-06,
+      "loss": 0.2582,
+      "step": 10963
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.679885864257812,
+      "learning_rate": 1.7762141753904238e-06,
+      "loss": 0.2969,
+      "step": 10964
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.09830379486084,
+      "learning_rate": 1.7736399519478293e-06,
+      "loss": 0.2912,
+      "step": 10965
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.981791496276855,
+      "learning_rate": 1.7710657285052342e-06,
+      "loss": 0.3144,
+      "step": 10966
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.299041748046875,
+      "learning_rate": 1.7684915050626394e-06,
+      "loss": 0.4067,
+      "step": 10967
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.790837287902832,
+      "learning_rate": 1.7659172816200446e-06,
+      "loss": 0.3289,
+      "step": 10968
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.538524627685547,
+      "learning_rate": 1.7633430581774498e-06,
+      "loss": 0.3175,
+      "step": 10969
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.780279159545898,
+      "learning_rate": 1.760768834734855e-06,
+      "loss": 0.3629,
+      "step": 10970
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.61856746673584,
+      "learning_rate": 1.7581946112922602e-06,
+      "loss": 0.4479,
+      "step": 10971
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.927382469177246,
+      "learning_rate": 1.7556203878496652e-06,
+      "loss": 0.2956,
+      "step": 10972
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.458925247192383,
+      "learning_rate": 1.7530461644070706e-06,
+      "loss": 0.2484,
+      "step": 10973
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.202786445617676,
+      "learning_rate": 1.7504719409644758e-06,
+      "loss": 0.415,
+      "step": 10974
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.060401916503906,
+      "learning_rate": 1.747897717521881e-06,
+      "loss": 0.3452,
+      "step": 10975
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.941603660583496,
+      "learning_rate": 1.745323494079286e-06,
+      "loss": 0.48,
+      "step": 10976
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.705573081970215,
+      "learning_rate": 1.7427492706366912e-06,
+      "loss": 0.5061,
+      "step": 10977
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.721020698547363,
+      "learning_rate": 1.7401750471940966e-06,
+      "loss": 0.3106,
+      "step": 10978
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.752761840820312,
+      "learning_rate": 1.7376008237515016e-06,
+      "loss": 0.4474,
+      "step": 10979
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.939994812011719,
+      "learning_rate": 1.7350266003089068e-06,
+      "loss": 0.5227,
+      "step": 10980
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.127387046813965,
+      "learning_rate": 1.732452376866312e-06,
+      "loss": 0.6356,
+      "step": 10981
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.27786922454834,
+      "learning_rate": 1.7298781534237172e-06,
+      "loss": 0.4401,
+      "step": 10982
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.561688423156738,
+      "learning_rate": 1.7273039299811224e-06,
+      "loss": 0.328,
+      "step": 10983
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.691143035888672,
+      "learning_rate": 1.7247297065385276e-06,
+      "loss": 0.3278,
+      "step": 10984
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.351495742797852,
+      "learning_rate": 1.7221554830959326e-06,
+      "loss": 0.3805,
+      "step": 10985
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.489583969116211,
+      "learning_rate": 1.719581259653338e-06,
+      "loss": 0.2829,
+      "step": 10986
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.070416450500488,
+      "learning_rate": 1.7170070362107432e-06,
+      "loss": 0.2392,
+      "step": 10987
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.467737197875977,
+      "learning_rate": 1.7144328127681484e-06,
+      "loss": 0.3205,
+      "step": 10988
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.04840087890625,
+      "learning_rate": 1.7118585893255534e-06,
+      "loss": 0.397,
+      "step": 10989
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.456292152404785,
+      "learning_rate": 1.7092843658829588e-06,
+      "loss": 0.6786,
+      "step": 10990
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.699073791503906,
+      "learning_rate": 1.706710142440364e-06,
+      "loss": 0.4608,
+      "step": 10991
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.682391166687012,
+      "learning_rate": 1.704135918997769e-06,
+      "loss": 0.5574,
+      "step": 10992
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.479904174804688,
+      "learning_rate": 1.7015616955551742e-06,
+      "loss": 0.3646,
+      "step": 10993
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.163212776184082,
+      "learning_rate": 1.6989874721125794e-06,
+      "loss": 0.4495,
+      "step": 10994
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.567753791809082,
+      "learning_rate": 1.6964132486699846e-06,
+      "loss": 0.3117,
+      "step": 10995
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.026771545410156,
+      "learning_rate": 1.6938390252273898e-06,
+      "loss": 0.2662,
+      "step": 10996
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.500985145568848,
+      "learning_rate": 1.691264801784795e-06,
+      "loss": 0.511,
+      "step": 10997
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.901535034179688,
+      "learning_rate": 1.6886905783422e-06,
+      "loss": 0.4055,
+      "step": 10998
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.854602813720703,
+      "learning_rate": 1.6861163548996054e-06,
+      "loss": 0.405,
+      "step": 10999
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.663098335266113,
+      "learning_rate": 1.6835421314570106e-06,
+      "loss": 0.2184,
+      "step": 11000
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.24438190460205,
+      "learning_rate": 1.6809679080144156e-06,
+      "loss": 0.3804,
+      "step": 11001
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.910487174987793,
+      "learning_rate": 1.6783936845718208e-06,
+      "loss": 0.3671,
+      "step": 11002
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.503626823425293,
+      "learning_rate": 1.6758194611292262e-06,
+      "loss": 0.3057,
+      "step": 11003
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.804804801940918,
+      "learning_rate": 1.6732452376866314e-06,
+      "loss": 0.3449,
+      "step": 11004
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.306525230407715,
+      "learning_rate": 1.6706710142440364e-06,
+      "loss": 0.3832,
+      "step": 11005
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.503520965576172,
+      "learning_rate": 1.6680967908014416e-06,
+      "loss": 0.3029,
+      "step": 11006
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.522000312805176,
+      "learning_rate": 1.6655225673588468e-06,
+      "loss": 0.3566,
+      "step": 11007
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.334946632385254,
+      "learning_rate": 1.662948343916252e-06,
+      "loss": 0.3406,
+      "step": 11008
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.630239486694336,
+      "learning_rate": 1.6603741204736572e-06,
+      "loss": 0.3247,
+      "step": 11009
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.048665046691895,
+      "learning_rate": 1.6577998970310624e-06,
+      "loss": 0.2981,
+      "step": 11010
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.395168304443359,
+      "learning_rate": 1.6552256735884674e-06,
+      "loss": 0.2665,
+      "step": 11011
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.834129333496094,
+      "learning_rate": 1.6526514501458728e-06,
+      "loss": 0.2973,
+      "step": 11012
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.7796783447265625,
+      "learning_rate": 1.650077226703278e-06,
+      "loss": 0.3329,
+      "step": 11013
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.606521606445312,
+      "learning_rate": 1.647503003260683e-06,
+      "loss": 0.5186,
+      "step": 11014
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.214790344238281,
+      "learning_rate": 1.6449287798180882e-06,
+      "loss": 0.41,
+      "step": 11015
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.67099380493164,
+      "learning_rate": 1.6423545563754936e-06,
+      "loss": 0.4352,
+      "step": 11016
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.937704086303711,
+      "learning_rate": 1.6397803329328988e-06,
+      "loss": 0.4888,
+      "step": 11017
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.618767738342285,
+      "learning_rate": 1.6372061094903038e-06,
+      "loss": 0.3589,
+      "step": 11018
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.796728134155273,
+      "learning_rate": 1.634631886047709e-06,
+      "loss": 0.3502,
+      "step": 11019
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.404500484466553,
+      "learning_rate": 1.6320576626051142e-06,
+      "loss": 0.2487,
+      "step": 11020
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.910385131835938,
+      "learning_rate": 1.6294834391625193e-06,
+      "loss": 0.3413,
+      "step": 11021
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.829666137695312,
+      "learning_rate": 1.6269092157199245e-06,
+      "loss": 0.3072,
+      "step": 11022
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.673471927642822,
+      "learning_rate": 1.6243349922773297e-06,
+      "loss": 0.2127,
+      "step": 11023
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.837130546569824,
+      "learning_rate": 1.6217607688347347e-06,
+      "loss": 0.3571,
+      "step": 11024
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.69570541381836,
+      "learning_rate": 1.6191865453921401e-06,
+      "loss": 0.4973,
+      "step": 11025
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.976789474487305,
+      "learning_rate": 1.6166123219495453e-06,
+      "loss": 0.4729,
+      "step": 11026
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.088257789611816,
+      "learning_rate": 1.6140380985069503e-06,
+      "loss": 0.3667,
+      "step": 11027
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.058648109436035,
+      "learning_rate": 1.6114638750643555e-06,
+      "loss": 0.2157,
+      "step": 11028
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.641792297363281,
+      "learning_rate": 1.608889651621761e-06,
+      "loss": 0.3405,
+      "step": 11029
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.001545429229736,
+      "learning_rate": 1.6063154281791661e-06,
+      "loss": 0.2472,
+      "step": 11030
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.410353660583496,
+      "learning_rate": 1.6037412047365711e-06,
+      "loss": 0.4894,
+      "step": 11031
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.552005767822266,
+      "learning_rate": 1.6011669812939763e-06,
+      "loss": 0.4725,
+      "step": 11032
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.280860900878906,
+      "learning_rate": 1.5985927578513815e-06,
+      "loss": 0.5346,
+      "step": 11033
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.028883934020996,
+      "learning_rate": 1.5960185344087867e-06,
+      "loss": 0.3666,
+      "step": 11034
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.481372833251953,
+      "learning_rate": 1.593444310966192e-06,
+      "loss": 0.4333,
+      "step": 11035
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.702973365783691,
+      "learning_rate": 1.5908700875235971e-06,
+      "loss": 0.2984,
+      "step": 11036
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.116095542907715,
+      "learning_rate": 1.5882958640810021e-06,
+      "loss": 0.3005,
+      "step": 11037
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.090060234069824,
+      "learning_rate": 1.5857216406384075e-06,
+      "loss": 0.2367,
+      "step": 11038
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 14.669819831848145,
+      "learning_rate": 1.5831474171958127e-06,
+      "loss": 0.48,
+      "step": 11039
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.91246223449707,
+      "learning_rate": 1.5805731937532177e-06,
+      "loss": 0.2103,
+      "step": 11040
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.193394660949707,
+      "learning_rate": 1.577998970310623e-06,
+      "loss": 0.2906,
+      "step": 11041
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.936698913574219,
+      "learning_rate": 1.5754247468680283e-06,
+      "loss": 0.3704,
+      "step": 11042
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.773835182189941,
+      "learning_rate": 1.5728505234254335e-06,
+      "loss": 0.3421,
+      "step": 11043
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.810704231262207,
+      "learning_rate": 1.5702762999828385e-06,
+      "loss": 0.3327,
+      "step": 11044
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.33499526977539,
+      "learning_rate": 1.5677020765402437e-06,
+      "loss": 0.3793,
+      "step": 11045
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.707942008972168,
+      "learning_rate": 1.565127853097649e-06,
+      "loss": 0.3876,
+      "step": 11046
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.534838676452637,
+      "learning_rate": 1.562553629655054e-06,
+      "loss": 0.3662,
+      "step": 11047
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.828907012939453,
+      "learning_rate": 1.5599794062124593e-06,
+      "loss": 0.3065,
+      "step": 11048
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.582770347595215,
+      "learning_rate": 1.5574051827698645e-06,
+      "loss": 0.3508,
+      "step": 11049
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.885123252868652,
+      "learning_rate": 1.5548309593272695e-06,
+      "loss": 0.4519,
+      "step": 11050
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.808953285217285,
+      "learning_rate": 1.552256735884675e-06,
+      "loss": 0.3399,
+      "step": 11051
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 15.436074256896973,
+      "learning_rate": 1.54968251244208e-06,
+      "loss": 0.5282,
+      "step": 11052
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.765783786773682,
+      "learning_rate": 1.547108288999485e-06,
+      "loss": 0.1748,
+      "step": 11053
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.8151216506958,
+      "learning_rate": 1.5445340655568903e-06,
+      "loss": 0.3572,
+      "step": 11054
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.112709045410156,
+      "learning_rate": 1.5419598421142957e-06,
+      "loss": 0.2839,
+      "step": 11055
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.86545181274414,
+      "learning_rate": 1.539385618671701e-06,
+      "loss": 0.3148,
+      "step": 11056
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.274297714233398,
+      "learning_rate": 1.5368113952291059e-06,
+      "loss": 0.2154,
+      "step": 11057
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.033451080322266,
+      "learning_rate": 1.534237171786511e-06,
+      "loss": 0.3413,
+      "step": 11058
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.844738006591797,
+      "learning_rate": 1.5316629483439163e-06,
+      "loss": 0.3977,
+      "step": 11059
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.7937650680542,
+      "learning_rate": 1.5290887249013215e-06,
+      "loss": 0.3504,
+      "step": 11060
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.066435813903809,
+      "learning_rate": 1.5265145014587267e-06,
+      "loss": 0.4311,
+      "step": 11061
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.249638557434082,
+      "learning_rate": 1.5239402780161319e-06,
+      "loss": 0.3725,
+      "step": 11062
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.696634292602539,
+      "learning_rate": 1.5213660545735369e-06,
+      "loss": 0.3621,
+      "step": 11063
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.671460151672363,
+      "learning_rate": 1.5187918311309423e-06,
+      "loss": 0.3871,
+      "step": 11064
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.716444969177246,
+      "learning_rate": 1.5162176076883475e-06,
+      "loss": 0.3548,
+      "step": 11065
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.064604759216309,
+      "learning_rate": 1.5136433842457525e-06,
+      "loss": 0.5071,
+      "step": 11066
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.194353103637695,
+      "learning_rate": 1.5110691608031577e-06,
+      "loss": 0.5182,
+      "step": 11067
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.161458969116211,
+      "learning_rate": 1.508494937360563e-06,
+      "loss": 0.4692,
+      "step": 11068
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.44462776184082,
+      "learning_rate": 1.5059207139179683e-06,
+      "loss": 0.2873,
+      "step": 11069
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.861928939819336,
+      "learning_rate": 1.5033464904753733e-06,
+      "loss": 0.4623,
+      "step": 11070
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.47918701171875,
+      "learning_rate": 1.5007722670327785e-06,
+      "loss": 0.3618,
+      "step": 11071
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.619352340698242,
+      "learning_rate": 1.4981980435901837e-06,
+      "loss": 0.3762,
+      "step": 11072
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.443218231201172,
+      "learning_rate": 1.4956238201475889e-06,
+      "loss": 0.5267,
+      "step": 11073
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.459564208984375,
+      "learning_rate": 1.493049596704994e-06,
+      "loss": 0.3748,
+      "step": 11074
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.884991645812988,
+      "learning_rate": 1.4904753732623993e-06,
+      "loss": 0.2525,
+      "step": 11075
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.193331718444824,
+      "learning_rate": 1.4879011498198042e-06,
+      "loss": 0.4248,
+      "step": 11076
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.723840713500977,
+      "learning_rate": 1.4853269263772097e-06,
+      "loss": 0.353,
+      "step": 11077
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.666188716888428,
+      "learning_rate": 1.4827527029346149e-06,
+      "loss": 0.2274,
+      "step": 11078
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.302162170410156,
+      "learning_rate": 1.4801784794920198e-06,
+      "loss": 0.3173,
+      "step": 11079
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.9813871383667,
+      "learning_rate": 1.477604256049425e-06,
+      "loss": 0.3018,
+      "step": 11080
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.430132865905762,
+      "learning_rate": 1.4750300326068304e-06,
+      "loss": 0.3852,
+      "step": 11081
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.488740921020508,
+      "learning_rate": 1.4724558091642354e-06,
+      "loss": 0.2931,
+      "step": 11082
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.134601593017578,
+      "learning_rate": 1.4698815857216406e-06,
+      "loss": 0.426,
+      "step": 11083
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.19254732131958,
+      "learning_rate": 1.4673073622790458e-06,
+      "loss": 0.2565,
+      "step": 11084
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.47159481048584,
+      "learning_rate": 1.464733138836451e-06,
+      "loss": 0.3586,
+      "step": 11085
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.662041664123535,
+      "learning_rate": 1.4621589153938562e-06,
+      "loss": 0.2908,
+      "step": 11086
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.238934516906738,
+      "learning_rate": 1.4595846919512614e-06,
+      "loss": 0.3454,
+      "step": 11087
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.97555923461914,
+      "learning_rate": 1.4570104685086666e-06,
+      "loss": 0.3494,
+      "step": 11088
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.415300369262695,
+      "learning_rate": 1.4544362450660716e-06,
+      "loss": 0.2691,
+      "step": 11089
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.385876655578613,
+      "learning_rate": 1.451862021623477e-06,
+      "loss": 0.2755,
+      "step": 11090
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.479072570800781,
+      "learning_rate": 1.4492877981808822e-06,
+      "loss": 0.3505,
+      "step": 11091
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.806800842285156,
+      "learning_rate": 1.4467135747382872e-06,
+      "loss": 0.5207,
+      "step": 11092
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.432601928710938,
+      "learning_rate": 1.4441393512956924e-06,
+      "loss": 0.4147,
+      "step": 11093
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.812528610229492,
+      "learning_rate": 1.4415651278530978e-06,
+      "loss": 0.2677,
+      "step": 11094
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.20866870880127,
+      "learning_rate": 1.4389909044105028e-06,
+      "loss": 0.4399,
+      "step": 11095
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.870978355407715,
+      "learning_rate": 1.436416680967908e-06,
+      "loss": 0.2924,
+      "step": 11096
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.04464054107666,
+      "learning_rate": 1.4338424575253132e-06,
+      "loss": 0.2668,
+      "step": 11097
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.724349498748779,
+      "learning_rate": 1.4312682340827184e-06,
+      "loss": 0.2505,
+      "step": 11098
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.885076522827148,
+      "learning_rate": 1.4286940106401236e-06,
+      "loss": 0.3736,
+      "step": 11099
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.8551607131958,
+      "learning_rate": 1.4261197871975288e-06,
+      "loss": 0.4088,
+      "step": 11100
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.75053882598877,
+      "learning_rate": 1.423545563754934e-06,
+      "loss": 0.4597,
+      "step": 11101
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.84624671936035,
+      "learning_rate": 1.420971340312339e-06,
+      "loss": 0.633,
+      "step": 11102
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.965681076049805,
+      "learning_rate": 1.4183971168697444e-06,
+      "loss": 0.3649,
+      "step": 11103
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.497034072875977,
+      "learning_rate": 1.4158228934271496e-06,
+      "loss": 0.3716,
+      "step": 11104
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.71844482421875,
+      "learning_rate": 1.4132486699845546e-06,
+      "loss": 0.3703,
+      "step": 11105
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.784379959106445,
+      "learning_rate": 1.4106744465419598e-06,
+      "loss": 0.2385,
+      "step": 11106
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.765949249267578,
+      "learning_rate": 1.4081002230993652e-06,
+      "loss": 0.3182,
+      "step": 11107
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.988414764404297,
+      "learning_rate": 1.4055259996567702e-06,
+      "loss": 0.3591,
+      "step": 11108
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.155403137207031,
+      "learning_rate": 1.4029517762141754e-06,
+      "loss": 0.6233,
+      "step": 11109
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.289401054382324,
+      "learning_rate": 1.4003775527715806e-06,
+      "loss": 0.2637,
+      "step": 11110
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.276763439178467,
+      "learning_rate": 1.3978033293289858e-06,
+      "loss": 0.382,
+      "step": 11111
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.864279747009277,
+      "learning_rate": 1.395229105886391e-06,
+      "loss": 0.2796,
+      "step": 11112
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.87149715423584,
+      "learning_rate": 1.3926548824437962e-06,
+      "loss": 0.3887,
+      "step": 11113
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.398486137390137,
+      "learning_rate": 1.3900806590012014e-06,
+      "loss": 0.3937,
+      "step": 11114
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.466192245483398,
+      "learning_rate": 1.3875064355586064e-06,
+      "loss": 0.4295,
+      "step": 11115
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.146299362182617,
+      "learning_rate": 1.3849322121160118e-06,
+      "loss": 0.2447,
+      "step": 11116
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.072288513183594,
+      "learning_rate": 1.382357988673417e-06,
+      "loss": 0.578,
+      "step": 11117
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.480597496032715,
+      "learning_rate": 1.379783765230822e-06,
+      "loss": 0.3693,
+      "step": 11118
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.855088233947754,
+      "learning_rate": 1.3772095417882272e-06,
+      "loss": 0.3826,
+      "step": 11119
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.57455062866211,
+      "learning_rate": 1.3746353183456326e-06,
+      "loss": 0.5271,
+      "step": 11120
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.553418159484863,
+      "learning_rate": 1.3720610949030376e-06,
+      "loss": 0.4354,
+      "step": 11121
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.02069091796875,
+      "learning_rate": 1.3694868714604428e-06,
+      "loss": 0.5303,
+      "step": 11122
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.128046035766602,
+      "learning_rate": 1.366912648017848e-06,
+      "loss": 0.279,
+      "step": 11123
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.339672088623047,
+      "learning_rate": 1.3643384245752532e-06,
+      "loss": 0.3368,
+      "step": 11124
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.95443058013916,
+      "learning_rate": 1.3617642011326584e-06,
+      "loss": 0.5186,
+      "step": 11125
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.590944290161133,
+      "learning_rate": 1.3591899776900636e-06,
+      "loss": 0.4422,
+      "step": 11126
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.974928855895996,
+      "learning_rate": 1.3566157542474688e-06,
+      "loss": 0.4472,
+      "step": 11127
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.079082489013672,
+      "learning_rate": 1.3540415308048737e-06,
+      "loss": 0.2701,
+      "step": 11128
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.837074279785156,
+      "learning_rate": 1.3514673073622792e-06,
+      "loss": 0.36,
+      "step": 11129
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.154749870300293,
+      "learning_rate": 1.3488930839196844e-06,
+      "loss": 0.2335,
+      "step": 11130
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.293107032775879,
+      "learning_rate": 1.3463188604770893e-06,
+      "loss": 0.3503,
+      "step": 11131
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.760516166687012,
+      "learning_rate": 1.3437446370344945e-06,
+      "loss": 0.4766,
+      "step": 11132
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.617878913879395,
+      "learning_rate": 1.3411704135919e-06,
+      "loss": 0.3759,
+      "step": 11133
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.617166519165039,
+      "learning_rate": 1.338596190149305e-06,
+      "loss": 0.3392,
+      "step": 11134
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.61210823059082,
+      "learning_rate": 1.3360219667067101e-06,
+      "loss": 0.6003,
+      "step": 11135
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.657895088195801,
+      "learning_rate": 1.3334477432641153e-06,
+      "loss": 0.2443,
+      "step": 11136
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.559221267700195,
+      "learning_rate": 1.3308735198215205e-06,
+      "loss": 0.3253,
+      "step": 11137
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.400186538696289,
+      "learning_rate": 1.3282992963789257e-06,
+      "loss": 0.158,
+      "step": 11138
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.523914337158203,
+      "learning_rate": 1.325725072936331e-06,
+      "loss": 0.4054,
+      "step": 11139
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.095930099487305,
+      "learning_rate": 1.3231508494937361e-06,
+      "loss": 0.2519,
+      "step": 11140
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.52552318572998,
+      "learning_rate": 1.3205766260511411e-06,
+      "loss": 0.3217,
+      "step": 11141
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.015631675720215,
+      "learning_rate": 1.3180024026085465e-06,
+      "loss": 0.3141,
+      "step": 11142
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.959336280822754,
+      "learning_rate": 1.3154281791659517e-06,
+      "loss": 0.2927,
+      "step": 11143
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.793444633483887,
+      "learning_rate": 1.3128539557233567e-06,
+      "loss": 0.4361,
+      "step": 11144
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.20043659210205,
+      "learning_rate": 1.310279732280762e-06,
+      "loss": 0.2795,
+      "step": 11145
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.522153854370117,
+      "learning_rate": 1.3077055088381673e-06,
+      "loss": 0.3831,
+      "step": 11146
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.44690227508545,
+      "learning_rate": 1.3051312853955723e-06,
+      "loss": 0.239,
+      "step": 11147
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.916020393371582,
+      "learning_rate": 1.3025570619529775e-06,
+      "loss": 0.2487,
+      "step": 11148
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.011587142944336,
+      "learning_rate": 1.2999828385103827e-06,
+      "loss": 0.5096,
+      "step": 11149
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.361563682556152,
+      "learning_rate": 1.297408615067788e-06,
+      "loss": 0.4364,
+      "step": 11150
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.001625061035156,
+      "learning_rate": 1.2948343916251931e-06,
+      "loss": 0.2891,
+      "step": 11151
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.219432830810547,
+      "learning_rate": 1.2922601681825983e-06,
+      "loss": 0.3015,
+      "step": 11152
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.579834938049316,
+      "learning_rate": 1.2896859447400035e-06,
+      "loss": 0.2665,
+      "step": 11153
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.218266487121582,
+      "learning_rate": 1.2871117212974085e-06,
+      "loss": 0.3915,
+      "step": 11154
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.835561752319336,
+      "learning_rate": 1.284537497854814e-06,
+      "loss": 0.2177,
+      "step": 11155
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.71733283996582,
+      "learning_rate": 1.2819632744122191e-06,
+      "loss": 0.3994,
+      "step": 11156
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.122781753540039,
+      "learning_rate": 1.279389050969624e-06,
+      "loss": 0.2778,
+      "step": 11157
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.859009742736816,
+      "learning_rate": 1.2768148275270293e-06,
+      "loss": 0.5948,
+      "step": 11158
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.164999961853027,
+      "learning_rate": 1.2742406040844347e-06,
+      "loss": 0.2499,
+      "step": 11159
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.167304039001465,
+      "learning_rate": 1.2716663806418397e-06,
+      "loss": 0.4165,
+      "step": 11160
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.132265090942383,
+      "learning_rate": 1.2690921571992449e-06,
+      "loss": 0.2645,
+      "step": 11161
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.252290725708008,
+      "learning_rate": 1.26651793375665e-06,
+      "loss": 0.3353,
+      "step": 11162
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.0601806640625,
+      "learning_rate": 1.2639437103140553e-06,
+      "loss": 0.4035,
+      "step": 11163
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.308443069458008,
+      "learning_rate": 1.2613694868714605e-06,
+      "loss": 0.4274,
+      "step": 11164
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.666142463684082,
+      "learning_rate": 1.2587952634288657e-06,
+      "loss": 0.611,
+      "step": 11165
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.5977201461792,
+      "learning_rate": 1.2562210399862709e-06,
+      "loss": 0.3677,
+      "step": 11166
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.810057640075684,
+      "learning_rate": 1.2536468165436759e-06,
+      "loss": 0.3956,
+      "step": 11167
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.103277206420898,
+      "learning_rate": 1.2510725931010813e-06,
+      "loss": 0.2853,
+      "step": 11168
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.228283882141113,
+      "learning_rate": 1.2484983696584865e-06,
+      "loss": 0.2452,
+      "step": 11169
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.345582962036133,
+      "learning_rate": 1.2459241462158915e-06,
+      "loss": 0.3881,
+      "step": 11170
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.230448722839355,
+      "learning_rate": 1.2433499227732967e-06,
+      "loss": 0.2997,
+      "step": 11171
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.033206939697266,
+      "learning_rate": 1.240775699330702e-06,
+      "loss": 0.333,
+      "step": 11172
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.548929214477539,
+      "learning_rate": 1.238201475888107e-06,
+      "loss": 0.3719,
+      "step": 11173
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.063002586364746,
+      "learning_rate": 1.2356272524455123e-06,
+      "loss": 0.374,
+      "step": 11174
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.224303245544434,
+      "learning_rate": 1.2330530290029175e-06,
+      "loss": 0.4004,
+      "step": 11175
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.494648933410645,
+      "learning_rate": 1.2304788055603227e-06,
+      "loss": 0.4149,
+      "step": 11176
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.038595199584961,
+      "learning_rate": 1.2279045821177279e-06,
+      "loss": 0.2261,
+      "step": 11177
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.000349998474121,
+      "learning_rate": 1.225330358675133e-06,
+      "loss": 0.4082,
+      "step": 11178
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.931147575378418,
+      "learning_rate": 1.2227561352325383e-06,
+      "loss": 0.2104,
+      "step": 11179
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.060903549194336,
+      "learning_rate": 1.2201819117899433e-06,
+      "loss": 0.2106,
+      "step": 11180
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.333086013793945,
+      "learning_rate": 1.2176076883473487e-06,
+      "loss": 0.382,
+      "step": 11181
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.011672973632812,
+      "learning_rate": 1.2150334649047539e-06,
+      "loss": 0.3036,
+      "step": 11182
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.26968002319336,
+      "learning_rate": 1.2124592414621588e-06,
+      "loss": 0.294,
+      "step": 11183
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.636792182922363,
+      "learning_rate": 1.209885018019564e-06,
+      "loss": 0.4169,
+      "step": 11184
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.713167190551758,
+      "learning_rate": 1.2073107945769695e-06,
+      "loss": 0.3351,
+      "step": 11185
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.600278854370117,
+      "learning_rate": 1.2047365711343744e-06,
+      "loss": 0.4731,
+      "step": 11186
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.932683944702148,
+      "learning_rate": 1.2021623476917796e-06,
+      "loss": 0.3507,
+      "step": 11187
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.079802513122559,
+      "learning_rate": 1.1995881242491848e-06,
+      "loss": 0.3308,
+      "step": 11188
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.066193580627441,
+      "learning_rate": 1.19701390080659e-06,
+      "loss": 0.4347,
+      "step": 11189
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.60873031616211,
+      "learning_rate": 1.1944396773639952e-06,
+      "loss": 0.4299,
+      "step": 11190
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.062819480895996,
+      "learning_rate": 1.1918654539214004e-06,
+      "loss": 0.3523,
+      "step": 11191
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.772015571594238,
+      "learning_rate": 1.1892912304788056e-06,
+      "loss": 0.2923,
+      "step": 11192
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.713652610778809,
+      "learning_rate": 1.1867170070362106e-06,
+      "loss": 0.3126,
+      "step": 11193
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.670111656188965,
+      "learning_rate": 1.184142783593616e-06,
+      "loss": 0.3533,
+      "step": 11194
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.896197319030762,
+      "learning_rate": 1.1815685601510212e-06,
+      "loss": 0.4245,
+      "step": 11195
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.288776397705078,
+      "learning_rate": 1.1789943367084262e-06,
+      "loss": 0.4578,
+      "step": 11196
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.529693603515625,
+      "learning_rate": 1.1764201132658314e-06,
+      "loss": 0.3531,
+      "step": 11197
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.612212181091309,
+      "learning_rate": 1.1738458898232368e-06,
+      "loss": 0.2996,
+      "step": 11198
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.519620895385742,
+      "learning_rate": 1.1712716663806418e-06,
+      "loss": 0.2407,
+      "step": 11199
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.2696943283081055,
+      "learning_rate": 1.168697442938047e-06,
+      "loss": 0.236,
+      "step": 11200
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.432975769042969,
+      "learning_rate": 1.1661232194954522e-06,
+      "loss": 0.2414,
+      "step": 11201
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.555484771728516,
+      "learning_rate": 1.1635489960528574e-06,
+      "loss": 0.4583,
+      "step": 11202
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.341927528381348,
+      "learning_rate": 1.1609747726102626e-06,
+      "loss": 0.3313,
+      "step": 11203
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.018180847167969,
+      "learning_rate": 1.1584005491676678e-06,
+      "loss": 0.423,
+      "step": 11204
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.9628214836120605,
+      "learning_rate": 1.155826325725073e-06,
+      "loss": 0.3947,
+      "step": 11205
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.402603149414062,
+      "learning_rate": 1.153252102282478e-06,
+      "loss": 0.2585,
+      "step": 11206
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.141619682312012,
+      "learning_rate": 1.1506778788398834e-06,
+      "loss": 0.3434,
+      "step": 11207
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.781804084777832,
+      "learning_rate": 1.1481036553972886e-06,
+      "loss": 0.2694,
+      "step": 11208
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.032614707946777,
+      "learning_rate": 1.1455294319546936e-06,
+      "loss": 0.4094,
+      "step": 11209
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.427877426147461,
+      "learning_rate": 1.1429552085120988e-06,
+      "loss": 0.2521,
+      "step": 11210
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.758294105529785,
+      "learning_rate": 1.1403809850695042e-06,
+      "loss": 0.3904,
+      "step": 11211
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.965252876281738,
+      "learning_rate": 1.1378067616269092e-06,
+      "loss": 0.3828,
+      "step": 11212
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.592247009277344,
+      "learning_rate": 1.1352325381843144e-06,
+      "loss": 0.3502,
+      "step": 11213
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.019633293151855,
+      "learning_rate": 1.1326583147417196e-06,
+      "loss": 0.3179,
+      "step": 11214
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.783926963806152,
+      "learning_rate": 1.1300840912991248e-06,
+      "loss": 0.3423,
+      "step": 11215
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.411629676818848,
+      "learning_rate": 1.12750986785653e-06,
+      "loss": 0.3384,
+      "step": 11216
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.852890014648438,
+      "learning_rate": 1.1249356444139352e-06,
+      "loss": 0.3336,
+      "step": 11217
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.638039588928223,
+      "learning_rate": 1.1223614209713404e-06,
+      "loss": 0.3948,
+      "step": 11218
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.062429428100586,
+      "learning_rate": 1.1197871975287454e-06,
+      "loss": 0.444,
+      "step": 11219
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.246325492858887,
+      "learning_rate": 1.1172129740861508e-06,
+      "loss": 0.3967,
+      "step": 11220
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.875120162963867,
+      "learning_rate": 1.114638750643556e-06,
+      "loss": 0.4064,
+      "step": 11221
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.381048202514648,
+      "learning_rate": 1.112064527200961e-06,
+      "loss": 0.2384,
+      "step": 11222
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.927454948425293,
+      "learning_rate": 1.1094903037583662e-06,
+      "loss": 0.3917,
+      "step": 11223
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.526918411254883,
+      "learning_rate": 1.1069160803157716e-06,
+      "loss": 0.3228,
+      "step": 11224
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.221765518188477,
+      "learning_rate": 1.1043418568731766e-06,
+      "loss": 0.3269,
+      "step": 11225
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.908487319946289,
+      "learning_rate": 1.1017676334305818e-06,
+      "loss": 0.2706,
+      "step": 11226
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.794954299926758,
+      "learning_rate": 1.099193409987987e-06,
+      "loss": 0.205,
+      "step": 11227
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.273452758789062,
+      "learning_rate": 1.0966191865453922e-06,
+      "loss": 0.4591,
+      "step": 11228
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.273874282836914,
+      "learning_rate": 1.0940449631027974e-06,
+      "loss": 0.4452,
+      "step": 11229
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.868019104003906,
+      "learning_rate": 1.0914707396602026e-06,
+      "loss": 0.3404,
+      "step": 11230
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.350021362304688,
+      "learning_rate": 1.0888965162176078e-06,
+      "loss": 0.3758,
+      "step": 11231
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.006874084472656,
+      "learning_rate": 1.0863222927750128e-06,
+      "loss": 0.4104,
+      "step": 11232
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.243325233459473,
+      "learning_rate": 1.0837480693324182e-06,
+      "loss": 0.3459,
+      "step": 11233
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.974726676940918,
+      "learning_rate": 1.0811738458898234e-06,
+      "loss": 0.3966,
+      "step": 11234
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.79494571685791,
+      "learning_rate": 1.0785996224472284e-06,
+      "loss": 0.4917,
+      "step": 11235
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.913002014160156,
+      "learning_rate": 1.0760253990046336e-06,
+      "loss": 0.5121,
+      "step": 11236
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.53303337097168,
+      "learning_rate": 1.073451175562039e-06,
+      "loss": 0.3243,
+      "step": 11237
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.39658546447754,
+      "learning_rate": 1.070876952119444e-06,
+      "loss": 0.269,
+      "step": 11238
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.369318008422852,
+      "learning_rate": 1.0683027286768492e-06,
+      "loss": 0.3039,
+      "step": 11239
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.533215522766113,
+      "learning_rate": 1.0657285052342544e-06,
+      "loss": 0.4033,
+      "step": 11240
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.617938041687012,
+      "learning_rate": 1.0631542817916596e-06,
+      "loss": 0.4906,
+      "step": 11241
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.701202392578125,
+      "learning_rate": 1.0605800583490647e-06,
+      "loss": 0.349,
+      "step": 11242
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.699152946472168,
+      "learning_rate": 1.05800583490647e-06,
+      "loss": 0.2965,
+      "step": 11243
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.104108810424805,
+      "learning_rate": 1.0554316114638751e-06,
+      "loss": 0.4275,
+      "step": 11244
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.235733032226562,
+      "learning_rate": 1.0528573880212801e-06,
+      "loss": 0.4241,
+      "step": 11245
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.381830215454102,
+      "learning_rate": 1.0502831645786855e-06,
+      "loss": 0.3633,
+      "step": 11246
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.662819862365723,
+      "learning_rate": 1.0477089411360907e-06,
+      "loss": 0.444,
+      "step": 11247
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.038917541503906,
+      "learning_rate": 1.0451347176934957e-06,
+      "loss": 0.3129,
+      "step": 11248
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.602962493896484,
+      "learning_rate": 1.042560494250901e-06,
+      "loss": 0.2215,
+      "step": 11249
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.234943389892578,
+      "learning_rate": 1.0399862708083063e-06,
+      "loss": 0.4591,
+      "step": 11250
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.435638427734375,
+      "learning_rate": 1.0374120473657113e-06,
+      "loss": 0.3963,
+      "step": 11251
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.230363845825195,
+      "learning_rate": 1.0348378239231165e-06,
+      "loss": 0.3475,
+      "step": 11252
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.6542887687683105,
+      "learning_rate": 1.0322636004805217e-06,
+      "loss": 0.2066,
+      "step": 11253
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.208277702331543,
+      "learning_rate": 1.029689377037927e-06,
+      "loss": 0.587,
+      "step": 11254
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.624720573425293,
+      "learning_rate": 1.0271151535953321e-06,
+      "loss": 0.4105,
+      "step": 11255
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.926789283752441,
+      "learning_rate": 1.0245409301527373e-06,
+      "loss": 0.2069,
+      "step": 11256
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.228850364685059,
+      "learning_rate": 1.0219667067101423e-06,
+      "loss": 0.2753,
+      "step": 11257
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.834033966064453,
+      "learning_rate": 1.0193924832675475e-06,
+      "loss": 0.5825,
+      "step": 11258
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.35616397857666,
+      "learning_rate": 1.016818259824953e-06,
+      "loss": 0.2947,
+      "step": 11259
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.541860580444336,
+      "learning_rate": 1.0142440363823581e-06,
+      "loss": 0.3636,
+      "step": 11260
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.137368202209473,
+      "learning_rate": 1.0116698129397631e-06,
+      "loss": 0.3446,
+      "step": 11261
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.032513618469238,
+      "learning_rate": 1.0090955894971683e-06,
+      "loss": 0.3341,
+      "step": 11262
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.584372520446777,
+      "learning_rate": 1.0065213660545737e-06,
+      "loss": 0.4229,
+      "step": 11263
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.587027549743652,
+      "learning_rate": 1.0039471426119787e-06,
+      "loss": 0.3627,
+      "step": 11264
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.200761795043945,
+      "learning_rate": 1.001372919169384e-06,
+      "loss": 0.3423,
+      "step": 11265
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.629083633422852,
+      "learning_rate": 9.98798695726789e-07,
+      "loss": 0.4505,
+      "step": 11266
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.095871925354004,
+      "learning_rate": 9.962244722841943e-07,
+      "loss": 0.3581,
+      "step": 11267
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.397126197814941,
+      "learning_rate": 9.936502488415995e-07,
+      "loss": 0.2732,
+      "step": 11268
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.796599388122559,
+      "learning_rate": 9.910760253990047e-07,
+      "loss": 0.2283,
+      "step": 11269
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.334604263305664,
+      "learning_rate": 9.885018019564097e-07,
+      "loss": 0.3054,
+      "step": 11270
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.330229759216309,
+      "learning_rate": 9.859275785138149e-07,
+      "loss": 0.3471,
+      "step": 11271
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.068267822265625,
+      "learning_rate": 9.833533550712203e-07,
+      "loss": 0.3788,
+      "step": 11272
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.766244411468506,
+      "learning_rate": 9.807791316286255e-07,
+      "loss": 0.2363,
+      "step": 11273
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.55380630493164,
+      "learning_rate": 9.782049081860305e-07,
+      "loss": 0.4018,
+      "step": 11274
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.42378044128418,
+      "learning_rate": 9.756306847434357e-07,
+      "loss": 0.3952,
+      "step": 11275
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.504379272460938,
+      "learning_rate": 9.73056461300841e-07,
+      "loss": 0.3481,
+      "step": 11276
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.353805541992188,
+      "learning_rate": 9.70482237858246e-07,
+      "loss": 0.5317,
+      "step": 11277
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.899829864501953,
+      "learning_rate": 9.679080144156513e-07,
+      "loss": 0.3624,
+      "step": 11278
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.905879974365234,
+      "learning_rate": 9.653337909730565e-07,
+      "loss": 0.4465,
+      "step": 11279
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.24138355255127,
+      "learning_rate": 9.627595675304617e-07,
+      "loss": 0.3696,
+      "step": 11280
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.197113990783691,
+      "learning_rate": 9.601853440878669e-07,
+      "loss": 0.3602,
+      "step": 11281
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.318521499633789,
+      "learning_rate": 9.57611120645272e-07,
+      "loss": 0.5945,
+      "step": 11282
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.276877403259277,
+      "learning_rate": 9.55036897202677e-07,
+      "loss": 0.2997,
+      "step": 11283
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.022278785705566,
+      "learning_rate": 9.524626737600824e-07,
+      "loss": 0.3282,
+      "step": 11284
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.468074798583984,
+      "learning_rate": 9.498884503174876e-07,
+      "loss": 0.318,
+      "step": 11285
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.843952178955078,
+      "learning_rate": 9.473142268748929e-07,
+      "loss": 0.3547,
+      "step": 11286
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.174036026000977,
+      "learning_rate": 9.447400034322979e-07,
+      "loss": 0.3074,
+      "step": 11287
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.60664176940918,
+      "learning_rate": 9.421657799897032e-07,
+      "loss": 0.4209,
+      "step": 11288
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.336174011230469,
+      "learning_rate": 9.395915565471084e-07,
+      "loss": 0.423,
+      "step": 11289
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.120474815368652,
+      "learning_rate": 9.370173331045136e-07,
+      "loss": 0.3692,
+      "step": 11290
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.143680572509766,
+      "learning_rate": 9.344431096619187e-07,
+      "loss": 0.3587,
+      "step": 11291
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.772634506225586,
+      "learning_rate": 9.318688862193239e-07,
+      "loss": 0.3199,
+      "step": 11292
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.608824729919434,
+      "learning_rate": 9.292946627767291e-07,
+      "loss": 0.3796,
+      "step": 11293
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.068592071533203,
+      "learning_rate": 9.267204393341343e-07,
+      "loss": 0.3352,
+      "step": 11294
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.383529663085938,
+      "learning_rate": 9.241462158915395e-07,
+      "loss": 0.318,
+      "step": 11295
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.84504508972168,
+      "learning_rate": 9.215719924489445e-07,
+      "loss": 0.3878,
+      "step": 11296
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.539249897003174,
+      "learning_rate": 9.189977690063497e-07,
+      "loss": 0.2165,
+      "step": 11297
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.671634674072266,
+      "learning_rate": 9.164235455637549e-07,
+      "loss": 0.3436,
+      "step": 11298
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.809889793395996,
+      "learning_rate": 9.138493221211601e-07,
+      "loss": 0.4075,
+      "step": 11299
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.600805282592773,
+      "learning_rate": 9.112750986785652e-07,
+      "loss": 0.3933,
+      "step": 11300
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.200301170349121,
+      "learning_rate": 9.087008752359705e-07,
+      "loss": 0.3252,
+      "step": 11301
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.493424415588379,
+      "learning_rate": 9.061266517933756e-07,
+      "loss": 0.3976,
+      "step": 11302
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.592477798461914,
+      "learning_rate": 9.035524283507809e-07,
+      "loss": 0.4543,
+      "step": 11303
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.08369255065918,
+      "learning_rate": 9.00978204908186e-07,
+      "loss": 0.3364,
+      "step": 11304
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.590408325195312,
+      "learning_rate": 8.984039814655912e-07,
+      "loss": 0.5278,
+      "step": 11305
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.600781440734863,
+      "learning_rate": 8.958297580229964e-07,
+      "loss": 0.2189,
+      "step": 11306
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.556259155273438,
+      "learning_rate": 8.932555345804016e-07,
+      "loss": 0.2317,
+      "step": 11307
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.474226951599121,
+      "learning_rate": 8.906813111378068e-07,
+      "loss": 0.3711,
+      "step": 11308
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.599727630615234,
+      "learning_rate": 8.881070876952119e-07,
+      "loss": 0.3542,
+      "step": 11309
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.618650436401367,
+      "learning_rate": 8.855328642526171e-07,
+      "loss": 0.3074,
+      "step": 11310
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.354700088500977,
+      "learning_rate": 8.829586408100223e-07,
+      "loss": 0.3029,
+      "step": 11311
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.634174346923828,
+      "learning_rate": 8.803844173674275e-07,
+      "loss": 0.336,
+      "step": 11312
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.879085540771484,
+      "learning_rate": 8.778101939248326e-07,
+      "loss": 0.3534,
+      "step": 11313
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.613602638244629,
+      "learning_rate": 8.752359704822379e-07,
+      "loss": 0.3587,
+      "step": 11314
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.624441146850586,
+      "learning_rate": 8.72661747039643e-07,
+      "loss": 0.268,
+      "step": 11315
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.6231050491333,
+      "learning_rate": 8.700875235970483e-07,
+      "loss": 0.2694,
+      "step": 11316
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.00921630859375,
+      "learning_rate": 8.675133001544534e-07,
+      "loss": 0.4135,
+      "step": 11317
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.385370254516602,
+      "learning_rate": 8.649390767118586e-07,
+      "loss": 0.4183,
+      "step": 11318
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.136366844177246,
+      "learning_rate": 8.623648532692638e-07,
+      "loss": 0.4017,
+      "step": 11319
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.964530944824219,
+      "learning_rate": 8.59790629826669e-07,
+      "loss": 0.3124,
+      "step": 11320
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.475838661193848,
+      "learning_rate": 8.572164063840742e-07,
+      "loss": 0.5261,
+      "step": 11321
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.14635181427002,
+      "learning_rate": 8.546421829414794e-07,
+      "loss": 0.3229,
+      "step": 11322
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.979029655456543,
+      "learning_rate": 8.520679594988845e-07,
+      "loss": 0.3312,
+      "step": 11323
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.051238059997559,
+      "learning_rate": 8.494937360562897e-07,
+      "loss": 0.443,
+      "step": 11324
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.99988842010498,
+      "learning_rate": 8.469195126136949e-07,
+      "loss": 0.3289,
+      "step": 11325
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.240226745605469,
+      "learning_rate": 8.443452891711e-07,
+      "loss": 0.3315,
+      "step": 11326
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.262712478637695,
+      "learning_rate": 8.417710657285053e-07,
+      "loss": 0.3172,
+      "step": 11327
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.68488597869873,
+      "learning_rate": 8.391968422859104e-07,
+      "loss": 0.3532,
+      "step": 11328
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.156920433044434,
+      "learning_rate": 8.366226188433157e-07,
+      "loss": 0.4471,
+      "step": 11329
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.043469429016113,
+      "learning_rate": 8.340483954007208e-07,
+      "loss": 0.343,
+      "step": 11330
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.056516647338867,
+      "learning_rate": 8.31474171958126e-07,
+      "loss": 0.2761,
+      "step": 11331
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.607675552368164,
+      "learning_rate": 8.288999485155312e-07,
+      "loss": 0.3714,
+      "step": 11332
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.879067420959473,
+      "learning_rate": 8.263257250729364e-07,
+      "loss": 0.3696,
+      "step": 11333
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.252850532531738,
+      "learning_rate": 8.237515016303415e-07,
+      "loss": 0.6638,
+      "step": 11334
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.101922988891602,
+      "learning_rate": 8.211772781877468e-07,
+      "loss": 0.326,
+      "step": 11335
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.902658462524414,
+      "learning_rate": 8.186030547451519e-07,
+      "loss": 0.3754,
+      "step": 11336
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.206715106964111,
+      "learning_rate": 8.160288313025571e-07,
+      "loss": 0.4128,
+      "step": 11337
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.075551509857178,
+      "learning_rate": 8.134546078599623e-07,
+      "loss": 0.1763,
+      "step": 11338
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.898223876953125,
+      "learning_rate": 8.108803844173674e-07,
+      "loss": 0.3561,
+      "step": 11339
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.690729141235352,
+      "learning_rate": 8.083061609747727e-07,
+      "loss": 0.3557,
+      "step": 11340
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.656914710998535,
+      "learning_rate": 8.057319375321778e-07,
+      "loss": 0.3916,
+      "step": 11341
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.575813293457031,
+      "learning_rate": 8.031577140895831e-07,
+      "loss": 0.3707,
+      "step": 11342
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.833656311035156,
+      "learning_rate": 8.005834906469882e-07,
+      "loss": 0.4552,
+      "step": 11343
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.008463859558105,
+      "learning_rate": 7.980092672043934e-07,
+      "loss": 0.3894,
+      "step": 11344
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.926922798156738,
+      "learning_rate": 7.954350437617986e-07,
+      "loss": 0.2971,
+      "step": 11345
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.602967262268066,
+      "learning_rate": 7.928608203192038e-07,
+      "loss": 0.25,
+      "step": 11346
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.169910430908203,
+      "learning_rate": 7.902865968766089e-07,
+      "loss": 0.3526,
+      "step": 11347
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.690129280090332,
+      "learning_rate": 7.877123734340142e-07,
+      "loss": 0.3324,
+      "step": 11348
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.213464736938477,
+      "learning_rate": 7.851381499914193e-07,
+      "loss": 0.2848,
+      "step": 11349
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.605606079101562,
+      "learning_rate": 7.825639265488245e-07,
+      "loss": 0.4787,
+      "step": 11350
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.166177749633789,
+      "learning_rate": 7.799897031062297e-07,
+      "loss": 0.5206,
+      "step": 11351
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.196728706359863,
+      "learning_rate": 7.774154796636347e-07,
+      "loss": 0.2789,
+      "step": 11352
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.934510231018066,
+      "learning_rate": 7.7484125622104e-07,
+      "loss": 0.4525,
+      "step": 11353
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.63411808013916,
+      "learning_rate": 7.722670327784451e-07,
+      "loss": 0.2259,
+      "step": 11354
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.784299850463867,
+      "learning_rate": 7.696928093358504e-07,
+      "loss": 0.28,
+      "step": 11355
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.000056266784668,
+      "learning_rate": 7.671185858932555e-07,
+      "loss": 0.5209,
+      "step": 11356
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.76236629486084,
+      "learning_rate": 7.645443624506607e-07,
+      "loss": 0.3218,
+      "step": 11357
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.138952255249023,
+      "learning_rate": 7.619701390080659e-07,
+      "loss": 0.3785,
+      "step": 11358
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.027183532714844,
+      "learning_rate": 7.593959155654711e-07,
+      "loss": 0.3568,
+      "step": 11359
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.467564582824707,
+      "learning_rate": 7.568216921228762e-07,
+      "loss": 0.5007,
+      "step": 11360
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.798078536987305,
+      "learning_rate": 7.542474686802815e-07,
+      "loss": 0.5194,
+      "step": 11361
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.058542251586914,
+      "learning_rate": 7.516732452376866e-07,
+      "loss": 0.3352,
+      "step": 11362
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.526508331298828,
+      "learning_rate": 7.490990217950918e-07,
+      "loss": 0.4279,
+      "step": 11363
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.043108940124512,
+      "learning_rate": 7.46524798352497e-07,
+      "loss": 0.2351,
+      "step": 11364
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.996644020080566,
+      "learning_rate": 7.439505749099021e-07,
+      "loss": 0.3735,
+      "step": 11365
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.366191864013672,
+      "learning_rate": 7.413763514673074e-07,
+      "loss": 0.4054,
+      "step": 11366
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.20408821105957,
+      "learning_rate": 7.388021280247125e-07,
+      "loss": 0.3294,
+      "step": 11367
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.842229843139648,
+      "learning_rate": 7.362279045821177e-07,
+      "loss": 0.2984,
+      "step": 11368
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.793534278869629,
+      "learning_rate": 7.336536811395229e-07,
+      "loss": 0.255,
+      "step": 11369
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.305765628814697,
+      "learning_rate": 7.310794576969281e-07,
+      "loss": 0.2969,
+      "step": 11370
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.11915397644043,
+      "learning_rate": 7.285052342543333e-07,
+      "loss": 0.3607,
+      "step": 11371
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.45128345489502,
+      "learning_rate": 7.259310108117385e-07,
+      "loss": 0.2797,
+      "step": 11372
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.961358070373535,
+      "learning_rate": 7.233567873691436e-07,
+      "loss": 0.2651,
+      "step": 11373
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.083110809326172,
+      "learning_rate": 7.207825639265489e-07,
+      "loss": 0.4404,
+      "step": 11374
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.668498992919922,
+      "learning_rate": 7.18208340483954e-07,
+      "loss": 0.2143,
+      "step": 11375
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.162538528442383,
+      "learning_rate": 7.156341170413592e-07,
+      "loss": 0.278,
+      "step": 11376
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.628896713256836,
+      "learning_rate": 7.130598935987644e-07,
+      "loss": 0.3406,
+      "step": 11377
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.505866050720215,
+      "learning_rate": 7.104856701561695e-07,
+      "loss": 0.3576,
+      "step": 11378
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.133254051208496,
+      "learning_rate": 7.079114467135748e-07,
+      "loss": 0.3205,
+      "step": 11379
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.557565689086914,
+      "learning_rate": 7.053372232709799e-07,
+      "loss": 0.4144,
+      "step": 11380
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.075443267822266,
+      "learning_rate": 7.027629998283851e-07,
+      "loss": 0.3218,
+      "step": 11381
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.629512786865234,
+      "learning_rate": 7.001887763857903e-07,
+      "loss": 0.3068,
+      "step": 11382
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.45313835144043,
+      "learning_rate": 6.976145529431955e-07,
+      "loss": 0.244,
+      "step": 11383
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.560997009277344,
+      "learning_rate": 6.950403295006007e-07,
+      "loss": 0.3099,
+      "step": 11384
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.60513973236084,
+      "learning_rate": 6.924661060580059e-07,
+      "loss": 0.3831,
+      "step": 11385
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.306890487670898,
+      "learning_rate": 6.89891882615411e-07,
+      "loss": 0.2036,
+      "step": 11386
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.6257123947143555,
+      "learning_rate": 6.873176591728163e-07,
+      "loss": 0.2673,
+      "step": 11387
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.490890502929688,
+      "learning_rate": 6.847434357302214e-07,
+      "loss": 0.3954,
+      "step": 11388
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.047118186950684,
+      "learning_rate": 6.821692122876266e-07,
+      "loss": 0.33,
+      "step": 11389
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.06814956665039,
+      "learning_rate": 6.795949888450318e-07,
+      "loss": 0.3274,
+      "step": 11390
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.393421173095703,
+      "learning_rate": 6.770207654024369e-07,
+      "loss": 0.4495,
+      "step": 11391
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.045222282409668,
+      "learning_rate": 6.744465419598422e-07,
+      "loss": 0.5531,
+      "step": 11392
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.564979553222656,
+      "learning_rate": 6.718723185172473e-07,
+      "loss": 0.3655,
+      "step": 11393
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.417673110961914,
+      "learning_rate": 6.692980950746525e-07,
+      "loss": 0.4051,
+      "step": 11394
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.536788940429688,
+      "learning_rate": 6.667238716320577e-07,
+      "loss": 0.3846,
+      "step": 11395
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.755039691925049,
+      "learning_rate": 6.641496481894629e-07,
+      "loss": 0.2846,
+      "step": 11396
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.713797569274902,
+      "learning_rate": 6.615754247468681e-07,
+      "loss": 0.4076,
+      "step": 11397
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.087739944458008,
+      "learning_rate": 6.590012013042733e-07,
+      "loss": 0.3359,
+      "step": 11398
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.078948020935059,
+      "learning_rate": 6.564269778616784e-07,
+      "loss": 0.3524,
+      "step": 11399
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.877887725830078,
+      "learning_rate": 6.538527544190837e-07,
+      "loss": 0.2715,
+      "step": 11400
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.709405899047852,
+      "learning_rate": 6.512785309764888e-07,
+      "loss": 0.1988,
+      "step": 11401
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.651058197021484,
+      "learning_rate": 6.48704307533894e-07,
+      "loss": 0.2806,
+      "step": 11402
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.915276527404785,
+      "learning_rate": 6.461300840912992e-07,
+      "loss": 0.3012,
+      "step": 11403
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.318134307861328,
+      "learning_rate": 6.435558606487042e-07,
+      "loss": 0.1687,
+      "step": 11404
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.65805721282959,
+      "learning_rate": 6.409816372061096e-07,
+      "loss": 0.4185,
+      "step": 11405
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.805819511413574,
+      "learning_rate": 6.384074137635146e-07,
+      "loss": 0.3746,
+      "step": 11406
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.475388526916504,
+      "learning_rate": 6.358331903209198e-07,
+      "loss": 0.3137,
+      "step": 11407
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.643163681030273,
+      "learning_rate": 6.33258966878325e-07,
+      "loss": 0.2668,
+      "step": 11408
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.919766426086426,
+      "learning_rate": 6.306847434357302e-07,
+      "loss": 0.3774,
+      "step": 11409
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.508401870727539,
+      "learning_rate": 6.281105199931354e-07,
+      "loss": 0.3102,
+      "step": 11410
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.121211051940918,
+      "learning_rate": 6.255362965505406e-07,
+      "loss": 0.4604,
+      "step": 11411
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.517192840576172,
+      "learning_rate": 6.229620731079457e-07,
+      "loss": 0.4134,
+      "step": 11412
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.313142776489258,
+      "learning_rate": 6.20387849665351e-07,
+      "loss": 0.3172,
+      "step": 11413
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.615242004394531,
+      "learning_rate": 6.178136262227561e-07,
+      "loss": 0.3419,
+      "step": 11414
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.35338020324707,
+      "learning_rate": 6.152394027801613e-07,
+      "loss": 0.3038,
+      "step": 11415
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.667108535766602,
+      "learning_rate": 6.126651793375665e-07,
+      "loss": 0.3588,
+      "step": 11416
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.784387588500977,
+      "learning_rate": 6.100909558949716e-07,
+      "loss": 0.307,
+      "step": 11417
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.328282356262207,
+      "learning_rate": 6.075167324523769e-07,
+      "loss": 0.3785,
+      "step": 11418
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.699987411499023,
+      "learning_rate": 6.04942509009782e-07,
+      "loss": 0.3058,
+      "step": 11419
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.649855613708496,
+      "learning_rate": 6.023682855671872e-07,
+      "loss": 0.3671,
+      "step": 11420
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.536017417907715,
+      "learning_rate": 5.997940621245924e-07,
+      "loss": 0.5921,
+      "step": 11421
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.392743110656738,
+      "learning_rate": 5.972198386819976e-07,
+      "loss": 0.4402,
+      "step": 11422
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.151330947875977,
+      "learning_rate": 5.946456152394028e-07,
+      "loss": 0.3639,
+      "step": 11423
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.538474082946777,
+      "learning_rate": 5.92071391796808e-07,
+      "loss": 0.591,
+      "step": 11424
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.823904037475586,
+      "learning_rate": 5.894971683542131e-07,
+      "loss": 0.2764,
+      "step": 11425
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.737130165100098,
+      "learning_rate": 5.869229449116184e-07,
+      "loss": 0.3678,
+      "step": 11426
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.101218223571777,
+      "learning_rate": 5.843487214690235e-07,
+      "loss": 0.5497,
+      "step": 11427
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.815122604370117,
+      "learning_rate": 5.817744980264287e-07,
+      "loss": 0.2557,
+      "step": 11428
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.70182991027832,
+      "learning_rate": 5.792002745838339e-07,
+      "loss": 0.4925,
+      "step": 11429
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.772957801818848,
+      "learning_rate": 5.76626051141239e-07,
+      "loss": 0.3222,
+      "step": 11430
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.422170162200928,
+      "learning_rate": 5.740518276986443e-07,
+      "loss": 0.1953,
+      "step": 11431
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.435391902923584,
+      "learning_rate": 5.714776042560494e-07,
+      "loss": 0.2463,
+      "step": 11432
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.697405815124512,
+      "learning_rate": 5.689033808134546e-07,
+      "loss": 0.3534,
+      "step": 11433
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.7539701461792,
+      "learning_rate": 5.663291573708598e-07,
+      "loss": 0.4687,
+      "step": 11434
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.636242866516113,
+      "learning_rate": 5.63754933928265e-07,
+      "loss": 0.2845,
+      "step": 11435
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.384602546691895,
+      "learning_rate": 5.611807104856702e-07,
+      "loss": 0.2499,
+      "step": 11436
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.37716293334961,
+      "learning_rate": 5.586064870430754e-07,
+      "loss": 0.3624,
+      "step": 11437
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.166197776794434,
+      "learning_rate": 5.560322636004805e-07,
+      "loss": 0.2766,
+      "step": 11438
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.577507019042969,
+      "learning_rate": 5.534580401578858e-07,
+      "loss": 0.4619,
+      "step": 11439
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.988529205322266,
+      "learning_rate": 5.508838167152909e-07,
+      "loss": 0.3986,
+      "step": 11440
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.408734321594238,
+      "learning_rate": 5.483095932726961e-07,
+      "loss": 0.3293,
+      "step": 11441
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.547457695007324,
+      "learning_rate": 5.457353698301013e-07,
+      "loss": 0.4507,
+      "step": 11442
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.997553825378418,
+      "learning_rate": 5.431611463875064e-07,
+      "loss": 0.4103,
+      "step": 11443
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.524253845214844,
+      "learning_rate": 5.405869229449117e-07,
+      "loss": 0.3093,
+      "step": 11444
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.29480504989624,
+      "learning_rate": 5.380126995023168e-07,
+      "loss": 0.2626,
+      "step": 11445
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 15.253386497497559,
+      "learning_rate": 5.35438476059722e-07,
+      "loss": 0.4418,
+      "step": 11446
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.01857852935791,
+      "learning_rate": 5.328642526171272e-07,
+      "loss": 0.3053,
+      "step": 11447
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.180018424987793,
+      "learning_rate": 5.302900291745324e-07,
+      "loss": 0.2964,
+      "step": 11448
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.832536697387695,
+      "learning_rate": 5.277158057319376e-07,
+      "loss": 0.256,
+      "step": 11449
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.6122407913208,
+      "learning_rate": 5.251415822893428e-07,
+      "loss": 0.4412,
+      "step": 11450
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.61803913116455,
+      "learning_rate": 5.225673588467479e-07,
+      "loss": 0.5073,
+      "step": 11451
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.499238967895508,
+      "learning_rate": 5.199931354041532e-07,
+      "loss": 0.4001,
+      "step": 11452
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.861466407775879,
+      "learning_rate": 5.174189119615583e-07,
+      "loss": 0.4356,
+      "step": 11453
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.302648544311523,
+      "learning_rate": 5.148446885189635e-07,
+      "loss": 0.285,
+      "step": 11454
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.945225715637207,
+      "learning_rate": 5.122704650763687e-07,
+      "loss": 0.4054,
+      "step": 11455
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.368644714355469,
+      "learning_rate": 5.096962416337738e-07,
+      "loss": 0.3205,
+      "step": 11456
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.485661506652832,
+      "learning_rate": 5.071220181911791e-07,
+      "loss": 0.2192,
+      "step": 11457
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.474787712097168,
+      "learning_rate": 5.045477947485842e-07,
+      "loss": 0.5874,
+      "step": 11458
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.382576942443848,
+      "learning_rate": 5.019735713059894e-07,
+      "loss": 0.3911,
+      "step": 11459
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.274835586547852,
+      "learning_rate": 4.993993478633946e-07,
+      "loss": 0.3661,
+      "step": 11460
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.426583766937256,
+      "learning_rate": 4.968251244207998e-07,
+      "loss": 0.2625,
+      "step": 11461
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.829323768615723,
+      "learning_rate": 4.942509009782048e-07,
+      "loss": 0.3157,
+      "step": 11462
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.745725631713867,
+      "learning_rate": 4.916766775356101e-07,
+      "loss": 0.3127,
+      "step": 11463
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.34257698059082,
+      "learning_rate": 4.891024540930152e-07,
+      "loss": 0.4637,
+      "step": 11464
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.127771377563477,
+      "learning_rate": 4.865282306504205e-07,
+      "loss": 0.3486,
+      "step": 11465
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.810133934020996,
+      "learning_rate": 4.839540072078256e-07,
+      "loss": 0.3189,
+      "step": 11466
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.137019634246826,
+      "learning_rate": 4.813797837652308e-07,
+      "loss": 0.2494,
+      "step": 11467
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.305639266967773,
+      "learning_rate": 4.78805560322636e-07,
+      "loss": 0.278,
+      "step": 11468
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.243700981140137,
+      "learning_rate": 4.762313368800412e-07,
+      "loss": 0.412,
+      "step": 11469
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.298053741455078,
+      "learning_rate": 4.7365711343744644e-07,
+      "loss": 0.4562,
+      "step": 11470
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.916495323181152,
+      "learning_rate": 4.710828899948516e-07,
+      "loss": 0.2969,
+      "step": 11471
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.318690299987793,
+      "learning_rate": 4.685086665522568e-07,
+      "loss": 0.3448,
+      "step": 11472
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.141204833984375,
+      "learning_rate": 4.6593444310966193e-07,
+      "loss": 0.4195,
+      "step": 11473
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.106829643249512,
+      "learning_rate": 4.6336021966706713e-07,
+      "loss": 0.5072,
+      "step": 11474
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.209422588348389,
+      "learning_rate": 4.607859962244723e-07,
+      "loss": 0.2639,
+      "step": 11475
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.521828651428223,
+      "learning_rate": 4.5821177278187747e-07,
+      "loss": 0.3201,
+      "step": 11476
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.716654777526855,
+      "learning_rate": 4.556375493392826e-07,
+      "loss": 0.424,
+      "step": 11477
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.334619522094727,
+      "learning_rate": 4.530633258966878e-07,
+      "loss": 0.348,
+      "step": 11478
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.950033187866211,
+      "learning_rate": 4.50489102454093e-07,
+      "loss": 0.2928,
+      "step": 11479
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.429141998291016,
+      "learning_rate": 4.479148790114982e-07,
+      "loss": 0.2919,
+      "step": 11480
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.359455108642578,
+      "learning_rate": 4.453406555689034e-07,
+      "loss": 0.3272,
+      "step": 11481
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.862103462219238,
+      "learning_rate": 4.4276643212630856e-07,
+      "loss": 0.3789,
+      "step": 11482
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.794952392578125,
+      "learning_rate": 4.4019220868371376e-07,
+      "loss": 0.5016,
+      "step": 11483
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.989912986755371,
+      "learning_rate": 4.3761798524111896e-07,
+      "loss": 0.2662,
+      "step": 11484
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.99366569519043,
+      "learning_rate": 4.3504376179852416e-07,
+      "loss": 0.4709,
+      "step": 11485
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.567079544067383,
+      "learning_rate": 4.324695383559293e-07,
+      "loss": 0.4098,
+      "step": 11486
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.405452728271484,
+      "learning_rate": 4.298953149133345e-07,
+      "loss": 0.2751,
+      "step": 11487
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.602827072143555,
+      "learning_rate": 4.273210914707397e-07,
+      "loss": 0.4951,
+      "step": 11488
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.505640983581543,
+      "learning_rate": 4.2474686802814485e-07,
+      "loss": 0.318,
+      "step": 11489
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.013648986816406,
+      "learning_rate": 4.2217264458555e-07,
+      "loss": 0.272,
+      "step": 11490
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.087907791137695,
+      "learning_rate": 4.195984211429552e-07,
+      "loss": 0.4187,
+      "step": 11491
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.0941162109375,
+      "learning_rate": 4.170241977003604e-07,
+      "loss": 0.4704,
+      "step": 11492
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.989205360412598,
+      "learning_rate": 4.144499742577656e-07,
+      "loss": 0.302,
+      "step": 11493
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.566054344177246,
+      "learning_rate": 4.1187575081517074e-07,
+      "loss": 0.4751,
+      "step": 11494
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.2603759765625,
+      "learning_rate": 4.0930152737257594e-07,
+      "loss": 0.2806,
+      "step": 11495
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.102025032043457,
+      "learning_rate": 4.0672730392998114e-07,
+      "loss": 0.356,
+      "step": 11496
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.15386962890625,
+      "learning_rate": 4.0415308048738634e-07,
+      "loss": 0.3403,
+      "step": 11497
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.41845703125,
+      "learning_rate": 4.0157885704479154e-07,
+      "loss": 0.3056,
+      "step": 11498
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.357941627502441,
+      "learning_rate": 3.990046336021967e-07,
+      "loss": 0.3618,
+      "step": 11499
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.907411575317383,
+      "learning_rate": 3.964304101596019e-07,
+      "loss": 0.331,
+      "step": 11500
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.477226257324219,
+      "learning_rate": 3.938561867170071e-07,
+      "loss": 0.3956,
+      "step": 11501
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.849809646606445,
+      "learning_rate": 3.912819632744122e-07,
+      "loss": 0.4089,
+      "step": 11502
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.161151885986328,
+      "learning_rate": 3.8870773983181737e-07,
+      "loss": 0.4616,
+      "step": 11503
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.105419158935547,
+      "learning_rate": 3.8613351638922257e-07,
+      "loss": 0.3344,
+      "step": 11504
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.301834106445312,
+      "learning_rate": 3.8355929294662777e-07,
+      "loss": 0.3531,
+      "step": 11505
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.82765007019043,
+      "learning_rate": 3.8098506950403297e-07,
+      "loss": 0.61,
+      "step": 11506
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.907184600830078,
+      "learning_rate": 3.784108460614381e-07,
+      "loss": 0.2997,
+      "step": 11507
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.38486385345459,
+      "learning_rate": 3.758366226188433e-07,
+      "loss": 0.3336,
+      "step": 11508
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.507502555847168,
+      "learning_rate": 3.732623991762485e-07,
+      "loss": 0.3728,
+      "step": 11509
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.295052528381348,
+      "learning_rate": 3.706881757336537e-07,
+      "loss": 0.3916,
+      "step": 11510
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.334773063659668,
+      "learning_rate": 3.6811395229105886e-07,
+      "loss": 0.3485,
+      "step": 11511
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.45616626739502,
+      "learning_rate": 3.6553972884846406e-07,
+      "loss": 0.3185,
+      "step": 11512
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.822600841522217,
+      "learning_rate": 3.6296550540586926e-07,
+      "loss": 0.3291,
+      "step": 11513
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.113861083984375,
+      "learning_rate": 3.6039128196327446e-07,
+      "loss": 0.389,
+      "step": 11514
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.824539184570312,
+      "learning_rate": 3.578170585206796e-07,
+      "loss": 0.2315,
+      "step": 11515
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.730411529541016,
+      "learning_rate": 3.5524283507808475e-07,
+      "loss": 0.2284,
+      "step": 11516
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.9005126953125,
+      "learning_rate": 3.5266861163548995e-07,
+      "loss": 0.4206,
+      "step": 11517
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.357040405273438,
+      "learning_rate": 3.5009438819289515e-07,
+      "loss": 0.4062,
+      "step": 11518
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.8555326461792,
+      "learning_rate": 3.4752016475030035e-07,
+      "loss": 0.4162,
+      "step": 11519
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.863471031188965,
+      "learning_rate": 3.449459413077055e-07,
+      "loss": 0.3734,
+      "step": 11520
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.063223838806152,
+      "learning_rate": 3.423717178651107e-07,
+      "loss": 0.364,
+      "step": 11521
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.022010803222656,
+      "learning_rate": 3.397974944225159e-07,
+      "loss": 0.5075,
+      "step": 11522
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.83696460723877,
+      "learning_rate": 3.372232709799211e-07,
+      "loss": 0.2929,
+      "step": 11523
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.155284881591797,
+      "learning_rate": 3.3464904753732624e-07,
+      "loss": 0.3201,
+      "step": 11524
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.256797790527344,
+      "learning_rate": 3.3207482409473143e-07,
+      "loss": 0.2956,
+      "step": 11525
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.252130508422852,
+      "learning_rate": 3.2950060065213663e-07,
+      "loss": 0.3463,
+      "step": 11526
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.993163108825684,
+      "learning_rate": 3.2692637720954183e-07,
+      "loss": 0.3171,
+      "step": 11527
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.930692672729492,
+      "learning_rate": 3.24352153766947e-07,
+      "loss": 0.3235,
+      "step": 11528
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.87640380859375,
+      "learning_rate": 3.217779303243521e-07,
+      "loss": 0.3065,
+      "step": 11529
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.991257667541504,
+      "learning_rate": 3.192037068817573e-07,
+      "loss": 0.3639,
+      "step": 11530
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.389147758483887,
+      "learning_rate": 3.166294834391625e-07,
+      "loss": 0.4295,
+      "step": 11531
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.487982749938965,
+      "learning_rate": 3.140552599965677e-07,
+      "loss": 0.327,
+      "step": 11532
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.464696884155273,
+      "learning_rate": 3.1148103655397287e-07,
+      "loss": 0.32,
+      "step": 11533
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.410228729248047,
+      "learning_rate": 3.0890681311137807e-07,
+      "loss": 0.3767,
+      "step": 11534
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.981040954589844,
+      "learning_rate": 3.0633258966878327e-07,
+      "loss": 0.3519,
+      "step": 11535
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.440129280090332,
+      "learning_rate": 3.0375836622618847e-07,
+      "loss": 0.3847,
+      "step": 11536
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.9198637008667,
+      "learning_rate": 3.011841427835936e-07,
+      "loss": 0.4402,
+      "step": 11537
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.214505195617676,
+      "learning_rate": 2.986099193409988e-07,
+      "loss": 0.2856,
+      "step": 11538
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.527960777282715,
+      "learning_rate": 2.96035695898404e-07,
+      "loss": 0.3397,
+      "step": 11539
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.586348533630371,
+      "learning_rate": 2.934614724558092e-07,
+      "loss": 0.2686,
+      "step": 11540
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.140045166015625,
+      "learning_rate": 2.9088724901321436e-07,
+      "loss": 0.3593,
+      "step": 11541
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.366416931152344,
+      "learning_rate": 2.883130255706195e-07,
+      "loss": 0.3621,
+      "step": 11542
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.255388259887695,
+      "learning_rate": 2.857388021280247e-07,
+      "loss": 0.2811,
+      "step": 11543
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.306804656982422,
+      "learning_rate": 2.831645786854299e-07,
+      "loss": 0.3178,
+      "step": 11544
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.351367950439453,
+      "learning_rate": 2.805903552428351e-07,
+      "loss": 0.2738,
+      "step": 11545
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.161386966705322,
+      "learning_rate": 2.7801613180024024e-07,
+      "loss": 0.2006,
+      "step": 11546
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.230098724365234,
+      "learning_rate": 2.7544190835764544e-07,
+      "loss": 0.3282,
+      "step": 11547
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.078783988952637,
+      "learning_rate": 2.7286768491505064e-07,
+      "loss": 0.4008,
+      "step": 11548
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.660420894622803,
+      "learning_rate": 2.7029346147245584e-07,
+      "loss": 0.2171,
+      "step": 11549
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.891387939453125,
+      "learning_rate": 2.67719238029861e-07,
+      "loss": 0.3924,
+      "step": 11550
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.31053638458252,
+      "learning_rate": 2.651450145872662e-07,
+      "loss": 0.5193,
+      "step": 11551
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.492470741271973,
+      "learning_rate": 2.625707911446714e-07,
+      "loss": 0.3821,
+      "step": 11552
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.105793952941895,
+      "learning_rate": 2.599965677020766e-07,
+      "loss": 0.3579,
+      "step": 11553
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.469000816345215,
+      "learning_rate": 2.5742234425948173e-07,
+      "loss": 0.416,
+      "step": 11554
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.040057182312012,
+      "learning_rate": 2.548481208168869e-07,
+      "loss": 0.3274,
+      "step": 11555
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.669187545776367,
+      "learning_rate": 2.522738973742921e-07,
+      "loss": 0.4086,
+      "step": 11556
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.471062660217285,
+      "learning_rate": 2.496996739316973e-07,
+      "loss": 0.3147,
+      "step": 11557
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.899096488952637,
+      "learning_rate": 2.471254504891024e-07,
+      "loss": 0.4109,
+      "step": 11558
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.59607982635498,
+      "learning_rate": 2.445512270465076e-07,
+      "loss": 0.3408,
+      "step": 11559
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.245586395263672,
+      "learning_rate": 2.419770036039128e-07,
+      "loss": 0.5837,
+      "step": 11560
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.268532752990723,
+      "learning_rate": 2.39402780161318e-07,
+      "loss": 0.2982,
+      "step": 11561
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.093299865722656,
+      "learning_rate": 2.3682855671872322e-07,
+      "loss": 0.4422,
+      "step": 11562
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.338587760925293,
+      "learning_rate": 2.342543332761284e-07,
+      "loss": 0.3539,
+      "step": 11563
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.079157829284668,
+      "learning_rate": 2.3168010983353356e-07,
+      "loss": 0.3752,
+      "step": 11564
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.305372714996338,
+      "learning_rate": 2.2910588639093874e-07,
+      "loss": 0.3345,
+      "step": 11565
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.811758041381836,
+      "learning_rate": 2.265316629483439e-07,
+      "loss": 0.2837,
+      "step": 11566
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.363420486450195,
+      "learning_rate": 2.239574395057491e-07,
+      "loss": 0.2655,
+      "step": 11567
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.844328880310059,
+      "learning_rate": 2.2138321606315428e-07,
+      "loss": 0.3373,
+      "step": 11568
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.258129119873047,
+      "learning_rate": 2.1880899262055948e-07,
+      "loss": 0.3602,
+      "step": 11569
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.300853729248047,
+      "learning_rate": 2.1623476917796465e-07,
+      "loss": 0.3643,
+      "step": 11570
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.862691402435303,
+      "learning_rate": 2.1366054573536985e-07,
+      "loss": 0.2271,
+      "step": 11571
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.394877433776855,
+      "learning_rate": 2.11086322292775e-07,
+      "loss": 0.3863,
+      "step": 11572
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.647296905517578,
+      "learning_rate": 2.085120988501802e-07,
+      "loss": 0.3639,
+      "step": 11573
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.53976058959961,
+      "learning_rate": 2.0593787540758537e-07,
+      "loss": 0.3572,
+      "step": 11574
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.477564334869385,
+      "learning_rate": 2.0336365196499057e-07,
+      "loss": 0.1911,
+      "step": 11575
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.660491943359375,
+      "learning_rate": 2.0078942852239577e-07,
+      "loss": 0.2693,
+      "step": 11576
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.215484619140625,
+      "learning_rate": 1.9821520507980094e-07,
+      "loss": 0.3393,
+      "step": 11577
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.43334674835205,
+      "learning_rate": 1.956409816372061e-07,
+      "loss": 0.3334,
+      "step": 11578
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.9304780960083,
+      "learning_rate": 1.9306675819461129e-07,
+      "loss": 0.4064,
+      "step": 11579
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.014012336730957,
+      "learning_rate": 1.9049253475201648e-07,
+      "loss": 0.317,
+      "step": 11580
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.0802583694458,
+      "learning_rate": 1.8791831130942166e-07,
+      "loss": 0.3601,
+      "step": 11581
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.10100269317627,
+      "learning_rate": 1.8534408786682686e-07,
+      "loss": 0.2967,
+      "step": 11582
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.166520118713379,
+      "learning_rate": 1.8276986442423203e-07,
+      "loss": 0.1733,
+      "step": 11583
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.802959442138672,
+      "learning_rate": 1.8019564098163723e-07,
+      "loss": 0.2037,
+      "step": 11584
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.004393577575684,
+      "learning_rate": 1.7762141753904237e-07,
+      "loss": 0.3187,
+      "step": 11585
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.01358699798584,
+      "learning_rate": 1.7504719409644757e-07,
+      "loss": 0.3986,
+      "step": 11586
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.093853950500488,
+      "learning_rate": 1.7247297065385275e-07,
+      "loss": 0.5979,
+      "step": 11587
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.381300926208496,
+      "learning_rate": 1.6989874721125795e-07,
+      "loss": 0.4252,
+      "step": 11588
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 41.779701232910156,
+      "learning_rate": 1.6732452376866312e-07,
+      "loss": 0.4688,
+      "step": 11589
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.901070594787598,
+      "learning_rate": 1.6475030032606832e-07,
+      "loss": 0.4482,
+      "step": 11590
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.64988899230957,
+      "learning_rate": 1.621760768834735e-07,
+      "loss": 0.3563,
+      "step": 11591
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.116276741027832,
+      "learning_rate": 1.5960185344087866e-07,
+      "loss": 0.3584,
+      "step": 11592
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.5116605758667,
+      "learning_rate": 1.5702762999828386e-07,
+      "loss": 0.4026,
+      "step": 11593
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.244938850402832,
+      "learning_rate": 1.5445340655568903e-07,
+      "loss": 0.4058,
+      "step": 11594
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.948546409606934,
+      "learning_rate": 1.5187918311309423e-07,
+      "loss": 0.4042,
+      "step": 11595
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.164525032043457,
+      "learning_rate": 1.493049596704994e-07,
+      "loss": 0.3639,
+      "step": 11596
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.158011436462402,
+      "learning_rate": 1.467307362279046e-07,
+      "loss": 0.3336,
+      "step": 11597
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.70641803741455,
+      "learning_rate": 1.4415651278530975e-07,
+      "loss": 0.4431,
+      "step": 11598
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.583722114562988,
+      "learning_rate": 1.4158228934271495e-07,
+      "loss": 0.2764,
+      "step": 11599
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.039341926574707,
+      "learning_rate": 1.3900806590012012e-07,
+      "loss": 0.2335,
+      "step": 11600
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.198591232299805,
+      "learning_rate": 1.3643384245752532e-07,
+      "loss": 0.3343,
+      "step": 11601
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.904633522033691,
+      "learning_rate": 1.338596190149305e-07,
+      "loss": 0.2966,
+      "step": 11602
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.212183952331543,
+      "learning_rate": 1.312853955723357e-07,
+      "loss": 0.3424,
+      "step": 11603
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.84640884399414,
+      "learning_rate": 1.2871117212974087e-07,
+      "loss": 0.4,
+      "step": 11604
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.661840915679932,
+      "learning_rate": 1.2613694868714604e-07,
+      "loss": 0.2402,
+      "step": 11605
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.088878631591797,
+      "learning_rate": 1.235627252445512e-07,
+      "loss": 0.2863,
+      "step": 11606
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.745158195495605,
+      "learning_rate": 1.209885018019564e-07,
+      "loss": 0.2985,
+      "step": 11607
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.851995468139648,
+      "learning_rate": 1.1841427835936161e-07,
+      "loss": 0.2763,
+      "step": 11608
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.716765403747559,
+      "learning_rate": 1.1584005491676678e-07,
+      "loss": 0.253,
+      "step": 11609
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.23376750946045,
+      "learning_rate": 1.1326583147417195e-07,
+      "loss": 0.4726,
+      "step": 11610
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.617305755615234,
+      "learning_rate": 1.1069160803157714e-07,
+      "loss": 0.389,
+      "step": 11611
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.755290985107422,
+      "learning_rate": 1.0811738458898233e-07,
+      "loss": 0.3865,
+      "step": 11612
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.882112503051758,
+      "learning_rate": 1.055431611463875e-07,
+      "loss": 0.4901,
+      "step": 11613
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.497066974639893,
+      "learning_rate": 1.0296893770379268e-07,
+      "loss": 0.1278,
+      "step": 11614
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.85369873046875,
+      "learning_rate": 1.0039471426119788e-07,
+      "loss": 0.322,
+      "step": 11615
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.569109916687012,
+      "learning_rate": 9.782049081860306e-08,
+      "loss": 0.4142,
+      "step": 11616
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.69202184677124,
+      "learning_rate": 9.524626737600824e-08,
+      "loss": 0.305,
+      "step": 11617
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.631950378417969,
+      "learning_rate": 9.267204393341343e-08,
+      "loss": 0.3786,
+      "step": 11618
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.579807758331299,
+      "learning_rate": 9.009782049081861e-08,
+      "loss": 0.2518,
+      "step": 11619
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.297823905944824,
+      "learning_rate": 8.752359704822379e-08,
+      "loss": 0.327,
+      "step": 11620
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.749382972717285,
+      "learning_rate": 8.494937360562897e-08,
+      "loss": 0.4666,
+      "step": 11621
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.204568862915039,
+      "learning_rate": 8.237515016303416e-08,
+      "loss": 0.3926,
+      "step": 11622
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.774746894836426,
+      "learning_rate": 7.980092672043933e-08,
+      "loss": 0.2924,
+      "step": 11623
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.393989086151123,
+      "learning_rate": 7.722670327784452e-08,
+      "loss": 0.1916,
+      "step": 11624
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.441868782043457,
+      "learning_rate": 7.46524798352497e-08,
+      "loss": 0.4196,
+      "step": 11625
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.14520263671875,
+      "learning_rate": 7.207825639265488e-08,
+      "loss": 0.3645,
+      "step": 11626
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.262258529663086,
+      "learning_rate": 6.950403295006006e-08,
+      "loss": 0.3976,
+      "step": 11627
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.72003173828125,
+      "learning_rate": 6.692980950746525e-08,
+      "loss": 0.2126,
+      "step": 11628
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.395048141479492,
+      "learning_rate": 6.435558606487043e-08,
+      "loss": 0.3385,
+      "step": 11629
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.435748100280762,
+      "learning_rate": 6.17813626222756e-08,
+      "loss": 0.3397,
+      "step": 11630
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.424450874328613,
+      "learning_rate": 5.9207139179680805e-08,
+      "loss": 0.3764,
+      "step": 11631
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.107622146606445,
+      "learning_rate": 5.663291573708598e-08,
+      "loss": 0.434,
+      "step": 11632
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.096739768981934,
+      "learning_rate": 5.405869229449116e-08,
+      "loss": 0.4002,
+      "step": 11633
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.735634803771973,
+      "learning_rate": 5.148446885189634e-08,
+      "loss": 0.4333,
+      "step": 11634
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.462284088134766,
+      "learning_rate": 4.891024540930153e-08,
+      "loss": 0.3572,
+      "step": 11635
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.17946720123291,
+      "learning_rate": 4.6336021966706714e-08,
+      "loss": 0.3477,
+      "step": 11636
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.753332138061523,
+      "learning_rate": 4.3761798524111893e-08,
+      "loss": 0.3611,
+      "step": 11637
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.240086555480957,
+      "learning_rate": 4.118757508151708e-08,
+      "loss": 0.3462,
+      "step": 11638
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.988096237182617,
+      "learning_rate": 3.861335163892226e-08,
+      "loss": 0.4948,
+      "step": 11639
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.858365058898926,
+      "learning_rate": 3.603912819632744e-08,
+      "loss": 0.2489,
+      "step": 11640
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.523008346557617,
+      "learning_rate": 3.3464904753732624e-08,
+      "loss": 0.3755,
+      "step": 11641
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.568328857421875,
+      "learning_rate": 3.08906813111378e-08,
+      "loss": 0.3955,
+      "step": 11642
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.07970905303955,
+      "learning_rate": 2.831645786854299e-08,
+      "loss": 0.3319,
+      "step": 11643
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.71036434173584,
+      "learning_rate": 2.574223442594817e-08,
+      "loss": 0.3837,
+      "step": 11644
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.206975936889648,
+      "learning_rate": 2.3168010983353357e-08,
+      "loss": 0.4032,
+      "step": 11645
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.85912036895752,
+      "learning_rate": 2.059378754075854e-08,
+      "loss": 0.3112,
+      "step": 11646
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.981271743774414,
+      "learning_rate": 1.801956409816372e-08,
+      "loss": 0.2749,
+      "step": 11647
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.14116382598877,
+      "learning_rate": 1.54453406555689e-08,
+      "loss": 0.4656,
+      "step": 11648
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.546759605407715,
+      "learning_rate": 1.2871117212974086e-08,
+      "loss": 0.296,
+      "step": 11649
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.54394817352295,
+      "learning_rate": 1.029689377037927e-08,
+      "loss": 0.4372,
+      "step": 11650
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.853440761566162,
+      "learning_rate": 7.72267032778445e-09,
+      "loss": 0.2876,
+      "step": 11651
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.25448989868164,
+      "learning_rate": 5.148446885189635e-09,
+      "loss": 0.5023,
+      "step": 11652
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.559902667999268,
+      "learning_rate": 2.5742234425948174e-09,
+      "loss": 0.2573,
+      "step": 11653
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.35457992553711,
+      "learning_rate": 0.0,
+      "loss": 0.5802,
+      "step": 11654
+    },
+    {
+      "epoch": 2.0,
+      "step": 11654,
+      "total_flos": 2.1922863214339584e+17,
+      "train_loss": 0.7619943554932388,
+      "train_runtime": 3561.7574,
+      "train_samples_per_second": 314.078,
+      "train_steps_per_second": 3.272
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11654,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 2.1922863214339584e+17,
+  "train_batch_size": 96,
+  "trial_name": null,
+  "trial_params": null
+}