diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 53.34782608695652,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.013377926421404682,
+      "grad_norm": 0.5197079181671143,
+      "learning_rate": 0.0,
+      "loss": 4.2636,
+      "step": 1
+    },
+    {
+      "epoch": 0.026755852842809364,
+      "grad_norm": 0.5626901984214783,
+      "learning_rate": 4e-05,
+      "loss": 4.3971,
+      "step": 2
+    },
+    {
+      "epoch": 0.04013377926421405,
+      "grad_norm": 0.5167903304100037,
+      "learning_rate": 8e-05,
+      "loss": 4.3249,
+      "step": 3
+    },
+    {
+      "epoch": 0.05351170568561873,
+      "grad_norm": 0.4764951169490814,
+      "learning_rate": 0.00012,
+      "loss": 4.2031,
+      "step": 4
+    },
+    {
+      "epoch": 0.06688963210702341,
+      "grad_norm": 0.45488491654396057,
+      "learning_rate": 0.00016,
+      "loss": 4.3914,
+      "step": 5
+    },
+    {
+      "epoch": 0.0802675585284281,
+      "grad_norm": 0.568274736404419,
+      "learning_rate": 0.0002,
+      "loss": 4.2346,
+      "step": 6
+    },
+    {
+      "epoch": 0.09364548494983277,
+      "grad_norm": 0.5974003076553345,
+      "learning_rate": 0.0001999555061179088,
+      "loss": 4.131,
+      "step": 7
+    },
+    {
+      "epoch": 0.10702341137123746,
+      "grad_norm": 0.6204471588134766,
+      "learning_rate": 0.00019991101223581757,
+      "loss": 4.2256,
+      "step": 8
+    },
+    {
+      "epoch": 0.12040133779264214,
+      "grad_norm": 0.7143808603286743,
+      "learning_rate": 0.00019986651835372636,
+      "loss": 3.8449,
+      "step": 9
+    },
+    {
+      "epoch": 0.13377926421404682,
+      "grad_norm": 0.7799420356750488,
+      "learning_rate": 0.00019982202447163517,
+      "loss": 4.4301,
+      "step": 10
+    },
+    {
+      "epoch": 0.14715719063545152,
+      "grad_norm": 0.8880407214164734,
+      "learning_rate": 0.00019977753058954395,
+      "loss": 4.2266,
+      "step": 11
+    },
+    {
+      "epoch": 0.1605351170568562,
+      "grad_norm": 0.7776209712028503,
+      "learning_rate": 0.00019973303670745273,
+      "loss": 4.3208,
+      "step": 12
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.9125858545303345,
+      "learning_rate": 0.0001996885428253615,
+      "loss": 4.4363,
+      "step": 13
+    },
+    {
+      "epoch": 0.18729096989966554,
+      "grad_norm": 0.9000256657600403,
+      "learning_rate": 0.00019964404894327032,
+      "loss": 4.2917,
+      "step": 14
+    },
+    {
+      "epoch": 0.20066889632107024,
+      "grad_norm": 0.9995108246803284,
+      "learning_rate": 0.00019959955506117908,
+      "loss": 4.1784,
+      "step": 15
+    },
+    {
+      "epoch": 0.2140468227424749,
+      "grad_norm": 0.9209024310112,
+      "learning_rate": 0.0001995550611790879,
+      "loss": 4.7852,
+      "step": 16
+    },
+    {
+      "epoch": 0.22742474916387959,
+      "grad_norm": 0.9421981573104858,
+      "learning_rate": 0.00019951056729699667,
+      "loss": 4.8501,
+      "step": 17
+    },
+    {
+      "epoch": 0.2408026755852843,
+      "grad_norm": 0.9213201403617859,
+      "learning_rate": 0.00019946607341490545,
+      "loss": 4.7923,
+      "step": 18
+    },
+    {
+      "epoch": 0.25418060200668896,
+      "grad_norm": 0.9378194212913513,
+      "learning_rate": 0.00019942157953281423,
+      "loss": 4.9593,
+      "step": 19
+    },
+    {
+      "epoch": 0.26755852842809363,
+      "grad_norm": 1.0096492767333984,
+      "learning_rate": 0.00019937708565072304,
+      "loss": 4.7099,
+      "step": 20
+    },
+    {
+      "epoch": 0.2809364548494983,
+      "grad_norm": 0.8903587460517883,
+      "learning_rate": 0.00019933259176863183,
+      "loss": 4.3746,
+      "step": 21
+    },
+    {
+      "epoch": 0.29431438127090304,
+      "grad_norm": 0.7808490991592407,
+      "learning_rate": 0.0001992880978865406,
+      "loss": 4.5873,
+      "step": 22
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.8145670294761658,
+      "learning_rate": 0.0001992436040044494,
+      "loss": 4.7924,
+      "step": 23
+    },
+    {
+      "epoch": 0.3210702341137124,
+      "grad_norm": 0.7945849299430847,
+      "learning_rate": 0.0001991991101223582,
+      "loss": 4.8881,
+      "step": 24
+    },
+    {
+      "epoch": 0.33444816053511706,
+      "grad_norm": 0.7871395349502563,
+      "learning_rate": 0.00019915461624026696,
+      "loss": 4.6922,
+      "step": 25
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.9111238718032837,
+      "learning_rate": 0.00019911012235817577,
+      "loss": 4.9982,
+      "step": 26
+    },
+    {
+      "epoch": 0.3612040133779264,
+      "grad_norm": 0.7121369242668152,
+      "learning_rate": 0.00019906562847608455,
+      "loss": 4.4756,
+      "step": 27
+    },
+    {
+      "epoch": 0.3745819397993311,
+      "grad_norm": 0.7118422389030457,
+      "learning_rate": 0.00019902113459399333,
+      "loss": 5.1389,
+      "step": 28
+    },
+    {
+      "epoch": 0.3879598662207358,
+      "grad_norm": 0.7100292444229126,
+      "learning_rate": 0.0001989766407119021,
+      "loss": 4.7691,
+      "step": 29
+    },
+    {
+      "epoch": 0.4013377926421405,
+      "grad_norm": 0.708591639995575,
+      "learning_rate": 0.00019893214682981092,
+      "loss": 4.8721,
+      "step": 30
+    },
+    {
+      "epoch": 0.41471571906354515,
+      "grad_norm": 0.6711616516113281,
+      "learning_rate": 0.0001988876529477197,
+      "loss": 4.9152,
+      "step": 31
+    },
+    {
+      "epoch": 0.4280936454849498,
+      "grad_norm": 0.7158232927322388,
+      "learning_rate": 0.0001988431590656285,
+      "loss": 4.828,
+      "step": 32
+    },
+    {
+      "epoch": 0.4414715719063545,
+      "grad_norm": 0.6246087551116943,
+      "learning_rate": 0.00019879866518353727,
+      "loss": 4.8452,
+      "step": 33
+    },
+    {
+      "epoch": 0.45484949832775917,
+      "grad_norm": 0.6088873744010925,
+      "learning_rate": 0.00019875417130144608,
+      "loss": 4.9702,
+      "step": 34
+    },
+    {
+      "epoch": 0.4682274247491639,
+      "grad_norm": 0.5798126459121704,
+      "learning_rate": 0.00019870967741935483,
+      "loss": 4.9838,
+      "step": 35
+    },
+    {
+      "epoch": 0.4816053511705686,
+      "grad_norm": 0.6268919706344604,
+      "learning_rate": 0.00019866518353726364,
+      "loss": 4.7636,
+      "step": 36
+    },
+    {
+      "epoch": 0.49498327759197325,
+      "grad_norm": 0.5649904012680054,
+      "learning_rate": 0.00019862068965517243,
+      "loss": 4.506,
+      "step": 37
+    },
+    {
+      "epoch": 0.5083612040133779,
+      "grad_norm": 0.5947792530059814,
+      "learning_rate": 0.0001985761957730812,
+      "loss": 4.8057,
+      "step": 38
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.6204257011413574,
+      "learning_rate": 0.00019853170189099,
+      "loss": 5.0511,
+      "step": 39
+    },
+    {
+      "epoch": 0.5351170568561873,
+      "grad_norm": 0.5972265601158142,
+      "learning_rate": 0.0001984872080088988,
+      "loss": 4.924,
+      "step": 40
+    },
+    {
+      "epoch": 0.5484949832775919,
+      "grad_norm": 0.6117077469825745,
+      "learning_rate": 0.00019844271412680758,
+      "loss": 4.8729,
+      "step": 41
+    },
+    {
+      "epoch": 0.5618729096989966,
+      "grad_norm": 0.5085508823394775,
+      "learning_rate": 0.00019839822024471637,
+      "loss": 4.3616,
+      "step": 42
+    },
+    {
+      "epoch": 0.5752508361204013,
+      "grad_norm": 0.550647497177124,
+      "learning_rate": 0.00019835372636262515,
+      "loss": 5.2512,
+      "step": 43
+    },
+    {
+      "epoch": 0.5886287625418061,
+      "grad_norm": 0.48329588770866394,
+      "learning_rate": 0.00019830923248053396,
+      "loss": 4.9501,
+      "step": 44
+    },
+    {
+      "epoch": 0.6020066889632107,
+      "grad_norm": 0.6313246488571167,
+      "learning_rate": 0.0001982647385984427,
+      "loss": 4.5767,
+      "step": 45
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5111928582191467,
+      "learning_rate": 0.00019822024471635152,
+      "loss": 4.5586,
+      "step": 46
+    },
+    {
+      "epoch": 0.6287625418060201,
+      "grad_norm": 0.5264492630958557,
+      "learning_rate": 0.0001981757508342603,
+      "loss": 4.7033,
+      "step": 47
+    },
+    {
+      "epoch": 0.6421404682274248,
+      "grad_norm": 0.5058289170265198,
+      "learning_rate": 0.0001981312569521691,
+      "loss": 4.8396,
+      "step": 48
+    },
+    {
+      "epoch": 0.6555183946488294,
+      "grad_norm": 0.5688439607620239,
+      "learning_rate": 0.00019808676307007787,
+      "loss": 5.1887,
+      "step": 49
+    },
+    {
+      "epoch": 0.6688963210702341,
+      "grad_norm": 0.5488842129707336,
+      "learning_rate": 0.00019804226918798665,
+      "loss": 4.6075,
+      "step": 50
+    },
+    {
+      "epoch": 0.6822742474916388,
+      "grad_norm": 0.5358632206916809,
+      "learning_rate": 0.00019799777530589546,
+      "loss": 5.0205,
+      "step": 51
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.47869494557380676,
+      "learning_rate": 0.00019795328142380422,
+      "loss": 4.8495,
+      "step": 52
+    },
+    {
+      "epoch": 0.7090301003344481,
+      "grad_norm": 0.49378660321235657,
+      "learning_rate": 0.00019790878754171303,
+      "loss": 4.6563,
+      "step": 53
+    },
+    {
+      "epoch": 0.7224080267558528,
+      "grad_norm": 0.5167868733406067,
+      "learning_rate": 0.0001978642936596218,
+      "loss": 5.2558,
+      "step": 54
+    },
+    {
+      "epoch": 0.7357859531772575,
+      "grad_norm": 0.5230040550231934,
+      "learning_rate": 0.0001978197997775306,
+      "loss": 4.7769,
+      "step": 55
+    },
+    {
+      "epoch": 0.7491638795986622,
+      "grad_norm": 0.4822310507297516,
+      "learning_rate": 0.00019777530589543937,
+      "loss": 4.9282,
+      "step": 56
+    },
+    {
+      "epoch": 0.7625418060200669,
+      "grad_norm": 0.500045895576477,
+      "learning_rate": 0.00019773081201334818,
+      "loss": 5.0399,
+      "step": 57
+    },
+    {
+      "epoch": 0.7759197324414716,
+      "grad_norm": 0.4740642309188843,
+      "learning_rate": 0.00019768631813125696,
+      "loss": 4.8041,
+      "step": 58
+    },
+    {
+      "epoch": 0.7892976588628763,
+      "grad_norm": 0.45918184518814087,
+      "learning_rate": 0.00019764182424916575,
+      "loss": 4.6304,
+      "step": 59
+    },
+    {
+      "epoch": 0.802675585284281,
+      "grad_norm": 0.53122878074646,
+      "learning_rate": 0.00019759733036707453,
+      "loss": 4.8377,
+      "step": 60
+    },
+    {
+      "epoch": 0.8160535117056856,
+      "grad_norm": 0.4925791919231415,
+      "learning_rate": 0.00019755283648498334,
+      "loss": 5.0919,
+      "step": 61
+    },
+    {
+      "epoch": 0.8294314381270903,
+      "grad_norm": 0.4777262806892395,
+      "learning_rate": 0.0001975083426028921,
+      "loss": 4.8379,
+      "step": 62
+    },
+    {
+      "epoch": 0.842809364548495,
+      "grad_norm": 0.49119675159454346,
+      "learning_rate": 0.0001974638487208009,
+      "loss": 5.0819,
+      "step": 63
+    },
+    {
+      "epoch": 0.8561872909698997,
+      "grad_norm": 0.4732685089111328,
+      "learning_rate": 0.00019741935483870969,
+      "loss": 4.8948,
+      "step": 64
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.46269145607948303,
+      "learning_rate": 0.00019737486095661847,
+      "loss": 4.824,
+      "step": 65
+    },
+    {
+      "epoch": 0.882943143812709,
+      "grad_norm": 0.49532708525657654,
+      "learning_rate": 0.00019733036707452725,
+      "loss": 4.8986,
+      "step": 66
+    },
+    {
+      "epoch": 0.8963210702341137,
+      "grad_norm": 0.5253002643585205,
+      "learning_rate": 0.00019728587319243606,
+      "loss": 4.9073,
+      "step": 67
+    },
+    {
+      "epoch": 0.9096989966555183,
+      "grad_norm": 0.5069419145584106,
+      "learning_rate": 0.00019724137931034484,
+      "loss": 4.8962,
+      "step": 68
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.5038817524909973,
+      "learning_rate": 0.00019719688542825363,
+      "loss": 4.8711,
+      "step": 69
+    },
+    {
+      "epoch": 0.9364548494983278,
+      "grad_norm": 0.4987100064754486,
+      "learning_rate": 0.0001971523915461624,
+      "loss": 4.8816,
+      "step": 70
+    },
+    {
+      "epoch": 0.9498327759197325,
+      "grad_norm": 0.47370976209640503,
+      "learning_rate": 0.00019710789766407122,
+      "loss": 4.9675,
+      "step": 71
+    },
+    {
+      "epoch": 0.9632107023411371,
+      "grad_norm": 0.5081727504730225,
+      "learning_rate": 0.00019706340378197997,
+      "loss": 4.2768,
+      "step": 72
+    },
+    {
+      "epoch": 0.9765886287625418,
+      "grad_norm": 0.45571258664131165,
+      "learning_rate": 0.00019701890989988878,
+      "loss": 4.6182,
+      "step": 73
+    },
+    {
+      "epoch": 0.9899665551839465,
+      "grad_norm": 0.5216127634048462,
+      "learning_rate": 0.00019697441601779756,
+      "loss": 4.7126,
+      "step": 74
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5393329858779907,
+      "learning_rate": 0.00019692992213570635,
+      "loss": 4.4919,
+      "step": 75
+    },
+    {
+      "epoch": 1.0133779264214047,
+      "grad_norm": 0.4506986737251282,
+      "learning_rate": 0.00019688542825361513,
+      "loss": 4.5089,
+      "step": 76
+    },
+    {
+      "epoch": 1.0267558528428093,
+      "grad_norm": 0.4328899085521698,
+      "learning_rate": 0.00019684093437152394,
+      "loss": 4.7518,
+      "step": 77
+    },
+    {
+      "epoch": 1.040133779264214,
+      "grad_norm": 0.4397362470626831,
+      "learning_rate": 0.00019679644048943272,
+      "loss": 4.5069,
+      "step": 78
+    },
+    {
+      "epoch": 1.0535117056856187,
+      "grad_norm": 0.4604664146900177,
+      "learning_rate": 0.0001967519466073415,
+      "loss": 4.7054,
+      "step": 79
+    },
+    {
+      "epoch": 1.0668896321070234,
+      "grad_norm": 0.4398234784603119,
+      "learning_rate": 0.00019670745272525029,
+      "loss": 4.2743,
+      "step": 80
+    },
+    {
+      "epoch": 1.080267558528428,
+      "grad_norm": 0.4570735692977905,
+      "learning_rate": 0.0001966629588431591,
+      "loss": 4.8012,
+      "step": 81
+    },
+    {
+      "epoch": 1.0936454849498327,
+      "grad_norm": 0.4814144968986511,
+      "learning_rate": 0.00019661846496106785,
+      "loss": 4.6449,
+      "step": 82
+    },
+    {
+      "epoch": 1.1070234113712374,
+      "grad_norm": 0.4526231288909912,
+      "learning_rate": 0.00019657397107897666,
+      "loss": 4.5546,
+      "step": 83
+    },
+    {
+      "epoch": 1.120401337792642,
+      "grad_norm": 0.4847906827926636,
+      "learning_rate": 0.00019652947719688544,
+      "loss": 4.4421,
+      "step": 84
+    },
+    {
+      "epoch": 1.1337792642140467,
+      "grad_norm": 0.5136271715164185,
+      "learning_rate": 0.00019648498331479422,
+      "loss": 4.7136,
+      "step": 85
+    },
+    {
+      "epoch": 1.1471571906354514,
+      "grad_norm": 0.49209895730018616,
+      "learning_rate": 0.000196440489432703,
+      "loss": 4.3145,
+      "step": 86
+    },
+    {
+      "epoch": 1.160535117056856,
+      "grad_norm": 0.4972032904624939,
+      "learning_rate": 0.00019639599555061182,
+      "loss": 4.0408,
+      "step": 87
+    },
+    {
+      "epoch": 1.1739130434782608,
+      "grad_norm": 0.5077862739562988,
+      "learning_rate": 0.0001963515016685206,
+      "loss": 4.4074,
+      "step": 88
+    },
+    {
+      "epoch": 1.1872909698996654,
+      "grad_norm": 0.5293861031532288,
+      "learning_rate": 0.00019630700778642935,
+      "loss": 4.5385,
+      "step": 89
+    },
+    {
+      "epoch": 1.2006688963210703,
+      "grad_norm": 0.5062645673751831,
+      "learning_rate": 0.00019626251390433816,
+      "loss": 4.5141,
+      "step": 90
+    },
+    {
+      "epoch": 1.214046822742475,
+      "grad_norm": 0.49655866622924805,
+      "learning_rate": 0.00019621802002224695,
+      "loss": 4.4765,
+      "step": 91
+    },
+    {
+      "epoch": 1.2274247491638797,
+      "grad_norm": 0.6059755086898804,
+      "learning_rate": 0.00019617352614015573,
+      "loss": 4.568,
+      "step": 92
+    },
+    {
+      "epoch": 1.2408026755852843,
+      "grad_norm": 0.5442761778831482,
+      "learning_rate": 0.0001961290322580645,
+      "loss": 4.7724,
+      "step": 93
+    },
+    {
+      "epoch": 1.254180602006689,
+      "grad_norm": 0.5426056385040283,
+      "learning_rate": 0.00019608453837597332,
+      "loss": 4.5308,
+      "step": 94
+    },
+    {
+      "epoch": 1.2675585284280937,
+      "grad_norm": 0.525372326374054,
+      "learning_rate": 0.0001960400444938821,
+      "loss": 4.394,
+      "step": 95
+    },
+    {
+      "epoch": 1.2809364548494984,
+      "grad_norm": 0.5407588481903076,
+      "learning_rate": 0.00019599555061179089,
+      "loss": 4.7347,
+      "step": 96
+    },
+    {
+      "epoch": 1.294314381270903,
+      "grad_norm": 0.5726659893989563,
+      "learning_rate": 0.00019595105672969967,
+      "loss": 4.9446,
+      "step": 97
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.6211283206939697,
+      "learning_rate": 0.00019590656284760848,
+      "loss": 4.697,
+      "step": 98
+    },
+    {
+      "epoch": 1.3210702341137124,
+      "grad_norm": 0.5627567172050476,
+      "learning_rate": 0.00019586206896551723,
+      "loss": 4.4892,
+      "step": 99
+    },
+    {
+      "epoch": 1.334448160535117,
+      "grad_norm": 0.6174790859222412,
+      "learning_rate": 0.00019581757508342604,
+      "loss": 4.5686,
+      "step": 100
+    },
+    {
+      "epoch": 1.3478260869565217,
+      "grad_norm": 0.5586990118026733,
+      "learning_rate": 0.00019577308120133482,
+      "loss": 4.3916,
+      "step": 101
+    },
+    {
+      "epoch": 1.3612040133779264,
+      "grad_norm": 0.5655365586280823,
+      "learning_rate": 0.0001957285873192436,
+      "loss": 4.1789,
+      "step": 102
+    },
+    {
+      "epoch": 1.374581939799331,
+      "grad_norm": 0.5834594964981079,
+      "learning_rate": 0.0001956840934371524,
+      "loss": 4.3316,
+      "step": 103
+    },
+    {
+      "epoch": 1.3879598662207357,
+      "grad_norm": 0.6065447926521301,
+      "learning_rate": 0.0001956395995550612,
+      "loss": 4.5167,
+      "step": 104
+    },
+    {
+      "epoch": 1.4013377926421404,
+      "grad_norm": 0.5250216722488403,
+      "learning_rate": 0.00019559510567296998,
+      "loss": 4.1718,
+      "step": 105
+    },
+    {
+      "epoch": 1.414715719063545,
+      "grad_norm": 0.5861116051673889,
+      "learning_rate": 0.00019555061179087876,
+      "loss": 4.3077,
+      "step": 106
+    },
+    {
+      "epoch": 1.4280936454849498,
+      "grad_norm": 0.6138104796409607,
+      "learning_rate": 0.00019550611790878755,
+      "loss": 4.4748,
+      "step": 107
+    },
+    {
+      "epoch": 1.4414715719063544,
+      "grad_norm": 0.6742071509361267,
+      "learning_rate": 0.00019546162402669636,
+      "loss": 4.8769,
+      "step": 108
+    },
+    {
+      "epoch": 1.4548494983277591,
+      "grad_norm": 0.6634951233863831,
+      "learning_rate": 0.0001954171301446051,
+      "loss": 4.6423,
+      "step": 109
+    },
+    {
+      "epoch": 1.468227424749164,
+      "grad_norm": 0.626646876335144,
+      "learning_rate": 0.00019537263626251392,
+      "loss": 4.4654,
+      "step": 110
+    },
+    {
+      "epoch": 1.4816053511705687,
+      "grad_norm": 0.6306963562965393,
+      "learning_rate": 0.0001953281423804227,
+      "loss": 4.7021,
+      "step": 111
+    },
+    {
+      "epoch": 1.4949832775919734,
+      "grad_norm": 0.620370626449585,
+      "learning_rate": 0.00019528364849833149,
+      "loss": 4.587,
+      "step": 112
+    },
+    {
+      "epoch": 1.508361204013378,
+      "grad_norm": 0.6410287618637085,
+      "learning_rate": 0.00019523915461624027,
+      "loss": 4.8089,
+      "step": 113
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 0.676434338092804,
+      "learning_rate": 0.00019519466073414908,
+      "loss": 4.668,
+      "step": 114
+    },
+    {
+      "epoch": 1.5351170568561874,
+      "grad_norm": 0.5756319761276245,
+      "learning_rate": 0.00019515016685205786,
+      "loss": 4.3223,
+      "step": 115
+    },
+    {
+      "epoch": 1.548494983277592,
+      "grad_norm": 0.5850693583488464,
+      "learning_rate": 0.00019510567296996664,
+      "loss": 4.2343,
+      "step": 116
+    },
+    {
+      "epoch": 1.5618729096989967,
+      "grad_norm": 0.6172360777854919,
+      "learning_rate": 0.00019506117908787542,
+      "loss": 4.6102,
+      "step": 117
+    },
+    {
+      "epoch": 1.5752508361204014,
+      "grad_norm": 0.5887568593025208,
+      "learning_rate": 0.00019501668520578423,
+      "loss": 4.8097,
+      "step": 118
+    },
+    {
+      "epoch": 1.588628762541806,
+      "grad_norm": 0.5763369798660278,
+      "learning_rate": 0.000194972191323693,
+      "loss": 4.2001,
+      "step": 119
+    },
+    {
+      "epoch": 1.6020066889632107,
+      "grad_norm": 0.6158986687660217,
+      "learning_rate": 0.0001949276974416018,
+      "loss": 4.7075,
+      "step": 120
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.5540957450866699,
+      "learning_rate": 0.00019488320355951058,
+      "loss": 4.452,
+      "step": 121
+    },
+    {
+      "epoch": 1.62876254180602,
+      "grad_norm": 0.6193795204162598,
+      "learning_rate": 0.00019483870967741936,
+      "loss": 4.4583,
+      "step": 122
+    },
+    {
+      "epoch": 1.6421404682274248,
+      "grad_norm": 0.6699966788291931,
+      "learning_rate": 0.00019479421579532815,
+      "loss": 4.3728,
+      "step": 123
+    },
+    {
+      "epoch": 1.6555183946488294,
+      "grad_norm": 0.5904677510261536,
+      "learning_rate": 0.00019474972191323696,
+      "loss": 4.5452,
+      "step": 124
+    },
+    {
+      "epoch": 1.6688963210702341,
+      "grad_norm": 0.6137760281562805,
+      "learning_rate": 0.00019470522803114574,
+      "loss": 4.2853,
+      "step": 125
+    },
+    {
+      "epoch": 1.6822742474916388,
+      "grad_norm": 0.6396192908287048,
+      "learning_rate": 0.00019466073414905452,
+      "loss": 4.4258,
+      "step": 126
+    },
+    {
+      "epoch": 1.6956521739130435,
+      "grad_norm": 0.6190487742424011,
+      "learning_rate": 0.0001946162402669633,
+      "loss": 4.9866,
+      "step": 127
+    },
+    {
+      "epoch": 1.7090301003344481,
+      "grad_norm": 0.6971675157546997,
+      "learning_rate": 0.0001945717463848721,
+      "loss": 4.2126,
+      "step": 128
+    },
+    {
+      "epoch": 1.7224080267558528,
+      "grad_norm": 0.6245931386947632,
+      "learning_rate": 0.00019452725250278087,
+      "loss": 4.8477,
+      "step": 129
+    },
+    {
+      "epoch": 1.7357859531772575,
+      "grad_norm": 0.5675052404403687,
+      "learning_rate": 0.00019448275862068965,
+      "loss": 4.4097,
+      "step": 130
+    },
+    {
+      "epoch": 1.7491638795986622,
+      "grad_norm": 0.6594040393829346,
+      "learning_rate": 0.00019443826473859846,
+      "loss": 4.3747,
+      "step": 131
+    },
+    {
+      "epoch": 1.7625418060200668,
+      "grad_norm": 0.6377655267715454,
+      "learning_rate": 0.00019439377085650724,
+      "loss": 4.2733,
+      "step": 132
+    },
+    {
+      "epoch": 1.7759197324414715,
+      "grad_norm": 0.6167862415313721,
+      "learning_rate": 0.00019434927697441602,
+      "loss": 4.5694,
+      "step": 133
+    },
+    {
+      "epoch": 1.7892976588628762,
+      "grad_norm": 0.577671468257904,
+      "learning_rate": 0.0001943047830923248,
+      "loss": 4.5006,
+      "step": 134
+    },
+    {
+      "epoch": 1.8026755852842808,
+      "grad_norm": 0.6361016035079956,
+      "learning_rate": 0.00019426028921023362,
+      "loss": 4.9907,
+      "step": 135
+    },
+    {
+      "epoch": 1.8160535117056855,
+      "grad_norm": 0.6445321440696716,
+      "learning_rate": 0.00019421579532814237,
+      "loss": 4.779,
+      "step": 136
+    },
+    {
+      "epoch": 1.8294314381270902,
+      "grad_norm": 0.5955402851104736,
+      "learning_rate": 0.00019417130144605118,
+      "loss": 4.6026,
+      "step": 137
+    },
+    {
+      "epoch": 1.8428093645484949,
+      "grad_norm": 0.6807080507278442,
+      "learning_rate": 0.00019412680756395996,
+      "loss": 4.7124,
+      "step": 138
+    },
+    {
+      "epoch": 1.8561872909698995,
+      "grad_norm": 0.5799288153648376,
+      "learning_rate": 0.00019408231368186875,
+      "loss": 4.0701,
+      "step": 139
+    },
+    {
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.6187757253646851,
+      "learning_rate": 0.00019403781979977753,
+      "loss": 4.705,
+      "step": 140
+    },
+    {
+      "epoch": 1.8829431438127089,
+      "grad_norm": 0.6614826917648315,
+      "learning_rate": 0.00019399332591768634,
+      "loss": 4.8146,
+      "step": 141
+    },
+    {
+      "epoch": 1.8963210702341136,
+      "grad_norm": 0.6204859614372253,
+      "learning_rate": 0.00019394883203559512,
+      "loss": 4.3041,
+      "step": 142
+    },
+    {
+      "epoch": 1.9096989966555182,
+      "grad_norm": 0.6527450680732727,
+      "learning_rate": 0.0001939043381535039,
+      "loss": 4.4493,
+      "step": 143
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.6470615267753601,
+      "learning_rate": 0.00019385984427141268,
+      "loss": 4.7771,
+      "step": 144
+    },
+    {
+      "epoch": 1.9364548494983278,
+      "grad_norm": 0.5642555952072144,
+      "learning_rate": 0.0001938153503893215,
+      "loss": 4.3344,
+      "step": 145
+    },
+    {
+      "epoch": 1.9498327759197325,
+      "grad_norm": 0.6206467151641846,
+      "learning_rate": 0.00019377085650723025,
+      "loss": 4.2191,
+      "step": 146
+    },
+    {
+      "epoch": 1.9632107023411371,
+      "grad_norm": 0.6079016923904419,
+      "learning_rate": 0.00019372636262513906,
+      "loss": 4.7397,
+      "step": 147
+    },
+    {
+      "epoch": 1.9765886287625418,
+      "grad_norm": 0.6197662353515625,
+      "learning_rate": 0.00019368186874304784,
+      "loss": 4.5342,
+      "step": 148
+    },
+    {
+      "epoch": 1.9899665551839465,
+      "grad_norm": 0.6556297540664673,
+      "learning_rate": 0.00019363737486095662,
+      "loss": 4.6709,
+      "step": 149
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7837930917739868,
+      "learning_rate": 0.0001935928809788654,
+      "loss": 4.6215,
+      "step": 150
+    },
+    {
+      "epoch": 2.0133779264214047,
+      "grad_norm": 0.5267267227172852,
+      "learning_rate": 0.00019354838709677422,
+      "loss": 4.2695,
+      "step": 151
+    },
+    {
+      "epoch": 2.0267558528428093,
+      "grad_norm": 0.5862157344818115,
+      "learning_rate": 0.000193503893214683,
+      "loss": 4.3702,
+      "step": 152
+    },
+    {
+      "epoch": 2.040133779264214,
+      "grad_norm": 0.538254976272583,
+      "learning_rate": 0.00019345939933259178,
+      "loss": 4.3953,
+      "step": 153
+    },
+    {
+      "epoch": 2.0535117056856187,
+      "grad_norm": 0.5977053642272949,
+      "learning_rate": 0.00019341490545050056,
+      "loss": 4.2156,
+      "step": 154
+    },
+    {
+      "epoch": 2.0668896321070234,
+      "grad_norm": 0.606006383895874,
+      "learning_rate": 0.00019337041156840937,
+      "loss": 4.2802,
+      "step": 155
+    },
+    {
+      "epoch": 2.080267558528428,
+      "grad_norm": 0.6071277856826782,
+      "learning_rate": 0.00019332591768631813,
+      "loss": 4.5545,
+      "step": 156
+    },
+    {
+      "epoch": 2.0936454849498327,
+      "grad_norm": 0.6281546354293823,
+      "learning_rate": 0.00019328142380422694,
+      "loss": 4.6105,
+      "step": 157
+    },
+    {
+      "epoch": 2.1070234113712374,
+      "grad_norm": 0.5703116655349731,
+      "learning_rate": 0.00019323692992213572,
+      "loss": 4.2751,
+      "step": 158
+    },
+    {
+      "epoch": 2.120401337792642,
+      "grad_norm": 0.6587452292442322,
+      "learning_rate": 0.0001931924360400445,
+      "loss": 4.6342,
+      "step": 159
+    },
+    {
+      "epoch": 2.1337792642140467,
+      "grad_norm": 0.6141905784606934,
+      "learning_rate": 0.00019314794215795328,
+      "loss": 4.4345,
+      "step": 160
+    },
+    {
+      "epoch": 2.1471571906354514,
+      "grad_norm": 0.6741939187049866,
+      "learning_rate": 0.0001931034482758621,
+      "loss": 4.0257,
+      "step": 161
+    },
+    {
+      "epoch": 2.160535117056856,
+      "grad_norm": 0.6468759179115295,
+      "learning_rate": 0.00019305895439377088,
+      "loss": 4.2313,
+      "step": 162
+    },
+    {
+      "epoch": 2.1739130434782608,
+      "grad_norm": 0.6703383326530457,
+      "learning_rate": 0.00019301446051167966,
+      "loss": 4.2164,
+      "step": 163
+    },
+    {
+      "epoch": 2.1872909698996654,
+      "grad_norm": 0.710967481136322,
+      "learning_rate": 0.00019296996662958844,
+      "loss": 4.3398,
+      "step": 164
+    },
+    {
+      "epoch": 2.20066889632107,
+      "grad_norm": 0.6862124800682068,
+      "learning_rate": 0.00019292547274749725,
+      "loss": 4.3379,
+      "step": 165
+    },
+    {
+      "epoch": 2.2140468227424748,
+      "grad_norm": 0.6288430690765381,
+      "learning_rate": 0.000192880978865406,
+      "loss": 4.3487,
+      "step": 166
+    },
+    {
+      "epoch": 2.2274247491638794,
+      "grad_norm": 0.6358796954154968,
+      "learning_rate": 0.00019283648498331481,
+      "loss": 4.1656,
+      "step": 167
+    },
+    {
+      "epoch": 2.240802675585284,
+      "grad_norm": 0.6818917393684387,
+      "learning_rate": 0.0001927919911012236,
+      "loss": 4.5363,
+      "step": 168
+    },
+    {
+      "epoch": 2.254180602006689,
+      "grad_norm": 0.6996105313301086,
+      "learning_rate": 0.00019274749721913238,
+      "loss": 4.3208,
+      "step": 169
+    },
+    {
+      "epoch": 2.2675585284280935,
+      "grad_norm": 0.6730326414108276,
+      "learning_rate": 0.00019270300333704116,
+      "loss": 4.1401,
+      "step": 170
+    },
+    {
+      "epoch": 2.280936454849498,
+      "grad_norm": 0.7022603750228882,
+      "learning_rate": 0.00019265850945494994,
+      "loss": 4.5761,
+      "step": 171
+    },
+    {
+      "epoch": 2.294314381270903,
+      "grad_norm": 0.6525995135307312,
+      "learning_rate": 0.00019261401557285875,
+      "loss": 4.4017,
+      "step": 172
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.7066033482551575,
+      "learning_rate": 0.0001925695216907675,
+      "loss": 4.0037,
+      "step": 173
+    },
+    {
+      "epoch": 2.321070234113712,
+      "grad_norm": 0.6708059310913086,
+      "learning_rate": 0.00019252502780867632,
+      "loss": 4.1947,
+      "step": 174
+    },
+    {
+      "epoch": 2.334448160535117,
+      "grad_norm": 0.8711172342300415,
+      "learning_rate": 0.0001924805339265851,
+      "loss": 3.9958,
+      "step": 175
+    },
+    {
+      "epoch": 2.3478260869565215,
+      "grad_norm": 0.7258634567260742,
+      "learning_rate": 0.00019243604004449388,
+      "loss": 4.4682,
+      "step": 176
+    },
+    {
+      "epoch": 2.361204013377926,
+      "grad_norm": 0.7693021893501282,
+      "learning_rate": 0.00019239154616240267,
+      "loss": 4.54,
+      "step": 177
+    },
+    {
+      "epoch": 2.374581939799331,
+      "grad_norm": 0.7271276116371155,
+      "learning_rate": 0.00019234705228031148,
+      "loss": 4.2942,
+      "step": 178
+    },
+    {
+      "epoch": 2.387959866220736,
+      "grad_norm": 0.6836609244346619,
+      "learning_rate": 0.00019230255839822026,
+      "loss": 4.3099,
+      "step": 179
+    },
+    {
+      "epoch": 2.4013377926421406,
+      "grad_norm": 0.731164813041687,
+      "learning_rate": 0.00019225806451612904,
+      "loss": 4.4077,
+      "step": 180
+    },
+    {
+      "epoch": 2.4147157190635453,
+      "grad_norm": 0.7575274109840393,
+      "learning_rate": 0.00019221357063403782,
+      "loss": 4.6572,
+      "step": 181
+    },
+    {
+      "epoch": 2.42809364548495,
+      "grad_norm": 0.8461325168609619,
+      "learning_rate": 0.00019216907675194663,
+      "loss": 4.4922,
+      "step": 182
+    },
+    {
+      "epoch": 2.4414715719063547,
+      "grad_norm": 0.7225251197814941,
+      "learning_rate": 0.0001921245828698554,
+      "loss": 4.0372,
+      "step": 183
+    },
+    {
+      "epoch": 2.4548494983277593,
+      "grad_norm": 3.563720703125,
+      "learning_rate": 0.0001920800889877642,
+      "loss": 4.5412,
+      "step": 184
+    },
+    {
+      "epoch": 2.468227424749164,
+      "grad_norm": 0.8452121019363403,
+      "learning_rate": 0.00019203559510567298,
+      "loss": 4.4961,
+      "step": 185
+    },
+    {
+      "epoch": 2.4816053511705687,
+      "grad_norm": 0.8734024167060852,
+      "learning_rate": 0.00019199110122358176,
+      "loss": 4.0884,
+      "step": 186
+    },
+    {
+      "epoch": 2.4949832775919734,
+      "grad_norm": 1.1765823364257812,
+      "learning_rate": 0.00019194660734149054,
+      "loss": 4.2228,
+      "step": 187
+    },
+    {
+      "epoch": 2.508361204013378,
+      "grad_norm": 0.750206708908081,
+      "learning_rate": 0.00019190211345939935,
+      "loss": 4.4305,
+      "step": 188
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 0.7574430704116821,
+      "learning_rate": 0.00019185761957730814,
+      "loss": 4.1511,
+      "step": 189
+    },
+    {
+      "epoch": 2.5351170568561874,
+      "grad_norm": 0.7105517387390137,
+      "learning_rate": 0.00019181312569521692,
+      "loss": 4.4793,
+      "step": 190
+    },
+    {
+      "epoch": 2.548494983277592,
+      "grad_norm": 0.7495557069778442,
+      "learning_rate": 0.0001917686318131257,
+      "loss": 4.1335,
+      "step": 191
+    },
+    {
+      "epoch": 2.5618729096989967,
+      "grad_norm": 0.8001168966293335,
+      "learning_rate": 0.0001917241379310345,
+      "loss": 4.7898,
+      "step": 192
+    },
+    {
+      "epoch": 2.5752508361204014,
+      "grad_norm": 0.7402104735374451,
+      "learning_rate": 0.00019167964404894327,
+      "loss": 4.4482,
+      "step": 193
+    },
+    {
+      "epoch": 2.588628762541806,
+      "grad_norm": 0.748267650604248,
+      "learning_rate": 0.00019163515016685207,
+      "loss": 4.3167,
+      "step": 194
+    },
+    {
+      "epoch": 2.6020066889632107,
+      "grad_norm": 0.8291250467300415,
+      "learning_rate": 0.00019159065628476086,
+      "loss": 4.058,
+      "step": 195
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.6945542693138123,
+      "learning_rate": 0.00019154616240266964,
+      "loss": 3.9751,
+      "step": 196
+    },
+    {
+      "epoch": 2.62876254180602,
+      "grad_norm": 0.7307319045066833,
+      "learning_rate": 0.00019150166852057842,
+      "loss": 4.2736,
+      "step": 197
+    },
+    {
+      "epoch": 2.6421404682274248,
+      "grad_norm": 0.7489168047904968,
+      "learning_rate": 0.00019145717463848723,
+      "loss": 4.3075,
+      "step": 198
+    },
+    {
+      "epoch": 2.6555183946488294,
+      "grad_norm": 0.9727582931518555,
+      "learning_rate": 0.00019141268075639601,
+      "loss": 4.6474,
+      "step": 199
+    },
+    {
+      "epoch": 2.668896321070234,
+      "grad_norm": 0.6776256561279297,
+      "learning_rate": 0.0001913681868743048,
+      "loss": 4.4217,
+      "step": 200
+    },
+    {
+      "epoch": 2.682274247491639,
+      "grad_norm": 0.7305111885070801,
+      "learning_rate": 0.00019132369299221358,
+      "loss": 4.2804,
+      "step": 201
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 0.7196978330612183,
+      "learning_rate": 0.0001912791991101224,
+      "loss": 4.3941,
+      "step": 202
+    },
+    {
+      "epoch": 2.709030100334448,
+      "grad_norm": 0.7988458871841431,
+      "learning_rate": 0.00019123470522803114,
+      "loss": 4.437,
+      "step": 203
+    },
+    {
+      "epoch": 2.722408026755853,
+      "grad_norm": 0.7004797458648682,
+      "learning_rate": 0.00019119021134593995,
+      "loss": 4.4986,
+      "step": 204
+    },
+    {
+      "epoch": 2.7357859531772575,
+      "grad_norm": 0.677796483039856,
+      "learning_rate": 0.00019114571746384874,
+      "loss": 4.0851,
+      "step": 205
+    },
+    {
+      "epoch": 2.749163879598662,
+      "grad_norm": 0.7527475357055664,
+      "learning_rate": 0.00019110122358175752,
+      "loss": 4.4469,
+      "step": 206
+    },
+    {
+      "epoch": 2.762541806020067,
+      "grad_norm": 1.1659115552902222,
+      "learning_rate": 0.0001910567296996663,
+      "loss": 4.3284,
+      "step": 207
+    },
+    {
+      "epoch": 2.7759197324414715,
+      "grad_norm": 0.7238364815711975,
+      "learning_rate": 0.0001910122358175751,
+      "loss": 4.2605,
+      "step": 208
+    },
+    {
+      "epoch": 2.789297658862876,
+      "grad_norm": 0.7537760734558105,
+      "learning_rate": 0.0001909677419354839,
+      "loss": 4.3775,
+      "step": 209
+    },
+    {
+      "epoch": 2.802675585284281,
+      "grad_norm": 0.6874127388000488,
+      "learning_rate": 0.00019092324805339267,
+      "loss": 4.3404,
+      "step": 210
+    },
+    {
+      "epoch": 2.8160535117056855,
+      "grad_norm": 0.7045959830284119,
+      "learning_rate": 0.00019087875417130146,
+      "loss": 4.1568,
+      "step": 211
+    },
+    {
+      "epoch": 2.82943143812709,
+      "grad_norm": 0.7249194383621216,
+      "learning_rate": 0.00019083426028921027,
+      "loss": 4.1969,
+      "step": 212
+    },
+    {
+      "epoch": 2.842809364548495,
+      "grad_norm": 0.8331268429756165,
+      "learning_rate": 0.00019078976640711902,
+      "loss": 4.3169,
+      "step": 213
+    },
+    {
+      "epoch": 2.8561872909698995,
+      "grad_norm": 0.7171936631202698,
+      "learning_rate": 0.0001907452725250278,
+      "loss": 4.5123,
+      "step": 214
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 0.759919285774231,
+      "learning_rate": 0.0001907007786429366,
+      "loss": 4.5412,
+      "step": 215
+    },
+    {
+      "epoch": 2.882943143812709,
+      "grad_norm": 0.7451274991035461,
+      "learning_rate": 0.0001906562847608454,
+      "loss": 4.5253,
+      "step": 216
+    },
+    {
+      "epoch": 2.8963210702341136,
+      "grad_norm": 0.6564481258392334,
+      "learning_rate": 0.00019061179087875418,
+      "loss": 4.1092,
+      "step": 217
+    },
+    {
+      "epoch": 2.9096989966555182,
+      "grad_norm": 0.7339865565299988,
+      "learning_rate": 0.00019056729699666296,
+      "loss": 4.5092,
+      "step": 218
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.7113937735557556,
+      "learning_rate": 0.00019052280311457177,
+      "loss": 4.3355,
+      "step": 219
+    },
+    {
+      "epoch": 2.936454849498328,
+      "grad_norm": 0.7306456565856934,
+      "learning_rate": 0.00019047830923248053,
+      "loss": 4.5745,
+      "step": 220
+    },
+    {
+      "epoch": 2.9498327759197327,
+      "grad_norm": 0.7971818447113037,
+      "learning_rate": 0.00019043381535038933,
+      "loss": 4.2903,
+      "step": 221
+    },
+    {
+      "epoch": 2.9632107023411374,
+      "grad_norm": 0.7757331728935242,
+      "learning_rate": 0.00019038932146829812,
+      "loss": 4.2832,
+      "step": 222
+    },
+    {
+      "epoch": 2.976588628762542,
+      "grad_norm": 0.7326288223266602,
+      "learning_rate": 0.0001903448275862069,
+      "loss": 4.2444,
+      "step": 223
+    },
+    {
+      "epoch": 2.9899665551839467,
+      "grad_norm": 0.7363834381103516,
+      "learning_rate": 0.00019030033370411568,
+      "loss": 4.6744,
+      "step": 224
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8835271596908569,
+      "learning_rate": 0.0001902558398220245,
+      "loss": 4.432,
+      "step": 225
+    },
+    {
+      "epoch": 3.0133779264214047,
+      "grad_norm": 0.6591921448707581,
+      "learning_rate": 0.00019021134593993327,
+      "loss": 4.1353,
+      "step": 226
+    },
+    {
+      "epoch": 3.0267558528428093,
+      "grad_norm": 0.6895263195037842,
+      "learning_rate": 0.00019016685205784206,
+      "loss": 4.1253,
+      "step": 227
+    },
+    {
+      "epoch": 3.040133779264214,
+      "grad_norm": 0.6476898789405823,
+      "learning_rate": 0.00019012235817575084,
+      "loss": 4.0354,
+      "step": 228
+    },
+    {
+      "epoch": 3.0535117056856187,
+      "grad_norm": 0.6398957967758179,
+      "learning_rate": 0.00019007786429365965,
+      "loss": 4.024,
+      "step": 229
+    },
+    {
+      "epoch": 3.0668896321070234,
+      "grad_norm": 0.7483389973640442,
+      "learning_rate": 0.0001900333704115684,
+      "loss": 4.1405,
+      "step": 230
+    },
+    {
+      "epoch": 3.080267558528428,
+      "grad_norm": 0.7003724575042725,
+      "learning_rate": 0.0001899888765294772,
+      "loss": 4.3593,
+      "step": 231
+    },
+    {
+      "epoch": 3.0936454849498327,
+      "grad_norm": 0.7426732182502747,
+      "learning_rate": 0.000189944382647386,
+      "loss": 4.485,
+      "step": 232
+    },
+    {
+      "epoch": 3.1070234113712374,
+      "grad_norm": 0.6957541108131409,
+      "learning_rate": 0.00018989988876529478,
+      "loss": 4.1017,
+      "step": 233
+    },
+    {
+      "epoch": 3.120401337792642,
+      "grad_norm": 0.8613067865371704,
+      "learning_rate": 0.00018985539488320356,
+      "loss": 4.3038,
+      "step": 234
+    },
+    {
+      "epoch": 3.1337792642140467,
+      "grad_norm": 0.8375754952430725,
+      "learning_rate": 0.00018981090100111237,
+      "loss": 4.4356,
+      "step": 235
+    },
+    {
+      "epoch": 3.1471571906354514,
+      "grad_norm": 0.7878522872924805,
+      "learning_rate": 0.00018976640711902115,
+      "loss": 3.9916,
+      "step": 236
+    },
+    {
+      "epoch": 3.160535117056856,
+      "grad_norm": 0.7463901042938232,
+      "learning_rate": 0.00018972191323692993,
+      "loss": 3.6761,
+      "step": 237
+    },
+    {
+      "epoch": 3.1739130434782608,
+      "grad_norm": 0.7360939979553223,
+      "learning_rate": 0.00018967741935483872,
+      "loss": 3.9573,
+      "step": 238
+    },
+    {
+      "epoch": 3.1872909698996654,
+      "grad_norm": 0.891861081123352,
+      "learning_rate": 0.00018963292547274753,
+      "loss": 4.1853,
+      "step": 239
+    },
+    {
+      "epoch": 3.20066889632107,
+      "grad_norm": 0.8589549660682678,
+      "learning_rate": 0.00018958843159065628,
+      "loss": 4.0679,
+      "step": 240
+    },
+    {
+      "epoch": 3.2140468227424748,
+      "grad_norm": 0.9534163475036621,
+      "learning_rate": 0.0001895439377085651,
+      "loss": 3.732,
+      "step": 241
+    },
+    {
+      "epoch": 3.2274247491638794,
+      "grad_norm": 0.8968185186386108,
+      "learning_rate": 0.00018949944382647387,
+      "loss": 4.2217,
+      "step": 242
+    },
+    {
+      "epoch": 3.240802675585284,
+      "grad_norm": 0.81589275598526,
+      "learning_rate": 0.00018945494994438266,
+      "loss": 4.428,
+      "step": 243
+    },
+    {
+      "epoch": 3.254180602006689,
+      "grad_norm": 0.929050862789154,
+      "learning_rate": 0.00018941045606229144,
+      "loss": 4.3468,
+      "step": 244
+    },
+    {
+      "epoch": 3.2675585284280935,
+      "grad_norm": 0.8535035252571106,
+      "learning_rate": 0.00018936596218020025,
+      "loss": 3.8489,
+      "step": 245
+    },
+    {
+      "epoch": 3.280936454849498,
+      "grad_norm": 0.9484681487083435,
+      "learning_rate": 0.00018932146829810903,
+      "loss": 4.0132,
+      "step": 246
+    },
+    {
+      "epoch": 3.294314381270903,
+      "grad_norm": 0.8190047144889832,
+      "learning_rate": 0.0001892769744160178,
+      "loss": 4.3574,
+      "step": 247
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.8764749765396118,
+      "learning_rate": 0.0001892324805339266,
+      "loss": 4.3103,
+      "step": 248
+    },
+    {
+      "epoch": 3.321070234113712,
+      "grad_norm": 0.8929185271263123,
+      "learning_rate": 0.0001891879866518354,
+      "loss": 4.3606,
+      "step": 249
+    },
+    {
+      "epoch": 3.334448160535117,
+      "grad_norm": 0.9096692204475403,
+      "learning_rate": 0.00018914349276974416,
+      "loss": 4.0047,
+      "step": 250
+    },
+    {
+      "epoch": 3.3478260869565215,
+      "grad_norm": 0.885143518447876,
+      "learning_rate": 0.00018909899888765297,
+      "loss": 4.182,
+      "step": 251
+    },
+    {
+      "epoch": 3.361204013377926,
+      "grad_norm": 0.7724215984344482,
+      "learning_rate": 0.00018905450500556175,
+      "loss": 3.9529,
+      "step": 252
+    },
+    {
+      "epoch": 3.374581939799331,
+      "grad_norm": 0.8351865410804749,
+      "learning_rate": 0.00018901001112347053,
+      "loss": 3.9533,
+      "step": 253
+    },
+    {
+      "epoch": 3.387959866220736,
+      "grad_norm": 0.8684999942779541,
+      "learning_rate": 0.00018896551724137932,
+      "loss": 3.8594,
+      "step": 254
+    },
+    {
+      "epoch": 3.4013377926421406,
+      "grad_norm": 0.8903334736824036,
+      "learning_rate": 0.0001889210233592881,
+      "loss": 3.9248,
+      "step": 255
+    },
+    {
+      "epoch": 3.4147157190635453,
+      "grad_norm": 0.826690137386322,
+      "learning_rate": 0.0001888765294771969,
+      "loss": 4.0389,
+      "step": 256
+    },
+    {
+      "epoch": 3.42809364548495,
+      "grad_norm": 0.8306142687797546,
+      "learning_rate": 0.00018883203559510566,
+      "loss": 3.8168,
+      "step": 257
+    },
+    {
+      "epoch": 3.4414715719063547,
+      "grad_norm": 0.9032199382781982,
+      "learning_rate": 0.00018878754171301447,
+      "loss": 4.178,
+      "step": 258
+    },
+    {
+      "epoch": 3.4548494983277593,
+      "grad_norm": 0.9081966280937195,
+      "learning_rate": 0.00018874304783092326,
+      "loss": 4.2583,
+      "step": 259
+    },
+    {
+      "epoch": 3.468227424749164,
+      "grad_norm": 0.8424077033996582,
+      "learning_rate": 0.00018869855394883204,
+      "loss": 4.3285,
+      "step": 260
+    },
+    {
+      "epoch": 3.4816053511705687,
+      "grad_norm": 0.8302170038223267,
+      "learning_rate": 0.00018865406006674082,
+      "loss": 4.1346,
+      "step": 261
+    },
+    {
+      "epoch": 3.4949832775919734,
+      "grad_norm": 0.8747193217277527,
+      "learning_rate": 0.00018860956618464963,
+      "loss": 4.0747,
+      "step": 262
+    },
+    {
+      "epoch": 3.508361204013378,
+      "grad_norm": 0.8613927364349365,
+      "learning_rate": 0.0001885650723025584,
+      "loss": 4.2346,
+      "step": 263
+    },
+    {
+      "epoch": 3.5217391304347827,
+      "grad_norm": 0.8321558833122253,
+      "learning_rate": 0.0001885205784204672,
+      "loss": 3.9781,
+      "step": 264
+    },
+    {
+      "epoch": 3.5351170568561874,
+      "grad_norm": 0.8961741328239441,
+      "learning_rate": 0.00018847608453837598,
+      "loss": 4.311,
+      "step": 265
+    },
+    {
+      "epoch": 3.548494983277592,
+      "grad_norm": 0.7703898549079895,
+      "learning_rate": 0.00018843159065628479,
+      "loss": 4.1163,
+      "step": 266
+    },
+    {
+      "epoch": 3.5618729096989967,
+      "grad_norm": 0.880051851272583,
+      "learning_rate": 0.00018838709677419354,
+      "loss": 3.8032,
+      "step": 267
+    },
+    {
+      "epoch": 3.5752508361204014,
+      "grad_norm": 0.8287038207054138,
+      "learning_rate": 0.00018834260289210235,
+      "loss": 4.1627,
+      "step": 268
+    },
+    {
+      "epoch": 3.588628762541806,
+      "grad_norm": 0.9726569652557373,
+      "learning_rate": 0.00018829810901001113,
+      "loss": 4.4055,
+      "step": 269
+    },
+    {
+      "epoch": 3.6020066889632107,
+      "grad_norm": 0.8071132898330688,
+      "learning_rate": 0.00018825361512791992,
+      "loss": 4.1709,
+      "step": 270
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.8310988545417786,
+      "learning_rate": 0.0001882091212458287,
+      "loss": 4.2359,
+      "step": 271
+    },
+    {
+      "epoch": 3.62876254180602,
+      "grad_norm": 0.8713561296463013,
+      "learning_rate": 0.0001881646273637375,
+      "loss": 4.1247,
+      "step": 272
+    },
+    {
+      "epoch": 3.6421404682274248,
+      "grad_norm": 0.8964342474937439,
+      "learning_rate": 0.0001881201334816463,
+      "loss": 4.0794,
+      "step": 273
+    },
+    {
+      "epoch": 3.6555183946488294,
+      "grad_norm": 0.9901681542396545,
+      "learning_rate": 0.00018807563959955507,
+      "loss": 4.0217,
+      "step": 274
+    },
+    {
+      "epoch": 3.668896321070234,
+      "grad_norm": 0.9279042482376099,
+      "learning_rate": 0.00018803114571746385,
+      "loss": 4.3244,
+      "step": 275
+    },
+    {
+      "epoch": 3.682274247491639,
+      "grad_norm": 0.8105964660644531,
+      "learning_rate": 0.00018798665183537266,
+      "loss": 3.9041,
+      "step": 276
+    },
+    {
+      "epoch": 3.6956521739130435,
+      "grad_norm": 0.8511622548103333,
+      "learning_rate": 0.00018794215795328142,
+      "loss": 3.8969,
+      "step": 277
+    },
+    {
+      "epoch": 3.709030100334448,
+      "grad_norm": 0.9072037935256958,
+      "learning_rate": 0.00018789766407119023,
+      "loss": 4.2185,
+      "step": 278
+    },
+    {
+      "epoch": 3.722408026755853,
+      "grad_norm": 0.9792962670326233,
+      "learning_rate": 0.000187853170189099,
+      "loss": 4.1915,
+      "step": 279
+    },
+    {
+      "epoch": 3.7357859531772575,
+      "grad_norm": 0.8579828143119812,
+      "learning_rate": 0.0001878086763070078,
+      "loss": 3.8903,
+      "step": 280
+    },
+    {
+      "epoch": 3.749163879598662,
+      "grad_norm": 0.9866719841957092,
+      "learning_rate": 0.00018776418242491658,
+      "loss": 4.2022,
+      "step": 281
+    },
+    {
+      "epoch": 3.762541806020067,
+      "grad_norm": 0.9251964688301086,
+      "learning_rate": 0.00018771968854282539,
+      "loss": 3.9536,
+      "step": 282
+    },
+    {
+      "epoch": 3.7759197324414715,
+      "grad_norm": 1.0300836563110352,
+      "learning_rate": 0.00018767519466073417,
+      "loss": 4.2908,
+      "step": 283
+    },
+    {
+      "epoch": 3.789297658862876,
+      "grad_norm": 1.0194575786590576,
+      "learning_rate": 0.00018763070077864295,
+      "loss": 4.1851,
+      "step": 284
+    },
+    {
+      "epoch": 3.802675585284281,
+      "grad_norm": 0.8165330290794373,
+      "learning_rate": 0.00018758620689655173,
+      "loss": 4.138,
+      "step": 285
+    },
+    {
+      "epoch": 3.8160535117056855,
+      "grad_norm": 1.0104280710220337,
+      "learning_rate": 0.00018754171301446054,
+      "loss": 3.9481,
+      "step": 286
+    },
+    {
+      "epoch": 3.82943143812709,
+      "grad_norm": 0.9972538352012634,
+      "learning_rate": 0.0001874972191323693,
+      "loss": 4.3932,
+      "step": 287
+    },
+    {
+      "epoch": 3.842809364548495,
+      "grad_norm": 0.96323162317276,
+      "learning_rate": 0.0001874527252502781,
+      "loss": 4.1133,
+      "step": 288
+    },
+    {
+      "epoch": 3.8561872909698995,
+      "grad_norm": 0.8500615954399109,
+      "learning_rate": 0.0001874082313681869,
+      "loss": 4.2205,
+      "step": 289
+    },
+    {
+      "epoch": 3.869565217391304,
+      "grad_norm": 0.8451250195503235,
+      "learning_rate": 0.00018736373748609567,
+      "loss": 4.1371,
+      "step": 290
+    },
+    {
+      "epoch": 3.882943143812709,
+      "grad_norm": 0.9399815201759338,
+      "learning_rate": 0.00018731924360400445,
+      "loss": 4.5237,
+      "step": 291
+    },
+    {
+      "epoch": 3.8963210702341136,
+      "grad_norm": 0.8061622977256775,
+      "learning_rate": 0.00018727474972191326,
+      "loss": 4.1033,
+      "step": 292
+    },
+    {
+      "epoch": 3.9096989966555182,
+      "grad_norm": 0.7987121343612671,
+      "learning_rate": 0.00018723025583982205,
+      "loss": 3.9311,
+      "step": 293
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.9041138291358948,
+      "learning_rate": 0.00018718576195773083,
+      "loss": 4.0252,
+      "step": 294
+    },
+    {
+      "epoch": 3.936454849498328,
+      "grad_norm": 1.0002484321594238,
+      "learning_rate": 0.0001871412680756396,
+      "loss": 4.4605,
+      "step": 295
+    },
+    {
+      "epoch": 3.9498327759197327,
+      "grad_norm": 0.9991098046302795,
+      "learning_rate": 0.0001870967741935484,
+      "loss": 4.1528,
+      "step": 296
+    },
+    {
+      "epoch": 3.9632107023411374,
+      "grad_norm": 1.2179397344589233,
+      "learning_rate": 0.00018705228031145718,
+      "loss": 4.5224,
+      "step": 297
+    },
+    {
+      "epoch": 3.976588628762542,
+      "grad_norm": 0.8279774785041809,
+      "learning_rate": 0.00018700778642936596,
+      "loss": 3.9464,
+      "step": 298
+    },
+    {
+      "epoch": 3.9899665551839467,
+      "grad_norm": 0.8012803792953491,
+      "learning_rate": 0.00018696329254727477,
+      "loss": 4.0139,
+      "step": 299
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9700272083282471,
+      "learning_rate": 0.00018691879866518355,
+      "loss": 3.8306,
+      "step": 300
+    },
+    {
+      "epoch": 4.013377926421405,
+      "grad_norm": 0.7136749625205994,
+      "learning_rate": 0.00018687430478309233,
+      "loss": 3.9253,
+      "step": 301
+    },
+    {
+      "epoch": 4.026755852842809,
+      "grad_norm": 0.7885096669197083,
+      "learning_rate": 0.00018682981090100111,
+      "loss": 3.927,
+      "step": 302
+    },
+    {
+      "epoch": 4.040133779264214,
+      "grad_norm": 0.7801666855812073,
+      "learning_rate": 0.00018678531701890992,
+      "loss": 3.6482,
+      "step": 303
+    },
+    {
+      "epoch": 4.053511705685619,
+      "grad_norm": 0.7857955098152161,
+      "learning_rate": 0.00018674082313681868,
+      "loss": 4.0665,
+      "step": 304
+    },
+    {
+      "epoch": 4.066889632107023,
+      "grad_norm": 0.707421064376831,
+      "learning_rate": 0.0001866963292547275,
+      "loss": 3.9142,
+      "step": 305
+    },
+    {
+      "epoch": 4.080267558528428,
+      "grad_norm": 0.7936912775039673,
+      "learning_rate": 0.00018665183537263627,
+      "loss": 4.1227,
+      "step": 306
+    },
+    {
+      "epoch": 4.093645484949833,
+      "grad_norm": 0.8899754881858826,
+      "learning_rate": 0.00018660734149054505,
+      "loss": 3.7661,
+      "step": 307
+    },
+    {
+      "epoch": 4.107023411371237,
+      "grad_norm": 0.7760347723960876,
+      "learning_rate": 0.00018656284760845384,
+      "loss": 3.8921,
+      "step": 308
+    },
+    {
+      "epoch": 4.120401337792642,
+      "grad_norm": 0.8672968745231628,
+      "learning_rate": 0.00018651835372636265,
+      "loss": 3.6037,
+      "step": 309
+    },
+    {
+      "epoch": 4.133779264214047,
+      "grad_norm": 0.8046863675117493,
+      "learning_rate": 0.0001864738598442714,
+      "loss": 3.9117,
+      "step": 310
+    },
+    {
+      "epoch": 4.147157190635451,
+      "grad_norm": 0.9172897934913635,
+      "learning_rate": 0.0001864293659621802,
+      "loss": 3.7229,
+      "step": 311
+    },
+    {
+      "epoch": 4.160535117056856,
+      "grad_norm": 0.9616653919219971,
+      "learning_rate": 0.000186384872080089,
+      "loss": 3.8851,
+      "step": 312
+    },
+    {
+      "epoch": 4.173913043478261,
+      "grad_norm": 0.9659278988838196,
+      "learning_rate": 0.0001863403781979978,
+      "loss": 4.005,
+      "step": 313
+    },
+    {
+      "epoch": 4.187290969899665,
+      "grad_norm": 0.9171205163002014,
+      "learning_rate": 0.00018629588431590656,
+      "loss": 3.8634,
+      "step": 314
+    },
+    {
+      "epoch": 4.20066889632107,
+      "grad_norm": 0.9968683123588562,
+      "learning_rate": 0.00018625139043381537,
+      "loss": 3.7321,
+      "step": 315
+    },
+    {
+      "epoch": 4.214046822742475,
+      "grad_norm": 0.8762083053588867,
+      "learning_rate": 0.00018620689655172415,
+      "loss": 3.931,
+      "step": 316
+    },
+    {
+      "epoch": 4.2274247491638794,
+      "grad_norm": 0.9815887212753296,
+      "learning_rate": 0.00018616240266963293,
+      "loss": 3.9975,
+      "step": 317
+    },
+    {
+      "epoch": 4.240802675585284,
+      "grad_norm": 1.0065505504608154,
+      "learning_rate": 0.00018611790878754171,
+      "loss": 3.8364,
+      "step": 318
+    },
+    {
+      "epoch": 4.254180602006689,
+      "grad_norm": 0.9785431623458862,
+      "learning_rate": 0.00018607341490545052,
+      "loss": 3.8822,
+      "step": 319
+    },
+    {
+      "epoch": 4.2675585284280935,
+      "grad_norm": 1.077799677848816,
+      "learning_rate": 0.00018602892102335928,
+      "loss": 3.8299,
+      "step": 320
+    },
+    {
+      "epoch": 4.280936454849498,
+      "grad_norm": 0.8109619617462158,
+      "learning_rate": 0.0001859844271412681,
+      "loss": 3.8096,
+      "step": 321
+    },
+    {
+      "epoch": 4.294314381270903,
+      "grad_norm": 0.967856764793396,
+      "learning_rate": 0.00018593993325917687,
+      "loss": 3.8639,
+      "step": 322
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.8657905459403992,
+      "learning_rate": 0.00018589543937708568,
+      "loss": 3.7556,
+      "step": 323
+    },
+    {
+      "epoch": 4.321070234113712,
+      "grad_norm": 0.9641517400741577,
+      "learning_rate": 0.00018585094549499444,
+      "loss": 3.9702,
+      "step": 324
+    },
+    {
+      "epoch": 4.334448160535117,
+      "grad_norm": 0.9664435982704163,
+      "learning_rate": 0.00018580645161290325,
+      "loss": 3.8754,
+      "step": 325
+    },
+    {
+      "epoch": 4.3478260869565215,
+      "grad_norm": 0.8322617411613464,
+      "learning_rate": 0.00018576195773081203,
+      "loss": 3.83,
+      "step": 326
+    },
+    {
+      "epoch": 4.361204013377926,
+      "grad_norm": 1.0363450050354004,
+      "learning_rate": 0.0001857174638487208,
+      "loss": 3.9825,
+      "step": 327
+    },
+    {
+      "epoch": 4.374581939799331,
+      "grad_norm": 1.0125840902328491,
+      "learning_rate": 0.0001856729699666296,
+      "loss": 3.6525,
+      "step": 328
+    },
+    {
+      "epoch": 4.3879598662207355,
+      "grad_norm": 0.9922601580619812,
+      "learning_rate": 0.0001856284760845384,
+      "loss": 4.2373,
+      "step": 329
+    },
+    {
+      "epoch": 4.40133779264214,
+      "grad_norm": 0.9070426225662231,
+      "learning_rate": 0.00018558398220244716,
+      "loss": 3.9623,
+      "step": 330
+    },
+    {
+      "epoch": 4.414715719063545,
+      "grad_norm": 0.9369637370109558,
+      "learning_rate": 0.00018553948832035597,
+      "loss": 3.9297,
+      "step": 331
+    },
+    {
+      "epoch": 4.4280936454849495,
+      "grad_norm": 1.108876347541809,
+      "learning_rate": 0.00018549499443826475,
+      "loss": 3.7325,
+      "step": 332
+    },
+    {
+      "epoch": 4.441471571906354,
+      "grad_norm": 0.9405660629272461,
+      "learning_rate": 0.00018545050055617356,
+      "loss": 3.8615,
+      "step": 333
+    },
+    {
+      "epoch": 4.454849498327759,
+      "grad_norm": 0.9730128645896912,
+      "learning_rate": 0.00018540600667408231,
+      "loss": 4.1794,
+      "step": 334
+    },
+    {
+      "epoch": 4.468227424749164,
+      "grad_norm": 0.9341335892677307,
+      "learning_rate": 0.00018536151279199112,
+      "loss": 3.9422,
+      "step": 335
+    },
+    {
+      "epoch": 4.481605351170568,
+      "grad_norm": 0.9262625575065613,
+      "learning_rate": 0.0001853170189098999,
+      "loss": 3.9819,
+      "step": 336
+    },
+    {
+      "epoch": 4.494983277591973,
+      "grad_norm": 1.0419141054153442,
+      "learning_rate": 0.00018527252502780866,
+      "loss": 3.7481,
+      "step": 337
+    },
+    {
+      "epoch": 4.508361204013378,
+      "grad_norm": 0.8986826539039612,
+      "learning_rate": 0.00018522803114571747,
+      "loss": 4.1195,
+      "step": 338
+    },
+    {
+      "epoch": 4.521739130434782,
+      "grad_norm": 0.9502431154251099,
+      "learning_rate": 0.00018518353726362625,
+      "loss": 3.8521,
+      "step": 339
+    },
+    {
+      "epoch": 4.535117056856187,
+      "grad_norm": 0.8936267495155334,
+      "learning_rate": 0.00018513904338153504,
+      "loss": 3.647,
+      "step": 340
+    },
+    {
+      "epoch": 4.548494983277592,
+      "grad_norm": 0.8870158195495605,
+      "learning_rate": 0.00018509454949944382,
+      "loss": 3.7624,
+      "step": 341
+    },
+    {
+      "epoch": 4.561872909698996,
+      "grad_norm": 0.9030978679656982,
+      "learning_rate": 0.00018505005561735263,
+      "loss": 3.8018,
+      "step": 342
+    },
+    {
+      "epoch": 4.575250836120401,
+      "grad_norm": 0.8690946698188782,
+      "learning_rate": 0.0001850055617352614,
+      "loss": 3.6183,
+      "step": 343
+    },
+    {
+      "epoch": 4.588628762541806,
+      "grad_norm": 0.9812071323394775,
+      "learning_rate": 0.0001849610678531702,
+      "loss": 4.1231,
+      "step": 344
+    },
+    {
+      "epoch": 4.602006688963211,
+      "grad_norm": 0.9404383301734924,
+      "learning_rate": 0.00018491657397107897,
+      "loss": 3.9645,
+      "step": 345
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 1.0422123670578003,
+      "learning_rate": 0.00018487208008898778,
+      "loss": 3.9031,
+      "step": 346
+    },
+    {
+      "epoch": 4.6287625418060205,
+      "grad_norm": 0.9838129281997681,
+      "learning_rate": 0.00018482758620689654,
+      "loss": 3.9985,
+      "step": 347
+    },
+    {
+      "epoch": 4.642140468227424,
+      "grad_norm": 0.9232532978057861,
+      "learning_rate": 0.00018478309232480535,
+      "loss": 4.0343,
+      "step": 348
+    },
+    {
+      "epoch": 4.65551839464883,
+      "grad_norm": 0.9242956042289734,
+      "learning_rate": 0.00018473859844271413,
+      "loss": 4.0669,
+      "step": 349
+    },
+    {
+      "epoch": 4.668896321070234,
+      "grad_norm": 0.919269859790802,
+      "learning_rate": 0.0001846941045606229,
+      "loss": 4.0549,
+      "step": 350
+    },
+    {
+      "epoch": 4.682274247491639,
+      "grad_norm": 0.93565833568573,
+      "learning_rate": 0.0001846496106785317,
+      "loss": 4.1306,
+      "step": 351
+    },
+    {
+      "epoch": 4.695652173913043,
+      "grad_norm": 0.9001899361610413,
+      "learning_rate": 0.0001846051167964405,
+      "loss": 3.8916,
+      "step": 352
+    },
+    {
+      "epoch": 4.709030100334449,
+      "grad_norm": 0.8896821737289429,
+      "learning_rate": 0.0001845606229143493,
+      "loss": 3.8377,
+      "step": 353
+    },
+    {
+      "epoch": 4.722408026755852,
+      "grad_norm": 1.0137807130813599,
+      "learning_rate": 0.00018451612903225807,
+      "loss": 3.9923,
+      "step": 354
+    },
+    {
+      "epoch": 4.735785953177258,
+      "grad_norm": 1.075823426246643,
+      "learning_rate": 0.00018447163515016685,
+      "loss": 4.0706,
+      "step": 355
+    },
+    {
+      "epoch": 4.749163879598662,
+      "grad_norm": 1.0076895952224731,
+      "learning_rate": 0.00018442714126807566,
+      "loss": 4.0759,
+      "step": 356
+    },
+    {
+      "epoch": 4.762541806020067,
+      "grad_norm": 0.9387428164482117,
+      "learning_rate": 0.00018438264738598442,
+      "loss": 3.6959,
+      "step": 357
+    },
+    {
+      "epoch": 4.775919732441472,
+      "grad_norm": 0.8920648097991943,
+      "learning_rate": 0.00018433815350389323,
+      "loss": 3.9213,
+      "step": 358
+    },
+    {
+      "epoch": 4.789297658862877,
+      "grad_norm": 1.0252491235733032,
+      "learning_rate": 0.000184293659621802,
+      "loss": 3.9118,
+      "step": 359
+    },
+    {
+      "epoch": 4.802675585284281,
+      "grad_norm": 1.0382707118988037,
+      "learning_rate": 0.0001842491657397108,
+      "loss": 4.0172,
+      "step": 360
+    },
+    {
+      "epoch": 4.816053511705686,
+      "grad_norm": 1.07838773727417,
+      "learning_rate": 0.00018420467185761957,
+      "loss": 3.8531,
+      "step": 361
+    },
+    {
+      "epoch": 4.829431438127091,
+      "grad_norm": 0.9974546432495117,
+      "learning_rate": 0.00018416017797552838,
+      "loss": 4.0387,
+      "step": 362
+    },
+    {
+      "epoch": 4.842809364548495,
+      "grad_norm": 1.024491548538208,
+      "learning_rate": 0.00018411568409343717,
+      "loss": 3.9504,
+      "step": 363
+    },
+    {
+      "epoch": 4.8561872909699,
+      "grad_norm": 0.9236369132995605,
+      "learning_rate": 0.00018407119021134595,
+      "loss": 3.7119,
+      "step": 364
+    },
+    {
+      "epoch": 4.869565217391305,
+      "grad_norm": 0.935644268989563,
+      "learning_rate": 0.00018402669632925473,
+      "loss": 4.0077,
+      "step": 365
+    },
+    {
+      "epoch": 4.882943143812709,
+      "grad_norm": 0.9328681230545044,
+      "learning_rate": 0.00018398220244716354,
+      "loss": 3.9133,
+      "step": 366
+    },
+    {
+      "epoch": 4.896321070234114,
+      "grad_norm": 0.9596607089042664,
+      "learning_rate": 0.0001839377085650723,
+      "loss": 3.8003,
+      "step": 367
+    },
+    {
+      "epoch": 4.909698996655519,
+      "grad_norm": 0.9878052473068237,
+      "learning_rate": 0.0001838932146829811,
+      "loss": 3.8805,
+      "step": 368
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 1.00381600856781,
+      "learning_rate": 0.0001838487208008899,
+      "loss": 4.0264,
+      "step": 369
+    },
+    {
+      "epoch": 4.936454849498328,
+      "grad_norm": 1.024754524230957,
+      "learning_rate": 0.00018380422691879867,
+      "loss": 3.7291,
+      "step": 370
+    },
+    {
+      "epoch": 4.949832775919733,
+      "grad_norm": 0.9670823812484741,
+      "learning_rate": 0.00018375973303670745,
+      "loss": 3.9418,
+      "step": 371
+    },
+    {
+      "epoch": 4.963210702341137,
+      "grad_norm": 0.9736581444740295,
+      "learning_rate": 0.00018371523915461626,
+      "loss": 3.8813,
+      "step": 372
+    },
+    {
+      "epoch": 4.976588628762542,
+      "grad_norm": 0.9752672910690308,
+      "learning_rate": 0.00018367074527252504,
+      "loss": 3.6717,
+      "step": 373
+    },
+    {
+      "epoch": 4.989966555183947,
+      "grad_norm": 1.1268304586410522,
+      "learning_rate": 0.00018362625139043383,
+      "loss": 3.9782,
+      "step": 374
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.7933701276779175,
+      "learning_rate": 0.0001835817575083426,
+      "loss": 3.001,
+      "step": 375
+    },
+    {
+      "epoch": 5.013377926421405,
+      "grad_norm": 0.8035010099411011,
+      "learning_rate": 0.00018353726362625142,
+      "loss": 3.7943,
+      "step": 376
+    },
+    {
+      "epoch": 5.026755852842809,
+      "grad_norm": 0.8016420006752014,
+      "learning_rate": 0.00018349276974416017,
+      "loss": 3.7454,
+      "step": 377
+    },
+    {
+      "epoch": 5.040133779264214,
+      "grad_norm": 0.6844643354415894,
+      "learning_rate": 0.00018344827586206896,
+      "loss": 3.699,
+      "step": 378
+    },
+    {
+      "epoch": 5.053511705685619,
+      "grad_norm": 0.8649943470954895,
+      "learning_rate": 0.00018340378197997777,
+      "loss": 3.7197,
+      "step": 379
+    },
+    {
+      "epoch": 5.066889632107023,
+      "grad_norm": 0.9685015678405762,
+      "learning_rate": 0.00018335928809788655,
+      "loss": 3.6952,
+      "step": 380
+    },
+    {
+      "epoch": 5.080267558528428,
+      "grad_norm": 0.8728330135345459,
+      "learning_rate": 0.00018331479421579533,
+      "loss": 3.7164,
+      "step": 381
+    },
+    {
+      "epoch": 5.093645484949833,
+      "grad_norm": 0.962504506111145,
+      "learning_rate": 0.0001832703003337041,
+      "loss": 3.6123,
+      "step": 382
+    },
+    {
+      "epoch": 5.107023411371237,
+      "grad_norm": 0.9194462895393372,
+      "learning_rate": 0.00018322580645161292,
+      "loss": 3.275,
+      "step": 383
+    },
+    {
+      "epoch": 5.120401337792642,
+      "grad_norm": 0.9851329326629639,
+      "learning_rate": 0.00018318131256952168,
+      "loss": 3.6222,
+      "step": 384
+    },
+    {
+      "epoch": 5.133779264214047,
+      "grad_norm": 1.0702580213546753,
+      "learning_rate": 0.0001831368186874305,
+      "loss": 3.8728,
+      "step": 385
+    },
+    {
+      "epoch": 5.147157190635451,
+      "grad_norm": 1.3237228393554688,
+      "learning_rate": 0.00018309232480533927,
+      "loss": 3.8948,
+      "step": 386
+    },
+    {
+      "epoch": 5.160535117056856,
+      "grad_norm": 1.0076218843460083,
+      "learning_rate": 0.00018304783092324805,
+      "loss": 3.8894,
+      "step": 387
+    },
+    {
+      "epoch": 5.173913043478261,
+      "grad_norm": 1.084722876548767,
+      "learning_rate": 0.00018300333704115683,
+      "loss": 3.7398,
+      "step": 388
+    },
+    {
+      "epoch": 5.187290969899665,
+      "grad_norm": 0.9112711548805237,
+      "learning_rate": 0.00018295884315906564,
+      "loss": 3.5901,
+      "step": 389
+    },
+    {
+      "epoch": 5.20066889632107,
+      "grad_norm": 0.9451406002044678,
+      "learning_rate": 0.00018291434927697443,
+      "loss": 3.6313,
+      "step": 390
+    },
+    {
+      "epoch": 5.214046822742475,
+      "grad_norm": 0.8901047706604004,
+      "learning_rate": 0.0001828698553948832,
+      "loss": 3.3191,
+      "step": 391
+    },
+    {
+      "epoch": 5.2274247491638794,
+      "grad_norm": 0.9838565587997437,
+      "learning_rate": 0.000182825361512792,
+      "loss": 3.882,
+      "step": 392
+    },
+    {
+      "epoch": 5.240802675585284,
+      "grad_norm": 0.9839156866073608,
+      "learning_rate": 0.0001827808676307008,
+      "loss": 3.6068,
+      "step": 393
+    },
+    {
+      "epoch": 5.254180602006689,
+      "grad_norm": 0.9328583478927612,
+      "learning_rate": 0.00018273637374860956,
+      "loss": 3.6856,
+      "step": 394
+    },
+    {
+      "epoch": 5.2675585284280935,
+      "grad_norm": 0.8705796003341675,
+      "learning_rate": 0.00018269187986651837,
+      "loss": 3.7282,
+      "step": 395
+    },
+    {
+      "epoch": 5.280936454849498,
+      "grad_norm": 0.9675374031066895,
+      "learning_rate": 0.00018264738598442715,
+      "loss": 3.6588,
+      "step": 396
+    },
+    {
+      "epoch": 5.294314381270903,
+      "grad_norm": 1.145280361175537,
+      "learning_rate": 0.00018260289210233593,
+      "loss": 3.8843,
+      "step": 397
+    },
+    {
+      "epoch": 5.3076923076923075,
+      "grad_norm": 0.9769694805145264,
+      "learning_rate": 0.0001825583982202447,
+      "loss": 3.7207,
+      "step": 398
+    },
+    {
+      "epoch": 5.321070234113712,
+      "grad_norm": 0.9277816414833069,
+      "learning_rate": 0.00018251390433815352,
+      "loss": 3.712,
+      "step": 399
+    },
+    {
+      "epoch": 5.334448160535117,
+      "grad_norm": 1.1015180349349976,
+      "learning_rate": 0.0001824694104560623,
+      "loss": 3.7941,
+      "step": 400
+    },
+    {
+      "epoch": 5.3478260869565215,
+      "grad_norm": 1.2234200239181519,
+      "learning_rate": 0.0001824249165739711,
+      "loss": 3.6559,
+      "step": 401
+    },
+    {
+      "epoch": 5.361204013377926,
+      "grad_norm": 0.9358471035957336,
+      "learning_rate": 0.00018238042269187987,
+      "loss": 3.7665,
+      "step": 402
+    },
+    {
+      "epoch": 5.374581939799331,
+      "grad_norm": 0.8287034630775452,
+      "learning_rate": 0.00018233592880978868,
+      "loss": 3.8265,
+      "step": 403
+    },
+    {
+      "epoch": 5.3879598662207355,
+      "grad_norm": 1.0219204425811768,
+      "learning_rate": 0.00018229143492769743,
+      "loss": 3.6829,
+      "step": 404
+    },
+    {
+      "epoch": 5.40133779264214,
+      "grad_norm": 1.0601041316986084,
+      "learning_rate": 0.00018224694104560624,
+      "loss": 3.5879,
+      "step": 405
+    },
+    {
+      "epoch": 5.414715719063545,
+      "grad_norm": 1.2221566438674927,
+      "learning_rate": 0.00018220244716351503,
+      "loss": 3.6023,
+      "step": 406
+    },
+    {
+      "epoch": 5.4280936454849495,
+      "grad_norm": 0.9589087963104248,
+      "learning_rate": 0.0001821579532814238,
+      "loss": 3.9109,
+      "step": 407
+    },
+    {
+      "epoch": 5.441471571906354,
+      "grad_norm": 1.088295340538025,
+      "learning_rate": 0.0001821134593993326,
+      "loss": 3.5695,
+      "step": 408
+    },
+    {
+      "epoch": 5.454849498327759,
+      "grad_norm": 1.1284915208816528,
+      "learning_rate": 0.0001820689655172414,
+      "loss": 3.6004,
+      "step": 409
+    },
+    {
+      "epoch": 5.468227424749164,
+      "grad_norm": 1.0108689069747925,
+      "learning_rate": 0.00018202447163515018,
+      "loss": 3.7797,
+      "step": 410
+    },
+    {
+      "epoch": 5.481605351170568,
+      "grad_norm": 0.8550918102264404,
+      "learning_rate": 0.00018197997775305896,
+      "loss": 3.5758,
+      "step": 411
+    },
+    {
+      "epoch": 5.494983277591973,
+      "grad_norm": 0.8765145540237427,
+      "learning_rate": 0.00018193548387096775,
+      "loss": 3.7625,
+      "step": 412
+    },
+    {
+      "epoch": 5.508361204013378,
+      "grad_norm": 1.0253541469573975,
+      "learning_rate": 0.00018189098998887656,
+      "loss": 3.6033,
+      "step": 413
+    },
+    {
+      "epoch": 5.521739130434782,
+      "grad_norm": 1.0475622415542603,
+      "learning_rate": 0.0001818464961067853,
+      "loss": 3.813,
+      "step": 414
+    },
+    {
+      "epoch": 5.535117056856187,
+      "grad_norm": 1.053133249282837,
+      "learning_rate": 0.00018180200222469412,
+      "loss": 3.4779,
+      "step": 415
+    },
+    {
+      "epoch": 5.548494983277592,
+      "grad_norm": 1.0151216983795166,
+      "learning_rate": 0.0001817575083426029,
+      "loss": 3.9038,
+      "step": 416
+    },
+    {
+      "epoch": 5.561872909698996,
+      "grad_norm": 1.4666434526443481,
+      "learning_rate": 0.00018171301446051169,
+      "loss": 3.4735,
+      "step": 417
+    },
+    {
+      "epoch": 5.575250836120401,
+      "grad_norm": 1.1043344736099243,
+      "learning_rate": 0.00018166852057842047,
+      "loss": 3.7449,
+      "step": 418
+    },
+    {
+      "epoch": 5.588628762541806,
+      "grad_norm": 0.900745153427124,
+      "learning_rate": 0.00018162402669632925,
+      "loss": 3.7401,
+      "step": 419
+    },
+    {
+      "epoch": 5.602006688963211,
+      "grad_norm": 0.9771101474761963,
+      "learning_rate": 0.00018157953281423806,
+      "loss": 3.8328,
+      "step": 420
+    },
+    {
+      "epoch": 5.615384615384615,
+      "grad_norm": 0.9099516272544861,
+      "learning_rate": 0.00018153503893214682,
+      "loss": 3.6245,
+      "step": 421
+    },
+    {
+      "epoch": 5.6287625418060205,
+      "grad_norm": 0.9844585657119751,
+      "learning_rate": 0.00018149054505005563,
+      "loss": 3.5776,
+      "step": 422
+    },
+    {
+      "epoch": 5.642140468227424,
+      "grad_norm": 1.0481154918670654,
+      "learning_rate": 0.0001814460511679644,
+      "loss": 3.6304,
+      "step": 423
+    },
+    {
+      "epoch": 5.65551839464883,
+      "grad_norm": 0.9971081614494324,
+      "learning_rate": 0.0001814015572858732,
+      "loss": 3.7863,
+      "step": 424
+    },
+    {
+      "epoch": 5.668896321070234,
+      "grad_norm": 0.9247872829437256,
+      "learning_rate": 0.00018135706340378197,
+      "loss": 3.7319,
+      "step": 425
+    },
+    {
+      "epoch": 5.682274247491639,
+      "grad_norm": 0.9895725846290588,
+      "learning_rate": 0.00018131256952169078,
+      "loss": 3.78,
+      "step": 426
+    },
+    {
+      "epoch": 5.695652173913043,
+      "grad_norm": 1.0847641229629517,
+      "learning_rate": 0.00018126807563959956,
+      "loss": 3.8662,
+      "step": 427
+    },
+    {
+      "epoch": 5.709030100334449,
+      "grad_norm": 0.986259937286377,
+      "learning_rate": 0.00018122358175750835,
+      "loss": 3.5621,
+      "step": 428
+    },
+    {
+      "epoch": 5.722408026755852,
+      "grad_norm": 0.9166681170463562,
+      "learning_rate": 0.00018117908787541713,
+      "loss": 3.6153,
+      "step": 429
+    },
+    {
+      "epoch": 5.735785953177258,
+      "grad_norm": 1.1331177949905396,
+      "learning_rate": 0.00018113459399332594,
+      "loss": 3.5976,
+      "step": 430
+    },
+    {
+      "epoch": 5.749163879598662,
+      "grad_norm": 0.8743540644645691,
+      "learning_rate": 0.0001810901001112347,
+      "loss": 3.2511,
+      "step": 431
+    },
+    {
+      "epoch": 5.762541806020067,
+      "grad_norm": 1.0700207948684692,
+      "learning_rate": 0.0001810456062291435,
+      "loss": 3.7634,
+      "step": 432
+    },
+    {
+      "epoch": 5.775919732441472,
+      "grad_norm": 0.9412694573402405,
+      "learning_rate": 0.00018100111234705229,
+      "loss": 3.6264,
+      "step": 433
+    },
+    {
+      "epoch": 5.789297658862877,
+      "grad_norm": 1.0398496389389038,
+      "learning_rate": 0.00018095661846496107,
+      "loss": 3.744,
+      "step": 434
+    },
+    {
+      "epoch": 5.802675585284281,
+      "grad_norm": 0.9605004787445068,
+      "learning_rate": 0.00018091212458286985,
+      "loss": 3.5532,
+      "step": 435
+    },
+    {
+      "epoch": 5.816053511705686,
+      "grad_norm": 1.0449095964431763,
+      "learning_rate": 0.00018086763070077866,
+      "loss": 4.0611,
+      "step": 436
+    },
+    {
+      "epoch": 5.829431438127091,
+      "grad_norm": 0.9342606663703918,
+      "learning_rate": 0.00018082313681868744,
+      "loss": 3.9957,
+      "step": 437
+    },
+    {
+      "epoch": 5.842809364548495,
+      "grad_norm": 0.9687880873680115,
+      "learning_rate": 0.00018077864293659622,
+      "loss": 3.9299,
+      "step": 438
+    },
+    {
+      "epoch": 5.8561872909699,
+      "grad_norm": 1.1390576362609863,
+      "learning_rate": 0.000180734149054505,
+      "loss": 3.6552,
+      "step": 439
+    },
+    {
+      "epoch": 5.869565217391305,
+      "grad_norm": 0.9280988574028015,
+      "learning_rate": 0.00018068965517241382,
+      "loss": 3.7828,
+      "step": 440
+    },
+    {
+      "epoch": 5.882943143812709,
+      "grad_norm": 1.2928193807601929,
+      "learning_rate": 0.00018064516129032257,
+      "loss": 3.6292,
+      "step": 441
+    },
+    {
+      "epoch": 5.896321070234114,
+      "grad_norm": 1.0959875583648682,
+      "learning_rate": 0.00018060066740823138,
+      "loss": 3.4293,
+      "step": 442
+    },
+    {
+      "epoch": 5.909698996655519,
+      "grad_norm": 1.0713289976119995,
+      "learning_rate": 0.00018055617352614016,
+      "loss": 3.7767,
+      "step": 443
+    },
+    {
+      "epoch": 5.923076923076923,
+      "grad_norm": 0.9309440851211548,
+      "learning_rate": 0.00018051167964404895,
+      "loss": 3.5473,
+      "step": 444
+    },
+    {
+      "epoch": 5.936454849498328,
+      "grad_norm": 1.0999056100845337,
+      "learning_rate": 0.00018046718576195773,
+      "loss": 3.9694,
+      "step": 445
+    },
+    {
+      "epoch": 5.949832775919733,
+      "grad_norm": 1.1073781251907349,
+      "learning_rate": 0.00018042269187986654,
+      "loss": 3.6882,
+      "step": 446
+    },
+    {
+      "epoch": 5.963210702341137,
+      "grad_norm": 1.0430257320404053,
+      "learning_rate": 0.00018037819799777532,
+      "loss": 3.4009,
+      "step": 447
+    },
+    {
+      "epoch": 5.976588628762542,
+      "grad_norm": 1.1132690906524658,
+      "learning_rate": 0.0001803337041156841,
+      "loss": 3.8832,
+      "step": 448
+    },
+    {
+      "epoch": 5.989966555183947,
+      "grad_norm": 1.0147771835327148,
+      "learning_rate": 0.00018028921023359289,
+      "loss": 3.7117,
+      "step": 449
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.458959698677063,
+      "learning_rate": 0.0001802447163515017,
+      "loss": 3.5745,
+      "step": 450
+    },
+    {
+      "epoch": 6.013377926421405,
+      "grad_norm": 0.8363592028617859,
+      "learning_rate": 0.00018020022246941045,
+      "loss": 3.5835,
+      "step": 451
+    },
+    {
+      "epoch": 6.026755852842809,
+      "grad_norm": 0.8071937561035156,
+      "learning_rate": 0.00018015572858731926,
+      "loss": 3.5923,
+      "step": 452
+    },
+    {
+      "epoch": 6.040133779264214,
+      "grad_norm": 0.7746313214302063,
+      "learning_rate": 0.00018011123470522804,
+      "loss": 3.5688,
+      "step": 453
+    },
+    {
+      "epoch": 6.053511705685619,
+      "grad_norm": 0.689179539680481,
+      "learning_rate": 0.00018006674082313682,
+      "loss": 3.412,
+      "step": 454
+    },
+    {
+      "epoch": 6.066889632107023,
+      "grad_norm": 0.8438050746917725,
+      "learning_rate": 0.0001800222469410456,
+      "loss": 3.403,
+      "step": 455
+    },
+    {
+      "epoch": 6.080267558528428,
+      "grad_norm": 0.7670062780380249,
+      "learning_rate": 0.00017997775305895442,
+      "loss": 3.5029,
+      "step": 456
+    },
+    {
+      "epoch": 6.093645484949833,
+      "grad_norm": 0.8185870051383972,
+      "learning_rate": 0.0001799332591768632,
+      "loss": 3.4584,
+      "step": 457
+    },
+    {
+      "epoch": 6.107023411371237,
+      "grad_norm": 0.9618543386459351,
+      "learning_rate": 0.00017988876529477198,
+      "loss": 3.6538,
+      "step": 458
+    },
+    {
+      "epoch": 6.120401337792642,
+      "grad_norm": 0.959724485874176,
+      "learning_rate": 0.00017984427141268076,
+      "loss": 3.5284,
+      "step": 459
+    },
+    {
+      "epoch": 6.133779264214047,
+      "grad_norm": 0.8044765591621399,
+      "learning_rate": 0.00017979977753058955,
+      "loss": 3.2198,
+      "step": 460
+    },
+    {
+      "epoch": 6.147157190635451,
+      "grad_norm": 0.8287092447280884,
+      "learning_rate": 0.00017975528364849833,
+      "loss": 3.4977,
+      "step": 461
+    },
+    {
+      "epoch": 6.160535117056856,
+      "grad_norm": 0.8855329155921936,
+      "learning_rate": 0.0001797107897664071,
+      "loss": 3.5008,
+      "step": 462
+    },
+    {
+      "epoch": 6.173913043478261,
+      "grad_norm": 0.8839483857154846,
+      "learning_rate": 0.00017966629588431592,
+      "loss": 3.6135,
+      "step": 463
+    },
+    {
+      "epoch": 6.187290969899665,
+      "grad_norm": 0.963446319103241,
+      "learning_rate": 0.0001796218020022247,
+      "loss": 3.6156,
+      "step": 464
+    },
+    {
+      "epoch": 6.20066889632107,
+      "grad_norm": 0.896743655204773,
+      "learning_rate": 0.00017957730812013348,
+      "loss": 3.6623,
+      "step": 465
+    },
+    {
+      "epoch": 6.214046822742475,
+      "grad_norm": 0.9268617033958435,
+      "learning_rate": 0.00017953281423804227,
+      "loss": 3.4343,
+      "step": 466
+    },
+    {
+      "epoch": 6.2274247491638794,
+      "grad_norm": 0.8335449695587158,
+      "learning_rate": 0.00017948832035595108,
+      "loss": 3.5716,
+      "step": 467
+    },
+    {
+      "epoch": 6.240802675585284,
+      "grad_norm": 0.7771849036216736,
+      "learning_rate": 0.00017944382647385983,
+      "loss": 3.5191,
+      "step": 468
+    },
+    {
+      "epoch": 6.254180602006689,
+      "grad_norm": 0.9157488346099854,
+      "learning_rate": 0.00017939933259176864,
+      "loss": 3.5583,
+      "step": 469
+    },
+    {
+      "epoch": 6.2675585284280935,
+      "grad_norm": 0.9348477721214294,
+      "learning_rate": 0.00017935483870967742,
+      "loss": 3.3137,
+      "step": 470
+    },
+    {
+      "epoch": 6.280936454849498,
+      "grad_norm": 0.8791135549545288,
+      "learning_rate": 0.0001793103448275862,
+      "loss": 3.5111,
+      "step": 471
+    },
+    {
+      "epoch": 6.294314381270903,
+      "grad_norm": 0.9963672757148743,
+      "learning_rate": 0.000179265850945495,
+      "loss": 3.7518,
+      "step": 472
+    },
+    {
+      "epoch": 6.3076923076923075,
+      "grad_norm": 0.9291539192199707,
+      "learning_rate": 0.0001792213570634038,
+      "loss": 3.4524,
+      "step": 473
+    },
+    {
+      "epoch": 6.321070234113712,
+      "grad_norm": 0.9349279403686523,
+      "learning_rate": 0.00017917686318131258,
+      "loss": 3.4753,
+      "step": 474
+    },
+    {
+      "epoch": 6.334448160535117,
+      "grad_norm": 0.8984476327896118,
+      "learning_rate": 0.00017913236929922136,
+      "loss": 3.7325,
+      "step": 475
+    },
+    {
+      "epoch": 6.3478260869565215,
+      "grad_norm": 0.8452139496803284,
+      "learning_rate": 0.00017908787541713015,
+      "loss": 3.8021,
+      "step": 476
+    },
+    {
+      "epoch": 6.361204013377926,
+      "grad_norm": 0.9418376088142395,
+      "learning_rate": 0.00017904338153503895,
+      "loss": 3.7426,
+      "step": 477
+    },
+    {
+      "epoch": 6.374581939799331,
+      "grad_norm": 1.0661097764968872,
+      "learning_rate": 0.0001789988876529477,
+      "loss": 3.7556,
+      "step": 478
+    },
+    {
+      "epoch": 6.3879598662207355,
+      "grad_norm": 0.9645984768867493,
+      "learning_rate": 0.00017895439377085652,
+      "loss": 3.3353,
+      "step": 479
+    },
+    {
+      "epoch": 6.40133779264214,
+      "grad_norm": 0.9243470430374146,
+      "learning_rate": 0.0001789098998887653,
+      "loss": 3.5729,
+      "step": 480
+    },
+    {
+      "epoch": 6.414715719063545,
+      "grad_norm": 0.885061502456665,
+      "learning_rate": 0.00017886540600667408,
+      "loss": 3.5699,
+      "step": 481
+    },
+    {
+      "epoch": 6.4280936454849495,
+      "grad_norm": 0.9025402069091797,
+      "learning_rate": 0.00017882091212458287,
+      "loss": 3.4532,
+      "step": 482
+    },
+    {
+      "epoch": 6.441471571906354,
+      "grad_norm": 0.9760842323303223,
+      "learning_rate": 0.00017877641824249168,
+      "loss": 3.7222,
+      "step": 483
+    },
+    {
+      "epoch": 6.454849498327759,
+      "grad_norm": 1.2709609270095825,
+      "learning_rate": 0.00017873192436040046,
+      "loss": 3.756,
+      "step": 484
+    },
+    {
+      "epoch": 6.468227424749164,
+      "grad_norm": 0.904513955116272,
+      "learning_rate": 0.00017868743047830924,
+      "loss": 3.657,
+      "step": 485
+    },
+    {
+      "epoch": 6.481605351170568,
+      "grad_norm": 1.158915400505066,
+      "learning_rate": 0.00017864293659621802,
+      "loss": 3.5897,
+      "step": 486
+    },
+    {
+      "epoch": 6.494983277591973,
+      "grad_norm": 0.9457879066467285,
+      "learning_rate": 0.00017859844271412683,
+      "loss": 3.4394,
+      "step": 487
+    },
+    {
+      "epoch": 6.508361204013378,
+      "grad_norm": 1.04762601852417,
+      "learning_rate": 0.0001785539488320356,
+      "loss": 3.39,
+      "step": 488
+    },
+    {
+      "epoch": 6.521739130434782,
+      "grad_norm": 0.9370948076248169,
+      "learning_rate": 0.0001785094549499444,
+      "loss": 3.4816,
+      "step": 489
+    },
+    {
+      "epoch": 6.535117056856187,
+      "grad_norm": 1.1307988166809082,
+      "learning_rate": 0.00017846496106785318,
+      "loss": 3.5079,
+      "step": 490
+    },
+    {
+      "epoch": 6.548494983277592,
+      "grad_norm": 1.0025054216384888,
+      "learning_rate": 0.00017842046718576196,
+      "loss": 3.3447,
+      "step": 491
+    },
+    {
+      "epoch": 6.561872909698996,
+      "grad_norm": 0.9893412590026855,
+      "learning_rate": 0.00017837597330367074,
+      "loss": 3.4698,
+      "step": 492
+    },
+    {
+      "epoch": 6.575250836120401,
+      "grad_norm": 1.0476347208023071,
+      "learning_rate": 0.00017833147942157955,
+      "loss": 3.4955,
+      "step": 493
+    },
+    {
+      "epoch": 6.588628762541806,
+      "grad_norm": 1.0508891344070435,
+      "learning_rate": 0.00017828698553948834,
+      "loss": 3.6568,
+      "step": 494
+    },
+    {
+      "epoch": 6.602006688963211,
+      "grad_norm": 1.0397465229034424,
+      "learning_rate": 0.00017824249165739712,
+      "loss": 3.5087,
+      "step": 495
+    },
+    {
+      "epoch": 6.615384615384615,
+      "grad_norm": 0.9884181022644043,
+      "learning_rate": 0.0001781979977753059,
+      "loss": 3.6778,
+      "step": 496
+    },
+    {
+      "epoch": 6.6287625418060205,
+      "grad_norm": 1.1187562942504883,
+      "learning_rate": 0.0001781535038932147,
+      "loss": 3.4345,
+      "step": 497
+    },
+    {
+      "epoch": 6.642140468227424,
+      "grad_norm": 1.1133880615234375,
+      "learning_rate": 0.00017810901001112347,
+      "loss": 3.4959,
+      "step": 498
+    },
+    {
+      "epoch": 6.65551839464883,
+      "grad_norm": 0.9368589520454407,
+      "learning_rate": 0.00017806451612903228,
+      "loss": 3.628,
+      "step": 499
+    },
+    {
+      "epoch": 6.668896321070234,
+      "grad_norm": 1.0427212715148926,
+      "learning_rate": 0.00017802002224694106,
+      "loss": 3.6073,
+      "step": 500
+    },
+    {
+      "epoch": 6.682274247491639,
+      "grad_norm": 0.9564261436462402,
+      "learning_rate": 0.00017797552836484984,
+      "loss": 3.4406,
+      "step": 501
+    },
+    {
+      "epoch": 6.695652173913043,
+      "grad_norm": 0.9754629135131836,
+      "learning_rate": 0.00017793103448275862,
+      "loss": 3.7025,
+      "step": 502
+    },
+    {
+      "epoch": 6.709030100334449,
+      "grad_norm": 0.9669683575630188,
+      "learning_rate": 0.0001778865406006674,
+      "loss": 3.5872,
+      "step": 503
+    },
+    {
+      "epoch": 6.722408026755852,
+      "grad_norm": 0.959619402885437,
+      "learning_rate": 0.00017784204671857621,
+      "loss": 3.5124,
+      "step": 504
+    },
+    {
+      "epoch": 6.735785953177258,
+      "grad_norm": 0.981737494468689,
+      "learning_rate": 0.00017779755283648497,
+      "loss": 3.4038,
+      "step": 505
+    },
+    {
+      "epoch": 6.749163879598662,
+      "grad_norm": 0.9372640252113342,
+      "learning_rate": 0.00017775305895439378,
+      "loss": 3.6725,
+      "step": 506
+    },
+    {
+      "epoch": 6.762541806020067,
+      "grad_norm": 0.947066605091095,
+      "learning_rate": 0.00017770856507230256,
+      "loss": 3.807,
+      "step": 507
+    },
+    {
+      "epoch": 6.775919732441472,
+      "grad_norm": 0.7659755349159241,
+      "learning_rate": 0.00017766407119021134,
+      "loss": 3.0411,
+      "step": 508
+    },
+    {
+      "epoch": 6.789297658862877,
+      "grad_norm": 1.0432168245315552,
+      "learning_rate": 0.00017761957730812013,
+      "loss": 3.6259,
+      "step": 509
+    },
+    {
+      "epoch": 6.802675585284281,
+      "grad_norm": 1.0104693174362183,
+      "learning_rate": 0.00017757508342602894,
+      "loss": 3.5971,
+      "step": 510
+    },
+    {
+      "epoch": 6.816053511705686,
+      "grad_norm": 0.9007440805435181,
+      "learning_rate": 0.00017753058954393772,
+      "loss": 3.5894,
+      "step": 511
+    },
+    {
+      "epoch": 6.829431438127091,
+      "grad_norm": 0.8829946517944336,
+      "learning_rate": 0.0001774860956618465,
+      "loss": 3.6094,
+      "step": 512
+    },
+    {
+      "epoch": 6.842809364548495,
+      "grad_norm": 0.9823127388954163,
+      "learning_rate": 0.00017744160177975528,
+      "loss": 3.6568,
+      "step": 513
+    },
+    {
+      "epoch": 6.8561872909699,
+      "grad_norm": 1.0247899293899536,
+      "learning_rate": 0.0001773971078976641,
+      "loss": 3.259,
+      "step": 514
+    },
+    {
+      "epoch": 6.869565217391305,
+      "grad_norm": 0.9435336589813232,
+      "learning_rate": 0.00017735261401557285,
+      "loss": 3.4978,
+      "step": 515
+    },
+    {
+      "epoch": 6.882943143812709,
+      "grad_norm": 1.0135424137115479,
+      "learning_rate": 0.00017730812013348166,
+      "loss": 3.4765,
+      "step": 516
+    },
+    {
+      "epoch": 6.896321070234114,
+      "grad_norm": 1.1327738761901855,
+      "learning_rate": 0.00017726362625139044,
+      "loss": 3.6927,
+      "step": 517
+    },
+    {
+      "epoch": 6.909698996655519,
+      "grad_norm": 0.9335159659385681,
+      "learning_rate": 0.00017721913236929922,
+      "loss": 3.7278,
+      "step": 518
+    },
+    {
+      "epoch": 6.923076923076923,
+      "grad_norm": 0.8229056000709534,
+      "learning_rate": 0.000177174638487208,
+      "loss": 3.549,
+      "step": 519
+    },
+    {
+      "epoch": 6.936454849498328,
+      "grad_norm": 0.8996124267578125,
+      "learning_rate": 0.00017713014460511681,
+      "loss": 3.5093,
+      "step": 520
+    },
+    {
+      "epoch": 6.949832775919733,
+      "grad_norm": 0.9984102845191956,
+      "learning_rate": 0.0001770856507230256,
+      "loss": 3.4772,
+      "step": 521
+    },
+    {
+      "epoch": 6.963210702341137,
+      "grad_norm": 0.9136049747467041,
+      "learning_rate": 0.00017704115684093438,
+      "loss": 3.6487,
+      "step": 522
+    },
+    {
+      "epoch": 6.976588628762542,
+      "grad_norm": 0.9707899689674377,
+      "learning_rate": 0.00017699666295884316,
+      "loss": 3.7471,
+      "step": 523
+    },
+    {
+      "epoch": 6.989966555183947,
+      "grad_norm": 0.919865071773529,
+      "learning_rate": 0.00017695216907675197,
+      "loss": 3.8013,
+      "step": 524
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.125541090965271,
+      "learning_rate": 0.00017690767519466073,
+      "loss": 3.4869,
+      "step": 525
+    },
+    {
+      "epoch": 7.013377926421405,
+      "grad_norm": 0.8512810468673706,
+      "learning_rate": 0.00017686318131256954,
+      "loss": 3.3334,
+      "step": 526
+    },
+    {
+      "epoch": 7.026755852842809,
+      "grad_norm": 0.8450623750686646,
+      "learning_rate": 0.00017681868743047832,
+      "loss": 3.5769,
+      "step": 527
+    },
+    {
+      "epoch": 7.040133779264214,
+      "grad_norm": 0.8526298403739929,
+      "learning_rate": 0.0001767741935483871,
+      "loss": 3.4546,
+      "step": 528
+    },
+    {
+      "epoch": 7.053511705685619,
+      "grad_norm": 0.75905442237854,
+      "learning_rate": 0.00017672969966629588,
+      "loss": 3.4228,
+      "step": 529
+    },
+    {
+      "epoch": 7.066889632107023,
+      "grad_norm": 0.8442811965942383,
+      "learning_rate": 0.0001766852057842047,
+      "loss": 3.5766,
+      "step": 530
+    },
+    {
+      "epoch": 7.080267558528428,
+      "grad_norm": 0.9584814310073853,
+      "learning_rate": 0.00017664071190211347,
+      "loss": 3.5312,
+      "step": 531
+    },
+    {
+      "epoch": 7.093645484949833,
+      "grad_norm": 0.9741052985191345,
+      "learning_rate": 0.00017659621802002226,
+      "loss": 3.4877,
+      "step": 532
+    },
+    {
+      "epoch": 7.107023411371237,
+      "grad_norm": 0.8638135194778442,
+      "learning_rate": 0.00017655172413793104,
+      "loss": 3.5701,
+      "step": 533
+    },
+    {
+      "epoch": 7.120401337792642,
+      "grad_norm": 1.0128440856933594,
+      "learning_rate": 0.00017650723025583985,
+      "loss": 3.3629,
+      "step": 534
+    },
+    {
+      "epoch": 7.133779264214047,
+      "grad_norm": 0.9763593673706055,
+      "learning_rate": 0.0001764627363737486,
+      "loss": 3.5651,
+      "step": 535
+    },
+    {
+      "epoch": 7.147157190635451,
+      "grad_norm": 0.8706293702125549,
+      "learning_rate": 0.00017641824249165741,
+      "loss": 3.454,
+      "step": 536
+    },
+    {
+      "epoch": 7.160535117056856,
+      "grad_norm": 0.9227468967437744,
+      "learning_rate": 0.0001763737486095662,
+      "loss": 3.5528,
+      "step": 537
+    },
+    {
+      "epoch": 7.173913043478261,
+      "grad_norm": 0.7493206262588501,
+      "learning_rate": 0.00017632925472747498,
+      "loss": 3.2662,
+      "step": 538
+    },
+    {
+      "epoch": 7.187290969899665,
+      "grad_norm": 0.8414123058319092,
+      "learning_rate": 0.00017628476084538376,
+      "loss": 3.3864,
+      "step": 539
+    },
+    {
+      "epoch": 7.20066889632107,
+      "grad_norm": 0.8352764248847961,
+      "learning_rate": 0.00017624026696329257,
+      "loss": 3.2407,
+      "step": 540
+    },
+    {
+      "epoch": 7.214046822742475,
+      "grad_norm": 0.7413480281829834,
+      "learning_rate": 0.00017619577308120135,
+      "loss": 3.3989,
+      "step": 541
+    },
+    {
+      "epoch": 7.2274247491638794,
+      "grad_norm": 0.7661281228065491,
+      "learning_rate": 0.00017615127919911014,
+      "loss": 3.3792,
+      "step": 542
+    },
+    {
+      "epoch": 7.240802675585284,
+      "grad_norm": 0.86900395154953,
+      "learning_rate": 0.00017610678531701892,
+      "loss": 3.2191,
+      "step": 543
+    },
+    {
+      "epoch": 7.254180602006689,
+      "grad_norm": 0.8536344170570374,
+      "learning_rate": 0.0001760622914349277,
+      "loss": 3.3366,
+      "step": 544
+    },
+    {
+      "epoch": 7.2675585284280935,
+      "grad_norm": 0.8729544878005981,
+      "learning_rate": 0.00017601779755283648,
+      "loss": 3.4976,
+      "step": 545
+    },
+    {
+      "epoch": 7.280936454849498,
+      "grad_norm": 0.8263023495674133,
+      "learning_rate": 0.00017597330367074526,
+      "loss": 3.4102,
+      "step": 546
+    },
+    {
+      "epoch": 7.294314381270903,
+      "grad_norm": 0.748373806476593,
+      "learning_rate": 0.00017592880978865407,
+      "loss": 3.332,
+      "step": 547
+    },
+    {
+      "epoch": 7.3076923076923075,
+      "grad_norm": 0.7606791853904724,
+      "learning_rate": 0.00017588431590656286,
+      "loss": 3.361,
+      "step": 548
+    },
+    {
+      "epoch": 7.321070234113712,
+      "grad_norm": 0.9155070781707764,
+      "learning_rate": 0.00017583982202447164,
+      "loss": 3.657,
+      "step": 549
+    },
+    {
+      "epoch": 7.334448160535117,
+      "grad_norm": 0.7440597414970398,
+      "learning_rate": 0.00017579532814238042,
+      "loss": 3.335,
+      "step": 550
+    },
+    {
+      "epoch": 7.3478260869565215,
+      "grad_norm": 0.8781002759933472,
+      "learning_rate": 0.00017575083426028923,
+      "loss": 3.5579,
+      "step": 551
+    },
+    {
+      "epoch": 7.361204013377926,
+      "grad_norm": 0.7886701822280884,
+      "learning_rate": 0.00017570634037819799,
+      "loss": 3.4636,
+      "step": 552
+    },
+    {
+      "epoch": 7.374581939799331,
+      "grad_norm": 0.8931376934051514,
+      "learning_rate": 0.0001756618464961068,
+      "loss": 3.4375,
+      "step": 553
+    },
+    {
+      "epoch": 7.3879598662207355,
+      "grad_norm": 0.7599623799324036,
+      "learning_rate": 0.00017561735261401558,
+      "loss": 3.6551,
+      "step": 554
+    },
+    {
+      "epoch": 7.40133779264214,
+      "grad_norm": 0.7692762613296509,
+      "learning_rate": 0.00017557285873192436,
+      "loss": 3.6373,
+      "step": 555
+    },
+    {
+      "epoch": 7.414715719063545,
+      "grad_norm": 0.8861828446388245,
+      "learning_rate": 0.00017552836484983314,
+      "loss": 3.4791,
+      "step": 556
+    },
+    {
+      "epoch": 7.4280936454849495,
+      "grad_norm": 0.9560372829437256,
+      "learning_rate": 0.00017548387096774195,
+      "loss": 3.7291,
+      "step": 557
+    },
+    {
+      "epoch": 7.441471571906354,
+      "grad_norm": 0.8745344281196594,
+      "learning_rate": 0.00017543937708565073,
+      "loss": 3.3071,
+      "step": 558
+    },
+    {
+      "epoch": 7.454849498327759,
+      "grad_norm": 0.8178285360336304,
+      "learning_rate": 0.00017539488320355952,
+      "loss": 3.4738,
+      "step": 559
+    },
+    {
+      "epoch": 7.468227424749164,
+      "grad_norm": 0.8611259460449219,
+      "learning_rate": 0.0001753503893214683,
+      "loss": 3.25,
+      "step": 560
+    },
+    {
+      "epoch": 7.481605351170568,
+      "grad_norm": 0.8623505234718323,
+      "learning_rate": 0.0001753058954393771,
+      "loss": 3.3701,
+      "step": 561
+    },
+    {
+      "epoch": 7.494983277591973,
+      "grad_norm": 0.76930171251297,
+      "learning_rate": 0.00017526140155728586,
+      "loss": 3.332,
+      "step": 562
+    },
+    {
+      "epoch": 7.508361204013378,
+      "grad_norm": 0.8986758589744568,
+      "learning_rate": 0.00017521690767519467,
+      "loss": 3.3927,
+      "step": 563
+    },
+    {
+      "epoch": 7.521739130434782,
+      "grad_norm": 0.9844257831573486,
+      "learning_rate": 0.00017517241379310346,
+      "loss": 3.5664,
+      "step": 564
+    },
+    {
+      "epoch": 7.535117056856187,
+      "grad_norm": 0.983921229839325,
+      "learning_rate": 0.00017512791991101224,
+      "loss": 3.3888,
+      "step": 565
+    },
+    {
+      "epoch": 7.548494983277592,
+      "grad_norm": 0.8052308559417725,
+      "learning_rate": 0.00017508342602892102,
+      "loss": 3.4809,
+      "step": 566
+    },
+    {
+      "epoch": 7.561872909698996,
+      "grad_norm": 0.7996425032615662,
+      "learning_rate": 0.00017503893214682983,
+      "loss": 3.4793,
+      "step": 567
+    },
+    {
+      "epoch": 7.575250836120401,
+      "grad_norm": 0.8453391194343567,
+      "learning_rate": 0.0001749944382647386,
+      "loss": 3.3199,
+      "step": 568
+    },
+    {
+      "epoch": 7.588628762541806,
+      "grad_norm": 0.8720147013664246,
+      "learning_rate": 0.0001749499443826474,
+      "loss": 3.5612,
+      "step": 569
+    },
+    {
+      "epoch": 7.602006688963211,
+      "grad_norm": 0.9093672633171082,
+      "learning_rate": 0.00017490545050055618,
+      "loss": 3.0509,
+      "step": 570
+    },
+    {
+      "epoch": 7.615384615384615,
+      "grad_norm": 0.8936579823493958,
+      "learning_rate": 0.000174860956618465,
+      "loss": 3.4408,
+      "step": 571
+    },
+    {
+      "epoch": 7.6287625418060205,
+      "grad_norm": 0.7683162689208984,
+      "learning_rate": 0.00017481646273637374,
+      "loss": 3.3536,
+      "step": 572
+    },
+    {
+      "epoch": 7.642140468227424,
+      "grad_norm": 0.7943581342697144,
+      "learning_rate": 0.00017477196885428255,
+      "loss": 3.4542,
+      "step": 573
+    },
+    {
+      "epoch": 7.65551839464883,
+      "grad_norm": 0.8183353543281555,
+      "learning_rate": 0.00017472747497219133,
+      "loss": 3.603,
+      "step": 574
+    },
+    {
+      "epoch": 7.668896321070234,
+      "grad_norm": 0.7816463708877563,
+      "learning_rate": 0.00017468298109010012,
+      "loss": 3.7388,
+      "step": 575
+    },
+    {
+      "epoch": 7.682274247491639,
+      "grad_norm": 0.8167930245399475,
+      "learning_rate": 0.0001746384872080089,
+      "loss": 3.7743,
+      "step": 576
+    },
+    {
+      "epoch": 7.695652173913043,
+      "grad_norm": 0.832392156124115,
+      "learning_rate": 0.0001745939933259177,
+      "loss": 3.7488,
+      "step": 577
+    },
+    {
+      "epoch": 7.709030100334449,
+      "grad_norm": 0.9362333416938782,
+      "learning_rate": 0.0001745494994438265,
+      "loss": 3.5722,
+      "step": 578
+    },
+    {
+      "epoch": 7.722408026755852,
+      "grad_norm": 1.0247248411178589,
+      "learning_rate": 0.00017450500556173527,
+      "loss": 3.2048,
+      "step": 579
+    },
+    {
+      "epoch": 7.735785953177258,
+      "grad_norm": 0.8833767175674438,
+      "learning_rate": 0.00017446051167964406,
+      "loss": 3.389,
+      "step": 580
+    },
+    {
+      "epoch": 7.749163879598662,
+      "grad_norm": 0.8344758749008179,
+      "learning_rate": 0.00017441601779755287,
+      "loss": 3.5264,
+      "step": 581
+    },
+    {
+      "epoch": 7.762541806020067,
+      "grad_norm": 0.9771448373794556,
+      "learning_rate": 0.00017437152391546162,
+      "loss": 3.041,
+      "step": 582
+    },
+    {
+      "epoch": 7.775919732441472,
+      "grad_norm": 0.8279567956924438,
+      "learning_rate": 0.00017432703003337043,
+      "loss": 3.6273,
+      "step": 583
+    },
+    {
+      "epoch": 7.789297658862877,
+      "grad_norm": 0.957206130027771,
+      "learning_rate": 0.0001742825361512792,
+      "loss": 3.3831,
+      "step": 584
+    },
+    {
+      "epoch": 7.802675585284281,
+      "grad_norm": 0.860619843006134,
+      "learning_rate": 0.000174238042269188,
+      "loss": 3.4566,
+      "step": 585
+    },
+    {
+      "epoch": 7.816053511705686,
+      "grad_norm": 0.8725448846817017,
+      "learning_rate": 0.00017419354838709678,
+      "loss": 3.6594,
+      "step": 586
+    },
+    {
+      "epoch": 7.829431438127091,
+      "grad_norm": 0.8343111276626587,
+      "learning_rate": 0.00017414905450500556,
+      "loss": 3.3423,
+      "step": 587
+    },
+    {
+      "epoch": 7.842809364548495,
+      "grad_norm": 0.9043267965316772,
+      "learning_rate": 0.00017410456062291437,
+      "loss": 3.221,
+      "step": 588
+    },
+    {
+      "epoch": 7.8561872909699,
+      "grad_norm": 0.9563114643096924,
+      "learning_rate": 0.00017406006674082312,
+      "loss": 3.5143,
+      "step": 589
+    },
+    {
+      "epoch": 7.869565217391305,
+      "grad_norm": 0.9726302027702332,
+      "learning_rate": 0.00017401557285873193,
+      "loss": 3.4373,
+      "step": 590
+    },
+    {
+      "epoch": 7.882943143812709,
+      "grad_norm": 0.9203178882598877,
+      "learning_rate": 0.00017397107897664072,
+      "loss": 3.6014,
+      "step": 591
+    },
+    {
+      "epoch": 7.896321070234114,
+      "grad_norm": 0.9120233654975891,
+      "learning_rate": 0.0001739265850945495,
+      "loss": 3.1429,
+      "step": 592
+    },
+    {
+      "epoch": 7.909698996655519,
+      "grad_norm": 0.7576518058776855,
+      "learning_rate": 0.00017388209121245828,
+      "loss": 3.2065,
+      "step": 593
+    },
+    {
+      "epoch": 7.923076923076923,
+      "grad_norm": 0.9629240036010742,
+      "learning_rate": 0.0001738375973303671,
+      "loss": 3.4788,
+      "step": 594
+    },
+    {
+      "epoch": 7.936454849498328,
+      "grad_norm": 0.8390881419181824,
+      "learning_rate": 0.00017379310344827587,
+      "loss": 3.2857,
+      "step": 595
+    },
+    {
+      "epoch": 7.949832775919733,
+      "grad_norm": 0.8708979487419128,
+      "learning_rate": 0.00017374860956618466,
+      "loss": 3.3321,
+      "step": 596
+    },
+    {
+      "epoch": 7.963210702341137,
+      "grad_norm": 0.7076835632324219,
+      "learning_rate": 0.00017370411568409344,
+      "loss": 3.5905,
+      "step": 597
+    },
+    {
+      "epoch": 7.976588628762542,
+      "grad_norm": 1.016526222229004,
+      "learning_rate": 0.00017365962180200225,
+      "loss": 3.4053,
+      "step": 598
+    },
+    {
+      "epoch": 7.989966555183947,
+      "grad_norm": 0.7592278718948364,
+      "learning_rate": 0.000173615127919911,
+      "loss": 3.3968,
+      "step": 599
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.0106462240219116,
+      "learning_rate": 0.0001735706340378198,
+      "loss": 3.3722,
+      "step": 600
+    },
+    {
+      "epoch": 8.013377926421406,
+      "grad_norm": 0.740808367729187,
+      "learning_rate": 0.0001735261401557286,
+      "loss": 3.394,
+      "step": 601
+    },
+    {
+      "epoch": 8.02675585284281,
+      "grad_norm": 0.6732498407363892,
+      "learning_rate": 0.00017348164627363738,
+      "loss": 3.297,
+      "step": 602
+    },
+    {
+      "epoch": 8.040133779264215,
+      "grad_norm": 0.8319197297096252,
+      "learning_rate": 0.00017343715239154616,
+      "loss": 3.2898,
+      "step": 603
+    },
+    {
+      "epoch": 8.053511705685619,
+      "grad_norm": 0.7834349870681763,
+      "learning_rate": 0.00017339265850945497,
+      "loss": 3.2341,
+      "step": 604
+    },
+    {
+      "epoch": 8.066889632107024,
+      "grad_norm": 0.705737292766571,
+      "learning_rate": 0.00017334816462736375,
+      "loss": 3.3429,
+      "step": 605
+    },
+    {
+      "epoch": 8.080267558528428,
+      "grad_norm": 0.8270958065986633,
+      "learning_rate": 0.00017330367074527253,
+      "loss": 3.0458,
+      "step": 606
+    },
+    {
+      "epoch": 8.093645484949834,
+      "grad_norm": 0.7254801392555237,
+      "learning_rate": 0.00017325917686318132,
+      "loss": 3.5143,
+      "step": 607
+    },
+    {
+      "epoch": 8.107023411371237,
+      "grad_norm": 0.8450751900672913,
+      "learning_rate": 0.00017321468298109013,
+      "loss": 3.1507,
+      "step": 608
+    },
+    {
+      "epoch": 8.120401337792643,
+      "grad_norm": 0.7936837673187256,
+      "learning_rate": 0.00017317018909899888,
+      "loss": 3.3979,
+      "step": 609
+    },
+    {
+      "epoch": 8.133779264214047,
+      "grad_norm": 0.6496401429176331,
+      "learning_rate": 0.0001731256952169077,
+      "loss": 3.3613,
+      "step": 610
+    },
+    {
+      "epoch": 8.147157190635452,
+      "grad_norm": 0.8721235990524292,
+      "learning_rate": 0.00017308120133481647,
+      "loss": 3.4299,
+      "step": 611
+    },
+    {
+      "epoch": 8.160535117056856,
+      "grad_norm": 0.7671874761581421,
+      "learning_rate": 0.00017303670745272525,
+      "loss": 3.3333,
+      "step": 612
+    },
+    {
+      "epoch": 8.173913043478262,
+      "grad_norm": 0.6427144408226013,
+      "learning_rate": 0.00017299221357063404,
+      "loss": 3.3304,
+      "step": 613
+    },
+    {
+      "epoch": 8.187290969899665,
+      "grad_norm": 0.7999966144561768,
+      "learning_rate": 0.00017294771968854285,
+      "loss": 3.4016,
+      "step": 614
+    },
+    {
+      "epoch": 8.200668896321071,
+      "grad_norm": 0.8216206431388855,
+      "learning_rate": 0.00017290322580645163,
+      "loss": 3.1724,
+      "step": 615
+    },
+    {
+      "epoch": 8.214046822742475,
+      "grad_norm": 0.7364024519920349,
+      "learning_rate": 0.0001728587319243604,
+      "loss": 3.34,
+      "step": 616
+    },
+    {
+      "epoch": 8.22742474916388,
+      "grad_norm": 0.7688239812850952,
+      "learning_rate": 0.0001728142380422692,
+      "loss": 3.2114,
+      "step": 617
+    },
+    {
+      "epoch": 8.240802675585284,
+      "grad_norm": 0.8786870837211609,
+      "learning_rate": 0.000172769744160178,
+      "loss": 3.378,
+      "step": 618
+    },
+    {
+      "epoch": 8.25418060200669,
+      "grad_norm": 0.9048855900764465,
+      "learning_rate": 0.00017272525027808676,
+      "loss": 3.1801,
+      "step": 619
+    },
+    {
+      "epoch": 8.267558528428093,
+      "grad_norm": 0.657189130783081,
+      "learning_rate": 0.00017268075639599557,
+      "loss": 3.1389,
+      "step": 620
+    },
+    {
+      "epoch": 8.280936454849499,
+      "grad_norm": 0.8015987873077393,
+      "learning_rate": 0.00017263626251390435,
+      "loss": 3.4621,
+      "step": 621
+    },
+    {
+      "epoch": 8.294314381270903,
+      "grad_norm": 0.8232793807983398,
+      "learning_rate": 0.00017259176863181313,
+      "loss": 3.1763,
+      "step": 622
+    },
+    {
+      "epoch": 8.307692307692308,
+      "grad_norm": 0.7447130680084229,
+      "learning_rate": 0.00017254727474972192,
+      "loss": 3.1266,
+      "step": 623
+    },
+    {
+      "epoch": 8.321070234113712,
+      "grad_norm": 0.7649840116500854,
+      "learning_rate": 0.00017250278086763072,
+      "loss": 3.1959,
+      "step": 624
+    },
+    {
+      "epoch": 8.334448160535118,
+      "grad_norm": 0.7119699120521545,
+      "learning_rate": 0.0001724582869855395,
+      "loss": 3.5626,
+      "step": 625
+    },
+    {
+      "epoch": 8.347826086956522,
+      "grad_norm": 0.8238518834114075,
+      "learning_rate": 0.00017241379310344826,
+      "loss": 3.1873,
+      "step": 626
+    },
+    {
+      "epoch": 8.361204013377927,
+      "grad_norm": 0.8248497843742371,
+      "learning_rate": 0.00017236929922135707,
+      "loss": 3.5686,
+      "step": 627
+    },
+    {
+      "epoch": 8.37458193979933,
+      "grad_norm": 0.8704475164413452,
+      "learning_rate": 0.00017232480533926585,
+      "loss": 3.3195,
+      "step": 628
+    },
+    {
+      "epoch": 8.387959866220736,
+      "grad_norm": 0.8160929083824158,
+      "learning_rate": 0.00017228031145717464,
+      "loss": 3.3808,
+      "step": 629
+    },
+    {
+      "epoch": 8.40133779264214,
+      "grad_norm": 0.8537085652351379,
+      "learning_rate": 0.00017223581757508342,
+      "loss": 3.3638,
+      "step": 630
+    },
+    {
+      "epoch": 8.414715719063546,
+      "grad_norm": 0.876519501209259,
+      "learning_rate": 0.00017219132369299223,
+      "loss": 3.2019,
+      "step": 631
+    },
+    {
+      "epoch": 8.42809364548495,
+      "grad_norm": 0.6573703289031982,
+      "learning_rate": 0.000172146829810901,
+      "loss": 3.4998,
+      "step": 632
+    },
+    {
+      "epoch": 8.441471571906355,
+      "grad_norm": 0.8822937607765198,
+      "learning_rate": 0.0001721023359288098,
+      "loss": 3.4281,
+      "step": 633
+    },
+    {
+      "epoch": 8.454849498327759,
+      "grad_norm": 0.764872670173645,
+      "learning_rate": 0.00017205784204671858,
+      "loss": 3.3693,
+      "step": 634
+    },
+    {
+      "epoch": 8.468227424749164,
+      "grad_norm": 0.7492384910583496,
+      "learning_rate": 0.00017201334816462739,
+      "loss": 3.5672,
+      "step": 635
+    },
+    {
+      "epoch": 8.481605351170568,
+      "grad_norm": 0.8037416934967041,
+      "learning_rate": 0.00017196885428253614,
+      "loss": 3.3804,
+      "step": 636
+    },
+    {
+      "epoch": 8.494983277591974,
+      "grad_norm": 0.8380945324897766,
+      "learning_rate": 0.00017192436040044495,
+      "loss": 3.2272,
+      "step": 637
+    },
+    {
+      "epoch": 8.508361204013378,
+      "grad_norm": 0.8467932939529419,
+      "learning_rate": 0.00017187986651835373,
+      "loss": 3.2649,
+      "step": 638
+    },
+    {
+      "epoch": 8.521739130434783,
+      "grad_norm": 0.751542866230011,
+      "learning_rate": 0.00017183537263626252,
+      "loss": 3.4135,
+      "step": 639
+    },
+    {
+      "epoch": 8.535117056856187,
+      "grad_norm": 0.7618190050125122,
+      "learning_rate": 0.0001717908787541713,
+      "loss": 3.483,
+      "step": 640
+    },
+    {
+      "epoch": 8.548494983277592,
+      "grad_norm": 0.9661890864372253,
+      "learning_rate": 0.0001717463848720801,
+      "loss": 3.2201,
+      "step": 641
+    },
+    {
+      "epoch": 8.561872909698996,
+      "grad_norm": 0.8655393719673157,
+      "learning_rate": 0.0001717018909899889,
+      "loss": 3.4142,
+      "step": 642
+    },
+    {
+      "epoch": 8.575250836120402,
+      "grad_norm": 0.796047031879425,
+      "learning_rate": 0.00017165739710789767,
+      "loss": 3.3558,
+      "step": 643
+    },
+    {
+      "epoch": 8.588628762541806,
+      "grad_norm": 1.0098161697387695,
+      "learning_rate": 0.00017161290322580645,
+      "loss": 3.553,
+      "step": 644
+    },
+    {
+      "epoch": 8.602006688963211,
+      "grad_norm": 1.1880302429199219,
+      "learning_rate": 0.00017156840934371526,
+      "loss": 3.3581,
+      "step": 645
+    },
+    {
+      "epoch": 8.615384615384615,
+      "grad_norm": 0.9361609220504761,
+      "learning_rate": 0.00017152391546162402,
+      "loss": 3.4543,
+      "step": 646
+    },
+    {
+      "epoch": 8.62876254180602,
+      "grad_norm": 0.8794479966163635,
+      "learning_rate": 0.00017147942157953283,
+      "loss": 3.3654,
+      "step": 647
+    },
+    {
+      "epoch": 8.642140468227424,
+      "grad_norm": 0.9263080954551697,
+      "learning_rate": 0.0001714349276974416,
+      "loss": 3.4376,
+      "step": 648
+    },
+    {
+      "epoch": 8.65551839464883,
+      "grad_norm": 1.0015815496444702,
+      "learning_rate": 0.0001713904338153504,
+      "loss": 3.3533,
+      "step": 649
+    },
+    {
+      "epoch": 8.668896321070234,
+      "grad_norm": 0.8525484204292297,
+      "learning_rate": 0.00017134593993325918,
+      "loss": 3.3897,
+      "step": 650
+    },
+    {
+      "epoch": 8.68227424749164,
+      "grad_norm": 0.7196484804153442,
+      "learning_rate": 0.00017130144605116799,
+      "loss": 3.2428,
+      "step": 651
+    },
+    {
+      "epoch": 8.695652173913043,
+      "grad_norm": 0.8779593706130981,
+      "learning_rate": 0.00017125695216907677,
+      "loss": 3.5471,
+      "step": 652
+    },
+    {
+      "epoch": 8.709030100334449,
+      "grad_norm": 0.9256909489631653,
+      "learning_rate": 0.00017121245828698555,
+      "loss": 3.1776,
+      "step": 653
+    },
+    {
+      "epoch": 8.722408026755852,
+      "grad_norm": 0.7774620652198792,
+      "learning_rate": 0.00017116796440489433,
+      "loss": 3.4836,
+      "step": 654
+    },
+    {
+      "epoch": 8.735785953177258,
+      "grad_norm": 0.8112596273422241,
+      "learning_rate": 0.00017112347052280314,
+      "loss": 3.5194,
+      "step": 655
+    },
+    {
+      "epoch": 8.749163879598662,
+      "grad_norm": 0.7350602746009827,
+      "learning_rate": 0.0001710789766407119,
+      "loss": 3.2248,
+      "step": 656
+    },
+    {
+      "epoch": 8.762541806020067,
+      "grad_norm": 0.8231781125068665,
+      "learning_rate": 0.0001710344827586207,
+      "loss": 3.4659,
+      "step": 657
+    },
+    {
+      "epoch": 8.775919732441471,
+      "grad_norm": 0.8921564221382141,
+      "learning_rate": 0.0001709899888765295,
+      "loss": 3.2712,
+      "step": 658
+    },
+    {
+      "epoch": 8.789297658862877,
+      "grad_norm": 0.8921830058097839,
+      "learning_rate": 0.00017094549499443827,
+      "loss": 3.5071,
+      "step": 659
+    },
+    {
+      "epoch": 8.80267558528428,
+      "grad_norm": 0.7809077501296997,
+      "learning_rate": 0.00017090100111234705,
+      "loss": 3.6639,
+      "step": 660
+    },
+    {
+      "epoch": 8.816053511705686,
+      "grad_norm": 0.9431234002113342,
+      "learning_rate": 0.00017085650723025586,
+      "loss": 3.2795,
+      "step": 661
+    },
+    {
+      "epoch": 8.82943143812709,
+      "grad_norm": 0.9707314968109131,
+      "learning_rate": 0.00017081201334816465,
+      "loss": 3.3395,
+      "step": 662
+    },
+    {
+      "epoch": 8.842809364548495,
+      "grad_norm": 0.7547470331192017,
+      "learning_rate": 0.00017076751946607343,
+      "loss": 3.5316,
+      "step": 663
+    },
+    {
+      "epoch": 8.856187290969899,
+      "grad_norm": 0.8989250659942627,
+      "learning_rate": 0.0001707230255839822,
+      "loss": 3.4029,
+      "step": 664
+    },
+    {
+      "epoch": 8.869565217391305,
+      "grad_norm": 1.0237400531768799,
+      "learning_rate": 0.00017067853170189102,
+      "loss": 3.5014,
+      "step": 665
+    },
+    {
+      "epoch": 8.882943143812708,
+      "grad_norm": 0.7289263010025024,
+      "learning_rate": 0.00017063403781979978,
+      "loss": 3.4211,
+      "step": 666
+    },
+    {
+      "epoch": 8.896321070234114,
+      "grad_norm": 0.7978695034980774,
+      "learning_rate": 0.00017058954393770856,
+      "loss": 3.5421,
+      "step": 667
+    },
+    {
+      "epoch": 8.909698996655518,
+      "grad_norm": 0.7401835918426514,
+      "learning_rate": 0.00017054505005561737,
+      "loss": 3.3419,
+      "step": 668
+    },
+    {
+      "epoch": 8.923076923076923,
+      "grad_norm": 0.8952983617782593,
+      "learning_rate": 0.00017050055617352615,
+      "loss": 3.1322,
+      "step": 669
+    },
+    {
+      "epoch": 8.936454849498327,
+      "grad_norm": 0.6922047734260559,
+      "learning_rate": 0.00017045606229143493,
+      "loss": 3.5872,
+      "step": 670
+    },
+    {
+      "epoch": 8.949832775919733,
+      "grad_norm": 0.8618977665901184,
+      "learning_rate": 0.00017041156840934371,
+      "loss": 3.1366,
+      "step": 671
+    },
+    {
+      "epoch": 8.963210702341136,
+      "grad_norm": 0.7933799624443054,
+      "learning_rate": 0.00017036707452725252,
+      "loss": 3.3108,
+      "step": 672
+    },
+    {
+      "epoch": 8.976588628762542,
+      "grad_norm": 0.718401312828064,
+      "learning_rate": 0.00017032258064516128,
+      "loss": 3.3771,
+      "step": 673
+    },
+    {
+      "epoch": 8.989966555183946,
+      "grad_norm": 0.8096804618835449,
+      "learning_rate": 0.0001702780867630701,
+      "loss": 3.3225,
+      "step": 674
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.0055694580078125,
+      "learning_rate": 0.00017023359288097887,
+      "loss": 3.5545,
+      "step": 675
+    },
+    {
+      "epoch": 9.013377926421406,
+      "grad_norm": 0.710986852645874,
+      "learning_rate": 0.00017018909899888765,
+      "loss": 3.3333,
+      "step": 676
+    },
+    {
+      "epoch": 9.02675585284281,
+      "grad_norm": 0.672132134437561,
+      "learning_rate": 0.00017014460511679644,
+      "loss": 2.9995,
+      "step": 677
+    },
+    {
+      "epoch": 9.040133779264215,
+      "grad_norm": 0.6752933263778687,
+      "learning_rate": 0.00017010011123470525,
+      "loss": 3.3571,
+      "step": 678
+    },
+    {
+      "epoch": 9.053511705685619,
+      "grad_norm": 0.6553521156311035,
+      "learning_rate": 0.00017005561735261403,
+      "loss": 3.3407,
+      "step": 679
+    },
+    {
+      "epoch": 9.066889632107024,
+      "grad_norm": 0.7492311596870422,
+      "learning_rate": 0.0001700111234705228,
+      "loss": 3.325,
+      "step": 680
+    },
+    {
+      "epoch": 9.080267558528428,
+      "grad_norm": 0.736139714717865,
+      "learning_rate": 0.0001699666295884316,
+      "loss": 3.2626,
+      "step": 681
+    },
+    {
+      "epoch": 9.093645484949834,
+      "grad_norm": 0.7131486535072327,
+      "learning_rate": 0.0001699221357063404,
+      "loss": 3.2612,
+      "step": 682
+    },
+    {
+      "epoch": 9.107023411371237,
+      "grad_norm": 0.7037603855133057,
+      "learning_rate": 0.00016987764182424916,
+      "loss": 3.2418,
+      "step": 683
+    },
+    {
+      "epoch": 9.120401337792643,
+      "grad_norm": 0.685518205165863,
+      "learning_rate": 0.00016983314794215797,
+      "loss": 3.4854,
+      "step": 684
+    },
+    {
+      "epoch": 9.133779264214047,
+      "grad_norm": 0.6050254106521606,
+      "learning_rate": 0.00016978865406006675,
+      "loss": 3.2312,
+      "step": 685
+    },
+    {
+      "epoch": 9.147157190635452,
+      "grad_norm": 0.6932830810546875,
+      "learning_rate": 0.00016974416017797553,
+      "loss": 3.4634,
+      "step": 686
+    },
+    {
+      "epoch": 9.160535117056856,
+      "grad_norm": 0.7055158615112305,
+      "learning_rate": 0.00016969966629588431,
+      "loss": 3.1408,
+      "step": 687
+    },
+    {
+      "epoch": 9.173913043478262,
+      "grad_norm": 0.6887643337249756,
+      "learning_rate": 0.00016965517241379312,
+      "loss": 3.0697,
+      "step": 688
+    },
+    {
+      "epoch": 9.187290969899665,
+      "grad_norm": 0.7201237082481384,
+      "learning_rate": 0.0001696106785317019,
+      "loss": 3.303,
+      "step": 689
+    },
+    {
+      "epoch": 9.200668896321071,
+      "grad_norm": 0.6617894768714905,
+      "learning_rate": 0.0001695661846496107,
+      "loss": 3.4846,
+      "step": 690
+    },
+    {
+      "epoch": 9.214046822742475,
+      "grad_norm": 0.8979818224906921,
+      "learning_rate": 0.00016952169076751947,
+      "loss": 3.1898,
+      "step": 691
+    },
+    {
+      "epoch": 9.22742474916388,
+      "grad_norm": 0.9507981538772583,
+      "learning_rate": 0.00016947719688542828,
+      "loss": 3.4748,
+      "step": 692
+    },
+    {
+      "epoch": 9.240802675585284,
+      "grad_norm": 0.7935391068458557,
+      "learning_rate": 0.00016943270300333704,
+      "loss": 3.3661,
+      "step": 693
+    },
+    {
+      "epoch": 9.25418060200669,
+      "grad_norm": 0.7437114715576172,
+      "learning_rate": 0.00016938820912124584,
+      "loss": 3.4407,
+      "step": 694
+    },
+    {
+      "epoch": 9.267558528428093,
+      "grad_norm": 0.680610179901123,
+      "learning_rate": 0.00016934371523915463,
+      "loss": 3.3135,
+      "step": 695
+    },
+    {
+      "epoch": 9.280936454849499,
+      "grad_norm": 0.846716582775116,
+      "learning_rate": 0.0001692992213570634,
+      "loss": 3.3923,
+      "step": 696
+    },
+    {
+      "epoch": 9.294314381270903,
+      "grad_norm": 0.9567786455154419,
+      "learning_rate": 0.0001692547274749722,
+      "loss": 3.1405,
+      "step": 697
+    },
+    {
+      "epoch": 9.307692307692308,
+      "grad_norm": 0.7509344816207886,
+      "learning_rate": 0.000169210233592881,
+      "loss": 3.4031,
+      "step": 698
+    },
+    {
+      "epoch": 9.321070234113712,
+      "grad_norm": 0.8118243217468262,
+      "learning_rate": 0.00016916573971078978,
+      "loss": 3.317,
+      "step": 699
+    },
+    {
+      "epoch": 9.334448160535118,
+      "grad_norm": 0.7445617318153381,
+      "learning_rate": 0.00016912124582869857,
+      "loss": 3.3989,
+      "step": 700
+    },
+    {
+      "epoch": 9.347826086956522,
+      "grad_norm": 0.7520869970321655,
+      "learning_rate": 0.00016907675194660735,
+      "loss": 3.08,
+      "step": 701
+    },
+    {
+      "epoch": 9.361204013377927,
+      "grad_norm": 0.7466426491737366,
+      "learning_rate": 0.00016903225806451616,
+      "loss": 3.3338,
+      "step": 702
+    },
+    {
+      "epoch": 9.37458193979933,
+      "grad_norm": 0.7595514059066772,
+      "learning_rate": 0.0001689877641824249,
+      "loss": 3.08,
+      "step": 703
+    },
+    {
+      "epoch": 9.387959866220736,
+      "grad_norm": 0.713771641254425,
+      "learning_rate": 0.00016894327030033372,
+      "loss": 3.236,
+      "step": 704
+    },
+    {
+      "epoch": 9.40133779264214,
+      "grad_norm": 0.670863687992096,
+      "learning_rate": 0.0001688987764182425,
+      "loss": 3.4491,
+      "step": 705
+    },
+    {
+      "epoch": 9.414715719063546,
+      "grad_norm": 0.8842789530754089,
+      "learning_rate": 0.0001688542825361513,
+      "loss": 3.3444,
+      "step": 706
+    },
+    {
+      "epoch": 9.42809364548495,
+      "grad_norm": 0.8298172950744629,
+      "learning_rate": 0.00016880978865406007,
+      "loss": 3.3818,
+      "step": 707
+    },
+    {
+      "epoch": 9.441471571906355,
+      "grad_norm": 0.7407504320144653,
+      "learning_rate": 0.00016876529477196885,
+      "loss": 3.2294,
+      "step": 708
+    },
+    {
+      "epoch": 9.454849498327759,
+      "grad_norm": 0.6642070412635803,
+      "learning_rate": 0.00016872080088987766,
+      "loss": 3.3009,
+      "step": 709
+    },
+    {
+      "epoch": 9.468227424749164,
+      "grad_norm": 0.7627503275871277,
+      "learning_rate": 0.00016867630700778642,
+      "loss": 3.3486,
+      "step": 710
+    },
+    {
+      "epoch": 9.481605351170568,
+      "grad_norm": 0.7307603359222412,
+      "learning_rate": 0.00016863181312569523,
+      "loss": 3.0257,
+      "step": 711
+    },
+    {
+      "epoch": 9.494983277591974,
+      "grad_norm": 0.7932866215705872,
+      "learning_rate": 0.000168587319243604,
+      "loss": 3.4191,
+      "step": 712
+    },
+    {
+      "epoch": 9.508361204013378,
+      "grad_norm": 0.7457575798034668,
+      "learning_rate": 0.0001685428253615128,
+      "loss": 3.2283,
+      "step": 713
+    },
+    {
+      "epoch": 9.521739130434783,
+      "grad_norm": 0.6718200445175171,
+      "learning_rate": 0.00016849833147942157,
+      "loss": 3.4569,
+      "step": 714
+    },
+    {
+      "epoch": 9.535117056856187,
+      "grad_norm": 0.8189072608947754,
+      "learning_rate": 0.00016845383759733038,
+      "loss": 3.2585,
+      "step": 715
+    },
+    {
+      "epoch": 9.548494983277592,
+      "grad_norm": 0.6895336508750916,
+      "learning_rate": 0.00016840934371523917,
+      "loss": 3.2417,
+      "step": 716
+    },
+    {
+      "epoch": 9.561872909698996,
+      "grad_norm": 0.723173975944519,
+      "learning_rate": 0.00016836484983314795,
+      "loss": 3.3243,
+      "step": 717
+    },
+    {
+      "epoch": 9.575250836120402,
+      "grad_norm": 0.8354344964027405,
+      "learning_rate": 0.00016832035595105673,
+      "loss": 3.3585,
+      "step": 718
+    },
+    {
+      "epoch": 9.588628762541806,
+      "grad_norm": 0.6736294031143188,
+      "learning_rate": 0.00016827586206896554,
+      "loss": 3.188,
+      "step": 719
+    },
+    {
+      "epoch": 9.602006688963211,
+      "grad_norm": 0.7790263295173645,
+      "learning_rate": 0.0001682313681868743,
+      "loss": 3.3171,
+      "step": 720
+    },
+    {
+      "epoch": 9.615384615384615,
+      "grad_norm": 0.6426937580108643,
+      "learning_rate": 0.0001681868743047831,
+      "loss": 3.0854,
+      "step": 721
+    },
+    {
+      "epoch": 9.62876254180602,
+      "grad_norm": 0.7029106020927429,
+      "learning_rate": 0.0001681423804226919,
+      "loss": 3.3629,
+      "step": 722
+    },
+    {
+      "epoch": 9.642140468227424,
+      "grad_norm": 0.8353022933006287,
+      "learning_rate": 0.00016809788654060067,
+      "loss": 3.3715,
+      "step": 723
+    },
+    {
+      "epoch": 9.65551839464883,
+      "grad_norm": 0.8578335642814636,
+      "learning_rate": 0.00016805339265850945,
+      "loss": 3.3554,
+      "step": 724
+    },
+    {
+      "epoch": 9.668896321070234,
+      "grad_norm": 0.6998556852340698,
+      "learning_rate": 0.00016800889877641826,
+      "loss": 3.3043,
+      "step": 725
+    },
+    {
+      "epoch": 9.68227424749164,
+      "grad_norm": 0.7134855389595032,
+      "learning_rate": 0.00016796440489432704,
+      "loss": 3.5856,
+      "step": 726
+    },
+    {
+      "epoch": 9.695652173913043,
+      "grad_norm": 0.6636050939559937,
+      "learning_rate": 0.00016791991101223583,
+      "loss": 3.2156,
+      "step": 727
+    },
+    {
+      "epoch": 9.709030100334449,
+      "grad_norm": 0.7757130861282349,
+      "learning_rate": 0.0001678754171301446,
+      "loss": 3.4974,
+      "step": 728
+    },
+    {
+      "epoch": 9.722408026755852,
+      "grad_norm": 0.770648181438446,
+      "learning_rate": 0.00016783092324805342,
+      "loss": 3.3251,
+      "step": 729
+    },
+    {
+      "epoch": 9.735785953177258,
+      "grad_norm": 0.7728201746940613,
+      "learning_rate": 0.00016778642936596217,
+      "loss": 3.2666,
+      "step": 730
+    },
+    {
+      "epoch": 9.749163879598662,
+      "grad_norm": 0.8277239203453064,
+      "learning_rate": 0.00016774193548387098,
+      "loss": 3.3867,
+      "step": 731
+    },
+    {
+      "epoch": 9.762541806020067,
+      "grad_norm": 0.6534886360168457,
+      "learning_rate": 0.00016769744160177977,
+      "loss": 3.175,
+      "step": 732
+    },
+    {
+      "epoch": 9.775919732441471,
+      "grad_norm": 0.8508428335189819,
+      "learning_rate": 0.00016765294771968855,
+      "loss": 3.2084,
+      "step": 733
+    },
+    {
+      "epoch": 9.789297658862877,
+      "grad_norm": 0.7656721472740173,
+      "learning_rate": 0.00016760845383759733,
+      "loss": 3.1426,
+      "step": 734
+    },
+    {
+      "epoch": 9.80267558528428,
+      "grad_norm": 0.9495553970336914,
+      "learning_rate": 0.00016756395995550614,
+      "loss": 3.1623,
+      "step": 735
+    },
+    {
+      "epoch": 9.816053511705686,
+      "grad_norm": 0.7998641729354858,
+      "learning_rate": 0.00016751946607341492,
+      "loss": 3.3893,
+      "step": 736
+    },
+    {
+      "epoch": 9.82943143812709,
+      "grad_norm": 0.8124551177024841,
+      "learning_rate": 0.0001674749721913237,
+      "loss": 3.2012,
+      "step": 737
+    },
+    {
+      "epoch": 9.842809364548495,
+      "grad_norm": 0.6332049369812012,
+      "learning_rate": 0.0001674304783092325,
+      "loss": 3.3384,
+      "step": 738
+    },
+    {
+      "epoch": 9.856187290969899,
+      "grad_norm": 0.7114555835723877,
+      "learning_rate": 0.0001673859844271413,
+      "loss": 3.0802,
+      "step": 739
+    },
+    {
+      "epoch": 9.869565217391305,
+      "grad_norm": 0.7175182700157166,
+      "learning_rate": 0.00016734149054505005,
+      "loss": 3.2572,
+      "step": 740
+    },
+    {
+      "epoch": 9.882943143812708,
+      "grad_norm": 0.7724816799163818,
+      "learning_rate": 0.00016729699666295886,
+      "loss": 3.1078,
+      "step": 741
+    },
+    {
+      "epoch": 9.896321070234114,
+      "grad_norm": 0.7834901213645935,
+      "learning_rate": 0.00016725250278086764,
+      "loss": 3.2513,
+      "step": 742
+    },
+    {
+      "epoch": 9.909698996655518,
+      "grad_norm": 0.663495659828186,
+      "learning_rate": 0.00016720800889877643,
+      "loss": 3.327,
+      "step": 743
+    },
+    {
+      "epoch": 9.923076923076923,
+      "grad_norm": 0.7828975319862366,
+      "learning_rate": 0.0001671635150166852,
+      "loss": 3.609,
+      "step": 744
+    },
+    {
+      "epoch": 9.936454849498327,
+      "grad_norm": 0.6747825145721436,
+      "learning_rate": 0.00016711902113459402,
+      "loss": 3.4479,
+      "step": 745
+    },
+    {
+      "epoch": 9.949832775919733,
+      "grad_norm": 0.7816379070281982,
+      "learning_rate": 0.0001670745272525028,
+      "loss": 3.2369,
+      "step": 746
+    },
+    {
+      "epoch": 9.963210702341136,
+      "grad_norm": 0.7011098265647888,
+      "learning_rate": 0.00016703003337041158,
+      "loss": 2.9103,
+      "step": 747
+    },
+    {
+      "epoch": 9.976588628762542,
+      "grad_norm": 0.7165176868438721,
+      "learning_rate": 0.00016698553948832036,
+      "loss": 3.1669,
+      "step": 748
+    },
+    {
+      "epoch": 9.989966555183946,
+      "grad_norm": 0.766315758228302,
+      "learning_rate": 0.00016694104560622915,
+      "loss": 3.1138,
+      "step": 749
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.876315176486969,
+      "learning_rate": 0.00016689655172413793,
+      "loss": 3.4851,
+      "step": 750
+    },
+    {
+      "epoch": 10.013377926421406,
+      "grad_norm": 0.807686984539032,
+      "learning_rate": 0.0001668520578420467,
+      "loss": 3.193,
+      "step": 751
+    },
+    {
+      "epoch": 10.02675585284281,
+      "grad_norm": 0.7085704803466797,
+      "learning_rate": 0.00016680756395995552,
+      "loss": 3.4797,
+      "step": 752
+    },
+    {
+      "epoch": 10.040133779264215,
+      "grad_norm": 0.7119605541229248,
+      "learning_rate": 0.0001667630700778643,
+      "loss": 3.1713,
+      "step": 753
+    },
+    {
+      "epoch": 10.053511705685619,
+      "grad_norm": 0.6569423675537109,
+      "learning_rate": 0.00016671857619577309,
+      "loss": 3.1661,
+      "step": 754
+    },
+    {
+      "epoch": 10.066889632107024,
+      "grad_norm": 0.8173550367355347,
+      "learning_rate": 0.00016667408231368187,
+      "loss": 2.8467,
+      "step": 755
+    },
+    {
+      "epoch": 10.080267558528428,
+      "grad_norm": 0.7261365056037903,
+      "learning_rate": 0.00016662958843159068,
+      "loss": 3.3679,
+      "step": 756
+    },
+    {
+      "epoch": 10.093645484949834,
+      "grad_norm": 0.7997227311134338,
+      "learning_rate": 0.00016658509454949943,
+      "loss": 3.0985,
+      "step": 757
+    },
+    {
+      "epoch": 10.107023411371237,
+      "grad_norm": 0.653391420841217,
+      "learning_rate": 0.00016654060066740824,
+      "loss": 3.2156,
+      "step": 758
+    },
+    {
+      "epoch": 10.120401337792643,
+      "grad_norm": 0.6799002289772034,
+      "learning_rate": 0.00016649610678531703,
+      "loss": 3.3302,
+      "step": 759
+    },
+    {
+      "epoch": 10.133779264214047,
+      "grad_norm": 0.6444498896598816,
+      "learning_rate": 0.0001664516129032258,
+      "loss": 3.2813,
+      "step": 760
+    },
+    {
+      "epoch": 10.147157190635452,
+      "grad_norm": 1.064769983291626,
+      "learning_rate": 0.0001664071190211346,
+      "loss": 3.1852,
+      "step": 761
+    },
+    {
+      "epoch": 10.160535117056856,
+      "grad_norm": 0.6534339189529419,
+      "learning_rate": 0.0001663626251390434,
+      "loss": 3.1563,
+      "step": 762
+    },
+    {
+      "epoch": 10.173913043478262,
+      "grad_norm": 0.6909127235412598,
+      "learning_rate": 0.00016631813125695218,
+      "loss": 3.2728,
+      "step": 763
+    },
+    {
+      "epoch": 10.187290969899665,
+      "grad_norm": 0.6549767851829529,
+      "learning_rate": 0.00016627363737486096,
+      "loss": 3.0491,
+      "step": 764
+    },
+    {
+      "epoch": 10.200668896321071,
+      "grad_norm": 0.678054928779602,
+      "learning_rate": 0.00016622914349276975,
+      "loss": 3.4807,
+      "step": 765
+    },
+    {
+      "epoch": 10.214046822742475,
+      "grad_norm": 0.613358199596405,
+      "learning_rate": 0.00016618464961067856,
+      "loss": 3.1746,
+      "step": 766
+    },
+    {
+      "epoch": 10.22742474916388,
+      "grad_norm": 0.6624737977981567,
+      "learning_rate": 0.0001661401557285873,
+      "loss": 2.8528,
+      "step": 767
+    },
+    {
+      "epoch": 10.240802675585284,
+      "grad_norm": 0.65067458152771,
+      "learning_rate": 0.00016609566184649612,
+      "loss": 3.1843,
+      "step": 768
+    },
+    {
+      "epoch": 10.25418060200669,
+      "grad_norm": 0.6192435622215271,
+      "learning_rate": 0.0001660511679644049,
+      "loss": 3.3162,
+      "step": 769
+    },
+    {
+      "epoch": 10.267558528428093,
+      "grad_norm": 0.6456341743469238,
+      "learning_rate": 0.00016600667408231369,
+      "loss": 3.2302,
+      "step": 770
+    },
+    {
+      "epoch": 10.280936454849499,
+      "grad_norm": 2.357724189758301,
+      "learning_rate": 0.00016596218020022247,
+      "loss": 3.2741,
+      "step": 771
+    },
+    {
+      "epoch": 10.294314381270903,
+      "grad_norm": 0.6833475828170776,
+      "learning_rate": 0.00016591768631813128,
+      "loss": 3.1516,
+      "step": 772
+    },
+    {
+      "epoch": 10.307692307692308,
+      "grad_norm": 0.5557199716567993,
+      "learning_rate": 0.00016587319243604006,
+      "loss": 3.281,
+      "step": 773
+    },
+    {
+      "epoch": 10.321070234113712,
+      "grad_norm": 0.6617905497550964,
+      "learning_rate": 0.00016582869855394884,
+      "loss": 3.375,
+      "step": 774
+    },
+    {
+      "epoch": 10.334448160535118,
+      "grad_norm": 0.5671921372413635,
+      "learning_rate": 0.00016578420467185762,
+      "loss": 3.4335,
+      "step": 775
+    },
+    {
+      "epoch": 10.347826086956522,
+      "grad_norm": 0.8487278819084167,
+      "learning_rate": 0.00016573971078976643,
+      "loss": 3.143,
+      "step": 776
+    },
+    {
+      "epoch": 10.361204013377927,
+      "grad_norm": 0.6489982604980469,
+      "learning_rate": 0.0001656952169076752,
+      "loss": 3.3258,
+      "step": 777
+    },
+    {
+      "epoch": 10.37458193979933,
+      "grad_norm": 0.8773537278175354,
+      "learning_rate": 0.000165650723025584,
+      "loss": 3.0466,
+      "step": 778
+    },
+    {
+      "epoch": 10.387959866220736,
+      "grad_norm": 0.5961865782737732,
+      "learning_rate": 0.00016560622914349278,
+      "loss": 3.3417,
+      "step": 779
+    },
+    {
+      "epoch": 10.40133779264214,
+      "grad_norm": 0.6149600148200989,
+      "learning_rate": 0.00016556173526140156,
+      "loss": 3.0622,
+      "step": 780
+    },
+    {
+      "epoch": 10.414715719063546,
+      "grad_norm": 0.7591158151626587,
+      "learning_rate": 0.00016551724137931035,
+      "loss": 3.2078,
+      "step": 781
+    },
+    {
+      "epoch": 10.42809364548495,
+      "grad_norm": 0.7915151119232178,
+      "learning_rate": 0.00016547274749721916,
+      "loss": 3.3082,
+      "step": 782
+    },
+    {
+      "epoch": 10.441471571906355,
+      "grad_norm": 0.8709903359413147,
+      "learning_rate": 0.00016542825361512794,
+      "loss": 3.3073,
+      "step": 783
+    },
+    {
+      "epoch": 10.454849498327759,
+      "grad_norm": 0.6593959331512451,
+      "learning_rate": 0.00016538375973303672,
+      "loss": 3.1574,
+      "step": 784
+    },
+    {
+      "epoch": 10.468227424749164,
+      "grad_norm": 0.8101013898849487,
+      "learning_rate": 0.0001653392658509455,
+      "loss": 3.3631,
+      "step": 785
+    },
+    {
+      "epoch": 10.481605351170568,
+      "grad_norm": 0.8200273513793945,
+      "learning_rate": 0.0001652947719688543,
+      "loss": 3.0447,
+      "step": 786
+    },
+    {
+      "epoch": 10.494983277591974,
+      "grad_norm": 0.7304090857505798,
+      "learning_rate": 0.00016525027808676307,
+      "loss": 3.3382,
+      "step": 787
+    },
+    {
+      "epoch": 10.508361204013378,
+      "grad_norm": 0.7059088349342346,
+      "learning_rate": 0.00016520578420467188,
+      "loss": 3.085,
+      "step": 788
+    },
+    {
+      "epoch": 10.521739130434783,
+      "grad_norm": 0.6664522886276245,
+      "learning_rate": 0.00016516129032258066,
+      "loss": 3.3198,
+      "step": 789
+    },
+    {
+      "epoch": 10.535117056856187,
+      "grad_norm": 0.9230799078941345,
+      "learning_rate": 0.00016511679644048944,
+      "loss": 3.2502,
+      "step": 790
+    },
+    {
+      "epoch": 10.548494983277592,
+      "grad_norm": 0.6974027752876282,
+      "learning_rate": 0.00016507230255839822,
+      "loss": 3.2432,
+      "step": 791
+    },
+    {
+      "epoch": 10.561872909698996,
+      "grad_norm": 0.7186788320541382,
+      "learning_rate": 0.000165027808676307,
+      "loss": 3.2273,
+      "step": 792
+    },
+    {
+      "epoch": 10.575250836120402,
+      "grad_norm": 0.6168047189712524,
+      "learning_rate": 0.00016498331479421582,
+      "loss": 3.2319,
+      "step": 793
+    },
+    {
+      "epoch": 10.588628762541806,
+      "grad_norm": 0.6219142079353333,
+      "learning_rate": 0.00016493882091212457,
+      "loss": 2.9733,
+      "step": 794
+    },
+    {
+      "epoch": 10.602006688963211,
+      "grad_norm": 0.573359489440918,
+      "learning_rate": 0.00016489432703003338,
+      "loss": 3.214,
+      "step": 795
+    },
+    {
+      "epoch": 10.615384615384615,
+      "grad_norm": 0.678263783454895,
+      "learning_rate": 0.00016484983314794216,
+      "loss": 3.2331,
+      "step": 796
+    },
+    {
+      "epoch": 10.62876254180602,
+      "grad_norm": 0.6593761444091797,
+      "learning_rate": 0.00016480533926585095,
+      "loss": 3.3605,
+      "step": 797
+    },
+    {
+      "epoch": 10.642140468227424,
+      "grad_norm": 0.8732627034187317,
+      "learning_rate": 0.00016476084538375973,
+      "loss": 3.3531,
+      "step": 798
+    },
+    {
+      "epoch": 10.65551839464883,
+      "grad_norm": 0.7198925614356995,
+      "learning_rate": 0.00016471635150166854,
+      "loss": 3.2927,
+      "step": 799
+    },
+    {
+      "epoch": 10.668896321070234,
+      "grad_norm": 0.7275107502937317,
+      "learning_rate": 0.00016467185761957732,
+      "loss": 3.3695,
+      "step": 800
+    },
+    {
+      "epoch": 10.68227424749164,
+      "grad_norm": 0.7077828049659729,
+      "learning_rate": 0.0001646273637374861,
+      "loss": 2.846,
+      "step": 801
+    },
+    {
+      "epoch": 10.695652173913043,
+      "grad_norm": 0.7579251527786255,
+      "learning_rate": 0.00016458286985539488,
+      "loss": 3.1917,
+      "step": 802
+    },
+    {
+      "epoch": 10.709030100334449,
+      "grad_norm": 0.7607265114784241,
+      "learning_rate": 0.0001645383759733037,
+      "loss": 3.2919,
+      "step": 803
+    },
+    {
+      "epoch": 10.722408026755852,
+      "grad_norm": 0.7122685313224792,
+      "learning_rate": 0.00016449388209121245,
+      "loss": 3.7108,
+      "step": 804
+    },
+    {
+      "epoch": 10.735785953177258,
+      "grad_norm": 0.7256726622581482,
+      "learning_rate": 0.00016444938820912126,
+      "loss": 3.3209,
+      "step": 805
+    },
+    {
+      "epoch": 10.749163879598662,
+      "grad_norm": 0.7903631925582886,
+      "learning_rate": 0.00016440489432703004,
+      "loss": 3.398,
+      "step": 806
+    },
+    {
+      "epoch": 10.762541806020067,
+      "grad_norm": 1.78204345703125,
+      "learning_rate": 0.00016436040044493882,
+      "loss": 3.1118,
+      "step": 807
+    },
+    {
+      "epoch": 10.775919732441471,
+      "grad_norm": 0.7647016644477844,
+      "learning_rate": 0.0001643159065628476,
+      "loss": 3.1889,
+      "step": 808
+    },
+    {
+      "epoch": 10.789297658862877,
+      "grad_norm": 0.8039811253547668,
+      "learning_rate": 0.00016427141268075642,
+      "loss": 3.0104,
+      "step": 809
+    },
+    {
+      "epoch": 10.80267558528428,
+      "grad_norm": 0.6011155843734741,
+      "learning_rate": 0.0001642269187986652,
+      "loss": 3.0988,
+      "step": 810
+    },
+    {
+      "epoch": 10.816053511705686,
+      "grad_norm": 0.8137276768684387,
+      "learning_rate": 0.00016418242491657398,
+      "loss": 3.4308,
+      "step": 811
+    },
+    {
+      "epoch": 10.82943143812709,
+      "grad_norm": 0.6501771807670593,
+      "learning_rate": 0.00016413793103448276,
+      "loss": 3.1405,
+      "step": 812
+    },
+    {
+      "epoch": 10.842809364548495,
+      "grad_norm": 0.678032636642456,
+      "learning_rate": 0.00016409343715239157,
+      "loss": 3.2243,
+      "step": 813
+    },
+    {
+      "epoch": 10.856187290969899,
+      "grad_norm": 0.6830305457115173,
+      "learning_rate": 0.00016404894327030033,
+      "loss": 3.2685,
+      "step": 814
+    },
+    {
+      "epoch": 10.869565217391305,
+      "grad_norm": 0.7482068538665771,
+      "learning_rate": 0.00016400444938820914,
+      "loss": 3.3363,
+      "step": 815
+    },
+    {
+      "epoch": 10.882943143812708,
+      "grad_norm": 0.6592227816581726,
+      "learning_rate": 0.00016395995550611792,
+      "loss": 3.2914,
+      "step": 816
+    },
+    {
+      "epoch": 10.896321070234114,
+      "grad_norm": 0.7520759105682373,
+      "learning_rate": 0.0001639154616240267,
+      "loss": 3.1371,
+      "step": 817
+    },
+    {
+      "epoch": 10.909698996655518,
+      "grad_norm": 0.6802201271057129,
+      "learning_rate": 0.00016387096774193548,
+      "loss": 3.2925,
+      "step": 818
+    },
+    {
+      "epoch": 10.923076923076923,
+      "grad_norm": 0.7528939247131348,
+      "learning_rate": 0.0001638264738598443,
+      "loss": 3.2147,
+      "step": 819
+    },
+    {
+      "epoch": 10.936454849498327,
+      "grad_norm": 0.7070727348327637,
+      "learning_rate": 0.00016378197997775308,
+      "loss": 3.2649,
+      "step": 820
+    },
+    {
+      "epoch": 10.949832775919733,
+      "grad_norm": 0.6121620535850525,
+      "learning_rate": 0.00016373748609566186,
+      "loss": 3.3999,
+      "step": 821
+    },
+    {
+      "epoch": 10.963210702341136,
+      "grad_norm": 0.7355679273605347,
+      "learning_rate": 0.00016369299221357064,
+      "loss": 3.2561,
+      "step": 822
+    },
+    {
+      "epoch": 10.976588628762542,
+      "grad_norm": 0.7294445037841797,
+      "learning_rate": 0.00016364849833147945,
+      "loss": 2.7019,
+      "step": 823
+    },
+    {
+      "epoch": 10.989966555183946,
+      "grad_norm": 0.8628729581832886,
+      "learning_rate": 0.0001636040044493882,
+      "loss": 3.3655,
+      "step": 824
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.8784325122833252,
+      "learning_rate": 0.00016355951056729702,
+      "loss": 3.504,
+      "step": 825
+    },
+    {
+      "epoch": 11.013377926421406,
+      "grad_norm": 0.6880869269371033,
+      "learning_rate": 0.0001635150166852058,
+      "loss": 3.1878,
+      "step": 826
+    },
+    {
+      "epoch": 11.02675585284281,
+      "grad_norm": 0.5625393390655518,
+      "learning_rate": 0.00016347052280311458,
+      "loss": 3.2757,
+      "step": 827
+    },
+    {
+      "epoch": 11.040133779264215,
+      "grad_norm": 0.5854038596153259,
+      "learning_rate": 0.00016342602892102336,
+      "loss": 3.0994,
+      "step": 828
+    },
+    {
+      "epoch": 11.053511705685619,
+      "grad_norm": 0.6682130098342896,
+      "learning_rate": 0.00016338153503893217,
+      "loss": 3.1935,
+      "step": 829
+    },
+    {
+      "epoch": 11.066889632107024,
+      "grad_norm": 0.6216278076171875,
+      "learning_rate": 0.00016333704115684095,
+      "loss": 3.1273,
+      "step": 830
+    },
+    {
+      "epoch": 11.080267558528428,
+      "grad_norm": 0.61285001039505,
+      "learning_rate": 0.0001632925472747497,
+      "loss": 3.2377,
+      "step": 831
+    },
+    {
+      "epoch": 11.093645484949834,
+      "grad_norm": 0.6559188365936279,
+      "learning_rate": 0.00016324805339265852,
+      "loss": 3.1084,
+      "step": 832
+    },
+    {
+      "epoch": 11.107023411371237,
+      "grad_norm": 0.6322848200798035,
+      "learning_rate": 0.0001632035595105673,
+      "loss": 3.3178,
+      "step": 833
+    },
+    {
+      "epoch": 11.120401337792643,
+      "grad_norm": 0.6306194067001343,
+      "learning_rate": 0.00016315906562847608,
+      "loss": 3.2012,
+      "step": 834
+    },
+    {
+      "epoch": 11.133779264214047,
+      "grad_norm": 0.6923161149024963,
+      "learning_rate": 0.00016311457174638487,
+      "loss": 2.9328,
+      "step": 835
+    },
+    {
+      "epoch": 11.147157190635452,
+      "grad_norm": 0.6900002360343933,
+      "learning_rate": 0.00016307007786429368,
+      "loss": 3.1436,
+      "step": 836
+    },
+    {
+      "epoch": 11.160535117056856,
+      "grad_norm": 0.817669153213501,
+      "learning_rate": 0.00016302558398220246,
+      "loss": 3.0935,
+      "step": 837
+    },
+    {
+      "epoch": 11.173913043478262,
+      "grad_norm": 0.7544119954109192,
+      "learning_rate": 0.00016298109010011124,
+      "loss": 3.0646,
+      "step": 838
+    },
+    {
+      "epoch": 11.187290969899665,
+      "grad_norm": 0.7996231913566589,
+      "learning_rate": 0.00016293659621802002,
+      "loss": 3.1975,
+      "step": 839
+    },
+    {
+      "epoch": 11.200668896321071,
+      "grad_norm": 0.6186792850494385,
+      "learning_rate": 0.00016289210233592883,
+      "loss": 3.31,
+      "step": 840
+    },
+    {
+      "epoch": 11.214046822742475,
+      "grad_norm": 0.6926666498184204,
+      "learning_rate": 0.0001628476084538376,
+      "loss": 3.0765,
+      "step": 841
+    },
+    {
+      "epoch": 11.22742474916388,
+      "grad_norm": 0.7475588917732239,
+      "learning_rate": 0.0001628031145717464,
+      "loss": 3.1743,
+      "step": 842
+    },
+    {
+      "epoch": 11.240802675585284,
+      "grad_norm": 0.5520989298820496,
+      "learning_rate": 0.00016275862068965518,
+      "loss": 3.4243,
+      "step": 843
+    },
+    {
+      "epoch": 11.25418060200669,
+      "grad_norm": 0.6556730270385742,
+      "learning_rate": 0.00016271412680756396,
+      "loss": 3.3293,
+      "step": 844
+    },
+    {
+      "epoch": 11.267558528428093,
+      "grad_norm": 0.6509199738502502,
+      "learning_rate": 0.00016266963292547274,
+      "loss": 2.999,
+      "step": 845
+    },
+    {
+      "epoch": 11.280936454849499,
+      "grad_norm": 0.6254273653030396,
+      "learning_rate": 0.00016262513904338155,
+      "loss": 3.1869,
+      "step": 846
+    },
+    {
+      "epoch": 11.294314381270903,
+      "grad_norm": 0.7454530000686646,
+      "learning_rate": 0.00016258064516129034,
+      "loss": 3.3694,
+      "step": 847
+    },
+    {
+      "epoch": 11.307692307692308,
+      "grad_norm": 0.7563592791557312,
+      "learning_rate": 0.00016253615127919912,
+      "loss": 3.0057,
+      "step": 848
+    },
+    {
+      "epoch": 11.321070234113712,
+      "grad_norm": 0.6986783742904663,
+      "learning_rate": 0.0001624916573971079,
+      "loss": 2.9893,
+      "step": 849
+    },
+    {
+      "epoch": 11.334448160535118,
+      "grad_norm": 0.7260631322860718,
+      "learning_rate": 0.0001624471635150167,
+      "loss": 3.1733,
+      "step": 850
+    },
+    {
+      "epoch": 11.347826086956522,
+      "grad_norm": 0.7522863745689392,
+      "learning_rate": 0.00016240266963292547,
+      "loss": 2.9829,
+      "step": 851
+    },
+    {
+      "epoch": 11.361204013377927,
+      "grad_norm": 0.7290140986442566,
+      "learning_rate": 0.00016235817575083428,
+      "loss": 3.1887,
+      "step": 852
+    },
+    {
+      "epoch": 11.37458193979933,
+      "grad_norm": 0.6470169425010681,
+      "learning_rate": 0.00016231368186874306,
+      "loss": 3.2537,
+      "step": 853
+    },
+    {
+      "epoch": 11.387959866220736,
+      "grad_norm": 0.863742470741272,
+      "learning_rate": 0.00016226918798665184,
+      "loss": 3.3443,
+      "step": 854
+    },
+    {
+      "epoch": 11.40133779264214,
+      "grad_norm": 0.7363939881324768,
+      "learning_rate": 0.00016222469410456062,
+      "loss": 3.3653,
+      "step": 855
+    },
+    {
+      "epoch": 11.414715719063546,
+      "grad_norm": 0.6548926830291748,
+      "learning_rate": 0.00016218020022246943,
+      "loss": 3.0373,
+      "step": 856
+    },
+    {
+      "epoch": 11.42809364548495,
+      "grad_norm": 0.8087872862815857,
+      "learning_rate": 0.00016213570634037821,
+      "loss": 3.0118,
+      "step": 857
+    },
+    {
+      "epoch": 11.441471571906355,
+      "grad_norm": 0.677811324596405,
+      "learning_rate": 0.000162091212458287,
+      "loss": 3.0339,
+      "step": 858
+    },
+    {
+      "epoch": 11.454849498327759,
+      "grad_norm": 0.6907945275306702,
+      "learning_rate": 0.00016204671857619578,
+      "loss": 2.9496,
+      "step": 859
+    },
+    {
+      "epoch": 11.468227424749164,
+      "grad_norm": 0.6940027475357056,
+      "learning_rate": 0.0001620022246941046,
+      "loss": 3.2825,
+      "step": 860
+    },
+    {
+      "epoch": 11.481605351170568,
+      "grad_norm": 0.7132136225700378,
+      "learning_rate": 0.00016195773081201334,
+      "loss": 3.1271,
+      "step": 861
+    },
+    {
+      "epoch": 11.494983277591974,
+      "grad_norm": 0.5997372269630432,
+      "learning_rate": 0.00016191323692992215,
+      "loss": 3.1292,
+      "step": 862
+    },
+    {
+      "epoch": 11.508361204013378,
+      "grad_norm": 0.6468494534492493,
+      "learning_rate": 0.00016186874304783094,
+      "loss": 3.32,
+      "step": 863
+    },
+    {
+      "epoch": 11.521739130434783,
+      "grad_norm": 0.5792532563209534,
+      "learning_rate": 0.00016182424916573972,
+      "loss": 3.3657,
+      "step": 864
+    },
+    {
+      "epoch": 11.535117056856187,
+      "grad_norm": 0.8242068290710449,
+      "learning_rate": 0.0001617797552836485,
+      "loss": 3.0094,
+      "step": 865
+    },
+    {
+      "epoch": 11.548494983277592,
+      "grad_norm": 0.9260333776473999,
+      "learning_rate": 0.0001617352614015573,
+      "loss": 3.1123,
+      "step": 866
+    },
+    {
+      "epoch": 11.561872909698996,
+      "grad_norm": 0.6337956786155701,
+      "learning_rate": 0.0001616907675194661,
+      "loss": 3.1761,
+      "step": 867
+    },
+    {
+      "epoch": 11.575250836120402,
+      "grad_norm": 0.6010364294052124,
+      "learning_rate": 0.00016164627363737487,
+      "loss": 3.2987,
+      "step": 868
+    },
+    {
+      "epoch": 11.588628762541806,
+      "grad_norm": 0.7492111921310425,
+      "learning_rate": 0.00016160177975528366,
+      "loss": 2.9085,
+      "step": 869
+    },
+    {
+      "epoch": 11.602006688963211,
+      "grad_norm": 0.6329553127288818,
+      "learning_rate": 0.00016155728587319247,
+      "loss": 3.1841,
+      "step": 870
+    },
+    {
+      "epoch": 11.615384615384615,
+      "grad_norm": 0.768527626991272,
+      "learning_rate": 0.00016151279199110122,
+      "loss": 3.0061,
+      "step": 871
+    },
+    {
+      "epoch": 11.62876254180602,
+      "grad_norm": 0.6333640813827515,
+      "learning_rate": 0.00016146829810901003,
+      "loss": 3.3046,
+      "step": 872
+    },
+    {
+      "epoch": 11.642140468227424,
+      "grad_norm": 0.7457571625709534,
+      "learning_rate": 0.00016142380422691881,
+      "loss": 3.2728,
+      "step": 873
+    },
+    {
+      "epoch": 11.65551839464883,
+      "grad_norm": 0.6389586925506592,
+      "learning_rate": 0.0001613793103448276,
+      "loss": 3.419,
+      "step": 874
+    },
+    {
+      "epoch": 11.668896321070234,
+      "grad_norm": 0.8885436058044434,
+      "learning_rate": 0.00016133481646273638,
+      "loss": 3.0938,
+      "step": 875
+    },
+    {
+      "epoch": 11.68227424749164,
+      "grad_norm": 0.7936431765556335,
+      "learning_rate": 0.00016129032258064516,
+      "loss": 3.305,
+      "step": 876
+    },
+    {
+      "epoch": 11.695652173913043,
+      "grad_norm": 0.6133994460105896,
+      "learning_rate": 0.00016124582869855397,
+      "loss": 3.3474,
+      "step": 877
+    },
+    {
+      "epoch": 11.709030100334449,
+      "grad_norm": 0.6638192534446716,
+      "learning_rate": 0.00016120133481646273,
+      "loss": 3.1418,
+      "step": 878
+    },
+    {
+      "epoch": 11.722408026755852,
+      "grad_norm": 0.6820496320724487,
+      "learning_rate": 0.00016115684093437154,
+      "loss": 3.1421,
+      "step": 879
+    },
+    {
+      "epoch": 11.735785953177258,
+      "grad_norm": 0.6057732105255127,
+      "learning_rate": 0.00016111234705228032,
+      "loss": 3.091,
+      "step": 880
+    },
+    {
+      "epoch": 11.749163879598662,
+      "grad_norm": 0.6267048716545105,
+      "learning_rate": 0.0001610678531701891,
+      "loss": 3.1289,
+      "step": 881
+    },
+    {
+      "epoch": 11.762541806020067,
+      "grad_norm": 0.6822847723960876,
+      "learning_rate": 0.00016102335928809788,
+      "loss": 3.157,
+      "step": 882
+    },
+    {
+      "epoch": 11.775919732441471,
+      "grad_norm": 0.6809714436531067,
+      "learning_rate": 0.0001609788654060067,
+      "loss": 3.0806,
+      "step": 883
+    },
+    {
+      "epoch": 11.789297658862877,
+      "grad_norm": 0.5546092391014099,
+      "learning_rate": 0.00016093437152391547,
+      "loss": 3.1853,
+      "step": 884
+    },
+    {
+      "epoch": 11.80267558528428,
+      "grad_norm": 0.7375029921531677,
+      "learning_rate": 0.00016088987764182426,
+      "loss": 3.1287,
+      "step": 885
+    },
+    {
+      "epoch": 11.816053511705686,
+      "grad_norm": 0.6246840953826904,
+      "learning_rate": 0.00016084538375973304,
+      "loss": 3.1331,
+      "step": 886
+    },
+    {
+      "epoch": 11.82943143812709,
+      "grad_norm": 0.6088026762008667,
+      "learning_rate": 0.00016080088987764185,
+      "loss": 3.3781,
+      "step": 887
+    },
+    {
+      "epoch": 11.842809364548495,
+      "grad_norm": 0.7996237874031067,
+      "learning_rate": 0.0001607563959955506,
+      "loss": 3.0161,
+      "step": 888
+    },
+    {
+      "epoch": 11.856187290969899,
+      "grad_norm": 0.6221441626548767,
+      "learning_rate": 0.0001607119021134594,
+      "loss": 3.1491,
+      "step": 889
+    },
+    {
+      "epoch": 11.869565217391305,
+      "grad_norm": 0.6276041269302368,
+      "learning_rate": 0.0001606674082313682,
+      "loss": 3.2575,
+      "step": 890
+    },
+    {
+      "epoch": 11.882943143812708,
+      "grad_norm": 0.6394500136375427,
+      "learning_rate": 0.00016062291434927698,
+      "loss": 3.3437,
+      "step": 891
+    },
+    {
+      "epoch": 11.896321070234114,
+      "grad_norm": 0.7674509286880493,
+      "learning_rate": 0.00016057842046718576,
+      "loss": 3.1995,
+      "step": 892
+    },
+    {
+      "epoch": 11.909698996655518,
+      "grad_norm": 0.7502215504646301,
+      "learning_rate": 0.00016053392658509457,
+      "loss": 3.3129,
+      "step": 893
+    },
+    {
+      "epoch": 11.923076923076923,
+      "grad_norm": 0.6078189611434937,
+      "learning_rate": 0.00016048943270300335,
+      "loss": 3.1623,
+      "step": 894
+    },
+    {
+      "epoch": 11.936454849498327,
+      "grad_norm": 0.6113708019256592,
+      "learning_rate": 0.00016044493882091213,
+      "loss": 3.6063,
+      "step": 895
+    },
+    {
+      "epoch": 11.949832775919733,
+      "grad_norm": 0.6606878638267517,
+      "learning_rate": 0.00016040044493882092,
+      "loss": 3.3216,
+      "step": 896
+    },
+    {
+      "epoch": 11.963210702341136,
+      "grad_norm": 0.7055956125259399,
+      "learning_rate": 0.00016035595105672973,
+      "loss": 3.2006,
+      "step": 897
+    },
+    {
+      "epoch": 11.976588628762542,
+      "grad_norm": 0.7424116730690002,
+      "learning_rate": 0.00016031145717463848,
+      "loss": 3.1298,
+      "step": 898
+    },
+    {
+      "epoch": 11.989966555183946,
+      "grad_norm": 0.6675695180892944,
+      "learning_rate": 0.0001602669632925473,
+      "loss": 3.1116,
+      "step": 899
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.9356181621551514,
+      "learning_rate": 0.00016022246941045607,
+      "loss": 3.2461,
+      "step": 900
+    },
+    {
+      "epoch": 12.013377926421406,
+      "grad_norm": 0.8539507985115051,
+      "learning_rate": 0.00016017797552836486,
+      "loss": 3.0671,
+      "step": 901
+    },
+    {
+      "epoch": 12.02675585284281,
+      "grad_norm": 0.573266327381134,
+      "learning_rate": 0.00016013348164627364,
+      "loss": 3.0789,
+      "step": 902
+    },
+    {
+      "epoch": 12.040133779264215,
+      "grad_norm": 0.5849746465682983,
+      "learning_rate": 0.00016008898776418245,
+      "loss": 3.1623,
+      "step": 903
+    },
+    {
+      "epoch": 12.053511705685619,
+      "grad_norm": 0.6523334980010986,
+      "learning_rate": 0.00016004449388209123,
+      "loss": 3.0243,
+      "step": 904
+    },
+    {
+      "epoch": 12.066889632107024,
+      "grad_norm": 0.6428223848342896,
+      "learning_rate": 0.00016,
+      "loss": 3.1241,
+      "step": 905
+    },
+    {
+      "epoch": 12.080267558528428,
+      "grad_norm": 0.5881937742233276,
+      "learning_rate": 0.0001599555061179088,
+      "loss": 3.1744,
+      "step": 906
+    },
+    {
+      "epoch": 12.093645484949834,
+      "grad_norm": 0.7523583173751831,
+      "learning_rate": 0.0001599110122358176,
+      "loss": 3.0821,
+      "step": 907
+    },
+    {
+      "epoch": 12.107023411371237,
+      "grad_norm": 0.6120390295982361,
+      "learning_rate": 0.00015986651835372636,
+      "loss": 3.16,
+      "step": 908
+    },
+    {
+      "epoch": 12.120401337792643,
+      "grad_norm": 0.6486253142356873,
+      "learning_rate": 0.00015982202447163517,
+      "loss": 3.1402,
+      "step": 909
+    },
+    {
+      "epoch": 12.133779264214047,
+      "grad_norm": 0.6322839260101318,
+      "learning_rate": 0.00015977753058954395,
+      "loss": 3.1325,
+      "step": 910
+    },
+    {
+      "epoch": 12.147157190635452,
+      "grad_norm": 0.5858875513076782,
+      "learning_rate": 0.00015973303670745273,
+      "loss": 3.1705,
+      "step": 911
+    },
+    {
+      "epoch": 12.160535117056856,
+      "grad_norm": 0.6495780348777771,
+      "learning_rate": 0.00015968854282536152,
+      "loss": 3.1977,
+      "step": 912
+    },
+    {
+      "epoch": 12.173913043478262,
+      "grad_norm": 0.6483474969863892,
+      "learning_rate": 0.00015964404894327033,
+      "loss": 2.9101,
+      "step": 913
+    },
+    {
+      "epoch": 12.187290969899665,
+      "grad_norm": 0.6021110415458679,
+      "learning_rate": 0.0001595995550611791,
+      "loss": 3.2106,
+      "step": 914
+    },
+    {
+      "epoch": 12.200668896321071,
+      "grad_norm": 0.5101630687713623,
+      "learning_rate": 0.00015955506117908786,
+      "loss": 3.1316,
+      "step": 915
+    },
+    {
+      "epoch": 12.214046822742475,
+      "grad_norm": 0.6226193904876709,
+      "learning_rate": 0.00015951056729699667,
+      "loss": 3.0897,
+      "step": 916
+    },
+    {
+      "epoch": 12.22742474916388,
+      "grad_norm": 0.6268473267555237,
+      "learning_rate": 0.00015946607341490546,
+      "loss": 3.2036,
+      "step": 917
+    },
+    {
+      "epoch": 12.240802675585284,
+      "grad_norm": 0.7825391292572021,
+      "learning_rate": 0.00015942157953281424,
+      "loss": 3.0082,
+      "step": 918
+    },
+    {
+      "epoch": 12.25418060200669,
+      "grad_norm": 0.7881148457527161,
+      "learning_rate": 0.00015937708565072302,
+      "loss": 3.0779,
+      "step": 919
+    },
+    {
+      "epoch": 12.267558528428093,
+      "grad_norm": 0.6725586652755737,
+      "learning_rate": 0.00015933259176863183,
+      "loss": 3.04,
+      "step": 920
+    },
+    {
+      "epoch": 12.280936454849499,
+      "grad_norm": 0.5831689238548279,
+      "learning_rate": 0.0001592880978865406,
+      "loss": 3.3319,
+      "step": 921
+    },
+    {
+      "epoch": 12.294314381270903,
+      "grad_norm": 0.6057907342910767,
+      "learning_rate": 0.0001592436040044494,
+      "loss": 3.0869,
+      "step": 922
+    },
+    {
+      "epoch": 12.307692307692308,
+      "grad_norm": 0.771857500076294,
+      "learning_rate": 0.00015919911012235818,
+      "loss": 3.1472,
+      "step": 923
+    },
+    {
+      "epoch": 12.321070234113712,
+      "grad_norm": 0.7447528839111328,
+      "learning_rate": 0.000159154616240267,
+      "loss": 3.174,
+      "step": 924
+    },
+    {
+      "epoch": 12.334448160535118,
+      "grad_norm": 0.5772632956504822,
+      "learning_rate": 0.00015911012235817574,
+      "loss": 3.1767,
+      "step": 925
+    },
+    {
+      "epoch": 12.347826086956522,
+      "grad_norm": 0.6952618956565857,
+      "learning_rate": 0.00015906562847608455,
+      "loss": 3.1597,
+      "step": 926
+    },
+    {
+      "epoch": 12.361204013377927,
+      "grad_norm": 0.600922703742981,
+      "learning_rate": 0.00015902113459399333,
+      "loss": 3.3612,
+      "step": 927
+    },
+    {
+      "epoch": 12.37458193979933,
+      "grad_norm": 0.7571472525596619,
+      "learning_rate": 0.00015897664071190212,
+      "loss": 2.9405,
+      "step": 928
+    },
+    {
+      "epoch": 12.387959866220736,
+      "grad_norm": 0.9343985915184021,
+      "learning_rate": 0.0001589321468298109,
+      "loss": 3.2886,
+      "step": 929
+    },
+    {
+      "epoch": 12.40133779264214,
+      "grad_norm": 0.7046729922294617,
+      "learning_rate": 0.0001588876529477197,
+      "loss": 3.4421,
+      "step": 930
+    },
+    {
+      "epoch": 12.414715719063546,
+      "grad_norm": 0.7591777443885803,
+      "learning_rate": 0.0001588431590656285,
+      "loss": 3.005,
+      "step": 931
+    },
+    {
+      "epoch": 12.42809364548495,
+      "grad_norm": 0.6508903503417969,
+      "learning_rate": 0.00015879866518353727,
+      "loss": 2.8554,
+      "step": 932
+    },
+    {
+      "epoch": 12.441471571906355,
+      "grad_norm": 0.6557784676551819,
+      "learning_rate": 0.00015875417130144606,
+      "loss": 3.3268,
+      "step": 933
+    },
+    {
+      "epoch": 12.454849498327759,
+      "grad_norm": 0.6941578984260559,
+      "learning_rate": 0.00015870967741935487,
+      "loss": 3.2088,
+      "step": 934
+    },
+    {
+      "epoch": 12.468227424749164,
+      "grad_norm": 0.6824263334274292,
+      "learning_rate": 0.00015866518353726362,
+      "loss": 3.0897,
+      "step": 935
+    },
+    {
+      "epoch": 12.481605351170568,
+      "grad_norm": 0.7324599027633667,
+      "learning_rate": 0.00015862068965517243,
+      "loss": 3.1443,
+      "step": 936
+    },
+    {
+      "epoch": 12.494983277591974,
+      "grad_norm": 0.577022135257721,
+      "learning_rate": 0.0001585761957730812,
+      "loss": 3.0896,
+      "step": 937
+    },
+    {
+      "epoch": 12.508361204013378,
+      "grad_norm": 0.6165060997009277,
+      "learning_rate": 0.00015853170189099,
+      "loss": 2.7546,
+      "step": 938
+    },
+    {
+      "epoch": 12.521739130434783,
+      "grad_norm": 0.561906635761261,
+      "learning_rate": 0.00015848720800889878,
+      "loss": 3.4192,
+      "step": 939
+    },
+    {
+      "epoch": 12.535117056856187,
+      "grad_norm": 0.5894923806190491,
+      "learning_rate": 0.0001584427141268076,
+      "loss": 3.0388,
+      "step": 940
+    },
+    {
+      "epoch": 12.548494983277592,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.00015839822024471637,
+      "loss": 3.0705,
+      "step": 941
+    },
+    {
+      "epoch": 12.561872909698996,
+      "grad_norm": 0.695101261138916,
+      "learning_rate": 0.00015835372636262515,
+      "loss": 3.1684,
+      "step": 942
+    },
+    {
+      "epoch": 12.575250836120402,
+      "grad_norm": 0.6176817417144775,
+      "learning_rate": 0.00015830923248053393,
+      "loss": 3.0708,
+      "step": 943
+    },
+    {
+      "epoch": 12.588628762541806,
+      "grad_norm": 0.6548507213592529,
+      "learning_rate": 0.00015826473859844274,
+      "loss": 3.1569,
+      "step": 944
+    },
+    {
+      "epoch": 12.602006688963211,
+      "grad_norm": 0.6046382188796997,
+      "learning_rate": 0.0001582202447163515,
+      "loss": 3.2479,
+      "step": 945
+    },
+    {
+      "epoch": 12.615384615384615,
+      "grad_norm": 0.7103912234306335,
+      "learning_rate": 0.0001581757508342603,
+      "loss": 3.127,
+      "step": 946
+    },
+    {
+      "epoch": 12.62876254180602,
+      "grad_norm": 0.7131765484809875,
+      "learning_rate": 0.0001581312569521691,
+      "loss": 3.0975,
+      "step": 947
+    },
+    {
+      "epoch": 12.642140468227424,
+      "grad_norm": 0.6442859768867493,
+      "learning_rate": 0.00015808676307007787,
+      "loss": 3.2885,
+      "step": 948
+    },
+    {
+      "epoch": 12.65551839464883,
+      "grad_norm": 0.6430050134658813,
+      "learning_rate": 0.00015804226918798666,
+      "loss": 3.0397,
+      "step": 949
+    },
+    {
+      "epoch": 12.668896321070234,
+      "grad_norm": 0.6894303560256958,
+      "learning_rate": 0.00015799777530589546,
+      "loss": 3.0756,
+      "step": 950
+    },
+    {
+      "epoch": 12.68227424749164,
+      "grad_norm": 0.7319600582122803,
+      "learning_rate": 0.00015795328142380425,
+      "loss": 3.0457,
+      "step": 951
+    },
+    {
+      "epoch": 12.695652173913043,
+      "grad_norm": 0.6445140838623047,
+      "learning_rate": 0.00015790878754171303,
+      "loss": 2.9828,
+      "step": 952
+    },
+    {
+      "epoch": 12.709030100334449,
+      "grad_norm": 0.7522070407867432,
+      "learning_rate": 0.0001578642936596218,
+      "loss": 2.8942,
+      "step": 953
+    },
+    {
+      "epoch": 12.722408026755852,
+      "grad_norm": 0.7962691783905029,
+      "learning_rate": 0.00015781979977753062,
+      "loss": 3.0985,
+      "step": 954
+    },
+    {
+      "epoch": 12.735785953177258,
+      "grad_norm": 0.6391687393188477,
+      "learning_rate": 0.00015777530589543938,
+      "loss": 3.1752,
+      "step": 955
+    },
+    {
+      "epoch": 12.749163879598662,
+      "grad_norm": 0.7632976174354553,
+      "learning_rate": 0.00015773081201334816,
+      "loss": 3.3505,
+      "step": 956
+    },
+    {
+      "epoch": 12.762541806020067,
+      "grad_norm": 0.7491022944450378,
+      "learning_rate": 0.00015768631813125697,
+      "loss": 3.0721,
+      "step": 957
+    },
+    {
+      "epoch": 12.775919732441471,
+      "grad_norm": 0.6163421273231506,
+      "learning_rate": 0.00015764182424916572,
+      "loss": 3.3242,
+      "step": 958
+    },
+    {
+      "epoch": 12.789297658862877,
+      "grad_norm": 0.6831198334693909,
+      "learning_rate": 0.00015759733036707453,
+      "loss": 3.1409,
+      "step": 959
+    },
+    {
+      "epoch": 12.80267558528428,
+      "grad_norm": 0.812300980091095,
+      "learning_rate": 0.00015755283648498332,
+      "loss": 2.9606,
+      "step": 960
+    },
+    {
+      "epoch": 12.816053511705686,
+      "grad_norm": 0.6904334425926208,
+      "learning_rate": 0.00015750834260289213,
+      "loss": 3.0398,
+      "step": 961
+    },
+    {
+      "epoch": 12.82943143812709,
+      "grad_norm": 0.6349720358848572,
+      "learning_rate": 0.00015746384872080088,
+      "loss": 3.1033,
+      "step": 962
+    },
+    {
+      "epoch": 12.842809364548495,
+      "grad_norm": 0.6837566494941711,
+      "learning_rate": 0.0001574193548387097,
+      "loss": 3.2353,
+      "step": 963
+    },
+    {
+      "epoch": 12.856187290969899,
+      "grad_norm": 0.5852749943733215,
+      "learning_rate": 0.00015737486095661847,
+      "loss": 3.0972,
+      "step": 964
+    },
+    {
+      "epoch": 12.869565217391305,
+      "grad_norm": 0.6641372442245483,
+      "learning_rate": 0.00015733036707452725,
+      "loss": 3.2243,
+      "step": 965
+    },
+    {
+      "epoch": 12.882943143812708,
+      "grad_norm": 0.6613900065422058,
+      "learning_rate": 0.00015728587319243604,
+      "loss": 3.1263,
+      "step": 966
+    },
+    {
+      "epoch": 12.896321070234114,
+      "grad_norm": 0.6126120090484619,
+      "learning_rate": 0.00015724137931034485,
+      "loss": 3.069,
+      "step": 967
+    },
+    {
+      "epoch": 12.909698996655518,
+      "grad_norm": 0.6764604449272156,
+      "learning_rate": 0.0001571968854282536,
+      "loss": 3.1397,
+      "step": 968
+    },
+    {
+      "epoch": 12.923076923076923,
+      "grad_norm": 0.6447578072547913,
+      "learning_rate": 0.0001571523915461624,
+      "loss": 3.1839,
+      "step": 969
+    },
+    {
+      "epoch": 12.936454849498327,
+      "grad_norm": 0.5872016549110413,
+      "learning_rate": 0.0001571078976640712,
+      "loss": 3.3144,
+      "step": 970
+    },
+    {
+      "epoch": 12.949832775919733,
+      "grad_norm": 0.626276969909668,
+      "learning_rate": 0.00015706340378198,
+      "loss": 3.1295,
+      "step": 971
+    },
+    {
+      "epoch": 12.963210702341136,
+      "grad_norm": 0.6829231381416321,
+      "learning_rate": 0.00015701890989988876,
+      "loss": 3.2261,
+      "step": 972
+    },
+    {
+      "epoch": 12.976588628762542,
+      "grad_norm": 0.6197345852851868,
+      "learning_rate": 0.00015697441601779757,
+      "loss": 3.1117,
+      "step": 973
+    },
+    {
+      "epoch": 12.989966555183946,
+      "grad_norm": 0.6137062907218933,
+      "learning_rate": 0.00015692992213570635,
+      "loss": 3.1548,
+      "step": 974
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.7483121752738953,
+      "learning_rate": 0.00015688542825361513,
+      "loss": 3.4944,
+      "step": 975
+    },
+    {
+      "epoch": 13.013377926421406,
+      "grad_norm": 0.6102525591850281,
+      "learning_rate": 0.00015684093437152392,
+      "loss": 3.1403,
+      "step": 976
+    },
+    {
+      "epoch": 13.02675585284281,
+      "grad_norm": 0.7258747220039368,
+      "learning_rate": 0.00015679644048943272,
+      "loss": 2.8474,
+      "step": 977
+    },
+    {
+      "epoch": 13.040133779264215,
+      "grad_norm": 0.6891087293624878,
+      "learning_rate": 0.00015675194660734148,
+      "loss": 2.9722,
+      "step": 978
+    },
+    {
+      "epoch": 13.053511705685619,
+      "grad_norm": 0.6320910453796387,
+      "learning_rate": 0.0001567074527252503,
+      "loss": 2.9807,
+      "step": 979
+    },
+    {
+      "epoch": 13.066889632107024,
+      "grad_norm": 0.8684266209602356,
+      "learning_rate": 0.00015666295884315907,
+      "loss": 2.8448,
+      "step": 980
+    },
+    {
+      "epoch": 13.080267558528428,
+      "grad_norm": 0.7126099467277527,
+      "learning_rate": 0.00015661846496106788,
+      "loss": 3.1915,
+      "step": 981
+    },
+    {
+      "epoch": 13.093645484949834,
+      "grad_norm": 0.7001529335975647,
+      "learning_rate": 0.00015657397107897664,
+      "loss": 3.2017,
+      "step": 982
+    },
+    {
+      "epoch": 13.107023411371237,
+      "grad_norm": 0.7901191711425781,
+      "learning_rate": 0.00015652947719688545,
+      "loss": 3.1273,
+      "step": 983
+    },
+    {
+      "epoch": 13.120401337792643,
+      "grad_norm": 0.5769410729408264,
+      "learning_rate": 0.00015648498331479423,
+      "loss": 2.8582,
+      "step": 984
+    },
+    {
+      "epoch": 13.133779264214047,
+      "grad_norm": 0.5969700813293457,
+      "learning_rate": 0.000156440489432703,
+      "loss": 3.118,
+      "step": 985
+    },
+    {
+      "epoch": 13.147157190635452,
+      "grad_norm": 0.5789377093315125,
+      "learning_rate": 0.0001563959955506118,
+      "loss": 2.8121,
+      "step": 986
+    },
+    {
+      "epoch": 13.160535117056856,
+      "grad_norm": 0.7945278882980347,
+      "learning_rate": 0.0001563515016685206,
+      "loss": 2.9212,
+      "step": 987
+    },
+    {
+      "epoch": 13.173913043478262,
+      "grad_norm": 0.60884690284729,
+      "learning_rate": 0.00015630700778642936,
+      "loss": 3.1993,
+      "step": 988
+    },
+    {
+      "epoch": 13.187290969899665,
+      "grad_norm": 0.616142213344574,
+      "learning_rate": 0.00015626251390433817,
+      "loss": 3.107,
+      "step": 989
+    },
+    {
+      "epoch": 13.200668896321071,
+      "grad_norm": 0.6428812146186829,
+      "learning_rate": 0.00015621802002224695,
+      "loss": 3.1401,
+      "step": 990
+    },
+    {
+      "epoch": 13.214046822742475,
+      "grad_norm": 0.5723693370819092,
+      "learning_rate": 0.00015617352614015576,
+      "loss": 3.0474,
+      "step": 991
+    },
+    {
+      "epoch": 13.22742474916388,
+      "grad_norm": 0.5820907950401306,
+      "learning_rate": 0.00015612903225806451,
+      "loss": 3.2211,
+      "step": 992
+    },
+    {
+      "epoch": 13.240802675585284,
+      "grad_norm": 0.555957555770874,
+      "learning_rate": 0.00015608453837597332,
+      "loss": 3.1306,
+      "step": 993
+    },
+    {
+      "epoch": 13.25418060200669,
+      "grad_norm": 0.528698205947876,
+      "learning_rate": 0.0001560400444938821,
+      "loss": 3.0989,
+      "step": 994
+    },
+    {
+      "epoch": 13.267558528428093,
+      "grad_norm": 0.5959749817848206,
+      "learning_rate": 0.0001559955506117909,
+      "loss": 2.9128,
+      "step": 995
+    },
+    {
+      "epoch": 13.280936454849499,
+      "grad_norm": 0.6702240705490112,
+      "learning_rate": 0.00015595105672969967,
+      "loss": 2.9969,
+      "step": 996
+    },
+    {
+      "epoch": 13.294314381270903,
+      "grad_norm": 0.5363825559616089,
+      "learning_rate": 0.00015590656284760845,
+      "loss": 3.2027,
+      "step": 997
+    },
+    {
+      "epoch": 13.307692307692308,
+      "grad_norm": 0.6402661204338074,
+      "learning_rate": 0.00015586206896551724,
+      "loss": 3.2733,
+      "step": 998
+    },
+    {
+      "epoch": 13.321070234113712,
+      "grad_norm": 0.700517475605011,
+      "learning_rate": 0.00015581757508342602,
+      "loss": 3.2629,
+      "step": 999
+    },
+    {
+      "epoch": 13.334448160535118,
+      "grad_norm": 0.5500949621200562,
+      "learning_rate": 0.00015577308120133483,
+      "loss": 3.172,
+      "step": 1000
+    },
+    {
+      "epoch": 13.347826086956522,
+      "grad_norm": 0.6199147701263428,
+      "learning_rate": 0.0001557285873192436,
+      "loss": 3.0556,
+      "step": 1001
+    },
+    {
+      "epoch": 13.361204013377927,
+      "grad_norm": 0.5900529026985168,
+      "learning_rate": 0.0001556840934371524,
+      "loss": 3.1248,
+      "step": 1002
+    },
+    {
+      "epoch": 13.37458193979933,
+      "grad_norm": 0.7272413372993469,
+      "learning_rate": 0.00015563959955506118,
+      "loss": 3.1142,
+      "step": 1003
+    },
+    {
+      "epoch": 13.387959866220736,
+      "grad_norm": 0.6461951732635498,
+      "learning_rate": 0.00015559510567296998,
+      "loss": 3.1145,
+      "step": 1004
+    },
+    {
+      "epoch": 13.40133779264214,
+      "grad_norm": 0.5750373005867004,
+      "learning_rate": 0.00015555061179087874,
+      "loss": 3.2117,
+      "step": 1005
+    },
+    {
+      "epoch": 13.414715719063546,
+      "grad_norm": 0.6486302614212036,
+      "learning_rate": 0.00015550611790878755,
+      "loss": 2.9018,
+      "step": 1006
+    },
+    {
+      "epoch": 13.42809364548495,
+      "grad_norm": 0.6897476315498352,
+      "learning_rate": 0.00015546162402669633,
+      "loss": 3.2987,
+      "step": 1007
+    },
+    {
+      "epoch": 13.441471571906355,
+      "grad_norm": 0.5997576713562012,
+      "learning_rate": 0.00015541713014460511,
+      "loss": 2.974,
+      "step": 1008
+    },
+    {
+      "epoch": 13.454849498327759,
+      "grad_norm": 0.6484793424606323,
+      "learning_rate": 0.0001553726362625139,
+      "loss": 2.9827,
+      "step": 1009
+    },
+    {
+      "epoch": 13.468227424749164,
+      "grad_norm": 0.562312126159668,
+      "learning_rate": 0.0001553281423804227,
+      "loss": 3.2243,
+      "step": 1010
+    },
+    {
+      "epoch": 13.481605351170568,
+      "grad_norm": 0.7257137298583984,
+      "learning_rate": 0.0001552836484983315,
+      "loss": 3.1081,
+      "step": 1011
+    },
+    {
+      "epoch": 13.494983277591974,
+      "grad_norm": 0.7201404571533203,
+      "learning_rate": 0.00015523915461624027,
+      "loss": 3.1468,
+      "step": 1012
+    },
+    {
+      "epoch": 13.508361204013378,
+      "grad_norm": 0.666539192199707,
+      "learning_rate": 0.00015519466073414905,
+      "loss": 3.0081,
+      "step": 1013
+    },
+    {
+      "epoch": 13.521739130434783,
+      "grad_norm": 0.6867642998695374,
+      "learning_rate": 0.00015515016685205786,
+      "loss": 2.7889,
+      "step": 1014
+    },
+    {
+      "epoch": 13.535117056856187,
+      "grad_norm": 0.5799785256385803,
+      "learning_rate": 0.00015510567296996662,
+      "loss": 3.2995,
+      "step": 1015
+    },
+    {
+      "epoch": 13.548494983277592,
+      "grad_norm": 0.6155371069908142,
+      "learning_rate": 0.00015506117908787543,
+      "loss": 3.2227,
+      "step": 1016
+    },
+    {
+      "epoch": 13.561872909698996,
+      "grad_norm": 0.7604040503501892,
+      "learning_rate": 0.0001550166852057842,
+      "loss": 3.0384,
+      "step": 1017
+    },
+    {
+      "epoch": 13.575250836120402,
+      "grad_norm": 0.8445917963981628,
+      "learning_rate": 0.000154972191323693,
+      "loss": 2.9566,
+      "step": 1018
+    },
+    {
+      "epoch": 13.588628762541806,
+      "grad_norm": 0.7978566288948059,
+      "learning_rate": 0.00015492769744160177,
+      "loss": 3.0175,
+      "step": 1019
+    },
+    {
+      "epoch": 13.602006688963211,
+      "grad_norm": 0.5899437069892883,
+      "learning_rate": 0.00015488320355951058,
+      "loss": 3.2418,
+      "step": 1020
+    },
+    {
+      "epoch": 13.615384615384615,
+      "grad_norm": 0.7204627990722656,
+      "learning_rate": 0.00015483870967741937,
+      "loss": 3.1568,
+      "step": 1021
+    },
+    {
+      "epoch": 13.62876254180602,
+      "grad_norm": 0.6504855155944824,
+      "learning_rate": 0.00015479421579532815,
+      "loss": 2.8952,
+      "step": 1022
+    },
+    {
+      "epoch": 13.642140468227424,
+      "grad_norm": 0.8101251125335693,
+      "learning_rate": 0.00015474972191323693,
+      "loss": 3.1975,
+      "step": 1023
+    },
+    {
+      "epoch": 13.65551839464883,
+      "grad_norm": 0.6161416172981262,
+      "learning_rate": 0.00015470522803114574,
+      "loss": 3.1565,
+      "step": 1024
+    },
+    {
+      "epoch": 13.668896321070234,
+      "grad_norm": 0.6131258606910706,
+      "learning_rate": 0.0001546607341490545,
+      "loss": 3.0382,
+      "step": 1025
+    },
+    {
+      "epoch": 13.68227424749164,
+      "grad_norm": 0.8008583784103394,
+      "learning_rate": 0.0001546162402669633,
+      "loss": 2.995,
+      "step": 1026
+    },
+    {
+      "epoch": 13.695652173913043,
+      "grad_norm": 0.7101227045059204,
+      "learning_rate": 0.0001545717463848721,
+      "loss": 3.0704,
+      "step": 1027
+    },
+    {
+      "epoch": 13.709030100334449,
+      "grad_norm": 0.7988458275794983,
+      "learning_rate": 0.00015452725250278087,
+      "loss": 3.1424,
+      "step": 1028
+    },
+    {
+      "epoch": 13.722408026755852,
+      "grad_norm": 0.6013655662536621,
+      "learning_rate": 0.00015448275862068965,
+      "loss": 2.9986,
+      "step": 1029
+    },
+    {
+      "epoch": 13.735785953177258,
+      "grad_norm": 0.6368236541748047,
+      "learning_rate": 0.00015443826473859846,
+      "loss": 3.0193,
+      "step": 1030
+    },
+    {
+      "epoch": 13.749163879598662,
+      "grad_norm": 0.8222694396972656,
+      "learning_rate": 0.00015439377085650724,
+      "loss": 3.0888,
+      "step": 1031
+    },
+    {
+      "epoch": 13.762541806020067,
+      "grad_norm": 0.7270404696464539,
+      "learning_rate": 0.00015434927697441603,
+      "loss": 3.0929,
+      "step": 1032
+    },
+    {
+      "epoch": 13.775919732441471,
+      "grad_norm": 0.7292355298995972,
+      "learning_rate": 0.0001543047830923248,
+      "loss": 2.7676,
+      "step": 1033
+    },
+    {
+      "epoch": 13.789297658862877,
+      "grad_norm": 0.6662157773971558,
+      "learning_rate": 0.00015426028921023362,
+      "loss": 3.1984,
+      "step": 1034
+    },
+    {
+      "epoch": 13.80267558528428,
+      "grad_norm": 0.6350163817405701,
+      "learning_rate": 0.00015421579532814237,
+      "loss": 3.3042,
+      "step": 1035
+    },
+    {
+      "epoch": 13.816053511705686,
+      "grad_norm": 0.5999907851219177,
+      "learning_rate": 0.00015417130144605118,
+      "loss": 3.0983,
+      "step": 1036
+    },
+    {
+      "epoch": 13.82943143812709,
+      "grad_norm": 0.5942257642745972,
+      "learning_rate": 0.00015412680756395997,
+      "loss": 3.0474,
+      "step": 1037
+    },
+    {
+      "epoch": 13.842809364548495,
+      "grad_norm": 0.662589430809021,
+      "learning_rate": 0.00015408231368186875,
+      "loss": 2.9251,
+      "step": 1038
+    },
+    {
+      "epoch": 13.856187290969899,
+      "grad_norm": 0.5817089080810547,
+      "learning_rate": 0.00015403781979977753,
+      "loss": 3.0716,
+      "step": 1039
+    },
+    {
+      "epoch": 13.869565217391305,
+      "grad_norm": 0.6019257307052612,
+      "learning_rate": 0.0001539933259176863,
+      "loss": 3.1754,
+      "step": 1040
+    },
+    {
+      "epoch": 13.882943143812708,
+      "grad_norm": 0.6301860213279724,
+      "learning_rate": 0.00015394883203559512,
+      "loss": 3.2066,
+      "step": 1041
+    },
+    {
+      "epoch": 13.896321070234114,
+      "grad_norm": 0.6468888521194458,
+      "learning_rate": 0.00015390433815350388,
+      "loss": 3.3001,
+      "step": 1042
+    },
+    {
+      "epoch": 13.909698996655518,
+      "grad_norm": 0.6510801911354065,
+      "learning_rate": 0.0001538598442714127,
+      "loss": 3.2198,
+      "step": 1043
+    },
+    {
+      "epoch": 13.923076923076923,
+      "grad_norm": 0.5692014694213867,
+      "learning_rate": 0.00015381535038932147,
+      "loss": 3.3257,
+      "step": 1044
+    },
+    {
+      "epoch": 13.936454849498327,
+      "grad_norm": 0.594219982624054,
+      "learning_rate": 0.00015377085650723025,
+      "loss": 2.9918,
+      "step": 1045
+    },
+    {
+      "epoch": 13.949832775919733,
+      "grad_norm": 0.6501769423484802,
+      "learning_rate": 0.00015372636262513903,
+      "loss": 2.9653,
+      "step": 1046
+    },
+    {
+      "epoch": 13.963210702341136,
+      "grad_norm": 0.6310623288154602,
+      "learning_rate": 0.00015368186874304784,
+      "loss": 3.19,
+      "step": 1047
+    },
+    {
+      "epoch": 13.976588628762542,
+      "grad_norm": 0.5795436501502991,
+      "learning_rate": 0.00015363737486095663,
+      "loss": 3.272,
+      "step": 1048
+    },
+    {
+      "epoch": 13.989966555183946,
+      "grad_norm": 0.5421392917633057,
+      "learning_rate": 0.0001535928809788654,
+      "loss": 3.2109,
+      "step": 1049
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.833959698677063,
+      "learning_rate": 0.0001535483870967742,
+      "loss": 3.3017,
+      "step": 1050
+    },
+    {
+      "epoch": 14.013377926421406,
+      "grad_norm": 0.6856208443641663,
+      "learning_rate": 0.000153503893214683,
+      "loss": 3.0784,
+      "step": 1051
+    },
+    {
+      "epoch": 14.02675585284281,
+      "grad_norm": 0.5841811895370483,
+      "learning_rate": 0.00015345939933259176,
+      "loss": 2.848,
+      "step": 1052
+    },
+    {
+      "epoch": 14.040133779264215,
+      "grad_norm": 0.557906985282898,
+      "learning_rate": 0.00015341490545050057,
+      "loss": 3.1564,
+      "step": 1053
+    },
+    {
+      "epoch": 14.053511705685619,
+      "grad_norm": 0.5468619465827942,
+      "learning_rate": 0.00015337041156840935,
+      "loss": 3.1237,
+      "step": 1054
+    },
+    {
+      "epoch": 14.066889632107024,
+      "grad_norm": 0.7213225364685059,
+      "learning_rate": 0.00015332591768631813,
+      "loss": 2.8993,
+      "step": 1055
+    },
+    {
+      "epoch": 14.080267558528428,
+      "grad_norm": 0.7413175106048584,
+      "learning_rate": 0.0001532814238042269,
+      "loss": 3.0028,
+      "step": 1056
+    },
+    {
+      "epoch": 14.093645484949834,
+      "grad_norm": 0.6072244644165039,
+      "learning_rate": 0.00015323692992213572,
+      "loss": 2.8534,
+      "step": 1057
+    },
+    {
+      "epoch": 14.107023411371237,
+      "grad_norm": 0.683262288570404,
+      "learning_rate": 0.0001531924360400445,
+      "loss": 2.8982,
+      "step": 1058
+    },
+    {
+      "epoch": 14.120401337792643,
+      "grad_norm": 0.5880157351493835,
+      "learning_rate": 0.0001531479421579533,
+      "loss": 2.9567,
+      "step": 1059
+    },
+    {
+      "epoch": 14.133779264214047,
+      "grad_norm": 0.7519298195838928,
+      "learning_rate": 0.00015310344827586207,
+      "loss": 2.9615,
+      "step": 1060
+    },
+    {
+      "epoch": 14.147157190635452,
+      "grad_norm": 0.7747945189476013,
+      "learning_rate": 0.00015305895439377088,
+      "loss": 2.7823,
+      "step": 1061
+    },
+    {
+      "epoch": 14.160535117056856,
+      "grad_norm": 0.6560395956039429,
+      "learning_rate": 0.00015301446051167963,
+      "loss": 3.1791,
+      "step": 1062
+    },
+    {
+      "epoch": 14.173913043478262,
+      "grad_norm": 0.6388076543807983,
+      "learning_rate": 0.00015296996662958844,
+      "loss": 2.9868,
+      "step": 1063
+    },
+    {
+      "epoch": 14.187290969899665,
+      "grad_norm": 0.7349525690078735,
+      "learning_rate": 0.00015292547274749723,
+      "loss": 2.9291,
+      "step": 1064
+    },
+    {
+      "epoch": 14.200668896321071,
+      "grad_norm": 0.7184433341026306,
+      "learning_rate": 0.000152880978865406,
+      "loss": 2.7889,
+      "step": 1065
+    },
+    {
+      "epoch": 14.214046822742475,
+      "grad_norm": 0.6776930093765259,
+      "learning_rate": 0.0001528364849833148,
+      "loss": 3.1572,
+      "step": 1066
+    },
+    {
+      "epoch": 14.22742474916388,
+      "grad_norm": 0.818756103515625,
+      "learning_rate": 0.0001527919911012236,
+      "loss": 2.767,
+      "step": 1067
+    },
+    {
+      "epoch": 14.240802675585284,
+      "grad_norm": 0.6005066633224487,
+      "learning_rate": 0.00015274749721913238,
+      "loss": 3.1796,
+      "step": 1068
+    },
+    {
+      "epoch": 14.25418060200669,
+      "grad_norm": 0.6367926001548767,
+      "learning_rate": 0.00015270300333704117,
+      "loss": 2.9999,
+      "step": 1069
+    },
+    {
+      "epoch": 14.267558528428093,
+      "grad_norm": 0.6823679208755493,
+      "learning_rate": 0.00015265850945494995,
+      "loss": 2.7663,
+      "step": 1070
+    },
+    {
+      "epoch": 14.280936454849499,
+      "grad_norm": 0.6238808631896973,
+      "learning_rate": 0.00015261401557285876,
+      "loss": 3.1908,
+      "step": 1071
+    },
+    {
+      "epoch": 14.294314381270903,
+      "grad_norm": 0.6983721256256104,
+      "learning_rate": 0.0001525695216907675,
+      "loss": 3.0264,
+      "step": 1072
+    },
+    {
+      "epoch": 14.307692307692308,
+      "grad_norm": 0.7568501234054565,
+      "learning_rate": 0.00015252502780867632,
+      "loss": 3.0474,
+      "step": 1073
+    },
+    {
+      "epoch": 14.321070234113712,
+      "grad_norm": 0.6250051259994507,
+      "learning_rate": 0.0001524805339265851,
+      "loss": 3.114,
+      "step": 1074
+    },
+    {
+      "epoch": 14.334448160535118,
+      "grad_norm": 0.5907386541366577,
+      "learning_rate": 0.0001524360400444939,
+      "loss": 3.1823,
+      "step": 1075
+    },
+    {
+      "epoch": 14.347826086956522,
+      "grad_norm": 0.6719332337379456,
+      "learning_rate": 0.00015239154616240267,
+      "loss": 2.994,
+      "step": 1076
+    },
+    {
+      "epoch": 14.361204013377927,
+      "grad_norm": 0.5911534428596497,
+      "learning_rate": 0.00015234705228031148,
+      "loss": 3.4178,
+      "step": 1077
+    },
+    {
+      "epoch": 14.37458193979933,
+      "grad_norm": 0.8071689009666443,
+      "learning_rate": 0.00015230255839822026,
+      "loss": 2.9633,
+      "step": 1078
+    },
+    {
+      "epoch": 14.387959866220736,
+      "grad_norm": 0.5957038998603821,
+      "learning_rate": 0.00015225806451612902,
+      "loss": 3.1126,
+      "step": 1079
+    },
+    {
+      "epoch": 14.40133779264214,
+      "grad_norm": 0.6604459285736084,
+      "learning_rate": 0.00015221357063403783,
+      "loss": 2.9193,
+      "step": 1080
+    },
+    {
+      "epoch": 14.414715719063546,
+      "grad_norm": 0.626081109046936,
+      "learning_rate": 0.0001521690767519466,
+      "loss": 2.9321,
+      "step": 1081
+    },
+    {
+      "epoch": 14.42809364548495,
+      "grad_norm": 0.5767174959182739,
+      "learning_rate": 0.0001521245828698554,
+      "loss": 3.1518,
+      "step": 1082
+    },
+    {
+      "epoch": 14.441471571906355,
+      "grad_norm": 0.6444874405860901,
+      "learning_rate": 0.00015208008898776417,
+      "loss": 3.1948,
+      "step": 1083
+    },
+    {
+      "epoch": 14.454849498327759,
+      "grad_norm": 0.668171763420105,
+      "learning_rate": 0.00015203559510567298,
+      "loss": 3.016,
+      "step": 1084
+    },
+    {
+      "epoch": 14.468227424749164,
+      "grad_norm": 0.6998944878578186,
+      "learning_rate": 0.00015199110122358176,
+      "loss": 2.9558,
+      "step": 1085
+    },
+    {
+      "epoch": 14.481605351170568,
+      "grad_norm": 0.5896235704421997,
+      "learning_rate": 0.00015194660734149055,
+      "loss": 3.0997,
+      "step": 1086
+    },
+    {
+      "epoch": 14.494983277591974,
+      "grad_norm": 0.6724826097488403,
+      "learning_rate": 0.00015190211345939933,
+      "loss": 2.9369,
+      "step": 1087
+    },
+    {
+      "epoch": 14.508361204013378,
+      "grad_norm": 0.5710486769676208,
+      "learning_rate": 0.00015185761957730814,
+      "loss": 3.4724,
+      "step": 1088
+    },
+    {
+      "epoch": 14.521739130434783,
+      "grad_norm": 0.9997962117195129,
+      "learning_rate": 0.0001518131256952169,
+      "loss": 2.9058,
+      "step": 1089
+    },
+    {
+      "epoch": 14.535117056856187,
+      "grad_norm": 0.668074905872345,
+      "learning_rate": 0.0001517686318131257,
+      "loss": 3.2401,
+      "step": 1090
+    },
+    {
+      "epoch": 14.548494983277592,
+      "grad_norm": 0.6180433630943298,
+      "learning_rate": 0.00015172413793103449,
+      "loss": 2.8975,
+      "step": 1091
+    },
+    {
+      "epoch": 14.561872909698996,
+      "grad_norm": 0.6412661075592041,
+      "learning_rate": 0.00015167964404894327,
+      "loss": 2.9882,
+      "step": 1092
+    },
+    {
+      "epoch": 14.575250836120402,
+      "grad_norm": 0.715288519859314,
+      "learning_rate": 0.00015163515016685205,
+      "loss": 3.4004,
+      "step": 1093
+    },
+    {
+      "epoch": 14.588628762541806,
+      "grad_norm": 0.689164400100708,
+      "learning_rate": 0.00015159065628476086,
+      "loss": 3.0705,
+      "step": 1094
+    },
+    {
+      "epoch": 14.602006688963211,
+      "grad_norm": 0.7713497281074524,
+      "learning_rate": 0.00015154616240266964,
+      "loss": 3.159,
+      "step": 1095
+    },
+    {
+      "epoch": 14.615384615384615,
+      "grad_norm": 0.6725841164588928,
+      "learning_rate": 0.00015150166852057843,
+      "loss": 2.8938,
+      "step": 1096
+    },
+    {
+      "epoch": 14.62876254180602,
+      "grad_norm": 0.658108651638031,
+      "learning_rate": 0.0001514571746384872,
+      "loss": 2.8747,
+      "step": 1097
+    },
+    {
+      "epoch": 14.642140468227424,
+      "grad_norm": 0.5711888074874878,
+      "learning_rate": 0.00015141268075639602,
+      "loss": 2.8989,
+      "step": 1098
+    },
+    {
+      "epoch": 14.65551839464883,
+      "grad_norm": 0.6184161305427551,
+      "learning_rate": 0.00015136818687430477,
+      "loss": 3.0904,
+      "step": 1099
+    },
+    {
+      "epoch": 14.668896321070234,
+      "grad_norm": 0.5937799215316772,
+      "learning_rate": 0.00015132369299221358,
+      "loss": 3.1637,
+      "step": 1100
+    },
+    {
+      "epoch": 14.68227424749164,
+      "grad_norm": 0.591673731803894,
+      "learning_rate": 0.00015127919911012236,
+      "loss": 2.9547,
+      "step": 1101
+    },
+    {
+      "epoch": 14.695652173913043,
+      "grad_norm": 0.7317401170730591,
+      "learning_rate": 0.00015123470522803115,
+      "loss": 3.2043,
+      "step": 1102
+    },
+    {
+      "epoch": 14.709030100334449,
+      "grad_norm": 0.5784003734588623,
+      "learning_rate": 0.00015119021134593993,
+      "loss": 3.102,
+      "step": 1103
+    },
+    {
+      "epoch": 14.722408026755852,
+      "grad_norm": 0.7077385187149048,
+      "learning_rate": 0.00015114571746384874,
+      "loss": 3.011,
+      "step": 1104
+    },
+    {
+      "epoch": 14.735785953177258,
+      "grad_norm": 0.6472675204277039,
+      "learning_rate": 0.00015110122358175752,
+      "loss": 3.2075,
+      "step": 1105
+    },
+    {
+      "epoch": 14.749163879598662,
+      "grad_norm": 0.6789306998252869,
+      "learning_rate": 0.0001510567296996663,
+      "loss": 2.9458,
+      "step": 1106
+    },
+    {
+      "epoch": 14.762541806020067,
+      "grad_norm": 0.6602732539176941,
+      "learning_rate": 0.00015101223581757509,
+      "loss": 2.9941,
+      "step": 1107
+    },
+    {
+      "epoch": 14.775919732441471,
+      "grad_norm": 0.7484832406044006,
+      "learning_rate": 0.0001509677419354839,
+      "loss": 3.0358,
+      "step": 1108
+    },
+    {
+      "epoch": 14.789297658862877,
+      "grad_norm": 0.704139769077301,
+      "learning_rate": 0.00015092324805339265,
+      "loss": 3.1006,
+      "step": 1109
+    },
+    {
+      "epoch": 14.80267558528428,
+      "grad_norm": 0.6545978784561157,
+      "learning_rate": 0.00015087875417130146,
+      "loss": 3.0369,
+      "step": 1110
+    },
+    {
+      "epoch": 14.816053511705686,
+      "grad_norm": 0.5718163847923279,
+      "learning_rate": 0.00015083426028921024,
+      "loss": 3.1683,
+      "step": 1111
+    },
+    {
+      "epoch": 14.82943143812709,
+      "grad_norm": 0.5773367285728455,
+      "learning_rate": 0.00015078976640711902,
+      "loss": 3.2753,
+      "step": 1112
+    },
+    {
+      "epoch": 14.842809364548495,
+      "grad_norm": 0.6617185473442078,
+      "learning_rate": 0.0001507452725250278,
+      "loss": 2.9713,
+      "step": 1113
+    },
+    {
+      "epoch": 14.856187290969899,
+      "grad_norm": 0.6748194098472595,
+      "learning_rate": 0.00015070077864293662,
+      "loss": 3.0961,
+      "step": 1114
+    },
+    {
+      "epoch": 14.869565217391305,
+      "grad_norm": 0.6942034959793091,
+      "learning_rate": 0.0001506562847608454,
+      "loss": 3.0778,
+      "step": 1115
+    },
+    {
+      "epoch": 14.882943143812708,
+      "grad_norm": 1.0203640460968018,
+      "learning_rate": 0.00015061179087875418,
+      "loss": 3.0705,
+      "step": 1116
+    },
+    {
+      "epoch": 14.896321070234114,
+      "grad_norm": 0.5746601223945618,
+      "learning_rate": 0.00015056729699666296,
+      "loss": 3.1204,
+      "step": 1117
+    },
+    {
+      "epoch": 14.909698996655518,
+      "grad_norm": 0.7374005317687988,
+      "learning_rate": 0.00015052280311457177,
+      "loss": 3.1289,
+      "step": 1118
+    },
+    {
+      "epoch": 14.923076923076923,
+      "grad_norm": 0.5524411201477051,
+      "learning_rate": 0.00015047830923248053,
+      "loss": 3.2795,
+      "step": 1119
+    },
+    {
+      "epoch": 14.936454849498327,
+      "grad_norm": 0.7024741768836975,
+      "learning_rate": 0.0001504338153503893,
+      "loss": 3.0675,
+      "step": 1120
+    },
+    {
+      "epoch": 14.949832775919733,
+      "grad_norm": 0.7431137561798096,
+      "learning_rate": 0.00015038932146829812,
+      "loss": 3.1222,
+      "step": 1121
+    },
+    {
+      "epoch": 14.963210702341136,
+      "grad_norm": 0.6568113565444946,
+      "learning_rate": 0.0001503448275862069,
+      "loss": 3.1523,
+      "step": 1122
+    },
+    {
+      "epoch": 14.976588628762542,
+      "grad_norm": 0.6193330883979797,
+      "learning_rate": 0.00015030033370411569,
+      "loss": 3.1632,
+      "step": 1123
+    },
+    {
+      "epoch": 14.989966555183946,
+      "grad_norm": 0.6371363401412964,
+      "learning_rate": 0.00015025583982202447,
+      "loss": 3.0525,
+      "step": 1124
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.652542233467102,
+      "learning_rate": 0.00015021134593993328,
+      "loss": 3.1846,
+      "step": 1125
+    },
+    {
+      "epoch": 15.013377926421406,
+      "grad_norm": 0.6337831616401672,
+      "learning_rate": 0.00015016685205784203,
+      "loss": 2.9292,
+      "step": 1126
+    },
+    {
+      "epoch": 15.02675585284281,
+      "grad_norm": 0.85350501537323,
+      "learning_rate": 0.00015012235817575084,
+      "loss": 2.913,
+      "step": 1127
+    },
+    {
+      "epoch": 15.040133779264215,
+      "grad_norm": 0.6439313888549805,
+      "learning_rate": 0.00015007786429365962,
+      "loss": 3.0809,
+      "step": 1128
+    },
+    {
+      "epoch": 15.053511705685619,
+      "grad_norm": 0.5232247114181519,
+      "learning_rate": 0.0001500333704115684,
+      "loss": 3.1348,
+      "step": 1129
+    },
+    {
+      "epoch": 15.066889632107024,
+      "grad_norm": 0.6082741618156433,
+      "learning_rate": 0.0001499888765294772,
+      "loss": 2.9282,
+      "step": 1130
+    },
+    {
+      "epoch": 15.080267558528428,
+      "grad_norm": 0.5736444592475891,
+      "learning_rate": 0.000149944382647386,
+      "loss": 2.9891,
+      "step": 1131
+    },
+    {
+      "epoch": 15.093645484949834,
+      "grad_norm": 0.7732790112495422,
+      "learning_rate": 0.00014989988876529478,
+      "loss": 3.0461,
+      "step": 1132
+    },
+    {
+      "epoch": 15.107023411371237,
+      "grad_norm": 0.618357241153717,
+      "learning_rate": 0.00014985539488320356,
+      "loss": 3.0296,
+      "step": 1133
+    },
+    {
+      "epoch": 15.120401337792643,
+      "grad_norm": 0.7245836853981018,
+      "learning_rate": 0.00014981090100111235,
+      "loss": 3.0078,
+      "step": 1134
+    },
+    {
+      "epoch": 15.133779264214047,
+      "grad_norm": 0.6738787293434143,
+      "learning_rate": 0.00014976640711902116,
+      "loss": 3.1171,
+      "step": 1135
+    },
+    {
+      "epoch": 15.147157190635452,
+      "grad_norm": 0.5802761316299438,
+      "learning_rate": 0.0001497219132369299,
+      "loss": 3.1403,
+      "step": 1136
+    },
+    {
+      "epoch": 15.160535117056856,
+      "grad_norm": 0.5941367745399475,
+      "learning_rate": 0.00014967741935483872,
+      "loss": 2.9149,
+      "step": 1137
+    },
+    {
+      "epoch": 15.173913043478262,
+      "grad_norm": 0.8884940147399902,
+      "learning_rate": 0.0001496329254727475,
+      "loss": 2.8983,
+      "step": 1138
+    },
+    {
+      "epoch": 15.187290969899665,
+      "grad_norm": 0.7229192852973938,
+      "learning_rate": 0.00014958843159065628,
+      "loss": 2.9774,
+      "step": 1139
+    },
+    {
+      "epoch": 15.200668896321071,
+      "grad_norm": 0.6714467406272888,
+      "learning_rate": 0.00014954393770856507,
+      "loss": 2.9219,
+      "step": 1140
+    },
+    {
+      "epoch": 15.214046822742475,
+      "grad_norm": 0.6785704493522644,
+      "learning_rate": 0.00014949944382647388,
+      "loss": 3.0205,
+      "step": 1141
+    },
+    {
+      "epoch": 15.22742474916388,
+      "grad_norm": 0.6349677443504333,
+      "learning_rate": 0.00014945494994438266,
+      "loss": 3.1601,
+      "step": 1142
+    },
+    {
+      "epoch": 15.240802675585284,
+      "grad_norm": 0.557123064994812,
+      "learning_rate": 0.00014941045606229144,
+      "loss": 2.6297,
+      "step": 1143
+    },
+    {
+      "epoch": 15.25418060200669,
+      "grad_norm": 0.6714944243431091,
+      "learning_rate": 0.00014936596218020022,
+      "loss": 2.7951,
+      "step": 1144
+    },
+    {
+      "epoch": 15.267558528428093,
+      "grad_norm": 0.6747463345527649,
+      "learning_rate": 0.00014932146829810903,
+      "loss": 2.7909,
+      "step": 1145
+    },
+    {
+      "epoch": 15.280936454849499,
+      "grad_norm": 0.5717387199401855,
+      "learning_rate": 0.0001492769744160178,
+      "loss": 3.2896,
+      "step": 1146
+    },
+    {
+      "epoch": 15.294314381270903,
+      "grad_norm": 0.6589123010635376,
+      "learning_rate": 0.0001492324805339266,
+      "loss": 2.8332,
+      "step": 1147
+    },
+    {
+      "epoch": 15.307692307692308,
+      "grad_norm": 0.6273646950721741,
+      "learning_rate": 0.00014918798665183538,
+      "loss": 3.0084,
+      "step": 1148
+    },
+    {
+      "epoch": 15.321070234113712,
+      "grad_norm": 0.6551377773284912,
+      "learning_rate": 0.00014914349276974416,
+      "loss": 2.8147,
+      "step": 1149
+    },
+    {
+      "epoch": 15.334448160535118,
+      "grad_norm": 0.6751659512519836,
+      "learning_rate": 0.00014909899888765295,
+      "loss": 3.1345,
+      "step": 1150
+    },
+    {
+      "epoch": 15.347826086956522,
+      "grad_norm": 0.677094042301178,
+      "learning_rate": 0.00014905450500556175,
+      "loss": 3.1958,
+      "step": 1151
+    },
+    {
+      "epoch": 15.361204013377927,
+      "grad_norm": 0.6613426804542542,
+      "learning_rate": 0.00014901001112347054,
+      "loss": 3.081,
+      "step": 1152
+    },
+    {
+      "epoch": 15.37458193979933,
+      "grad_norm": 0.7645783424377441,
+      "learning_rate": 0.00014896551724137932,
+      "loss": 2.8799,
+      "step": 1153
+    },
+    {
+      "epoch": 15.387959866220736,
+      "grad_norm": 0.5698953866958618,
+      "learning_rate": 0.0001489210233592881,
+      "loss": 3.1691,
+      "step": 1154
+    },
+    {
+      "epoch": 15.40133779264214,
+      "grad_norm": 0.6581351161003113,
+      "learning_rate": 0.0001488765294771969,
+      "loss": 3.2365,
+      "step": 1155
+    },
+    {
+      "epoch": 15.414715719063546,
+      "grad_norm": 0.7809271812438965,
+      "learning_rate": 0.00014883203559510567,
+      "loss": 2.833,
+      "step": 1156
+    },
+    {
+      "epoch": 15.42809364548495,
+      "grad_norm": 0.6226280927658081,
+      "learning_rate": 0.00014878754171301448,
+      "loss": 3.1502,
+      "step": 1157
+    },
+    {
+      "epoch": 15.441471571906355,
+      "grad_norm": 0.5494824051856995,
+      "learning_rate": 0.00014874304783092326,
+      "loss": 3.2195,
+      "step": 1158
+    },
+    {
+      "epoch": 15.454849498327759,
+      "grad_norm": 0.5729116797447205,
+      "learning_rate": 0.00014869855394883204,
+      "loss": 2.9463,
+      "step": 1159
+    },
+    {
+      "epoch": 15.468227424749164,
+      "grad_norm": 0.6673750877380371,
+      "learning_rate": 0.00014865406006674082,
+      "loss": 3.1514,
+      "step": 1160
+    },
+    {
+      "epoch": 15.481605351170568,
+      "grad_norm": 0.6746686697006226,
+      "learning_rate": 0.0001486095661846496,
+      "loss": 3.0088,
+      "step": 1161
+    },
+    {
+      "epoch": 15.494983277591974,
+      "grad_norm": 0.6898564100265503,
+      "learning_rate": 0.00014856507230255842,
+      "loss": 2.8468,
+      "step": 1162
+    },
+    {
+      "epoch": 15.508361204013378,
+      "grad_norm": 0.7262438535690308,
+      "learning_rate": 0.00014852057842046717,
+      "loss": 2.8017,
+      "step": 1163
+    },
+    {
+      "epoch": 15.521739130434783,
+      "grad_norm": 0.6878666877746582,
+      "learning_rate": 0.00014847608453837598,
+      "loss": 2.9486,
+      "step": 1164
+    },
+    {
+      "epoch": 15.535117056856187,
+      "grad_norm": 0.6375080347061157,
+      "learning_rate": 0.00014843159065628476,
+      "loss": 3.0757,
+      "step": 1165
+    },
+    {
+      "epoch": 15.548494983277592,
+      "grad_norm": 0.6540268063545227,
+      "learning_rate": 0.00014838709677419355,
+      "loss": 3.1262,
+      "step": 1166
+    },
+    {
+      "epoch": 15.561872909698996,
+      "grad_norm": 0.6036689877510071,
+      "learning_rate": 0.00014834260289210233,
+      "loss": 3.0607,
+      "step": 1167
+    },
+    {
+      "epoch": 15.575250836120402,
+      "grad_norm": 0.5899893641471863,
+      "learning_rate": 0.00014829810901001114,
+      "loss": 3.1222,
+      "step": 1168
+    },
+    {
+      "epoch": 15.588628762541806,
+      "grad_norm": 0.7268028259277344,
+      "learning_rate": 0.00014825361512791992,
+      "loss": 3.1777,
+      "step": 1169
+    },
+    {
+      "epoch": 15.602006688963211,
+      "grad_norm": 0.6990141272544861,
+      "learning_rate": 0.0001482091212458287,
+      "loss": 3.2142,
+      "step": 1170
+    },
+    {
+      "epoch": 15.615384615384615,
+      "grad_norm": 0.6009657382965088,
+      "learning_rate": 0.00014816462736373748,
+      "loss": 3.141,
+      "step": 1171
+    },
+    {
+      "epoch": 15.62876254180602,
+      "grad_norm": 0.6287830471992493,
+      "learning_rate": 0.0001481201334816463,
+      "loss": 3.0621,
+      "step": 1172
+    },
+    {
+      "epoch": 15.642140468227424,
+      "grad_norm": 0.6720128655433655,
+      "learning_rate": 0.00014807563959955505,
+      "loss": 3.0353,
+      "step": 1173
+    },
+    {
+      "epoch": 15.65551839464883,
+      "grad_norm": 0.6694427132606506,
+      "learning_rate": 0.00014803114571746386,
+      "loss": 3.0171,
+      "step": 1174
+    },
+    {
+      "epoch": 15.668896321070234,
+      "grad_norm": 0.5630237460136414,
+      "learning_rate": 0.00014798665183537264,
+      "loss": 3.1444,
+      "step": 1175
+    },
+    {
+      "epoch": 15.68227424749164,
+      "grad_norm": 0.7139558792114258,
+      "learning_rate": 0.00014794215795328142,
+      "loss": 3.0012,
+      "step": 1176
+    },
+    {
+      "epoch": 15.695652173913043,
+      "grad_norm": 0.6374551057815552,
+      "learning_rate": 0.0001478976640711902,
+      "loss": 2.9123,
+      "step": 1177
+    },
+    {
+      "epoch": 15.709030100334449,
+      "grad_norm": 0.5957819223403931,
+      "learning_rate": 0.00014785317018909902,
+      "loss": 3.09,
+      "step": 1178
+    },
+    {
+      "epoch": 15.722408026755852,
+      "grad_norm": 0.6083621382713318,
+      "learning_rate": 0.0001478086763070078,
+      "loss": 3.0231,
+      "step": 1179
+    },
+    {
+      "epoch": 15.735785953177258,
+      "grad_norm": 0.6169192790985107,
+      "learning_rate": 0.00014776418242491658,
+      "loss": 2.9863,
+      "step": 1180
+    },
+    {
+      "epoch": 15.749163879598662,
+      "grad_norm": 0.6058081984519958,
+      "learning_rate": 0.00014771968854282536,
+      "loss": 3.0261,
+      "step": 1181
+    },
+    {
+      "epoch": 15.762541806020067,
+      "grad_norm": 0.5816760659217834,
+      "learning_rate": 0.00014767519466073417,
+      "loss": 3.1593,
+      "step": 1182
+    },
+    {
+      "epoch": 15.775919732441471,
+      "grad_norm": 0.6246895790100098,
+      "learning_rate": 0.00014763070077864293,
+      "loss": 3.1029,
+      "step": 1183
+    },
+    {
+      "epoch": 15.789297658862877,
+      "grad_norm": 0.56280517578125,
+      "learning_rate": 0.00014758620689655174,
+      "loss": 2.9778,
+      "step": 1184
+    },
+    {
+      "epoch": 15.80267558528428,
+      "grad_norm": 0.5743212699890137,
+      "learning_rate": 0.00014754171301446052,
+      "loss": 2.8799,
+      "step": 1185
+    },
+    {
+      "epoch": 15.816053511705686,
+      "grad_norm": 0.6163922548294067,
+      "learning_rate": 0.0001474972191323693,
+      "loss": 3.0226,
+      "step": 1186
+    },
+    {
+      "epoch": 15.82943143812709,
+      "grad_norm": 0.5892409682273865,
+      "learning_rate": 0.00014745272525027808,
+      "loss": 3.1167,
+      "step": 1187
+    },
+    {
+      "epoch": 15.842809364548495,
+      "grad_norm": 0.7977785468101501,
+      "learning_rate": 0.0001474082313681869,
+      "loss": 2.8427,
+      "step": 1188
+    },
+    {
+      "epoch": 15.856187290969899,
+      "grad_norm": 0.7396023273468018,
+      "learning_rate": 0.00014736373748609568,
+      "loss": 2.7809,
+      "step": 1189
+    },
+    {
+      "epoch": 15.869565217391305,
+      "grad_norm": 0.58844393491745,
+      "learning_rate": 0.00014731924360400446,
+      "loss": 3.0624,
+      "step": 1190
+    },
+    {
+      "epoch": 15.882943143812708,
+      "grad_norm": 0.6903204321861267,
+      "learning_rate": 0.00014727474972191324,
+      "loss": 3.1246,
+      "step": 1191
+    },
+    {
+      "epoch": 15.896321070234114,
+      "grad_norm": 0.5902391672134399,
+      "learning_rate": 0.00014723025583982205,
+      "loss": 3.0505,
+      "step": 1192
+    },
+    {
+      "epoch": 15.909698996655518,
+      "grad_norm": 0.575752317905426,
+      "learning_rate": 0.0001471857619577308,
+      "loss": 2.8508,
+      "step": 1193
+    },
+    {
+      "epoch": 15.923076923076923,
+      "grad_norm": 0.7248224020004272,
+      "learning_rate": 0.00014714126807563961,
+      "loss": 3.1438,
+      "step": 1194
+    },
+    {
+      "epoch": 15.936454849498327,
+      "grad_norm": 0.5669791102409363,
+      "learning_rate": 0.0001470967741935484,
+      "loss": 3.1765,
+      "step": 1195
+    },
+    {
+      "epoch": 15.949832775919733,
+      "grad_norm": 0.6656806468963623,
+      "learning_rate": 0.00014705228031145718,
+      "loss": 3.1456,
+      "step": 1196
+    },
+    {
+      "epoch": 15.963210702341136,
+      "grad_norm": 0.6073266863822937,
+      "learning_rate": 0.00014700778642936596,
+      "loss": 3.1836,
+      "step": 1197
+    },
+    {
+      "epoch": 15.976588628762542,
+      "grad_norm": 0.8209658861160278,
+      "learning_rate": 0.00014696329254727477,
+      "loss": 2.8457,
+      "step": 1198
+    },
+    {
+      "epoch": 15.989966555183946,
+      "grad_norm": 0.6495081186294556,
+      "learning_rate": 0.00014691879866518355,
+      "loss": 3.0161,
+      "step": 1199
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.7522635459899902,
+      "learning_rate": 0.00014687430478309234,
+      "loss": 2.8293,
+      "step": 1200
+    },
+    {
+      "epoch": 16.013377926421406,
+      "grad_norm": 0.8024417161941528,
+      "learning_rate": 0.00014682981090100112,
+      "loss": 2.909,
+      "step": 1201
+    },
+    {
+      "epoch": 16.02675585284281,
+      "grad_norm": 0.7147983908653259,
+      "learning_rate": 0.0001467853170189099,
+      "loss": 2.7664,
+      "step": 1202
+    },
+    {
+      "epoch": 16.040133779264213,
+      "grad_norm": 0.602427065372467,
+      "learning_rate": 0.00014674082313681868,
+      "loss": 2.8189,
+      "step": 1203
+    },
+    {
+      "epoch": 16.05351170568562,
+      "grad_norm": 0.7274264097213745,
+      "learning_rate": 0.00014669632925472747,
+      "loss": 2.9957,
+      "step": 1204
+    },
+    {
+      "epoch": 16.066889632107024,
+      "grad_norm": 0.6042147278785706,
+      "learning_rate": 0.00014665183537263628,
+      "loss": 2.8925,
+      "step": 1205
+    },
+    {
+      "epoch": 16.08026755852843,
+      "grad_norm": 0.592339813709259,
+      "learning_rate": 0.00014660734149054506,
+      "loss": 3.1463,
+      "step": 1206
+    },
+    {
+      "epoch": 16.093645484949832,
+      "grad_norm": 0.6099987030029297,
+      "learning_rate": 0.00014656284760845384,
+      "loss": 3.2264,
+      "step": 1207
+    },
+    {
+      "epoch": 16.107023411371237,
+      "grad_norm": 0.9324999451637268,
+      "learning_rate": 0.00014651835372636262,
+      "loss": 2.916,
+      "step": 1208
+    },
+    {
+      "epoch": 16.120401337792643,
+      "grad_norm": 0.6000798344612122,
+      "learning_rate": 0.00014647385984427143,
+      "loss": 2.9242,
+      "step": 1209
+    },
+    {
+      "epoch": 16.13377926421405,
+      "grad_norm": 0.609199047088623,
+      "learning_rate": 0.0001464293659621802,
+      "loss": 2.9405,
+      "step": 1210
+    },
+    {
+      "epoch": 16.14715719063545,
+      "grad_norm": 0.6789796948432922,
+      "learning_rate": 0.000146384872080089,
+      "loss": 3.003,
+      "step": 1211
+    },
+    {
+      "epoch": 16.160535117056856,
+      "grad_norm": 0.6567651629447937,
+      "learning_rate": 0.00014634037819799778,
+      "loss": 2.8733,
+      "step": 1212
+    },
+    {
+      "epoch": 16.17391304347826,
+      "grad_norm": 0.5860549211502075,
+      "learning_rate": 0.00014629588431590656,
+      "loss": 3.1429,
+      "step": 1213
+    },
+    {
+      "epoch": 16.187290969899667,
+      "grad_norm": 0.6623414158821106,
+      "learning_rate": 0.00014625139043381534,
+      "loss": 3.1741,
+      "step": 1214
+    },
+    {
+      "epoch": 16.20066889632107,
+      "grad_norm": 0.8366180062294006,
+      "learning_rate": 0.00014620689655172415,
+      "loss": 2.8627,
+      "step": 1215
+    },
+    {
+      "epoch": 16.214046822742475,
+      "grad_norm": 0.616780698299408,
+      "learning_rate": 0.00014616240266963294,
+      "loss": 2.7619,
+      "step": 1216
+    },
+    {
+      "epoch": 16.22742474916388,
+      "grad_norm": 0.6345306634902954,
+      "learning_rate": 0.00014611790878754172,
+      "loss": 2.9104,
+      "step": 1217
+    },
+    {
+      "epoch": 16.240802675585286,
+      "grad_norm": 0.6326844096183777,
+      "learning_rate": 0.0001460734149054505,
+      "loss": 2.8453,
+      "step": 1218
+    },
+    {
+      "epoch": 16.254180602006688,
+      "grad_norm": 0.5441793203353882,
+      "learning_rate": 0.0001460289210233593,
+      "loss": 3.2924,
+      "step": 1219
+    },
+    {
+      "epoch": 16.267558528428093,
+      "grad_norm": 0.604637086391449,
+      "learning_rate": 0.00014598442714126807,
+      "loss": 3.0362,
+      "step": 1220
+    },
+    {
+      "epoch": 16.2809364548495,
+      "grad_norm": 0.6100621819496155,
+      "learning_rate": 0.00014593993325917687,
+      "loss": 2.917,
+      "step": 1221
+    },
+    {
+      "epoch": 16.294314381270905,
+      "grad_norm": 0.6323224902153015,
+      "learning_rate": 0.00014589543937708566,
+      "loss": 3.144,
+      "step": 1222
+    },
+    {
+      "epoch": 16.307692307692307,
+      "grad_norm": 0.595485270023346,
+      "learning_rate": 0.00014585094549499444,
+      "loss": 2.8869,
+      "step": 1223
+    },
+    {
+      "epoch": 16.321070234113712,
+      "grad_norm": 0.6350538730621338,
+      "learning_rate": 0.00014580645161290322,
+      "loss": 2.7253,
+      "step": 1224
+    },
+    {
+      "epoch": 16.334448160535118,
+      "grad_norm": 0.5804395079612732,
+      "learning_rate": 0.00014576195773081203,
+      "loss": 3.1523,
+      "step": 1225
+    },
+    {
+      "epoch": 16.347826086956523,
+      "grad_norm": 0.5905717015266418,
+      "learning_rate": 0.00014571746384872081,
+      "loss": 3.1419,
+      "step": 1226
+    },
+    {
+      "epoch": 16.361204013377925,
+      "grad_norm": 0.6824894547462463,
+      "learning_rate": 0.0001456729699666296,
+      "loss": 2.8953,
+      "step": 1227
+    },
+    {
+      "epoch": 16.37458193979933,
+      "grad_norm": 0.5840978622436523,
+      "learning_rate": 0.00014562847608453838,
+      "loss": 3.1229,
+      "step": 1228
+    },
+    {
+      "epoch": 16.387959866220736,
+      "grad_norm": 0.7102469801902771,
+      "learning_rate": 0.0001455839822024472,
+      "loss": 2.6571,
+      "step": 1229
+    },
+    {
+      "epoch": 16.401337792642142,
+      "grad_norm": 0.6148349046707153,
+      "learning_rate": 0.00014553948832035594,
+      "loss": 3.0084,
+      "step": 1230
+    },
+    {
+      "epoch": 16.414715719063544,
+      "grad_norm": 0.60859215259552,
+      "learning_rate": 0.00014549499443826475,
+      "loss": 3.1348,
+      "step": 1231
+    },
+    {
+      "epoch": 16.42809364548495,
+      "grad_norm": 0.6059513688087463,
+      "learning_rate": 0.00014545050055617354,
+      "loss": 2.9678,
+      "step": 1232
+    },
+    {
+      "epoch": 16.441471571906355,
+      "grad_norm": 0.6012231707572937,
+      "learning_rate": 0.00014540600667408232,
+      "loss": 2.9279,
+      "step": 1233
+    },
+    {
+      "epoch": 16.45484949832776,
+      "grad_norm": 0.6185587644577026,
+      "learning_rate": 0.0001453615127919911,
+      "loss": 2.9922,
+      "step": 1234
+    },
+    {
+      "epoch": 16.468227424749163,
+      "grad_norm": 0.5989127159118652,
+      "learning_rate": 0.0001453170189098999,
+      "loss": 3.0736,
+      "step": 1235
+    },
+    {
+      "epoch": 16.48160535117057,
+      "grad_norm": 0.673633337020874,
+      "learning_rate": 0.0001452725250278087,
+      "loss": 2.9703,
+      "step": 1236
+    },
+    {
+      "epoch": 16.494983277591974,
+      "grad_norm": 0.6815900206565857,
+      "learning_rate": 0.00014522803114571747,
+      "loss": 3.0171,
+      "step": 1237
+    },
+    {
+      "epoch": 16.50836120401338,
+      "grad_norm": 0.7687232494354248,
+      "learning_rate": 0.00014518353726362626,
+      "loss": 2.8988,
+      "step": 1238
+    },
+    {
+      "epoch": 16.52173913043478,
+      "grad_norm": 0.7992755174636841,
+      "learning_rate": 0.00014513904338153507,
+      "loss": 3.0347,
+      "step": 1239
+    },
+    {
+      "epoch": 16.535117056856187,
+      "grad_norm": 0.8673639893531799,
+      "learning_rate": 0.00014509454949944382,
+      "loss": 2.9654,
+      "step": 1240
+    },
+    {
+      "epoch": 16.548494983277592,
+      "grad_norm": 0.6200671792030334,
+      "learning_rate": 0.00014505005561735263,
+      "loss": 3.1075,
+      "step": 1241
+    },
+    {
+      "epoch": 16.561872909698998,
+      "grad_norm": 0.8055624961853027,
+      "learning_rate": 0.0001450055617352614,
+      "loss": 2.877,
+      "step": 1242
+    },
+    {
+      "epoch": 16.5752508361204,
+      "grad_norm": 0.6428245902061462,
+      "learning_rate": 0.0001449610678531702,
+      "loss": 3.0326,
+      "step": 1243
+    },
+    {
+      "epoch": 16.588628762541806,
+      "grad_norm": 0.844804584980011,
+      "learning_rate": 0.00014491657397107898,
+      "loss": 2.8093,
+      "step": 1244
+    },
+    {
+      "epoch": 16.60200668896321,
+      "grad_norm": 0.5699613690376282,
+      "learning_rate": 0.00014487208008898776,
+      "loss": 3.1914,
+      "step": 1245
+    },
+    {
+      "epoch": 16.615384615384617,
+      "grad_norm": 0.6638582348823547,
+      "learning_rate": 0.00014482758620689657,
+      "loss": 2.7484,
+      "step": 1246
+    },
+    {
+      "epoch": 16.62876254180602,
+      "grad_norm": 0.7390914559364319,
+      "learning_rate": 0.00014478309232480533,
+      "loss": 3.0355,
+      "step": 1247
+    },
+    {
+      "epoch": 16.642140468227424,
+      "grad_norm": 0.6177923083305359,
+      "learning_rate": 0.00014473859844271413,
+      "loss": 3.0067,
+      "step": 1248
+    },
+    {
+      "epoch": 16.65551839464883,
+      "grad_norm": 0.6234062314033508,
+      "learning_rate": 0.00014469410456062292,
+      "loss": 2.8153,
+      "step": 1249
+    },
+    {
+      "epoch": 16.668896321070235,
+      "grad_norm": 0.8505418300628662,
+      "learning_rate": 0.0001446496106785317,
+      "loss": 2.8292,
+      "step": 1250
+    },
+    {
+      "epoch": 16.682274247491637,
+      "grad_norm": 0.8339266180992126,
+      "learning_rate": 0.00014460511679644048,
+      "loss": 2.8072,
+      "step": 1251
+    },
+    {
+      "epoch": 16.695652173913043,
+      "grad_norm": 0.5782635807991028,
+      "learning_rate": 0.0001445606229143493,
+      "loss": 3.1079,
+      "step": 1252
+    },
+    {
+      "epoch": 16.70903010033445,
+      "grad_norm": 0.687126874923706,
+      "learning_rate": 0.00014451612903225807,
+      "loss": 3.1401,
+      "step": 1253
+    },
+    {
+      "epoch": 16.722408026755854,
+      "grad_norm": 0.7313762307167053,
+      "learning_rate": 0.00014447163515016686,
+      "loss": 2.9167,
+      "step": 1254
+    },
+    {
+      "epoch": 16.735785953177256,
+      "grad_norm": 0.8815247416496277,
+      "learning_rate": 0.00014442714126807564,
+      "loss": 3.0294,
+      "step": 1255
+    },
+    {
+      "epoch": 16.74916387959866,
+      "grad_norm": 0.7636277675628662,
+      "learning_rate": 0.00014438264738598445,
+      "loss": 3.0758,
+      "step": 1256
+    },
+    {
+      "epoch": 16.762541806020067,
+      "grad_norm": 0.5961578488349915,
+      "learning_rate": 0.0001443381535038932,
+      "loss": 3.1027,
+      "step": 1257
+    },
+    {
+      "epoch": 16.775919732441473,
+      "grad_norm": 0.6840028762817383,
+      "learning_rate": 0.000144293659621802,
+      "loss": 2.9192,
+      "step": 1258
+    },
+    {
+      "epoch": 16.789297658862875,
+      "grad_norm": 0.7895340323448181,
+      "learning_rate": 0.0001442491657397108,
+      "loss": 2.8772,
+      "step": 1259
+    },
+    {
+      "epoch": 16.80267558528428,
+      "grad_norm": 0.8516091704368591,
+      "learning_rate": 0.00014420467185761958,
+      "loss": 2.9168,
+      "step": 1260
+    },
+    {
+      "epoch": 16.816053511705686,
+      "grad_norm": 0.6745076179504395,
+      "learning_rate": 0.00014416017797552836,
+      "loss": 3.1352,
+      "step": 1261
+    },
+    {
+      "epoch": 16.82943143812709,
+      "grad_norm": 0.6744667887687683,
+      "learning_rate": 0.00014411568409343717,
+      "loss": 2.8439,
+      "step": 1262
+    },
+    {
+      "epoch": 16.842809364548494,
+      "grad_norm": 0.6307089924812317,
+      "learning_rate": 0.00014407119021134595,
+      "loss": 3.2054,
+      "step": 1263
+    },
+    {
+      "epoch": 16.8561872909699,
+      "grad_norm": 0.6480753421783447,
+      "learning_rate": 0.00014402669632925473,
+      "loss": 3.039,
+      "step": 1264
+    },
+    {
+      "epoch": 16.869565217391305,
+      "grad_norm": 0.6143667697906494,
+      "learning_rate": 0.00014398220244716352,
+      "loss": 2.9475,
+      "step": 1265
+    },
+    {
+      "epoch": 16.88294314381271,
+      "grad_norm": 0.6289299130439758,
+      "learning_rate": 0.00014393770856507233,
+      "loss": 3.1437,
+      "step": 1266
+    },
+    {
+      "epoch": 16.896321070234112,
+      "grad_norm": 0.6618160009384155,
+      "learning_rate": 0.00014389321468298108,
+      "loss": 2.8641,
+      "step": 1267
+    },
+    {
+      "epoch": 16.909698996655518,
+      "grad_norm": 0.6053375601768494,
+      "learning_rate": 0.0001438487208008899,
+      "loss": 2.9129,
+      "step": 1268
+    },
+    {
+      "epoch": 16.923076923076923,
+      "grad_norm": 0.5706185102462769,
+      "learning_rate": 0.00014380422691879867,
+      "loss": 3.005,
+      "step": 1269
+    },
+    {
+      "epoch": 16.93645484949833,
+      "grad_norm": 0.6779253482818604,
+      "learning_rate": 0.00014375973303670746,
+      "loss": 3.1071,
+      "step": 1270
+    },
+    {
+      "epoch": 16.94983277591973,
+      "grad_norm": 0.6679616570472717,
+      "learning_rate": 0.00014371523915461624,
+      "loss": 3.1792,
+      "step": 1271
+    },
+    {
+      "epoch": 16.963210702341136,
+      "grad_norm": 0.6018584966659546,
+      "learning_rate": 0.00014367074527252505,
+      "loss": 2.9947,
+      "step": 1272
+    },
+    {
+      "epoch": 16.976588628762542,
+      "grad_norm": 0.6106094717979431,
+      "learning_rate": 0.00014362625139043383,
+      "loss": 3.1965,
+      "step": 1273
+    },
+    {
+      "epoch": 16.989966555183948,
+      "grad_norm": 0.5486257672309875,
+      "learning_rate": 0.0001435817575083426,
+      "loss": 3.0975,
+      "step": 1274
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.6516265273094177,
+      "learning_rate": 0.0001435372636262514,
+      "loss": 3.1064,
+      "step": 1275
+    },
+    {
+      "epoch": 17.013377926421406,
+      "grad_norm": 0.6434454917907715,
+      "learning_rate": 0.0001434927697441602,
+      "loss": 2.8903,
+      "step": 1276
+    },
+    {
+      "epoch": 17.02675585284281,
+      "grad_norm": 0.6237186193466187,
+      "learning_rate": 0.00014344827586206896,
+      "loss": 3.135,
+      "step": 1277
+    },
+    {
+      "epoch": 17.040133779264213,
+      "grad_norm": 0.5920026302337646,
+      "learning_rate": 0.00014340378197997777,
+      "loss": 2.8612,
+      "step": 1278
+    },
+    {
+      "epoch": 17.05351170568562,
+      "grad_norm": 0.6545232534408569,
+      "learning_rate": 0.00014335928809788655,
+      "loss": 2.8934,
+      "step": 1279
+    },
+    {
+      "epoch": 17.066889632107024,
+      "grad_norm": 0.7839710116386414,
+      "learning_rate": 0.00014331479421579533,
+      "loss": 2.965,
+      "step": 1280
+    },
+    {
+      "epoch": 17.08026755852843,
+      "grad_norm": 0.6448978781700134,
+      "learning_rate": 0.00014327030033370412,
+      "loss": 2.9343,
+      "step": 1281
+    },
+    {
+      "epoch": 17.093645484949832,
+      "grad_norm": 0.5713958144187927,
+      "learning_rate": 0.00014322580645161293,
+      "loss": 3.1566,
+      "step": 1282
+    },
+    {
+      "epoch": 17.107023411371237,
+      "grad_norm": 0.666409969329834,
+      "learning_rate": 0.0001431813125695217,
+      "loss": 2.7705,
+      "step": 1283
+    },
+    {
+      "epoch": 17.120401337792643,
+      "grad_norm": 0.6068354249000549,
+      "learning_rate": 0.0001431368186874305,
+      "loss": 3.088,
+      "step": 1284
+    },
+    {
+      "epoch": 17.13377926421405,
+      "grad_norm": 0.8292580246925354,
+      "learning_rate": 0.00014309232480533927,
+      "loss": 2.8939,
+      "step": 1285
+    },
+    {
+      "epoch": 17.14715719063545,
+      "grad_norm": 0.6789494156837463,
+      "learning_rate": 0.00014304783092324806,
+      "loss": 3.2317,
+      "step": 1286
+    },
+    {
+      "epoch": 17.160535117056856,
+      "grad_norm": 0.6030963063240051,
+      "learning_rate": 0.00014300333704115684,
+      "loss": 2.9248,
+      "step": 1287
+    },
+    {
+      "epoch": 17.17391304347826,
+      "grad_norm": 0.7090041041374207,
+      "learning_rate": 0.00014295884315906562,
+      "loss": 2.9097,
+      "step": 1288
+    },
+    {
+      "epoch": 17.187290969899667,
+      "grad_norm": 0.5750879645347595,
+      "learning_rate": 0.00014291434927697443,
+      "loss": 2.9585,
+      "step": 1289
+    },
+    {
+      "epoch": 17.20066889632107,
+      "grad_norm": 0.6379792094230652,
+      "learning_rate": 0.0001428698553948832,
+      "loss": 2.794,
+      "step": 1290
+    },
+    {
+      "epoch": 17.214046822742475,
+      "grad_norm": 0.6736400127410889,
+      "learning_rate": 0.000142825361512792,
+      "loss": 2.8812,
+      "step": 1291
+    },
+    {
+      "epoch": 17.22742474916388,
+      "grad_norm": 0.6580933332443237,
+      "learning_rate": 0.00014278086763070078,
+      "loss": 2.8029,
+      "step": 1292
+    },
+    {
+      "epoch": 17.240802675585286,
+      "grad_norm": 0.6550527215003967,
+      "learning_rate": 0.00014273637374860959,
+      "loss": 3.1135,
+      "step": 1293
+    },
+    {
+      "epoch": 17.254180602006688,
+      "grad_norm": 0.6616796255111694,
+      "learning_rate": 0.00014269187986651834,
+      "loss": 2.8916,
+      "step": 1294
+    },
+    {
+      "epoch": 17.267558528428093,
+      "grad_norm": 0.7247623801231384,
+      "learning_rate": 0.00014264738598442715,
+      "loss": 2.8482,
+      "step": 1295
+    },
+    {
+      "epoch": 17.2809364548495,
+      "grad_norm": 0.7138639092445374,
+      "learning_rate": 0.00014260289210233593,
+      "loss": 2.9283,
+      "step": 1296
+    },
+    {
+      "epoch": 17.294314381270905,
+      "grad_norm": 0.6413894891738892,
+      "learning_rate": 0.00014255839822024472,
+      "loss": 3.0273,
+      "step": 1297
+    },
+    {
+      "epoch": 17.307692307692307,
+      "grad_norm": 0.6106882095336914,
+      "learning_rate": 0.0001425139043381535,
+      "loss": 3.062,
+      "step": 1298
+    },
+    {
+      "epoch": 17.321070234113712,
+      "grad_norm": 0.6762199997901917,
+      "learning_rate": 0.0001424694104560623,
+      "loss": 2.7864,
+      "step": 1299
+    },
+    {
+      "epoch": 17.334448160535118,
+      "grad_norm": 0.65083909034729,
+      "learning_rate": 0.0001424249165739711,
+      "loss": 3.115,
+      "step": 1300
+    },
+    {
+      "epoch": 17.347826086956523,
+      "grad_norm": 0.7381249666213989,
+      "learning_rate": 0.00014238042269187987,
+      "loss": 2.6981,
+      "step": 1301
+    },
+    {
+      "epoch": 17.361204013377925,
+      "grad_norm": 0.5674475431442261,
+      "learning_rate": 0.00014233592880978865,
+      "loss": 2.8646,
+      "step": 1302
+    },
+    {
+      "epoch": 17.37458193979933,
+      "grad_norm": 0.6201330423355103,
+      "learning_rate": 0.00014229143492769746,
+      "loss": 3.207,
+      "step": 1303
+    },
+    {
+      "epoch": 17.387959866220736,
+      "grad_norm": 0.7004446983337402,
+      "learning_rate": 0.00014224694104560622,
+      "loss": 2.7623,
+      "step": 1304
+    },
+    {
+      "epoch": 17.401337792642142,
+      "grad_norm": 0.7278717756271362,
+      "learning_rate": 0.00014220244716351503,
+      "loss": 2.7967,
+      "step": 1305
+    },
+    {
+      "epoch": 17.414715719063544,
+      "grad_norm": 0.6384133696556091,
+      "learning_rate": 0.0001421579532814238,
+      "loss": 2.7542,
+      "step": 1306
+    },
+    {
+      "epoch": 17.42809364548495,
+      "grad_norm": 0.6443619132041931,
+      "learning_rate": 0.0001421134593993326,
+      "loss": 2.9376,
+      "step": 1307
+    },
+    {
+      "epoch": 17.441471571906355,
+      "grad_norm": 0.6889017224311829,
+      "learning_rate": 0.00014206896551724138,
+      "loss": 2.8868,
+      "step": 1308
+    },
+    {
+      "epoch": 17.45484949832776,
+      "grad_norm": 0.624864935874939,
+      "learning_rate": 0.00014202447163515019,
+      "loss": 3.0459,
+      "step": 1309
+    },
+    {
+      "epoch": 17.468227424749163,
+      "grad_norm": 0.5964120626449585,
+      "learning_rate": 0.00014197997775305897,
+      "loss": 2.8604,
+      "step": 1310
+    },
+    {
+      "epoch": 17.48160535117057,
+      "grad_norm": 0.5584320425987244,
+      "learning_rate": 0.00014193548387096775,
+      "loss": 3.1267,
+      "step": 1311
+    },
+    {
+      "epoch": 17.494983277591974,
+      "grad_norm": 0.6566155552864075,
+      "learning_rate": 0.00014189098998887653,
+      "loss": 2.7059,
+      "step": 1312
+    },
+    {
+      "epoch": 17.50836120401338,
+      "grad_norm": 0.7166509032249451,
+      "learning_rate": 0.00014184649610678534,
+      "loss": 3.1076,
+      "step": 1313
+    },
+    {
+      "epoch": 17.52173913043478,
+      "grad_norm": 0.6076642870903015,
+      "learning_rate": 0.0001418020022246941,
+      "loss": 3.1121,
+      "step": 1314
+    },
+    {
+      "epoch": 17.535117056856187,
+      "grad_norm": 0.6293672323226929,
+      "learning_rate": 0.0001417575083426029,
+      "loss": 3.0719,
+      "step": 1315
+    },
+    {
+      "epoch": 17.548494983277592,
+      "grad_norm": 0.6696231365203857,
+      "learning_rate": 0.0001417130144605117,
+      "loss": 2.4281,
+      "step": 1316
+    },
+    {
+      "epoch": 17.561872909698998,
+      "grad_norm": 0.7928171157836914,
+      "learning_rate": 0.00014166852057842047,
+      "loss": 2.8602,
+      "step": 1317
+    },
+    {
+      "epoch": 17.5752508361204,
+      "grad_norm": 0.5897494554519653,
+      "learning_rate": 0.00014162402669632925,
+      "loss": 3.1556,
+      "step": 1318
+    },
+    {
+      "epoch": 17.588628762541806,
+      "grad_norm": 0.6028451323509216,
+      "learning_rate": 0.00014157953281423806,
+      "loss": 3.1228,
+      "step": 1319
+    },
+    {
+      "epoch": 17.60200668896321,
+      "grad_norm": 0.6237207651138306,
+      "learning_rate": 0.00014153503893214685,
+      "loss": 2.9691,
+      "step": 1320
+    },
+    {
+      "epoch": 17.615384615384617,
+      "grad_norm": 0.6401494741439819,
+      "learning_rate": 0.00014149054505005563,
+      "loss": 3.0659,
+      "step": 1321
+    },
+    {
+      "epoch": 17.62876254180602,
+      "grad_norm": 0.7098166942596436,
+      "learning_rate": 0.0001414460511679644,
+      "loss": 2.9606,
+      "step": 1322
+    },
+    {
+      "epoch": 17.642140468227424,
+      "grad_norm": 0.6416228413581848,
+      "learning_rate": 0.00014140155728587322,
+      "loss": 3.1056,
+      "step": 1323
+    },
+    {
+      "epoch": 17.65551839464883,
+      "grad_norm": 0.7303211092948914,
+      "learning_rate": 0.00014135706340378198,
+      "loss": 2.8604,
+      "step": 1324
+    },
+    {
+      "epoch": 17.668896321070235,
+      "grad_norm": 0.64544677734375,
+      "learning_rate": 0.00014131256952169079,
+      "loss": 2.6751,
+      "step": 1325
+    },
+    {
+      "epoch": 17.682274247491637,
+      "grad_norm": 0.6870211362838745,
+      "learning_rate": 0.00014126807563959957,
+      "loss": 2.9802,
+      "step": 1326
+    },
+    {
+      "epoch": 17.695652173913043,
+      "grad_norm": 0.6570687294006348,
+      "learning_rate": 0.00014122358175750835,
+      "loss": 3.0496,
+      "step": 1327
+    },
+    {
+      "epoch": 17.70903010033445,
+      "grad_norm": 0.7057302594184875,
+      "learning_rate": 0.00014117908787541713,
+      "loss": 2.845,
+      "step": 1328
+    },
+    {
+      "epoch": 17.722408026755854,
+      "grad_norm": 0.8613574504852295,
+      "learning_rate": 0.00014113459399332591,
+      "loss": 3.0314,
+      "step": 1329
+    },
+    {
+      "epoch": 17.735785953177256,
+      "grad_norm": 0.7408957481384277,
+      "learning_rate": 0.00014109010011123472,
+      "loss": 2.8651,
+      "step": 1330
+    },
+    {
+      "epoch": 17.74916387959866,
+      "grad_norm": 0.6553664803504944,
+      "learning_rate": 0.00014104560622914348,
+      "loss": 2.9647,
+      "step": 1331
+    },
+    {
+      "epoch": 17.762541806020067,
+      "grad_norm": 0.5991332530975342,
+      "learning_rate": 0.0001410011123470523,
+      "loss": 2.9796,
+      "step": 1332
+    },
+    {
+      "epoch": 17.775919732441473,
+      "grad_norm": 0.6124044060707092,
+      "learning_rate": 0.00014095661846496107,
+      "loss": 3.0682,
+      "step": 1333
+    },
+    {
+      "epoch": 17.789297658862875,
+      "grad_norm": 0.5788628458976746,
+      "learning_rate": 0.00014091212458286985,
+      "loss": 3.0357,
+      "step": 1334
+    },
+    {
+      "epoch": 17.80267558528428,
+      "grad_norm": 0.6785842776298523,
+      "learning_rate": 0.00014086763070077864,
+      "loss": 3.1128,
+      "step": 1335
+    },
+    {
+      "epoch": 17.816053511705686,
+      "grad_norm": 0.5994388461112976,
+      "learning_rate": 0.00014082313681868745,
+      "loss": 2.9087,
+      "step": 1336
+    },
+    {
+      "epoch": 17.82943143812709,
+      "grad_norm": 0.6150069236755371,
+      "learning_rate": 0.00014077864293659623,
+      "loss": 2.9488,
+      "step": 1337
+    },
+    {
+      "epoch": 17.842809364548494,
+      "grad_norm": 0.6211126446723938,
+      "learning_rate": 0.000140734149054505,
+      "loss": 2.8006,
+      "step": 1338
+    },
+    {
+      "epoch": 17.8561872909699,
+      "grad_norm": 0.6093603372573853,
+      "learning_rate": 0.0001406896551724138,
+      "loss": 3.0013,
+      "step": 1339
+    },
+    {
+      "epoch": 17.869565217391305,
+      "grad_norm": 0.6861109137535095,
+      "learning_rate": 0.0001406451612903226,
+      "loss": 2.9423,
+      "step": 1340
+    },
+    {
+      "epoch": 17.88294314381271,
+      "grad_norm": 0.6148517727851868,
+      "learning_rate": 0.00014060066740823136,
+      "loss": 2.8471,
+      "step": 1341
+    },
+    {
+      "epoch": 17.896321070234112,
+      "grad_norm": 0.8285694718360901,
+      "learning_rate": 0.00014055617352614017,
+      "loss": 3.055,
+      "step": 1342
+    },
+    {
+      "epoch": 17.909698996655518,
+      "grad_norm": 0.5955973863601685,
+      "learning_rate": 0.00014051167964404895,
+      "loss": 3.0088,
+      "step": 1343
+    },
+    {
+      "epoch": 17.923076923076923,
+      "grad_norm": 0.6020825505256653,
+      "learning_rate": 0.00014046718576195773,
+      "loss": 3.1445,
+      "step": 1344
+    },
+    {
+      "epoch": 17.93645484949833,
+      "grad_norm": 0.6115384101867676,
+      "learning_rate": 0.00014042269187986651,
+      "loss": 3.0889,
+      "step": 1345
+    },
+    {
+      "epoch": 17.94983277591973,
+      "grad_norm": 0.6469634175300598,
+      "learning_rate": 0.00014037819799777532,
+      "loss": 3.1916,
+      "step": 1346
+    },
+    {
+      "epoch": 17.963210702341136,
+      "grad_norm": 0.6653386354446411,
+      "learning_rate": 0.0001403337041156841,
+      "loss": 2.8805,
+      "step": 1347
+    },
+    {
+      "epoch": 17.976588628762542,
+      "grad_norm": 0.6167243719100952,
+      "learning_rate": 0.0001402892102335929,
+      "loss": 2.9819,
+      "step": 1348
+    },
+    {
+      "epoch": 17.989966555183948,
+      "grad_norm": 0.6281883716583252,
+      "learning_rate": 0.00014024471635150167,
+      "loss": 3.1109,
+      "step": 1349
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.6747295260429382,
+      "learning_rate": 0.00014020022246941048,
+      "loss": 2.8772,
+      "step": 1350
+    },
+    {
+      "epoch": 18.013377926421406,
+      "grad_norm": 0.5834116339683533,
+      "learning_rate": 0.00014015572858731924,
+      "loss": 2.9399,
+      "step": 1351
+    },
+    {
+      "epoch": 18.02675585284281,
+      "grad_norm": 0.620858371257782,
+      "learning_rate": 0.00014011123470522805,
+      "loss": 2.7351,
+      "step": 1352
+    },
+    {
+      "epoch": 18.040133779264213,
+      "grad_norm": 0.5407689809799194,
+      "learning_rate": 0.00014006674082313683,
+      "loss": 2.8329,
+      "step": 1353
+    },
+    {
+      "epoch": 18.05351170568562,
+      "grad_norm": 0.6045056581497192,
+      "learning_rate": 0.0001400222469410456,
+      "loss": 2.7382,
+      "step": 1354
+    },
+    {
+      "epoch": 18.066889632107024,
+      "grad_norm": 0.5433570146560669,
+      "learning_rate": 0.0001399777530589544,
+      "loss": 2.9154,
+      "step": 1355
+    },
+    {
+      "epoch": 18.08026755852843,
+      "grad_norm": 0.6174083352088928,
+      "learning_rate": 0.0001399332591768632,
+      "loss": 2.7438,
+      "step": 1356
+    },
+    {
+      "epoch": 18.093645484949832,
+      "grad_norm": 0.6720690727233887,
+      "learning_rate": 0.00013988876529477198,
+      "loss": 3.1324,
+      "step": 1357
+    },
+    {
+      "epoch": 18.107023411371237,
+      "grad_norm": 0.648423433303833,
+      "learning_rate": 0.00013984427141268077,
+      "loss": 2.7802,
+      "step": 1358
+    },
+    {
+      "epoch": 18.120401337792643,
+      "grad_norm": 0.6625978350639343,
+      "learning_rate": 0.00013979977753058955,
+      "loss": 2.9152,
+      "step": 1359
+    },
+    {
+      "epoch": 18.13377926421405,
+      "grad_norm": 0.5362007021903992,
+      "learning_rate": 0.00013975528364849836,
+      "loss": 2.9552,
+      "step": 1360
+    },
+    {
+      "epoch": 18.14715719063545,
+      "grad_norm": 0.6275555491447449,
+      "learning_rate": 0.00013971078976640711,
+      "loss": 3.0742,
+      "step": 1361
+    },
+    {
+      "epoch": 18.160535117056856,
+      "grad_norm": 0.5755884647369385,
+      "learning_rate": 0.00013966629588431592,
+      "loss": 2.8705,
+      "step": 1362
+    },
+    {
+      "epoch": 18.17391304347826,
+      "grad_norm": 0.5092719793319702,
+      "learning_rate": 0.0001396218020022247,
+      "loss": 2.9684,
+      "step": 1363
+    },
+    {
+      "epoch": 18.187290969899667,
+      "grad_norm": 0.7400075197219849,
+      "learning_rate": 0.0001395773081201335,
+      "loss": 2.9726,
+      "step": 1364
+    },
+    {
+      "epoch": 18.20066889632107,
+      "grad_norm": 0.6478124260902405,
+      "learning_rate": 0.00013953281423804227,
+      "loss": 2.7427,
+      "step": 1365
+    },
+    {
+      "epoch": 18.214046822742475,
+      "grad_norm": 0.6313418745994568,
+      "learning_rate": 0.00013948832035595108,
+      "loss": 2.7713,
+      "step": 1366
+    },
+    {
+      "epoch": 18.22742474916388,
+      "grad_norm": 0.5571421980857849,
+      "learning_rate": 0.00013944382647385986,
+      "loss": 2.9221,
+      "step": 1367
+    },
+    {
+      "epoch": 18.240802675585286,
+      "grad_norm": 0.5346395373344421,
+      "learning_rate": 0.00013939933259176862,
+      "loss": 2.9842,
+      "step": 1368
+    },
+    {
+      "epoch": 18.254180602006688,
+      "grad_norm": 0.5828048586845398,
+      "learning_rate": 0.00013935483870967743,
+      "loss": 2.8683,
+      "step": 1369
+    },
+    {
+      "epoch": 18.267558528428093,
+      "grad_norm": 0.6446037888526917,
+      "learning_rate": 0.0001393103448275862,
+      "loss": 2.8,
+      "step": 1370
+    },
+    {
+      "epoch": 18.2809364548495,
+      "grad_norm": 0.612689197063446,
+      "learning_rate": 0.000139265850945495,
+      "loss": 3.0642,
+      "step": 1371
+    },
+    {
+      "epoch": 18.294314381270905,
+      "grad_norm": 0.5511941909790039,
+      "learning_rate": 0.00013922135706340377,
+      "loss": 3.0633,
+      "step": 1372
+    },
+    {
+      "epoch": 18.307692307692307,
+      "grad_norm": 0.7538149356842041,
+      "learning_rate": 0.00013917686318131258,
+      "loss": 2.7224,
+      "step": 1373
+    },
+    {
+      "epoch": 18.321070234113712,
+      "grad_norm": 0.6194874048233032,
+      "learning_rate": 0.00013913236929922137,
+      "loss": 2.9357,
+      "step": 1374
+    },
+    {
+      "epoch": 18.334448160535118,
+      "grad_norm": 0.5770833492279053,
+      "learning_rate": 0.00013908787541713015,
+      "loss": 2.8585,
+      "step": 1375
+    },
+    {
+      "epoch": 18.347826086956523,
+      "grad_norm": 0.609080970287323,
+      "learning_rate": 0.00013904338153503893,
+      "loss": 2.7763,
+      "step": 1376
+    },
+    {
+      "epoch": 18.361204013377925,
+      "grad_norm": 0.5578462481498718,
+      "learning_rate": 0.00013899888765294774,
+      "loss": 3.0324,
+      "step": 1377
+    },
+    {
+      "epoch": 18.37458193979933,
+      "grad_norm": 0.5949610471725464,
+      "learning_rate": 0.0001389543937708565,
+      "loss": 2.7332,
+      "step": 1378
+    },
+    {
+      "epoch": 18.387959866220736,
+      "grad_norm": 0.6248785257339478,
+      "learning_rate": 0.0001389098998887653,
+      "loss": 2.7355,
+      "step": 1379
+    },
+    {
+      "epoch": 18.401337792642142,
+      "grad_norm": 0.606239378452301,
+      "learning_rate": 0.0001388654060066741,
+      "loss": 2.9883,
+      "step": 1380
+    },
+    {
+      "epoch": 18.414715719063544,
+      "grad_norm": 0.6222496628761292,
+      "learning_rate": 0.00013882091212458287,
+      "loss": 2.7384,
+      "step": 1381
+    },
+    {
+      "epoch": 18.42809364548495,
+      "grad_norm": 0.6253412365913391,
+      "learning_rate": 0.00013877641824249165,
+      "loss": 2.7555,
+      "step": 1382
+    },
+    {
+      "epoch": 18.441471571906355,
+      "grad_norm": 0.6204626560211182,
+      "learning_rate": 0.00013873192436040046,
+      "loss": 2.7279,
+      "step": 1383
+    },
+    {
+      "epoch": 18.45484949832776,
+      "grad_norm": 0.7254919409751892,
+      "learning_rate": 0.00013868743047830924,
+      "loss": 2.9053,
+      "step": 1384
+    },
+    {
+      "epoch": 18.468227424749163,
+      "grad_norm": 0.6207154393196106,
+      "learning_rate": 0.00013864293659621803,
+      "loss": 3.0648,
+      "step": 1385
+    },
+    {
+      "epoch": 18.48160535117057,
+      "grad_norm": 0.6959066390991211,
+      "learning_rate": 0.0001385984427141268,
+      "loss": 2.9641,
+      "step": 1386
+    },
+    {
+      "epoch": 18.494983277591974,
+      "grad_norm": 0.6345707774162292,
+      "learning_rate": 0.00013855394883203562,
+      "loss": 2.9452,
+      "step": 1387
+    },
+    {
+      "epoch": 18.50836120401338,
+      "grad_norm": 0.5806639790534973,
+      "learning_rate": 0.00013850945494994437,
+      "loss": 2.9375,
+      "step": 1388
+    },
+    {
+      "epoch": 18.52173913043478,
+      "grad_norm": 0.6498666405677795,
+      "learning_rate": 0.00013846496106785318,
+      "loss": 2.867,
+      "step": 1389
+    },
+    {
+      "epoch": 18.535117056856187,
+      "grad_norm": 0.629264771938324,
+      "learning_rate": 0.00013842046718576197,
+      "loss": 2.822,
+      "step": 1390
+    },
+    {
+      "epoch": 18.548494983277592,
+      "grad_norm": 0.6734644174575806,
+      "learning_rate": 0.00013837597330367075,
+      "loss": 2.9065,
+      "step": 1391
+    },
+    {
+      "epoch": 18.561872909698998,
+      "grad_norm": 0.5705899000167847,
+      "learning_rate": 0.00013833147942157953,
+      "loss": 2.9909,
+      "step": 1392
+    },
+    {
+      "epoch": 18.5752508361204,
+      "grad_norm": 0.6786744594573975,
+      "learning_rate": 0.00013828698553948834,
+      "loss": 3.0667,
+      "step": 1393
+    },
+    {
+      "epoch": 18.588628762541806,
+      "grad_norm": 0.6044118404388428,
+      "learning_rate": 0.00013824249165739712,
+      "loss": 2.8345,
+      "step": 1394
+    },
+    {
+      "epoch": 18.60200668896321,
+      "grad_norm": 0.5928333401679993,
+      "learning_rate": 0.0001381979977753059,
+      "loss": 2.974,
+      "step": 1395
+    },
+    {
+      "epoch": 18.615384615384617,
+      "grad_norm": 0.636883556842804,
+      "learning_rate": 0.0001381535038932147,
+      "loss": 3.1034,
+      "step": 1396
+    },
+    {
+      "epoch": 18.62876254180602,
+      "grad_norm": 0.6029159426689148,
+      "learning_rate": 0.0001381090100111235,
+      "loss": 3.0303,
+      "step": 1397
+    },
+    {
+      "epoch": 18.642140468227424,
+      "grad_norm": 0.6479122638702393,
+      "learning_rate": 0.00013806451612903225,
+      "loss": 3.1782,
+      "step": 1398
+    },
+    {
+      "epoch": 18.65551839464883,
+      "grad_norm": 0.6547753810882568,
+      "learning_rate": 0.00013802002224694106,
+      "loss": 3.0594,
+      "step": 1399
+    },
+    {
+      "epoch": 18.668896321070235,
+      "grad_norm": 0.6506614089012146,
+      "learning_rate": 0.00013797552836484984,
+      "loss": 3.0955,
+      "step": 1400
+    },
+    {
+      "epoch": 18.682274247491637,
+      "grad_norm": 0.6073411107063293,
+      "learning_rate": 0.00013793103448275863,
+      "loss": 2.7452,
+      "step": 1401
+    },
+    {
+      "epoch": 18.695652173913043,
+      "grad_norm": 0.6307429075241089,
+      "learning_rate": 0.0001378865406006674,
+      "loss": 2.8329,
+      "step": 1402
+    },
+    {
+      "epoch": 18.70903010033445,
+      "grad_norm": 0.7205286026000977,
+      "learning_rate": 0.00013784204671857622,
+      "loss": 2.8228,
+      "step": 1403
+    },
+    {
+      "epoch": 18.722408026755854,
+      "grad_norm": 0.6154866814613342,
+      "learning_rate": 0.000137797552836485,
+      "loss": 2.7081,
+      "step": 1404
+    },
+    {
+      "epoch": 18.735785953177256,
+      "grad_norm": 0.548799455165863,
+      "learning_rate": 0.00013775305895439378,
+      "loss": 2.9336,
+      "step": 1405
+    },
+    {
+      "epoch": 18.74916387959866,
+      "grad_norm": 0.5723371505737305,
+      "learning_rate": 0.00013770856507230257,
+      "loss": 2.9661,
+      "step": 1406
+    },
+    {
+      "epoch": 18.762541806020067,
+      "grad_norm": 0.5917731523513794,
+      "learning_rate": 0.00013766407119021137,
+      "loss": 2.8458,
+      "step": 1407
+    },
+    {
+      "epoch": 18.775919732441473,
+      "grad_norm": 0.6160814166069031,
+      "learning_rate": 0.00013761957730812013,
+      "loss": 2.8453,
+      "step": 1408
+    },
+    {
+      "epoch": 18.789297658862875,
+      "grad_norm": 0.5995500683784485,
+      "learning_rate": 0.0001375750834260289,
+      "loss": 3.1361,
+      "step": 1409
+    },
+    {
+      "epoch": 18.80267558528428,
+      "grad_norm": 0.5574924945831299,
+      "learning_rate": 0.00013753058954393772,
+      "loss": 2.9808,
+      "step": 1410
+    },
+    {
+      "epoch": 18.816053511705686,
+      "grad_norm": 0.6444510221481323,
+      "learning_rate": 0.0001374860956618465,
+      "loss": 2.9204,
+      "step": 1411
+    },
+    {
+      "epoch": 18.82943143812709,
+      "grad_norm": 0.6153264045715332,
+      "learning_rate": 0.0001374416017797553,
+      "loss": 3.036,
+      "step": 1412
+    },
+    {
+      "epoch": 18.842809364548494,
+      "grad_norm": 0.5951060652732849,
+      "learning_rate": 0.00013739710789766407,
+      "loss": 3.1012,
+      "step": 1413
+    },
+    {
+      "epoch": 18.8561872909699,
+      "grad_norm": 0.5688861012458801,
+      "learning_rate": 0.00013735261401557288,
+      "loss": 3.1568,
+      "step": 1414
+    },
+    {
+      "epoch": 18.869565217391305,
+      "grad_norm": 0.6256094574928284,
+      "learning_rate": 0.00013730812013348163,
+      "loss": 2.7721,
+      "step": 1415
+    },
+    {
+      "epoch": 18.88294314381271,
+      "grad_norm": 0.646250307559967,
+      "learning_rate": 0.00013726362625139044,
+      "loss": 3.0091,
+      "step": 1416
+    },
+    {
+      "epoch": 18.896321070234112,
+      "grad_norm": 0.6805879473686218,
+      "learning_rate": 0.00013721913236929923,
+      "loss": 2.8397,
+      "step": 1417
+    },
+    {
+      "epoch": 18.909698996655518,
+      "grad_norm": 0.6164728999137878,
+      "learning_rate": 0.000137174638487208,
+      "loss": 2.8318,
+      "step": 1418
+    },
+    {
+      "epoch": 18.923076923076923,
+      "grad_norm": 0.6298549771308899,
+      "learning_rate": 0.0001371301446051168,
+      "loss": 2.8493,
+      "step": 1419
+    },
+    {
+      "epoch": 18.93645484949833,
+      "grad_norm": 0.5760109424591064,
+      "learning_rate": 0.0001370856507230256,
+      "loss": 3.1736,
+      "step": 1420
+    },
+    {
+      "epoch": 18.94983277591973,
+      "grad_norm": 0.6126035451889038,
+      "learning_rate": 0.00013704115684093438,
+      "loss": 3.1313,
+      "step": 1421
+    },
+    {
+      "epoch": 18.963210702341136,
+      "grad_norm": 0.6092283129692078,
+      "learning_rate": 0.00013699666295884316,
+      "loss": 3.0601,
+      "step": 1422
+    },
+    {
+      "epoch": 18.976588628762542,
+      "grad_norm": 0.6506980657577515,
+      "learning_rate": 0.00013695216907675195,
+      "loss": 2.7787,
+      "step": 1423
+    },
+    {
+      "epoch": 18.989966555183948,
+      "grad_norm": 0.6060482263565063,
+      "learning_rate": 0.00013690767519466076,
+      "loss": 3.062,
+      "step": 1424
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 0.7881284952163696,
+      "learning_rate": 0.0001368631813125695,
+      "loss": 2.9882,
+      "step": 1425
+    },
+    {
+      "epoch": 19.013377926421406,
+      "grad_norm": 0.5459823608398438,
+      "learning_rate": 0.00013681868743047832,
+      "loss": 2.9312,
+      "step": 1426
+    },
+    {
+      "epoch": 19.02675585284281,
+      "grad_norm": 0.71108078956604,
+      "learning_rate": 0.0001367741935483871,
+      "loss": 2.6003,
+      "step": 1427
+    },
+    {
+      "epoch": 19.040133779264213,
+      "grad_norm": 0.6824572682380676,
+      "learning_rate": 0.00013672969966629589,
+      "loss": 2.8354,
+      "step": 1428
+    },
+    {
+      "epoch": 19.05351170568562,
+      "grad_norm": 0.607400119304657,
+      "learning_rate": 0.00013668520578420467,
+      "loss": 2.95,
+      "step": 1429
+    },
+    {
+      "epoch": 19.066889632107024,
+      "grad_norm": 0.5925526022911072,
+      "learning_rate": 0.00013664071190211348,
+      "loss": 2.7101,
+      "step": 1430
+    },
+    {
+      "epoch": 19.08026755852843,
+      "grad_norm": 0.6208476424217224,
+      "learning_rate": 0.00013659621802002226,
+      "loss": 2.7759,
+      "step": 1431
+    },
+    {
+      "epoch": 19.093645484949832,
+      "grad_norm": 0.6047778725624084,
+      "learning_rate": 0.00013655172413793104,
+      "loss": 3.0377,
+      "step": 1432
+    },
+    {
+      "epoch": 19.107023411371237,
+      "grad_norm": 0.5979378819465637,
+      "learning_rate": 0.00013650723025583983,
+      "loss": 2.8671,
+      "step": 1433
+    },
+    {
+      "epoch": 19.120401337792643,
+      "grad_norm": 0.6338753700256348,
+      "learning_rate": 0.00013646273637374863,
+      "loss": 2.7869,
+      "step": 1434
+    },
+    {
+      "epoch": 19.13377926421405,
+      "grad_norm": 0.6263737678527832,
+      "learning_rate": 0.0001364182424916574,
+      "loss": 2.6591,
+      "step": 1435
+    },
+    {
+      "epoch": 19.14715719063545,
+      "grad_norm": 0.646990180015564,
+      "learning_rate": 0.0001363737486095662,
+      "loss": 2.896,
+      "step": 1436
+    },
+    {
+      "epoch": 19.160535117056856,
+      "grad_norm": 0.5691138505935669,
+      "learning_rate": 0.00013632925472747498,
+      "loss": 2.8774,
+      "step": 1437
+    },
+    {
+      "epoch": 19.17391304347826,
+      "grad_norm": 0.5838844180107117,
+      "learning_rate": 0.00013628476084538376,
+      "loss": 2.9722,
+      "step": 1438
+    },
+    {
+      "epoch": 19.187290969899667,
+      "grad_norm": 0.6154463887214661,
+      "learning_rate": 0.00013624026696329255,
+      "loss": 3.0143,
+      "step": 1439
+    },
+    {
+      "epoch": 19.20066889632107,
+      "grad_norm": 0.6480549573898315,
+      "learning_rate": 0.00013619577308120136,
+      "loss": 2.8118,
+      "step": 1440
+    },
+    {
+      "epoch": 19.214046822742475,
+      "grad_norm": 0.7675592303276062,
+      "learning_rate": 0.00013615127919911014,
+      "loss": 2.865,
+      "step": 1441
+    },
+    {
+      "epoch": 19.22742474916388,
+      "grad_norm": 0.7231382131576538,
+      "learning_rate": 0.00013610678531701892,
+      "loss": 2.6942,
+      "step": 1442
+    },
+    {
+      "epoch": 19.240802675585286,
+      "grad_norm": 0.6359425187110901,
+      "learning_rate": 0.0001360622914349277,
+      "loss": 2.5976,
+      "step": 1443
+    },
+    {
+      "epoch": 19.254180602006688,
+      "grad_norm": 0.6486908793449402,
+      "learning_rate": 0.0001360177975528365,
+      "loss": 2.9494,
+      "step": 1444
+    },
+    {
+      "epoch": 19.267558528428093,
+      "grad_norm": 0.5930846929550171,
+      "learning_rate": 0.00013597330367074527,
+      "loss": 3.1295,
+      "step": 1445
+    },
+    {
+      "epoch": 19.2809364548495,
+      "grad_norm": 0.6988996267318726,
+      "learning_rate": 0.00013592880978865408,
+      "loss": 2.9627,
+      "step": 1446
+    },
+    {
+      "epoch": 19.294314381270905,
+      "grad_norm": 0.5971337556838989,
+      "learning_rate": 0.00013588431590656286,
+      "loss": 2.9636,
+      "step": 1447
+    },
+    {
+      "epoch": 19.307692307692307,
+      "grad_norm": 0.6479155421257019,
+      "learning_rate": 0.00013583982202447164,
+      "loss": 2.6575,
+      "step": 1448
+    },
+    {
+      "epoch": 19.321070234113712,
+      "grad_norm": 0.6771759986877441,
+      "learning_rate": 0.00013579532814238043,
+      "loss": 3.0667,
+      "step": 1449
+    },
+    {
+      "epoch": 19.334448160535118,
+      "grad_norm": 0.5785907506942749,
+      "learning_rate": 0.0001357508342602892,
+      "loss": 2.8839,
+      "step": 1450
+    },
+    {
+      "epoch": 19.347826086956523,
+      "grad_norm": 0.6315357089042664,
+      "learning_rate": 0.00013570634037819802,
+      "loss": 2.6925,
+      "step": 1451
+    },
+    {
+      "epoch": 19.361204013377925,
+      "grad_norm": 0.6047807931900024,
+      "learning_rate": 0.00013566184649610677,
+      "loss": 2.7948,
+      "step": 1452
+    },
+    {
+      "epoch": 19.37458193979933,
+      "grad_norm": 0.6079906225204468,
+      "learning_rate": 0.00013561735261401558,
+      "loss": 3.1798,
+      "step": 1453
+    },
+    {
+      "epoch": 19.387959866220736,
+      "grad_norm": 0.5820274353027344,
+      "learning_rate": 0.00013557285873192436,
+      "loss": 3.1159,
+      "step": 1454
+    },
+    {
+      "epoch": 19.401337792642142,
+      "grad_norm": 0.562022864818573,
+      "learning_rate": 0.00013552836484983315,
+      "loss": 2.9869,
+      "step": 1455
+    },
+    {
+      "epoch": 19.414715719063544,
+      "grad_norm": 0.6663182973861694,
+      "learning_rate": 0.00013548387096774193,
+      "loss": 2.8038,
+      "step": 1456
+    },
+    {
+      "epoch": 19.42809364548495,
+      "grad_norm": 0.7092719078063965,
+      "learning_rate": 0.00013543937708565074,
+      "loss": 3.0159,
+      "step": 1457
+    },
+    {
+      "epoch": 19.441471571906355,
+      "grad_norm": 0.5990714430809021,
+      "learning_rate": 0.00013539488320355952,
+      "loss": 2.8858,
+      "step": 1458
+    },
+    {
+      "epoch": 19.45484949832776,
+      "grad_norm": 0.600847065448761,
+      "learning_rate": 0.0001353503893214683,
+      "loss": 2.9352,
+      "step": 1459
+    },
+    {
+      "epoch": 19.468227424749163,
+      "grad_norm": 0.6393849849700928,
+      "learning_rate": 0.00013530589543937709,
+      "loss": 2.6911,
+      "step": 1460
+    },
+    {
+      "epoch": 19.48160535117057,
+      "grad_norm": 0.6170421838760376,
+      "learning_rate": 0.0001352614015572859,
+      "loss": 2.749,
+      "step": 1461
+    },
+    {
+      "epoch": 19.494983277591974,
+      "grad_norm": 0.53690105676651,
+      "learning_rate": 0.00013521690767519465,
+      "loss": 2.9516,
+      "step": 1462
+    },
+    {
+      "epoch": 19.50836120401338,
+      "grad_norm": 0.5976501703262329,
+      "learning_rate": 0.00013517241379310346,
+      "loss": 2.9996,
+      "step": 1463
+    },
+    {
+      "epoch": 19.52173913043478,
+      "grad_norm": 0.5892135500907898,
+      "learning_rate": 0.00013512791991101224,
+      "loss": 3.2039,
+      "step": 1464
+    },
+    {
+      "epoch": 19.535117056856187,
+      "grad_norm": 0.65968918800354,
+      "learning_rate": 0.00013508342602892102,
+      "loss": 2.5897,
+      "step": 1465
+    },
+    {
+      "epoch": 19.548494983277592,
+      "grad_norm": 0.591454267501831,
+      "learning_rate": 0.0001350389321468298,
+      "loss": 3.0104,
+      "step": 1466
+    },
+    {
+      "epoch": 19.561872909698998,
+      "grad_norm": 0.6272184252738953,
+      "learning_rate": 0.00013499443826473862,
+      "loss": 2.7266,
+      "step": 1467
+    },
+    {
+      "epoch": 19.5752508361204,
+      "grad_norm": 0.6142420172691345,
+      "learning_rate": 0.0001349499443826474,
+      "loss": 2.8358,
+      "step": 1468
+    },
+    {
+      "epoch": 19.588628762541806,
+      "grad_norm": 0.6268441677093506,
+      "learning_rate": 0.00013490545050055618,
+      "loss": 3.0196,
+      "step": 1469
+    },
+    {
+      "epoch": 19.60200668896321,
+      "grad_norm": 0.6512436866760254,
+      "learning_rate": 0.00013486095661846496,
+      "loss": 2.9558,
+      "step": 1470
+    },
+    {
+      "epoch": 19.615384615384617,
+      "grad_norm": 0.5983771681785583,
+      "learning_rate": 0.00013481646273637377,
+      "loss": 2.9958,
+      "step": 1471
+    },
+    {
+      "epoch": 19.62876254180602,
+      "grad_norm": 0.6994190216064453,
+      "learning_rate": 0.00013477196885428253,
+      "loss": 2.8019,
+      "step": 1472
+    },
+    {
+      "epoch": 19.642140468227424,
+      "grad_norm": 0.5878567695617676,
+      "learning_rate": 0.00013472747497219134,
+      "loss": 2.7007,
+      "step": 1473
+    },
+    {
+      "epoch": 19.65551839464883,
+      "grad_norm": 0.6140199303627014,
+      "learning_rate": 0.00013468298109010012,
+      "loss": 2.9212,
+      "step": 1474
+    },
+    {
+      "epoch": 19.668896321070235,
+      "grad_norm": 0.648714542388916,
+      "learning_rate": 0.0001346384872080089,
+      "loss": 3.0053,
+      "step": 1475
+    },
+    {
+      "epoch": 19.682274247491637,
+      "grad_norm": 0.5991750359535217,
+      "learning_rate": 0.00013459399332591769,
+      "loss": 2.9129,
+      "step": 1476
+    },
+    {
+      "epoch": 19.695652173913043,
+      "grad_norm": 0.5538223385810852,
+      "learning_rate": 0.0001345494994438265,
+      "loss": 2.9097,
+      "step": 1477
+    },
+    {
+      "epoch": 19.70903010033445,
+      "grad_norm": 0.5864409804344177,
+      "learning_rate": 0.00013450500556173528,
+      "loss": 2.9348,
+      "step": 1478
+    },
+    {
+      "epoch": 19.722408026755854,
+      "grad_norm": 0.6004533767700195,
+      "learning_rate": 0.00013446051167964406,
+      "loss": 2.8845,
+      "step": 1479
+    },
+    {
+      "epoch": 19.735785953177256,
+      "grad_norm": 0.6316581964492798,
+      "learning_rate": 0.00013441601779755284,
+      "loss": 2.8619,
+      "step": 1480
+    },
+    {
+      "epoch": 19.74916387959866,
+      "grad_norm": 0.593138575553894,
+      "learning_rate": 0.00013437152391546165,
+      "loss": 3.041,
+      "step": 1481
+    },
+    {
+      "epoch": 19.762541806020067,
+      "grad_norm": 0.5826678276062012,
+      "learning_rate": 0.0001343270300333704,
+      "loss": 2.7432,
+      "step": 1482
+    },
+    {
+      "epoch": 19.775919732441473,
+      "grad_norm": 0.6341697573661804,
+      "learning_rate": 0.00013428253615127922,
+      "loss": 2.9755,
+      "step": 1483
+    },
+    {
+      "epoch": 19.789297658862875,
+      "grad_norm": 0.5894901156425476,
+      "learning_rate": 0.000134238042269188,
+      "loss": 2.9754,
+      "step": 1484
+    },
+    {
+      "epoch": 19.80267558528428,
+      "grad_norm": 0.5840655565261841,
+      "learning_rate": 0.00013419354838709678,
+      "loss": 2.9959,
+      "step": 1485
+    },
+    {
+      "epoch": 19.816053511705686,
+      "grad_norm": 0.6006319522857666,
+      "learning_rate": 0.00013414905450500556,
+      "loss": 3.0196,
+      "step": 1486
+    },
+    {
+      "epoch": 19.82943143812709,
+      "grad_norm": 0.5647453665733337,
+      "learning_rate": 0.00013410456062291437,
+      "loss": 2.9662,
+      "step": 1487
+    },
+    {
+      "epoch": 19.842809364548494,
+      "grad_norm": 0.6583006978034973,
+      "learning_rate": 0.00013406006674082316,
+      "loss": 2.8719,
+      "step": 1488
+    },
+    {
+      "epoch": 19.8561872909699,
+      "grad_norm": 0.6041131615638733,
+      "learning_rate": 0.00013401557285873194,
+      "loss": 3.0019,
+      "step": 1489
+    },
+    {
+      "epoch": 19.869565217391305,
+      "grad_norm": 0.5524600148200989,
+      "learning_rate": 0.00013397107897664072,
+      "loss": 2.7409,
+      "step": 1490
+    },
+    {
+      "epoch": 19.88294314381271,
+      "grad_norm": 0.6532869338989258,
+      "learning_rate": 0.0001339265850945495,
+      "loss": 2.9336,
+      "step": 1491
+    },
+    {
+      "epoch": 19.896321070234112,
+      "grad_norm": 0.6459875106811523,
+      "learning_rate": 0.00013388209121245828,
+      "loss": 2.9572,
+      "step": 1492
+    },
+    {
+      "epoch": 19.909698996655518,
+      "grad_norm": 0.6051417589187622,
+      "learning_rate": 0.00013383759733036707,
+      "loss": 2.864,
+      "step": 1493
+    },
+    {
+      "epoch": 19.923076923076923,
+      "grad_norm": 0.6565695405006409,
+      "learning_rate": 0.00013379310344827588,
+      "loss": 2.865,
+      "step": 1494
+    },
+    {
+      "epoch": 19.93645484949833,
+      "grad_norm": 0.6118014454841614,
+      "learning_rate": 0.00013374860956618466,
+      "loss": 2.7594,
+      "step": 1495
+    },
+    {
+      "epoch": 19.94983277591973,
+      "grad_norm": 0.6801209449768066,
+      "learning_rate": 0.00013370411568409344,
+      "loss": 2.7532,
+      "step": 1496
+    },
+    {
+      "epoch": 19.963210702341136,
+      "grad_norm": 0.5785267353057861,
+      "learning_rate": 0.00013365962180200222,
+      "loss": 3.0618,
+      "step": 1497
+    },
+    {
+      "epoch": 19.976588628762542,
+      "grad_norm": 0.6344903707504272,
+      "learning_rate": 0.00013361512791991103,
+      "loss": 2.7908,
+      "step": 1498
+    },
+    {
+      "epoch": 19.989966555183948,
+      "grad_norm": 0.6073011159896851,
+      "learning_rate": 0.0001335706340378198,
+      "loss": 3.0699,
+      "step": 1499
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 0.6989748477935791,
+      "learning_rate": 0.0001335261401557286,
+      "loss": 2.7932,
+      "step": 1500
+    },
+    {
+      "epoch": 20.013377926421406,
+      "grad_norm": 0.5897710919380188,
+      "learning_rate": 0.00013348164627363738,
+      "loss": 3.0379,
+      "step": 1501
+    },
+    {
+      "epoch": 20.02675585284281,
+      "grad_norm": 0.5845353603363037,
+      "learning_rate": 0.00013343715239154616,
+      "loss": 2.7123,
+      "step": 1502
+    },
+    {
+      "epoch": 20.040133779264213,
+      "grad_norm": 0.5720913410186768,
+      "learning_rate": 0.00013339265850945495,
+      "loss": 3.0196,
+      "step": 1503
+    },
+    {
+      "epoch": 20.05351170568562,
+      "grad_norm": 0.7616726756095886,
+      "learning_rate": 0.00013334816462736375,
+      "loss": 2.6719,
+      "step": 1504
+    },
+    {
+      "epoch": 20.066889632107024,
+      "grad_norm": 0.6211048364639282,
+      "learning_rate": 0.00013330367074527254,
+      "loss": 2.8263,
+      "step": 1505
+    },
+    {
+      "epoch": 20.08026755852843,
+      "grad_norm": 0.6016023755073547,
+      "learning_rate": 0.00013325917686318132,
+      "loss": 2.7226,
+      "step": 1506
+    },
+    {
+      "epoch": 20.093645484949832,
+      "grad_norm": 0.6265879273414612,
+      "learning_rate": 0.0001332146829810901,
+      "loss": 2.8597,
+      "step": 1507
+    },
+    {
+      "epoch": 20.107023411371237,
+      "grad_norm": 0.6129719614982605,
+      "learning_rate": 0.0001331701890989989,
+      "loss": 2.9449,
+      "step": 1508
+    },
+    {
+      "epoch": 20.120401337792643,
+      "grad_norm": 0.6351513266563416,
+      "learning_rate": 0.00013312569521690767,
+      "loss": 2.9768,
+      "step": 1509
+    },
+    {
+      "epoch": 20.13377926421405,
+      "grad_norm": 0.5772795677185059,
+      "learning_rate": 0.00013308120133481648,
+      "loss": 3.0424,
+      "step": 1510
+    },
+    {
+      "epoch": 20.14715719063545,
+      "grad_norm": 0.600697934627533,
+      "learning_rate": 0.00013303670745272526,
+      "loss": 2.793,
+      "step": 1511
+    },
+    {
+      "epoch": 20.160535117056856,
+      "grad_norm": 0.5418747663497925,
+      "learning_rate": 0.00013299221357063404,
+      "loss": 2.8706,
+      "step": 1512
+    },
+    {
+      "epoch": 20.17391304347826,
+      "grad_norm": 0.5500038862228394,
+      "learning_rate": 0.00013294771968854282,
+      "loss": 2.7352,
+      "step": 1513
+    },
+    {
+      "epoch": 20.187290969899667,
+      "grad_norm": 0.5478414297103882,
+      "learning_rate": 0.00013290322580645163,
+      "loss": 2.9113,
+      "step": 1514
+    },
+    {
+      "epoch": 20.20066889632107,
+      "grad_norm": 0.5382576584815979,
+      "learning_rate": 0.00013285873192436042,
+      "loss": 2.8276,
+      "step": 1515
+    },
+    {
+      "epoch": 20.214046822742475,
+      "grad_norm": 0.6003616452217102,
+      "learning_rate": 0.0001328142380422692,
+      "loss": 2.9112,
+      "step": 1516
+    },
+    {
+      "epoch": 20.22742474916388,
+      "grad_norm": 0.6674323678016663,
+      "learning_rate": 0.00013276974416017798,
+      "loss": 2.7382,
+      "step": 1517
+    },
+    {
+      "epoch": 20.240802675585286,
+      "grad_norm": 0.591314435005188,
+      "learning_rate": 0.0001327252502780868,
+      "loss": 2.8003,
+      "step": 1518
+    },
+    {
+      "epoch": 20.254180602006688,
+      "grad_norm": 0.5268637537956238,
+      "learning_rate": 0.00013268075639599554,
+      "loss": 2.7134,
+      "step": 1519
+    },
+    {
+      "epoch": 20.267558528428093,
+      "grad_norm": 0.5217694640159607,
+      "learning_rate": 0.00013263626251390435,
+      "loss": 2.6157,
+      "step": 1520
+    },
+    {
+      "epoch": 20.2809364548495,
+      "grad_norm": 0.6171165704727173,
+      "learning_rate": 0.00013259176863181314,
+      "loss": 2.8738,
+      "step": 1521
+    },
+    {
+      "epoch": 20.294314381270905,
+      "grad_norm": 0.5457054972648621,
+      "learning_rate": 0.00013254727474972192,
+      "loss": 2.6106,
+      "step": 1522
+    },
+    {
+      "epoch": 20.307692307692307,
+      "grad_norm": 0.6596150994300842,
+      "learning_rate": 0.0001325027808676307,
+      "loss": 2.9974,
+      "step": 1523
+    },
+    {
+      "epoch": 20.321070234113712,
+      "grad_norm": 0.7236288785934448,
+      "learning_rate": 0.0001324582869855395,
+      "loss": 2.8669,
+      "step": 1524
+    },
+    {
+      "epoch": 20.334448160535118,
+      "grad_norm": 0.6390851736068726,
+      "learning_rate": 0.0001324137931034483,
+      "loss": 2.909,
+      "step": 1525
+    },
+    {
+      "epoch": 20.347826086956523,
+      "grad_norm": 0.6010439991950989,
+      "learning_rate": 0.00013236929922135708,
+      "loss": 2.9567,
+      "step": 1526
+    },
+    {
+      "epoch": 20.361204013377925,
+      "grad_norm": 0.5825399160385132,
+      "learning_rate": 0.00013232480533926586,
+      "loss": 2.7688,
+      "step": 1527
+    },
+    {
+      "epoch": 20.37458193979933,
+      "grad_norm": 0.6000121831893921,
+      "learning_rate": 0.00013228031145717467,
+      "loss": 3.0515,
+      "step": 1528
+    },
+    {
+      "epoch": 20.387959866220736,
+      "grad_norm": 0.5775492787361145,
+      "learning_rate": 0.00013223581757508342,
+      "loss": 2.8958,
+      "step": 1529
+    },
+    {
+      "epoch": 20.401337792642142,
+      "grad_norm": 0.6193161010742188,
+      "learning_rate": 0.00013219132369299223,
+      "loss": 2.9648,
+      "step": 1530
+    },
+    {
+      "epoch": 20.414715719063544,
+      "grad_norm": 0.7751132249832153,
+      "learning_rate": 0.00013214682981090101,
+      "loss": 2.677,
+      "step": 1531
+    },
+    {
+      "epoch": 20.42809364548495,
+      "grad_norm": 0.6269053220748901,
+      "learning_rate": 0.0001321023359288098,
+      "loss": 2.9483,
+      "step": 1532
+    },
+    {
+      "epoch": 20.441471571906355,
+      "grad_norm": 0.5793836116790771,
+      "learning_rate": 0.00013205784204671858,
+      "loss": 2.8025,
+      "step": 1533
+    },
+    {
+      "epoch": 20.45484949832776,
+      "grad_norm": 0.5609649419784546,
+      "learning_rate": 0.00013201334816462736,
+      "loss": 2.8104,
+      "step": 1534
+    },
+    {
+      "epoch": 20.468227424749163,
+      "grad_norm": 0.6055050492286682,
+      "learning_rate": 0.00013196885428253617,
+      "loss": 2.8649,
+      "step": 1535
+    },
+    {
+      "epoch": 20.48160535117057,
+      "grad_norm": 0.5952653884887695,
+      "learning_rate": 0.00013192436040044493,
+      "loss": 2.6815,
+      "step": 1536
+    },
+    {
+      "epoch": 20.494983277591974,
+      "grad_norm": 0.5716878175735474,
+      "learning_rate": 0.00013187986651835374,
+      "loss": 2.637,
+      "step": 1537
+    },
+    {
+      "epoch": 20.50836120401338,
+      "grad_norm": 0.5684781074523926,
+      "learning_rate": 0.00013183537263626252,
+      "loss": 2.8601,
+      "step": 1538
+    },
+    {
+      "epoch": 20.52173913043478,
+      "grad_norm": 0.5603588223457336,
+      "learning_rate": 0.0001317908787541713,
+      "loss": 2.9173,
+      "step": 1539
+    },
+    {
+      "epoch": 20.535117056856187,
+      "grad_norm": 0.6022379398345947,
+      "learning_rate": 0.00013174638487208008,
+      "loss": 2.873,
+      "step": 1540
+    },
+    {
+      "epoch": 20.548494983277592,
+      "grad_norm": 0.6124081611633301,
+      "learning_rate": 0.0001317018909899889,
+      "loss": 3.05,
+      "step": 1541
+    },
+    {
+      "epoch": 20.561872909698998,
+      "grad_norm": 0.5928300023078918,
+      "learning_rate": 0.00013165739710789768,
+      "loss": 2.7411,
+      "step": 1542
+    },
+    {
+      "epoch": 20.5752508361204,
+      "grad_norm": 0.5931242108345032,
+      "learning_rate": 0.00013161290322580646,
+      "loss": 3.0032,
+      "step": 1543
+    },
+    {
+      "epoch": 20.588628762541806,
+      "grad_norm": 0.7123291492462158,
+      "learning_rate": 0.00013156840934371524,
+      "loss": 2.6485,
+      "step": 1544
+    },
+    {
+      "epoch": 20.60200668896321,
+      "grad_norm": 0.565980851650238,
+      "learning_rate": 0.00013152391546162405,
+      "loss": 2.7143,
+      "step": 1545
+    },
+    {
+      "epoch": 20.615384615384617,
+      "grad_norm": 0.6504361033439636,
+      "learning_rate": 0.0001314794215795328,
+      "loss": 2.6285,
+      "step": 1546
+    },
+    {
+      "epoch": 20.62876254180602,
+      "grad_norm": 0.6558433175086975,
+      "learning_rate": 0.00013143492769744161,
+      "loss": 3.0403,
+      "step": 1547
+    },
+    {
+      "epoch": 20.642140468227424,
+      "grad_norm": 0.6159015893936157,
+      "learning_rate": 0.0001313904338153504,
+      "loss": 2.8836,
+      "step": 1548
+    },
+    {
+      "epoch": 20.65551839464883,
+      "grad_norm": 0.6061471104621887,
+      "learning_rate": 0.00013134593993325918,
+      "loss": 2.9076,
+      "step": 1549
+    },
+    {
+      "epoch": 20.668896321070235,
+      "grad_norm": 0.6891756057739258,
+      "learning_rate": 0.00013130144605116796,
+      "loss": 3.0326,
+      "step": 1550
+    },
+    {
+      "epoch": 20.682274247491637,
+      "grad_norm": 0.5920016765594482,
+      "learning_rate": 0.00013125695216907677,
+      "loss": 2.949,
+      "step": 1551
+    },
+    {
+      "epoch": 20.695652173913043,
+      "grad_norm": 0.7947055697441101,
+      "learning_rate": 0.00013121245828698555,
+      "loss": 2.7796,
+      "step": 1552
+    },
+    {
+      "epoch": 20.70903010033445,
+      "grad_norm": 0.6409865021705627,
+      "learning_rate": 0.00013116796440489434,
+      "loss": 2.9103,
+      "step": 1553
+    },
+    {
+      "epoch": 20.722408026755854,
+      "grad_norm": 0.5864197611808777,
+      "learning_rate": 0.00013112347052280312,
+      "loss": 3.0493,
+      "step": 1554
+    },
+    {
+      "epoch": 20.735785953177256,
+      "grad_norm": 0.6567059755325317,
+      "learning_rate": 0.00013107897664071193,
+      "loss": 2.8359,
+      "step": 1555
+    },
+    {
+      "epoch": 20.74916387959866,
+      "grad_norm": 0.5808839201927185,
+      "learning_rate": 0.00013103448275862068,
+      "loss": 2.9595,
+      "step": 1556
+    },
+    {
+      "epoch": 20.762541806020067,
+      "grad_norm": 0.6176792979240417,
+      "learning_rate": 0.0001309899888765295,
+      "loss": 2.6975,
+      "step": 1557
+    },
+    {
+      "epoch": 20.775919732441473,
+      "grad_norm": 0.5763049721717834,
+      "learning_rate": 0.00013094549499443827,
+      "loss": 2.9752,
+      "step": 1558
+    },
+    {
+      "epoch": 20.789297658862875,
+      "grad_norm": 0.5877807140350342,
+      "learning_rate": 0.00013090100111234706,
+      "loss": 2.7214,
+      "step": 1559
+    },
+    {
+      "epoch": 20.80267558528428,
+      "grad_norm": 0.6624999642372131,
+      "learning_rate": 0.00013085650723025584,
+      "loss": 2.8642,
+      "step": 1560
+    },
+    {
+      "epoch": 20.816053511705686,
+      "grad_norm": 0.7411154508590698,
+      "learning_rate": 0.00013081201334816465,
+      "loss": 2.7526,
+      "step": 1561
+    },
+    {
+      "epoch": 20.82943143812709,
+      "grad_norm": 0.6246466636657715,
+      "learning_rate": 0.00013076751946607343,
+      "loss": 3.0441,
+      "step": 1562
+    },
+    {
+      "epoch": 20.842809364548494,
+      "grad_norm": 0.5515304803848267,
+      "learning_rate": 0.00013072302558398221,
+      "loss": 2.7211,
+      "step": 1563
+    },
+    {
+      "epoch": 20.8561872909699,
+      "grad_norm": 0.5537006258964539,
+      "learning_rate": 0.000130678531701891,
+      "loss": 2.9619,
+      "step": 1564
+    },
+    {
+      "epoch": 20.869565217391305,
+      "grad_norm": 0.6326210498809814,
+      "learning_rate": 0.0001306340378197998,
+      "loss": 3.2433,
+      "step": 1565
+    },
+    {
+      "epoch": 20.88294314381271,
+      "grad_norm": 0.6463941335678101,
+      "learning_rate": 0.00013058954393770856,
+      "loss": 2.7263,
+      "step": 1566
+    },
+    {
+      "epoch": 20.896321070234112,
+      "grad_norm": 0.6212389469146729,
+      "learning_rate": 0.00013054505005561737,
+      "loss": 2.9775,
+      "step": 1567
+    },
+    {
+      "epoch": 20.909698996655518,
+      "grad_norm": 0.5878139734268188,
+      "learning_rate": 0.00013050055617352615,
+      "loss": 2.7427,
+      "step": 1568
+    },
+    {
+      "epoch": 20.923076923076923,
+      "grad_norm": 0.5821187496185303,
+      "learning_rate": 0.00013045606229143494,
+      "loss": 2.9058,
+      "step": 1569
+    },
+    {
+      "epoch": 20.93645484949833,
+      "grad_norm": 0.6192734837532043,
+      "learning_rate": 0.00013041156840934372,
+      "loss": 2.9744,
+      "step": 1570
+    },
+    {
+      "epoch": 20.94983277591973,
+      "grad_norm": 0.5540494322776794,
+      "learning_rate": 0.00013036707452725253,
+      "loss": 2.8299,
+      "step": 1571
+    },
+    {
+      "epoch": 20.963210702341136,
+      "grad_norm": 0.6620298624038696,
+      "learning_rate": 0.0001303225806451613,
+      "loss": 2.8291,
+      "step": 1572
+    },
+    {
+      "epoch": 20.976588628762542,
+      "grad_norm": 0.7161823511123657,
+      "learning_rate": 0.0001302780867630701,
+      "loss": 2.9829,
+      "step": 1573
+    },
+    {
+      "epoch": 20.989966555183948,
+      "grad_norm": 1.1449205875396729,
+      "learning_rate": 0.00013023359288097887,
+      "loss": 2.8807,
+      "step": 1574
+    },
+    {
+      "epoch": 21.0,
+      "grad_norm": 0.6721773743629456,
+      "learning_rate": 0.00013018909899888766,
+      "loss": 2.8398,
+      "step": 1575
+    },
+    {
+      "epoch": 21.013377926421406,
+      "grad_norm": 0.5506662130355835,
+      "learning_rate": 0.00013014460511679644,
+      "loss": 2.6858,
+      "step": 1576
+    },
+    {
+      "epoch": 21.02675585284281,
+      "grad_norm": 0.5651558041572571,
+      "learning_rate": 0.00013010011123470522,
+      "loss": 2.8692,
+      "step": 1577
+    },
+    {
+      "epoch": 21.040133779264213,
+      "grad_norm": 0.5379538536071777,
+      "learning_rate": 0.00013005561735261403,
+      "loss": 2.8221,
+      "step": 1578
+    },
+    {
+      "epoch": 21.05351170568562,
+      "grad_norm": 0.5756275653839111,
+      "learning_rate": 0.0001300111234705228,
+      "loss": 2.5469,
+      "step": 1579
+    },
+    {
+      "epoch": 21.066889632107024,
+      "grad_norm": 0.672498881816864,
+      "learning_rate": 0.0001299666295884316,
+      "loss": 2.825,
+      "step": 1580
+    },
+    {
+      "epoch": 21.08026755852843,
+      "grad_norm": 0.5217440128326416,
+      "learning_rate": 0.00012992213570634038,
+      "loss": 2.732,
+      "step": 1581
+    },
+    {
+      "epoch": 21.093645484949832,
+      "grad_norm": 0.6326786875724792,
+      "learning_rate": 0.0001298776418242492,
+      "loss": 2.6537,
+      "step": 1582
+    },
+    {
+      "epoch": 21.107023411371237,
+      "grad_norm": 0.6899091005325317,
+      "learning_rate": 0.00012983314794215794,
+      "loss": 2.7133,
+      "step": 1583
+    },
+    {
+      "epoch": 21.120401337792643,
+      "grad_norm": 0.604834794998169,
+      "learning_rate": 0.00012978865406006675,
+      "loss": 2.9341,
+      "step": 1584
+    },
+    {
+      "epoch": 21.13377926421405,
+      "grad_norm": 0.7540157437324524,
+      "learning_rate": 0.00012974416017797553,
+      "loss": 2.8224,
+      "step": 1585
+    },
+    {
+      "epoch": 21.14715719063545,
+      "grad_norm": 0.6161774396896362,
+      "learning_rate": 0.00012969966629588432,
+      "loss": 2.9968,
+      "step": 1586
+    },
+    {
+      "epoch": 21.160535117056856,
+      "grad_norm": 0.618781328201294,
+      "learning_rate": 0.0001296551724137931,
+      "loss": 2.8074,
+      "step": 1587
+    },
+    {
+      "epoch": 21.17391304347826,
+      "grad_norm": 0.6708270907402039,
+      "learning_rate": 0.0001296106785317019,
+      "loss": 2.9327,
+      "step": 1588
+    },
+    {
+      "epoch": 21.187290969899667,
+      "grad_norm": 0.6362271904945374,
+      "learning_rate": 0.0001295661846496107,
+      "loss": 2.46,
+      "step": 1589
+    },
+    {
+      "epoch": 21.20066889632107,
+      "grad_norm": 0.6039800047874451,
+      "learning_rate": 0.00012952169076751947,
+      "loss": 2.7728,
+      "step": 1590
+    },
+    {
+      "epoch": 21.214046822742475,
+      "grad_norm": 0.6119550466537476,
+      "learning_rate": 0.00012947719688542826,
+      "loss": 2.7123,
+      "step": 1591
+    },
+    {
+      "epoch": 21.22742474916388,
+      "grad_norm": 0.8122309446334839,
+      "learning_rate": 0.00012943270300333707,
+      "loss": 2.7424,
+      "step": 1592
+    },
+    {
+      "epoch": 21.240802675585286,
+      "grad_norm": 0.5658085346221924,
+      "learning_rate": 0.00012938820912124582,
+      "loss": 2.4553,
+      "step": 1593
+    },
+    {
+      "epoch": 21.254180602006688,
+      "grad_norm": 0.7365813851356506,
+      "learning_rate": 0.00012934371523915463,
+      "loss": 2.9437,
+      "step": 1594
+    },
+    {
+      "epoch": 21.267558528428093,
+      "grad_norm": 0.5746303796768188,
+      "learning_rate": 0.0001292992213570634,
+      "loss": 2.8262,
+      "step": 1595
+    },
+    {
+      "epoch": 21.2809364548495,
+      "grad_norm": 0.5708132386207581,
+      "learning_rate": 0.0001292547274749722,
+      "loss": 2.9298,
+      "step": 1596
+    },
+    {
+      "epoch": 21.294314381270905,
+      "grad_norm": 0.6003934741020203,
+      "learning_rate": 0.00012921023359288098,
+      "loss": 2.7943,
+      "step": 1597
+    },
+    {
+      "epoch": 21.307692307692307,
+      "grad_norm": 0.5722339749336243,
+      "learning_rate": 0.0001291657397107898,
+      "loss": 3.0547,
+      "step": 1598
+    },
+    {
+      "epoch": 21.321070234113712,
+      "grad_norm": 0.6287402510643005,
+      "learning_rate": 0.00012912124582869857,
+      "loss": 2.8814,
+      "step": 1599
+    },
+    {
+      "epoch": 21.334448160535118,
+      "grad_norm": 0.6594595313072205,
+      "learning_rate": 0.00012907675194660735,
+      "loss": 2.7927,
+      "step": 1600
+    },
+    {
+      "epoch": 21.347826086956523,
+      "grad_norm": 0.6678686141967773,
+      "learning_rate": 0.00012903225806451613,
+      "loss": 2.7981,
+      "step": 1601
+    },
+    {
+      "epoch": 21.361204013377925,
+      "grad_norm": 0.6103408336639404,
+      "learning_rate": 0.00012898776418242494,
+      "loss": 2.9009,
+      "step": 1602
+    },
+    {
+      "epoch": 21.37458193979933,
+      "grad_norm": 0.5791922807693481,
+      "learning_rate": 0.0001289432703003337,
+      "loss": 2.6689,
+      "step": 1603
+    },
+    {
+      "epoch": 21.387959866220736,
+      "grad_norm": 0.5776186585426331,
+      "learning_rate": 0.0001288987764182425,
+      "loss": 2.903,
+      "step": 1604
+    },
+    {
+      "epoch": 21.401337792642142,
+      "grad_norm": 0.5746363401412964,
+      "learning_rate": 0.0001288542825361513,
+      "loss": 2.832,
+      "step": 1605
+    },
+    {
+      "epoch": 21.414715719063544,
+      "grad_norm": 0.5495322346687317,
+      "learning_rate": 0.00012880978865406007,
+      "loss": 2.7305,
+      "step": 1606
+    },
+    {
+      "epoch": 21.42809364548495,
+      "grad_norm": 0.6944610476493835,
+      "learning_rate": 0.00012876529477196886,
+      "loss": 2.7534,
+      "step": 1607
+    },
+    {
+      "epoch": 21.441471571906355,
+      "grad_norm": 0.6640026569366455,
+      "learning_rate": 0.00012872080088987767,
+      "loss": 2.814,
+      "step": 1608
+    },
+    {
+      "epoch": 21.45484949832776,
+      "grad_norm": 0.7469935417175293,
+      "learning_rate": 0.00012867630700778642,
+      "loss": 2.8523,
+      "step": 1609
+    },
+    {
+      "epoch": 21.468227424749163,
+      "grad_norm": 0.600739061832428,
+      "learning_rate": 0.00012863181312569523,
+      "loss": 3.0074,
+      "step": 1610
+    },
+    {
+      "epoch": 21.48160535117057,
+      "grad_norm": 0.6701086759567261,
+      "learning_rate": 0.000128587319243604,
+      "loss": 2.6693,
+      "step": 1611
+    },
+    {
+      "epoch": 21.494983277591974,
+      "grad_norm": 0.6365100145339966,
+      "learning_rate": 0.00012854282536151282,
+      "loss": 2.9558,
+      "step": 1612
+    },
+    {
+      "epoch": 21.50836120401338,
+      "grad_norm": 0.6170171499252319,
+      "learning_rate": 0.00012849833147942158,
+      "loss": 2.605,
+      "step": 1613
+    },
+    {
+      "epoch": 21.52173913043478,
+      "grad_norm": 0.5899895429611206,
+      "learning_rate": 0.0001284538375973304,
+      "loss": 2.7816,
+      "step": 1614
+    },
+    {
+      "epoch": 21.535117056856187,
+      "grad_norm": 0.6815734505653381,
+      "learning_rate": 0.00012840934371523917,
+      "loss": 2.6663,
+      "step": 1615
+    },
+    {
+      "epoch": 21.548494983277592,
+      "grad_norm": 0.6620442271232605,
+      "learning_rate": 0.00012836484983314792,
+      "loss": 2.9333,
+      "step": 1616
+    },
+    {
+      "epoch": 21.561872909698998,
+      "grad_norm": 0.6552561521530151,
+      "learning_rate": 0.00012832035595105673,
+      "loss": 2.6833,
+      "step": 1617
+    },
+    {
+      "epoch": 21.5752508361204,
+      "grad_norm": 0.6288586854934692,
+      "learning_rate": 0.00012827586206896552,
+      "loss": 2.959,
+      "step": 1618
+    },
+    {
+      "epoch": 21.588628762541806,
+      "grad_norm": 0.6969814300537109,
+      "learning_rate": 0.0001282313681868743,
+      "loss": 2.7129,
+      "step": 1619
+    },
+    {
+      "epoch": 21.60200668896321,
+      "grad_norm": 0.5918965339660645,
+      "learning_rate": 0.00012818687430478308,
+      "loss": 2.9057,
+      "step": 1620
+    },
+    {
+      "epoch": 21.615384615384617,
+      "grad_norm": 0.5739157199859619,
+      "learning_rate": 0.0001281423804226919,
+      "loss": 2.9275,
+      "step": 1621
+    },
+    {
+      "epoch": 21.62876254180602,
+      "grad_norm": 0.6649355888366699,
+      "learning_rate": 0.00012809788654060067,
+      "loss": 2.7629,
+      "step": 1622
+    },
+    {
+      "epoch": 21.642140468227424,
+      "grad_norm": 0.6449686884880066,
+      "learning_rate": 0.00012805339265850946,
+      "loss": 2.9173,
+      "step": 1623
+    },
+    {
+      "epoch": 21.65551839464883,
+      "grad_norm": 0.5940249562263489,
+      "learning_rate": 0.00012800889877641824,
+      "loss": 2.875,
+      "step": 1624
+    },
+    {
+      "epoch": 21.668896321070235,
+      "grad_norm": 0.8104184865951538,
+      "learning_rate": 0.00012796440489432705,
+      "loss": 2.7381,
+      "step": 1625
+    },
+    {
+      "epoch": 21.682274247491637,
+      "grad_norm": 0.6892338395118713,
+      "learning_rate": 0.0001279199110122358,
+      "loss": 2.7516,
+      "step": 1626
+    },
+    {
+      "epoch": 21.695652173913043,
+      "grad_norm": 0.5888476967811584,
+      "learning_rate": 0.0001278754171301446,
+      "loss": 2.9234,
+      "step": 1627
+    },
+    {
+      "epoch": 21.70903010033445,
+      "grad_norm": 0.6148846745491028,
+      "learning_rate": 0.0001278309232480534,
+      "loss": 3.0824,
+      "step": 1628
+    },
+    {
+      "epoch": 21.722408026755854,
+      "grad_norm": 0.6559250354766846,
+      "learning_rate": 0.00012778642936596218,
+      "loss": 3.0316,
+      "step": 1629
+    },
+    {
+      "epoch": 21.735785953177256,
+      "grad_norm": 0.5679346919059753,
+      "learning_rate": 0.00012774193548387096,
+      "loss": 3.0012,
+      "step": 1630
+    },
+    {
+      "epoch": 21.74916387959866,
+      "grad_norm": 0.6776624917984009,
+      "learning_rate": 0.00012769744160177977,
+      "loss": 2.8331,
+      "step": 1631
+    },
+    {
+      "epoch": 21.762541806020067,
+      "grad_norm": 0.6982457637786865,
+      "learning_rate": 0.00012765294771968855,
+      "loss": 2.8175,
+      "step": 1632
+    },
+    {
+      "epoch": 21.775919732441473,
+      "grad_norm": 0.6834619045257568,
+      "learning_rate": 0.00012760845383759733,
+      "loss": 3.1136,
+      "step": 1633
+    },
+    {
+      "epoch": 21.789297658862875,
+      "grad_norm": 0.6460130214691162,
+      "learning_rate": 0.00012756395995550612,
+      "loss": 2.7737,
+      "step": 1634
+    },
+    {
+      "epoch": 21.80267558528428,
+      "grad_norm": 0.6818533539772034,
+      "learning_rate": 0.00012751946607341493,
+      "loss": 2.5013,
+      "step": 1635
+    },
+    {
+      "epoch": 21.816053511705686,
+      "grad_norm": 0.7875826954841614,
+      "learning_rate": 0.00012747497219132368,
+      "loss": 2.9702,
+      "step": 1636
+    },
+    {
+      "epoch": 21.82943143812709,
+      "grad_norm": 0.5845924019813538,
+      "learning_rate": 0.0001274304783092325,
+      "loss": 2.8592,
+      "step": 1637
+    },
+    {
+      "epoch": 21.842809364548494,
+      "grad_norm": 0.6136084794998169,
+      "learning_rate": 0.00012738598442714127,
+      "loss": 2.8456,
+      "step": 1638
+    },
+    {
+      "epoch": 21.8561872909699,
+      "grad_norm": 0.7345831394195557,
+      "learning_rate": 0.00012734149054505005,
+      "loss": 2.9846,
+      "step": 1639
+    },
+    {
+      "epoch": 21.869565217391305,
+      "grad_norm": 0.5709069967269897,
+      "learning_rate": 0.00012729699666295884,
+      "loss": 2.9736,
+      "step": 1640
+    },
+    {
+      "epoch": 21.88294314381271,
+      "grad_norm": 0.7338367104530334,
+      "learning_rate": 0.00012725250278086765,
+      "loss": 2.8215,
+      "step": 1641
+    },
+    {
+      "epoch": 21.896321070234112,
+      "grad_norm": 0.6029645800590515,
+      "learning_rate": 0.00012720800889877643,
+      "loss": 2.7419,
+      "step": 1642
+    },
+    {
+      "epoch": 21.909698996655518,
+      "grad_norm": 0.7492797374725342,
+      "learning_rate": 0.0001271635150166852,
+      "loss": 2.8256,
+      "step": 1643
+    },
+    {
+      "epoch": 21.923076923076923,
+      "grad_norm": 0.6877094507217407,
+      "learning_rate": 0.000127119021134594,
+      "loss": 2.8241,
+      "step": 1644
+    },
+    {
+      "epoch": 21.93645484949833,
+      "grad_norm": 0.6263371109962463,
+      "learning_rate": 0.0001270745272525028,
+      "loss": 2.8459,
+      "step": 1645
+    },
+    {
+      "epoch": 21.94983277591973,
+      "grad_norm": 0.5821470618247986,
+      "learning_rate": 0.00012703003337041156,
+      "loss": 3.0513,
+      "step": 1646
+    },
+    {
+      "epoch": 21.963210702341136,
+      "grad_norm": 0.5494033098220825,
+      "learning_rate": 0.00012698553948832037,
+      "loss": 3.0951,
+      "step": 1647
+    },
+    {
+      "epoch": 21.976588628762542,
+      "grad_norm": 0.5838212370872498,
+      "learning_rate": 0.00012694104560622915,
+      "loss": 3.0089,
+      "step": 1648
+    },
+    {
+      "epoch": 21.989966555183948,
+      "grad_norm": 0.7115112543106079,
+      "learning_rate": 0.00012689655172413793,
+      "loss": 2.9137,
+      "step": 1649
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 0.7208350896835327,
+      "learning_rate": 0.00012685205784204672,
+      "loss": 2.8793,
+      "step": 1650
+    },
+    {
+      "epoch": 22.013377926421406,
+      "grad_norm": 0.912466824054718,
+      "learning_rate": 0.00012680756395995552,
+      "loss": 2.7563,
+      "step": 1651
+    },
+    {
+      "epoch": 22.02675585284281,
+      "grad_norm": 0.7778460383415222,
+      "learning_rate": 0.0001267630700778643,
+      "loss": 2.7859,
+      "step": 1652
+    },
+    {
+      "epoch": 22.040133779264213,
+      "grad_norm": 0.6658245921134949,
+      "learning_rate": 0.0001267185761957731,
+      "loss": 2.8353,
+      "step": 1653
+    },
+    {
+      "epoch": 22.05351170568562,
+      "grad_norm": 0.5711331963539124,
+      "learning_rate": 0.00012667408231368187,
+      "loss": 2.7731,
+      "step": 1654
+    },
+    {
+      "epoch": 22.066889632107024,
+      "grad_norm": 0.6227294206619263,
+      "learning_rate": 0.00012662958843159068,
+      "loss": 2.6841,
+      "step": 1655
+    },
+    {
+      "epoch": 22.08026755852843,
+      "grad_norm": 0.6221343874931335,
+      "learning_rate": 0.00012658509454949944,
+      "loss": 2.6303,
+      "step": 1656
+    },
+    {
+      "epoch": 22.093645484949832,
+      "grad_norm": 0.6461536288261414,
+      "learning_rate": 0.00012654060066740822,
+      "loss": 2.7591,
+      "step": 1657
+    },
+    {
+      "epoch": 22.107023411371237,
+      "grad_norm": 0.9253934025764465,
+      "learning_rate": 0.00012649610678531703,
+      "loss": 2.7619,
+      "step": 1658
+    },
+    {
+      "epoch": 22.120401337792643,
+      "grad_norm": 1.0310887098312378,
+      "learning_rate": 0.0001264516129032258,
+      "loss": 2.7614,
+      "step": 1659
+    },
+    {
+      "epoch": 22.13377926421405,
+      "grad_norm": 0.682077944278717,
+      "learning_rate": 0.0001264071190211346,
+      "loss": 2.9013,
+      "step": 1660
+    },
+    {
+      "epoch": 22.14715719063545,
+      "grad_norm": 0.6438060998916626,
+      "learning_rate": 0.00012636262513904338,
+      "loss": 2.8715,
+      "step": 1661
+    },
+    {
+      "epoch": 22.160535117056856,
+      "grad_norm": 0.5626085996627808,
+      "learning_rate": 0.00012631813125695219,
+      "loss": 2.7459,
+      "step": 1662
+    },
+    {
+      "epoch": 22.17391304347826,
+      "grad_norm": 0.6211214065551758,
+      "learning_rate": 0.00012627363737486094,
+      "loss": 2.9223,
+      "step": 1663
+    },
+    {
+      "epoch": 22.187290969899667,
+      "grad_norm": 0.5843247175216675,
+      "learning_rate": 0.00012622914349276975,
+      "loss": 3.0176,
+      "step": 1664
+    },
+    {
+      "epoch": 22.20066889632107,
+      "grad_norm": 0.6780831813812256,
+      "learning_rate": 0.00012618464961067853,
+      "loss": 3.0641,
+      "step": 1665
+    },
+    {
+      "epoch": 22.214046822742475,
+      "grad_norm": 0.6507664918899536,
+      "learning_rate": 0.00012614015572858731,
+      "loss": 2.9043,
+      "step": 1666
+    },
+    {
+      "epoch": 22.22742474916388,
+      "grad_norm": 0.6472734212875366,
+      "learning_rate": 0.0001260956618464961,
+      "loss": 2.678,
+      "step": 1667
+    },
+    {
+      "epoch": 22.240802675585286,
+      "grad_norm": 0.5920576453208923,
+      "learning_rate": 0.0001260511679644049,
+      "loss": 2.9801,
+      "step": 1668
+    },
+    {
+      "epoch": 22.254180602006688,
+      "grad_norm": 0.5785727500915527,
+      "learning_rate": 0.0001260066740823137,
+      "loss": 2.6949,
+      "step": 1669
+    },
+    {
+      "epoch": 22.267558528428093,
+      "grad_norm": 0.8260320425033569,
+      "learning_rate": 0.00012596218020022247,
+      "loss": 2.939,
+      "step": 1670
+    },
+    {
+      "epoch": 22.2809364548495,
+      "grad_norm": 0.6697145700454712,
+      "learning_rate": 0.00012591768631813125,
+      "loss": 2.6997,
+      "step": 1671
+    },
+    {
+      "epoch": 22.294314381270905,
+      "grad_norm": 0.6255594491958618,
+      "learning_rate": 0.00012587319243604006,
+      "loss": 2.7875,
+      "step": 1672
+    },
+    {
+      "epoch": 22.307692307692307,
+      "grad_norm": 0.707083523273468,
+      "learning_rate": 0.00012582869855394882,
+      "loss": 2.8843,
+      "step": 1673
+    },
+    {
+      "epoch": 22.321070234113712,
+      "grad_norm": 0.5888648629188538,
+      "learning_rate": 0.00012578420467185763,
+      "loss": 2.8862,
+      "step": 1674
+    },
+    {
+      "epoch": 22.334448160535118,
+      "grad_norm": 0.6249898672103882,
+      "learning_rate": 0.0001257397107897664,
+      "loss": 2.6108,
+      "step": 1675
+    },
+    {
+      "epoch": 22.347826086956523,
+      "grad_norm": 0.6932422518730164,
+      "learning_rate": 0.0001256952169076752,
+      "loss": 2.8393,
+      "step": 1676
+    },
+    {
+      "epoch": 22.361204013377925,
+      "grad_norm": 0.6037236452102661,
+      "learning_rate": 0.00012565072302558398,
+      "loss": 2.6517,
+      "step": 1677
+    },
+    {
+      "epoch": 22.37458193979933,
+      "grad_norm": 0.5549351572990417,
+      "learning_rate": 0.00012560622914349278,
+      "loss": 2.7725,
+      "step": 1678
+    },
+    {
+      "epoch": 22.387959866220736,
+      "grad_norm": 0.6309815645217896,
+      "learning_rate": 0.00012556173526140157,
+      "loss": 2.7539,
+      "step": 1679
+    },
+    {
+      "epoch": 22.401337792642142,
+      "grad_norm": 0.6618108153343201,
+      "learning_rate": 0.00012551724137931035,
+      "loss": 2.9342,
+      "step": 1680
+    },
+    {
+      "epoch": 22.414715719063544,
+      "grad_norm": 0.7427678108215332,
+      "learning_rate": 0.00012547274749721913,
+      "loss": 2.9707,
+      "step": 1681
+    },
+    {
+      "epoch": 22.42809364548495,
+      "grad_norm": 0.6226203441619873,
+      "learning_rate": 0.00012542825361512794,
+      "loss": 2.728,
+      "step": 1682
+    },
+    {
+      "epoch": 22.441471571906355,
+      "grad_norm": 0.6432873606681824,
+      "learning_rate": 0.0001253837597330367,
+      "loss": 2.6271,
+      "step": 1683
+    },
+    {
+      "epoch": 22.45484949832776,
+      "grad_norm": 0.618368923664093,
+      "learning_rate": 0.0001253392658509455,
+      "loss": 2.7364,
+      "step": 1684
+    },
+    {
+      "epoch": 22.468227424749163,
+      "grad_norm": 0.601243257522583,
+      "learning_rate": 0.0001252947719688543,
+      "loss": 2.7071,
+      "step": 1685
+    },
+    {
+      "epoch": 22.48160535117057,
+      "grad_norm": 0.5549631714820862,
+      "learning_rate": 0.00012525027808676307,
+      "loss": 2.7703,
+      "step": 1686
+    },
+    {
+      "epoch": 22.494983277591974,
+      "grad_norm": 0.8125869035720825,
+      "learning_rate": 0.00012520578420467185,
+      "loss": 2.6798,
+      "step": 1687
+    },
+    {
+      "epoch": 22.50836120401338,
+      "grad_norm": 0.5712392330169678,
+      "learning_rate": 0.00012516129032258066,
+      "loss": 2.7694,
+      "step": 1688
+    },
+    {
+      "epoch": 22.52173913043478,
+      "grad_norm": 0.5955860614776611,
+      "learning_rate": 0.00012511679644048945,
+      "loss": 2.8128,
+      "step": 1689
+    },
+    {
+      "epoch": 22.535117056856187,
+      "grad_norm": 0.6090961694717407,
+      "learning_rate": 0.00012507230255839823,
+      "loss": 2.9096,
+      "step": 1690
+    },
+    {
+      "epoch": 22.548494983277592,
+      "grad_norm": 0.6430270671844482,
+      "learning_rate": 0.000125027808676307,
+      "loss": 2.8046,
+      "step": 1691
+    },
+    {
+      "epoch": 22.561872909698998,
+      "grad_norm": 0.6325036883354187,
+      "learning_rate": 0.00012498331479421582,
+      "loss": 2.9619,
+      "step": 1692
+    },
+    {
+      "epoch": 22.5752508361204,
+      "grad_norm": 0.5997435450553894,
+      "learning_rate": 0.00012493882091212458,
+      "loss": 2.9809,
+      "step": 1693
+    },
+    {
+      "epoch": 22.588628762541806,
+      "grad_norm": 0.6140502095222473,
+      "learning_rate": 0.00012489432703003338,
+      "loss": 2.9622,
+      "step": 1694
+    },
+    {
+      "epoch": 22.60200668896321,
+      "grad_norm": 0.5908461809158325,
+      "learning_rate": 0.00012484983314794217,
+      "loss": 2.8282,
+      "step": 1695
+    },
+    {
+      "epoch": 22.615384615384617,
+      "grad_norm": 0.5987250208854675,
+      "learning_rate": 0.00012480533926585095,
+      "loss": 2.8239,
+      "step": 1696
+    },
+    {
+      "epoch": 22.62876254180602,
+      "grad_norm": 0.6643062233924866,
+      "learning_rate": 0.00012476084538375973,
+      "loss": 2.8677,
+      "step": 1697
+    },
+    {
+      "epoch": 22.642140468227424,
+      "grad_norm": 0.640986442565918,
+      "learning_rate": 0.00012471635150166851,
+      "loss": 2.6359,
+      "step": 1698
+    },
+    {
+      "epoch": 22.65551839464883,
+      "grad_norm": 0.6315325498580933,
+      "learning_rate": 0.00012467185761957732,
+      "loss": 2.7919,
+      "step": 1699
+    },
+    {
+      "epoch": 22.668896321070235,
+      "grad_norm": 0.6384916305541992,
+      "learning_rate": 0.00012462736373748608,
+      "loss": 3.0353,
+      "step": 1700
+    },
+    {
+      "epoch": 22.682274247491637,
+      "grad_norm": 0.6934499144554138,
+      "learning_rate": 0.0001245828698553949,
+      "loss": 2.8694,
+      "step": 1701
+    },
+    {
+      "epoch": 22.695652173913043,
+      "grad_norm": 0.6261676549911499,
+      "learning_rate": 0.00012453837597330367,
+      "loss": 2.6893,
+      "step": 1702
+    },
+    {
+      "epoch": 22.70903010033445,
+      "grad_norm": 0.6069616079330444,
+      "learning_rate": 0.00012449388209121245,
+      "loss": 2.9214,
+      "step": 1703
+    },
+    {
+      "epoch": 22.722408026755854,
+      "grad_norm": 0.6354973316192627,
+      "learning_rate": 0.00012444938820912124,
+      "loss": 2.754,
+      "step": 1704
+    },
+    {
+      "epoch": 22.735785953177256,
+      "grad_norm": 0.7385901808738708,
+      "learning_rate": 0.00012440489432703004,
+      "loss": 2.7633,
+      "step": 1705
+    },
+    {
+      "epoch": 22.74916387959866,
+      "grad_norm": 0.680008053779602,
+      "learning_rate": 0.00012436040044493883,
+      "loss": 2.851,
+      "step": 1706
+    },
+    {
+      "epoch": 22.762541806020067,
+      "grad_norm": 0.5894846320152283,
+      "learning_rate": 0.0001243159065628476,
+      "loss": 2.8797,
+      "step": 1707
+    },
+    {
+      "epoch": 22.775919732441473,
+      "grad_norm": 0.5705847144126892,
+      "learning_rate": 0.0001242714126807564,
+      "loss": 2.7471,
+      "step": 1708
+    },
+    {
+      "epoch": 22.789297658862875,
+      "grad_norm": 0.6021112203598022,
+      "learning_rate": 0.0001242269187986652,
+      "loss": 3.0716,
+      "step": 1709
+    },
+    {
+      "epoch": 22.80267558528428,
+      "grad_norm": 0.63507080078125,
+      "learning_rate": 0.00012418242491657396,
+      "loss": 2.8873,
+      "step": 1710
+    },
+    {
+      "epoch": 22.816053511705686,
+      "grad_norm": 0.6159088015556335,
+      "learning_rate": 0.00012413793103448277,
+      "loss": 2.5226,
+      "step": 1711
+    },
+    {
+      "epoch": 22.82943143812709,
+      "grad_norm": 0.6099200248718262,
+      "learning_rate": 0.00012409343715239155,
+      "loss": 2.9937,
+      "step": 1712
+    },
+    {
+      "epoch": 22.842809364548494,
+      "grad_norm": 0.5942363142967224,
+      "learning_rate": 0.00012404894327030033,
+      "loss": 2.5113,
+      "step": 1713
+    },
+    {
+      "epoch": 22.8561872909699,
+      "grad_norm": 0.6209394335746765,
+      "learning_rate": 0.00012400444938820911,
+      "loss": 3.0766,
+      "step": 1714
+    },
+    {
+      "epoch": 22.869565217391305,
+      "grad_norm": 0.6850425601005554,
+      "learning_rate": 0.00012395995550611792,
+      "loss": 2.8807,
+      "step": 1715
+    },
+    {
+      "epoch": 22.88294314381271,
+      "grad_norm": 0.5698277354240417,
+      "learning_rate": 0.0001239154616240267,
+      "loss": 2.6821,
+      "step": 1716
+    },
+    {
+      "epoch": 22.896321070234112,
+      "grad_norm": 0.7140303254127502,
+      "learning_rate": 0.0001238709677419355,
+      "loss": 2.7047,
+      "step": 1717
+    },
+    {
+      "epoch": 22.909698996655518,
+      "grad_norm": 0.6129325032234192,
+      "learning_rate": 0.00012382647385984427,
+      "loss": 2.705,
+      "step": 1718
+    },
+    {
+      "epoch": 22.923076923076923,
+      "grad_norm": 0.6409560441970825,
+      "learning_rate": 0.00012378197997775308,
+      "loss": 2.7126,
+      "step": 1719
+    },
+    {
+      "epoch": 22.93645484949833,
+      "grad_norm": 0.5590237975120544,
+      "learning_rate": 0.00012373748609566184,
+      "loss": 2.7945,
+      "step": 1720
+    },
+    {
+      "epoch": 22.94983277591973,
+      "grad_norm": 0.5763843655586243,
+      "learning_rate": 0.00012369299221357064,
+      "loss": 2.6802,
+      "step": 1721
+    },
+    {
+      "epoch": 22.963210702341136,
+      "grad_norm": 0.6224062442779541,
+      "learning_rate": 0.00012364849833147943,
+      "loss": 2.7035,
+      "step": 1722
+    },
+    {
+      "epoch": 22.976588628762542,
+      "grad_norm": 0.5638879537582397,
+      "learning_rate": 0.0001236040044493882,
+      "loss": 2.8259,
+      "step": 1723
+    },
+    {
+      "epoch": 22.989966555183948,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.000123559510567297,
+      "loss": 2.9997,
+      "step": 1724
+    },
+    {
+      "epoch": 23.0,
+      "grad_norm": 0.6950054168701172,
+      "learning_rate": 0.0001235150166852058,
+      "loss": 2.7497,
+      "step": 1725
+    },
+    {
+      "epoch": 23.013377926421406,
+      "grad_norm": 0.67231285572052,
+      "learning_rate": 0.00012347052280311458,
+      "loss": 2.859,
+      "step": 1726
+    },
+    {
+      "epoch": 23.02675585284281,
+      "grad_norm": 0.6401329040527344,
+      "learning_rate": 0.00012342602892102337,
+      "loss": 2.7056,
+      "step": 1727
+    },
+    {
+      "epoch": 23.040133779264213,
+      "grad_norm": 0.6234462261199951,
+      "learning_rate": 0.00012338153503893215,
+      "loss": 2.8271,
+      "step": 1728
+    },
+    {
+      "epoch": 23.05351170568562,
+      "grad_norm": 0.6482179164886475,
+      "learning_rate": 0.00012333704115684096,
+      "loss": 2.898,
+      "step": 1729
+    },
+    {
+      "epoch": 23.066889632107024,
+      "grad_norm": 0.6077515482902527,
+      "learning_rate": 0.0001232925472747497,
+      "loss": 2.6865,
+      "step": 1730
+    },
+    {
+      "epoch": 23.08026755852843,
+      "grad_norm": 0.5866036415100098,
+      "learning_rate": 0.00012324805339265852,
+      "loss": 2.9001,
+      "step": 1731
+    },
+    {
+      "epoch": 23.093645484949832,
+      "grad_norm": 0.6499370336532593,
+      "learning_rate": 0.0001232035595105673,
+      "loss": 2.8718,
+      "step": 1732
+    },
+    {
+      "epoch": 23.107023411371237,
+      "grad_norm": 0.6188756227493286,
+      "learning_rate": 0.0001231590656284761,
+      "loss": 2.7928,
+      "step": 1733
+    },
+    {
+      "epoch": 23.120401337792643,
+      "grad_norm": 0.5990676879882812,
+      "learning_rate": 0.00012311457174638487,
+      "loss": 2.7918,
+      "step": 1734
+    },
+    {
+      "epoch": 23.13377926421405,
+      "grad_norm": 0.5592599511146545,
+      "learning_rate": 0.00012307007786429368,
+      "loss": 2.7948,
+      "step": 1735
+    },
+    {
+      "epoch": 23.14715719063545,
+      "grad_norm": 0.5751471519470215,
+      "learning_rate": 0.00012302558398220246,
+      "loss": 2.6472,
+      "step": 1736
+    },
+    {
+      "epoch": 23.160535117056856,
+      "grad_norm": 0.6300846338272095,
+      "learning_rate": 0.00012298109010011124,
+      "loss": 2.7491,
+      "step": 1737
+    },
+    {
+      "epoch": 23.17391304347826,
+      "grad_norm": 0.5754296779632568,
+      "learning_rate": 0.00012293659621802003,
+      "loss": 2.7557,
+      "step": 1738
+    },
+    {
+      "epoch": 23.187290969899667,
+      "grad_norm": 0.6483306884765625,
+      "learning_rate": 0.0001228921023359288,
+      "loss": 2.776,
+      "step": 1739
+    },
+    {
+      "epoch": 23.20066889632107,
+      "grad_norm": 0.5819816589355469,
+      "learning_rate": 0.0001228476084538376,
+      "loss": 2.9192,
+      "step": 1740
+    },
+    {
+      "epoch": 23.214046822742475,
+      "grad_norm": 0.6570101380348206,
+      "learning_rate": 0.00012280311457174637,
+      "loss": 2.8571,
+      "step": 1741
+    },
+    {
+      "epoch": 23.22742474916388,
+      "grad_norm": 0.6435232162475586,
+      "learning_rate": 0.00012275862068965518,
+      "loss": 2.6134,
+      "step": 1742
+    },
+    {
+      "epoch": 23.240802675585286,
+      "grad_norm": 0.5817757844924927,
+      "learning_rate": 0.00012271412680756397,
+      "loss": 2.8384,
+      "step": 1743
+    },
+    {
+      "epoch": 23.254180602006688,
+      "grad_norm": 0.5564932823181152,
+      "learning_rate": 0.00012266963292547275,
+      "loss": 2.4606,
+      "step": 1744
+    },
+    {
+      "epoch": 23.267558528428093,
+      "grad_norm": 0.6028578281402588,
+      "learning_rate": 0.00012262513904338153,
+      "loss": 2.824,
+      "step": 1745
+    },
+    {
+      "epoch": 23.2809364548495,
+      "grad_norm": 0.5562644600868225,
+      "learning_rate": 0.00012258064516129034,
+      "loss": 2.8195,
+      "step": 1746
+    },
+    {
+      "epoch": 23.294314381270905,
+      "grad_norm": 0.6838886141777039,
+      "learning_rate": 0.0001225361512791991,
+      "loss": 2.8453,
+      "step": 1747
+    },
+    {
+      "epoch": 23.307692307692307,
+      "grad_norm": 0.6102856993675232,
+      "learning_rate": 0.0001224916573971079,
+      "loss": 2.8331,
+      "step": 1748
+    },
+    {
+      "epoch": 23.321070234113712,
+      "grad_norm": 0.5895605087280273,
+      "learning_rate": 0.0001224471635150167,
+      "loss": 2.7479,
+      "step": 1749
+    },
+    {
+      "epoch": 23.334448160535118,
+      "grad_norm": 0.7318586111068726,
+      "learning_rate": 0.00012240266963292547,
+      "loss": 2.7201,
+      "step": 1750
+    },
+    {
+      "epoch": 23.347826086956523,
+      "grad_norm": 0.6560878157615662,
+      "learning_rate": 0.00012235817575083425,
+      "loss": 2.931,
+      "step": 1751
+    },
+    {
+      "epoch": 23.361204013377925,
+      "grad_norm": 0.5879004001617432,
+      "learning_rate": 0.00012231368186874306,
+      "loss": 2.8607,
+      "step": 1752
+    },
+    {
+      "epoch": 23.37458193979933,
+      "grad_norm": 0.5916743278503418,
+      "learning_rate": 0.00012226918798665184,
+      "loss": 2.6923,
+      "step": 1753
+    },
+    {
+      "epoch": 23.387959866220736,
+      "grad_norm": 0.5569814443588257,
+      "learning_rate": 0.00012222469410456063,
+      "loss": 2.634,
+      "step": 1754
+    },
+    {
+      "epoch": 23.401337792642142,
+      "grad_norm": 0.7285422682762146,
+      "learning_rate": 0.0001221802002224694,
+      "loss": 2.7135,
+      "step": 1755
+    },
+    {
+      "epoch": 23.414715719063544,
+      "grad_norm": 0.7323710918426514,
+      "learning_rate": 0.00012213570634037822,
+      "loss": 2.6855,
+      "step": 1756
+    },
+    {
+      "epoch": 23.42809364548495,
+      "grad_norm": 0.65497887134552,
+      "learning_rate": 0.00012209121245828697,
+      "loss": 2.8777,
+      "step": 1757
+    },
+    {
+      "epoch": 23.441471571906355,
+      "grad_norm": 0.5919074416160583,
+      "learning_rate": 0.00012204671857619578,
+      "loss": 2.657,
+      "step": 1758
+    },
+    {
+      "epoch": 23.45484949832776,
+      "grad_norm": 0.6749523282051086,
+      "learning_rate": 0.00012200222469410457,
+      "loss": 2.8142,
+      "step": 1759
+    },
+    {
+      "epoch": 23.468227424749163,
+      "grad_norm": 0.6859252452850342,
+      "learning_rate": 0.00012195773081201336,
+      "loss": 2.8572,
+      "step": 1760
+    },
+    {
+      "epoch": 23.48160535117057,
+      "grad_norm": 0.5956023931503296,
+      "learning_rate": 0.00012191323692992213,
+      "loss": 2.8385,
+      "step": 1761
+    },
+    {
+      "epoch": 23.494983277591974,
+      "grad_norm": 0.684101939201355,
+      "learning_rate": 0.00012186874304783094,
+      "loss": 2.4779,
+      "step": 1762
+    },
+    {
+      "epoch": 23.50836120401338,
+      "grad_norm": 0.726864755153656,
+      "learning_rate": 0.00012182424916573971,
+      "loss": 2.8523,
+      "step": 1763
+    },
+    {
+      "epoch": 23.52173913043478,
+      "grad_norm": 0.5874570608139038,
+      "learning_rate": 0.00012177975528364852,
+      "loss": 2.7866,
+      "step": 1764
+    },
+    {
+      "epoch": 23.535117056856187,
+      "grad_norm": 0.6077170968055725,
+      "learning_rate": 0.00012173526140155729,
+      "loss": 2.711,
+      "step": 1765
+    },
+    {
+      "epoch": 23.548494983277592,
+      "grad_norm": 0.6854214668273926,
+      "learning_rate": 0.00012169076751946608,
+      "loss": 2.655,
+      "step": 1766
+    },
+    {
+      "epoch": 23.561872909698998,
+      "grad_norm": 0.6409576535224915,
+      "learning_rate": 0.00012164627363737486,
+      "loss": 2.7309,
+      "step": 1767
+    },
+    {
+      "epoch": 23.5752508361204,
+      "grad_norm": 0.6241947412490845,
+      "learning_rate": 0.00012160177975528366,
+      "loss": 2.67,
+      "step": 1768
+    },
+    {
+      "epoch": 23.588628762541806,
+      "grad_norm": 0.5962851643562317,
+      "learning_rate": 0.00012155728587319244,
+      "loss": 2.6147,
+      "step": 1769
+    },
+    {
+      "epoch": 23.60200668896321,
+      "grad_norm": 0.6183756589889526,
+      "learning_rate": 0.00012151279199110124,
+      "loss": 2.9772,
+      "step": 1770
+    },
+    {
+      "epoch": 23.615384615384617,
+      "grad_norm": 0.6092197299003601,
+      "learning_rate": 0.00012146829810901001,
+      "loss": 2.7146,
+      "step": 1771
+    },
+    {
+      "epoch": 23.62876254180602,
+      "grad_norm": 0.6660104990005493,
+      "learning_rate": 0.00012142380422691882,
+      "loss": 2.6339,
+      "step": 1772
+    },
+    {
+      "epoch": 23.642140468227424,
+      "grad_norm": 0.9198834896087646,
+      "learning_rate": 0.00012137931034482759,
+      "loss": 2.7188,
+      "step": 1773
+    },
+    {
+      "epoch": 23.65551839464883,
+      "grad_norm": 0.7053755521774292,
+      "learning_rate": 0.0001213348164627364,
+      "loss": 2.7484,
+      "step": 1774
+    },
+    {
+      "epoch": 23.668896321070235,
+      "grad_norm": 0.7443736791610718,
+      "learning_rate": 0.00012129032258064516,
+      "loss": 2.6805,
+      "step": 1775
+    },
+    {
+      "epoch": 23.682274247491637,
+      "grad_norm": 0.579801082611084,
+      "learning_rate": 0.00012124582869855396,
+      "loss": 2.6467,
+      "step": 1776
+    },
+    {
+      "epoch": 23.695652173913043,
+      "grad_norm": 0.6014502644538879,
+      "learning_rate": 0.00012120133481646274,
+      "loss": 2.5816,
+      "step": 1777
+    },
+    {
+      "epoch": 23.70903010033445,
+      "grad_norm": 0.6318315863609314,
+      "learning_rate": 0.00012115684093437154,
+      "loss": 2.8137,
+      "step": 1778
+    },
+    {
+      "epoch": 23.722408026755854,
+      "grad_norm": 0.6172413229942322,
+      "learning_rate": 0.00012111234705228032,
+      "loss": 2.7789,
+      "step": 1779
+    },
+    {
+      "epoch": 23.735785953177256,
+      "grad_norm": 0.6184534430503845,
+      "learning_rate": 0.00012106785317018909,
+      "loss": 2.6802,
+      "step": 1780
+    },
+    {
+      "epoch": 23.74916387959866,
+      "grad_norm": 0.6380288600921631,
+      "learning_rate": 0.00012102335928809789,
+      "loss": 2.8755,
+      "step": 1781
+    },
+    {
+      "epoch": 23.762541806020067,
+      "grad_norm": 0.5941389799118042,
+      "learning_rate": 0.00012097886540600667,
+      "loss": 2.945,
+      "step": 1782
+    },
+    {
+      "epoch": 23.775919732441473,
+      "grad_norm": 0.6913108825683594,
+      "learning_rate": 0.00012093437152391546,
+      "loss": 2.7036,
+      "step": 1783
+    },
+    {
+      "epoch": 23.789297658862875,
+      "grad_norm": 0.563119113445282,
+      "learning_rate": 0.00012088987764182425,
+      "loss": 2.6608,
+      "step": 1784
+    },
+    {
+      "epoch": 23.80267558528428,
+      "grad_norm": 0.6387828588485718,
+      "learning_rate": 0.00012084538375973304,
+      "loss": 2.8933,
+      "step": 1785
+    },
+    {
+      "epoch": 23.816053511705686,
+      "grad_norm": 0.7530612945556641,
+      "learning_rate": 0.00012080088987764183,
+      "loss": 2.8399,
+      "step": 1786
+    },
+    {
+      "epoch": 23.82943143812709,
+      "grad_norm": 0.6401646733283997,
+      "learning_rate": 0.00012075639599555062,
+      "loss": 2.8757,
+      "step": 1787
+    },
+    {
+      "epoch": 23.842809364548494,
+      "grad_norm": 0.7403398752212524,
+      "learning_rate": 0.00012071190211345939,
+      "loss": 2.629,
+      "step": 1788
+    },
+    {
+      "epoch": 23.8561872909699,
+      "grad_norm": 0.6479887366294861,
+      "learning_rate": 0.0001206674082313682,
+      "loss": 2.9069,
+      "step": 1789
+    },
+    {
+      "epoch": 23.869565217391305,
+      "grad_norm": 0.588141679763794,
+      "learning_rate": 0.00012062291434927697,
+      "loss": 2.732,
+      "step": 1790
+    },
+    {
+      "epoch": 23.88294314381271,
+      "grad_norm": 0.6330631971359253,
+      "learning_rate": 0.00012057842046718576,
+      "loss": 2.9348,
+      "step": 1791
+    },
+    {
+      "epoch": 23.896321070234112,
+      "grad_norm": 0.6657344698905945,
+      "learning_rate": 0.00012053392658509455,
+      "loss": 3.0499,
+      "step": 1792
+    },
+    {
+      "epoch": 23.909698996655518,
+      "grad_norm": 0.6816025972366333,
+      "learning_rate": 0.00012048943270300334,
+      "loss": 2.9961,
+      "step": 1793
+    },
+    {
+      "epoch": 23.923076923076923,
+      "grad_norm": 0.6501593589782715,
+      "learning_rate": 0.00012044493882091212,
+      "loss": 2.8555,
+      "step": 1794
+    },
+    {
+      "epoch": 23.93645484949833,
+      "grad_norm": 0.6157960295677185,
+      "learning_rate": 0.00012040044493882092,
+      "loss": 2.8133,
+      "step": 1795
+    },
+    {
+      "epoch": 23.94983277591973,
+      "grad_norm": 0.5868191719055176,
+      "learning_rate": 0.0001203559510567297,
+      "loss": 2.7887,
+      "step": 1796
+    },
+    {
+      "epoch": 23.963210702341136,
+      "grad_norm": 0.6586911678314209,
+      "learning_rate": 0.0001203114571746385,
+      "loss": 3.0916,
+      "step": 1797
+    },
+    {
+      "epoch": 23.976588628762542,
+      "grad_norm": 0.6382162570953369,
+      "learning_rate": 0.00012026696329254727,
+      "loss": 2.7469,
+      "step": 1798
+    },
+    {
+      "epoch": 23.989966555183948,
+      "grad_norm": 0.6481708288192749,
+      "learning_rate": 0.00012022246941045608,
+      "loss": 2.9216,
+      "step": 1799
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 0.7272862792015076,
+      "learning_rate": 0.00012017797552836485,
+      "loss": 2.686,
+      "step": 1800
+    },
+    {
+      "epoch": 24.013377926421406,
+      "grad_norm": 0.743366003036499,
+      "learning_rate": 0.00012013348164627364,
+      "loss": 2.6791,
+      "step": 1801
+    },
+    {
+      "epoch": 24.02675585284281,
+      "grad_norm": 0.6428125500679016,
+      "learning_rate": 0.00012008898776418242,
+      "loss": 2.782,
+      "step": 1802
+    },
+    {
+      "epoch": 24.040133779264213,
+      "grad_norm": 0.6210600137710571,
+      "learning_rate": 0.00012004449388209122,
+      "loss": 2.9352,
+      "step": 1803
+    },
+    {
+      "epoch": 24.05351170568562,
+      "grad_norm": 0.9589283466339111,
+      "learning_rate": 0.00012,
+      "loss": 2.7686,
+      "step": 1804
+    },
+    {
+      "epoch": 24.066889632107024,
+      "grad_norm": 0.5503745675086975,
+      "learning_rate": 0.0001199555061179088,
+      "loss": 2.9241,
+      "step": 1805
+    },
+    {
+      "epoch": 24.08026755852843,
+      "grad_norm": 0.6018356084823608,
+      "learning_rate": 0.00011991101223581758,
+      "loss": 2.7512,
+      "step": 1806
+    },
+    {
+      "epoch": 24.093645484949832,
+      "grad_norm": 0.7458929419517517,
+      "learning_rate": 0.00011986651835372638,
+      "loss": 2.5795,
+      "step": 1807
+    },
+    {
+      "epoch": 24.107023411371237,
+      "grad_norm": 0.705740213394165,
+      "learning_rate": 0.00011982202447163515,
+      "loss": 2.8537,
+      "step": 1808
+    },
+    {
+      "epoch": 24.120401337792643,
+      "grad_norm": 0.6299166679382324,
+      "learning_rate": 0.00011977753058954396,
+      "loss": 2.7815,
+      "step": 1809
+    },
+    {
+      "epoch": 24.13377926421405,
+      "grad_norm": 1.6271346807479858,
+      "learning_rate": 0.00011973303670745272,
+      "loss": 2.6373,
+      "step": 1810
+    },
+    {
+      "epoch": 24.14715719063545,
+      "grad_norm": 0.6286007165908813,
+      "learning_rate": 0.00011968854282536152,
+      "loss": 2.5587,
+      "step": 1811
+    },
+    {
+      "epoch": 24.160535117056856,
+      "grad_norm": 0.6259005069732666,
+      "learning_rate": 0.0001196440489432703,
+      "loss": 2.6414,
+      "step": 1812
+    },
+    {
+      "epoch": 24.17391304347826,
+      "grad_norm": 0.6058273911476135,
+      "learning_rate": 0.0001195995550611791,
+      "loss": 2.8571,
+      "step": 1813
+    },
+    {
+      "epoch": 24.187290969899667,
+      "grad_norm": 0.6144312620162964,
+      "learning_rate": 0.00011955506117908788,
+      "loss": 2.7777,
+      "step": 1814
+    },
+    {
+      "epoch": 24.20066889632107,
+      "grad_norm": 0.7302254438400269,
+      "learning_rate": 0.00011951056729699668,
+      "loss": 2.6298,
+      "step": 1815
+    },
+    {
+      "epoch": 24.214046822742475,
+      "grad_norm": 0.7253098487854004,
+      "learning_rate": 0.00011946607341490546,
+      "loss": 2.9066,
+      "step": 1816
+    },
+    {
+      "epoch": 24.22742474916388,
+      "grad_norm": 0.7629284858703613,
+      "learning_rate": 0.00011942157953281426,
+      "loss": 2.5391,
+      "step": 1817
+    },
+    {
+      "epoch": 24.240802675585286,
+      "grad_norm": 0.6672796607017517,
+      "learning_rate": 0.00011937708565072302,
+      "loss": 2.8014,
+      "step": 1818
+    },
+    {
+      "epoch": 24.254180602006688,
+      "grad_norm": 0.6049585938453674,
+      "learning_rate": 0.00011933259176863183,
+      "loss": 2.5539,
+      "step": 1819
+    },
+    {
+      "epoch": 24.267558528428093,
+      "grad_norm": 0.5817275047302246,
+      "learning_rate": 0.0001192880978865406,
+      "loss": 2.5771,
+      "step": 1820
+    },
+    {
+      "epoch": 24.2809364548495,
+      "grad_norm": 0.6785464882850647,
+      "learning_rate": 0.00011924360400444938,
+      "loss": 2.8043,
+      "step": 1821
+    },
+    {
+      "epoch": 24.294314381270905,
+      "grad_norm": 0.6705557107925415,
+      "learning_rate": 0.00011919911012235818,
+      "loss": 2.7126,
+      "step": 1822
+    },
+    {
+      "epoch": 24.307692307692307,
+      "grad_norm": 0.6267027854919434,
+      "learning_rate": 0.00011915461624026696,
+      "loss": 2.8739,
+      "step": 1823
+    },
+    {
+      "epoch": 24.321070234113712,
+      "grad_norm": 0.6952410340309143,
+      "learning_rate": 0.00011911012235817576,
+      "loss": 2.8402,
+      "step": 1824
+    },
+    {
+      "epoch": 24.334448160535118,
+      "grad_norm": 0.6892711520195007,
+      "learning_rate": 0.00011906562847608453,
+      "loss": 2.7314,
+      "step": 1825
+    },
+    {
+      "epoch": 24.347826086956523,
+      "grad_norm": 0.5816205739974976,
+      "learning_rate": 0.00011902113459399334,
+      "loss": 2.6115,
+      "step": 1826
+    },
+    {
+      "epoch": 24.361204013377925,
+      "grad_norm": 0.6549028158187866,
+      "learning_rate": 0.0001189766407119021,
+      "loss": 2.7503,
+      "step": 1827
+    },
+    {
+      "epoch": 24.37458193979933,
+      "grad_norm": 0.5902572870254517,
+      "learning_rate": 0.0001189321468298109,
+      "loss": 2.7549,
+      "step": 1828
+    },
+    {
+      "epoch": 24.387959866220736,
+      "grad_norm": 0.8412529230117798,
+      "learning_rate": 0.00011888765294771968,
+      "loss": 2.9654,
+      "step": 1829
+    },
+    {
+      "epoch": 24.401337792642142,
+      "grad_norm": 0.5785757303237915,
+      "learning_rate": 0.00011884315906562848,
+      "loss": 2.8101,
+      "step": 1830
+    },
+    {
+      "epoch": 24.414715719063544,
+      "grad_norm": 0.5835216045379639,
+      "learning_rate": 0.00011879866518353726,
+      "loss": 2.5368,
+      "step": 1831
+    },
+    {
+      "epoch": 24.42809364548495,
+      "grad_norm": 0.6589221954345703,
+      "learning_rate": 0.00011875417130144606,
+      "loss": 2.7917,
+      "step": 1832
+    },
+    {
+      "epoch": 24.441471571906355,
+      "grad_norm": 0.6370606422424316,
+      "learning_rate": 0.00011870967741935484,
+      "loss": 2.7514,
+      "step": 1833
+    },
+    {
+      "epoch": 24.45484949832776,
+      "grad_norm": 0.644980788230896,
+      "learning_rate": 0.00011866518353726364,
+      "loss": 2.6756,
+      "step": 1834
+    },
+    {
+      "epoch": 24.468227424749163,
+      "grad_norm": 0.7281529307365417,
+      "learning_rate": 0.0001186206896551724,
+      "loss": 2.5308,
+      "step": 1835
+    },
+    {
+      "epoch": 24.48160535117057,
+      "grad_norm": 0.5669592618942261,
+      "learning_rate": 0.00011857619577308122,
+      "loss": 2.7197,
+      "step": 1836
+    },
+    {
+      "epoch": 24.494983277591974,
+      "grad_norm": 0.7093492150306702,
+      "learning_rate": 0.00011853170189098998,
+      "loss": 2.9053,
+      "step": 1837
+    },
+    {
+      "epoch": 24.50836120401338,
+      "grad_norm": 0.5949118733406067,
+      "learning_rate": 0.00011848720800889878,
+      "loss": 2.8594,
+      "step": 1838
+    },
+    {
+      "epoch": 24.52173913043478,
+      "grad_norm": 0.5725157260894775,
+      "learning_rate": 0.00011844271412680756,
+      "loss": 2.5906,
+      "step": 1839
+    },
+    {
+      "epoch": 24.535117056856187,
+      "grad_norm": 0.7810790538787842,
+      "learning_rate": 0.00011839822024471636,
+      "loss": 2.7852,
+      "step": 1840
+    },
+    {
+      "epoch": 24.548494983277592,
+      "grad_norm": 0.6285178661346436,
+      "learning_rate": 0.00011835372636262514,
+      "loss": 2.7201,
+      "step": 1841
+    },
+    {
+      "epoch": 24.561872909698998,
+      "grad_norm": 0.6262674927711487,
+      "learning_rate": 0.00011830923248053394,
+      "loss": 2.7249,
+      "step": 1842
+    },
+    {
+      "epoch": 24.5752508361204,
+      "grad_norm": 0.5526009202003479,
+      "learning_rate": 0.00011826473859844272,
+      "loss": 2.6668,
+      "step": 1843
+    },
+    {
+      "epoch": 24.588628762541806,
+      "grad_norm": 0.6295923590660095,
+      "learning_rate": 0.00011822024471635152,
+      "loss": 2.8788,
+      "step": 1844
+    },
+    {
+      "epoch": 24.60200668896321,
+      "grad_norm": 0.6965957283973694,
+      "learning_rate": 0.00011817575083426028,
+      "loss": 2.8676,
+      "step": 1845
+    },
+    {
+      "epoch": 24.615384615384617,
+      "grad_norm": 0.6246476173400879,
+      "learning_rate": 0.0001181312569521691,
+      "loss": 2.8326,
+      "step": 1846
+    },
+    {
+      "epoch": 24.62876254180602,
+      "grad_norm": 0.6312415599822998,
+      "learning_rate": 0.00011808676307007786,
+      "loss": 2.5748,
+      "step": 1847
+    },
+    {
+      "epoch": 24.642140468227424,
+      "grad_norm": 0.7262862920761108,
+      "learning_rate": 0.00011804226918798666,
+      "loss": 2.922,
+      "step": 1848
+    },
+    {
+      "epoch": 24.65551839464883,
+      "grad_norm": 0.6640278697013855,
+      "learning_rate": 0.00011799777530589544,
+      "loss": 2.7494,
+      "step": 1849
+    },
+    {
+      "epoch": 24.668896321070235,
+      "grad_norm": 0.6154195070266724,
+      "learning_rate": 0.00011795328142380424,
+      "loss": 2.6168,
+      "step": 1850
+    },
+    {
+      "epoch": 24.682274247491637,
+      "grad_norm": 0.6020949482917786,
+      "learning_rate": 0.00011790878754171302,
+      "loss": 2.7566,
+      "step": 1851
+    },
+    {
+      "epoch": 24.695652173913043,
+      "grad_norm": 0.5775970816612244,
+      "learning_rate": 0.00011786429365962182,
+      "loss": 2.6862,
+      "step": 1852
+    },
+    {
+      "epoch": 24.70903010033445,
+      "grad_norm": 0.6687362194061279,
+      "learning_rate": 0.0001178197997775306,
+      "loss": 2.554,
+      "step": 1853
+    },
+    {
+      "epoch": 24.722408026755854,
+      "grad_norm": 0.6184269785881042,
+      "learning_rate": 0.0001177753058954394,
+      "loss": 2.9353,
+      "step": 1854
+    },
+    {
+      "epoch": 24.735785953177256,
+      "grad_norm": 0.6340799331665039,
+      "learning_rate": 0.00011773081201334816,
+      "loss": 2.857,
+      "step": 1855
+    },
+    {
+      "epoch": 24.74916387959866,
+      "grad_norm": 0.8216782808303833,
+      "learning_rate": 0.00011768631813125697,
+      "loss": 2.6295,
+      "step": 1856
+    },
+    {
+      "epoch": 24.762541806020067,
+      "grad_norm": 0.5764971375465393,
+      "learning_rate": 0.00011764182424916574,
+      "loss": 2.8431,
+      "step": 1857
+    },
+    {
+      "epoch": 24.775919732441473,
+      "grad_norm": 0.6233363747596741,
+      "learning_rate": 0.00011759733036707454,
+      "loss": 2.6867,
+      "step": 1858
+    },
+    {
+      "epoch": 24.789297658862875,
+      "grad_norm": 0.6669492721557617,
+      "learning_rate": 0.00011755283648498332,
+      "loss": 2.5405,
+      "step": 1859
+    },
+    {
+      "epoch": 24.80267558528428,
+      "grad_norm": 0.6156982183456421,
+      "learning_rate": 0.00011750834260289212,
+      "loss": 2.8071,
+      "step": 1860
+    },
+    {
+      "epoch": 24.816053511705686,
+      "grad_norm": 0.6210219860076904,
+      "learning_rate": 0.0001174638487208009,
+      "loss": 2.8308,
+      "step": 1861
+    },
+    {
+      "epoch": 24.82943143812709,
+      "grad_norm": 0.6786196231842041,
+      "learning_rate": 0.00011741935483870967,
+      "loss": 2.9039,
+      "step": 1862
+    },
+    {
+      "epoch": 24.842809364548494,
+      "grad_norm": 0.6467169523239136,
+      "learning_rate": 0.00011737486095661848,
+      "loss": 2.6601,
+      "step": 1863
+    },
+    {
+      "epoch": 24.8561872909699,
+      "grad_norm": 0.6734811663627625,
+      "learning_rate": 0.00011733036707452724,
+      "loss": 2.8614,
+      "step": 1864
+    },
+    {
+      "epoch": 24.869565217391305,
+      "grad_norm": 0.6223945021629333,
+      "learning_rate": 0.00011728587319243604,
+      "loss": 2.9397,
+      "step": 1865
+    },
+    {
+      "epoch": 24.88294314381271,
+      "grad_norm": 0.6042306423187256,
+      "learning_rate": 0.00011724137931034482,
+      "loss": 2.8984,
+      "step": 1866
+    },
+    {
+      "epoch": 24.896321070234112,
+      "grad_norm": 0.945598840713501,
+      "learning_rate": 0.00011719688542825362,
+      "loss": 2.6882,
+      "step": 1867
+    },
+    {
+      "epoch": 24.909698996655518,
+      "grad_norm": 0.6739409565925598,
+      "learning_rate": 0.0001171523915461624,
+      "loss": 2.6285,
+      "step": 1868
+    },
+    {
+      "epoch": 24.923076923076923,
+      "grad_norm": 0.6158167719841003,
+      "learning_rate": 0.0001171078976640712,
+      "loss": 2.8019,
+      "step": 1869
+    },
+    {
+      "epoch": 24.93645484949833,
+      "grad_norm": 0.5984783172607422,
+      "learning_rate": 0.00011706340378197998,
+      "loss": 2.785,
+      "step": 1870
+    },
+    {
+      "epoch": 24.94983277591973,
+      "grad_norm": 0.594236433506012,
+      "learning_rate": 0.00011701890989988878,
+      "loss": 3.0126,
+      "step": 1871
+    },
+    {
+      "epoch": 24.963210702341136,
+      "grad_norm": 0.6268919706344604,
+      "learning_rate": 0.00011697441601779754,
+      "loss": 2.7908,
+      "step": 1872
+    },
+    {
+      "epoch": 24.976588628762542,
+      "grad_norm": 0.6119821071624756,
+      "learning_rate": 0.00011692992213570635,
+      "loss": 2.5886,
+      "step": 1873
+    },
+    {
+      "epoch": 24.989966555183948,
+      "grad_norm": 0.8935804963111877,
+      "learning_rate": 0.00011688542825361512,
+      "loss": 2.9702,
+      "step": 1874
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 0.6583001613616943,
+      "learning_rate": 0.00011684093437152392,
+      "loss": 2.7691,
+      "step": 1875
+    },
+    {
+      "epoch": 25.013377926421406,
+      "grad_norm": 0.5952125787734985,
+      "learning_rate": 0.0001167964404894327,
+      "loss": 2.6709,
+      "step": 1876
+    },
+    {
+      "epoch": 25.02675585284281,
+      "grad_norm": 0.6740133762359619,
+      "learning_rate": 0.0001167519466073415,
+      "loss": 2.6548,
+      "step": 1877
+    },
+    {
+      "epoch": 25.040133779264213,
+      "grad_norm": 0.6665695309638977,
+      "learning_rate": 0.00011670745272525028,
+      "loss": 2.6141,
+      "step": 1878
+    },
+    {
+      "epoch": 25.05351170568562,
+      "grad_norm": 0.698957085609436,
+      "learning_rate": 0.00011666295884315908,
+      "loss": 2.7752,
+      "step": 1879
+    },
+    {
+      "epoch": 25.066889632107024,
+      "grad_norm": 0.7277541160583496,
+      "learning_rate": 0.00011661846496106786,
+      "loss": 2.8876,
+      "step": 1880
+    },
+    {
+      "epoch": 25.08026755852843,
+      "grad_norm": 0.5987577438354492,
+      "learning_rate": 0.00011657397107897665,
+      "loss": 2.5601,
+      "step": 1881
+    },
+    {
+      "epoch": 25.093645484949832,
+      "grad_norm": 0.5897870063781738,
+      "learning_rate": 0.00011652947719688542,
+      "loss": 2.5854,
+      "step": 1882
+    },
+    {
+      "epoch": 25.107023411371237,
+      "grad_norm": 0.573273241519928,
+      "learning_rate": 0.00011648498331479423,
+      "loss": 2.7141,
+      "step": 1883
+    },
+    {
+      "epoch": 25.120401337792643,
+      "grad_norm": 0.6173187494277954,
+      "learning_rate": 0.000116440489432703,
+      "loss": 2.6697,
+      "step": 1884
+    },
+    {
+      "epoch": 25.13377926421405,
+      "grad_norm": 0.5728760361671448,
+      "learning_rate": 0.0001163959955506118,
+      "loss": 2.7058,
+      "step": 1885
+    },
+    {
+      "epoch": 25.14715719063545,
+      "grad_norm": 0.543645441532135,
+      "learning_rate": 0.00011635150166852058,
+      "loss": 2.441,
+      "step": 1886
+    },
+    {
+      "epoch": 25.160535117056856,
+      "grad_norm": 0.6582958698272705,
+      "learning_rate": 0.00011630700778642938,
+      "loss": 2.8462,
+      "step": 1887
+    },
+    {
+      "epoch": 25.17391304347826,
+      "grad_norm": 0.6733880043029785,
+      "learning_rate": 0.00011626251390433816,
+      "loss": 2.7294,
+      "step": 1888
+    },
+    {
+      "epoch": 25.187290969899667,
+      "grad_norm": 0.8249802589416504,
+      "learning_rate": 0.00011621802002224695,
+      "loss": 2.7738,
+      "step": 1889
+    },
+    {
+      "epoch": 25.20066889632107,
+      "grad_norm": 0.6004601120948792,
+      "learning_rate": 0.00011617352614015574,
+      "loss": 2.6992,
+      "step": 1890
+    },
+    {
+      "epoch": 25.214046822742475,
+      "grad_norm": 0.6188962459564209,
+      "learning_rate": 0.00011612903225806453,
+      "loss": 2.5715,
+      "step": 1891
+    },
+    {
+      "epoch": 25.22742474916388,
+      "grad_norm": 0.6831389665603638,
+      "learning_rate": 0.0001160845383759733,
+      "loss": 2.8777,
+      "step": 1892
+    },
+    {
+      "epoch": 25.240802675585286,
+      "grad_norm": 0.5578925013542175,
+      "learning_rate": 0.00011604004449388211,
+      "loss": 2.4875,
+      "step": 1893
+    },
+    {
+      "epoch": 25.254180602006688,
+      "grad_norm": 0.5995929837226868,
+      "learning_rate": 0.00011599555061179088,
+      "loss": 2.5166,
+      "step": 1894
+    },
+    {
+      "epoch": 25.267558528428093,
+      "grad_norm": 0.6817669868469238,
+      "learning_rate": 0.00011595105672969967,
+      "loss": 2.7628,
+      "step": 1895
+    },
+    {
+      "epoch": 25.2809364548495,
+      "grad_norm": 0.671890139579773,
+      "learning_rate": 0.00011590656284760846,
+      "loss": 2.5406,
+      "step": 1896
+    },
+    {
+      "epoch": 25.294314381270905,
+      "grad_norm": 0.6167321801185608,
+      "learning_rate": 0.00011586206896551725,
+      "loss": 2.8789,
+      "step": 1897
+    },
+    {
+      "epoch": 25.307692307692307,
+      "grad_norm": 0.6145638227462769,
+      "learning_rate": 0.00011581757508342604,
+      "loss": 2.7629,
+      "step": 1898
+    },
+    {
+      "epoch": 25.321070234113712,
+      "grad_norm": 0.6117168068885803,
+      "learning_rate": 0.00011577308120133483,
+      "loss": 2.7048,
+      "step": 1899
+    },
+    {
+      "epoch": 25.334448160535118,
+      "grad_norm": 0.6903496384620667,
+      "learning_rate": 0.00011572858731924361,
+      "loss": 2.9096,
+      "step": 1900
+    },
+    {
+      "epoch": 25.347826086956523,
+      "grad_norm": 0.5942792296409607,
+      "learning_rate": 0.00011568409343715241,
+      "loss": 2.4525,
+      "step": 1901
+    },
+    {
+      "epoch": 25.361204013377925,
+      "grad_norm": 0.5714770555496216,
+      "learning_rate": 0.00011563959955506118,
+      "loss": 2.6605,
+      "step": 1902
+    },
+    {
+      "epoch": 25.37458193979933,
+      "grad_norm": 1.0252799987792969,
+      "learning_rate": 0.00011559510567296996,
+      "loss": 2.4599,
+      "step": 1903
+    },
+    {
+      "epoch": 25.387959866220736,
+      "grad_norm": 0.6459619998931885,
+      "learning_rate": 0.00011555061179087876,
+      "loss": 2.657,
+      "step": 1904
+    },
+    {
+      "epoch": 25.401337792642142,
+      "grad_norm": 0.6513925790786743,
+      "learning_rate": 0.00011550611790878754,
+      "loss": 2.6491,
+      "step": 1905
+    },
+    {
+      "epoch": 25.414715719063544,
+      "grad_norm": 0.838390588760376,
+      "learning_rate": 0.00011546162402669634,
+      "loss": 2.8909,
+      "step": 1906
+    },
+    {
+      "epoch": 25.42809364548495,
+      "grad_norm": 0.6873288750648499,
+      "learning_rate": 0.00011541713014460512,
+      "loss": 2.9252,
+      "step": 1907
+    },
+    {
+      "epoch": 25.441471571906355,
+      "grad_norm": 0.6325615644454956,
+      "learning_rate": 0.00011537263626251391,
+      "loss": 2.8784,
+      "step": 1908
+    },
+    {
+      "epoch": 25.45484949832776,
+      "grad_norm": 0.7683016061782837,
+      "learning_rate": 0.00011532814238042268,
+      "loss": 2.7951,
+      "step": 1909
+    },
+    {
+      "epoch": 25.468227424749163,
+      "grad_norm": 0.6581037044525146,
+      "learning_rate": 0.00011528364849833149,
+      "loss": 2.7674,
+      "step": 1910
+    },
+    {
+      "epoch": 25.48160535117057,
+      "grad_norm": 0.6761939525604248,
+      "learning_rate": 0.00011523915461624026,
+      "loss": 2.7315,
+      "step": 1911
+    },
+    {
+      "epoch": 25.494983277591974,
+      "grad_norm": 0.621744692325592,
+      "learning_rate": 0.00011519466073414906,
+      "loss": 2.6977,
+      "step": 1912
+    },
+    {
+      "epoch": 25.50836120401338,
+      "grad_norm": 0.5744712352752686,
+      "learning_rate": 0.00011515016685205784,
+      "loss": 2.6568,
+      "step": 1913
+    },
+    {
+      "epoch": 25.52173913043478,
+      "grad_norm": 0.6587238907814026,
+      "learning_rate": 0.00011510567296996664,
+      "loss": 3.0222,
+      "step": 1914
+    },
+    {
+      "epoch": 25.535117056856187,
+      "grad_norm": 0.6338929533958435,
+      "learning_rate": 0.00011506117908787542,
+      "loss": 2.4611,
+      "step": 1915
+    },
+    {
+      "epoch": 25.548494983277592,
+      "grad_norm": 0.6468850374221802,
+      "learning_rate": 0.00011501668520578421,
+      "loss": 2.5532,
+      "step": 1916
+    },
+    {
+      "epoch": 25.561872909698998,
+      "grad_norm": 0.6900503635406494,
+      "learning_rate": 0.000114972191323693,
+      "loss": 2.7291,
+      "step": 1917
+    },
+    {
+      "epoch": 25.5752508361204,
+      "grad_norm": 0.9081000089645386,
+      "learning_rate": 0.00011492769744160179,
+      "loss": 2.528,
+      "step": 1918
+    },
+    {
+      "epoch": 25.588628762541806,
+      "grad_norm": 0.6108705401420593,
+      "learning_rate": 0.00011488320355951056,
+      "loss": 2.7942,
+      "step": 1919
+    },
+    {
+      "epoch": 25.60200668896321,
+      "grad_norm": 0.5745365023612976,
+      "learning_rate": 0.00011483870967741937,
+      "loss": 2.7273,
+      "step": 1920
+    },
+    {
+      "epoch": 25.615384615384617,
+      "grad_norm": 0.7215850949287415,
+      "learning_rate": 0.00011479421579532814,
+      "loss": 2.85,
+      "step": 1921
+    },
+    {
+      "epoch": 25.62876254180602,
+      "grad_norm": 0.6535587310791016,
+      "learning_rate": 0.00011474972191323693,
+      "loss": 2.7488,
+      "step": 1922
+    },
+    {
+      "epoch": 25.642140468227424,
+      "grad_norm": 0.6285836696624756,
+      "learning_rate": 0.00011470522803114572,
+      "loss": 2.6874,
+      "step": 1923
+    },
+    {
+      "epoch": 25.65551839464883,
+      "grad_norm": 0.652277410030365,
+      "learning_rate": 0.00011466073414905451,
+      "loss": 2.889,
+      "step": 1924
+    },
+    {
+      "epoch": 25.668896321070235,
+      "grad_norm": 0.6526362299919128,
+      "learning_rate": 0.0001146162402669633,
+      "loss": 2.8342,
+      "step": 1925
+    },
+    {
+      "epoch": 25.682274247491637,
+      "grad_norm": 0.6387808918952942,
+      "learning_rate": 0.00011457174638487209,
+      "loss": 2.6713,
+      "step": 1926
+    },
+    {
+      "epoch": 25.695652173913043,
+      "grad_norm": 0.5775367617607117,
+      "learning_rate": 0.00011452725250278087,
+      "loss": 2.6848,
+      "step": 1927
+    },
+    {
+      "epoch": 25.70903010033445,
+      "grad_norm": 0.7041560411453247,
+      "learning_rate": 0.00011448275862068967,
+      "loss": 2.7532,
+      "step": 1928
+    },
+    {
+      "epoch": 25.722408026755854,
+      "grad_norm": 0.6833249926567078,
+      "learning_rate": 0.00011443826473859844,
+      "loss": 2.8257,
+      "step": 1929
+    },
+    {
+      "epoch": 25.735785953177256,
+      "grad_norm": 0.6916714310646057,
+      "learning_rate": 0.00011439377085650725,
+      "loss": 2.5491,
+      "step": 1930
+    },
+    {
+      "epoch": 25.74916387959866,
+      "grad_norm": 0.8915151357650757,
+      "learning_rate": 0.00011434927697441602,
+      "loss": 2.8361,
+      "step": 1931
+    },
+    {
+      "epoch": 25.762541806020067,
+      "grad_norm": 0.6278401613235474,
+      "learning_rate": 0.00011430478309232481,
+      "loss": 2.8022,
+      "step": 1932
+    },
+    {
+      "epoch": 25.775919732441473,
+      "grad_norm": 0.6929000020027161,
+      "learning_rate": 0.0001142602892102336,
+      "loss": 2.8938,
+      "step": 1933
+    },
+    {
+      "epoch": 25.789297658862875,
+      "grad_norm": 0.6363662481307983,
+      "learning_rate": 0.00011421579532814239,
+      "loss": 2.8736,
+      "step": 1934
+    },
+    {
+      "epoch": 25.80267558528428,
+      "grad_norm": 0.6596313118934631,
+      "learning_rate": 0.00011417130144605117,
+      "loss": 2.9864,
+      "step": 1935
+    },
+    {
+      "epoch": 25.816053511705686,
+      "grad_norm": 0.9694854617118835,
+      "learning_rate": 0.00011412680756395997,
+      "loss": 2.7362,
+      "step": 1936
+    },
+    {
+      "epoch": 25.82943143812709,
+      "grad_norm": 0.721339225769043,
+      "learning_rate": 0.00011408231368186875,
+      "loss": 2.9919,
+      "step": 1937
+    },
+    {
+      "epoch": 25.842809364548494,
+      "grad_norm": 0.6164969205856323,
+      "learning_rate": 0.00011403781979977755,
+      "loss": 3.0446,
+      "step": 1938
+    },
+    {
+      "epoch": 25.8561872909699,
+      "grad_norm": 0.8240790367126465,
+      "learning_rate": 0.00011399332591768632,
+      "loss": 2.7113,
+      "step": 1939
+    },
+    {
+      "epoch": 25.869565217391305,
+      "grad_norm": 0.865219235420227,
+      "learning_rate": 0.00011394883203559513,
+      "loss": 2.6692,
+      "step": 1940
+    },
+    {
+      "epoch": 25.88294314381271,
+      "grad_norm": 0.6014530062675476,
+      "learning_rate": 0.0001139043381535039,
+      "loss": 2.7527,
+      "step": 1941
+    },
+    {
+      "epoch": 25.896321070234112,
+      "grad_norm": 0.6110913157463074,
+      "learning_rate": 0.00011385984427141269,
+      "loss": 2.6213,
+      "step": 1942
+    },
+    {
+      "epoch": 25.909698996655518,
+      "grad_norm": 0.6967377662658691,
+      "learning_rate": 0.00011381535038932147,
+      "loss": 3.0013,
+      "step": 1943
+    },
+    {
+      "epoch": 25.923076923076923,
+      "grad_norm": 0.6235397458076477,
+      "learning_rate": 0.00011377085650723027,
+      "loss": 2.8256,
+      "step": 1944
+    },
+    {
+      "epoch": 25.93645484949833,
+      "grad_norm": 0.6463499665260315,
+      "learning_rate": 0.00011372636262513905,
+      "loss": 2.6573,
+      "step": 1945
+    },
+    {
+      "epoch": 25.94983277591973,
+      "grad_norm": 0.7898368239402771,
+      "learning_rate": 0.00011368186874304782,
+      "loss": 2.6484,
+      "step": 1946
+    },
+    {
+      "epoch": 25.963210702341136,
+      "grad_norm": 0.6130072474479675,
+      "learning_rate": 0.00011363737486095663,
+      "loss": 2.6912,
+      "step": 1947
+    },
+    {
+      "epoch": 25.976588628762542,
+      "grad_norm": 0.6541060209274292,
+      "learning_rate": 0.0001135928809788654,
+      "loss": 2.789,
+      "step": 1948
+    },
+    {
+      "epoch": 25.989966555183948,
+      "grad_norm": 0.6722179651260376,
+      "learning_rate": 0.0001135483870967742,
+      "loss": 2.7197,
+      "step": 1949
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 0.6884087920188904,
+      "learning_rate": 0.00011350389321468298,
+      "loss": 2.83,
+      "step": 1950
+    },
+    {
+      "epoch": 26.013377926421406,
+      "grad_norm": 0.5991981029510498,
+      "learning_rate": 0.00011345939933259177,
+      "loss": 2.8456,
+      "step": 1951
+    },
+    {
+      "epoch": 26.02675585284281,
+      "grad_norm": 0.660169243812561,
+      "learning_rate": 0.00011341490545050056,
+      "loss": 2.8782,
+      "step": 1952
+    },
+    {
+      "epoch": 26.040133779264213,
+      "grad_norm": 0.651779294013977,
+      "learning_rate": 0.00011337041156840935,
+      "loss": 2.7286,
+      "step": 1953
+    },
+    {
+      "epoch": 26.05351170568562,
+      "grad_norm": 0.6962876915931702,
+      "learning_rate": 0.00011332591768631813,
+      "loss": 2.5009,
+      "step": 1954
+    },
+    {
+      "epoch": 26.066889632107024,
+      "grad_norm": 0.6752526164054871,
+      "learning_rate": 0.00011328142380422693,
+      "loss": 2.6137,
+      "step": 1955
+    },
+    {
+      "epoch": 26.08026755852843,
+      "grad_norm": 0.6335856914520264,
+      "learning_rate": 0.0001132369299221357,
+      "loss": 2.4744,
+      "step": 1956
+    },
+    {
+      "epoch": 26.093645484949832,
+      "grad_norm": 0.6627519726753235,
+      "learning_rate": 0.00011319243604004451,
+      "loss": 2.9633,
+      "step": 1957
+    },
+    {
+      "epoch": 26.107023411371237,
+      "grad_norm": 0.5948102474212646,
+      "learning_rate": 0.00011314794215795328,
+      "loss": 2.6333,
+      "step": 1958
+    },
+    {
+      "epoch": 26.120401337792643,
+      "grad_norm": 0.5609325766563416,
+      "learning_rate": 0.00011310344827586207,
+      "loss": 2.5427,
+      "step": 1959
+    },
+    {
+      "epoch": 26.13377926421405,
+      "grad_norm": 0.6280467510223389,
+      "learning_rate": 0.00011305895439377086,
+      "loss": 2.6838,
+      "step": 1960
+    },
+    {
+      "epoch": 26.14715719063545,
+      "grad_norm": 0.5900541543960571,
+      "learning_rate": 0.00011301446051167965,
+      "loss": 2.8057,
+      "step": 1961
+    },
+    {
+      "epoch": 26.160535117056856,
+      "grad_norm": 0.6185488700866699,
+      "learning_rate": 0.00011296996662958843,
+      "loss": 2.7778,
+      "step": 1962
+    },
+    {
+      "epoch": 26.17391304347826,
+      "grad_norm": 0.6149762272834778,
+      "learning_rate": 0.00011292547274749723,
+      "loss": 2.7384,
+      "step": 1963
+    },
+    {
+      "epoch": 26.187290969899667,
+      "grad_norm": 0.5975783467292786,
+      "learning_rate": 0.00011288097886540601,
+      "loss": 2.6957,
+      "step": 1964
+    },
+    {
+      "epoch": 26.20066889632107,
+      "grad_norm": 0.7946596741676331,
+      "learning_rate": 0.00011283648498331481,
+      "loss": 2.6071,
+      "step": 1965
+    },
+    {
+      "epoch": 26.214046822742475,
+      "grad_norm": 0.5987370014190674,
+      "learning_rate": 0.00011279199110122358,
+      "loss": 2.5812,
+      "step": 1966
+    },
+    {
+      "epoch": 26.22742474916388,
+      "grad_norm": 0.7613934278488159,
+      "learning_rate": 0.00011274749721913239,
+      "loss": 2.7886,
+      "step": 1967
+    },
+    {
+      "epoch": 26.240802675585286,
+      "grad_norm": 0.6035755276679993,
+      "learning_rate": 0.00011270300333704116,
+      "loss": 2.4709,
+      "step": 1968
+    },
+    {
+      "epoch": 26.254180602006688,
+      "grad_norm": 0.9545117616653442,
+      "learning_rate": 0.00011265850945494995,
+      "loss": 2.6039,
+      "step": 1969
+    },
+    {
+      "epoch": 26.267558528428093,
+      "grad_norm": 0.5982191562652588,
+      "learning_rate": 0.00011261401557285873,
+      "loss": 2.664,
+      "step": 1970
+    },
+    {
+      "epoch": 26.2809364548495,
+      "grad_norm": 0.6070300340652466,
+      "learning_rate": 0.00011256952169076753,
+      "loss": 2.671,
+      "step": 1971
+    },
+    {
+      "epoch": 26.294314381270905,
+      "grad_norm": 0.6093387603759766,
+      "learning_rate": 0.00011252502780867631,
+      "loss": 2.6441,
+      "step": 1972
+    },
+    {
+      "epoch": 26.307692307692307,
+      "grad_norm": 0.6499563455581665,
+      "learning_rate": 0.00011248053392658511,
+      "loss": 2.6851,
+      "step": 1973
+    },
+    {
+      "epoch": 26.321070234113712,
+      "grad_norm": 0.6848227977752686,
+      "learning_rate": 0.00011243604004449389,
+      "loss": 2.6448,
+      "step": 1974
+    },
+    {
+      "epoch": 26.334448160535118,
+      "grad_norm": 0.6483498811721802,
+      "learning_rate": 0.00011239154616240269,
+      "loss": 2.5361,
+      "step": 1975
+    },
+    {
+      "epoch": 26.347826086956523,
+      "grad_norm": 0.6412089467048645,
+      "learning_rate": 0.00011234705228031146,
+      "loss": 2.6625,
+      "step": 1976
+    },
+    {
+      "epoch": 26.361204013377925,
+      "grad_norm": 0.6891600489616394,
+      "learning_rate": 0.00011230255839822026,
+      "loss": 2.6901,
+      "step": 1977
+    },
+    {
+      "epoch": 26.37458193979933,
+      "grad_norm": 0.606422483921051,
+      "learning_rate": 0.00011225806451612903,
+      "loss": 2.7778,
+      "step": 1978
+    },
+    {
+      "epoch": 26.387959866220736,
+      "grad_norm": 0.6754788756370544,
+      "learning_rate": 0.00011221357063403783,
+      "loss": 2.7757,
+      "step": 1979
+    },
+    {
+      "epoch": 26.401337792642142,
+      "grad_norm": 0.6534073352813721,
+      "learning_rate": 0.00011216907675194661,
+      "loss": 2.7682,
+      "step": 1980
+    },
+    {
+      "epoch": 26.414715719063544,
+      "grad_norm": 0.6333116888999939,
+      "learning_rate": 0.00011212458286985541,
+      "loss": 2.739,
+      "step": 1981
+    },
+    {
+      "epoch": 26.42809364548495,
+      "grad_norm": 0.6784356236457825,
+      "learning_rate": 0.00011208008898776419,
+      "loss": 2.499,
+      "step": 1982
+    },
+    {
+      "epoch": 26.441471571906355,
+      "grad_norm": 0.6247740983963013,
+      "learning_rate": 0.00011203559510567299,
+      "loss": 2.8451,
+      "step": 1983
+    },
+    {
+      "epoch": 26.45484949832776,
+      "grad_norm": 0.6405763030052185,
+      "learning_rate": 0.00011199110122358177,
+      "loss": 2.7237,
+      "step": 1984
+    },
+    {
+      "epoch": 26.468227424749163,
+      "grad_norm": 0.7398512959480286,
+      "learning_rate": 0.00011194660734149056,
+      "loss": 2.6931,
+      "step": 1985
+    },
+    {
+      "epoch": 26.48160535117057,
+      "grad_norm": 0.6067454814910889,
+      "learning_rate": 0.00011190211345939933,
+      "loss": 2.8153,
+      "step": 1986
+    },
+    {
+      "epoch": 26.494983277591974,
+      "grad_norm": 0.5860223770141602,
+      "learning_rate": 0.00011185761957730812,
+      "loss": 2.9522,
+      "step": 1987
+    },
+    {
+      "epoch": 26.50836120401338,
+      "grad_norm": 0.6325815320014954,
+      "learning_rate": 0.00011181312569521691,
+      "loss": 2.7294,
+      "step": 1988
+    },
+    {
+      "epoch": 26.52173913043478,
+      "grad_norm": 0.6049959063529968,
+      "learning_rate": 0.0001117686318131257,
+      "loss": 2.7989,
+      "step": 1989
+    },
+    {
+      "epoch": 26.535117056856187,
+      "grad_norm": 0.5589177012443542,
+      "learning_rate": 0.00011172413793103449,
+      "loss": 2.4431,
+      "step": 1990
+    },
+    {
+      "epoch": 26.548494983277592,
+      "grad_norm": 0.6467594504356384,
+      "learning_rate": 0.00011167964404894327,
+      "loss": 2.7166,
+      "step": 1991
+    },
+    {
+      "epoch": 26.561872909698998,
+      "grad_norm": 0.7082380056381226,
+      "learning_rate": 0.00011163515016685207,
+      "loss": 2.7768,
+      "step": 1992
+    },
+    {
+      "epoch": 26.5752508361204,
+      "grad_norm": 0.7438964247703552,
+      "learning_rate": 0.00011159065628476084,
+      "loss": 2.8655,
+      "step": 1993
+    },
+    {
+      "epoch": 26.588628762541806,
+      "grad_norm": 0.7755714654922485,
+      "learning_rate": 0.00011154616240266965,
+      "loss": 2.6956,
+      "step": 1994
+    },
+    {
+      "epoch": 26.60200668896321,
+      "grad_norm": 0.576988160610199,
+      "learning_rate": 0.00011150166852057842,
+      "loss": 2.5831,
+      "step": 1995
+    },
+    {
+      "epoch": 26.615384615384617,
+      "grad_norm": 0.6044856309890747,
+      "learning_rate": 0.00011145717463848721,
+      "loss": 2.6834,
+      "step": 1996
+    },
+    {
+      "epoch": 26.62876254180602,
+      "grad_norm": 0.5846719741821289,
+      "learning_rate": 0.000111412680756396,
+      "loss": 2.5553,
+      "step": 1997
+    },
+    {
+      "epoch": 26.642140468227424,
+      "grad_norm": 0.6511263251304626,
+      "learning_rate": 0.00011136818687430479,
+      "loss": 2.5404,
+      "step": 1998
+    },
+    {
+      "epoch": 26.65551839464883,
+      "grad_norm": 0.6784057021141052,
+      "learning_rate": 0.00011132369299221357,
+      "loss": 2.5896,
+      "step": 1999
+    },
+    {
+      "epoch": 26.668896321070235,
+      "grad_norm": 0.7156029939651489,
+      "learning_rate": 0.00011127919911012237,
+      "loss": 2.6683,
+      "step": 2000
+    },
+    {
+      "epoch": 26.682274247491637,
+      "grad_norm": 1.5313527584075928,
+      "learning_rate": 0.00011123470522803115,
+      "loss": 3.1002,
+      "step": 2001
+    },
+    {
+      "epoch": 26.695652173913043,
+      "grad_norm": 1.570999264717102,
+      "learning_rate": 0.00011119021134593995,
+      "loss": 3.0022,
+      "step": 2002
+    },
+    {
+      "epoch": 26.70903010033445,
+      "grad_norm": 1.672279953956604,
+      "learning_rate": 0.00011114571746384872,
+      "loss": 2.9443,
+      "step": 2003
+    },
+    {
+      "epoch": 26.722408026755854,
+      "grad_norm": 1.3200056552886963,
+      "learning_rate": 0.00011110122358175752,
+      "loss": 3.3101,
+      "step": 2004
+    },
+    {
+      "epoch": 26.735785953177256,
+      "grad_norm": 1.9621714353561401,
+      "learning_rate": 0.0001110567296996663,
+      "loss": 3.0294,
+      "step": 2005
+    },
+    {
+      "epoch": 26.74916387959866,
+      "grad_norm": 1.5567591190338135,
+      "learning_rate": 0.00011101223581757509,
+      "loss": 3.1361,
+      "step": 2006
+    },
+    {
+      "epoch": 26.762541806020067,
+      "grad_norm": 1.6681092977523804,
+      "learning_rate": 0.00011096774193548387,
+      "loss": 3.2709,
+      "step": 2007
+    },
+    {
+      "epoch": 26.775919732441473,
+      "grad_norm": 1.461212158203125,
+      "learning_rate": 0.00011092324805339267,
+      "loss": 3.0585,
+      "step": 2008
+    },
+    {
+      "epoch": 26.789297658862875,
+      "grad_norm": 1.4997388124465942,
+      "learning_rate": 0.00011087875417130145,
+      "loss": 3.0182,
+      "step": 2009
+    },
+    {
+      "epoch": 26.80267558528428,
+      "grad_norm": 1.3930811882019043,
+      "learning_rate": 0.00011083426028921025,
+      "loss": 2.9796,
+      "step": 2010
+    },
+    {
+      "epoch": 26.816053511705686,
+      "grad_norm": 1.3426967859268188,
+      "learning_rate": 0.00011078976640711903,
+      "loss": 3.2029,
+      "step": 2011
+    },
+    {
+      "epoch": 26.82943143812709,
+      "grad_norm": 1.7054353952407837,
+      "learning_rate": 0.00011074527252502782,
+      "loss": 3.0889,
+      "step": 2012
+    },
+    {
+      "epoch": 26.842809364548494,
+      "grad_norm": 1.6260415315628052,
+      "learning_rate": 0.00011070077864293659,
+      "loss": 3.2461,
+      "step": 2013
+    },
+    {
+      "epoch": 26.8561872909699,
+      "grad_norm": 1.7708278894424438,
+      "learning_rate": 0.0001106562847608454,
+      "loss": 3.1737,
+      "step": 2014
+    },
+    {
+      "epoch": 26.869565217391305,
+      "grad_norm": 1.4107633829116821,
+      "learning_rate": 0.00011061179087875417,
+      "loss": 3.0006,
+      "step": 2015
+    },
+    {
+      "epoch": 26.88294314381271,
+      "grad_norm": 1.929655909538269,
+      "learning_rate": 0.00011056729699666297,
+      "loss": 3.0805,
+      "step": 2016
+    },
+    {
+      "epoch": 26.896321070234112,
+      "grad_norm": 1.835286021232605,
+      "learning_rate": 0.00011052280311457175,
+      "loss": 3.0995,
+      "step": 2017
+    },
+    {
+      "epoch": 26.909698996655518,
+      "grad_norm": 1.9293112754821777,
+      "learning_rate": 0.00011047830923248055,
+      "loss": 3.1714,
+      "step": 2018
+    },
+    {
+      "epoch": 26.923076923076923,
+      "grad_norm": 1.5640552043914795,
+      "learning_rate": 0.00011043381535038933,
+      "loss": 3.1381,
+      "step": 2019
+    },
+    {
+      "epoch": 26.93645484949833,
+      "grad_norm": 1.7890485525131226,
+      "learning_rate": 0.00011038932146829812,
+      "loss": 3.0785,
+      "step": 2020
+    },
+    {
+      "epoch": 26.94983277591973,
+      "grad_norm": 1.3167777061462402,
+      "learning_rate": 0.0001103448275862069,
+      "loss": 3.2041,
+      "step": 2021
+    },
+    {
+      "epoch": 26.963210702341136,
+      "grad_norm": 1.3006356954574585,
+      "learning_rate": 0.0001103003337041157,
+      "loss": 2.6744,
+      "step": 2022
+    },
+    {
+      "epoch": 26.976588628762542,
+      "grad_norm": 1.3111026287078857,
+      "learning_rate": 0.00011025583982202447,
+      "loss": 2.9279,
+      "step": 2023
+    },
+    {
+      "epoch": 26.989966555183948,
+      "grad_norm": 1.9163484573364258,
+      "learning_rate": 0.00011021134593993328,
+      "loss": 3.0983,
+      "step": 2024
+    },
+    {
+      "epoch": 27.013377926421406,
+      "grad_norm": 2.121852159500122,
+      "learning_rate": 0.00011016685205784205,
+      "loss": 6.1272,
+      "step": 2025
+    },
+    {
+      "epoch": 27.02675585284281,
+      "grad_norm": 1.2503304481506348,
+      "learning_rate": 0.00011012235817575085,
+      "loss": 2.8418,
+      "step": 2026
+    },
+    {
+      "epoch": 27.040133779264213,
+      "grad_norm": 1.2677263021469116,
+      "learning_rate": 0.00011007786429365963,
+      "loss": 3.166,
+      "step": 2027
+    },
+    {
+      "epoch": 27.05351170568562,
+      "grad_norm": 1.1323914527893066,
+      "learning_rate": 0.00011003337041156841,
+      "loss": 2.881,
+      "step": 2028
+    },
+    {
+      "epoch": 27.066889632107024,
+      "grad_norm": 1.32306969165802,
+      "learning_rate": 0.0001099888765294772,
+      "loss": 3.1736,
+      "step": 2029
+    },
+    {
+      "epoch": 27.08026755852843,
+      "grad_norm": 1.1854408979415894,
+      "learning_rate": 0.00010994438264738598,
+      "loss": 3.0644,
+      "step": 2030
+    },
+    {
+      "epoch": 27.093645484949832,
+      "grad_norm": 1.3617606163024902,
+      "learning_rate": 0.00010989988876529478,
+      "loss": 3.0845,
+      "step": 2031
+    },
+    {
+      "epoch": 27.107023411371237,
+      "grad_norm": 1.3771755695343018,
+      "learning_rate": 0.00010985539488320355,
+      "loss": 3.215,
+      "step": 2032
+    },
+    {
+      "epoch": 27.120401337792643,
+      "grad_norm": 1.4441111087799072,
+      "learning_rate": 0.00010981090100111235,
+      "loss": 3.1639,
+      "step": 2033
+    },
+    {
+      "epoch": 27.13377926421405,
+      "grad_norm": 1.4004138708114624,
+      "learning_rate": 0.00010976640711902113,
+      "loss": 3.2066,
+      "step": 2034
+    },
+    {
+      "epoch": 27.14715719063545,
+      "grad_norm": 1.3546315431594849,
+      "learning_rate": 0.00010972191323692993,
+      "loss": 2.8154,
+      "step": 2035
+    },
+    {
+      "epoch": 27.160535117056856,
+      "grad_norm": 1.122201681137085,
+      "learning_rate": 0.00010967741935483871,
+      "loss": 2.8852,
+      "step": 2036
+    },
+    {
+      "epoch": 27.17391304347826,
+      "grad_norm": 1.4592616558074951,
+      "learning_rate": 0.0001096329254727475,
+      "loss": 2.9766,
+      "step": 2037
+    },
+    {
+      "epoch": 27.187290969899667,
+      "grad_norm": 1.1791727542877197,
+      "learning_rate": 0.00010958843159065629,
+      "loss": 2.8115,
+      "step": 2038
+    },
+    {
+      "epoch": 27.20066889632107,
+      "grad_norm": 1.3347991704940796,
+      "learning_rate": 0.00010954393770856508,
+      "loss": 2.839,
+      "step": 2039
+    },
+    {
+      "epoch": 27.214046822742475,
+      "grad_norm": 1.6515034437179565,
+      "learning_rate": 0.00010949944382647385,
+      "loss": 3.2088,
+      "step": 2040
+    },
+    {
+      "epoch": 27.22742474916388,
+      "grad_norm": 1.4179280996322632,
+      "learning_rate": 0.00010945494994438266,
+      "loss": 3.0364,
+      "step": 2041
+    },
+    {
+      "epoch": 27.240802675585286,
+      "grad_norm": 1.5073821544647217,
+      "learning_rate": 0.00010941045606229143,
+      "loss": 2.9634,
+      "step": 2042
+    },
+    {
+      "epoch": 27.254180602006688,
+      "grad_norm": 1.3612418174743652,
+      "learning_rate": 0.00010936596218020023,
+      "loss": 3.0204,
+      "step": 2043
+    },
+    {
+      "epoch": 27.267558528428093,
+      "grad_norm": 1.3515981435775757,
+      "learning_rate": 0.00010932146829810901,
+      "loss": 2.7506,
+      "step": 2044
+    },
+    {
+      "epoch": 27.2809364548495,
+      "grad_norm": 1.394405722618103,
+      "learning_rate": 0.0001092769744160178,
+      "loss": 2.8979,
+      "step": 2045
+    },
+    {
+      "epoch": 27.294314381270905,
+      "grad_norm": 1.4995383024215698,
+      "learning_rate": 0.00010923248053392659,
+      "loss": 3.2501,
+      "step": 2046
+    },
+    {
+      "epoch": 27.307692307692307,
+      "grad_norm": 1.4658801555633545,
+      "learning_rate": 0.00010918798665183538,
+      "loss": 3.1901,
+      "step": 2047
+    },
+    {
+      "epoch": 27.321070234113712,
+      "grad_norm": 1.4346429109573364,
+      "learning_rate": 0.00010914349276974417,
+      "loss": 2.9835,
+      "step": 2048
+    },
+    {
+      "epoch": 27.334448160535118,
+      "grad_norm": 1.455994963645935,
+      "learning_rate": 0.00010909899888765296,
+      "loss": 3.0354,
+      "step": 2049
+    },
+    {
+      "epoch": 27.347826086956523,
+      "grad_norm": 1.2409014701843262,
+      "learning_rate": 0.00010905450500556173,
+      "loss": 3.1903,
+      "step": 2050
+    },
+    {
+      "epoch": 27.361204013377925,
+      "grad_norm": 1.417441487312317,
+      "learning_rate": 0.00010901001112347054,
+      "loss": 2.9081,
+      "step": 2051
+    },
+    {
+      "epoch": 27.37458193979933,
+      "grad_norm": 1.4116290807724,
+      "learning_rate": 0.00010896551724137931,
+      "loss": 2.8466,
+      "step": 2052
+    },
+    {
+      "epoch": 27.387959866220736,
+      "grad_norm": 1.58635675907135,
+      "learning_rate": 0.0001089210233592881,
+      "loss": 3.134,
+      "step": 2053
+    },
+    {
+      "epoch": 27.401337792642142,
+      "grad_norm": 1.4947638511657715,
+      "learning_rate": 0.00010887652947719689,
+      "loss": 2.9585,
+      "step": 2054
+    },
+    {
+      "epoch": 27.414715719063544,
+      "grad_norm": 1.5971535444259644,
+      "learning_rate": 0.00010883203559510568,
+      "loss": 2.954,
+      "step": 2055
+    },
+    {
+      "epoch": 27.42809364548495,
+      "grad_norm": 1.5000510215759277,
+      "learning_rate": 0.00010878754171301447,
+      "loss": 3.1699,
+      "step": 2056
+    },
+    {
+      "epoch": 27.441471571906355,
+      "grad_norm": 1.3633960485458374,
+      "learning_rate": 0.00010874304783092326,
+      "loss": 3.0109,
+      "step": 2057
+    },
+    {
+      "epoch": 27.45484949832776,
+      "grad_norm": 1.2857588529586792,
+      "learning_rate": 0.00010869855394883204,
+      "loss": 3.1932,
+      "step": 2058
+    },
+    {
+      "epoch": 27.468227424749163,
+      "grad_norm": 1.6654231548309326,
+      "learning_rate": 0.00010865406006674084,
+      "loss": 3.2411,
+      "step": 2059
+    },
+    {
+      "epoch": 27.48160535117057,
+      "grad_norm": 1.398990511894226,
+      "learning_rate": 0.00010860956618464961,
+      "loss": 3.2183,
+      "step": 2060
+    },
+    {
+      "epoch": 27.494983277591974,
+      "grad_norm": 1.4811097383499146,
+      "learning_rate": 0.00010856507230255842,
+      "loss": 2.9565,
+      "step": 2061
+    },
+    {
+      "epoch": 27.50836120401338,
+      "grad_norm": 1.5817798376083374,
+      "learning_rate": 0.00010852057842046719,
+      "loss": 3.0652,
+      "step": 2062
+    },
+    {
+      "epoch": 27.52173913043478,
+      "grad_norm": 1.2948859930038452,
+      "learning_rate": 0.00010847608453837598,
+      "loss": 2.8978,
+      "step": 2063
+    },
+    {
+      "epoch": 27.535117056856187,
+      "grad_norm": 1.3314701318740845,
+      "learning_rate": 0.00010843159065628477,
+      "loss": 2.8764,
+      "step": 2064
+    },
+    {
+      "epoch": 27.548494983277592,
+      "grad_norm": 1.531958818435669,
+      "learning_rate": 0.00010838709677419356,
+      "loss": 3.3317,
+      "step": 2065
+    },
+    {
+      "epoch": 27.561872909698998,
+      "grad_norm": 1.5060667991638184,
+      "learning_rate": 0.00010834260289210234,
+      "loss": 3.0724,
+      "step": 2066
+    },
+    {
+      "epoch": 27.5752508361204,
+      "grad_norm": 1.4489496946334839,
+      "learning_rate": 0.00010829810901001114,
+      "loss": 3.0841,
+      "step": 2067
+    },
+    {
+      "epoch": 27.588628762541806,
+      "grad_norm": 1.4431540966033936,
+      "learning_rate": 0.00010825361512791992,
+      "loss": 3.145,
+      "step": 2068
+    },
+    {
+      "epoch": 27.60200668896321,
+      "grad_norm": 1.6104519367218018,
+      "learning_rate": 0.00010820912124582869,
+      "loss": 3.0226,
+      "step": 2069
+    },
+    {
+      "epoch": 27.615384615384617,
+      "grad_norm": 1.249314546585083,
+      "learning_rate": 0.00010816462736373749,
+      "loss": 3.1273,
+      "step": 2070
+    },
+    {
+      "epoch": 27.62876254180602,
+      "grad_norm": 1.1928060054779053,
+      "learning_rate": 0.00010812013348164627,
+      "loss": 2.8877,
+      "step": 2071
+    },
+    {
+      "epoch": 27.642140468227424,
+      "grad_norm": 1.5014128684997559,
+      "learning_rate": 0.00010807563959955507,
+      "loss": 3.4129,
+      "step": 2072
+    },
+    {
+      "epoch": 27.65551839464883,
+      "grad_norm": 1.344773530960083,
+      "learning_rate": 0.00010803114571746385,
+      "loss": 3.1112,
+      "step": 2073
+    },
+    {
+      "epoch": 27.668896321070235,
+      "grad_norm": 1.616158366203308,
+      "learning_rate": 0.00010798665183537264,
+      "loss": 3.1799,
+      "step": 2074
+    },
+    {
+      "epoch": 27.682274247491637,
+      "grad_norm": 1.4087249040603638,
+      "learning_rate": 0.00010794215795328141,
+      "loss": 2.9315,
+      "step": 2075
+    },
+    {
+      "epoch": 27.695652173913043,
+      "grad_norm": 1.48402738571167,
+      "learning_rate": 0.00010789766407119022,
+      "loss": 3.1755,
+      "step": 2076
+    },
+    {
+      "epoch": 27.70903010033445,
+      "grad_norm": 1.3316876888275146,
+      "learning_rate": 0.00010785317018909899,
+      "loss": 3.087,
+      "step": 2077
+    },
+    {
+      "epoch": 27.722408026755854,
+      "grad_norm": 1.3765549659729004,
+      "learning_rate": 0.0001078086763070078,
+      "loss": 3.2595,
+      "step": 2078
+    },
+    {
+      "epoch": 27.735785953177256,
+      "grad_norm": 1.646952748298645,
+      "learning_rate": 0.00010776418242491657,
+      "loss": 3.0627,
+      "step": 2079
+    },
+    {
+      "epoch": 27.74916387959866,
+      "grad_norm": 1.2064934968948364,
+      "learning_rate": 0.00010771968854282537,
+      "loss": 3.0569,
+      "step": 2080
+    },
+    {
+      "epoch": 27.762541806020067,
+      "grad_norm": 1.3027973175048828,
+      "learning_rate": 0.00010767519466073415,
+      "loss": 3.0246,
+      "step": 2081
+    },
+    {
+      "epoch": 27.775919732441473,
+      "grad_norm": 1.4610519409179688,
+      "learning_rate": 0.00010763070077864294,
+      "loss": 3.0123,
+      "step": 2082
+    },
+    {
+      "epoch": 27.789297658862875,
+      "grad_norm": 1.376183032989502,
+      "learning_rate": 0.00010758620689655173,
+      "loss": 3.1312,
+      "step": 2083
+    },
+    {
+      "epoch": 27.80267558528428,
+      "grad_norm": 1.3715664148330688,
+      "learning_rate": 0.00010754171301446052,
+      "loss": 3.0104,
+      "step": 2084
+    },
+    {
+      "epoch": 27.816053511705686,
+      "grad_norm": 1.2951513528823853,
+      "learning_rate": 0.00010749721913236929,
+      "loss": 3.0353,
+      "step": 2085
+    },
+    {
+      "epoch": 27.82943143812709,
+      "grad_norm": 1.1614545583724976,
+      "learning_rate": 0.0001074527252502781,
+      "loss": 2.8977,
+      "step": 2086
+    },
+    {
+      "epoch": 27.842809364548494,
+      "grad_norm": 1.3948044776916504,
+      "learning_rate": 0.00010740823136818687,
+      "loss": 2.8611,
+      "step": 2087
+    },
+    {
+      "epoch": 27.8561872909699,
+      "grad_norm": 1.3053209781646729,
+      "learning_rate": 0.00010736373748609568,
+      "loss": 2.7965,
+      "step": 2088
+    },
+    {
+      "epoch": 27.869565217391305,
+      "grad_norm": 1.289396047592163,
+      "learning_rate": 0.00010731924360400445,
+      "loss": 2.8663,
+      "step": 2089
+    },
+    {
+      "epoch": 27.88294314381271,
+      "grad_norm": 1.4188008308410645,
+      "learning_rate": 0.00010727474972191324,
+      "loss": 3.0139,
+      "step": 2090
+    },
+    {
+      "epoch": 27.896321070234112,
+      "grad_norm": 1.5196731090545654,
+      "learning_rate": 0.00010723025583982203,
+      "loss": 2.9593,
+      "step": 2091
+    },
+    {
+      "epoch": 27.909698996655518,
+      "grad_norm": 1.5437039136886597,
+      "learning_rate": 0.00010718576195773082,
+      "loss": 2.8009,
+      "step": 2092
+    },
+    {
+      "epoch": 27.923076923076923,
+      "grad_norm": 1.4248530864715576,
+      "learning_rate": 0.0001071412680756396,
+      "loss": 2.8469,
+      "step": 2093
+    },
+    {
+      "epoch": 27.93645484949833,
+      "grad_norm": 1.4018962383270264,
+      "learning_rate": 0.0001070967741935484,
+      "loss": 3.0897,
+      "step": 2094
+    },
+    {
+      "epoch": 27.94983277591973,
+      "grad_norm": 1.2483067512512207,
+      "learning_rate": 0.00010705228031145717,
+      "loss": 3.1738,
+      "step": 2095
+    },
+    {
+      "epoch": 27.963210702341136,
+      "grad_norm": 1.25924813747406,
+      "learning_rate": 0.00010700778642936598,
+      "loss": 3.263,
+      "step": 2096
+    },
+    {
+      "epoch": 27.976588628762542,
+      "grad_norm": 1.2528305053710938,
+      "learning_rate": 0.00010696329254727475,
+      "loss": 3.0374,
+      "step": 2097
+    },
+    {
+      "epoch": 27.989966555183948,
+      "grad_norm": 1.7555912733078003,
+      "learning_rate": 0.00010691879866518356,
+      "loss": 3.1906,
+      "step": 2098
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 1.7080843448638916,
+      "learning_rate": 0.00010687430478309233,
+      "loss": 2.9748,
+      "step": 2099
+    },
+    {
+      "epoch": 28.013377926421406,
+      "grad_norm": 1.120409369468689,
+      "learning_rate": 0.00010682981090100112,
+      "loss": 2.9385,
+      "step": 2100
+    },
+    {
+      "epoch": 28.02675585284281,
+      "grad_norm": 0.8793032169342041,
+      "learning_rate": 0.0001067853170189099,
+      "loss": 2.9205,
+      "step": 2101
+    },
+    {
+      "epoch": 28.040133779264213,
+      "grad_norm": 1.018744707107544,
+      "learning_rate": 0.0001067408231368187,
+      "loss": 3.1447,
+      "step": 2102
+    },
+    {
+      "epoch": 28.05351170568562,
+      "grad_norm": 0.9645024538040161,
+      "learning_rate": 0.00010669632925472748,
+      "loss": 2.783,
+      "step": 2103
+    },
+    {
+      "epoch": 28.066889632107024,
+      "grad_norm": 1.0723059177398682,
+      "learning_rate": 0.00010665183537263628,
+      "loss": 2.9627,
+      "step": 2104
+    },
+    {
+      "epoch": 28.08026755852843,
+      "grad_norm": 1.0304062366485596,
+      "learning_rate": 0.00010660734149054505,
+      "loss": 2.8118,
+      "step": 2105
+    },
+    {
+      "epoch": 28.093645484949832,
+      "grad_norm": 1.0894114971160889,
+      "learning_rate": 0.00010656284760845386,
+      "loss": 2.9643,
+      "step": 2106
+    },
+    {
+      "epoch": 28.107023411371237,
+      "grad_norm": 1.1615228652954102,
+      "learning_rate": 0.00010651835372636263,
+      "loss": 2.9402,
+      "step": 2107
+    },
+    {
+      "epoch": 28.120401337792643,
+      "grad_norm": 1.066375732421875,
+      "learning_rate": 0.00010647385984427144,
+      "loss": 2.8733,
+      "step": 2108
+    },
+    {
+      "epoch": 28.13377926421405,
+      "grad_norm": 1.1981687545776367,
+      "learning_rate": 0.0001064293659621802,
+      "loss": 2.9301,
+      "step": 2109
+    },
+    {
+      "epoch": 28.14715719063545,
+      "grad_norm": 1.0867598056793213,
+      "learning_rate": 0.00010638487208008899,
+      "loss": 3.1676,
+      "step": 2110
+    },
+    {
+      "epoch": 28.160535117056856,
+      "grad_norm": 1.156256079673767,
+      "learning_rate": 0.00010634037819799778,
+      "loss": 2.9555,
+      "step": 2111
+    },
+    {
+      "epoch": 28.17391304347826,
+      "grad_norm": 0.9931294322013855,
+      "learning_rate": 0.00010629588431590655,
+      "loss": 2.7119,
+      "step": 2112
+    },
+    {
+      "epoch": 28.187290969899667,
+      "grad_norm": 0.9383543729782104,
+      "learning_rate": 0.00010625139043381536,
+      "loss": 2.9251,
+      "step": 2113
+    },
+    {
+      "epoch": 28.20066889632107,
+      "grad_norm": 0.859805166721344,
+      "learning_rate": 0.00010620689655172413,
+      "loss": 2.8734,
+      "step": 2114
+    },
+    {
+      "epoch": 28.214046822742475,
+      "grad_norm": 1.081536889076233,
+      "learning_rate": 0.00010616240266963293,
+      "loss": 3.0356,
+      "step": 2115
+    },
+    {
+      "epoch": 28.22742474916388,
+      "grad_norm": 1.132004976272583,
+      "learning_rate": 0.00010611790878754171,
+      "loss": 3.1751,
+      "step": 2116
+    },
+    {
+      "epoch": 28.240802675585286,
+      "grad_norm": 1.0406700372695923,
+      "learning_rate": 0.0001060734149054505,
+      "loss": 2.913,
+      "step": 2117
+    },
+    {
+      "epoch": 28.254180602006688,
+      "grad_norm": 1.4362869262695312,
+      "learning_rate": 0.00010602892102335929,
+      "loss": 3.0823,
+      "step": 2118
+    },
+    {
+      "epoch": 28.267558528428093,
+      "grad_norm": 1.0609610080718994,
+      "learning_rate": 0.00010598442714126808,
+      "loss": 3.0693,
+      "step": 2119
+    },
+    {
+      "epoch": 28.2809364548495,
+      "grad_norm": 1.0952329635620117,
+      "learning_rate": 0.00010593993325917686,
+      "loss": 2.9953,
+      "step": 2120
+    },
+    {
+      "epoch": 28.294314381270905,
+      "grad_norm": 1.0358316898345947,
+      "learning_rate": 0.00010589543937708566,
+      "loss": 2.7737,
+      "step": 2121
+    },
+    {
+      "epoch": 28.307692307692307,
+      "grad_norm": 1.1332812309265137,
+      "learning_rate": 0.00010585094549499443,
+      "loss": 2.9874,
+      "step": 2122
+    },
+    {
+      "epoch": 28.321070234113712,
+      "grad_norm": 1.0737558603286743,
+      "learning_rate": 0.00010580645161290324,
+      "loss": 3.0547,
+      "step": 2123
+    },
+    {
+      "epoch": 28.334448160535118,
+      "grad_norm": 1.030110478401184,
+      "learning_rate": 0.00010576195773081201,
+      "loss": 2.9786,
+      "step": 2124
+    },
+    {
+      "epoch": 28.347826086956523,
+      "grad_norm": 1.09152352809906,
+      "learning_rate": 0.0001057174638487208,
+      "loss": 2.6407,
+      "step": 2125
+    },
+    {
+      "epoch": 28.361204013377925,
+      "grad_norm": 0.9880086183547974,
+      "learning_rate": 0.00010567296996662959,
+      "loss": 3.1367,
+      "step": 2126
+    },
+    {
+      "epoch": 28.37458193979933,
+      "grad_norm": 1.1486655473709106,
+      "learning_rate": 0.00010562847608453838,
+      "loss": 3.1705,
+      "step": 2127
+    },
+    {
+      "epoch": 28.387959866220736,
+      "grad_norm": 1.199541449546814,
+      "learning_rate": 0.00010558398220244716,
+      "loss": 3.0254,
+      "step": 2128
+    },
+    {
+      "epoch": 28.401337792642142,
+      "grad_norm": 0.9470932483673096,
+      "learning_rate": 0.00010553948832035596,
+      "loss": 3.1296,
+      "step": 2129
+    },
+    {
+      "epoch": 28.414715719063544,
+      "grad_norm": 1.1546344757080078,
+      "learning_rate": 0.00010549499443826474,
+      "loss": 2.9981,
+      "step": 2130
+    },
+    {
+      "epoch": 28.42809364548495,
+      "grad_norm": 1.0525349378585815,
+      "learning_rate": 0.00010545050055617354,
+      "loss": 2.884,
+      "step": 2131
+    },
+    {
+      "epoch": 28.441471571906355,
+      "grad_norm": 0.8965997695922852,
+      "learning_rate": 0.00010540600667408231,
+      "loss": 3.1666,
+      "step": 2132
+    },
+    {
+      "epoch": 28.45484949832776,
+      "grad_norm": 1.0740344524383545,
+      "learning_rate": 0.00010536151279199112,
+      "loss": 2.9477,
+      "step": 2133
+    },
+    {
+      "epoch": 28.468227424749163,
+      "grad_norm": 1.0058174133300781,
+      "learning_rate": 0.00010531701890989989,
+      "loss": 2.9355,
+      "step": 2134
+    },
+    {
+      "epoch": 28.48160535117057,
+      "grad_norm": 1.0790162086486816,
+      "learning_rate": 0.00010527252502780868,
+      "loss": 2.8728,
+      "step": 2135
+    },
+    {
+      "epoch": 28.494983277591974,
+      "grad_norm": 1.0231183767318726,
+      "learning_rate": 0.00010522803114571746,
+      "loss": 2.6208,
+      "step": 2136
+    },
+    {
+      "epoch": 28.50836120401338,
+      "grad_norm": 1.1527316570281982,
+      "learning_rate": 0.00010518353726362626,
+      "loss": 3.0255,
+      "step": 2137
+    },
+    {
+      "epoch": 28.52173913043478,
+      "grad_norm": 1.0380228757858276,
+      "learning_rate": 0.00010513904338153504,
+      "loss": 3.1382,
+      "step": 2138
+    },
+    {
+      "epoch": 28.535117056856187,
+      "grad_norm": 1.1528582572937012,
+      "learning_rate": 0.00010509454949944384,
+      "loss": 2.9921,
+      "step": 2139
+    },
+    {
+      "epoch": 28.548494983277592,
+      "grad_norm": 1.0610437393188477,
+      "learning_rate": 0.00010505005561735262,
+      "loss": 2.832,
+      "step": 2140
+    },
+    {
+      "epoch": 28.561872909698998,
+      "grad_norm": 1.1252180337905884,
+      "learning_rate": 0.00010500556173526142,
+      "loss": 3.0318,
+      "step": 2141
+    },
+    {
+      "epoch": 28.5752508361204,
+      "grad_norm": 0.910602331161499,
+      "learning_rate": 0.00010496106785317019,
+      "loss": 3.0174,
+      "step": 2142
+    },
+    {
+      "epoch": 28.588628762541806,
+      "grad_norm": 1.1615829467773438,
+      "learning_rate": 0.000104916573971079,
+      "loss": 3.1465,
+      "step": 2143
+    },
+    {
+      "epoch": 28.60200668896321,
+      "grad_norm": 1.0431264638900757,
+      "learning_rate": 0.00010487208008898776,
+      "loss": 2.9406,
+      "step": 2144
+    },
+    {
+      "epoch": 28.615384615384617,
+      "grad_norm": 1.1530067920684814,
+      "learning_rate": 0.00010482758620689656,
+      "loss": 2.7672,
+      "step": 2145
+    },
+    {
+      "epoch": 28.62876254180602,
+      "grad_norm": 1.0598475933074951,
+      "learning_rate": 0.00010478309232480534,
+      "loss": 3.0477,
+      "step": 2146
+    },
+    {
+      "epoch": 28.642140468227424,
+      "grad_norm": 0.9107276797294617,
+      "learning_rate": 0.00010473859844271414,
+      "loss": 2.949,
+      "step": 2147
+    },
+    {
+      "epoch": 28.65551839464883,
+      "grad_norm": 1.0739660263061523,
+      "learning_rate": 0.00010469410456062292,
+      "loss": 3.0195,
+      "step": 2148
+    },
+    {
+      "epoch": 28.668896321070235,
+      "grad_norm": 1.0200115442276,
+      "learning_rate": 0.00010464961067853172,
+      "loss": 2.9086,
+      "step": 2149
+    },
+    {
+      "epoch": 28.682274247491637,
+      "grad_norm": 0.9945818185806274,
+      "learning_rate": 0.0001046051167964405,
+      "loss": 3.3103,
+      "step": 2150
+    },
+    {
+      "epoch": 28.695652173913043,
+      "grad_norm": 1.231123447418213,
+      "learning_rate": 0.00010456062291434927,
+      "loss": 2.9266,
+      "step": 2151
+    },
+    {
+      "epoch": 28.70903010033445,
+      "grad_norm": 1.02529776096344,
+      "learning_rate": 0.00010451612903225806,
+      "loss": 2.8796,
+      "step": 2152
+    },
+    {
+      "epoch": 28.722408026755854,
+      "grad_norm": 1.1735590696334839,
+      "learning_rate": 0.00010447163515016685,
+      "loss": 2.8582,
+      "step": 2153
+    },
+    {
+      "epoch": 28.735785953177256,
+      "grad_norm": 1.0260053873062134,
+      "learning_rate": 0.00010442714126807564,
+      "loss": 3.1459,
+      "step": 2154
+    },
+    {
+      "epoch": 28.74916387959866,
+      "grad_norm": 1.0473296642303467,
+      "learning_rate": 0.00010438264738598442,
+      "loss": 3.1804,
+      "step": 2155
+    },
+    {
+      "epoch": 28.762541806020067,
+      "grad_norm": 1.0932636260986328,
+      "learning_rate": 0.00010433815350389322,
+      "loss": 2.8114,
+      "step": 2156
+    },
+    {
+      "epoch": 28.775919732441473,
+      "grad_norm": 1.115605115890503,
+      "learning_rate": 0.000104293659621802,
+      "loss": 2.925,
+      "step": 2157
+    },
+    {
+      "epoch": 28.789297658862875,
+      "grad_norm": 1.0479992628097534,
+      "learning_rate": 0.0001042491657397108,
+      "loss": 3.0461,
+      "step": 2158
+    },
+    {
+      "epoch": 28.80267558528428,
+      "grad_norm": 0.9241504669189453,
+      "learning_rate": 0.00010420467185761957,
+      "loss": 3.0755,
+      "step": 2159
+    },
+    {
+      "epoch": 28.816053511705686,
+      "grad_norm": 0.9609296917915344,
+      "learning_rate": 0.00010416017797552838,
+      "loss": 3.077,
+      "step": 2160
+    },
+    {
+      "epoch": 28.82943143812709,
+      "grad_norm": 0.9561198949813843,
+      "learning_rate": 0.00010411568409343715,
+      "loss": 2.9041,
+      "step": 2161
+    },
+    {
+      "epoch": 28.842809364548494,
+      "grad_norm": 1.1552175283432007,
+      "learning_rate": 0.00010407119021134594,
+      "loss": 2.8253,
+      "step": 2162
+    },
+    {
+      "epoch": 28.8561872909699,
+      "grad_norm": 0.8945892453193665,
+      "learning_rate": 0.00010402669632925472,
+      "loss": 2.9228,
+      "step": 2163
+    },
+    {
+      "epoch": 28.869565217391305,
+      "grad_norm": 1.0247336626052856,
+      "learning_rate": 0.00010398220244716352,
+      "loss": 2.6202,
+      "step": 2164
+    },
+    {
+      "epoch": 28.88294314381271,
+      "grad_norm": 1.0500261783599854,
+      "learning_rate": 0.0001039377085650723,
+      "loss": 2.777,
+      "step": 2165
+    },
+    {
+      "epoch": 28.896321070234112,
+      "grad_norm": 1.0250235795974731,
+      "learning_rate": 0.0001038932146829811,
+      "loss": 3.0606,
+      "step": 2166
+    },
+    {
+      "epoch": 28.909698996655518,
+      "grad_norm": 1.0018290281295776,
+      "learning_rate": 0.00010384872080088988,
+      "loss": 3.0099,
+      "step": 2167
+    },
+    {
+      "epoch": 28.923076923076923,
+      "grad_norm": 0.8957269191741943,
+      "learning_rate": 0.00010380422691879868,
+      "loss": 2.7081,
+      "step": 2168
+    },
+    {
+      "epoch": 28.93645484949833,
+      "grad_norm": 1.060746431350708,
+      "learning_rate": 0.00010375973303670745,
+      "loss": 2.9771,
+      "step": 2169
+    },
+    {
+      "epoch": 28.94983277591973,
+      "grad_norm": 0.9782920479774475,
+      "learning_rate": 0.00010371523915461626,
+      "loss": 3.035,
+      "step": 2170
+    },
+    {
+      "epoch": 28.963210702341136,
+      "grad_norm": 0.9972925186157227,
+      "learning_rate": 0.00010367074527252502,
+      "loss": 3.0531,
+      "step": 2171
+    },
+    {
+      "epoch": 28.976588628762542,
+      "grad_norm": 1.1463640928268433,
+      "learning_rate": 0.00010362625139043382,
+      "loss": 2.9485,
+      "step": 2172
+    },
+    {
+      "epoch": 28.989966555183948,
+      "grad_norm": 1.0142980813980103,
+      "learning_rate": 0.0001035817575083426,
+      "loss": 2.8852,
+      "step": 2173
+    },
+    {
+      "epoch": 29.0,
+      "grad_norm": 1.347244381904602,
+      "learning_rate": 0.0001035372636262514,
+      "loss": 2.898,
+      "step": 2174
+    },
+    {
+      "epoch": 29.013377926421406,
+      "grad_norm": 0.9205636978149414,
+      "learning_rate": 0.00010349276974416018,
+      "loss": 2.7324,
+      "step": 2175
+    },
+    {
+      "epoch": 29.02675585284281,
+      "grad_norm": 0.9123748540878296,
+      "learning_rate": 0.00010344827586206898,
+      "loss": 2.943,
+      "step": 2176
+    },
+    {
+      "epoch": 29.040133779264213,
+      "grad_norm": 0.8594069480895996,
+      "learning_rate": 0.00010340378197997776,
+      "loss": 2.8531,
+      "step": 2177
+    },
+    {
+      "epoch": 29.05351170568562,
+      "grad_norm": 0.7439792156219482,
+      "learning_rate": 0.00010335928809788655,
+      "loss": 2.521,
+      "step": 2178
+    },
+    {
+      "epoch": 29.066889632107024,
+      "grad_norm": 0.7489947080612183,
+      "learning_rate": 0.00010331479421579532,
+      "loss": 2.9317,
+      "step": 2179
+    },
+    {
+      "epoch": 29.08026755852843,
+      "grad_norm": 0.8311702013015747,
+      "learning_rate": 0.00010327030033370413,
+      "loss": 2.7137,
+      "step": 2180
+    },
+    {
+      "epoch": 29.093645484949832,
+      "grad_norm": 0.9763099551200867,
+      "learning_rate": 0.0001032258064516129,
+      "loss": 2.7502,
+      "step": 2181
+    },
+    {
+      "epoch": 29.107023411371237,
+      "grad_norm": 0.8277449607849121,
+      "learning_rate": 0.0001031813125695217,
+      "loss": 2.9792,
+      "step": 2182
+    },
+    {
+      "epoch": 29.120401337792643,
+      "grad_norm": 0.7491191029548645,
+      "learning_rate": 0.00010313681868743048,
+      "loss": 2.8479,
+      "step": 2183
+    },
+    {
+      "epoch": 29.13377926421405,
+      "grad_norm": 0.7314183115959167,
+      "learning_rate": 0.00010309232480533928,
+      "loss": 2.8962,
+      "step": 2184
+    },
+    {
+      "epoch": 29.14715719063545,
+      "grad_norm": 0.8334509134292603,
+      "learning_rate": 0.00010304783092324806,
+      "loss": 2.86,
+      "step": 2185
+    },
+    {
+      "epoch": 29.160535117056856,
+      "grad_norm": 0.9736613631248474,
+      "learning_rate": 0.00010300333704115685,
+      "loss": 2.9289,
+      "step": 2186
+    },
+    {
+      "epoch": 29.17391304347826,
+      "grad_norm": 0.9118865132331848,
+      "learning_rate": 0.00010295884315906564,
+      "loss": 3.0865,
+      "step": 2187
+    },
+    {
+      "epoch": 29.187290969899667,
+      "grad_norm": 0.992799699306488,
+      "learning_rate": 0.00010291434927697443,
+      "loss": 2.7811,
+      "step": 2188
+    },
+    {
+      "epoch": 29.20066889632107,
+      "grad_norm": 1.0415345430374146,
+      "learning_rate": 0.0001028698553948832,
+      "loss": 2.7707,
+      "step": 2189
+    },
+    {
+      "epoch": 29.214046822742475,
+      "grad_norm": 0.7599477171897888,
+      "learning_rate": 0.00010282536151279201,
+      "loss": 2.8162,
+      "step": 2190
+    },
+    {
+      "epoch": 29.22742474916388,
+      "grad_norm": 1.17939293384552,
+      "learning_rate": 0.00010278086763070078,
+      "loss": 2.7939,
+      "step": 2191
+    },
+    {
+      "epoch": 29.240802675585286,
+      "grad_norm": 0.8793815970420837,
+      "learning_rate": 0.00010273637374860956,
+      "loss": 2.7704,
+      "step": 2192
+    },
+    {
+      "epoch": 29.254180602006688,
+      "grad_norm": 0.838680624961853,
+      "learning_rate": 0.00010269187986651836,
+      "loss": 2.8513,
+      "step": 2193
+    },
+    {
+      "epoch": 29.267558528428093,
+      "grad_norm": 0.8918489217758179,
+      "learning_rate": 0.00010264738598442714,
+      "loss": 2.7748,
+      "step": 2194
+    },
+    {
+      "epoch": 29.2809364548495,
+      "grad_norm": 0.7513167858123779,
+      "learning_rate": 0.00010260289210233594,
+      "loss": 2.8151,
+      "step": 2195
+    },
+    {
+      "epoch": 29.294314381270905,
+      "grad_norm": 0.7921515703201294,
+      "learning_rate": 0.0001025583982202447,
+      "loss": 2.8004,
+      "step": 2196
+    },
+    {
+      "epoch": 29.307692307692307,
+      "grad_norm": 0.7862170934677124,
+      "learning_rate": 0.00010251390433815352,
+      "loss": 2.9749,
+      "step": 2197
+    },
+    {
+      "epoch": 29.321070234113712,
+      "grad_norm": 0.8512644171714783,
+      "learning_rate": 0.00010246941045606228,
+      "loss": 2.9936,
+      "step": 2198
+    },
+    {
+      "epoch": 29.334448160535118,
+      "grad_norm": 0.9074665307998657,
+      "learning_rate": 0.00010242491657397108,
+      "loss": 2.814,
+      "step": 2199
+    },
+    {
+      "epoch": 29.347826086956523,
+      "grad_norm": 0.9875670075416565,
+      "learning_rate": 0.00010238042269187986,
+      "loss": 2.9862,
+      "step": 2200
+    },
+    {
+      "epoch": 29.361204013377925,
+      "grad_norm": 0.7445062398910522,
+      "learning_rate": 0.00010233592880978866,
+      "loss": 2.8233,
+      "step": 2201
+    },
+    {
+      "epoch": 29.37458193979933,
+      "grad_norm": 0.8388944268226624,
+      "learning_rate": 0.00010229143492769744,
+      "loss": 2.8983,
+      "step": 2202
+    },
+    {
+      "epoch": 29.387959866220736,
+      "grad_norm": 0.7882921099662781,
+      "learning_rate": 0.00010224694104560624,
+      "loss": 2.9521,
+      "step": 2203
+    },
+    {
+      "epoch": 29.401337792642142,
+      "grad_norm": 0.8031525015830994,
+      "learning_rate": 0.00010220244716351502,
+      "loss": 2.8722,
+      "step": 2204
+    },
+    {
+      "epoch": 29.414715719063544,
+      "grad_norm": 0.8128193616867065,
+      "learning_rate": 0.00010215795328142381,
+      "loss": 2.972,
+      "step": 2205
+    },
+    {
+      "epoch": 29.42809364548495,
+      "grad_norm": 1.153869867324829,
+      "learning_rate": 0.00010211345939933258,
+      "loss": 2.8711,
+      "step": 2206
+    },
+    {
+      "epoch": 29.441471571906355,
+      "grad_norm": 0.8942824602127075,
+      "learning_rate": 0.0001020689655172414,
+      "loss": 3.2308,
+      "step": 2207
+    },
+    {
+      "epoch": 29.45484949832776,
+      "grad_norm": 0.8985738158226013,
+      "learning_rate": 0.00010202447163515016,
+      "loss": 2.8042,
+      "step": 2208
+    },
+    {
+      "epoch": 29.468227424749163,
+      "grad_norm": 0.8202070593833923,
+      "learning_rate": 0.00010197997775305896,
+      "loss": 2.7193,
+      "step": 2209
+    },
+    {
+      "epoch": 29.48160535117057,
+      "grad_norm": 0.7752913236618042,
+      "learning_rate": 0.00010193548387096774,
+      "loss": 2.8162,
+      "step": 2210
+    },
+    {
+      "epoch": 29.494983277591974,
+      "grad_norm": 0.8399960994720459,
+      "learning_rate": 0.00010189098998887654,
+      "loss": 3.0488,
+      "step": 2211
+    },
+    {
+      "epoch": 29.50836120401338,
+      "grad_norm": 0.9991989731788635,
+      "learning_rate": 0.00010184649610678532,
+      "loss": 2.8647,
+      "step": 2212
+    },
+    {
+      "epoch": 29.52173913043478,
+      "grad_norm": 0.8456683158874512,
+      "learning_rate": 0.00010180200222469411,
+      "loss": 2.9635,
+      "step": 2213
+    },
+    {
+      "epoch": 29.535117056856187,
+      "grad_norm": 0.7978901863098145,
+      "learning_rate": 0.0001017575083426029,
+      "loss": 2.8927,
+      "step": 2214
+    },
+    {
+      "epoch": 29.548494983277592,
+      "grad_norm": 0.8560264706611633,
+      "learning_rate": 0.00010171301446051169,
+      "loss": 2.9403,
+      "step": 2215
+    },
+    {
+      "epoch": 29.561872909698998,
+      "grad_norm": 0.7215510010719299,
+      "learning_rate": 0.00010166852057842046,
+      "loss": 2.7367,
+      "step": 2216
+    },
+    {
+      "epoch": 29.5752508361204,
+      "grad_norm": 0.8395666480064392,
+      "learning_rate": 0.00010162402669632927,
+      "loss": 2.9857,
+      "step": 2217
+    },
+    {
+      "epoch": 29.588628762541806,
+      "grad_norm": 0.9163089394569397,
+      "learning_rate": 0.00010157953281423804,
+      "loss": 2.9605,
+      "step": 2218
+    },
+    {
+      "epoch": 29.60200668896321,
+      "grad_norm": 0.8586741089820862,
+      "learning_rate": 0.00010153503893214684,
+      "loss": 2.9543,
+      "step": 2219
+    },
+    {
+      "epoch": 29.615384615384617,
+      "grad_norm": 0.7774202227592468,
+      "learning_rate": 0.00010149054505005562,
+      "loss": 3.0245,
+      "step": 2220
+    },
+    {
+      "epoch": 29.62876254180602,
+      "grad_norm": 0.8288631439208984,
+      "learning_rate": 0.00010144605116796441,
+      "loss": 2.975,
+      "step": 2221
+    },
+    {
+      "epoch": 29.642140468227424,
+      "grad_norm": 0.7886912226676941,
+      "learning_rate": 0.0001014015572858732,
+      "loss": 2.6679,
+      "step": 2222
+    },
+    {
+      "epoch": 29.65551839464883,
+      "grad_norm": 0.7964534759521484,
+      "learning_rate": 0.00010135706340378199,
+      "loss": 2.9989,
+      "step": 2223
+    },
+    {
+      "epoch": 29.668896321070235,
+      "grad_norm": 1.0234196186065674,
+      "learning_rate": 0.00010131256952169078,
+      "loss": 3.2245,
+      "step": 2224
+    },
+    {
+      "epoch": 29.682274247491637,
+      "grad_norm": 0.8414390683174133,
+      "learning_rate": 0.00010126807563959957,
+      "loss": 3.0525,
+      "step": 2225
+    },
+    {
+      "epoch": 29.695652173913043,
+      "grad_norm": 0.8034384250640869,
+      "learning_rate": 0.00010122358175750834,
+      "loss": 2.9012,
+      "step": 2226
+    },
+    {
+      "epoch": 29.70903010033445,
+      "grad_norm": 0.7711749076843262,
+      "learning_rate": 0.00010117908787541715,
+      "loss": 2.7548,
+      "step": 2227
+    },
+    {
+      "epoch": 29.722408026755854,
+      "grad_norm": 0.9712610840797424,
+      "learning_rate": 0.00010113459399332592,
+      "loss": 3.1162,
+      "step": 2228
+    },
+    {
+      "epoch": 29.735785953177256,
+      "grad_norm": 0.8966991305351257,
+      "learning_rate": 0.00010109010011123471,
+      "loss": 3.0883,
+      "step": 2229
+    },
+    {
+      "epoch": 29.74916387959866,
+      "grad_norm": 0.8569051027297974,
+      "learning_rate": 0.0001010456062291435,
+      "loss": 2.8909,
+      "step": 2230
+    },
+    {
+      "epoch": 29.762541806020067,
+      "grad_norm": 0.9149661660194397,
+      "learning_rate": 0.00010100111234705229,
+      "loss": 2.7559,
+      "step": 2231
+    },
+    {
+      "epoch": 29.775919732441473,
+      "grad_norm": 1.1889090538024902,
+      "learning_rate": 0.00010095661846496107,
+      "loss": 2.7872,
+      "step": 2232
+    },
+    {
+      "epoch": 29.789297658862875,
+      "grad_norm": 0.8464928865432739,
+      "learning_rate": 0.00010091212458286984,
+      "loss": 2.8496,
+      "step": 2233
+    },
+    {
+      "epoch": 29.80267558528428,
+      "grad_norm": 0.8210122585296631,
+      "learning_rate": 0.00010086763070077865,
+      "loss": 3.0782,
+      "step": 2234
+    },
+    {
+      "epoch": 29.816053511705686,
+      "grad_norm": 0.7978618741035461,
+      "learning_rate": 0.00010082313681868742,
+      "loss": 2.9633,
+      "step": 2235
+    },
+    {
+      "epoch": 29.82943143812709,
+      "grad_norm": 0.7863582968711853,
+      "learning_rate": 0.00010077864293659622,
+      "loss": 2.9536,
+      "step": 2236
+    },
+    {
+      "epoch": 29.842809364548494,
+      "grad_norm": 0.9451448917388916,
+      "learning_rate": 0.000100734149054505,
+      "loss": 2.945,
+      "step": 2237
+    },
+    {
+      "epoch": 29.8561872909699,
+      "grad_norm": 0.9478015303611755,
+      "learning_rate": 0.0001006896551724138,
+      "loss": 2.8408,
+      "step": 2238
+    },
+    {
+      "epoch": 29.869565217391305,
+      "grad_norm": 0.804568886756897,
+      "learning_rate": 0.00010064516129032258,
+      "loss": 3.093,
+      "step": 2239
+    },
+    {
+      "epoch": 29.88294314381271,
+      "grad_norm": 0.8370556235313416,
+      "learning_rate": 0.00010060066740823137,
+      "loss": 3.0561,
+      "step": 2240
+    },
+    {
+      "epoch": 29.896321070234112,
+      "grad_norm": 0.9402264952659607,
+      "learning_rate": 0.00010055617352614016,
+      "loss": 3.0024,
+      "step": 2241
+    },
+    {
+      "epoch": 29.909698996655518,
+      "grad_norm": 0.7789230942726135,
+      "learning_rate": 0.00010051167964404895,
+      "loss": 2.8388,
+      "step": 2242
+    },
+    {
+      "epoch": 29.923076923076923,
+      "grad_norm": 0.8228744268417358,
+      "learning_rate": 0.00010046718576195772,
+      "loss": 3.1335,
+      "step": 2243
+    },
+    {
+      "epoch": 29.93645484949833,
+      "grad_norm": 0.9612992405891418,
+      "learning_rate": 0.00010042269187986653,
+      "loss": 3.0355,
+      "step": 2244
+    },
+    {
+      "epoch": 29.94983277591973,
+      "grad_norm": 0.989894688129425,
+      "learning_rate": 0.0001003781979977753,
+      "loss": 2.6879,
+      "step": 2245
+    },
+    {
+      "epoch": 29.963210702341136,
+      "grad_norm": 0.7333267331123352,
+      "learning_rate": 0.0001003337041156841,
+      "loss": 3.004,
+      "step": 2246
+    },
+    {
+      "epoch": 29.976588628762542,
+      "grad_norm": 0.802925705909729,
+      "learning_rate": 0.00010028921023359288,
+      "loss": 3.0526,
+      "step": 2247
+    },
+    {
+      "epoch": 29.989966555183948,
+      "grad_norm": 0.8313626646995544,
+      "learning_rate": 0.00010024471635150167,
+      "loss": 2.554,
+      "step": 2248
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 1.1626179218292236,
+      "learning_rate": 0.00010020022246941046,
+      "loss": 3.0765,
+      "step": 2249
+    },
+    {
+      "epoch": 30.013377926421406,
+      "grad_norm": 0.6676865220069885,
+      "learning_rate": 0.00010015572858731925,
+      "loss": 2.8693,
+      "step": 2250
+    },
+    {
+      "epoch": 30.02675585284281,
+      "grad_norm": 0.719235897064209,
+      "learning_rate": 0.00010011123470522804,
+      "loss": 2.8247,
+      "step": 2251
+    },
+    {
+      "epoch": 30.040133779264213,
+      "grad_norm": 0.8199754953384399,
+      "learning_rate": 0.00010006674082313683,
+      "loss": 3.0088,
+      "step": 2252
+    },
+    {
+      "epoch": 30.05351170568562,
+      "grad_norm": 0.7587671279907227,
+      "learning_rate": 0.0001000222469410456,
+      "loss": 3.118,
+      "step": 2253
+    },
+    {
+      "epoch": 30.066889632107024,
+      "grad_norm": 0.7287217974662781,
+      "learning_rate": 9.99777530589544e-05,
+      "loss": 2.6612,
+      "step": 2254
+    },
+    {
+      "epoch": 30.08026755852843,
+      "grad_norm": 0.6484748125076294,
+      "learning_rate": 9.993325917686318e-05,
+      "loss": 3.0679,
+      "step": 2255
+    },
+    {
+      "epoch": 30.093645484949832,
+      "grad_norm": 0.6975134015083313,
+      "learning_rate": 9.988876529477197e-05,
+      "loss": 2.9633,
+      "step": 2256
+    },
+    {
+      "epoch": 30.107023411371237,
+      "grad_norm": 0.7293218970298767,
+      "learning_rate": 9.984427141268076e-05,
+      "loss": 2.9374,
+      "step": 2257
+    },
+    {
+      "epoch": 30.120401337792643,
+      "grad_norm": 0.8367610573768616,
+      "learning_rate": 9.979977753058954e-05,
+      "loss": 2.8749,
+      "step": 2258
+    },
+    {
+      "epoch": 30.13377926421405,
+      "grad_norm": 0.6851879954338074,
+      "learning_rate": 9.975528364849834e-05,
+      "loss": 2.4921,
+      "step": 2259
+    },
+    {
+      "epoch": 30.14715719063545,
+      "grad_norm": 0.6916420459747314,
+      "learning_rate": 9.971078976640712e-05,
+      "loss": 2.6425,
+      "step": 2260
+    },
+    {
+      "epoch": 30.160535117056856,
+      "grad_norm": 0.6615001559257507,
+      "learning_rate": 9.966629588431591e-05,
+      "loss": 2.6696,
+      "step": 2261
+    },
+    {
+      "epoch": 30.17391304347826,
+      "grad_norm": 0.7402496933937073,
+      "learning_rate": 9.96218020022247e-05,
+      "loss": 2.7479,
+      "step": 2262
+    },
+    {
+      "epoch": 30.187290969899667,
+      "grad_norm": 0.6824116110801697,
+      "learning_rate": 9.957730812013348e-05,
+      "loss": 2.427,
+      "step": 2263
+    },
+    {
+      "epoch": 30.20066889632107,
+      "grad_norm": 0.6979151964187622,
+      "learning_rate": 9.953281423804227e-05,
+      "loss": 3.0677,
+      "step": 2264
+    },
+    {
+      "epoch": 30.214046822742475,
+      "grad_norm": 0.7541396617889404,
+      "learning_rate": 9.948832035595106e-05,
+      "loss": 2.9202,
+      "step": 2265
+    },
+    {
+      "epoch": 30.22742474916388,
+      "grad_norm": 0.671625554561615,
+      "learning_rate": 9.944382647385985e-05,
+      "loss": 2.6953,
+      "step": 2266
+    },
+    {
+      "epoch": 30.240802675585286,
+      "grad_norm": 0.6994143128395081,
+      "learning_rate": 9.939933259176863e-05,
+      "loss": 2.877,
+      "step": 2267
+    },
+    {
+      "epoch": 30.254180602006688,
+      "grad_norm": 0.7016029953956604,
+      "learning_rate": 9.935483870967742e-05,
+      "loss": 2.7883,
+      "step": 2268
+    },
+    {
+      "epoch": 30.267558528428093,
+      "grad_norm": 0.7190927267074585,
+      "learning_rate": 9.931034482758621e-05,
+      "loss": 2.9066,
+      "step": 2269
+    },
+    {
+      "epoch": 30.2809364548495,
+      "grad_norm": 0.8793424963951111,
+      "learning_rate": 9.9265850945495e-05,
+      "loss": 2.7433,
+      "step": 2270
+    },
+    {
+      "epoch": 30.294314381270905,
+      "grad_norm": 0.7914022207260132,
+      "learning_rate": 9.922135706340379e-05,
+      "loss": 2.6283,
+      "step": 2271
+    },
+    {
+      "epoch": 30.307692307692307,
+      "grad_norm": 0.7672899961471558,
+      "learning_rate": 9.917686318131257e-05,
+      "loss": 3.0767,
+      "step": 2272
+    },
+    {
+      "epoch": 30.321070234113712,
+      "grad_norm": 0.7170990109443665,
+      "learning_rate": 9.913236929922136e-05,
+      "loss": 2.9666,
+      "step": 2273
+    },
+    {
+      "epoch": 30.334448160535118,
+      "grad_norm": 0.7215063571929932,
+      "learning_rate": 9.908787541713015e-05,
+      "loss": 2.9646,
+      "step": 2274
+    },
+    {
+      "epoch": 30.347826086956523,
+      "grad_norm": 0.6509780287742615,
+      "learning_rate": 9.904338153503893e-05,
+      "loss": 2.9134,
+      "step": 2275
+    },
+    {
+      "epoch": 30.361204013377925,
+      "grad_norm": 0.6361657977104187,
+      "learning_rate": 9.899888765294773e-05,
+      "loss": 2.6955,
+      "step": 2276
+    },
+    {
+      "epoch": 30.37458193979933,
+      "grad_norm": 0.6722748279571533,
+      "learning_rate": 9.895439377085651e-05,
+      "loss": 3.1694,
+      "step": 2277
+    },
+    {
+      "epoch": 30.387959866220736,
+      "grad_norm": 0.796759307384491,
+      "learning_rate": 9.89098998887653e-05,
+      "loss": 2.9863,
+      "step": 2278
+    },
+    {
+      "epoch": 30.401337792642142,
+      "grad_norm": 0.8303789496421814,
+      "learning_rate": 9.886540600667409e-05,
+      "loss": 2.6192,
+      "step": 2279
+    },
+    {
+      "epoch": 30.414715719063544,
+      "grad_norm": 0.7565797567367554,
+      "learning_rate": 9.882091212458287e-05,
+      "loss": 2.8709,
+      "step": 2280
+    },
+    {
+      "epoch": 30.42809364548495,
+      "grad_norm": 0.7000163197517395,
+      "learning_rate": 9.877641824249167e-05,
+      "loss": 2.9017,
+      "step": 2281
+    },
+    {
+      "epoch": 30.441471571906355,
+      "grad_norm": 0.7144603729248047,
+      "learning_rate": 9.873192436040045e-05,
+      "loss": 2.9685,
+      "step": 2282
+    },
+    {
+      "epoch": 30.45484949832776,
+      "grad_norm": 0.8616060018539429,
+      "learning_rate": 9.868743047830923e-05,
+      "loss": 2.554,
+      "step": 2283
+    },
+    {
+      "epoch": 30.468227424749163,
+      "grad_norm": 0.6722779273986816,
+      "learning_rate": 9.864293659621803e-05,
+      "loss": 2.7132,
+      "step": 2284
+    },
+    {
+      "epoch": 30.48160535117057,
+      "grad_norm": 0.8574376702308655,
+      "learning_rate": 9.859844271412681e-05,
+      "loss": 2.8295,
+      "step": 2285
+    },
+    {
+      "epoch": 30.494983277591974,
+      "grad_norm": 0.6491283178329468,
+      "learning_rate": 9.855394883203561e-05,
+      "loss": 2.6042,
+      "step": 2286
+    },
+    {
+      "epoch": 30.50836120401338,
+      "grad_norm": 0.677865743637085,
+      "learning_rate": 9.850945494994439e-05,
+      "loss": 2.9783,
+      "step": 2287
+    },
+    {
+      "epoch": 30.52173913043478,
+      "grad_norm": 0.6842278242111206,
+      "learning_rate": 9.846496106785317e-05,
+      "loss": 2.8781,
+      "step": 2288
+    },
+    {
+      "epoch": 30.535117056856187,
+      "grad_norm": 0.7441837787628174,
+      "learning_rate": 9.842046718576197e-05,
+      "loss": 2.73,
+      "step": 2289
+    },
+    {
+      "epoch": 30.548494983277592,
+      "grad_norm": 0.8213462233543396,
+      "learning_rate": 9.837597330367075e-05,
+      "loss": 3.0742,
+      "step": 2290
+    },
+    {
+      "epoch": 30.561872909698998,
+      "grad_norm": 0.7320848703384399,
+      "learning_rate": 9.833147942157955e-05,
+      "loss": 3.0144,
+      "step": 2291
+    },
+    {
+      "epoch": 30.5752508361204,
+      "grad_norm": 0.6924611330032349,
+      "learning_rate": 9.828698553948833e-05,
+      "loss": 2.825,
+      "step": 2292
+    },
+    {
+      "epoch": 30.588628762541806,
+      "grad_norm": 0.6979595422744751,
+      "learning_rate": 9.824249165739711e-05,
+      "loss": 2.5886,
+      "step": 2293
+    },
+    {
+      "epoch": 30.60200668896321,
+      "grad_norm": 0.772278368473053,
+      "learning_rate": 9.819799777530591e-05,
+      "loss": 2.9113,
+      "step": 2294
+    },
+    {
+      "epoch": 30.615384615384617,
+      "grad_norm": 0.7430849075317383,
+      "learning_rate": 9.815350389321468e-05,
+      "loss": 2.8571,
+      "step": 2295
+    },
+    {
+      "epoch": 30.62876254180602,
+      "grad_norm": 0.7554115056991577,
+      "learning_rate": 9.810901001112347e-05,
+      "loss": 2.8272,
+      "step": 2296
+    },
+    {
+      "epoch": 30.642140468227424,
+      "grad_norm": 0.8022170662879944,
+      "learning_rate": 9.806451612903226e-05,
+      "loss": 2.8113,
+      "step": 2297
+    },
+    {
+      "epoch": 30.65551839464883,
+      "grad_norm": 0.7076969146728516,
+      "learning_rate": 9.802002224694105e-05,
+      "loss": 2.8282,
+      "step": 2298
+    },
+    {
+      "epoch": 30.668896321070235,
+      "grad_norm": 0.8672876954078674,
+      "learning_rate": 9.797552836484983e-05,
+      "loss": 2.8059,
+      "step": 2299
+    },
+    {
+      "epoch": 30.682274247491637,
+      "grad_norm": 0.7063407301902771,
+      "learning_rate": 9.793103448275862e-05,
+      "loss": 3.0516,
+      "step": 2300
+    },
+    {
+      "epoch": 30.695652173913043,
+      "grad_norm": 0.6833634376525879,
+      "learning_rate": 9.788654060066741e-05,
+      "loss": 3.0762,
+      "step": 2301
+    },
+    {
+      "epoch": 30.70903010033445,
+      "grad_norm": 0.7486119270324707,
+      "learning_rate": 9.78420467185762e-05,
+      "loss": 2.7423,
+      "step": 2302
+    },
+    {
+      "epoch": 30.722408026755854,
+      "grad_norm": 0.7409520149230957,
+      "learning_rate": 9.779755283648499e-05,
+      "loss": 2.9579,
+      "step": 2303
+    },
+    {
+      "epoch": 30.735785953177256,
+      "grad_norm": 0.7436200380325317,
+      "learning_rate": 9.775305895439377e-05,
+      "loss": 2.796,
+      "step": 2304
+    },
+    {
+      "epoch": 30.74916387959866,
+      "grad_norm": 0.7009103298187256,
+      "learning_rate": 9.770856507230256e-05,
+      "loss": 2.9099,
+      "step": 2305
+    },
+    {
+      "epoch": 30.762541806020067,
+      "grad_norm": 0.7946734428405762,
+      "learning_rate": 9.766407119021135e-05,
+      "loss": 3.014,
+      "step": 2306
+    },
+    {
+      "epoch": 30.775919732441473,
+      "grad_norm": 0.7147461175918579,
+      "learning_rate": 9.761957730812013e-05,
+      "loss": 2.5407,
+      "step": 2307
+    },
+    {
+      "epoch": 30.789297658862875,
+      "grad_norm": 0.7347055673599243,
+      "learning_rate": 9.757508342602893e-05,
+      "loss": 3.0734,
+      "step": 2308
+    },
+    {
+      "epoch": 30.80267558528428,
+      "grad_norm": 0.7037932872772217,
+      "learning_rate": 9.753058954393771e-05,
+      "loss": 2.8854,
+      "step": 2309
+    },
+    {
+      "epoch": 30.816053511705686,
+      "grad_norm": 0.6475211977958679,
+      "learning_rate": 9.74860956618465e-05,
+      "loss": 2.8844,
+      "step": 2310
+    },
+    {
+      "epoch": 30.82943143812709,
+      "grad_norm": 0.7679370045661926,
+      "learning_rate": 9.744160177975529e-05,
+      "loss": 2.8667,
+      "step": 2311
+    },
+    {
+      "epoch": 30.842809364548494,
+      "grad_norm": 0.7334820032119751,
+      "learning_rate": 9.739710789766407e-05,
+      "loss": 2.8936,
+      "step": 2312
+    },
+    {
+      "epoch": 30.8561872909699,
+      "grad_norm": 0.6336019039154053,
+      "learning_rate": 9.735261401557287e-05,
+      "loss": 2.7476,
+      "step": 2313
+    },
+    {
+      "epoch": 30.869565217391305,
+      "grad_norm": 0.6566322445869446,
+      "learning_rate": 9.730812013348165e-05,
+      "loss": 2.8259,
+      "step": 2314
+    },
+    {
+      "epoch": 30.88294314381271,
+      "grad_norm": 0.6635501384735107,
+      "learning_rate": 9.726362625139043e-05,
+      "loss": 2.7668,
+      "step": 2315
+    },
+    {
+      "epoch": 30.896321070234112,
+      "grad_norm": 0.7602748274803162,
+      "learning_rate": 9.721913236929923e-05,
+      "loss": 2.8645,
+      "step": 2316
+    },
+    {
+      "epoch": 30.909698996655518,
+      "grad_norm": 0.8196545243263245,
+      "learning_rate": 9.717463848720801e-05,
+      "loss": 2.8413,
+      "step": 2317
+    },
+    {
+      "epoch": 30.923076923076923,
+      "grad_norm": 0.6762588620185852,
+      "learning_rate": 9.713014460511681e-05,
+      "loss": 2.6832,
+      "step": 2318
+    },
+    {
+      "epoch": 30.93645484949833,
+      "grad_norm": 0.665920078754425,
+      "learning_rate": 9.708565072302559e-05,
+      "loss": 3.0145,
+      "step": 2319
+    },
+    {
+      "epoch": 30.94983277591973,
+      "grad_norm": 0.6335712671279907,
+      "learning_rate": 9.704115684093437e-05,
+      "loss": 2.9063,
+      "step": 2320
+    },
+    {
+      "epoch": 30.963210702341136,
+      "grad_norm": 0.7016931176185608,
+      "learning_rate": 9.699666295884317e-05,
+      "loss": 2.7531,
+      "step": 2321
+    },
+    {
+      "epoch": 30.976588628762542,
+      "grad_norm": 0.8611680865287781,
+      "learning_rate": 9.695216907675195e-05,
+      "loss": 2.9685,
+      "step": 2322
+    },
+    {
+      "epoch": 30.989966555183948,
+      "grad_norm": 0.6724279522895813,
+      "learning_rate": 9.690767519466075e-05,
+      "loss": 2.8643,
+      "step": 2323
+    },
+    {
+      "epoch": 31.0,
+      "grad_norm": 0.7072698473930359,
+      "learning_rate": 9.686318131256953e-05,
+      "loss": 2.8678,
+      "step": 2324
+    },
+    {
+      "epoch": 31.013377926421406,
+      "grad_norm": 0.7127351760864258,
+      "learning_rate": 9.681868743047831e-05,
+      "loss": 2.7762,
+      "step": 2325
+    },
+    {
+      "epoch": 31.02675585284281,
+      "grad_norm": 0.6769905686378479,
+      "learning_rate": 9.677419354838711e-05,
+      "loss": 2.3606,
+      "step": 2326
+    },
+    {
+      "epoch": 31.040133779264213,
+      "grad_norm": 0.6301751732826233,
+      "learning_rate": 9.672969966629589e-05,
+      "loss": 2.8011,
+      "step": 2327
+    },
+    {
+      "epoch": 31.05351170568562,
+      "grad_norm": 0.6441762447357178,
+      "learning_rate": 9.668520578420469e-05,
+      "loss": 2.902,
+      "step": 2328
+    },
+    {
+      "epoch": 31.066889632107024,
+      "grad_norm": 0.6481513381004333,
+      "learning_rate": 9.664071190211347e-05,
+      "loss": 2.7218,
+      "step": 2329
+    },
+    {
+      "epoch": 31.08026755852843,
+      "grad_norm": 0.5800172686576843,
+      "learning_rate": 9.659621802002225e-05,
+      "loss": 2.6825,
+      "step": 2330
+    },
+    {
+      "epoch": 31.093645484949832,
+      "grad_norm": 0.574657142162323,
+      "learning_rate": 9.655172413793105e-05,
+      "loss": 3.0232,
+      "step": 2331
+    },
+    {
+      "epoch": 31.107023411371237,
+      "grad_norm": 0.7543913125991821,
+      "learning_rate": 9.650723025583983e-05,
+      "loss": 2.8385,
+      "step": 2332
+    },
+    {
+      "epoch": 31.120401337792643,
+      "grad_norm": 0.5946618914604187,
+      "learning_rate": 9.646273637374862e-05,
+      "loss": 2.9112,
+      "step": 2333
+    },
+    {
+      "epoch": 31.13377926421405,
+      "grad_norm": 0.6467399001121521,
+      "learning_rate": 9.641824249165741e-05,
+      "loss": 2.6113,
+      "step": 2334
+    },
+    {
+      "epoch": 31.14715719063545,
+      "grad_norm": 0.6605884432792664,
+      "learning_rate": 9.637374860956619e-05,
+      "loss": 2.6679,
+      "step": 2335
+    },
+    {
+      "epoch": 31.160535117056856,
+      "grad_norm": 0.6213564276695251,
+      "learning_rate": 9.632925472747497e-05,
+      "loss": 2.7759,
+      "step": 2336
+    },
+    {
+      "epoch": 31.17391304347826,
+      "grad_norm": 0.5951142907142639,
+      "learning_rate": 9.628476084538375e-05,
+      "loss": 2.7004,
+      "step": 2337
+    },
+    {
+      "epoch": 31.187290969899667,
+      "grad_norm": 0.7088178396224976,
+      "learning_rate": 9.624026696329255e-05,
+      "loss": 2.6991,
+      "step": 2338
+    },
+    {
+      "epoch": 31.20066889632107,
+      "grad_norm": 0.6114148497581482,
+      "learning_rate": 9.619577308120133e-05,
+      "loss": 2.8156,
+      "step": 2339
+    },
+    {
+      "epoch": 31.214046822742475,
+      "grad_norm": 0.6434885859489441,
+      "learning_rate": 9.615127919911013e-05,
+      "loss": 2.7579,
+      "step": 2340
+    },
+    {
+      "epoch": 31.22742474916388,
+      "grad_norm": 0.6285985708236694,
+      "learning_rate": 9.610678531701891e-05,
+      "loss": 2.748,
+      "step": 2341
+    },
+    {
+      "epoch": 31.240802675585286,
+      "grad_norm": 0.6324411630630493,
+      "learning_rate": 9.60622914349277e-05,
+      "loss": 3.0152,
+      "step": 2342
+    },
+    {
+      "epoch": 31.254180602006688,
+      "grad_norm": 0.6331474781036377,
+      "learning_rate": 9.601779755283649e-05,
+      "loss": 2.6819,
+      "step": 2343
+    },
+    {
+      "epoch": 31.267558528428093,
+      "grad_norm": 0.6276957392692566,
+      "learning_rate": 9.597330367074527e-05,
+      "loss": 2.8349,
+      "step": 2344
+    },
+    {
+      "epoch": 31.2809364548495,
+      "grad_norm": 0.6469705104827881,
+      "learning_rate": 9.592880978865407e-05,
+      "loss": 2.9038,
+      "step": 2345
+    },
+    {
+      "epoch": 31.294314381270905,
+      "grad_norm": 0.6182751059532166,
+      "learning_rate": 9.588431590656285e-05,
+      "loss": 2.6157,
+      "step": 2346
+    },
+    {
+      "epoch": 31.307692307692307,
+      "grad_norm": 0.6365599632263184,
+      "learning_rate": 9.583982202447163e-05,
+      "loss": 2.6948,
+      "step": 2347
+    },
+    {
+      "epoch": 31.321070234113712,
+      "grad_norm": 0.6663181781768799,
+      "learning_rate": 9.579532814238043e-05,
+      "loss": 2.7856,
+      "step": 2348
+    },
+    {
+      "epoch": 31.334448160535118,
+      "grad_norm": 0.6316637396812439,
+      "learning_rate": 9.575083426028921e-05,
+      "loss": 2.7797,
+      "step": 2349
+    },
+    {
+      "epoch": 31.347826086956523,
+      "grad_norm": 0.6328778862953186,
+      "learning_rate": 9.570634037819801e-05,
+      "loss": 2.856,
+      "step": 2350
+    },
+    {
+      "epoch": 31.361204013377925,
+      "grad_norm": 0.6759666204452515,
+      "learning_rate": 9.566184649610679e-05,
+      "loss": 2.8553,
+      "step": 2351
+    },
+    {
+      "epoch": 31.37458193979933,
+      "grad_norm": 0.6212480664253235,
+      "learning_rate": 9.561735261401557e-05,
+      "loss": 2.9275,
+      "step": 2352
+    },
+    {
+      "epoch": 31.387959866220736,
+      "grad_norm": 0.6709821224212646,
+      "learning_rate": 9.557285873192437e-05,
+      "loss": 2.6805,
+      "step": 2353
+    },
+    {
+      "epoch": 31.401337792642142,
+      "grad_norm": 0.883128821849823,
+      "learning_rate": 9.552836484983315e-05,
+      "loss": 2.9176,
+      "step": 2354
+    },
+    {
+      "epoch": 31.414715719063544,
+      "grad_norm": 0.6021890640258789,
+      "learning_rate": 9.548387096774195e-05,
+      "loss": 2.9855,
+      "step": 2355
+    },
+    {
+      "epoch": 31.42809364548495,
+      "grad_norm": 0.6398679614067078,
+      "learning_rate": 9.543937708565073e-05,
+      "loss": 2.6652,
+      "step": 2356
+    },
+    {
+      "epoch": 31.441471571906355,
+      "grad_norm": 0.6757022738456726,
+      "learning_rate": 9.539488320355951e-05,
+      "loss": 2.7315,
+      "step": 2357
+    },
+    {
+      "epoch": 31.45484949832776,
+      "grad_norm": 0.6144642233848572,
+      "learning_rate": 9.53503893214683e-05,
+      "loss": 2.9645,
+      "step": 2358
+    },
+    {
+      "epoch": 31.468227424749163,
+      "grad_norm": 0.6620282530784607,
+      "learning_rate": 9.530589543937709e-05,
+      "loss": 2.6869,
+      "step": 2359
+    },
+    {
+      "epoch": 31.48160535117057,
+      "grad_norm": 0.7360846996307373,
+      "learning_rate": 9.526140155728588e-05,
+      "loss": 2.8083,
+      "step": 2360
+    },
+    {
+      "epoch": 31.494983277591974,
+      "grad_norm": 0.6184893250465393,
+      "learning_rate": 9.521690767519467e-05,
+      "loss": 2.9699,
+      "step": 2361
+    },
+    {
+      "epoch": 31.50836120401338,
+      "grad_norm": 0.5998436808586121,
+      "learning_rate": 9.517241379310345e-05,
+      "loss": 2.6591,
+      "step": 2362
+    },
+    {
+      "epoch": 31.52173913043478,
+      "grad_norm": 0.6120555996894836,
+      "learning_rate": 9.512791991101225e-05,
+      "loss": 3.0051,
+      "step": 2363
+    },
+    {
+      "epoch": 31.535117056856187,
+      "grad_norm": 0.6484112739562988,
+      "learning_rate": 9.508342602892103e-05,
+      "loss": 2.8998,
+      "step": 2364
+    },
+    {
+      "epoch": 31.548494983277592,
+      "grad_norm": 0.7333430051803589,
+      "learning_rate": 9.503893214682982e-05,
+      "loss": 2.8552,
+      "step": 2365
+    },
+    {
+      "epoch": 31.561872909698998,
+      "grad_norm": 0.6599973440170288,
+      "learning_rate": 9.49944382647386e-05,
+      "loss": 2.8119,
+      "step": 2366
+    },
+    {
+      "epoch": 31.5752508361204,
+      "grad_norm": 0.6654923558235168,
+      "learning_rate": 9.494994438264739e-05,
+      "loss": 2.6518,
+      "step": 2367
+    },
+    {
+      "epoch": 31.588628762541806,
+      "grad_norm": 0.6521298289299011,
+      "learning_rate": 9.490545050055618e-05,
+      "loss": 2.9758,
+      "step": 2368
+    },
+    {
+      "epoch": 31.60200668896321,
+      "grad_norm": 0.6431354284286499,
+      "learning_rate": 9.486095661846497e-05,
+      "loss": 2.7054,
+      "step": 2369
+    },
+    {
+      "epoch": 31.615384615384617,
+      "grad_norm": 0.7166205644607544,
+      "learning_rate": 9.481646273637376e-05,
+      "loss": 2.8292,
+      "step": 2370
+    },
+    {
+      "epoch": 31.62876254180602,
+      "grad_norm": 0.6184803247451782,
+      "learning_rate": 9.477196885428255e-05,
+      "loss": 2.8888,
+      "step": 2371
+    },
+    {
+      "epoch": 31.642140468227424,
+      "grad_norm": 0.6148157119750977,
+      "learning_rate": 9.472747497219133e-05,
+      "loss": 3.0765,
+      "step": 2372
+    },
+    {
+      "epoch": 31.65551839464883,
+      "grad_norm": 0.5624946355819702,
+      "learning_rate": 9.468298109010012e-05,
+      "loss": 2.4355,
+      "step": 2373
+    },
+    {
+      "epoch": 31.668896321070235,
+      "grad_norm": 0.6958364844322205,
+      "learning_rate": 9.46384872080089e-05,
+      "loss": 2.7913,
+      "step": 2374
+    },
+    {
+      "epoch": 31.682274247491637,
+      "grad_norm": 0.5998165011405945,
+      "learning_rate": 9.45939933259177e-05,
+      "loss": 2.7362,
+      "step": 2375
+    },
+    {
+      "epoch": 31.695652173913043,
+      "grad_norm": 0.7454515695571899,
+      "learning_rate": 9.454949944382648e-05,
+      "loss": 2.6624,
+      "step": 2376
+    },
+    {
+      "epoch": 31.70903010033445,
+      "grad_norm": 0.7100658416748047,
+      "learning_rate": 9.450500556173527e-05,
+      "loss": 3.0165,
+      "step": 2377
+    },
+    {
+      "epoch": 31.722408026755854,
+      "grad_norm": 0.6329564452171326,
+      "learning_rate": 9.446051167964405e-05,
+      "loss": 2.6719,
+      "step": 2378
+    },
+    {
+      "epoch": 31.735785953177256,
+      "grad_norm": 0.6892597675323486,
+      "learning_rate": 9.441601779755283e-05,
+      "loss": 2.9397,
+      "step": 2379
+    },
+    {
+      "epoch": 31.74916387959866,
+      "grad_norm": 0.7385604381561279,
+      "learning_rate": 9.437152391546163e-05,
+      "loss": 2.8576,
+      "step": 2380
+    },
+    {
+      "epoch": 31.762541806020067,
+      "grad_norm": 0.6018693447113037,
+      "learning_rate": 9.432703003337041e-05,
+      "loss": 2.6954,
+      "step": 2381
+    },
+    {
+      "epoch": 31.775919732441473,
+      "grad_norm": 0.8231506943702698,
+      "learning_rate": 9.42825361512792e-05,
+      "loss": 2.6043,
+      "step": 2382
+    },
+    {
+      "epoch": 31.789297658862875,
+      "grad_norm": 0.6664673686027527,
+      "learning_rate": 9.423804226918799e-05,
+      "loss": 2.6166,
+      "step": 2383
+    },
+    {
+      "epoch": 31.80267558528428,
+      "grad_norm": 0.6987932324409485,
+      "learning_rate": 9.419354838709677e-05,
+      "loss": 2.8913,
+      "step": 2384
+    },
+    {
+      "epoch": 31.816053511705686,
+      "grad_norm": 0.6309788823127747,
+      "learning_rate": 9.414905450500557e-05,
+      "loss": 2.9431,
+      "step": 2385
+    },
+    {
+      "epoch": 31.82943143812709,
+      "grad_norm": 0.6008022427558899,
+      "learning_rate": 9.410456062291435e-05,
+      "loss": 2.8104,
+      "step": 2386
+    },
+    {
+      "epoch": 31.842809364548494,
+      "grad_norm": 0.5797815918922424,
+      "learning_rate": 9.406006674082315e-05,
+      "loss": 2.685,
+      "step": 2387
+    },
+    {
+      "epoch": 31.8561872909699,
+      "grad_norm": 0.6194100379943848,
+      "learning_rate": 9.401557285873193e-05,
+      "loss": 2.9663,
+      "step": 2388
+    },
+    {
+      "epoch": 31.869565217391305,
+      "grad_norm": 0.633310079574585,
+      "learning_rate": 9.397107897664071e-05,
+      "loss": 2.9512,
+      "step": 2389
+    },
+    {
+      "epoch": 31.88294314381271,
+      "grad_norm": 0.6811971664428711,
+      "learning_rate": 9.39265850945495e-05,
+      "loss": 2.8803,
+      "step": 2390
+    },
+    {
+      "epoch": 31.896321070234112,
+      "grad_norm": 0.6683096885681152,
+      "learning_rate": 9.388209121245829e-05,
+      "loss": 2.8334,
+      "step": 2391
+    },
+    {
+      "epoch": 31.909698996655518,
+      "grad_norm": 0.6561327576637268,
+      "learning_rate": 9.383759733036708e-05,
+      "loss": 2.7997,
+      "step": 2392
+    },
+    {
+      "epoch": 31.923076923076923,
+      "grad_norm": 0.7540359497070312,
+      "learning_rate": 9.379310344827587e-05,
+      "loss": 2.9634,
+      "step": 2393
+    },
+    {
+      "epoch": 31.93645484949833,
+      "grad_norm": 0.6072854399681091,
+      "learning_rate": 9.374860956618465e-05,
+      "loss": 2.8317,
+      "step": 2394
+    },
+    {
+      "epoch": 31.94983277591973,
+      "grad_norm": 0.6614176630973816,
+      "learning_rate": 9.370411568409344e-05,
+      "loss": 2.9773,
+      "step": 2395
+    },
+    {
+      "epoch": 31.963210702341136,
+      "grad_norm": 0.6928074955940247,
+      "learning_rate": 9.365962180200223e-05,
+      "loss": 2.8665,
+      "step": 2396
+    },
+    {
+      "epoch": 31.976588628762542,
+      "grad_norm": 0.644250750541687,
+      "learning_rate": 9.361512791991102e-05,
+      "loss": 2.7206,
+      "step": 2397
+    },
+    {
+      "epoch": 31.989966555183948,
+      "grad_norm": 0.6529290080070496,
+      "learning_rate": 9.35706340378198e-05,
+      "loss": 2.7764,
+      "step": 2398
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 0.7077373266220093,
+      "learning_rate": 9.352614015572859e-05,
+      "loss": 2.9435,
+      "step": 2399
+    },
+    {
+      "epoch": 32.013377926421406,
+      "grad_norm": 0.614982545375824,
+      "learning_rate": 9.348164627363738e-05,
+      "loss": 2.9358,
+      "step": 2400
+    },
+    {
+      "epoch": 32.02675585284281,
+      "grad_norm": 0.6013137102127075,
+      "learning_rate": 9.343715239154617e-05,
+      "loss": 2.7346,
+      "step": 2401
+    },
+    {
+      "epoch": 32.04013377926422,
+      "grad_norm": 0.653290867805481,
+      "learning_rate": 9.339265850945496e-05,
+      "loss": 2.9841,
+      "step": 2402
+    },
+    {
+      "epoch": 32.05351170568562,
+      "grad_norm": 0.6256197094917297,
+      "learning_rate": 9.334816462736374e-05,
+      "loss": 2.8465,
+      "step": 2403
+    },
+    {
+      "epoch": 32.06688963210702,
+      "grad_norm": 0.5798116326332092,
+      "learning_rate": 9.330367074527253e-05,
+      "loss": 2.5661,
+      "step": 2404
+    },
+    {
+      "epoch": 32.080267558528426,
+      "grad_norm": 0.6489708423614502,
+      "learning_rate": 9.325917686318132e-05,
+      "loss": 2.8186,
+      "step": 2405
+    },
+    {
+      "epoch": 32.09364548494983,
+      "grad_norm": 0.6225152015686035,
+      "learning_rate": 9.32146829810901e-05,
+      "loss": 2.6598,
+      "step": 2406
+    },
+    {
+      "epoch": 32.10702341137124,
+      "grad_norm": 0.6490108966827393,
+      "learning_rate": 9.31701890989989e-05,
+      "loss": 2.7589,
+      "step": 2407
+    },
+    {
+      "epoch": 32.12040133779264,
+      "grad_norm": 0.5600370764732361,
+      "learning_rate": 9.312569521690768e-05,
+      "loss": 2.6324,
+      "step": 2408
+    },
+    {
+      "epoch": 32.13377926421405,
+      "grad_norm": 0.6141876578330994,
+      "learning_rate": 9.308120133481647e-05,
+      "loss": 2.7371,
+      "step": 2409
+    },
+    {
+      "epoch": 32.147157190635454,
+      "grad_norm": 0.5927343368530273,
+      "learning_rate": 9.303670745272526e-05,
+      "loss": 2.9498,
+      "step": 2410
+    },
+    {
+      "epoch": 32.16053511705686,
+      "grad_norm": 0.686123788356781,
+      "learning_rate": 9.299221357063404e-05,
+      "loss": 2.698,
+      "step": 2411
+    },
+    {
+      "epoch": 32.17391304347826,
+      "grad_norm": 0.6357244253158569,
+      "learning_rate": 9.294771968854284e-05,
+      "loss": 2.7431,
+      "step": 2412
+    },
+    {
+      "epoch": 32.187290969899664,
+      "grad_norm": 0.6475897431373596,
+      "learning_rate": 9.290322580645162e-05,
+      "loss": 2.7808,
+      "step": 2413
+    },
+    {
+      "epoch": 32.20066889632107,
+      "grad_norm": 0.5825409293174744,
+      "learning_rate": 9.28587319243604e-05,
+      "loss": 2.6867,
+      "step": 2414
+    },
+    {
+      "epoch": 32.214046822742475,
+      "grad_norm": 0.6462129354476929,
+      "learning_rate": 9.28142380422692e-05,
+      "loss": 2.5075,
+      "step": 2415
+    },
+    {
+      "epoch": 32.22742474916388,
+      "grad_norm": 0.6217963099479675,
+      "learning_rate": 9.276974416017798e-05,
+      "loss": 3.1276,
+      "step": 2416
+    },
+    {
+      "epoch": 32.240802675585286,
+      "grad_norm": 1.1910638809204102,
+      "learning_rate": 9.272525027808678e-05,
+      "loss": 2.7246,
+      "step": 2417
+    },
+    {
+      "epoch": 32.25418060200669,
+      "grad_norm": 0.6929856538772583,
+      "learning_rate": 9.268075639599556e-05,
+      "loss": 2.9576,
+      "step": 2418
+    },
+    {
+      "epoch": 32.2675585284281,
+      "grad_norm": 0.6331362128257751,
+      "learning_rate": 9.263626251390433e-05,
+      "loss": 2.7267,
+      "step": 2419
+    },
+    {
+      "epoch": 32.280936454849495,
+      "grad_norm": 0.6223586797714233,
+      "learning_rate": 9.259176863181313e-05,
+      "loss": 2.5904,
+      "step": 2420
+    },
+    {
+      "epoch": 32.2943143812709,
+      "grad_norm": 0.6478082537651062,
+      "learning_rate": 9.254727474972191e-05,
+      "loss": 2.7874,
+      "step": 2421
+    },
+    {
+      "epoch": 32.30769230769231,
+      "grad_norm": 0.6407091617584229,
+      "learning_rate": 9.25027808676307e-05,
+      "loss": 2.7926,
+      "step": 2422
+    },
+    {
+      "epoch": 32.32107023411371,
+      "grad_norm": 0.5770341753959656,
+      "learning_rate": 9.245828698553949e-05,
+      "loss": 2.9566,
+      "step": 2423
+    },
+    {
+      "epoch": 32.33444816053512,
+      "grad_norm": 0.6170133948326111,
+      "learning_rate": 9.241379310344827e-05,
+      "loss": 2.8071,
+      "step": 2424
+    },
+    {
+      "epoch": 32.34782608695652,
+      "grad_norm": 0.6636959910392761,
+      "learning_rate": 9.236929922135707e-05,
+      "loss": 2.8502,
+      "step": 2425
+    },
+    {
+      "epoch": 32.36120401337793,
+      "grad_norm": 0.5865075588226318,
+      "learning_rate": 9.232480533926585e-05,
+      "loss": 2.9213,
+      "step": 2426
+    },
+    {
+      "epoch": 32.374581939799334,
+      "grad_norm": 0.6461585760116577,
+      "learning_rate": 9.228031145717464e-05,
+      "loss": 2.7773,
+      "step": 2427
+    },
+    {
+      "epoch": 32.38795986622073,
+      "grad_norm": 0.6047717928886414,
+      "learning_rate": 9.223581757508343e-05,
+      "loss": 2.7859,
+      "step": 2428
+    },
+    {
+      "epoch": 32.40133779264214,
+      "grad_norm": 0.6110160946846008,
+      "learning_rate": 9.219132369299221e-05,
+      "loss": 3.0866,
+      "step": 2429
+    },
+    {
+      "epoch": 32.414715719063544,
+      "grad_norm": 0.6165122389793396,
+      "learning_rate": 9.2146829810901e-05,
+      "loss": 2.5561,
+      "step": 2430
+    },
+    {
+      "epoch": 32.42809364548495,
+      "grad_norm": 0.6536113023757935,
+      "learning_rate": 9.210233592880979e-05,
+      "loss": 2.6207,
+      "step": 2431
+    },
+    {
+      "epoch": 32.441471571906355,
+      "grad_norm": 0.6606733202934265,
+      "learning_rate": 9.205784204671858e-05,
+      "loss": 2.8492,
+      "step": 2432
+    },
+    {
+      "epoch": 32.45484949832776,
+      "grad_norm": 0.5955353379249573,
+      "learning_rate": 9.201334816462737e-05,
+      "loss": 2.517,
+      "step": 2433
+    },
+    {
+      "epoch": 32.468227424749166,
+      "grad_norm": 0.6918089985847473,
+      "learning_rate": 9.196885428253615e-05,
+      "loss": 2.7207,
+      "step": 2434
+    },
+    {
+      "epoch": 32.48160535117057,
+      "grad_norm": 0.6295514702796936,
+      "learning_rate": 9.192436040044494e-05,
+      "loss": 2.8519,
+      "step": 2435
+    },
+    {
+      "epoch": 32.49498327759197,
+      "grad_norm": 0.5654889345169067,
+      "learning_rate": 9.187986651835373e-05,
+      "loss": 2.6741,
+      "step": 2436
+    },
+    {
+      "epoch": 32.508361204013376,
+      "grad_norm": 0.6156261563301086,
+      "learning_rate": 9.183537263626252e-05,
+      "loss": 2.8508,
+      "step": 2437
+    },
+    {
+      "epoch": 32.52173913043478,
+      "grad_norm": 0.578255832195282,
+      "learning_rate": 9.17908787541713e-05,
+      "loss": 2.7162,
+      "step": 2438
+    },
+    {
+      "epoch": 32.53511705685619,
+      "grad_norm": 0.6073182225227356,
+      "learning_rate": 9.174638487208009e-05,
+      "loss": 2.9915,
+      "step": 2439
+    },
+    {
+      "epoch": 32.54849498327759,
+      "grad_norm": 0.6045039892196655,
+      "learning_rate": 9.170189098998888e-05,
+      "loss": 2.7507,
+      "step": 2440
+    },
+    {
+      "epoch": 32.561872909699,
+      "grad_norm": 0.6131772398948669,
+      "learning_rate": 9.165739710789767e-05,
+      "loss": 2.8352,
+      "step": 2441
+    },
+    {
+      "epoch": 32.575250836120404,
+      "grad_norm": 0.5804582834243774,
+      "learning_rate": 9.161290322580646e-05,
+      "loss": 2.844,
+      "step": 2442
+    },
+    {
+      "epoch": 32.58862876254181,
+      "grad_norm": 0.6363027095794678,
+      "learning_rate": 9.156840934371524e-05,
+      "loss": 2.7884,
+      "step": 2443
+    },
+    {
+      "epoch": 32.60200668896321,
+      "grad_norm": 0.599229097366333,
+      "learning_rate": 9.152391546162403e-05,
+      "loss": 2.5051,
+      "step": 2444
+    },
+    {
+      "epoch": 32.61538461538461,
+      "grad_norm": 0.6206624507904053,
+      "learning_rate": 9.147942157953282e-05,
+      "loss": 2.9053,
+      "step": 2445
+    },
+    {
+      "epoch": 32.62876254180602,
+      "grad_norm": 0.6836562752723694,
+      "learning_rate": 9.14349276974416e-05,
+      "loss": 2.8332,
+      "step": 2446
+    },
+    {
+      "epoch": 32.642140468227424,
+      "grad_norm": 0.6113333702087402,
+      "learning_rate": 9.13904338153504e-05,
+      "loss": 2.8872,
+      "step": 2447
+    },
+    {
+      "epoch": 32.65551839464883,
+      "grad_norm": 0.5945793986320496,
+      "learning_rate": 9.134593993325918e-05,
+      "loss": 2.7071,
+      "step": 2448
+    },
+    {
+      "epoch": 32.668896321070235,
+      "grad_norm": 0.647544264793396,
+      "learning_rate": 9.130144605116796e-05,
+      "loss": 2.7016,
+      "step": 2449
+    },
+    {
+      "epoch": 32.68227424749164,
+      "grad_norm": 0.5891870856285095,
+      "learning_rate": 9.125695216907676e-05,
+      "loss": 2.7178,
+      "step": 2450
+    },
+    {
+      "epoch": 32.69565217391305,
+      "grad_norm": 0.5718060731887817,
+      "learning_rate": 9.121245828698554e-05,
+      "loss": 2.5202,
+      "step": 2451
+    },
+    {
+      "epoch": 32.709030100334445,
+      "grad_norm": 0.6582256555557251,
+      "learning_rate": 9.116796440489434e-05,
+      "loss": 2.9955,
+      "step": 2452
+    },
+    {
+      "epoch": 32.72240802675585,
+      "grad_norm": 0.6060263514518738,
+      "learning_rate": 9.112347052280312e-05,
+      "loss": 2.7279,
+      "step": 2453
+    },
+    {
+      "epoch": 32.735785953177256,
+      "grad_norm": 0.6045027375221252,
+      "learning_rate": 9.10789766407119e-05,
+      "loss": 2.7305,
+      "step": 2454
+    },
+    {
+      "epoch": 32.74916387959866,
+      "grad_norm": 0.5832952857017517,
+      "learning_rate": 9.10344827586207e-05,
+      "loss": 2.8703,
+      "step": 2455
+    },
+    {
+      "epoch": 32.76254180602007,
+      "grad_norm": 0.5899837613105774,
+      "learning_rate": 9.098998887652948e-05,
+      "loss": 2.5243,
+      "step": 2456
+    },
+    {
+      "epoch": 32.77591973244147,
+      "grad_norm": 0.6895177960395813,
+      "learning_rate": 9.094549499443828e-05,
+      "loss": 2.7974,
+      "step": 2457
+    },
+    {
+      "epoch": 32.78929765886288,
+      "grad_norm": 0.5995526909828186,
+      "learning_rate": 9.090100111234706e-05,
+      "loss": 2.6908,
+      "step": 2458
+    },
+    {
+      "epoch": 32.802675585284284,
+      "grad_norm": 0.592761754989624,
+      "learning_rate": 9.085650723025584e-05,
+      "loss": 2.7314,
+      "step": 2459
+    },
+    {
+      "epoch": 32.81605351170568,
+      "grad_norm": 0.5758498311042786,
+      "learning_rate": 9.081201334816463e-05,
+      "loss": 2.5282,
+      "step": 2460
+    },
+    {
+      "epoch": 32.82943143812709,
+      "grad_norm": 0.6093677282333374,
+      "learning_rate": 9.076751946607341e-05,
+      "loss": 2.9135,
+      "step": 2461
+    },
+    {
+      "epoch": 32.84280936454849,
+      "grad_norm": 0.617378830909729,
+      "learning_rate": 9.07230255839822e-05,
+      "loss": 2.6457,
+      "step": 2462
+    },
+    {
+      "epoch": 32.8561872909699,
+      "grad_norm": 0.5618733763694763,
+      "learning_rate": 9.067853170189099e-05,
+      "loss": 2.4018,
+      "step": 2463
+    },
+    {
+      "epoch": 32.869565217391305,
+      "grad_norm": 0.6228212714195251,
+      "learning_rate": 9.063403781979978e-05,
+      "loss": 2.7449,
+      "step": 2464
+    },
+    {
+      "epoch": 32.88294314381271,
+      "grad_norm": 0.6647076606750488,
+      "learning_rate": 9.058954393770856e-05,
+      "loss": 2.7208,
+      "step": 2465
+    },
+    {
+      "epoch": 32.896321070234116,
+      "grad_norm": 0.6458041071891785,
+      "learning_rate": 9.054505005561735e-05,
+      "loss": 2.9473,
+      "step": 2466
+    },
+    {
+      "epoch": 32.90969899665552,
+      "grad_norm": 0.5941137671470642,
+      "learning_rate": 9.050055617352614e-05,
+      "loss": 2.807,
+      "step": 2467
+    },
+    {
+      "epoch": 32.92307692307692,
+      "grad_norm": 0.7032221555709839,
+      "learning_rate": 9.045606229143493e-05,
+      "loss": 2.7777,
+      "step": 2468
+    },
+    {
+      "epoch": 32.936454849498325,
+      "grad_norm": 0.5803418755531311,
+      "learning_rate": 9.041156840934372e-05,
+      "loss": 2.7822,
+      "step": 2469
+    },
+    {
+      "epoch": 32.94983277591973,
+      "grad_norm": 0.5971820950508118,
+      "learning_rate": 9.03670745272525e-05,
+      "loss": 2.8719,
+      "step": 2470
+    },
+    {
+      "epoch": 32.96321070234114,
+      "grad_norm": 0.6287712454795837,
+      "learning_rate": 9.032258064516129e-05,
+      "loss": 2.9398,
+      "step": 2471
+    },
+    {
+      "epoch": 32.97658862876254,
+      "grad_norm": 0.5914057493209839,
+      "learning_rate": 9.027808676307008e-05,
+      "loss": 2.8867,
+      "step": 2472
+    },
+    {
+      "epoch": 32.98996655518395,
+      "grad_norm": 0.5863922834396362,
+      "learning_rate": 9.023359288097886e-05,
+      "loss": 2.8591,
+      "step": 2473
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 0.7483989000320435,
+      "learning_rate": 9.018909899888766e-05,
+      "loss": 2.6477,
+      "step": 2474
+    },
+    {
+      "epoch": 33.013377926421406,
+      "grad_norm": 0.6205896139144897,
+      "learning_rate": 9.014460511679644e-05,
+      "loss": 2.5337,
+      "step": 2475
+    },
+    {
+      "epoch": 33.02675585284281,
+      "grad_norm": 0.5723957419395447,
+      "learning_rate": 9.010011123470522e-05,
+      "loss": 2.5932,
+      "step": 2476
+    },
+    {
+      "epoch": 33.04013377926422,
+      "grad_norm": 0.5575253963470459,
+      "learning_rate": 9.005561735261402e-05,
+      "loss": 2.6566,
+      "step": 2477
+    },
+    {
+      "epoch": 33.05351170568562,
+      "grad_norm": 0.5518794655799866,
+      "learning_rate": 9.00111234705228e-05,
+      "loss": 2.6017,
+      "step": 2478
+    },
+    {
+      "epoch": 33.06688963210702,
+      "grad_norm": 0.5580832958221436,
+      "learning_rate": 8.99666295884316e-05,
+      "loss": 2.5346,
+      "step": 2479
+    },
+    {
+      "epoch": 33.080267558528426,
+      "grad_norm": 0.5831477046012878,
+      "learning_rate": 8.992213570634038e-05,
+      "loss": 2.4491,
+      "step": 2480
+    },
+    {
+      "epoch": 33.09364548494983,
+      "grad_norm": 0.5773186683654785,
+      "learning_rate": 8.987764182424916e-05,
+      "loss": 2.6822,
+      "step": 2481
+    },
+    {
+      "epoch": 33.10702341137124,
+      "grad_norm": 0.63813316822052,
+      "learning_rate": 8.983314794215796e-05,
+      "loss": 2.6537,
+      "step": 2482
+    },
+    {
+      "epoch": 33.12040133779264,
+      "grad_norm": 0.6081259250640869,
+      "learning_rate": 8.978865406006674e-05,
+      "loss": 2.6482,
+      "step": 2483
+    },
+    {
+      "epoch": 33.13377926421405,
+      "grad_norm": 0.5737481117248535,
+      "learning_rate": 8.974416017797554e-05,
+      "loss": 2.6858,
+      "step": 2484
+    },
+    {
+      "epoch": 33.147157190635454,
+      "grad_norm": 0.6300746202468872,
+      "learning_rate": 8.969966629588432e-05,
+      "loss": 2.7791,
+      "step": 2485
+    },
+    {
+      "epoch": 33.16053511705686,
+      "grad_norm": 0.6003954410552979,
+      "learning_rate": 8.96551724137931e-05,
+      "loss": 2.8088,
+      "step": 2486
+    },
+    {
+      "epoch": 33.17391304347826,
+      "grad_norm": 0.5618523955345154,
+      "learning_rate": 8.96106785317019e-05,
+      "loss": 2.6345,
+      "step": 2487
+    },
+    {
+      "epoch": 33.187290969899664,
+      "grad_norm": 0.6398965120315552,
+      "learning_rate": 8.956618464961068e-05,
+      "loss": 2.7798,
+      "step": 2488
+    },
+    {
+      "epoch": 33.20066889632107,
+      "grad_norm": 0.6220288276672363,
+      "learning_rate": 8.952169076751948e-05,
+      "loss": 2.8217,
+      "step": 2489
+    },
+    {
+      "epoch": 33.214046822742475,
+      "grad_norm": 0.6096678376197815,
+      "learning_rate": 8.947719688542826e-05,
+      "loss": 2.6255,
+      "step": 2490
+    },
+    {
+      "epoch": 33.22742474916388,
+      "grad_norm": 0.5997774600982666,
+      "learning_rate": 8.943270300333704e-05,
+      "loss": 2.8361,
+      "step": 2491
+    },
+    {
+      "epoch": 33.240802675585286,
+      "grad_norm": 0.6017264127731323,
+      "learning_rate": 8.938820912124584e-05,
+      "loss": 2.7003,
+      "step": 2492
+    },
+    {
+      "epoch": 33.25418060200669,
+      "grad_norm": 0.6213170289993286,
+      "learning_rate": 8.934371523915462e-05,
+      "loss": 2.8992,
+      "step": 2493
+    },
+    {
+      "epoch": 33.2675585284281,
+      "grad_norm": 0.6049240231513977,
+      "learning_rate": 8.929922135706342e-05,
+      "loss": 2.79,
+      "step": 2494
+    },
+    {
+      "epoch": 33.280936454849495,
+      "grad_norm": 0.6120469570159912,
+      "learning_rate": 8.92547274749722e-05,
+      "loss": 2.7695,
+      "step": 2495
+    },
+    {
+      "epoch": 33.2943143812709,
+      "grad_norm": 0.5764347314834595,
+      "learning_rate": 8.921023359288098e-05,
+      "loss": 2.5934,
+      "step": 2496
+    },
+    {
+      "epoch": 33.30769230769231,
+      "grad_norm": 0.5952526926994324,
+      "learning_rate": 8.916573971078978e-05,
+      "loss": 2.8907,
+      "step": 2497
+    },
+    {
+      "epoch": 33.32107023411371,
+      "grad_norm": 0.5542176961898804,
+      "learning_rate": 8.912124582869856e-05,
+      "loss": 2.6922,
+      "step": 2498
+    },
+    {
+      "epoch": 33.33444816053512,
+      "grad_norm": 0.622314453125,
+      "learning_rate": 8.907675194660736e-05,
+      "loss": 2.7322,
+      "step": 2499
+    },
+    {
+      "epoch": 33.34782608695652,
+      "grad_norm": 0.5831202864646912,
+      "learning_rate": 8.903225806451614e-05,
+      "loss": 2.766,
+      "step": 2500
+    },
+    {
+      "epoch": 33.36120401337793,
+      "grad_norm": 0.6198363900184631,
+      "learning_rate": 8.898776418242492e-05,
+      "loss": 3.0431,
+      "step": 2501
+    },
+    {
+      "epoch": 33.374581939799334,
+      "grad_norm": 0.6069219708442688,
+      "learning_rate": 8.89432703003337e-05,
+      "loss": 2.8924,
+      "step": 2502
+    },
+    {
+      "epoch": 33.38795986622073,
+      "grad_norm": 0.6229302287101746,
+      "learning_rate": 8.889877641824249e-05,
+      "loss": 2.807,
+      "step": 2503
+    },
+    {
+      "epoch": 33.40133779264214,
+      "grad_norm": 0.6178746819496155,
+      "learning_rate": 8.885428253615128e-05,
+      "loss": 2.7833,
+      "step": 2504
+    },
+    {
+      "epoch": 33.414715719063544,
+      "grad_norm": 0.5880036950111389,
+      "learning_rate": 8.880978865406006e-05,
+      "loss": 2.7688,
+      "step": 2505
+    },
+    {
+      "epoch": 33.42809364548495,
+      "grad_norm": 0.6083427667617798,
+      "learning_rate": 8.876529477196886e-05,
+      "loss": 2.5995,
+      "step": 2506
+    },
+    {
+      "epoch": 33.441471571906355,
+      "grad_norm": 0.5846229195594788,
+      "learning_rate": 8.872080088987764e-05,
+      "loss": 2.7198,
+      "step": 2507
+    },
+    {
+      "epoch": 33.45484949832776,
+      "grad_norm": 0.602645754814148,
+      "learning_rate": 8.867630700778642e-05,
+      "loss": 2.8395,
+      "step": 2508
+    },
+    {
+      "epoch": 33.468227424749166,
+      "grad_norm": 0.5731549859046936,
+      "learning_rate": 8.863181312569522e-05,
+      "loss": 2.6083,
+      "step": 2509
+    },
+    {
+      "epoch": 33.48160535117057,
+      "grad_norm": 0.6011414527893066,
+      "learning_rate": 8.8587319243604e-05,
+      "loss": 2.6036,
+      "step": 2510
+    },
+    {
+      "epoch": 33.49498327759197,
+      "grad_norm": 0.5945054292678833,
+      "learning_rate": 8.85428253615128e-05,
+      "loss": 2.8461,
+      "step": 2511
+    },
+    {
+      "epoch": 33.508361204013376,
+      "grad_norm": 0.6700778007507324,
+      "learning_rate": 8.849833147942158e-05,
+      "loss": 2.9723,
+      "step": 2512
+    },
+    {
+      "epoch": 33.52173913043478,
+      "grad_norm": 0.5661737322807312,
+      "learning_rate": 8.845383759733036e-05,
+      "loss": 2.5432,
+      "step": 2513
+    },
+    {
+      "epoch": 33.53511705685619,
+      "grad_norm": 0.615665078163147,
+      "learning_rate": 8.840934371523916e-05,
+      "loss": 2.5919,
+      "step": 2514
+    },
+    {
+      "epoch": 33.54849498327759,
+      "grad_norm": 0.6195909380912781,
+      "learning_rate": 8.836484983314794e-05,
+      "loss": 2.6273,
+      "step": 2515
+    },
+    {
+      "epoch": 33.561872909699,
+      "grad_norm": 0.6008119583129883,
+      "learning_rate": 8.832035595105674e-05,
+      "loss": 2.6099,
+      "step": 2516
+    },
+    {
+      "epoch": 33.575250836120404,
+      "grad_norm": 0.6525030732154846,
+      "learning_rate": 8.827586206896552e-05,
+      "loss": 2.7715,
+      "step": 2517
+    },
+    {
+      "epoch": 33.58862876254181,
+      "grad_norm": 0.6317800879478455,
+      "learning_rate": 8.82313681868743e-05,
+      "loss": 3.1264,
+      "step": 2518
+    },
+    {
+      "epoch": 33.60200668896321,
+      "grad_norm": 0.5794599056243896,
+      "learning_rate": 8.81868743047831e-05,
+      "loss": 2.5186,
+      "step": 2519
+    },
+    {
+      "epoch": 33.61538461538461,
+      "grad_norm": 0.633000373840332,
+      "learning_rate": 8.814238042269188e-05,
+      "loss": 2.7709,
+      "step": 2520
+    },
+    {
+      "epoch": 33.62876254180602,
+      "grad_norm": 0.6132048964500427,
+      "learning_rate": 8.809788654060068e-05,
+      "loss": 2.6311,
+      "step": 2521
+    },
+    {
+      "epoch": 33.642140468227424,
+      "grad_norm": 0.6357754468917847,
+      "learning_rate": 8.805339265850946e-05,
+      "loss": 2.8635,
+      "step": 2522
+    },
+    {
+      "epoch": 33.65551839464883,
+      "grad_norm": 0.5678791403770447,
+      "learning_rate": 8.800889877641824e-05,
+      "loss": 2.773,
+      "step": 2523
+    },
+    {
+      "epoch": 33.668896321070235,
+      "grad_norm": 0.5938114523887634,
+      "learning_rate": 8.796440489432704e-05,
+      "loss": 2.6978,
+      "step": 2524
+    },
+    {
+      "epoch": 33.68227424749164,
+      "grad_norm": 0.6006829142570496,
+      "learning_rate": 8.791991101223582e-05,
+      "loss": 2.8147,
+      "step": 2525
+    },
+    {
+      "epoch": 33.69565217391305,
+      "grad_norm": 0.6094152331352234,
+      "learning_rate": 8.787541713014462e-05,
+      "loss": 2.9597,
+      "step": 2526
+    },
+    {
+      "epoch": 33.709030100334445,
+      "grad_norm": 0.6261194944381714,
+      "learning_rate": 8.78309232480534e-05,
+      "loss": 2.7116,
+      "step": 2527
+    },
+    {
+      "epoch": 33.72240802675585,
+      "grad_norm": 0.6670746803283691,
+      "learning_rate": 8.778642936596218e-05,
+      "loss": 2.8481,
+      "step": 2528
+    },
+    {
+      "epoch": 33.735785953177256,
+      "grad_norm": 0.5976547598838806,
+      "learning_rate": 8.774193548387098e-05,
+      "loss": 2.7368,
+      "step": 2529
+    },
+    {
+      "epoch": 33.74916387959866,
+      "grad_norm": 0.5903379321098328,
+      "learning_rate": 8.769744160177976e-05,
+      "loss": 2.784,
+      "step": 2530
+    },
+    {
+      "epoch": 33.76254180602007,
+      "grad_norm": 0.604728102684021,
+      "learning_rate": 8.765294771968855e-05,
+      "loss": 2.7636,
+      "step": 2531
+    },
+    {
+      "epoch": 33.77591973244147,
+      "grad_norm": 0.6338250041007996,
+      "learning_rate": 8.760845383759734e-05,
+      "loss": 2.8111,
+      "step": 2532
+    },
+    {
+      "epoch": 33.78929765886288,
+      "grad_norm": 0.6076844334602356,
+      "learning_rate": 8.756395995550612e-05,
+      "loss": 2.7592,
+      "step": 2533
+    },
+    {
+      "epoch": 33.802675585284284,
+      "grad_norm": 0.6152723431587219,
+      "learning_rate": 8.751946607341492e-05,
+      "loss": 2.889,
+      "step": 2534
+    },
+    {
+      "epoch": 33.81605351170568,
+      "grad_norm": 0.5518035292625427,
+      "learning_rate": 8.74749721913237e-05,
+      "loss": 2.5681,
+      "step": 2535
+    },
+    {
+      "epoch": 33.82943143812709,
+      "grad_norm": 0.624555766582489,
+      "learning_rate": 8.74304783092325e-05,
+      "loss": 2.6651,
+      "step": 2536
+    },
+    {
+      "epoch": 33.84280936454849,
+      "grad_norm": 0.6052064895629883,
+      "learning_rate": 8.738598442714128e-05,
+      "loss": 2.6292,
+      "step": 2537
+    },
+    {
+      "epoch": 33.8561872909699,
+      "grad_norm": 0.6039239168167114,
+      "learning_rate": 8.734149054505006e-05,
+      "loss": 2.9519,
+      "step": 2538
+    },
+    {
+      "epoch": 33.869565217391305,
+      "grad_norm": 0.633802056312561,
+      "learning_rate": 8.729699666295885e-05,
+      "loss": 2.8169,
+      "step": 2539
+    },
+    {
+      "epoch": 33.88294314381271,
+      "grad_norm": 0.6555929780006409,
+      "learning_rate": 8.725250278086764e-05,
+      "loss": 2.7988,
+      "step": 2540
+    },
+    {
+      "epoch": 33.896321070234116,
+      "grad_norm": 0.6146313548088074,
+      "learning_rate": 8.720800889877643e-05,
+      "loss": 2.7496,
+      "step": 2541
+    },
+    {
+      "epoch": 33.90969899665552,
+      "grad_norm": 0.5777727365493774,
+      "learning_rate": 8.716351501668522e-05,
+      "loss": 2.7106,
+      "step": 2542
+    },
+    {
+      "epoch": 33.92307692307692,
+      "grad_norm": 0.6095665097236633,
+      "learning_rate": 8.7119021134594e-05,
+      "loss": 2.8617,
+      "step": 2543
+    },
+    {
+      "epoch": 33.936454849498325,
+      "grad_norm": 0.5426962375640869,
+      "learning_rate": 8.707452725250278e-05,
+      "loss": 2.7434,
+      "step": 2544
+    },
+    {
+      "epoch": 33.94983277591973,
+      "grad_norm": 0.5962931513786316,
+      "learning_rate": 8.703003337041156e-05,
+      "loss": 2.6328,
+      "step": 2545
+    },
+    {
+      "epoch": 33.96321070234114,
+      "grad_norm": 0.609198808670044,
+      "learning_rate": 8.698553948832036e-05,
+      "loss": 3.0128,
+      "step": 2546
+    },
+    {
+      "epoch": 33.97658862876254,
+      "grad_norm": 0.5592045783996582,
+      "learning_rate": 8.694104560622914e-05,
+      "loss": 2.4845,
+      "step": 2547
+    },
+    {
+      "epoch": 33.98996655518395,
+      "grad_norm": 0.6074299812316895,
+      "learning_rate": 8.689655172413794e-05,
+      "loss": 2.76,
+      "step": 2548
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 0.7298697233200073,
+      "learning_rate": 8.685205784204672e-05,
+      "loss": 2.7132,
+      "step": 2549
+    },
+    {
+      "epoch": 34.013377926421406,
+      "grad_norm": 0.5669110417366028,
+      "learning_rate": 8.68075639599555e-05,
+      "loss": 2.6124,
+      "step": 2550
+    },
+    {
+      "epoch": 34.02675585284281,
+      "grad_norm": 0.6375609636306763,
+      "learning_rate": 8.67630700778643e-05,
+      "loss": 2.8555,
+      "step": 2551
+    },
+    {
+      "epoch": 34.04013377926422,
+      "grad_norm": 0.6021872162818909,
+      "learning_rate": 8.671857619577308e-05,
+      "loss": 2.6396,
+      "step": 2552
+    },
+    {
+      "epoch": 34.05351170568562,
+      "grad_norm": 0.5895376801490784,
+      "learning_rate": 8.667408231368188e-05,
+      "loss": 2.7079,
+      "step": 2553
+    },
+    {
+      "epoch": 34.06688963210702,
+      "grad_norm": 0.5841491222381592,
+      "learning_rate": 8.662958843159066e-05,
+      "loss": 2.5644,
+      "step": 2554
+    },
+    {
+      "epoch": 34.080267558528426,
+      "grad_norm": 0.6124491691589355,
+      "learning_rate": 8.658509454949944e-05,
+      "loss": 2.8281,
+      "step": 2555
+    },
+    {
+      "epoch": 34.09364548494983,
+      "grad_norm": 0.5991553664207458,
+      "learning_rate": 8.654060066740824e-05,
+      "loss": 2.7442,
+      "step": 2556
+    },
+    {
+      "epoch": 34.10702341137124,
+      "grad_norm": 0.5902360081672668,
+      "learning_rate": 8.649610678531702e-05,
+      "loss": 2.6759,
+      "step": 2557
+    },
+    {
+      "epoch": 34.12040133779264,
+      "grad_norm": 0.590696394443512,
+      "learning_rate": 8.645161290322581e-05,
+      "loss": 2.6966,
+      "step": 2558
+    },
+    {
+      "epoch": 34.13377926421405,
+      "grad_norm": 0.5867067575454712,
+      "learning_rate": 8.64071190211346e-05,
+      "loss": 2.7264,
+      "step": 2559
+    },
+    {
+      "epoch": 34.147157190635454,
+      "grad_norm": 0.6447772979736328,
+      "learning_rate": 8.636262513904338e-05,
+      "loss": 2.7659,
+      "step": 2560
+    },
+    {
+      "epoch": 34.16053511705686,
+      "grad_norm": 0.6006119251251221,
+      "learning_rate": 8.631813125695218e-05,
+      "loss": 2.7641,
+      "step": 2561
+    },
+    {
+      "epoch": 34.17391304347826,
+      "grad_norm": 0.593132495880127,
+      "learning_rate": 8.627363737486096e-05,
+      "loss": 2.6671,
+      "step": 2562
+    },
+    {
+      "epoch": 34.187290969899664,
+      "grad_norm": 0.5959640145301819,
+      "learning_rate": 8.622914349276975e-05,
+      "loss": 2.7752,
+      "step": 2563
+    },
+    {
+      "epoch": 34.20066889632107,
+      "grad_norm": 0.6102313995361328,
+      "learning_rate": 8.618464961067854e-05,
+      "loss": 2.8268,
+      "step": 2564
+    },
+    {
+      "epoch": 34.214046822742475,
+      "grad_norm": 0.6111622452735901,
+      "learning_rate": 8.614015572858732e-05,
+      "loss": 2.5703,
+      "step": 2565
+    },
+    {
+      "epoch": 34.22742474916388,
+      "grad_norm": 0.5814309120178223,
+      "learning_rate": 8.609566184649611e-05,
+      "loss": 2.5823,
+      "step": 2566
+    },
+    {
+      "epoch": 34.240802675585286,
+      "grad_norm": 0.5712332129478455,
+      "learning_rate": 8.60511679644049e-05,
+      "loss": 2.3881,
+      "step": 2567
+    },
+    {
+      "epoch": 34.25418060200669,
+      "grad_norm": 0.5990560054779053,
+      "learning_rate": 8.600667408231369e-05,
+      "loss": 2.6209,
+      "step": 2568
+    },
+    {
+      "epoch": 34.2675585284281,
+      "grad_norm": 0.6006229519844055,
+      "learning_rate": 8.596218020022248e-05,
+      "loss": 2.6623,
+      "step": 2569
+    },
+    {
+      "epoch": 34.280936454849495,
+      "grad_norm": 0.607763946056366,
+      "learning_rate": 8.591768631813126e-05,
+      "loss": 2.7403,
+      "step": 2570
+    },
+    {
+      "epoch": 34.2943143812709,
+      "grad_norm": 0.6283872127532959,
+      "learning_rate": 8.587319243604005e-05,
+      "loss": 2.9186,
+      "step": 2571
+    },
+    {
+      "epoch": 34.30769230769231,
+      "grad_norm": 0.6027917861938477,
+      "learning_rate": 8.582869855394884e-05,
+      "loss": 2.7604,
+      "step": 2572
+    },
+    {
+      "epoch": 34.32107023411371,
+      "grad_norm": 0.617796778678894,
+      "learning_rate": 8.578420467185763e-05,
+      "loss": 2.2285,
+      "step": 2573
+    },
+    {
+      "epoch": 34.33444816053512,
+      "grad_norm": 0.6242703199386597,
+      "learning_rate": 8.573971078976641e-05,
+      "loss": 2.7596,
+      "step": 2574
+    },
+    {
+      "epoch": 34.34782608695652,
+      "grad_norm": 0.6676938533782959,
+      "learning_rate": 8.56952169076752e-05,
+      "loss": 2.8307,
+      "step": 2575
+    },
+    {
+      "epoch": 34.36120401337793,
+      "grad_norm": 0.616120457649231,
+      "learning_rate": 8.565072302558399e-05,
+      "loss": 2.562,
+      "step": 2576
+    },
+    {
+      "epoch": 34.374581939799334,
+      "grad_norm": 0.6003774404525757,
+      "learning_rate": 8.560622914349277e-05,
+      "loss": 2.913,
+      "step": 2577
+    },
+    {
+      "epoch": 34.38795986622073,
+      "grad_norm": 0.615688145160675,
+      "learning_rate": 8.556173526140157e-05,
+      "loss": 2.707,
+      "step": 2578
+    },
+    {
+      "epoch": 34.40133779264214,
+      "grad_norm": 0.6033686995506287,
+      "learning_rate": 8.551724137931035e-05,
+      "loss": 2.8834,
+      "step": 2579
+    },
+    {
+      "epoch": 34.414715719063544,
+      "grad_norm": 0.5940545201301575,
+      "learning_rate": 8.547274749721914e-05,
+      "loss": 2.5229,
+      "step": 2580
+    },
+    {
+      "epoch": 34.42809364548495,
+      "grad_norm": 0.587794303894043,
+      "learning_rate": 8.542825361512793e-05,
+      "loss": 2.4714,
+      "step": 2581
+    },
+    {
+      "epoch": 34.441471571906355,
+      "grad_norm": 0.6004595160484314,
+      "learning_rate": 8.538375973303671e-05,
+      "loss": 2.7798,
+      "step": 2582
+    },
+    {
+      "epoch": 34.45484949832776,
+      "grad_norm": 0.5780850648880005,
+      "learning_rate": 8.533926585094551e-05,
+      "loss": 2.5977,
+      "step": 2583
+    },
+    {
+      "epoch": 34.468227424749166,
+      "grad_norm": 0.6270626187324524,
+      "learning_rate": 8.529477196885428e-05,
+      "loss": 2.8688,
+      "step": 2584
+    },
+    {
+      "epoch": 34.48160535117057,
+      "grad_norm": 0.5620662569999695,
+      "learning_rate": 8.525027808676307e-05,
+      "loss": 2.5738,
+      "step": 2585
+    },
+    {
+      "epoch": 34.49498327759197,
+      "grad_norm": 0.6294083595275879,
+      "learning_rate": 8.520578420467186e-05,
+      "loss": 2.9486,
+      "step": 2586
+    },
+    {
+      "epoch": 34.508361204013376,
+      "grad_norm": 0.596815824508667,
+      "learning_rate": 8.516129032258064e-05,
+      "loss": 2.7196,
+      "step": 2587
+    },
+    {
+      "epoch": 34.52173913043478,
+      "grad_norm": 0.5836464166641235,
+      "learning_rate": 8.511679644048944e-05,
+      "loss": 2.7812,
+      "step": 2588
+    },
+    {
+      "epoch": 34.53511705685619,
+      "grad_norm": 0.6028749942779541,
+      "learning_rate": 8.507230255839822e-05,
+      "loss": 2.7632,
+      "step": 2589
+    },
+    {
+      "epoch": 34.54849498327759,
+      "grad_norm": 0.602236807346344,
+      "learning_rate": 8.502780867630701e-05,
+      "loss": 2.7839,
+      "step": 2590
+    },
+    {
+      "epoch": 34.561872909699,
+      "grad_norm": 0.6275736093521118,
+      "learning_rate": 8.49833147942158e-05,
+      "loss": 2.7257,
+      "step": 2591
+    },
+    {
+      "epoch": 34.575250836120404,
+      "grad_norm": 0.5825592875480652,
+      "learning_rate": 8.493882091212458e-05,
+      "loss": 2.6373,
+      "step": 2592
+    },
+    {
+      "epoch": 34.58862876254181,
+      "grad_norm": 0.6081156134605408,
+      "learning_rate": 8.489432703003337e-05,
+      "loss": 2.7701,
+      "step": 2593
+    },
+    {
+      "epoch": 34.60200668896321,
+      "grad_norm": 0.5824214220046997,
+      "learning_rate": 8.484983314794216e-05,
+      "loss": 2.7086,
+      "step": 2594
+    },
+    {
+      "epoch": 34.61538461538461,
+      "grad_norm": 0.5969799160957336,
+      "learning_rate": 8.480533926585095e-05,
+      "loss": 2.6757,
+      "step": 2595
+    },
+    {
+      "epoch": 34.62876254180602,
+      "grad_norm": 0.6191763877868652,
+      "learning_rate": 8.476084538375974e-05,
+      "loss": 2.8289,
+      "step": 2596
+    },
+    {
+      "epoch": 34.642140468227424,
+      "grad_norm": 0.570646345615387,
+      "learning_rate": 8.471635150166852e-05,
+      "loss": 2.6874,
+      "step": 2597
+    },
+    {
+      "epoch": 34.65551839464883,
+      "grad_norm": 0.6193318963050842,
+      "learning_rate": 8.467185761957731e-05,
+      "loss": 2.7932,
+      "step": 2598
+    },
+    {
+      "epoch": 34.668896321070235,
+      "grad_norm": 0.607635498046875,
+      "learning_rate": 8.46273637374861e-05,
+      "loss": 2.7382,
+      "step": 2599
+    },
+    {
+      "epoch": 34.68227424749164,
+      "grad_norm": 0.5707620978355408,
+      "learning_rate": 8.458286985539489e-05,
+      "loss": 2.5729,
+      "step": 2600
+    },
+    {
+      "epoch": 34.69565217391305,
+      "grad_norm": 0.5609964728355408,
+      "learning_rate": 8.453837597330367e-05,
+      "loss": 2.6938,
+      "step": 2601
+    },
+    {
+      "epoch": 34.709030100334445,
+      "grad_norm": 0.6459850072860718,
+      "learning_rate": 8.449388209121246e-05,
+      "loss": 2.7604,
+      "step": 2602
+    },
+    {
+      "epoch": 34.72240802675585,
+      "grad_norm": 0.5685142874717712,
+      "learning_rate": 8.444938820912125e-05,
+      "loss": 2.6518,
+      "step": 2603
+    },
+    {
+      "epoch": 34.735785953177256,
+      "grad_norm": 0.5800818204879761,
+      "learning_rate": 8.440489432703003e-05,
+      "loss": 2.6193,
+      "step": 2604
+    },
+    {
+      "epoch": 34.74916387959866,
+      "grad_norm": 0.6133848428726196,
+      "learning_rate": 8.436040044493883e-05,
+      "loss": 2.8524,
+      "step": 2605
+    },
+    {
+      "epoch": 34.76254180602007,
+      "grad_norm": 0.5716528296470642,
+      "learning_rate": 8.431590656284761e-05,
+      "loss": 2.9175,
+      "step": 2606
+    },
+    {
+      "epoch": 34.77591973244147,
+      "grad_norm": 0.5581215023994446,
+      "learning_rate": 8.42714126807564e-05,
+      "loss": 2.6181,
+      "step": 2607
+    },
+    {
+      "epoch": 34.78929765886288,
+      "grad_norm": 0.5720329284667969,
+      "learning_rate": 8.422691879866519e-05,
+      "loss": 2.6401,
+      "step": 2608
+    },
+    {
+      "epoch": 34.802675585284284,
+      "grad_norm": 0.6074360609054565,
+      "learning_rate": 8.418242491657397e-05,
+      "loss": 2.8911,
+      "step": 2609
+    },
+    {
+      "epoch": 34.81605351170568,
+      "grad_norm": 0.5685964226722717,
+      "learning_rate": 8.413793103448277e-05,
+      "loss": 2.868,
+      "step": 2610
+    },
+    {
+      "epoch": 34.82943143812709,
+      "grad_norm": 0.5762805938720703,
+      "learning_rate": 8.409343715239155e-05,
+      "loss": 2.7641,
+      "step": 2611
+    },
+    {
+      "epoch": 34.84280936454849,
+      "grad_norm": 0.6570479869842529,
+      "learning_rate": 8.404894327030033e-05,
+      "loss": 2.8734,
+      "step": 2612
+    },
+    {
+      "epoch": 34.8561872909699,
+      "grad_norm": 0.6143054366111755,
+      "learning_rate": 8.400444938820913e-05,
+      "loss": 2.7792,
+      "step": 2613
+    },
+    {
+      "epoch": 34.869565217391305,
+      "grad_norm": 0.6144394278526306,
+      "learning_rate": 8.395995550611791e-05,
+      "loss": 2.8023,
+      "step": 2614
+    },
+    {
+      "epoch": 34.88294314381271,
+      "grad_norm": 0.6102853417396545,
+      "learning_rate": 8.391546162402671e-05,
+      "loss": 2.7508,
+      "step": 2615
+    },
+    {
+      "epoch": 34.896321070234116,
+      "grad_norm": 0.6034586429595947,
+      "learning_rate": 8.387096774193549e-05,
+      "loss": 2.5528,
+      "step": 2616
+    },
+    {
+      "epoch": 34.90969899665552,
+      "grad_norm": 0.6032466292381287,
+      "learning_rate": 8.382647385984427e-05,
+      "loss": 2.7771,
+      "step": 2617
+    },
+    {
+      "epoch": 34.92307692307692,
+      "grad_norm": 0.6024323105812073,
+      "learning_rate": 8.378197997775307e-05,
+      "loss": 2.5514,
+      "step": 2618
+    },
+    {
+      "epoch": 34.936454849498325,
+      "grad_norm": 0.6762076616287231,
+      "learning_rate": 8.373748609566185e-05,
+      "loss": 2.8336,
+      "step": 2619
+    },
+    {
+      "epoch": 34.94983277591973,
+      "grad_norm": 0.5701456069946289,
+      "learning_rate": 8.369299221357065e-05,
+      "loss": 2.4598,
+      "step": 2620
+    },
+    {
+      "epoch": 34.96321070234114,
+      "grad_norm": 0.5878699421882629,
+      "learning_rate": 8.364849833147943e-05,
+      "loss": 2.6511,
+      "step": 2621
+    },
+    {
+      "epoch": 34.97658862876254,
+      "grad_norm": 0.6187490820884705,
+      "learning_rate": 8.360400444938821e-05,
+      "loss": 2.6226,
+      "step": 2622
+    },
+    {
+      "epoch": 34.98996655518395,
+      "grad_norm": 0.5954105854034424,
+      "learning_rate": 8.355951056729701e-05,
+      "loss": 2.7115,
+      "step": 2623
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 0.7193446755409241,
+      "learning_rate": 8.351501668520579e-05,
+      "loss": 2.7859,
+      "step": 2624
+    },
+    {
+      "epoch": 35.013377926421406,
+      "grad_norm": 0.6013773083686829,
+      "learning_rate": 8.347052280311457e-05,
+      "loss": 2.8222,
+      "step": 2625
+    },
+    {
+      "epoch": 35.02675585284281,
+      "grad_norm": 0.594046950340271,
+      "learning_rate": 8.342602892102336e-05,
+      "loss": 2.6455,
+      "step": 2626
+    },
+    {
+      "epoch": 35.04013377926422,
+      "grad_norm": 0.5741755962371826,
+      "learning_rate": 8.338153503893215e-05,
+      "loss": 2.6076,
+      "step": 2627
+    },
+    {
+      "epoch": 35.05351170568562,
+      "grad_norm": 0.6248610019683838,
+      "learning_rate": 8.333704115684093e-05,
+      "loss": 2.8714,
+      "step": 2628
+    },
+    {
+      "epoch": 35.06688963210702,
+      "grad_norm": 0.6038011312484741,
+      "learning_rate": 8.329254727474972e-05,
+      "loss": 2.6117,
+      "step": 2629
+    },
+    {
+      "epoch": 35.080267558528426,
+      "grad_norm": 0.618163526058197,
+      "learning_rate": 8.324805339265851e-05,
+      "loss": 2.8792,
+      "step": 2630
+    },
+    {
+      "epoch": 35.09364548494983,
+      "grad_norm": 0.5919619798660278,
+      "learning_rate": 8.32035595105673e-05,
+      "loss": 2.7317,
+      "step": 2631
+    },
+    {
+      "epoch": 35.10702341137124,
+      "grad_norm": 0.5996441841125488,
+      "learning_rate": 8.315906562847609e-05,
+      "loss": 2.6743,
+      "step": 2632
+    },
+    {
+      "epoch": 35.12040133779264,
+      "grad_norm": 0.5751325488090515,
+      "learning_rate": 8.311457174638487e-05,
+      "loss": 2.7814,
+      "step": 2633
+    },
+    {
+      "epoch": 35.13377926421405,
+      "grad_norm": 0.5554592609405518,
+      "learning_rate": 8.307007786429366e-05,
+      "loss": 2.4874,
+      "step": 2634
+    },
+    {
+      "epoch": 35.147157190635454,
+      "grad_norm": 0.5908927321434021,
+      "learning_rate": 8.302558398220245e-05,
+      "loss": 2.6838,
+      "step": 2635
+    },
+    {
+      "epoch": 35.16053511705686,
+      "grad_norm": 0.6261016130447388,
+      "learning_rate": 8.298109010011123e-05,
+      "loss": 2.8281,
+      "step": 2636
+    },
+    {
+      "epoch": 35.17391304347826,
+      "grad_norm": 0.5825657844543457,
+      "learning_rate": 8.293659621802003e-05,
+      "loss": 2.806,
+      "step": 2637
+    },
+    {
+      "epoch": 35.187290969899664,
+      "grad_norm": 0.5951602458953857,
+      "learning_rate": 8.289210233592881e-05,
+      "loss": 2.728,
+      "step": 2638
+    },
+    {
+      "epoch": 35.20066889632107,
+      "grad_norm": 0.6045755743980408,
+      "learning_rate": 8.28476084538376e-05,
+      "loss": 2.5632,
+      "step": 2639
+    },
+    {
+      "epoch": 35.214046822742475,
+      "grad_norm": 0.5793343782424927,
+      "learning_rate": 8.280311457174639e-05,
+      "loss": 2.4912,
+      "step": 2640
+    },
+    {
+      "epoch": 35.22742474916388,
+      "grad_norm": 0.5778194665908813,
+      "learning_rate": 8.275862068965517e-05,
+      "loss": 2.649,
+      "step": 2641
+    },
+    {
+      "epoch": 35.240802675585286,
+      "grad_norm": 0.5981440544128418,
+      "learning_rate": 8.271412680756397e-05,
+      "loss": 2.7538,
+      "step": 2642
+    },
+    {
+      "epoch": 35.25418060200669,
+      "grad_norm": 0.601005494594574,
+      "learning_rate": 8.266963292547275e-05,
+      "loss": 2.5793,
+      "step": 2643
+    },
+    {
+      "epoch": 35.2675585284281,
+      "grad_norm": 0.5964012742042542,
+      "learning_rate": 8.262513904338153e-05,
+      "loss": 2.7611,
+      "step": 2644
+    },
+    {
+      "epoch": 35.280936454849495,
+      "grad_norm": 0.603792130947113,
+      "learning_rate": 8.258064516129033e-05,
+      "loss": 2.5607,
+      "step": 2645
+    },
+    {
+      "epoch": 35.2943143812709,
+      "grad_norm": 0.6139290928840637,
+      "learning_rate": 8.253615127919911e-05,
+      "loss": 2.7616,
+      "step": 2646
+    },
+    {
+      "epoch": 35.30769230769231,
+      "grad_norm": 0.5915001034736633,
+      "learning_rate": 8.249165739710791e-05,
+      "loss": 2.7107,
+      "step": 2647
+    },
+    {
+      "epoch": 35.32107023411371,
+      "grad_norm": 0.6058906316757202,
+      "learning_rate": 8.244716351501669e-05,
+      "loss": 2.7078,
+      "step": 2648
+    },
+    {
+      "epoch": 35.33444816053512,
+      "grad_norm": 0.6192322969436646,
+      "learning_rate": 8.240266963292547e-05,
+      "loss": 2.6736,
+      "step": 2649
+    },
+    {
+      "epoch": 35.34782608695652,
+      "grad_norm": 0.6301470994949341,
+      "learning_rate": 8.235817575083427e-05,
+      "loss": 2.6024,
+      "step": 2650
+    },
+    {
+      "epoch": 35.36120401337793,
+      "grad_norm": 0.59424889087677,
+      "learning_rate": 8.231368186874305e-05,
+      "loss": 2.7615,
+      "step": 2651
+    },
+    {
+      "epoch": 35.374581939799334,
+      "grad_norm": 0.5878474712371826,
+      "learning_rate": 8.226918798665185e-05,
+      "loss": 2.7163,
+      "step": 2652
+    },
+    {
+      "epoch": 35.38795986622073,
+      "grad_norm": 0.6637581586837769,
+      "learning_rate": 8.222469410456063e-05,
+      "loss": 2.3804,
+      "step": 2653
+    },
+    {
+      "epoch": 35.40133779264214,
+      "grad_norm": 0.6142652630805969,
+      "learning_rate": 8.218020022246941e-05,
+      "loss": 2.8305,
+      "step": 2654
+    },
+    {
+      "epoch": 35.414715719063544,
+      "grad_norm": 0.59095698595047,
+      "learning_rate": 8.213570634037821e-05,
+      "loss": 2.7017,
+      "step": 2655
+    },
+    {
+      "epoch": 35.42809364548495,
+      "grad_norm": 0.6011471748352051,
+      "learning_rate": 8.209121245828699e-05,
+      "loss": 2.5261,
+      "step": 2656
+    },
+    {
+      "epoch": 35.441471571906355,
+      "grad_norm": 0.6142131686210632,
+      "learning_rate": 8.204671857619579e-05,
+      "loss": 2.8304,
+      "step": 2657
+    },
+    {
+      "epoch": 35.45484949832776,
+      "grad_norm": 0.5916325449943542,
+      "learning_rate": 8.200222469410457e-05,
+      "loss": 2.5681,
+      "step": 2658
+    },
+    {
+      "epoch": 35.468227424749166,
+      "grad_norm": 0.6357722282409668,
+      "learning_rate": 8.195773081201335e-05,
+      "loss": 2.6106,
+      "step": 2659
+    },
+    {
+      "epoch": 35.48160535117057,
+      "grad_norm": 0.6173760890960693,
+      "learning_rate": 8.191323692992215e-05,
+      "loss": 2.644,
+      "step": 2660
+    },
+    {
+      "epoch": 35.49498327759197,
+      "grad_norm": 0.6166900396347046,
+      "learning_rate": 8.186874304783093e-05,
+      "loss": 2.6629,
+      "step": 2661
+    },
+    {
+      "epoch": 35.508361204013376,
+      "grad_norm": 0.593647837638855,
+      "learning_rate": 8.182424916573973e-05,
+      "loss": 2.722,
+      "step": 2662
+    },
+    {
+      "epoch": 35.52173913043478,
+      "grad_norm": 0.6336297988891602,
+      "learning_rate": 8.177975528364851e-05,
+      "loss": 2.7934,
+      "step": 2663
+    },
+    {
+      "epoch": 35.53511705685619,
+      "grad_norm": 0.6457211971282959,
+      "learning_rate": 8.173526140155729e-05,
+      "loss": 2.5204,
+      "step": 2664
+    },
+    {
+      "epoch": 35.54849498327759,
+      "grad_norm": 0.5988196730613708,
+      "learning_rate": 8.169076751946609e-05,
+      "loss": 2.8071,
+      "step": 2665
+    },
+    {
+      "epoch": 35.561872909699,
+      "grad_norm": 0.6295507550239563,
+      "learning_rate": 8.164627363737485e-05,
+      "loss": 2.6807,
+      "step": 2666
+    },
+    {
+      "epoch": 35.575250836120404,
+      "grad_norm": 0.5988739728927612,
+      "learning_rate": 8.160177975528365e-05,
+      "loss": 2.5091,
+      "step": 2667
+    },
+    {
+      "epoch": 35.58862876254181,
+      "grad_norm": 0.5924574732780457,
+      "learning_rate": 8.155728587319243e-05,
+      "loss": 2.5552,
+      "step": 2668
+    },
+    {
+      "epoch": 35.60200668896321,
+      "grad_norm": 0.6240471005439758,
+      "learning_rate": 8.151279199110123e-05,
+      "loss": 2.8841,
+      "step": 2669
+    },
+    {
+      "epoch": 35.61538461538461,
+      "grad_norm": 0.5939052700996399,
+      "learning_rate": 8.146829810901001e-05,
+      "loss": 2.4867,
+      "step": 2670
+    },
+    {
+      "epoch": 35.62876254180602,
+      "grad_norm": 0.6398495435714722,
+      "learning_rate": 8.14238042269188e-05,
+      "loss": 2.801,
+      "step": 2671
+    },
+    {
+      "epoch": 35.642140468227424,
+      "grad_norm": 0.6106517910957336,
+      "learning_rate": 8.137931034482759e-05,
+      "loss": 2.7849,
+      "step": 2672
+    },
+    {
+      "epoch": 35.65551839464883,
+      "grad_norm": 0.5723669528961182,
+      "learning_rate": 8.133481646273637e-05,
+      "loss": 2.6566,
+      "step": 2673
+    },
+    {
+      "epoch": 35.668896321070235,
+      "grad_norm": 0.616449236869812,
+      "learning_rate": 8.129032258064517e-05,
+      "loss": 2.823,
+      "step": 2674
+    },
+    {
+      "epoch": 35.68227424749164,
+      "grad_norm": 0.6001511812210083,
+      "learning_rate": 8.124582869855395e-05,
+      "loss": 2.6833,
+      "step": 2675
+    },
+    {
+      "epoch": 35.69565217391305,
+      "grad_norm": 0.5830066800117493,
+      "learning_rate": 8.120133481646273e-05,
+      "loss": 2.8018,
+      "step": 2676
+    },
+    {
+      "epoch": 35.709030100334445,
+      "grad_norm": 0.5999500155448914,
+      "learning_rate": 8.115684093437153e-05,
+      "loss": 2.7339,
+      "step": 2677
+    },
+    {
+      "epoch": 35.72240802675585,
+      "grad_norm": 0.6584502458572388,
+      "learning_rate": 8.111234705228031e-05,
+      "loss": 2.8347,
+      "step": 2678
+    },
+    {
+      "epoch": 35.735785953177256,
+      "grad_norm": 0.6561383008956909,
+      "learning_rate": 8.106785317018911e-05,
+      "loss": 2.6747,
+      "step": 2679
+    },
+    {
+      "epoch": 35.74916387959866,
+      "grad_norm": 0.6040836572647095,
+      "learning_rate": 8.102335928809789e-05,
+      "loss": 2.713,
+      "step": 2680
+    },
+    {
+      "epoch": 35.76254180602007,
+      "grad_norm": 0.6353011131286621,
+      "learning_rate": 8.097886540600667e-05,
+      "loss": 2.51,
+      "step": 2681
+    },
+    {
+      "epoch": 35.77591973244147,
+      "grad_norm": 0.5977175235748291,
+      "learning_rate": 8.093437152391547e-05,
+      "loss": 2.7502,
+      "step": 2682
+    },
+    {
+      "epoch": 35.78929765886288,
+      "grad_norm": 0.6166728734970093,
+      "learning_rate": 8.088987764182425e-05,
+      "loss": 2.8618,
+      "step": 2683
+    },
+    {
+      "epoch": 35.802675585284284,
+      "grad_norm": 0.5676133036613464,
+      "learning_rate": 8.084538375973305e-05,
+      "loss": 2.6979,
+      "step": 2684
+    },
+    {
+      "epoch": 35.81605351170568,
+      "grad_norm": 0.6121795773506165,
+      "learning_rate": 8.080088987764183e-05,
+      "loss": 2.7341,
+      "step": 2685
+    },
+    {
+      "epoch": 35.82943143812709,
+      "grad_norm": 0.6005775332450867,
+      "learning_rate": 8.075639599555061e-05,
+      "loss": 2.6737,
+      "step": 2686
+    },
+    {
+      "epoch": 35.84280936454849,
+      "grad_norm": 0.5900837182998657,
+      "learning_rate": 8.071190211345941e-05,
+      "loss": 2.7243,
+      "step": 2687
+    },
+    {
+      "epoch": 35.8561872909699,
+      "grad_norm": 0.5916823744773865,
+      "learning_rate": 8.066740823136819e-05,
+      "loss": 2.5637,
+      "step": 2688
+    },
+    {
+      "epoch": 35.869565217391305,
+      "grad_norm": 0.6083678603172302,
+      "learning_rate": 8.062291434927699e-05,
+      "loss": 2.7644,
+      "step": 2689
+    },
+    {
+      "epoch": 35.88294314381271,
+      "grad_norm": 0.5921754837036133,
+      "learning_rate": 8.057842046718577e-05,
+      "loss": 2.6599,
+      "step": 2690
+    },
+    {
+      "epoch": 35.896321070234116,
+      "grad_norm": 0.6174843907356262,
+      "learning_rate": 8.053392658509455e-05,
+      "loss": 2.6216,
+      "step": 2691
+    },
+    {
+      "epoch": 35.90969899665552,
+      "grad_norm": 0.5929603576660156,
+      "learning_rate": 8.048943270300335e-05,
+      "loss": 2.5201,
+      "step": 2692
+    },
+    {
+      "epoch": 35.92307692307692,
+      "grad_norm": 0.6305480599403381,
+      "learning_rate": 8.044493882091213e-05,
+      "loss": 2.726,
+      "step": 2693
+    },
+    {
+      "epoch": 35.936454849498325,
+      "grad_norm": 0.662198543548584,
+      "learning_rate": 8.040044493882092e-05,
+      "loss": 2.7794,
+      "step": 2694
+    },
+    {
+      "epoch": 35.94983277591973,
+      "grad_norm": 0.5887805223464966,
+      "learning_rate": 8.03559510567297e-05,
+      "loss": 2.5348,
+      "step": 2695
+    },
+    {
+      "epoch": 35.96321070234114,
+      "grad_norm": 0.5848047137260437,
+      "learning_rate": 8.031145717463849e-05,
+      "loss": 2.4549,
+      "step": 2696
+    },
+    {
+      "epoch": 35.97658862876254,
+      "grad_norm": 0.5758711695671082,
+      "learning_rate": 8.026696329254729e-05,
+      "loss": 2.5325,
+      "step": 2697
+    },
+    {
+      "epoch": 35.98996655518395,
+      "grad_norm": 0.6385326981544495,
+      "learning_rate": 8.022246941045607e-05,
+      "loss": 2.7417,
+      "step": 2698
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 0.7033568024635315,
+      "learning_rate": 8.017797552836486e-05,
+      "loss": 2.6844,
+      "step": 2699
+    },
+    {
+      "epoch": 36.013377926421406,
+      "grad_norm": 0.5869501233100891,
+      "learning_rate": 8.013348164627365e-05,
+      "loss": 2.5899,
+      "step": 2700
+    },
+    {
+      "epoch": 36.02675585284281,
+      "grad_norm": 0.6519103050231934,
+      "learning_rate": 8.008898776418243e-05,
+      "loss": 2.8153,
+      "step": 2701
+    },
+    {
+      "epoch": 36.04013377926422,
+      "grad_norm": 0.5899248123168945,
+      "learning_rate": 8.004449388209122e-05,
+      "loss": 2.8749,
+      "step": 2702
+    },
+    {
+      "epoch": 36.05351170568562,
+      "grad_norm": 0.6234768629074097,
+      "learning_rate": 8e-05,
+      "loss": 2.5244,
+      "step": 2703
+    },
+    {
+      "epoch": 36.06688963210702,
+      "grad_norm": 0.5662122368812561,
+      "learning_rate": 7.99555061179088e-05,
+      "loss": 2.4421,
+      "step": 2704
+    },
+    {
+      "epoch": 36.080267558528426,
+      "grad_norm": 0.584144651889801,
+      "learning_rate": 7.991101223581758e-05,
+      "loss": 2.7261,
+      "step": 2705
+    },
+    {
+      "epoch": 36.09364548494983,
+      "grad_norm": 0.6016060709953308,
+      "learning_rate": 7.986651835372637e-05,
+      "loss": 2.607,
+      "step": 2706
+    },
+    {
+      "epoch": 36.10702341137124,
+      "grad_norm": 0.6164819002151489,
+      "learning_rate": 7.982202447163516e-05,
+      "loss": 2.7332,
+      "step": 2707
+    },
+    {
+      "epoch": 36.12040133779264,
+      "grad_norm": 0.6115155816078186,
+      "learning_rate": 7.977753058954393e-05,
+      "loss": 2.706,
+      "step": 2708
+    },
+    {
+      "epoch": 36.13377926421405,
+      "grad_norm": 0.6102612018585205,
+      "learning_rate": 7.973303670745273e-05,
+      "loss": 2.8179,
+      "step": 2709
+    },
+    {
+      "epoch": 36.147157190635454,
+      "grad_norm": 0.5910987257957458,
+      "learning_rate": 7.968854282536151e-05,
+      "loss": 2.6354,
+      "step": 2710
+    },
+    {
+      "epoch": 36.16053511705686,
+      "grad_norm": 0.6355498433113098,
+      "learning_rate": 7.96440489432703e-05,
+      "loss": 2.705,
+      "step": 2711
+    },
+    {
+      "epoch": 36.17391304347826,
+      "grad_norm": 0.6213639378547668,
+      "learning_rate": 7.959955506117909e-05,
+      "loss": 2.481,
+      "step": 2712
+    },
+    {
+      "epoch": 36.187290969899664,
+      "grad_norm": 0.587666392326355,
+      "learning_rate": 7.955506117908787e-05,
+      "loss": 2.6295,
+      "step": 2713
+    },
+    {
+      "epoch": 36.20066889632107,
+      "grad_norm": 0.5778639316558838,
+      "learning_rate": 7.951056729699667e-05,
+      "loss": 2.5795,
+      "step": 2714
+    },
+    {
+      "epoch": 36.214046822742475,
+      "grad_norm": 0.6050285696983337,
+      "learning_rate": 7.946607341490545e-05,
+      "loss": 2.671,
+      "step": 2715
+    },
+    {
+      "epoch": 36.22742474916388,
+      "grad_norm": 0.6501610279083252,
+      "learning_rate": 7.942157953281425e-05,
+      "loss": 2.7712,
+      "step": 2716
+    },
+    {
+      "epoch": 36.240802675585286,
+      "grad_norm": 0.6189882159233093,
+      "learning_rate": 7.937708565072303e-05,
+      "loss": 2.5324,
+      "step": 2717
+    },
+    {
+      "epoch": 36.25418060200669,
+      "grad_norm": 0.6189801096916199,
+      "learning_rate": 7.933259176863181e-05,
+      "loss": 2.5716,
+      "step": 2718
+    },
+    {
+      "epoch": 36.2675585284281,
+      "grad_norm": 0.617070734500885,
+      "learning_rate": 7.92880978865406e-05,
+      "loss": 2.7485,
+      "step": 2719
+    },
+    {
+      "epoch": 36.280936454849495,
+      "grad_norm": 0.6042205095291138,
+      "learning_rate": 7.924360400444939e-05,
+      "loss": 2.4323,
+      "step": 2720
+    },
+    {
+      "epoch": 36.2943143812709,
+      "grad_norm": 0.6032358407974243,
+      "learning_rate": 7.919911012235818e-05,
+      "loss": 2.6852,
+      "step": 2721
+    },
+    {
+      "epoch": 36.30769230769231,
+      "grad_norm": 0.5686265230178833,
+      "learning_rate": 7.915461624026697e-05,
+      "loss": 2.551,
+      "step": 2722
+    },
+    {
+      "epoch": 36.32107023411371,
+      "grad_norm": 0.5876219868659973,
+      "learning_rate": 7.911012235817575e-05,
+      "loss": 2.3279,
+      "step": 2723
+    },
+    {
+      "epoch": 36.33444816053512,
+      "grad_norm": 0.624603807926178,
+      "learning_rate": 7.906562847608455e-05,
+      "loss": 2.6303,
+      "step": 2724
+    },
+    {
+      "epoch": 36.34782608695652,
+      "grad_norm": 0.5873613953590393,
+      "learning_rate": 7.902113459399333e-05,
+      "loss": 2.7109,
+      "step": 2725
+    },
+    {
+      "epoch": 36.36120401337793,
+      "grad_norm": 0.6027836203575134,
+      "learning_rate": 7.897664071190212e-05,
+      "loss": 2.6051,
+      "step": 2726
+    },
+    {
+      "epoch": 36.374581939799334,
+      "grad_norm": 0.5907491445541382,
+      "learning_rate": 7.89321468298109e-05,
+      "loss": 2.5136,
+      "step": 2727
+    },
+    {
+      "epoch": 36.38795986622073,
+      "grad_norm": 0.6281634569168091,
+      "learning_rate": 7.888765294771969e-05,
+      "loss": 2.6751,
+      "step": 2728
+    },
+    {
+      "epoch": 36.40133779264214,
+      "grad_norm": 0.6197676658630371,
+      "learning_rate": 7.884315906562848e-05,
+      "loss": 2.6621,
+      "step": 2729
+    },
+    {
+      "epoch": 36.414715719063544,
+      "grad_norm": 0.6759832501411438,
+      "learning_rate": 7.879866518353727e-05,
+      "loss": 2.9938,
+      "step": 2730
+    },
+    {
+      "epoch": 36.42809364548495,
+      "grad_norm": 0.6368512511253357,
+      "learning_rate": 7.875417130144606e-05,
+      "loss": 2.8629,
+      "step": 2731
+    },
+    {
+      "epoch": 36.441471571906355,
+      "grad_norm": 0.5914281606674194,
+      "learning_rate": 7.870967741935484e-05,
+      "loss": 2.6094,
+      "step": 2732
+    },
+    {
+      "epoch": 36.45484949832776,
+      "grad_norm": 0.60133957862854,
+      "learning_rate": 7.866518353726363e-05,
+      "loss": 2.7982,
+      "step": 2733
+    },
+    {
+      "epoch": 36.468227424749166,
+      "grad_norm": 0.621593177318573,
+      "learning_rate": 7.862068965517242e-05,
+      "loss": 2.595,
+      "step": 2734
+    },
+    {
+      "epoch": 36.48160535117057,
+      "grad_norm": 0.6033498048782349,
+      "learning_rate": 7.85761957730812e-05,
+      "loss": 2.7714,
+      "step": 2735
+    },
+    {
+      "epoch": 36.49498327759197,
+      "grad_norm": 0.6060223579406738,
+      "learning_rate": 7.853170189099e-05,
+      "loss": 2.5214,
+      "step": 2736
+    },
+    {
+      "epoch": 36.508361204013376,
+      "grad_norm": 0.6156677007675171,
+      "learning_rate": 7.848720800889878e-05,
+      "loss": 2.7125,
+      "step": 2737
+    },
+    {
+      "epoch": 36.52173913043478,
+      "grad_norm": 0.611400306224823,
+      "learning_rate": 7.844271412680757e-05,
+      "loss": 2.8178,
+      "step": 2738
+    },
+    {
+      "epoch": 36.53511705685619,
+      "grad_norm": 0.6582430601119995,
+      "learning_rate": 7.839822024471636e-05,
+      "loss": 2.9621,
+      "step": 2739
+    },
+    {
+      "epoch": 36.54849498327759,
+      "grad_norm": 0.6210145950317383,
+      "learning_rate": 7.835372636262514e-05,
+      "loss": 2.6619,
+      "step": 2740
+    },
+    {
+      "epoch": 36.561872909699,
+      "grad_norm": 0.6173349022865295,
+      "learning_rate": 7.830923248053394e-05,
+      "loss": 2.6075,
+      "step": 2741
+    },
+    {
+      "epoch": 36.575250836120404,
+      "grad_norm": 0.5770074129104614,
+      "learning_rate": 7.826473859844272e-05,
+      "loss": 2.5834,
+      "step": 2742
+    },
+    {
+      "epoch": 36.58862876254181,
+      "grad_norm": 0.6317800283432007,
+      "learning_rate": 7.82202447163515e-05,
+      "loss": 2.6134,
+      "step": 2743
+    },
+    {
+      "epoch": 36.60200668896321,
+      "grad_norm": 0.6122284531593323,
+      "learning_rate": 7.81757508342603e-05,
+      "loss": 2.7218,
+      "step": 2744
+    },
+    {
+      "epoch": 36.61538461538461,
+      "grad_norm": 0.5796928405761719,
+      "learning_rate": 7.813125695216908e-05,
+      "loss": 2.6459,
+      "step": 2745
+    },
+    {
+      "epoch": 36.62876254180602,
+      "grad_norm": 0.5904363393783569,
+      "learning_rate": 7.808676307007788e-05,
+      "loss": 2.5216,
+      "step": 2746
+    },
+    {
+      "epoch": 36.642140468227424,
+      "grad_norm": 0.6074969172477722,
+      "learning_rate": 7.804226918798666e-05,
+      "loss": 2.5776,
+      "step": 2747
+    },
+    {
+      "epoch": 36.65551839464883,
+      "grad_norm": 0.6080595850944519,
+      "learning_rate": 7.799777530589544e-05,
+      "loss": 2.7651,
+      "step": 2748
+    },
+    {
+      "epoch": 36.668896321070235,
+      "grad_norm": 0.5953335762023926,
+      "learning_rate": 7.795328142380423e-05,
+      "loss": 2.5139,
+      "step": 2749
+    },
+    {
+      "epoch": 36.68227424749164,
+      "grad_norm": 0.6192560195922852,
+      "learning_rate": 7.790878754171301e-05,
+      "loss": 2.5091,
+      "step": 2750
+    },
+    {
+      "epoch": 36.69565217391305,
+      "grad_norm": 0.6019810438156128,
+      "learning_rate": 7.78642936596218e-05,
+      "loss": 2.7217,
+      "step": 2751
+    },
+    {
+      "epoch": 36.709030100334445,
+      "grad_norm": 0.5785876512527466,
+      "learning_rate": 7.781979977753059e-05,
+      "loss": 2.4336,
+      "step": 2752
+    },
+    {
+      "epoch": 36.72240802675585,
+      "grad_norm": 0.6099944114685059,
+      "learning_rate": 7.777530589543937e-05,
+      "loss": 2.6614,
+      "step": 2753
+    },
+    {
+      "epoch": 36.735785953177256,
+      "grad_norm": 0.6206619143486023,
+      "learning_rate": 7.773081201334817e-05,
+      "loss": 2.7906,
+      "step": 2754
+    },
+    {
+      "epoch": 36.74916387959866,
+      "grad_norm": 0.5885736346244812,
+      "learning_rate": 7.768631813125695e-05,
+      "loss": 2.4836,
+      "step": 2755
+    },
+    {
+      "epoch": 36.76254180602007,
+      "grad_norm": 0.6022318005561829,
+      "learning_rate": 7.764182424916574e-05,
+      "loss": 2.6954,
+      "step": 2756
+    },
+    {
+      "epoch": 36.77591973244147,
+      "grad_norm": 0.6479224562644958,
+      "learning_rate": 7.759733036707453e-05,
+      "loss": 2.6576,
+      "step": 2757
+    },
+    {
+      "epoch": 36.78929765886288,
+      "grad_norm": 0.6288067102432251,
+      "learning_rate": 7.755283648498331e-05,
+      "loss": 2.6265,
+      "step": 2758
+    },
+    {
+      "epoch": 36.802675585284284,
+      "grad_norm": 0.6241925358772278,
+      "learning_rate": 7.75083426028921e-05,
+      "loss": 2.7554,
+      "step": 2759
+    },
+    {
+      "epoch": 36.81605351170568,
+      "grad_norm": 0.6327086687088013,
+      "learning_rate": 7.746384872080089e-05,
+      "loss": 2.6145,
+      "step": 2760
+    },
+    {
+      "epoch": 36.82943143812709,
+      "grad_norm": 0.6177266836166382,
+      "learning_rate": 7.741935483870968e-05,
+      "loss": 2.8887,
+      "step": 2761
+    },
+    {
+      "epoch": 36.84280936454849,
+      "grad_norm": 0.6014055013656616,
+      "learning_rate": 7.737486095661847e-05,
+      "loss": 2.6512,
+      "step": 2762
+    },
+    {
+      "epoch": 36.8561872909699,
+      "grad_norm": 0.5890874266624451,
+      "learning_rate": 7.733036707452725e-05,
+      "loss": 2.4959,
+      "step": 2763
+    },
+    {
+      "epoch": 36.869565217391305,
+      "grad_norm": 0.5938498973846436,
+      "learning_rate": 7.728587319243604e-05,
+      "loss": 2.6364,
+      "step": 2764
+    },
+    {
+      "epoch": 36.88294314381271,
+      "grad_norm": 0.587141752243042,
+      "learning_rate": 7.724137931034483e-05,
+      "loss": 2.804,
+      "step": 2765
+    },
+    {
+      "epoch": 36.896321070234116,
+      "grad_norm": 0.5674299597740173,
+      "learning_rate": 7.719688542825362e-05,
+      "loss": 2.5209,
+      "step": 2766
+    },
+    {
+      "epoch": 36.90969899665552,
+      "grad_norm": 0.616888165473938,
+      "learning_rate": 7.71523915461624e-05,
+      "loss": 2.6253,
+      "step": 2767
+    },
+    {
+      "epoch": 36.92307692307692,
+      "grad_norm": 0.6040933132171631,
+      "learning_rate": 7.710789766407119e-05,
+      "loss": 2.6645,
+      "step": 2768
+    },
+    {
+      "epoch": 36.936454849498325,
+      "grad_norm": 0.5765551328659058,
+      "learning_rate": 7.706340378197998e-05,
+      "loss": 2.5598,
+      "step": 2769
+    },
+    {
+      "epoch": 36.94983277591973,
+      "grad_norm": 0.598349928855896,
+      "learning_rate": 7.701890989988877e-05,
+      "loss": 2.6193,
+      "step": 2770
+    },
+    {
+      "epoch": 36.96321070234114,
+      "grad_norm": 0.5944122076034546,
+      "learning_rate": 7.697441601779756e-05,
+      "loss": 2.7833,
+      "step": 2771
+    },
+    {
+      "epoch": 36.97658862876254,
+      "grad_norm": 0.6495612263679504,
+      "learning_rate": 7.692992213570634e-05,
+      "loss": 2.8149,
+      "step": 2772
+    },
+    {
+      "epoch": 36.98996655518395,
+      "grad_norm": 0.6290954351425171,
+      "learning_rate": 7.688542825361513e-05,
+      "loss": 2.8039,
+      "step": 2773
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 0.7225489020347595,
+      "learning_rate": 7.684093437152392e-05,
+      "loss": 2.6676,
+      "step": 2774
+    },
+    {
+      "epoch": 37.013377926421406,
+      "grad_norm": 0.5684818029403687,
+      "learning_rate": 7.67964404894327e-05,
+      "loss": 2.5921,
+      "step": 2775
+    },
+    {
+      "epoch": 37.02675585284281,
+      "grad_norm": 0.6261118054389954,
+      "learning_rate": 7.67519466073415e-05,
+      "loss": 2.6034,
+      "step": 2776
+    },
+    {
+      "epoch": 37.04013377926422,
+      "grad_norm": 0.5897383689880371,
+      "learning_rate": 7.670745272525028e-05,
+      "loss": 2.6218,
+      "step": 2777
+    },
+    {
+      "epoch": 37.05351170568562,
+      "grad_norm": 0.6284380555152893,
+      "learning_rate": 7.666295884315907e-05,
+      "loss": 2.8228,
+      "step": 2778
+    },
+    {
+      "epoch": 37.06688963210702,
+      "grad_norm": 0.6081552505493164,
+      "learning_rate": 7.661846496106786e-05,
+      "loss": 2.4883,
+      "step": 2779
+    },
+    {
+      "epoch": 37.080267558528426,
+      "grad_norm": 0.5826512575149536,
+      "learning_rate": 7.657397107897664e-05,
+      "loss": 2.288,
+      "step": 2780
+    },
+    {
+      "epoch": 37.09364548494983,
+      "grad_norm": 0.609711766242981,
+      "learning_rate": 7.652947719688544e-05,
+      "loss": 2.7462,
+      "step": 2781
+    },
+    {
+      "epoch": 37.10702341137124,
+      "grad_norm": 0.6180025339126587,
+      "learning_rate": 7.648498331479422e-05,
+      "loss": 2.5538,
+      "step": 2782
+    },
+    {
+      "epoch": 37.12040133779264,
+      "grad_norm": 0.5980042815208435,
+      "learning_rate": 7.6440489432703e-05,
+      "loss": 2.5066,
+      "step": 2783
+    },
+    {
+      "epoch": 37.13377926421405,
+      "grad_norm": 0.5932002067565918,
+      "learning_rate": 7.63959955506118e-05,
+      "loss": 2.5456,
+      "step": 2784
+    },
+    {
+      "epoch": 37.147157190635454,
+      "grad_norm": 0.5952025055885315,
+      "learning_rate": 7.635150166852058e-05,
+      "loss": 2.5584,
+      "step": 2785
+    },
+    {
+      "epoch": 37.16053511705686,
+      "grad_norm": 0.6259213089942932,
+      "learning_rate": 7.630700778642938e-05,
+      "loss": 2.7745,
+      "step": 2786
+    },
+    {
+      "epoch": 37.17391304347826,
+      "grad_norm": 0.620085597038269,
+      "learning_rate": 7.626251390433816e-05,
+      "loss": 2.7472,
+      "step": 2787
+    },
+    {
+      "epoch": 37.187290969899664,
+      "grad_norm": 0.5972263216972351,
+      "learning_rate": 7.621802002224694e-05,
+      "loss": 2.4507,
+      "step": 2788
+    },
+    {
+      "epoch": 37.20066889632107,
+      "grad_norm": 0.6285966634750366,
+      "learning_rate": 7.617352614015574e-05,
+      "loss": 2.6729,
+      "step": 2789
+    },
+    {
+      "epoch": 37.214046822742475,
+      "grad_norm": 0.6250035762786865,
+      "learning_rate": 7.612903225806451e-05,
+      "loss": 2.7339,
+      "step": 2790
+    },
+    {
+      "epoch": 37.22742474916388,
+      "grad_norm": 0.6274656653404236,
+      "learning_rate": 7.60845383759733e-05,
+      "loss": 2.4538,
+      "step": 2791
+    },
+    {
+      "epoch": 37.240802675585286,
+      "grad_norm": 0.632949948310852,
+      "learning_rate": 7.604004449388209e-05,
+      "loss": 2.6448,
+      "step": 2792
+    },
+    {
+      "epoch": 37.25418060200669,
+      "grad_norm": 0.5869743824005127,
+      "learning_rate": 7.599555061179088e-05,
+      "loss": 2.6065,
+      "step": 2793
+    },
+    {
+      "epoch": 37.2675585284281,
+      "grad_norm": 0.6650564670562744,
+      "learning_rate": 7.595105672969966e-05,
+      "loss": 2.6521,
+      "step": 2794
+    },
+    {
+      "epoch": 37.280936454849495,
+      "grad_norm": 0.597977340221405,
+      "learning_rate": 7.590656284760845e-05,
+      "loss": 2.7055,
+      "step": 2795
+    },
+    {
+      "epoch": 37.2943143812709,
+      "grad_norm": 0.6276383996009827,
+      "learning_rate": 7.586206896551724e-05,
+      "loss": 2.6584,
+      "step": 2796
+    },
+    {
+      "epoch": 37.30769230769231,
+      "grad_norm": 0.637657105922699,
+      "learning_rate": 7.581757508342603e-05,
+      "loss": 2.763,
+      "step": 2797
+    },
+    {
+      "epoch": 37.32107023411371,
+      "grad_norm": 0.5994833707809448,
+      "learning_rate": 7.577308120133482e-05,
+      "loss": 2.6043,
+      "step": 2798
+    },
+    {
+      "epoch": 37.33444816053512,
+      "grad_norm": 0.6213527321815491,
+      "learning_rate": 7.57285873192436e-05,
+      "loss": 2.7679,
+      "step": 2799
+    },
+    {
+      "epoch": 37.34782608695652,
+      "grad_norm": 0.6137163639068604,
+      "learning_rate": 7.568409343715239e-05,
+      "loss": 2.6287,
+      "step": 2800
+    },
+    {
+      "epoch": 37.36120401337793,
+      "grad_norm": 0.6083935499191284,
+      "learning_rate": 7.563959955506118e-05,
+      "loss": 2.6636,
+      "step": 2801
+    },
+    {
+      "epoch": 37.374581939799334,
+      "grad_norm": 0.6070682406425476,
+      "learning_rate": 7.559510567296996e-05,
+      "loss": 2.5606,
+      "step": 2802
+    },
+    {
+      "epoch": 37.38795986622073,
+      "grad_norm": 0.6027565002441406,
+      "learning_rate": 7.555061179087876e-05,
+      "loss": 2.55,
+      "step": 2803
+    },
+    {
+      "epoch": 37.40133779264214,
+      "grad_norm": 0.6037593483924866,
+      "learning_rate": 7.550611790878754e-05,
+      "loss": 2.4951,
+      "step": 2804
+    },
+    {
+      "epoch": 37.414715719063544,
+      "grad_norm": 0.588081955909729,
+      "learning_rate": 7.546162402669633e-05,
+      "loss": 2.4888,
+      "step": 2805
+    },
+    {
+      "epoch": 37.42809364548495,
+      "grad_norm": 0.640067994594574,
+      "learning_rate": 7.541713014460512e-05,
+      "loss": 2.6508,
+      "step": 2806
+    },
+    {
+      "epoch": 37.441471571906355,
+      "grad_norm": 0.6123721599578857,
+      "learning_rate": 7.53726362625139e-05,
+      "loss": 2.6542,
+      "step": 2807
+    },
+    {
+      "epoch": 37.45484949832776,
+      "grad_norm": 0.6153370141983032,
+      "learning_rate": 7.53281423804227e-05,
+      "loss": 2.5986,
+      "step": 2808
+    },
+    {
+      "epoch": 37.468227424749166,
+      "grad_norm": 0.6331472396850586,
+      "learning_rate": 7.528364849833148e-05,
+      "loss": 2.5883,
+      "step": 2809
+    },
+    {
+      "epoch": 37.48160535117057,
+      "grad_norm": 0.6108154654502869,
+      "learning_rate": 7.523915461624026e-05,
+      "loss": 2.5254,
+      "step": 2810
+    },
+    {
+      "epoch": 37.49498327759197,
+      "grad_norm": 0.6167747974395752,
+      "learning_rate": 7.519466073414906e-05,
+      "loss": 2.5088,
+      "step": 2811
+    },
+    {
+      "epoch": 37.508361204013376,
+      "grad_norm": 0.6372016668319702,
+      "learning_rate": 7.515016685205784e-05,
+      "loss": 2.6996,
+      "step": 2812
+    },
+    {
+      "epoch": 37.52173913043478,
+      "grad_norm": 0.5725007057189941,
+      "learning_rate": 7.510567296996664e-05,
+      "loss": 2.5294,
+      "step": 2813
+    },
+    {
+      "epoch": 37.53511705685619,
+      "grad_norm": 0.5692083835601807,
+      "learning_rate": 7.506117908787542e-05,
+      "loss": 2.5158,
+      "step": 2814
+    },
+    {
+      "epoch": 37.54849498327759,
+      "grad_norm": 0.6351197361946106,
+      "learning_rate": 7.50166852057842e-05,
+      "loss": 2.7666,
+      "step": 2815
+    },
+    {
+      "epoch": 37.561872909699,
+      "grad_norm": 0.5791570544242859,
+      "learning_rate": 7.4972191323693e-05,
+      "loss": 2.4728,
+      "step": 2816
+    },
+    {
+      "epoch": 37.575250836120404,
+      "grad_norm": 0.6294588446617126,
+      "learning_rate": 7.492769744160178e-05,
+      "loss": 2.657,
+      "step": 2817
+    },
+    {
+      "epoch": 37.58862876254181,
+      "grad_norm": 0.6071592569351196,
+      "learning_rate": 7.488320355951058e-05,
+      "loss": 2.7995,
+      "step": 2818
+    },
+    {
+      "epoch": 37.60200668896321,
+      "grad_norm": 0.6079913377761841,
+      "learning_rate": 7.483870967741936e-05,
+      "loss": 2.8394,
+      "step": 2819
+    },
+    {
+      "epoch": 37.61538461538461,
+      "grad_norm": 0.5867500901222229,
+      "learning_rate": 7.479421579532814e-05,
+      "loss": 2.413,
+      "step": 2820
+    },
+    {
+      "epoch": 37.62876254180602,
+      "grad_norm": 0.6049195528030396,
+      "learning_rate": 7.474972191323694e-05,
+      "loss": 2.64,
+      "step": 2821
+    },
+    {
+      "epoch": 37.642140468227424,
+      "grad_norm": 0.5787717700004578,
+      "learning_rate": 7.470522803114572e-05,
+      "loss": 2.5029,
+      "step": 2822
+    },
+    {
+      "epoch": 37.65551839464883,
+      "grad_norm": 0.6378308534622192,
+      "learning_rate": 7.466073414905452e-05,
+      "loss": 2.6937,
+      "step": 2823
+    },
+    {
+      "epoch": 37.668896321070235,
+      "grad_norm": 0.6016746163368225,
+      "learning_rate": 7.46162402669633e-05,
+      "loss": 2.477,
+      "step": 2824
+    },
+    {
+      "epoch": 37.68227424749164,
+      "grad_norm": 0.5864558219909668,
+      "learning_rate": 7.457174638487208e-05,
+      "loss": 2.7177,
+      "step": 2825
+    },
+    {
+      "epoch": 37.69565217391305,
+      "grad_norm": 0.6352173686027527,
+      "learning_rate": 7.452725250278088e-05,
+      "loss": 2.73,
+      "step": 2826
+    },
+    {
+      "epoch": 37.709030100334445,
+      "grad_norm": 0.6432989239692688,
+      "learning_rate": 7.448275862068966e-05,
+      "loss": 2.5062,
+      "step": 2827
+    },
+    {
+      "epoch": 37.72240802675585,
+      "grad_norm": 0.6061347723007202,
+      "learning_rate": 7.443826473859846e-05,
+      "loss": 2.5776,
+      "step": 2828
+    },
+    {
+      "epoch": 37.735785953177256,
+      "grad_norm": 0.6208527684211731,
+      "learning_rate": 7.439377085650724e-05,
+      "loss": 2.4125,
+      "step": 2829
+    },
+    {
+      "epoch": 37.74916387959866,
+      "grad_norm": 0.6052190661430359,
+      "learning_rate": 7.434927697441602e-05,
+      "loss": 2.5331,
+      "step": 2830
+    },
+    {
+      "epoch": 37.76254180602007,
+      "grad_norm": 0.6361626982688904,
+      "learning_rate": 7.43047830923248e-05,
+      "loss": 2.7675,
+      "step": 2831
+    },
+    {
+      "epoch": 37.77591973244147,
+      "grad_norm": 0.6102336645126343,
+      "learning_rate": 7.426028921023359e-05,
+      "loss": 2.3707,
+      "step": 2832
+    },
+    {
+      "epoch": 37.78929765886288,
+      "grad_norm": 0.6248254776000977,
+      "learning_rate": 7.421579532814238e-05,
+      "loss": 2.8015,
+      "step": 2833
+    },
+    {
+      "epoch": 37.802675585284284,
+      "grad_norm": 0.6449552178382874,
+      "learning_rate": 7.417130144605116e-05,
+      "loss": 2.7157,
+      "step": 2834
+    },
+    {
+      "epoch": 37.81605351170568,
+      "grad_norm": 0.6328997611999512,
+      "learning_rate": 7.412680756395996e-05,
+      "loss": 2.8754,
+      "step": 2835
+    },
+    {
+      "epoch": 37.82943143812709,
+      "grad_norm": 0.617190957069397,
+      "learning_rate": 7.408231368186874e-05,
+      "loss": 2.8825,
+      "step": 2836
+    },
+    {
+      "epoch": 37.84280936454849,
+      "grad_norm": 0.6431034207344055,
+      "learning_rate": 7.403781979977752e-05,
+      "loss": 2.877,
+      "step": 2837
+    },
+    {
+      "epoch": 37.8561872909699,
+      "grad_norm": 0.6334032416343689,
+      "learning_rate": 7.399332591768632e-05,
+      "loss": 2.5787,
+      "step": 2838
+    },
+    {
+      "epoch": 37.869565217391305,
+      "grad_norm": 0.599069356918335,
+      "learning_rate": 7.39488320355951e-05,
+      "loss": 2.6774,
+      "step": 2839
+    },
+    {
+      "epoch": 37.88294314381271,
+      "grad_norm": 0.6226022243499756,
+      "learning_rate": 7.39043381535039e-05,
+      "loss": 2.7821,
+      "step": 2840
+    },
+    {
+      "epoch": 37.896321070234116,
+      "grad_norm": 0.6168680191040039,
+      "learning_rate": 7.385984427141268e-05,
+      "loss": 2.6552,
+      "step": 2841
+    },
+    {
+      "epoch": 37.90969899665552,
+      "grad_norm": 0.5987963080406189,
+      "learning_rate": 7.381535038932146e-05,
+      "loss": 2.6139,
+      "step": 2842
+    },
+    {
+      "epoch": 37.92307692307692,
+      "grad_norm": 0.5855274796485901,
+      "learning_rate": 7.377085650723026e-05,
+      "loss": 2.687,
+      "step": 2843
+    },
+    {
+      "epoch": 37.936454849498325,
+      "grad_norm": 0.6431803703308105,
+      "learning_rate": 7.372636262513904e-05,
+      "loss": 2.7411,
+      "step": 2844
+    },
+    {
+      "epoch": 37.94983277591973,
+      "grad_norm": 0.6025601029396057,
+      "learning_rate": 7.368186874304784e-05,
+      "loss": 2.6054,
+      "step": 2845
+    },
+    {
+      "epoch": 37.96321070234114,
+      "grad_norm": 0.6150123476982117,
+      "learning_rate": 7.363737486095662e-05,
+      "loss": 2.7483,
+      "step": 2846
+    },
+    {
+      "epoch": 37.97658862876254,
+      "grad_norm": 0.5857491493225098,
+      "learning_rate": 7.35928809788654e-05,
+      "loss": 2.68,
+      "step": 2847
+    },
+    {
+      "epoch": 37.98996655518395,
+      "grad_norm": 0.6244483590126038,
+      "learning_rate": 7.35483870967742e-05,
+      "loss": 2.7171,
+      "step": 2848
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 0.7437806725502014,
+      "learning_rate": 7.350389321468298e-05,
+      "loss": 2.5646,
+      "step": 2849
+    },
+    {
+      "epoch": 38.013377926421406,
+      "grad_norm": 0.6323840618133545,
+      "learning_rate": 7.345939933259178e-05,
+      "loss": 2.7457,
+      "step": 2850
+    },
+    {
+      "epoch": 38.02675585284281,
+      "grad_norm": 0.6000533103942871,
+      "learning_rate": 7.341490545050056e-05,
+      "loss": 2.6829,
+      "step": 2851
+    },
+    {
+      "epoch": 38.04013377926422,
+      "grad_norm": 0.621135950088501,
+      "learning_rate": 7.337041156840934e-05,
+      "loss": 2.795,
+      "step": 2852
+    },
+    {
+      "epoch": 38.05351170568562,
+      "grad_norm": 0.6256486773490906,
+      "learning_rate": 7.332591768631814e-05,
+      "loss": 2.6762,
+      "step": 2853
+    },
+    {
+      "epoch": 38.06688963210702,
+      "grad_norm": 0.6535592079162598,
+      "learning_rate": 7.328142380422692e-05,
+      "loss": 2.8339,
+      "step": 2854
+    },
+    {
+      "epoch": 38.080267558528426,
+      "grad_norm": 0.6324830055236816,
+      "learning_rate": 7.323692992213572e-05,
+      "loss": 2.6129,
+      "step": 2855
+    },
+    {
+      "epoch": 38.09364548494983,
+      "grad_norm": 0.5966958403587341,
+      "learning_rate": 7.31924360400445e-05,
+      "loss": 2.3738,
+      "step": 2856
+    },
+    {
+      "epoch": 38.10702341137124,
+      "grad_norm": 0.6118017435073853,
+      "learning_rate": 7.314794215795328e-05,
+      "loss": 2.5778,
+      "step": 2857
+    },
+    {
+      "epoch": 38.12040133779264,
+      "grad_norm": 0.5761687159538269,
+      "learning_rate": 7.310344827586208e-05,
+      "loss": 2.347,
+      "step": 2858
+    },
+    {
+      "epoch": 38.13377926421405,
+      "grad_norm": 0.6576936841011047,
+      "learning_rate": 7.305895439377086e-05,
+      "loss": 2.6811,
+      "step": 2859
+    },
+    {
+      "epoch": 38.147157190635454,
+      "grad_norm": 0.6086276173591614,
+      "learning_rate": 7.301446051167965e-05,
+      "loss": 2.713,
+      "step": 2860
+    },
+    {
+      "epoch": 38.16053511705686,
+      "grad_norm": 0.5824201107025146,
+      "learning_rate": 7.296996662958844e-05,
+      "loss": 2.457,
+      "step": 2861
+    },
+    {
+      "epoch": 38.17391304347826,
+      "grad_norm": 0.6397709250450134,
+      "learning_rate": 7.292547274749722e-05,
+      "loss": 2.5384,
+      "step": 2862
+    },
+    {
+      "epoch": 38.187290969899664,
+      "grad_norm": 0.6120484471321106,
+      "learning_rate": 7.288097886540602e-05,
+      "loss": 2.5742,
+      "step": 2863
+    },
+    {
+      "epoch": 38.20066889632107,
+      "grad_norm": 0.6424990892410278,
+      "learning_rate": 7.28364849833148e-05,
+      "loss": 2.8879,
+      "step": 2864
+    },
+    {
+      "epoch": 38.214046822742475,
+      "grad_norm": 0.610888659954071,
+      "learning_rate": 7.27919911012236e-05,
+      "loss": 2.5114,
+      "step": 2865
+    },
+    {
+      "epoch": 38.22742474916388,
+      "grad_norm": 0.5722895264625549,
+      "learning_rate": 7.274749721913238e-05,
+      "loss": 2.1089,
+      "step": 2866
+    },
+    {
+      "epoch": 38.240802675585286,
+      "grad_norm": 0.6000377535820007,
+      "learning_rate": 7.270300333704116e-05,
+      "loss": 2.3571,
+      "step": 2867
+    },
+    {
+      "epoch": 38.25418060200669,
+      "grad_norm": 0.6441863775253296,
+      "learning_rate": 7.265850945494995e-05,
+      "loss": 2.828,
+      "step": 2868
+    },
+    {
+      "epoch": 38.2675585284281,
+      "grad_norm": 0.6954602003097534,
+      "learning_rate": 7.261401557285874e-05,
+      "loss": 2.586,
+      "step": 2869
+    },
+    {
+      "epoch": 38.280936454849495,
+      "grad_norm": 0.6308737397193909,
+      "learning_rate": 7.256952169076753e-05,
+      "loss": 2.558,
+      "step": 2870
+    },
+    {
+      "epoch": 38.2943143812709,
+      "grad_norm": 0.5915984511375427,
+      "learning_rate": 7.252502780867632e-05,
+      "loss": 2.6369,
+      "step": 2871
+    },
+    {
+      "epoch": 38.30769230769231,
+      "grad_norm": 0.5993978381156921,
+      "learning_rate": 7.24805339265851e-05,
+      "loss": 2.5857,
+      "step": 2872
+    },
+    {
+      "epoch": 38.32107023411371,
+      "grad_norm": 0.5899291038513184,
+      "learning_rate": 7.243604004449388e-05,
+      "loss": 2.396,
+      "step": 2873
+    },
+    {
+      "epoch": 38.33444816053512,
+      "grad_norm": 0.6214590668678284,
+      "learning_rate": 7.239154616240266e-05,
+      "loss": 2.3691,
+      "step": 2874
+    },
+    {
+      "epoch": 38.34782608695652,
+      "grad_norm": 0.5765502452850342,
+      "learning_rate": 7.234705228031146e-05,
+      "loss": 2.5417,
+      "step": 2875
+    },
+    {
+      "epoch": 38.36120401337793,
+      "grad_norm": 0.6192215085029602,
+      "learning_rate": 7.230255839822024e-05,
+      "loss": 2.459,
+      "step": 2876
+    },
+    {
+      "epoch": 38.374581939799334,
+      "grad_norm": 0.6413353085517883,
+      "learning_rate": 7.225806451612904e-05,
+      "loss": 2.5345,
+      "step": 2877
+    },
+    {
+      "epoch": 38.38795986622073,
+      "grad_norm": 0.6169536709785461,
+      "learning_rate": 7.221357063403782e-05,
+      "loss": 2.5972,
+      "step": 2878
+    },
+    {
+      "epoch": 38.40133779264214,
+      "grad_norm": 0.6331331133842468,
+      "learning_rate": 7.21690767519466e-05,
+      "loss": 2.625,
+      "step": 2879
+    },
+    {
+      "epoch": 38.414715719063544,
+      "grad_norm": 0.6282365918159485,
+      "learning_rate": 7.21245828698554e-05,
+      "loss": 2.6334,
+      "step": 2880
+    },
+    {
+      "epoch": 38.42809364548495,
+      "grad_norm": 0.6211024522781372,
+      "learning_rate": 7.208008898776418e-05,
+      "loss": 2.4833,
+      "step": 2881
+    },
+    {
+      "epoch": 38.441471571906355,
+      "grad_norm": 0.6600573062896729,
+      "learning_rate": 7.203559510567298e-05,
+      "loss": 2.7515,
+      "step": 2882
+    },
+    {
+      "epoch": 38.45484949832776,
+      "grad_norm": 0.6234501004219055,
+      "learning_rate": 7.199110122358176e-05,
+      "loss": 2.4914,
+      "step": 2883
+    },
+    {
+      "epoch": 38.468227424749166,
+      "grad_norm": 0.6104183197021484,
+      "learning_rate": 7.194660734149054e-05,
+      "loss": 2.6202,
+      "step": 2884
+    },
+    {
+      "epoch": 38.48160535117057,
+      "grad_norm": 0.6139323711395264,
+      "learning_rate": 7.190211345939934e-05,
+      "loss": 2.6918,
+      "step": 2885
+    },
+    {
+      "epoch": 38.49498327759197,
+      "grad_norm": 0.6672762632369995,
+      "learning_rate": 7.185761957730812e-05,
+      "loss": 2.7207,
+      "step": 2886
+    },
+    {
+      "epoch": 38.508361204013376,
+      "grad_norm": 0.636197030544281,
+      "learning_rate": 7.181312569521691e-05,
+      "loss": 2.8393,
+      "step": 2887
+    },
+    {
+      "epoch": 38.52173913043478,
+      "grad_norm": 0.6464013457298279,
+      "learning_rate": 7.17686318131257e-05,
+      "loss": 2.6601,
+      "step": 2888
+    },
+    {
+      "epoch": 38.53511705685619,
+      "grad_norm": 0.5882225632667542,
+      "learning_rate": 7.172413793103448e-05,
+      "loss": 2.4206,
+      "step": 2889
+    },
+    {
+      "epoch": 38.54849498327759,
+      "grad_norm": 0.6114091873168945,
+      "learning_rate": 7.167964404894328e-05,
+      "loss": 2.6474,
+      "step": 2890
+    },
+    {
+      "epoch": 38.561872909699,
+      "grad_norm": 0.5976521968841553,
+      "learning_rate": 7.163515016685206e-05,
+      "loss": 2.5393,
+      "step": 2891
+    },
+    {
+      "epoch": 38.575250836120404,
+      "grad_norm": 0.6302757859230042,
+      "learning_rate": 7.159065628476085e-05,
+      "loss": 2.5687,
+      "step": 2892
+    },
+    {
+      "epoch": 38.58862876254181,
+      "grad_norm": 0.6598967909812927,
+      "learning_rate": 7.154616240266964e-05,
+      "loss": 2.8124,
+      "step": 2893
+    },
+    {
+      "epoch": 38.60200668896321,
+      "grad_norm": 0.6214116215705872,
+      "learning_rate": 7.150166852057842e-05,
+      "loss": 2.6531,
+      "step": 2894
+    },
+    {
+      "epoch": 38.61538461538461,
+      "grad_norm": 0.6264208555221558,
+      "learning_rate": 7.145717463848721e-05,
+      "loss": 2.6821,
+      "step": 2895
+    },
+    {
+      "epoch": 38.62876254180602,
+      "grad_norm": 0.647375226020813,
+      "learning_rate": 7.1412680756396e-05,
+      "loss": 2.5769,
+      "step": 2896
+    },
+    {
+      "epoch": 38.642140468227424,
+      "grad_norm": 0.6418511271476746,
+      "learning_rate": 7.136818687430479e-05,
+      "loss": 2.682,
+      "step": 2897
+    },
+    {
+      "epoch": 38.65551839464883,
+      "grad_norm": 0.6308450102806091,
+      "learning_rate": 7.132369299221358e-05,
+      "loss": 2.6151,
+      "step": 2898
+    },
+    {
+      "epoch": 38.668896321070235,
+      "grad_norm": 0.6675850749015808,
+      "learning_rate": 7.127919911012236e-05,
+      "loss": 2.7805,
+      "step": 2899
+    },
+    {
+      "epoch": 38.68227424749164,
+      "grad_norm": 0.5915642380714417,
+      "learning_rate": 7.123470522803115e-05,
+      "loss": 2.6036,
+      "step": 2900
+    },
+    {
+      "epoch": 38.69565217391305,
+      "grad_norm": 0.6425783038139343,
+      "learning_rate": 7.119021134593994e-05,
+      "loss": 2.805,
+      "step": 2901
+    },
+    {
+      "epoch": 38.709030100334445,
+      "grad_norm": 0.613692045211792,
+      "learning_rate": 7.114571746384873e-05,
+      "loss": 2.5263,
+      "step": 2902
+    },
+    {
+      "epoch": 38.72240802675585,
+      "grad_norm": 0.7906155586242676,
+      "learning_rate": 7.110122358175751e-05,
+      "loss": 2.7032,
+      "step": 2903
+    },
+    {
+      "epoch": 38.735785953177256,
+      "grad_norm": 0.6317926645278931,
+      "learning_rate": 7.10567296996663e-05,
+      "loss": 2.7303,
+      "step": 2904
+    },
+    {
+      "epoch": 38.74916387959866,
+      "grad_norm": 0.6563540101051331,
+      "learning_rate": 7.101223581757509e-05,
+      "loss": 2.6913,
+      "step": 2905
+    },
+    {
+      "epoch": 38.76254180602007,
+      "grad_norm": 0.6583380699157715,
+      "learning_rate": 7.096774193548388e-05,
+      "loss": 2.8355,
+      "step": 2906
+    },
+    {
+      "epoch": 38.77591973244147,
+      "grad_norm": 0.6026883125305176,
+      "learning_rate": 7.092324805339267e-05,
+      "loss": 2.4306,
+      "step": 2907
+    },
+    {
+      "epoch": 38.78929765886288,
+      "grad_norm": 0.6352461576461792,
+      "learning_rate": 7.087875417130145e-05,
+      "loss": 2.7963,
+      "step": 2908
+    },
+    {
+      "epoch": 38.802675585284284,
+      "grad_norm": 0.6036055684089661,
+      "learning_rate": 7.083426028921024e-05,
+      "loss": 2.6946,
+      "step": 2909
+    },
+    {
+      "epoch": 38.81605351170568,
+      "grad_norm": 0.5911431312561035,
+      "learning_rate": 7.078976640711903e-05,
+      "loss": 2.5936,
+      "step": 2910
+    },
+    {
+      "epoch": 38.82943143812709,
+      "grad_norm": 0.6125901341438293,
+      "learning_rate": 7.074527252502781e-05,
+      "loss": 2.4901,
+      "step": 2911
+    },
+    {
+      "epoch": 38.84280936454849,
+      "grad_norm": 0.600537896156311,
+      "learning_rate": 7.070077864293661e-05,
+      "loss": 2.6603,
+      "step": 2912
+    },
+    {
+      "epoch": 38.8561872909699,
+      "grad_norm": 0.6200169324874878,
+      "learning_rate": 7.065628476084539e-05,
+      "loss": 2.629,
+      "step": 2913
+    },
+    {
+      "epoch": 38.869565217391305,
+      "grad_norm": 0.5789446830749512,
+      "learning_rate": 7.061179087875418e-05,
+      "loss": 2.3878,
+      "step": 2914
+    },
+    {
+      "epoch": 38.88294314381271,
+      "grad_norm": 0.6316003799438477,
+      "learning_rate": 7.056729699666296e-05,
+      "loss": 2.4585,
+      "step": 2915
+    },
+    {
+      "epoch": 38.896321070234116,
+      "grad_norm": 0.6021048426628113,
+      "learning_rate": 7.052280311457174e-05,
+      "loss": 2.4609,
+      "step": 2916
+    },
+    {
+      "epoch": 38.90969899665552,
+      "grad_norm": 0.6088526844978333,
+      "learning_rate": 7.047830923248054e-05,
+      "loss": 2.6102,
+      "step": 2917
+    },
+    {
+      "epoch": 38.92307692307692,
+      "grad_norm": 0.658087432384491,
+      "learning_rate": 7.043381535038932e-05,
+      "loss": 2.6058,
+      "step": 2918
+    },
+    {
+      "epoch": 38.936454849498325,
+      "grad_norm": 0.6708016991615295,
+      "learning_rate": 7.038932146829811e-05,
+      "loss": 2.9315,
+      "step": 2919
+    },
+    {
+      "epoch": 38.94983277591973,
+      "grad_norm": 0.6475014686584473,
+      "learning_rate": 7.03448275862069e-05,
+      "loss": 2.8692,
+      "step": 2920
+    },
+    {
+      "epoch": 38.96321070234114,
+      "grad_norm": 0.6264966726303101,
+      "learning_rate": 7.030033370411568e-05,
+      "loss": 2.7866,
+      "step": 2921
+    },
+    {
+      "epoch": 38.97658862876254,
+      "grad_norm": 0.6225918531417847,
+      "learning_rate": 7.025583982202447e-05,
+      "loss": 2.4795,
+      "step": 2922
+    },
+    {
+      "epoch": 38.98996655518395,
+      "grad_norm": 0.6052890419960022,
+      "learning_rate": 7.021134593993326e-05,
+      "loss": 2.2833,
+      "step": 2923
+    },
+    {
+      "epoch": 39.0,
+      "grad_norm": 0.695241391658783,
+      "learning_rate": 7.016685205784205e-05,
+      "loss": 2.4251,
+      "step": 2924
+    },
+    {
+      "epoch": 39.013377926421406,
+      "grad_norm": 0.5891976952552795,
+      "learning_rate": 7.012235817575084e-05,
+      "loss": 2.561,
+      "step": 2925
+    },
+    {
+      "epoch": 39.02675585284281,
+      "grad_norm": 0.6551567316055298,
+      "learning_rate": 7.007786429365962e-05,
+      "loss": 2.4206,
+      "step": 2926
+    },
+    {
+      "epoch": 39.04013377926422,
+      "grad_norm": 0.5852656364440918,
+      "learning_rate": 7.003337041156841e-05,
+      "loss": 2.4159,
+      "step": 2927
+    },
+    {
+      "epoch": 39.05351170568562,
+      "grad_norm": 0.5813770294189453,
+      "learning_rate": 6.99888765294772e-05,
+      "loss": 2.5426,
+      "step": 2928
+    },
+    {
+      "epoch": 39.06688963210702,
+      "grad_norm": 0.5840299725532532,
+      "learning_rate": 6.994438264738599e-05,
+      "loss": 2.5013,
+      "step": 2929
+    },
+    {
+      "epoch": 39.080267558528426,
+      "grad_norm": 0.5940836071968079,
+      "learning_rate": 6.989988876529477e-05,
+      "loss": 2.4776,
+      "step": 2930
+    },
+    {
+      "epoch": 39.09364548494983,
+      "grad_norm": 0.6290647387504578,
+      "learning_rate": 6.985539488320356e-05,
+      "loss": 2.6289,
+      "step": 2931
+    },
+    {
+      "epoch": 39.10702341137124,
+      "grad_norm": 0.6070288419723511,
+      "learning_rate": 6.981090100111235e-05,
+      "loss": 2.3952,
+      "step": 2932
+    },
+    {
+      "epoch": 39.12040133779264,
+      "grad_norm": 0.6416279077529907,
+      "learning_rate": 6.976640711902114e-05,
+      "loss": 2.8001,
+      "step": 2933
+    },
+    {
+      "epoch": 39.13377926421405,
+      "grad_norm": 0.6238275170326233,
+      "learning_rate": 6.972191323692993e-05,
+      "loss": 2.5927,
+      "step": 2934
+    },
+    {
+      "epoch": 39.147157190635454,
+      "grad_norm": 0.6179841756820679,
+      "learning_rate": 6.967741935483871e-05,
+      "loss": 2.4691,
+      "step": 2935
+    },
+    {
+      "epoch": 39.16053511705686,
+      "grad_norm": 0.5916646122932434,
+      "learning_rate": 6.96329254727475e-05,
+      "loss": 2.4734,
+      "step": 2936
+    },
+    {
+      "epoch": 39.17391304347826,
+      "grad_norm": 0.6340898275375366,
+      "learning_rate": 6.958843159065629e-05,
+      "loss": 2.6764,
+      "step": 2937
+    },
+    {
+      "epoch": 39.187290969899664,
+      "grad_norm": 0.6154446005821228,
+      "learning_rate": 6.954393770856507e-05,
+      "loss": 2.666,
+      "step": 2938
+    },
+    {
+      "epoch": 39.20066889632107,
+      "grad_norm": 0.6166310906410217,
+      "learning_rate": 6.949944382647387e-05,
+      "loss": 2.5198,
+      "step": 2939
+    },
+    {
+      "epoch": 39.214046822742475,
+      "grad_norm": 0.6393052935600281,
+      "learning_rate": 6.945494994438265e-05,
+      "loss": 2.6016,
+      "step": 2940
+    },
+    {
+      "epoch": 39.22742474916388,
+      "grad_norm": 0.6041097640991211,
+      "learning_rate": 6.941045606229144e-05,
+      "loss": 2.462,
+      "step": 2941
+    },
+    {
+      "epoch": 39.240802675585286,
+      "grad_norm": 0.6157374382019043,
+      "learning_rate": 6.936596218020023e-05,
+      "loss": 2.5844,
+      "step": 2942
+    },
+    {
+      "epoch": 39.25418060200669,
+      "grad_norm": 0.641880452632904,
+      "learning_rate": 6.932146829810901e-05,
+      "loss": 2.7159,
+      "step": 2943
+    },
+    {
+      "epoch": 39.2675585284281,
+      "grad_norm": 0.6598557233810425,
+      "learning_rate": 6.927697441601781e-05,
+      "loss": 2.5649,
+      "step": 2944
+    },
+    {
+      "epoch": 39.280936454849495,
+      "grad_norm": 0.6511268615722656,
+      "learning_rate": 6.923248053392659e-05,
+      "loss": 2.7623,
+      "step": 2945
+    },
+    {
+      "epoch": 39.2943143812709,
+      "grad_norm": 0.6559281945228577,
+      "learning_rate": 6.918798665183537e-05,
+      "loss": 2.3524,
+      "step": 2946
+    },
+    {
+      "epoch": 39.30769230769231,
+      "grad_norm": 0.601424515247345,
+      "learning_rate": 6.914349276974417e-05,
+      "loss": 2.2888,
+      "step": 2947
+    },
+    {
+      "epoch": 39.32107023411371,
+      "grad_norm": 0.6525271534919739,
+      "learning_rate": 6.909899888765295e-05,
+      "loss": 2.4266,
+      "step": 2948
+    },
+    {
+      "epoch": 39.33444816053512,
+      "grad_norm": 0.6467943787574768,
+      "learning_rate": 6.905450500556175e-05,
+      "loss": 2.654,
+      "step": 2949
+    },
+    {
+      "epoch": 39.34782608695652,
+      "grad_norm": 0.6527394652366638,
+      "learning_rate": 6.901001112347053e-05,
+      "loss": 2.6237,
+      "step": 2950
+    },
+    {
+      "epoch": 39.36120401337793,
+      "grad_norm": 0.615578830242157,
+      "learning_rate": 6.896551724137931e-05,
+      "loss": 2.6382,
+      "step": 2951
+    },
+    {
+      "epoch": 39.374581939799334,
+      "grad_norm": 0.6257203221321106,
+      "learning_rate": 6.892102335928811e-05,
+      "loss": 2.8014,
+      "step": 2952
+    },
+    {
+      "epoch": 39.38795986622073,
+      "grad_norm": 0.6732933521270752,
+      "learning_rate": 6.887652947719689e-05,
+      "loss": 2.7796,
+      "step": 2953
+    },
+    {
+      "epoch": 39.40133779264214,
+      "grad_norm": 0.5996230840682983,
+      "learning_rate": 6.883203559510569e-05,
+      "loss": 2.4852,
+      "step": 2954
+    },
+    {
+      "epoch": 39.414715719063544,
+      "grad_norm": 0.6640390157699585,
+      "learning_rate": 6.878754171301446e-05,
+      "loss": 2.4979,
+      "step": 2955
+    },
+    {
+      "epoch": 39.42809364548495,
+      "grad_norm": 0.6220453381538391,
+      "learning_rate": 6.874304783092325e-05,
+      "loss": 2.4849,
+      "step": 2956
+    },
+    {
+      "epoch": 39.441471571906355,
+      "grad_norm": 0.6283277273178101,
+      "learning_rate": 6.869855394883203e-05,
+      "loss": 2.5193,
+      "step": 2957
+    },
+    {
+      "epoch": 39.45484949832776,
+      "grad_norm": 0.6696292757987976,
+      "learning_rate": 6.865406006674082e-05,
+      "loss": 2.6485,
+      "step": 2958
+    },
+    {
+      "epoch": 39.468227424749166,
+      "grad_norm": 0.6544603705406189,
+      "learning_rate": 6.860956618464961e-05,
+      "loss": 2.5956,
+      "step": 2959
+    },
+    {
+      "epoch": 39.48160535117057,
+      "grad_norm": 0.6541810631752014,
+      "learning_rate": 6.85650723025584e-05,
+      "loss": 2.502,
+      "step": 2960
+    },
+    {
+      "epoch": 39.49498327759197,
+      "grad_norm": 0.6799680590629578,
+      "learning_rate": 6.852057842046719e-05,
+      "loss": 2.7545,
+      "step": 2961
+    },
+    {
+      "epoch": 39.508361204013376,
+      "grad_norm": 0.6350213289260864,
+      "learning_rate": 6.847608453837597e-05,
+      "loss": 2.707,
+      "step": 2962
+    },
+    {
+      "epoch": 39.52173913043478,
+      "grad_norm": 0.5974067449569702,
+      "learning_rate": 6.843159065628476e-05,
+      "loss": 2.5572,
+      "step": 2963
+    },
+    {
+      "epoch": 39.53511705685619,
+      "grad_norm": 0.5829086899757385,
+      "learning_rate": 6.838709677419355e-05,
+      "loss": 2.5003,
+      "step": 2964
+    },
+    {
+      "epoch": 39.54849498327759,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 6.834260289210233e-05,
+      "loss": 2.5413,
+      "step": 2965
+    },
+    {
+      "epoch": 39.561872909699,
+      "grad_norm": 0.6405021548271179,
+      "learning_rate": 6.829810901001113e-05,
+      "loss": 2.4371,
+      "step": 2966
+    },
+    {
+      "epoch": 39.575250836120404,
+      "grad_norm": 0.6281189322471619,
+      "learning_rate": 6.825361512791991e-05,
+      "loss": 2.6484,
+      "step": 2967
+    },
+    {
+      "epoch": 39.58862876254181,
+      "grad_norm": 0.7368245720863342,
+      "learning_rate": 6.82091212458287e-05,
+      "loss": 2.5467,
+      "step": 2968
+    },
+    {
+      "epoch": 39.60200668896321,
+      "grad_norm": 0.6534647941589355,
+      "learning_rate": 6.816462736373749e-05,
+      "loss": 2.7322,
+      "step": 2969
+    },
+    {
+      "epoch": 39.61538461538461,
+      "grad_norm": 0.821247935295105,
+      "learning_rate": 6.812013348164627e-05,
+      "loss": 2.4748,
+      "step": 2970
+    },
+    {
+      "epoch": 39.62876254180602,
+      "grad_norm": 0.6398321390151978,
+      "learning_rate": 6.807563959955507e-05,
+      "loss": 2.6597,
+      "step": 2971
+    },
+    {
+      "epoch": 39.642140468227424,
+      "grad_norm": 0.6529011130332947,
+      "learning_rate": 6.803114571746385e-05,
+      "loss": 2.5417,
+      "step": 2972
+    },
+    {
+      "epoch": 39.65551839464883,
+      "grad_norm": 0.6198508143424988,
+      "learning_rate": 6.798665183537263e-05,
+      "loss": 2.4684,
+      "step": 2973
+    },
+    {
+      "epoch": 39.668896321070235,
+      "grad_norm": 0.6332361698150635,
+      "learning_rate": 6.794215795328143e-05,
+      "loss": 2.4961,
+      "step": 2974
+    },
+    {
+      "epoch": 39.68227424749164,
+      "grad_norm": 0.6088337898254395,
+      "learning_rate": 6.789766407119021e-05,
+      "loss": 2.6169,
+      "step": 2975
+    },
+    {
+      "epoch": 39.69565217391305,
+      "grad_norm": 0.6401640176773071,
+      "learning_rate": 6.785317018909901e-05,
+      "loss": 2.4942,
+      "step": 2976
+    },
+    {
+      "epoch": 39.709030100334445,
+      "grad_norm": 0.6432260870933533,
+      "learning_rate": 6.780867630700779e-05,
+      "loss": 2.4886,
+      "step": 2977
+    },
+    {
+      "epoch": 39.72240802675585,
+      "grad_norm": 0.6184273362159729,
+      "learning_rate": 6.776418242491657e-05,
+      "loss": 2.6445,
+      "step": 2978
+    },
+    {
+      "epoch": 39.735785953177256,
+      "grad_norm": 0.6584762334823608,
+      "learning_rate": 6.771968854282537e-05,
+      "loss": 2.5727,
+      "step": 2979
+    },
+    {
+      "epoch": 39.74916387959866,
+      "grad_norm": 0.577503502368927,
+      "learning_rate": 6.767519466073415e-05,
+      "loss": 2.3277,
+      "step": 2980
+    },
+    {
+      "epoch": 39.76254180602007,
+      "grad_norm": 0.62856525182724,
+      "learning_rate": 6.763070077864295e-05,
+      "loss": 2.7912,
+      "step": 2981
+    },
+    {
+      "epoch": 39.77591973244147,
+      "grad_norm": 0.6405977606773376,
+      "learning_rate": 6.758620689655173e-05,
+      "loss": 2.6165,
+      "step": 2982
+    },
+    {
+      "epoch": 39.78929765886288,
+      "grad_norm": 0.5938370823860168,
+      "learning_rate": 6.754171301446051e-05,
+      "loss": 2.6061,
+      "step": 2983
+    },
+    {
+      "epoch": 39.802675585284284,
+      "grad_norm": 0.6175585985183716,
+      "learning_rate": 6.749721913236931e-05,
+      "loss": 2.4875,
+      "step": 2984
+    },
+    {
+      "epoch": 39.81605351170568,
+      "grad_norm": 0.7778329253196716,
+      "learning_rate": 6.745272525027809e-05,
+      "loss": 2.8471,
+      "step": 2985
+    },
+    {
+      "epoch": 39.82943143812709,
+      "grad_norm": 0.6055701375007629,
+      "learning_rate": 6.740823136818689e-05,
+      "loss": 2.4975,
+      "step": 2986
+    },
+    {
+      "epoch": 39.84280936454849,
+      "grad_norm": 0.68842613697052,
+      "learning_rate": 6.736373748609567e-05,
+      "loss": 2.7126,
+      "step": 2987
+    },
+    {
+      "epoch": 39.8561872909699,
+      "grad_norm": 0.6329212784767151,
+      "learning_rate": 6.731924360400445e-05,
+      "loss": 2.6741,
+      "step": 2988
+    },
+    {
+      "epoch": 39.869565217391305,
+      "grad_norm": 0.6239383220672607,
+      "learning_rate": 6.727474972191325e-05,
+      "loss": 2.6681,
+      "step": 2989
+    },
+    {
+      "epoch": 39.88294314381271,
+      "grad_norm": 0.6401740908622742,
+      "learning_rate": 6.723025583982203e-05,
+      "loss": 2.7272,
+      "step": 2990
+    },
+    {
+      "epoch": 39.896321070234116,
+      "grad_norm": 0.6988689303398132,
+      "learning_rate": 6.718576195773083e-05,
+      "loss": 2.7339,
+      "step": 2991
+    },
+    {
+      "epoch": 39.90969899665552,
+      "grad_norm": 0.6372131705284119,
+      "learning_rate": 6.714126807563961e-05,
+      "loss": 2.4379,
+      "step": 2992
+    },
+    {
+      "epoch": 39.92307692307692,
+      "grad_norm": 0.6818649172782898,
+      "learning_rate": 6.709677419354839e-05,
+      "loss": 2.8075,
+      "step": 2993
+    },
+    {
+      "epoch": 39.936454849498325,
+      "grad_norm": 0.653063952922821,
+      "learning_rate": 6.705228031145719e-05,
+      "loss": 2.5813,
+      "step": 2994
+    },
+    {
+      "epoch": 39.94983277591973,
+      "grad_norm": 0.657010018825531,
+      "learning_rate": 6.700778642936597e-05,
+      "loss": 2.7227,
+      "step": 2995
+    },
+    {
+      "epoch": 39.96321070234114,
+      "grad_norm": 0.6535374522209167,
+      "learning_rate": 6.696329254727475e-05,
+      "loss": 2.7369,
+      "step": 2996
+    },
+    {
+      "epoch": 39.97658862876254,
+      "grad_norm": 0.6069753766059875,
+      "learning_rate": 6.691879866518353e-05,
+      "loss": 2.5248,
+      "step": 2997
+    },
+    {
+      "epoch": 39.98996655518395,
+      "grad_norm": 0.6699077486991882,
+      "learning_rate": 6.687430478309233e-05,
+      "loss": 2.6428,
+      "step": 2998
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 0.7744498252868652,
+      "learning_rate": 6.682981090100111e-05,
+      "loss": 2.7199,
+      "step": 2999
+    },
+    {
+      "epoch": 40.013377926421406,
+      "grad_norm": 0.6341271996498108,
+      "learning_rate": 6.67853170189099e-05,
+      "loss": 2.5324,
+      "step": 3000
+    },
+    {
+      "epoch": 40.02675585284281,
+      "grad_norm": 0.6379808783531189,
+      "learning_rate": 6.674082313681869e-05,
+      "loss": 2.6118,
+      "step": 3001
+    },
+    {
+      "epoch": 40.04013377926422,
+      "grad_norm": 0.6141436696052551,
+      "learning_rate": 6.669632925472747e-05,
+      "loss": 2.5697,
+      "step": 3002
+    },
+    {
+      "epoch": 40.05351170568562,
+      "grad_norm": 0.6789229512214661,
+      "learning_rate": 6.665183537263627e-05,
+      "loss": 2.6907,
+      "step": 3003
+    },
+    {
+      "epoch": 40.06688963210702,
+      "grad_norm": 0.6370909214019775,
+      "learning_rate": 6.660734149054505e-05,
+      "loss": 2.7101,
+      "step": 3004
+    },
+    {
+      "epoch": 40.080267558528426,
+      "grad_norm": 0.6512884497642517,
+      "learning_rate": 6.656284760845383e-05,
+      "loss": 2.6242,
+      "step": 3005
+    },
+    {
+      "epoch": 40.09364548494983,
+      "grad_norm": 0.6215392351150513,
+      "learning_rate": 6.651835372636263e-05,
+      "loss": 2.513,
+      "step": 3006
+    },
+    {
+      "epoch": 40.10702341137124,
+      "grad_norm": 0.634795606136322,
+      "learning_rate": 6.647385984427141e-05,
+      "loss": 2.6536,
+      "step": 3007
+    },
+    {
+      "epoch": 40.12040133779264,
+      "grad_norm": 0.6299925446510315,
+      "learning_rate": 6.642936596218021e-05,
+      "loss": 2.5797,
+      "step": 3008
+    },
+    {
+      "epoch": 40.13377926421405,
+      "grad_norm": 0.602271318435669,
+      "learning_rate": 6.638487208008899e-05,
+      "loss": 2.4958,
+      "step": 3009
+    },
+    {
+      "epoch": 40.147157190635454,
+      "grad_norm": 0.6046922206878662,
+      "learning_rate": 6.634037819799777e-05,
+      "loss": 2.591,
+      "step": 3010
+    },
+    {
+      "epoch": 40.16053511705686,
+      "grad_norm": 0.6569716334342957,
+      "learning_rate": 6.629588431590657e-05,
+      "loss": 2.7014,
+      "step": 3011
+    },
+    {
+      "epoch": 40.17391304347826,
+      "grad_norm": 0.8074204325675964,
+      "learning_rate": 6.625139043381535e-05,
+      "loss": 2.38,
+      "step": 3012
+    },
+    {
+      "epoch": 40.187290969899664,
+      "grad_norm": 0.6500153541564941,
+      "learning_rate": 6.620689655172415e-05,
+      "loss": 2.5114,
+      "step": 3013
+    },
+    {
+      "epoch": 40.20066889632107,
+      "grad_norm": 0.6205062866210938,
+      "learning_rate": 6.616240266963293e-05,
+      "loss": 2.4542,
+      "step": 3014
+    },
+    {
+      "epoch": 40.214046822742475,
+      "grad_norm": 0.6093875169754028,
+      "learning_rate": 6.611790878754171e-05,
+      "loss": 2.4217,
+      "step": 3015
+    },
+    {
+      "epoch": 40.22742474916388,
+      "grad_norm": 0.6581112146377563,
+      "learning_rate": 6.607341490545051e-05,
+      "loss": 2.5728,
+      "step": 3016
+    },
+    {
+      "epoch": 40.240802675585286,
+      "grad_norm": 0.6695924401283264,
+      "learning_rate": 6.602892102335929e-05,
+      "loss": 2.7236,
+      "step": 3017
+    },
+    {
+      "epoch": 40.25418060200669,
+      "grad_norm": 0.7370408773422241,
+      "learning_rate": 6.598442714126809e-05,
+      "loss": 2.4101,
+      "step": 3018
+    },
+    {
+      "epoch": 40.2675585284281,
+      "grad_norm": 0.6048701405525208,
+      "learning_rate": 6.593993325917687e-05,
+      "loss": 2.517,
+      "step": 3019
+    },
+    {
+      "epoch": 40.280936454849495,
+      "grad_norm": 0.628162682056427,
+      "learning_rate": 6.589543937708565e-05,
+      "loss": 2.3851,
+      "step": 3020
+    },
+    {
+      "epoch": 40.2943143812709,
+      "grad_norm": 0.6344666481018066,
+      "learning_rate": 6.585094549499445e-05,
+      "loss": 2.6034,
+      "step": 3021
+    },
+    {
+      "epoch": 40.30769230769231,
+      "grad_norm": 0.6417535543441772,
+      "learning_rate": 6.580645161290323e-05,
+      "loss": 2.6893,
+      "step": 3022
+    },
+    {
+      "epoch": 40.32107023411371,
+      "grad_norm": 0.6148284077644348,
+      "learning_rate": 6.576195773081202e-05,
+      "loss": 2.4466,
+      "step": 3023
+    },
+    {
+      "epoch": 40.33444816053512,
+      "grad_norm": 0.6240962147712708,
+      "learning_rate": 6.571746384872081e-05,
+      "loss": 2.5427,
+      "step": 3024
+    },
+    {
+      "epoch": 40.34782608695652,
+      "grad_norm": 0.6034325957298279,
+      "learning_rate": 6.567296996662959e-05,
+      "loss": 2.588,
+      "step": 3025
+    },
+    {
+      "epoch": 40.36120401337793,
+      "grad_norm": 0.6440005302429199,
+      "learning_rate": 6.562847608453839e-05,
+      "loss": 2.5972,
+      "step": 3026
+    },
+    {
+      "epoch": 40.374581939799334,
+      "grad_norm": 0.6619619727134705,
+      "learning_rate": 6.558398220244717e-05,
+      "loss": 2.6194,
+      "step": 3027
+    },
+    {
+      "epoch": 40.38795986622073,
+      "grad_norm": 0.6329185962677002,
+      "learning_rate": 6.553948832035596e-05,
+      "loss": 2.463,
+      "step": 3028
+    },
+    {
+      "epoch": 40.40133779264214,
+      "grad_norm": 0.5924078822135925,
+      "learning_rate": 6.549499443826475e-05,
+      "loss": 2.5448,
+      "step": 3029
+    },
+    {
+      "epoch": 40.414715719063544,
+      "grad_norm": 0.6422038078308105,
+      "learning_rate": 6.545050055617353e-05,
+      "loss": 2.5527,
+      "step": 3030
+    },
+    {
+      "epoch": 40.42809364548495,
+      "grad_norm": 0.8861187696456909,
+      "learning_rate": 6.540600667408232e-05,
+      "loss": 2.5809,
+      "step": 3031
+    },
+    {
+      "epoch": 40.441471571906355,
+      "grad_norm": 0.6287595629692078,
+      "learning_rate": 6.536151279199111e-05,
+      "loss": 2.3957,
+      "step": 3032
+    },
+    {
+      "epoch": 40.45484949832776,
+      "grad_norm": 0.6341630816459656,
+      "learning_rate": 6.53170189098999e-05,
+      "loss": 2.7261,
+      "step": 3033
+    },
+    {
+      "epoch": 40.468227424749166,
+      "grad_norm": 0.6193274855613708,
+      "learning_rate": 6.527252502780869e-05,
+      "loss": 2.6581,
+      "step": 3034
+    },
+    {
+      "epoch": 40.48160535117057,
+      "grad_norm": 0.6899332404136658,
+      "learning_rate": 6.522803114571747e-05,
+      "loss": 2.5425,
+      "step": 3035
+    },
+    {
+      "epoch": 40.49498327759197,
+      "grad_norm": 0.6204500794410706,
+      "learning_rate": 6.518353726362626e-05,
+      "loss": 2.5298,
+      "step": 3036
+    },
+    {
+      "epoch": 40.508361204013376,
+      "grad_norm": 0.642670750617981,
+      "learning_rate": 6.513904338153505e-05,
+      "loss": 2.5961,
+      "step": 3037
+    },
+    {
+      "epoch": 40.52173913043478,
+      "grad_norm": 0.6434988975524902,
+      "learning_rate": 6.509454949944383e-05,
+      "loss": 2.5516,
+      "step": 3038
+    },
+    {
+      "epoch": 40.53511705685619,
+      "grad_norm": 0.62471604347229,
+      "learning_rate": 6.505005561735261e-05,
+      "loss": 2.4798,
+      "step": 3039
+    },
+    {
+      "epoch": 40.54849498327759,
+      "grad_norm": 0.6490535140037537,
+      "learning_rate": 6.50055617352614e-05,
+      "loss": 2.6373,
+      "step": 3040
+    },
+    {
+      "epoch": 40.561872909699,
+      "grad_norm": 0.6673942804336548,
+      "learning_rate": 6.496106785317019e-05,
+      "loss": 2.3785,
+      "step": 3041
+    },
+    {
+      "epoch": 40.575250836120404,
+      "grad_norm": 0.6491774320602417,
+      "learning_rate": 6.491657397107897e-05,
+      "loss": 2.7776,
+      "step": 3042
+    },
+    {
+      "epoch": 40.58862876254181,
+      "grad_norm": 0.6260173916816711,
+      "learning_rate": 6.487208008898777e-05,
+      "loss": 2.33,
+      "step": 3043
+    },
+    {
+      "epoch": 40.60200668896321,
+      "grad_norm": 0.6691784262657166,
+      "learning_rate": 6.482758620689655e-05,
+      "loss": 2.5681,
+      "step": 3044
+    },
+    {
+      "epoch": 40.61538461538461,
+      "grad_norm": 0.6271287798881531,
+      "learning_rate": 6.478309232480535e-05,
+      "loss": 2.5003,
+      "step": 3045
+    },
+    {
+      "epoch": 40.62876254180602,
+      "grad_norm": 0.5982306599617004,
+      "learning_rate": 6.473859844271413e-05,
+      "loss": 2.4117,
+      "step": 3046
+    },
+    {
+      "epoch": 40.642140468227424,
+      "grad_norm": 0.6666748523712158,
+      "learning_rate": 6.469410456062291e-05,
+      "loss": 2.9225,
+      "step": 3047
+    },
+    {
+      "epoch": 40.65551839464883,
+      "grad_norm": 0.7303210496902466,
+      "learning_rate": 6.46496106785317e-05,
+      "loss": 2.4966,
+      "step": 3048
+    },
+    {
+      "epoch": 40.668896321070235,
+      "grad_norm": 0.6557690501213074,
+      "learning_rate": 6.460511679644049e-05,
+      "loss": 2.7931,
+      "step": 3049
+    },
+    {
+      "epoch": 40.68227424749164,
+      "grad_norm": 0.6602864265441895,
+      "learning_rate": 6.456062291434928e-05,
+      "loss": 2.6292,
+      "step": 3050
+    },
+    {
+      "epoch": 40.69565217391305,
+      "grad_norm": 0.6641650199890137,
+      "learning_rate": 6.451612903225807e-05,
+      "loss": 2.4421,
+      "step": 3051
+    },
+    {
+      "epoch": 40.709030100334445,
+      "grad_norm": 0.6553093791007996,
+      "learning_rate": 6.447163515016685e-05,
+      "loss": 2.7653,
+      "step": 3052
+    },
+    {
+      "epoch": 40.72240802675585,
+      "grad_norm": 0.6673446893692017,
+      "learning_rate": 6.442714126807565e-05,
+      "loss": 2.7016,
+      "step": 3053
+    },
+    {
+      "epoch": 40.735785953177256,
+      "grad_norm": 0.6132397055625916,
+      "learning_rate": 6.438264738598443e-05,
+      "loss": 2.6634,
+      "step": 3054
+    },
+    {
+      "epoch": 40.74916387959866,
+      "grad_norm": 0.6243513822555542,
+      "learning_rate": 6.433815350389321e-05,
+      "loss": 2.3653,
+      "step": 3055
+    },
+    {
+      "epoch": 40.76254180602007,
+      "grad_norm": 0.6419558525085449,
+      "learning_rate": 6.4293659621802e-05,
+      "loss": 2.5712,
+      "step": 3056
+    },
+    {
+      "epoch": 40.77591973244147,
+      "grad_norm": 0.6716275215148926,
+      "learning_rate": 6.424916573971079e-05,
+      "loss": 2.645,
+      "step": 3057
+    },
+    {
+      "epoch": 40.78929765886288,
+      "grad_norm": 0.6217707991600037,
+      "learning_rate": 6.420467185761958e-05,
+      "loss": 2.427,
+      "step": 3058
+    },
+    {
+      "epoch": 40.802675585284284,
+      "grad_norm": 0.6085132956504822,
+      "learning_rate": 6.416017797552837e-05,
+      "loss": 2.6017,
+      "step": 3059
+    },
+    {
+      "epoch": 40.81605351170568,
+      "grad_norm": 0.6166917681694031,
+      "learning_rate": 6.411568409343715e-05,
+      "loss": 2.383,
+      "step": 3060
+    },
+    {
+      "epoch": 40.82943143812709,
+      "grad_norm": 0.607100248336792,
+      "learning_rate": 6.407119021134595e-05,
+      "loss": 2.5013,
+      "step": 3061
+    },
+    {
+      "epoch": 40.84280936454849,
+      "grad_norm": 0.641970694065094,
+      "learning_rate": 6.402669632925473e-05,
+      "loss": 2.4467,
+      "step": 3062
+    },
+    {
+      "epoch": 40.8561872909699,
+      "grad_norm": 0.598092257976532,
+      "learning_rate": 6.398220244716352e-05,
+      "loss": 2.3341,
+      "step": 3063
+    },
+    {
+      "epoch": 40.869565217391305,
+      "grad_norm": 0.6640897393226624,
+      "learning_rate": 6.39377085650723e-05,
+      "loss": 2.533,
+      "step": 3064
+    },
+    {
+      "epoch": 40.88294314381271,
+      "grad_norm": 0.6127122044563293,
+      "learning_rate": 6.389321468298109e-05,
+      "loss": 2.3842,
+      "step": 3065
+    },
+    {
+      "epoch": 40.896321070234116,
+      "grad_norm": 0.6670165657997131,
+      "learning_rate": 6.384872080088988e-05,
+      "loss": 2.5481,
+      "step": 3066
+    },
+    {
+      "epoch": 40.90969899665552,
+      "grad_norm": 0.6466744542121887,
+      "learning_rate": 6.380422691879867e-05,
+      "loss": 2.7235,
+      "step": 3067
+    },
+    {
+      "epoch": 40.92307692307692,
+      "grad_norm": 0.6578346490859985,
+      "learning_rate": 6.375973303670746e-05,
+      "loss": 2.779,
+      "step": 3068
+    },
+    {
+      "epoch": 40.936454849498325,
+      "grad_norm": 0.6433802843093872,
+      "learning_rate": 6.371523915461625e-05,
+      "loss": 2.3942,
+      "step": 3069
+    },
+    {
+      "epoch": 40.94983277591973,
+      "grad_norm": 0.6575675010681152,
+      "learning_rate": 6.367074527252503e-05,
+      "loss": 2.705,
+      "step": 3070
+    },
+    {
+      "epoch": 40.96321070234114,
+      "grad_norm": 0.6280662417411804,
+      "learning_rate": 6.362625139043382e-05,
+      "loss": 2.4813,
+      "step": 3071
+    },
+    {
+      "epoch": 40.97658862876254,
+      "grad_norm": 0.6336925625801086,
+      "learning_rate": 6.35817575083426e-05,
+      "loss": 2.6624,
+      "step": 3072
+    },
+    {
+      "epoch": 40.98996655518395,
+      "grad_norm": 0.6524997353553772,
+      "learning_rate": 6.35372636262514e-05,
+      "loss": 2.7407,
+      "step": 3073
+    },
+    {
+      "epoch": 41.0,
+      "grad_norm": 0.7705938816070557,
+      "learning_rate": 6.349276974416018e-05,
+      "loss": 2.4095,
+      "step": 3074
+    },
+    {
+      "epoch": 41.013377926421406,
+      "grad_norm": 0.6220411658287048,
+      "learning_rate": 6.344827586206897e-05,
+      "loss": 2.6689,
+      "step": 3075
+    },
+    {
+      "epoch": 41.02675585284281,
+      "grad_norm": 0.7290953397750854,
+      "learning_rate": 6.340378197997776e-05,
+      "loss": 2.5802,
+      "step": 3076
+    },
+    {
+      "epoch": 41.04013377926422,
+      "grad_norm": 0.8503483533859253,
+      "learning_rate": 6.335928809788654e-05,
+      "loss": 2.4214,
+      "step": 3077
+    },
+    {
+      "epoch": 41.05351170568562,
+      "grad_norm": 0.6377903819084167,
+      "learning_rate": 6.331479421579534e-05,
+      "loss": 2.4564,
+      "step": 3078
+    },
+    {
+      "epoch": 41.06688963210702,
+      "grad_norm": 0.6296485662460327,
+      "learning_rate": 6.327030033370411e-05,
+      "loss": 2.3787,
+      "step": 3079
+    },
+    {
+      "epoch": 41.080267558528426,
+      "grad_norm": 0.680673360824585,
+      "learning_rate": 6.32258064516129e-05,
+      "loss": 2.6136,
+      "step": 3080
+    },
+    {
+      "epoch": 41.09364548494983,
+      "grad_norm": 0.67911297082901,
+      "learning_rate": 6.318131256952169e-05,
+      "loss": 2.6889,
+      "step": 3081
+    },
+    {
+      "epoch": 41.10702341137124,
+      "grad_norm": 0.7195813655853271,
+      "learning_rate": 6.313681868743047e-05,
+      "loss": 2.5189,
+      "step": 3082
+    },
+    {
+      "epoch": 41.12040133779264,
+      "grad_norm": 0.644444465637207,
+      "learning_rate": 6.309232480533927e-05,
+      "loss": 2.4799,
+      "step": 3083
+    },
+    {
+      "epoch": 41.13377926421405,
+      "grad_norm": 0.6452529430389404,
+      "learning_rate": 6.304783092324805e-05,
+      "loss": 2.452,
+      "step": 3084
+    },
+    {
+      "epoch": 41.147157190635454,
+      "grad_norm": 0.6581646800041199,
+      "learning_rate": 6.300333704115684e-05,
+      "loss": 2.5695,
+      "step": 3085
+    },
+    {
+      "epoch": 41.16053511705686,
+      "grad_norm": 0.6414064168930054,
+      "learning_rate": 6.295884315906563e-05,
+      "loss": 2.4686,
+      "step": 3086
+    },
+    {
+      "epoch": 41.17391304347826,
+      "grad_norm": 0.620027482509613,
+      "learning_rate": 6.291434927697441e-05,
+      "loss": 2.4335,
+      "step": 3087
+    },
+    {
+      "epoch": 41.187290969899664,
+      "grad_norm": 0.6725049018859863,
+      "learning_rate": 6.28698553948832e-05,
+      "loss": 2.4423,
+      "step": 3088
+    },
+    {
+      "epoch": 41.20066889632107,
+      "grad_norm": 0.6626586318016052,
+      "learning_rate": 6.282536151279199e-05,
+      "loss": 2.3875,
+      "step": 3089
+    },
+    {
+      "epoch": 41.214046822742475,
+      "grad_norm": 0.6290739178657532,
+      "learning_rate": 6.278086763070078e-05,
+      "loss": 2.6957,
+      "step": 3090
+    },
+    {
+      "epoch": 41.22742474916388,
+      "grad_norm": 0.6320621967315674,
+      "learning_rate": 6.273637374860957e-05,
+      "loss": 2.6297,
+      "step": 3091
+    },
+    {
+      "epoch": 41.240802675585286,
+      "grad_norm": 0.6563973426818848,
+      "learning_rate": 6.269187986651835e-05,
+      "loss": 2.5361,
+      "step": 3092
+    },
+    {
+      "epoch": 41.25418060200669,
+      "grad_norm": 0.6116068363189697,
+      "learning_rate": 6.264738598442714e-05,
+      "loss": 2.644,
+      "step": 3093
+    },
+    {
+      "epoch": 41.2675585284281,
+      "grad_norm": 0.6175321340560913,
+      "learning_rate": 6.260289210233593e-05,
+      "loss": 2.2662,
+      "step": 3094
+    },
+    {
+      "epoch": 41.280936454849495,
+      "grad_norm": 0.6325392723083496,
+      "learning_rate": 6.255839822024472e-05,
+      "loss": 2.4981,
+      "step": 3095
+    },
+    {
+      "epoch": 41.2943143812709,
+      "grad_norm": 0.6011773943901062,
+      "learning_rate": 6.25139043381535e-05,
+      "loss": 2.4708,
+      "step": 3096
+    },
+    {
+      "epoch": 41.30769230769231,
+      "grad_norm": 0.6720942258834839,
+      "learning_rate": 6.246941045606229e-05,
+      "loss": 2.6639,
+      "step": 3097
+    },
+    {
+      "epoch": 41.32107023411371,
+      "grad_norm": 0.6427400708198547,
+      "learning_rate": 6.242491657397108e-05,
+      "loss": 2.3893,
+      "step": 3098
+    },
+    {
+      "epoch": 41.33444816053512,
+      "grad_norm": 0.6931129693984985,
+      "learning_rate": 6.238042269187987e-05,
+      "loss": 2.6919,
+      "step": 3099
+    },
+    {
+      "epoch": 41.34782608695652,
+      "grad_norm": 0.6366786956787109,
+      "learning_rate": 6.233592880978866e-05,
+      "loss": 2.3554,
+      "step": 3100
+    },
+    {
+      "epoch": 41.36120401337793,
+      "grad_norm": 0.5965732336044312,
+      "learning_rate": 6.229143492769744e-05,
+      "loss": 2.5368,
+      "step": 3101
+    },
+    {
+      "epoch": 41.374581939799334,
+      "grad_norm": 0.6699475646018982,
+      "learning_rate": 6.224694104560623e-05,
+      "loss": 2.614,
+      "step": 3102
+    },
+    {
+      "epoch": 41.38795986622073,
+      "grad_norm": 0.6514851450920105,
+      "learning_rate": 6.220244716351502e-05,
+      "loss": 2.611,
+      "step": 3103
+    },
+    {
+      "epoch": 41.40133779264214,
+      "grad_norm": 0.6431285738945007,
+      "learning_rate": 6.21579532814238e-05,
+      "loss": 2.5508,
+      "step": 3104
+    },
+    {
+      "epoch": 41.414715719063544,
+      "grad_norm": 0.6158801913261414,
+      "learning_rate": 6.21134593993326e-05,
+      "loss": 2.5008,
+      "step": 3105
+    },
+    {
+      "epoch": 41.42809364548495,
+      "grad_norm": 0.6571468114852905,
+      "learning_rate": 6.206896551724138e-05,
+      "loss": 2.5121,
+      "step": 3106
+    },
+    {
+      "epoch": 41.441471571906355,
+      "grad_norm": 0.6691536903381348,
+      "learning_rate": 6.202447163515017e-05,
+      "loss": 2.7706,
+      "step": 3107
+    },
+    {
+      "epoch": 41.45484949832776,
+      "grad_norm": 0.6521206498146057,
+      "learning_rate": 6.197997775305896e-05,
+      "loss": 2.5349,
+      "step": 3108
+    },
+    {
+      "epoch": 41.468227424749166,
+      "grad_norm": 0.6497208476066589,
+      "learning_rate": 6.193548387096774e-05,
+      "loss": 2.5069,
+      "step": 3109
+    },
+    {
+      "epoch": 41.48160535117057,
+      "grad_norm": 0.6708588004112244,
+      "learning_rate": 6.189098998887654e-05,
+      "loss": 2.5195,
+      "step": 3110
+    },
+    {
+      "epoch": 41.49498327759197,
+      "grad_norm": 0.6343464851379395,
+      "learning_rate": 6.184649610678532e-05,
+      "loss": 2.5025,
+      "step": 3111
+    },
+    {
+      "epoch": 41.508361204013376,
+      "grad_norm": 0.6211134195327759,
+      "learning_rate": 6.18020022246941e-05,
+      "loss": 2.3835,
+      "step": 3112
+    },
+    {
+      "epoch": 41.52173913043478,
+      "grad_norm": 0.7531595230102539,
+      "learning_rate": 6.17575083426029e-05,
+      "loss": 2.6583,
+      "step": 3113
+    },
+    {
+      "epoch": 41.53511705685619,
+      "grad_norm": 0.6848407983779907,
+      "learning_rate": 6.171301446051168e-05,
+      "loss": 2.5161,
+      "step": 3114
+    },
+    {
+      "epoch": 41.54849498327759,
+      "grad_norm": 0.6121324300765991,
+      "learning_rate": 6.166852057842048e-05,
+      "loss": 2.6031,
+      "step": 3115
+    },
+    {
+      "epoch": 41.561872909699,
+      "grad_norm": 0.6391879320144653,
+      "learning_rate": 6.162402669632926e-05,
+      "loss": 2.4762,
+      "step": 3116
+    },
+    {
+      "epoch": 41.575250836120404,
+      "grad_norm": 0.6806944012641907,
+      "learning_rate": 6.157953281423804e-05,
+      "loss": 2.4322,
+      "step": 3117
+    },
+    {
+      "epoch": 41.58862876254181,
+      "grad_norm": 0.5879799127578735,
+      "learning_rate": 6.153503893214684e-05,
+      "loss": 2.4121,
+      "step": 3118
+    },
+    {
+      "epoch": 41.60200668896321,
+      "grad_norm": 0.6548652052879333,
+      "learning_rate": 6.149054505005562e-05,
+      "loss": 2.516,
+      "step": 3119
+    },
+    {
+      "epoch": 41.61538461538461,
+      "grad_norm": 0.6357155442237854,
+      "learning_rate": 6.14460511679644e-05,
+      "loss": 2.6071,
+      "step": 3120
+    },
+    {
+      "epoch": 41.62876254180602,
+      "grad_norm": 0.6546043753623962,
+      "learning_rate": 6.140155728587319e-05,
+      "loss": 2.5423,
+      "step": 3121
+    },
+    {
+      "epoch": 41.642140468227424,
+      "grad_norm": 0.6443083882331848,
+      "learning_rate": 6.135706340378198e-05,
+      "loss": 2.4373,
+      "step": 3122
+    },
+    {
+      "epoch": 41.65551839464883,
+      "grad_norm": 0.599860429763794,
+      "learning_rate": 6.131256952169077e-05,
+      "loss": 2.3978,
+      "step": 3123
+    },
+    {
+      "epoch": 41.668896321070235,
+      "grad_norm": 0.6442713737487793,
+      "learning_rate": 6.126807563959955e-05,
+      "loss": 2.437,
+      "step": 3124
+    },
+    {
+      "epoch": 41.68227424749164,
+      "grad_norm": 0.704804003238678,
+      "learning_rate": 6.122358175750834e-05,
+      "loss": 2.5634,
+      "step": 3125
+    },
+    {
+      "epoch": 41.69565217391305,
+      "grad_norm": 0.6029696464538574,
+      "learning_rate": 6.117908787541713e-05,
+      "loss": 2.5276,
+      "step": 3126
+    },
+    {
+      "epoch": 41.709030100334445,
+      "grad_norm": 0.6845595240592957,
+      "learning_rate": 6.113459399332592e-05,
+      "loss": 2.6695,
+      "step": 3127
+    },
+    {
+      "epoch": 41.72240802675585,
+      "grad_norm": 0.6166744232177734,
+      "learning_rate": 6.10901001112347e-05,
+      "loss": 2.5261,
+      "step": 3128
+    },
+    {
+      "epoch": 41.735785953177256,
+      "grad_norm": 0.6824954152107239,
+      "learning_rate": 6.104560622914349e-05,
+      "loss": 2.6611,
+      "step": 3129
+    },
+    {
+      "epoch": 41.74916387959866,
+      "grad_norm": 0.6522831320762634,
+      "learning_rate": 6.100111234705228e-05,
+      "loss": 2.5558,
+      "step": 3130
+    },
+    {
+      "epoch": 41.76254180602007,
+      "grad_norm": 0.6635300517082214,
+      "learning_rate": 6.0956618464961065e-05,
+      "loss": 2.6164,
+      "step": 3131
+    },
+    {
+      "epoch": 41.77591973244147,
+      "grad_norm": 0.6519820094108582,
+      "learning_rate": 6.0912124582869854e-05,
+      "loss": 2.5598,
+      "step": 3132
+    },
+    {
+      "epoch": 41.78929765886288,
+      "grad_norm": 0.6299057006835938,
+      "learning_rate": 6.086763070077864e-05,
+      "loss": 2.3672,
+      "step": 3133
+    },
+    {
+      "epoch": 41.802675585284284,
+      "grad_norm": 0.6761189103126526,
+      "learning_rate": 6.082313681868743e-05,
+      "loss": 2.4973,
+      "step": 3134
+    },
+    {
+      "epoch": 41.81605351170568,
+      "grad_norm": 0.6459671258926392,
+      "learning_rate": 6.077864293659622e-05,
+      "loss": 2.5184,
+      "step": 3135
+    },
+    {
+      "epoch": 41.82943143812709,
+      "grad_norm": 0.6295469999313354,
+      "learning_rate": 6.0734149054505004e-05,
+      "loss": 2.5129,
+      "step": 3136
+    },
+    {
+      "epoch": 41.84280936454849,
+      "grad_norm": 0.6619572639465332,
+      "learning_rate": 6.068965517241379e-05,
+      "loss": 2.602,
+      "step": 3137
+    },
+    {
+      "epoch": 41.8561872909699,
+      "grad_norm": 0.619455873966217,
+      "learning_rate": 6.064516129032258e-05,
+      "loss": 2.3657,
+      "step": 3138
+    },
+    {
+      "epoch": 41.869565217391305,
+      "grad_norm": 0.6936202049255371,
+      "learning_rate": 6.060066740823137e-05,
+      "loss": 2.6134,
+      "step": 3139
+    },
+    {
+      "epoch": 41.88294314381271,
+      "grad_norm": 0.7031989097595215,
+      "learning_rate": 6.055617352614016e-05,
+      "loss": 2.7589,
+      "step": 3140
+    },
+    {
+      "epoch": 41.896321070234116,
+      "grad_norm": 0.6757580041885376,
+      "learning_rate": 6.051167964404894e-05,
+      "loss": 2.5874,
+      "step": 3141
+    },
+    {
+      "epoch": 41.90969899665552,
+      "grad_norm": 0.6400772333145142,
+      "learning_rate": 6.046718576195773e-05,
+      "loss": 2.5219,
+      "step": 3142
+    },
+    {
+      "epoch": 41.92307692307692,
+      "grad_norm": 0.652505099773407,
+      "learning_rate": 6.042269187986652e-05,
+      "loss": 2.8772,
+      "step": 3143
+    },
+    {
+      "epoch": 41.936454849498325,
+      "grad_norm": 0.6441666483879089,
+      "learning_rate": 6.037819799777531e-05,
+      "loss": 2.5647,
+      "step": 3144
+    },
+    {
+      "epoch": 41.94983277591973,
+      "grad_norm": 0.6462740898132324,
+      "learning_rate": 6.03337041156841e-05,
+      "loss": 2.755,
+      "step": 3145
+    },
+    {
+      "epoch": 41.96321070234114,
+      "grad_norm": 0.6710271835327148,
+      "learning_rate": 6.028921023359288e-05,
+      "loss": 2.5574,
+      "step": 3146
+    },
+    {
+      "epoch": 41.97658862876254,
+      "grad_norm": 0.6961186528205872,
+      "learning_rate": 6.024471635150167e-05,
+      "loss": 2.5408,
+      "step": 3147
+    },
+    {
+      "epoch": 41.98996655518395,
+      "grad_norm": 0.6388704180717468,
+      "learning_rate": 6.020022246941046e-05,
+      "loss": 2.4168,
+      "step": 3148
+    },
+    {
+      "epoch": 42.0,
+      "grad_norm": 0.8058933615684509,
+      "learning_rate": 6.015572858731925e-05,
+      "loss": 2.7006,
+      "step": 3149
+    },
+    {
+      "epoch": 42.013377926421406,
+      "grad_norm": 0.6950250864028931,
+      "learning_rate": 6.011123470522804e-05,
+      "loss": 2.5687,
+      "step": 3150
+    },
+    {
+      "epoch": 42.02675585284281,
+      "grad_norm": 0.6203860640525818,
+      "learning_rate": 6.006674082313682e-05,
+      "loss": 2.5466,
+      "step": 3151
+    },
+    {
+      "epoch": 42.04013377926422,
+      "grad_norm": 0.6443012952804565,
+      "learning_rate": 6.002224694104561e-05,
+      "loss": 2.6772,
+      "step": 3152
+    },
+    {
+      "epoch": 42.05351170568562,
+      "grad_norm": 0.6492132544517517,
+      "learning_rate": 5.99777530589544e-05,
+      "loss": 2.6421,
+      "step": 3153
+    },
+    {
+      "epoch": 42.06688963210702,
+      "grad_norm": 0.6684079766273499,
+      "learning_rate": 5.993325917686319e-05,
+      "loss": 2.4521,
+      "step": 3154
+    },
+    {
+      "epoch": 42.080267558528426,
+      "grad_norm": 0.6225764751434326,
+      "learning_rate": 5.988876529477198e-05,
+      "loss": 2.3612,
+      "step": 3155
+    },
+    {
+      "epoch": 42.09364548494983,
+      "grad_norm": 0.7330244183540344,
+      "learning_rate": 5.984427141268076e-05,
+      "loss": 2.6417,
+      "step": 3156
+    },
+    {
+      "epoch": 42.10702341137124,
+      "grad_norm": 0.671696662902832,
+      "learning_rate": 5.979977753058955e-05,
+      "loss": 2.4971,
+      "step": 3157
+    },
+    {
+      "epoch": 42.12040133779264,
+      "grad_norm": 0.6462723612785339,
+      "learning_rate": 5.975528364849834e-05,
+      "loss": 2.4472,
+      "step": 3158
+    },
+    {
+      "epoch": 42.13377926421405,
+      "grad_norm": 0.646503210067749,
+      "learning_rate": 5.971078976640713e-05,
+      "loss": 2.5354,
+      "step": 3159
+    },
+    {
+      "epoch": 42.147157190635454,
+      "grad_norm": 0.6394603848457336,
+      "learning_rate": 5.966629588431592e-05,
+      "loss": 2.4596,
+      "step": 3160
+    },
+    {
+      "epoch": 42.16053511705686,
+      "grad_norm": 0.6669955253601074,
+      "learning_rate": 5.962180200222469e-05,
+      "loss": 2.4145,
+      "step": 3161
+    },
+    {
+      "epoch": 42.17391304347826,
+      "grad_norm": 0.756793737411499,
+      "learning_rate": 5.957730812013348e-05,
+      "loss": 2.5212,
+      "step": 3162
+    },
+    {
+      "epoch": 42.187290969899664,
+      "grad_norm": 0.6672840714454651,
+      "learning_rate": 5.9532814238042264e-05,
+      "loss": 2.4857,
+      "step": 3163
+    },
+    {
+      "epoch": 42.20066889632107,
+      "grad_norm": 0.6417210698127747,
+      "learning_rate": 5.948832035595105e-05,
+      "loss": 2.2809,
+      "step": 3164
+    },
+    {
+      "epoch": 42.214046822742475,
+      "grad_norm": 0.8348161578178406,
+      "learning_rate": 5.944382647385984e-05,
+      "loss": 2.4916,
+      "step": 3165
+    },
+    {
+      "epoch": 42.22742474916388,
+      "grad_norm": 0.6534797549247742,
+      "learning_rate": 5.939933259176863e-05,
+      "loss": 2.5452,
+      "step": 3166
+    },
+    {
+      "epoch": 42.240802675585286,
+      "grad_norm": 0.6620765924453735,
+      "learning_rate": 5.935483870967742e-05,
+      "loss": 2.4918,
+      "step": 3167
+    },
+    {
+      "epoch": 42.25418060200669,
+      "grad_norm": 0.6680149435997009,
+      "learning_rate": 5.93103448275862e-05,
+      "loss": 2.5057,
+      "step": 3168
+    },
+    {
+      "epoch": 42.2675585284281,
+      "grad_norm": 0.6373328566551208,
+      "learning_rate": 5.926585094549499e-05,
+      "loss": 2.4261,
+      "step": 3169
+    },
+    {
+      "epoch": 42.280936454849495,
+      "grad_norm": 0.647787868976593,
+      "learning_rate": 5.922135706340378e-05,
+      "loss": 2.5882,
+      "step": 3170
+    },
+    {
+      "epoch": 42.2943143812709,
+      "grad_norm": 0.7209821343421936,
+      "learning_rate": 5.917686318131257e-05,
+      "loss": 2.5832,
+      "step": 3171
+    },
+    {
+      "epoch": 42.30769230769231,
+      "grad_norm": 0.650434672832489,
+      "learning_rate": 5.913236929922136e-05,
+      "loss": 2.6609,
+      "step": 3172
+    },
+    {
+      "epoch": 42.32107023411371,
+      "grad_norm": 0.6905040740966797,
+      "learning_rate": 5.908787541713014e-05,
+      "loss": 2.4769,
+      "step": 3173
+    },
+    {
+      "epoch": 42.33444816053512,
+      "grad_norm": 0.6566370725631714,
+      "learning_rate": 5.904338153503893e-05,
+      "loss": 2.3585,
+      "step": 3174
+    },
+    {
+      "epoch": 42.34782608695652,
+      "grad_norm": 0.6680585741996765,
+      "learning_rate": 5.899888765294772e-05,
+      "loss": 2.4853,
+      "step": 3175
+    },
+    {
+      "epoch": 42.36120401337793,
+      "grad_norm": 0.6622024774551392,
+      "learning_rate": 5.895439377085651e-05,
+      "loss": 2.525,
+      "step": 3176
+    },
+    {
+      "epoch": 42.374581939799334,
+      "grad_norm": 0.6646168828010559,
+      "learning_rate": 5.89098998887653e-05,
+      "loss": 2.7159,
+      "step": 3177
+    },
+    {
+      "epoch": 42.38795986622073,
+      "grad_norm": 0.6752247214317322,
+      "learning_rate": 5.886540600667408e-05,
+      "loss": 2.6018,
+      "step": 3178
+    },
+    {
+      "epoch": 42.40133779264214,
+      "grad_norm": 0.7147390842437744,
+      "learning_rate": 5.882091212458287e-05,
+      "loss": 2.52,
+      "step": 3179
+    },
+    {
+      "epoch": 42.414715719063544,
+      "grad_norm": 0.7092293500900269,
+      "learning_rate": 5.877641824249166e-05,
+      "loss": 2.3825,
+      "step": 3180
+    },
+    {
+      "epoch": 42.42809364548495,
+      "grad_norm": 0.6321873664855957,
+      "learning_rate": 5.873192436040045e-05,
+      "loss": 2.5694,
+      "step": 3181
+    },
+    {
+      "epoch": 42.441471571906355,
+      "grad_norm": 0.7254124879837036,
+      "learning_rate": 5.868743047830924e-05,
+      "loss": 2.5501,
+      "step": 3182
+    },
+    {
+      "epoch": 42.45484949832776,
+      "grad_norm": 0.6851321458816528,
+      "learning_rate": 5.864293659621802e-05,
+      "loss": 2.3981,
+      "step": 3183
+    },
+    {
+      "epoch": 42.468227424749166,
+      "grad_norm": 0.7157125473022461,
+      "learning_rate": 5.859844271412681e-05,
+      "loss": 2.4062,
+      "step": 3184
+    },
+    {
+      "epoch": 42.48160535117057,
+      "grad_norm": 0.6364109516143799,
+      "learning_rate": 5.85539488320356e-05,
+      "loss": 2.4445,
+      "step": 3185
+    },
+    {
+      "epoch": 42.49498327759197,
+      "grad_norm": 0.7198134064674377,
+      "learning_rate": 5.850945494994439e-05,
+      "loss": 2.4817,
+      "step": 3186
+    },
+    {
+      "epoch": 42.508361204013376,
+      "grad_norm": 0.6572685837745667,
+      "learning_rate": 5.846496106785318e-05,
+      "loss": 2.7384,
+      "step": 3187
+    },
+    {
+      "epoch": 42.52173913043478,
+      "grad_norm": 0.6623586416244507,
+      "learning_rate": 5.842046718576196e-05,
+      "loss": 2.4099,
+      "step": 3188
+    },
+    {
+      "epoch": 42.53511705685619,
+      "grad_norm": 0.6533327102661133,
+      "learning_rate": 5.837597330367075e-05,
+      "loss": 2.4957,
+      "step": 3189
+    },
+    {
+      "epoch": 42.54849498327759,
+      "grad_norm": 0.706375241279602,
+      "learning_rate": 5.833147942157954e-05,
+      "loss": 2.6125,
+      "step": 3190
+    },
+    {
+      "epoch": 42.561872909699,
+      "grad_norm": 0.6902095675468445,
+      "learning_rate": 5.828698553948833e-05,
+      "loss": 2.616,
+      "step": 3191
+    },
+    {
+      "epoch": 42.575250836120404,
+      "grad_norm": 0.6547108888626099,
+      "learning_rate": 5.8242491657397116e-05,
+      "loss": 2.3856,
+      "step": 3192
+    },
+    {
+      "epoch": 42.58862876254181,
+      "grad_norm": 0.6478321552276611,
+      "learning_rate": 5.81979977753059e-05,
+      "loss": 2.5016,
+      "step": 3193
+    },
+    {
+      "epoch": 42.60200668896321,
+      "grad_norm": 0.6727077960968018,
+      "learning_rate": 5.815350389321469e-05,
+      "loss": 2.4331,
+      "step": 3194
+    },
+    {
+      "epoch": 42.61538461538461,
+      "grad_norm": 0.6678333878517151,
+      "learning_rate": 5.810901001112348e-05,
+      "loss": 2.5317,
+      "step": 3195
+    },
+    {
+      "epoch": 42.62876254180602,
+      "grad_norm": 0.7883992791175842,
+      "learning_rate": 5.8064516129032266e-05,
+      "loss": 2.5823,
+      "step": 3196
+    },
+    {
+      "epoch": 42.642140468227424,
+      "grad_norm": 0.6439074873924255,
+      "learning_rate": 5.8020022246941055e-05,
+      "loss": 2.4141,
+      "step": 3197
+    },
+    {
+      "epoch": 42.65551839464883,
+      "grad_norm": 0.6541165113449097,
+      "learning_rate": 5.797552836484984e-05,
+      "loss": 2.5384,
+      "step": 3198
+    },
+    {
+      "epoch": 42.668896321070235,
+      "grad_norm": 0.6592057943344116,
+      "learning_rate": 5.7931034482758627e-05,
+      "loss": 2.4035,
+      "step": 3199
+    },
+    {
+      "epoch": 42.68227424749164,
+      "grad_norm": 0.6413922905921936,
+      "learning_rate": 5.7886540600667416e-05,
+      "loss": 2.4653,
+      "step": 3200
+    },
+    {
+      "epoch": 42.69565217391305,
+      "grad_norm": 0.6790466904640198,
+      "learning_rate": 5.7842046718576205e-05,
+      "loss": 2.5069,
+      "step": 3201
+    },
+    {
+      "epoch": 42.709030100334445,
+      "grad_norm": 0.6454100608825684,
+      "learning_rate": 5.779755283648498e-05,
+      "loss": 2.4377,
+      "step": 3202
+    },
+    {
+      "epoch": 42.72240802675585,
+      "grad_norm": 0.6789502501487732,
+      "learning_rate": 5.775305895439377e-05,
+      "loss": 2.6028,
+      "step": 3203
+    },
+    {
+      "epoch": 42.735785953177256,
+      "grad_norm": 0.6526380777359009,
+      "learning_rate": 5.770856507230256e-05,
+      "loss": 2.544,
+      "step": 3204
+    },
+    {
+      "epoch": 42.74916387959866,
+      "grad_norm": 0.6715470552444458,
+      "learning_rate": 5.766407119021134e-05,
+      "loss": 2.5226,
+      "step": 3205
+    },
+    {
+      "epoch": 42.76254180602007,
+      "grad_norm": 0.6436430811882019,
+      "learning_rate": 5.761957730812013e-05,
+      "loss": 2.4716,
+      "step": 3206
+    },
+    {
+      "epoch": 42.77591973244147,
+      "grad_norm": 0.6560038328170776,
+      "learning_rate": 5.757508342602892e-05,
+      "loss": 2.575,
+      "step": 3207
+    },
+    {
+      "epoch": 42.78929765886288,
+      "grad_norm": 0.6760275363922119,
+      "learning_rate": 5.753058954393771e-05,
+      "loss": 2.6278,
+      "step": 3208
+    },
+    {
+      "epoch": 42.802675585284284,
+      "grad_norm": 0.6424068212509155,
+      "learning_rate": 5.74860956618465e-05,
+      "loss": 2.5249,
+      "step": 3209
+    },
+    {
+      "epoch": 42.81605351170568,
+      "grad_norm": 0.6159402132034302,
+      "learning_rate": 5.744160177975528e-05,
+      "loss": 2.4096,
+      "step": 3210
+    },
+    {
+      "epoch": 42.82943143812709,
+      "grad_norm": 0.6637108325958252,
+      "learning_rate": 5.739710789766407e-05,
+      "loss": 2.5771,
+      "step": 3211
+    },
+    {
+      "epoch": 42.84280936454849,
+      "grad_norm": 0.6485425233840942,
+      "learning_rate": 5.735261401557286e-05,
+      "loss": 2.4803,
+      "step": 3212
+    },
+    {
+      "epoch": 42.8561872909699,
+      "grad_norm": 0.6670514941215515,
+      "learning_rate": 5.730812013348165e-05,
+      "loss": 2.7293,
+      "step": 3213
+    },
+    {
+      "epoch": 42.869565217391305,
+      "grad_norm": 0.6216664910316467,
+      "learning_rate": 5.726362625139044e-05,
+      "loss": 2.488,
+      "step": 3214
+    },
+    {
+      "epoch": 42.88294314381271,
+      "grad_norm": 0.6202486157417297,
+      "learning_rate": 5.721913236929922e-05,
+      "loss": 2.4099,
+      "step": 3215
+    },
+    {
+      "epoch": 42.896321070234116,
+      "grad_norm": 0.6572627425193787,
+      "learning_rate": 5.717463848720801e-05,
+      "loss": 2.6004,
+      "step": 3216
+    },
+    {
+      "epoch": 42.90969899665552,
+      "grad_norm": 0.6133331060409546,
+      "learning_rate": 5.71301446051168e-05,
+      "loss": 2.4422,
+      "step": 3217
+    },
+    {
+      "epoch": 42.92307692307692,
+      "grad_norm": 0.6674253940582275,
+      "learning_rate": 5.708565072302559e-05,
+      "loss": 2.5856,
+      "step": 3218
+    },
+    {
+      "epoch": 42.936454849498325,
+      "grad_norm": 0.6581224799156189,
+      "learning_rate": 5.7041156840934376e-05,
+      "loss": 2.4517,
+      "step": 3219
+    },
+    {
+      "epoch": 42.94983277591973,
+      "grad_norm": 0.6469165682792664,
+      "learning_rate": 5.699666295884316e-05,
+      "loss": 2.6374,
+      "step": 3220
+    },
+    {
+      "epoch": 42.96321070234114,
+      "grad_norm": 0.6413242816925049,
+      "learning_rate": 5.695216907675195e-05,
+      "loss": 2.5076,
+      "step": 3221
+    },
+    {
+      "epoch": 42.97658862876254,
+      "grad_norm": 0.6252263188362122,
+      "learning_rate": 5.690767519466074e-05,
+      "loss": 2.4679,
+      "step": 3222
+    },
+    {
+      "epoch": 42.98996655518395,
+      "grad_norm": 0.7174087166786194,
+      "learning_rate": 5.6863181312569526e-05,
+      "loss": 2.6383,
+      "step": 3223
+    },
+    {
+      "epoch": 43.0,
+      "grad_norm": 0.7846776247024536,
+      "learning_rate": 5.6818687430478315e-05,
+      "loss": 2.7,
+      "step": 3224
+    },
+    {
+      "epoch": 43.013377926421406,
+      "grad_norm": 0.6958154439926147,
+      "learning_rate": 5.67741935483871e-05,
+      "loss": 2.5076,
+      "step": 3225
+    },
+    {
+      "epoch": 43.02675585284281,
+      "grad_norm": 0.6751554608345032,
+      "learning_rate": 5.672969966629589e-05,
+      "loss": 2.6423,
+      "step": 3226
+    },
+    {
+      "epoch": 43.04013377926422,
+      "grad_norm": 0.6690580248832703,
+      "learning_rate": 5.6685205784204676e-05,
+      "loss": 2.5008,
+      "step": 3227
+    },
+    {
+      "epoch": 43.05351170568562,
+      "grad_norm": 0.7116791605949402,
+      "learning_rate": 5.6640711902113465e-05,
+      "loss": 2.3609,
+      "step": 3228
+    },
+    {
+      "epoch": 43.06688963210702,
+      "grad_norm": 0.626956582069397,
+      "learning_rate": 5.6596218020022254e-05,
+      "loss": 2.471,
+      "step": 3229
+    },
+    {
+      "epoch": 43.080267558528426,
+      "grad_norm": 0.6250149607658386,
+      "learning_rate": 5.6551724137931037e-05,
+      "loss": 2.4207,
+      "step": 3230
+    },
+    {
+      "epoch": 43.09364548494983,
+      "grad_norm": 0.6632840037345886,
+      "learning_rate": 5.6507230255839826e-05,
+      "loss": 2.4791,
+      "step": 3231
+    },
+    {
+      "epoch": 43.10702341137124,
+      "grad_norm": 1.4978615045547485,
+      "learning_rate": 5.6462736373748615e-05,
+      "loss": 2.3859,
+      "step": 3232
+    },
+    {
+      "epoch": 43.12040133779264,
+      "grad_norm": 0.7302461862564087,
+      "learning_rate": 5.6418242491657404e-05,
+      "loss": 2.4571,
+      "step": 3233
+    },
+    {
+      "epoch": 43.13377926421405,
+      "grad_norm": 0.6785010099411011,
+      "learning_rate": 5.637374860956619e-05,
+      "loss": 2.4689,
+      "step": 3234
+    },
+    {
+      "epoch": 43.147157190635454,
+      "grad_norm": 0.6378248929977417,
+      "learning_rate": 5.6329254727474976e-05,
+      "loss": 2.5509,
+      "step": 3235
+    },
+    {
+      "epoch": 43.16053511705686,
+      "grad_norm": 0.6675599217414856,
+      "learning_rate": 5.6284760845383765e-05,
+      "loss": 2.6542,
+      "step": 3236
+    },
+    {
+      "epoch": 43.17391304347826,
+      "grad_norm": 0.6947749853134155,
+      "learning_rate": 5.6240266963292554e-05,
+      "loss": 2.7001,
+      "step": 3237
+    },
+    {
+      "epoch": 43.187290969899664,
+      "grad_norm": 0.6531405448913574,
+      "learning_rate": 5.619577308120134e-05,
+      "loss": 2.4718,
+      "step": 3238
+    },
+    {
+      "epoch": 43.20066889632107,
+      "grad_norm": 0.6625925898551941,
+      "learning_rate": 5.615127919911013e-05,
+      "loss": 2.344,
+      "step": 3239
+    },
+    {
+      "epoch": 43.214046822742475,
+      "grad_norm": 1.235804796218872,
+      "learning_rate": 5.6106785317018915e-05,
+      "loss": 2.5819,
+      "step": 3240
+    },
+    {
+      "epoch": 43.22742474916388,
+      "grad_norm": 0.6504685878753662,
+      "learning_rate": 5.6062291434927704e-05,
+      "loss": 2.3638,
+      "step": 3241
+    },
+    {
+      "epoch": 43.240802675585286,
+      "grad_norm": 0.7184932231903076,
+      "learning_rate": 5.601779755283649e-05,
+      "loss": 2.4465,
+      "step": 3242
+    },
+    {
+      "epoch": 43.25418060200669,
+      "grad_norm": 0.6429609656333923,
+      "learning_rate": 5.597330367074528e-05,
+      "loss": 2.1915,
+      "step": 3243
+    },
+    {
+      "epoch": 43.2675585284281,
+      "grad_norm": 0.7342157959938049,
+      "learning_rate": 5.592880978865406e-05,
+      "loss": 2.3047,
+      "step": 3244
+    },
+    {
+      "epoch": 43.280936454849495,
+      "grad_norm": 0.7221858501434326,
+      "learning_rate": 5.588431590656285e-05,
+      "loss": 2.5297,
+      "step": 3245
+    },
+    {
+      "epoch": 43.2943143812709,
+      "grad_norm": 0.7088733315467834,
+      "learning_rate": 5.5839822024471636e-05,
+      "loss": 2.6321,
+      "step": 3246
+    },
+    {
+      "epoch": 43.30769230769231,
+      "grad_norm": 0.6901667714118958,
+      "learning_rate": 5.579532814238042e-05,
+      "loss": 2.4381,
+      "step": 3247
+    },
+    {
+      "epoch": 43.32107023411371,
+      "grad_norm": 0.6635287404060364,
+      "learning_rate": 5.575083426028921e-05,
+      "loss": 2.5138,
+      "step": 3248
+    },
+    {
+      "epoch": 43.33444816053512,
+      "grad_norm": 0.6960332989692688,
+      "learning_rate": 5.5706340378198e-05,
+      "loss": 2.6753,
+      "step": 3249
+    },
+    {
+      "epoch": 43.34782608695652,
+      "grad_norm": 0.7167268395423889,
+      "learning_rate": 5.5661846496106786e-05,
+      "loss": 2.7389,
+      "step": 3250
+    },
+    {
+      "epoch": 43.36120401337793,
+      "grad_norm": 0.6687808036804199,
+      "learning_rate": 5.5617352614015575e-05,
+      "loss": 2.4048,
+      "step": 3251
+    },
+    {
+      "epoch": 43.374581939799334,
+      "grad_norm": 0.6490589380264282,
+      "learning_rate": 5.557285873192436e-05,
+      "loss": 2.5123,
+      "step": 3252
+    },
+    {
+      "epoch": 43.38795986622073,
+      "grad_norm": 0.6539404392242432,
+      "learning_rate": 5.552836484983315e-05,
+      "loss": 2.676,
+      "step": 3253
+    },
+    {
+      "epoch": 43.40133779264214,
+      "grad_norm": 0.6659723520278931,
+      "learning_rate": 5.5483870967741936e-05,
+      "loss": 2.3908,
+      "step": 3254
+    },
+    {
+      "epoch": 43.414715719063544,
+      "grad_norm": 0.6532467603683472,
+      "learning_rate": 5.5439377085650725e-05,
+      "loss": 2.4925,
+      "step": 3255
+    },
+    {
+      "epoch": 43.42809364548495,
+      "grad_norm": 0.8012048006057739,
+      "learning_rate": 5.5394883203559514e-05,
+      "loss": 2.5035,
+      "step": 3256
+    },
+    {
+      "epoch": 43.441471571906355,
+      "grad_norm": 0.6359190344810486,
+      "learning_rate": 5.5350389321468297e-05,
+      "loss": 2.4463,
+      "step": 3257
+    },
+    {
+      "epoch": 43.45484949832776,
+      "grad_norm": 0.6980485916137695,
+      "learning_rate": 5.5305895439377086e-05,
+      "loss": 2.6038,
+      "step": 3258
+    },
+    {
+      "epoch": 43.468227424749166,
+      "grad_norm": 0.757401168346405,
+      "learning_rate": 5.5261401557285875e-05,
+      "loss": 2.5263,
+      "step": 3259
+    },
+    {
+      "epoch": 43.48160535117057,
+      "grad_norm": 0.7248904705047607,
+      "learning_rate": 5.5216907675194664e-05,
+      "loss": 2.4536,
+      "step": 3260
+    },
+    {
+      "epoch": 43.49498327759197,
+      "grad_norm": 1.246519923210144,
+      "learning_rate": 5.517241379310345e-05,
+      "loss": 2.7486,
+      "step": 3261
+    },
+    {
+      "epoch": 43.508361204013376,
+      "grad_norm": 0.6592410802841187,
+      "learning_rate": 5.5127919911012236e-05,
+      "loss": 2.4943,
+      "step": 3262
+    },
+    {
+      "epoch": 43.52173913043478,
+      "grad_norm": 0.8287456035614014,
+      "learning_rate": 5.5083426028921025e-05,
+      "loss": 2.3626,
+      "step": 3263
+    },
+    {
+      "epoch": 43.53511705685619,
+      "grad_norm": 0.6615451574325562,
+      "learning_rate": 5.5038932146829814e-05,
+      "loss": 2.4929,
+      "step": 3264
+    },
+    {
+      "epoch": 43.54849498327759,
+      "grad_norm": 0.6615239977836609,
+      "learning_rate": 5.49944382647386e-05,
+      "loss": 2.2937,
+      "step": 3265
+    },
+    {
+      "epoch": 43.561872909699,
+      "grad_norm": 0.6681683659553528,
+      "learning_rate": 5.494994438264739e-05,
+      "loss": 2.5765,
+      "step": 3266
+    },
+    {
+      "epoch": 43.575250836120404,
+      "grad_norm": 0.6718909740447998,
+      "learning_rate": 5.4905450500556175e-05,
+      "loss": 2.4609,
+      "step": 3267
+    },
+    {
+      "epoch": 43.58862876254181,
+      "grad_norm": 0.6635290384292603,
+      "learning_rate": 5.4860956618464964e-05,
+      "loss": 2.3023,
+      "step": 3268
+    },
+    {
+      "epoch": 43.60200668896321,
+      "grad_norm": 0.6895936727523804,
+      "learning_rate": 5.481646273637375e-05,
+      "loss": 2.4802,
+      "step": 3269
+    },
+    {
+      "epoch": 43.61538461538461,
+      "grad_norm": 0.6714615821838379,
+      "learning_rate": 5.477196885428254e-05,
+      "loss": 2.4943,
+      "step": 3270
+    },
+    {
+      "epoch": 43.62876254180602,
+      "grad_norm": 0.6521191596984863,
+      "learning_rate": 5.472747497219133e-05,
+      "loss": 2.5329,
+      "step": 3271
+    },
+    {
+      "epoch": 43.642140468227424,
+      "grad_norm": 0.6796808242797852,
+      "learning_rate": 5.4682981090100114e-05,
+      "loss": 2.7126,
+      "step": 3272
+    },
+    {
+      "epoch": 43.65551839464883,
+      "grad_norm": 0.6271299123764038,
+      "learning_rate": 5.46384872080089e-05,
+      "loss": 2.3406,
+      "step": 3273
+    },
+    {
+      "epoch": 43.668896321070235,
+      "grad_norm": 0.6406214237213135,
+      "learning_rate": 5.459399332591769e-05,
+      "loss": 2.5473,
+      "step": 3274
+    },
+    {
+      "epoch": 43.68227424749164,
+      "grad_norm": 0.688023030757904,
+      "learning_rate": 5.454949944382648e-05,
+      "loss": 2.5628,
+      "step": 3275
+    },
+    {
+      "epoch": 43.69565217391305,
+      "grad_norm": 0.6261418461799622,
+      "learning_rate": 5.450500556173527e-05,
+      "loss": 2.3595,
+      "step": 3276
+    },
+    {
+      "epoch": 43.709030100334445,
+      "grad_norm": 0.6406996846199036,
+      "learning_rate": 5.446051167964405e-05,
+      "loss": 2.4768,
+      "step": 3277
+    },
+    {
+      "epoch": 43.72240802675585,
+      "grad_norm": 0.6346117258071899,
+      "learning_rate": 5.441601779755284e-05,
+      "loss": 2.3375,
+      "step": 3278
+    },
+    {
+      "epoch": 43.735785953177256,
+      "grad_norm": 0.6415894627571106,
+      "learning_rate": 5.437152391546163e-05,
+      "loss": 2.2956,
+      "step": 3279
+    },
+    {
+      "epoch": 43.74916387959866,
+      "grad_norm": 0.681527316570282,
+      "learning_rate": 5.432703003337042e-05,
+      "loss": 2.4731,
+      "step": 3280
+    },
+    {
+      "epoch": 43.76254180602007,
+      "grad_norm": 0.6715549826622009,
+      "learning_rate": 5.428253615127921e-05,
+      "loss": 2.4454,
+      "step": 3281
+    },
+    {
+      "epoch": 43.77591973244147,
+      "grad_norm": 0.6230939626693726,
+      "learning_rate": 5.423804226918799e-05,
+      "loss": 2.2612,
+      "step": 3282
+    },
+    {
+      "epoch": 43.78929765886288,
+      "grad_norm": 0.7007676362991333,
+      "learning_rate": 5.419354838709678e-05,
+      "loss": 2.5626,
+      "step": 3283
+    },
+    {
+      "epoch": 43.802675585284284,
+      "grad_norm": 0.6708741784095764,
+      "learning_rate": 5.414905450500557e-05,
+      "loss": 2.4612,
+      "step": 3284
+    },
+    {
+      "epoch": 43.81605351170568,
+      "grad_norm": 0.6595252156257629,
+      "learning_rate": 5.4104560622914346e-05,
+      "loss": 2.4559,
+      "step": 3285
+    },
+    {
+      "epoch": 43.82943143812709,
+      "grad_norm": 0.627967119216919,
+      "learning_rate": 5.4060066740823135e-05,
+      "loss": 2.5175,
+      "step": 3286
+    },
+    {
+      "epoch": 43.84280936454849,
+      "grad_norm": 0.6663414835929871,
+      "learning_rate": 5.4015572858731924e-05,
+      "loss": 2.3807,
+      "step": 3287
+    },
+    {
+      "epoch": 43.8561872909699,
+      "grad_norm": 0.6707153916358948,
+      "learning_rate": 5.3971078976640707e-05,
+      "loss": 2.4467,
+      "step": 3288
+    },
+    {
+      "epoch": 43.869565217391305,
+      "grad_norm": 0.6654629707336426,
+      "learning_rate": 5.3926585094549496e-05,
+      "loss": 2.5185,
+      "step": 3289
+    },
+    {
+      "epoch": 43.88294314381271,
+      "grad_norm": 0.6775473356246948,
+      "learning_rate": 5.3882091212458285e-05,
+      "loss": 2.5433,
+      "step": 3290
+    },
+    {
+      "epoch": 43.896321070234116,
+      "grad_norm": 0.7619598507881165,
+      "learning_rate": 5.3837597330367074e-05,
+      "loss": 2.5873,
+      "step": 3291
+    },
+    {
+      "epoch": 43.90969899665552,
+      "grad_norm": 0.6738859415054321,
+      "learning_rate": 5.379310344827586e-05,
+      "loss": 2.7934,
+      "step": 3292
+    },
+    {
+      "epoch": 43.92307692307692,
+      "grad_norm": 0.6576844453811646,
+      "learning_rate": 5.3748609566184646e-05,
+      "loss": 2.4892,
+      "step": 3293
+    },
+    {
+      "epoch": 43.936454849498325,
+      "grad_norm": 0.7075631022453308,
+      "learning_rate": 5.3704115684093435e-05,
+      "loss": 2.6011,
+      "step": 3294
+    },
+    {
+      "epoch": 43.94983277591973,
+      "grad_norm": 0.6499934196472168,
+      "learning_rate": 5.3659621802002224e-05,
+      "loss": 2.5615,
+      "step": 3295
+    },
+    {
+      "epoch": 43.96321070234114,
+      "grad_norm": 0.6628806591033936,
+      "learning_rate": 5.361512791991101e-05,
+      "loss": 2.4376,
+      "step": 3296
+    },
+    {
+      "epoch": 43.97658862876254,
+      "grad_norm": 0.6394050121307373,
+      "learning_rate": 5.35706340378198e-05,
+      "loss": 2.5635,
+      "step": 3297
+    },
+    {
+      "epoch": 43.98996655518395,
+      "grad_norm": 0.6746408343315125,
+      "learning_rate": 5.3526140155728585e-05,
+      "loss": 2.676,
+      "step": 3298
+    },
+    {
+      "epoch": 44.0,
+      "grad_norm": 0.8157567381858826,
+      "learning_rate": 5.3481646273637374e-05,
+      "loss": 2.7095,
+      "step": 3299
+    },
+    {
+      "epoch": 44.013377926421406,
+      "grad_norm": 0.648750901222229,
+      "learning_rate": 5.343715239154616e-05,
+      "loss": 2.511,
+      "step": 3300
+    },
+    {
+      "epoch": 44.02675585284281,
+      "grad_norm": 0.646703839302063,
+      "learning_rate": 5.339265850945495e-05,
+      "loss": 2.4419,
+      "step": 3301
+    },
+    {
+      "epoch": 44.04013377926422,
+      "grad_norm": 0.6128621697425842,
+      "learning_rate": 5.334816462736374e-05,
+      "loss": 2.3166,
+      "step": 3302
+    },
+    {
+      "epoch": 44.05351170568562,
+      "grad_norm": 0.7439555525779724,
+      "learning_rate": 5.3303670745272524e-05,
+      "loss": 2.4277,
+      "step": 3303
+    },
+    {
+      "epoch": 44.06688963210702,
+      "grad_norm": 0.6968475580215454,
+      "learning_rate": 5.325917686318131e-05,
+      "loss": 2.4495,
+      "step": 3304
+    },
+    {
+      "epoch": 44.080267558528426,
+      "grad_norm": 0.6829770803451538,
+      "learning_rate": 5.32146829810901e-05,
+      "loss": 2.5492,
+      "step": 3305
+    },
+    {
+      "epoch": 44.09364548494983,
+      "grad_norm": 0.7078229784965515,
+      "learning_rate": 5.317018909899889e-05,
+      "loss": 2.3142,
+      "step": 3306
+    },
+    {
+      "epoch": 44.10702341137124,
+      "grad_norm": 0.7869110107421875,
+      "learning_rate": 5.312569521690768e-05,
+      "loss": 2.6079,
+      "step": 3307
+    },
+    {
+      "epoch": 44.12040133779264,
+      "grad_norm": 0.6723588705062866,
+      "learning_rate": 5.308120133481646e-05,
+      "loss": 2.5012,
+      "step": 3308
+    },
+    {
+      "epoch": 44.13377926421405,
+      "grad_norm": 0.6307100057601929,
+      "learning_rate": 5.303670745272525e-05,
+      "loss": 2.2734,
+      "step": 3309
+    },
+    {
+      "epoch": 44.147157190635454,
+      "grad_norm": 0.7307317852973938,
+      "learning_rate": 5.299221357063404e-05,
+      "loss": 2.5035,
+      "step": 3310
+    },
+    {
+      "epoch": 44.16053511705686,
+      "grad_norm": 0.6610035300254822,
+      "learning_rate": 5.294771968854283e-05,
+      "loss": 2.5255,
+      "step": 3311
+    },
+    {
+      "epoch": 44.17391304347826,
+      "grad_norm": 0.6795505285263062,
+      "learning_rate": 5.290322580645162e-05,
+      "loss": 2.4404,
+      "step": 3312
+    },
+    {
+      "epoch": 44.187290969899664,
+      "grad_norm": 0.8478922843933105,
+      "learning_rate": 5.28587319243604e-05,
+      "loss": 2.5065,
+      "step": 3313
+    },
+    {
+      "epoch": 44.20066889632107,
+      "grad_norm": 0.6836498975753784,
+      "learning_rate": 5.281423804226919e-05,
+      "loss": 2.5934,
+      "step": 3314
+    },
+    {
+      "epoch": 44.214046822742475,
+      "grad_norm": 0.6741980314254761,
+      "learning_rate": 5.276974416017798e-05,
+      "loss": 2.4481,
+      "step": 3315
+    },
+    {
+      "epoch": 44.22742474916388,
+      "grad_norm": 0.6965019702911377,
+      "learning_rate": 5.272525027808677e-05,
+      "loss": 2.3167,
+      "step": 3316
+    },
+    {
+      "epoch": 44.240802675585286,
+      "grad_norm": 0.7769578695297241,
+      "learning_rate": 5.268075639599556e-05,
+      "loss": 2.3156,
+      "step": 3317
+    },
+    {
+      "epoch": 44.25418060200669,
+      "grad_norm": 0.6968414783477783,
+      "learning_rate": 5.263626251390434e-05,
+      "loss": 2.5261,
+      "step": 3318
+    },
+    {
+      "epoch": 44.2675585284281,
+      "grad_norm": 0.6269694566726685,
+      "learning_rate": 5.259176863181313e-05,
+      "loss": 2.3642,
+      "step": 3319
+    },
+    {
+      "epoch": 44.280936454849495,
+      "grad_norm": 0.6887113451957703,
+      "learning_rate": 5.254727474972192e-05,
+      "loss": 2.5911,
+      "step": 3320
+    },
+    {
+      "epoch": 44.2943143812709,
+      "grad_norm": 0.6772909760475159,
+      "learning_rate": 5.250278086763071e-05,
+      "loss": 2.5551,
+      "step": 3321
+    },
+    {
+      "epoch": 44.30769230769231,
+      "grad_norm": 0.7027330994606018,
+      "learning_rate": 5.24582869855395e-05,
+      "loss": 2.7092,
+      "step": 3322
+    },
+    {
+      "epoch": 44.32107023411371,
+      "grad_norm": 0.6497491598129272,
+      "learning_rate": 5.241379310344828e-05,
+      "loss": 2.4204,
+      "step": 3323
+    },
+    {
+      "epoch": 44.33444816053512,
+      "grad_norm": 0.9224070906639099,
+      "learning_rate": 5.236929922135707e-05,
+      "loss": 2.3124,
+      "step": 3324
+    },
+    {
+      "epoch": 44.34782608695652,
+      "grad_norm": 0.6555243730545044,
+      "learning_rate": 5.232480533926586e-05,
+      "loss": 2.1393,
+      "step": 3325
+    },
+    {
+      "epoch": 44.36120401337793,
+      "grad_norm": 1.082617163658142,
+      "learning_rate": 5.2280311457174634e-05,
+      "loss": 2.2029,
+      "step": 3326
+    },
+    {
+      "epoch": 44.374581939799334,
+      "grad_norm": 0.6710803508758545,
+      "learning_rate": 5.223581757508342e-05,
+      "loss": 2.5351,
+      "step": 3327
+    },
+    {
+      "epoch": 44.38795986622073,
+      "grad_norm": 0.6694113612174988,
+      "learning_rate": 5.219132369299221e-05,
+      "loss": 2.4342,
+      "step": 3328
+    },
+    {
+      "epoch": 44.40133779264214,
+      "grad_norm": 0.6941688656806946,
+      "learning_rate": 5.2146829810901e-05,
+      "loss": 2.3725,
+      "step": 3329
+    },
+    {
+      "epoch": 44.414715719063544,
+      "grad_norm": 0.6413290500640869,
+      "learning_rate": 5.2102335928809784e-05,
+      "loss": 2.4962,
+      "step": 3330
+    },
+    {
+      "epoch": 44.42809364548495,
+      "grad_norm": 0.6977859139442444,
+      "learning_rate": 5.205784204671857e-05,
+      "loss": 2.5549,
+      "step": 3331
+    },
+    {
+      "epoch": 44.441471571906355,
+      "grad_norm": 0.7250884771347046,
+      "learning_rate": 5.201334816462736e-05,
+      "loss": 2.6143,
+      "step": 3332
+    },
+    {
+      "epoch": 44.45484949832776,
+      "grad_norm": 0.7633907794952393,
+      "learning_rate": 5.196885428253615e-05,
+      "loss": 2.4557,
+      "step": 3333
+    },
+    {
+      "epoch": 44.468227424749166,
+      "grad_norm": 0.7065114378929138,
+      "learning_rate": 5.192436040044494e-05,
+      "loss": 2.5935,
+      "step": 3334
+    },
+    {
+      "epoch": 44.48160535117057,
+      "grad_norm": 0.6201562285423279,
+      "learning_rate": 5.187986651835372e-05,
+      "loss": 2.3543,
+      "step": 3335
+    },
+    {
+      "epoch": 44.49498327759197,
+      "grad_norm": 0.6835589408874512,
+      "learning_rate": 5.183537263626251e-05,
+      "loss": 2.5158,
+      "step": 3336
+    },
+    {
+      "epoch": 44.508361204013376,
+      "grad_norm": 0.7697896957397461,
+      "learning_rate": 5.17908787541713e-05,
+      "loss": 2.2409,
+      "step": 3337
+    },
+    {
+      "epoch": 44.52173913043478,
+      "grad_norm": 0.7114497423171997,
+      "learning_rate": 5.174638487208009e-05,
+      "loss": 2.6477,
+      "step": 3338
+    },
+    {
+      "epoch": 44.53511705685619,
+      "grad_norm": 0.6574714183807373,
+      "learning_rate": 5.170189098998888e-05,
+      "loss": 2.3153,
+      "step": 3339
+    },
+    {
+      "epoch": 44.54849498327759,
+      "grad_norm": 0.6645333170890808,
+      "learning_rate": 5.165739710789766e-05,
+      "loss": 2.4608,
+      "step": 3340
+    },
+    {
+      "epoch": 44.561872909699,
+      "grad_norm": 0.6737465262413025,
+      "learning_rate": 5.161290322580645e-05,
+      "loss": 2.3891,
+      "step": 3341
+    },
+    {
+      "epoch": 44.575250836120404,
+      "grad_norm": 0.732427179813385,
+      "learning_rate": 5.156840934371524e-05,
+      "loss": 2.5159,
+      "step": 3342
+    },
+    {
+      "epoch": 44.58862876254181,
+      "grad_norm": 0.6582529544830322,
+      "learning_rate": 5.152391546162403e-05,
+      "loss": 2.4066,
+      "step": 3343
+    },
+    {
+      "epoch": 44.60200668896321,
+      "grad_norm": 0.7296741008758545,
+      "learning_rate": 5.147942157953282e-05,
+      "loss": 2.5349,
+      "step": 3344
+    },
+    {
+      "epoch": 44.61538461538461,
+      "grad_norm": 0.6754583716392517,
+      "learning_rate": 5.14349276974416e-05,
+      "loss": 2.5574,
+      "step": 3345
+    },
+    {
+      "epoch": 44.62876254180602,
+      "grad_norm": 0.6714019775390625,
+      "learning_rate": 5.139043381535039e-05,
+      "loss": 2.3795,
+      "step": 3346
+    },
+    {
+      "epoch": 44.642140468227424,
+      "grad_norm": 0.818821132183075,
+      "learning_rate": 5.134593993325918e-05,
+      "loss": 2.6163,
+      "step": 3347
+    },
+    {
+      "epoch": 44.65551839464883,
+      "grad_norm": 0.6695653200149536,
+      "learning_rate": 5.130144605116797e-05,
+      "loss": 2.5223,
+      "step": 3348
+    },
+    {
+      "epoch": 44.668896321070235,
+      "grad_norm": 0.674275815486908,
+      "learning_rate": 5.125695216907676e-05,
+      "loss": 2.3804,
+      "step": 3349
+    },
+    {
+      "epoch": 44.68227424749164,
+      "grad_norm": 0.6844233870506287,
+      "learning_rate": 5.121245828698554e-05,
+      "loss": 2.5665,
+      "step": 3350
+    },
+    {
+      "epoch": 44.69565217391305,
+      "grad_norm": 0.7201416492462158,
+      "learning_rate": 5.116796440489433e-05,
+      "loss": 2.5106,
+      "step": 3351
+    },
+    {
+      "epoch": 44.709030100334445,
+      "grad_norm": 0.6668140888214111,
+      "learning_rate": 5.112347052280312e-05,
+      "loss": 2.6049,
+      "step": 3352
+    },
+    {
+      "epoch": 44.72240802675585,
+      "grad_norm": 0.7152537703514099,
+      "learning_rate": 5.107897664071191e-05,
+      "loss": 2.4075,
+      "step": 3353
+    },
+    {
+      "epoch": 44.735785953177256,
+      "grad_norm": 0.6661379933357239,
+      "learning_rate": 5.10344827586207e-05,
+      "loss": 2.5558,
+      "step": 3354
+    },
+    {
+      "epoch": 44.74916387959866,
+      "grad_norm": 0.6641297936439514,
+      "learning_rate": 5.098998887652948e-05,
+      "loss": 2.4548,
+      "step": 3355
+    },
+    {
+      "epoch": 44.76254180602007,
+      "grad_norm": 0.689754843711853,
+      "learning_rate": 5.094549499443827e-05,
+      "loss": 2.4735,
+      "step": 3356
+    },
+    {
+      "epoch": 44.77591973244147,
+      "grad_norm": 0.6680580377578735,
+      "learning_rate": 5.090100111234706e-05,
+      "loss": 2.483,
+      "step": 3357
+    },
+    {
+      "epoch": 44.78929765886288,
+      "grad_norm": 1.1155476570129395,
+      "learning_rate": 5.0856507230255847e-05,
+      "loss": 2.6322,
+      "step": 3358
+    },
+    {
+      "epoch": 44.802675585284284,
+      "grad_norm": 0.6346137523651123,
+      "learning_rate": 5.0812013348164636e-05,
+      "loss": 2.4057,
+      "step": 3359
+    },
+    {
+      "epoch": 44.81605351170568,
+      "grad_norm": 0.6553255319595337,
+      "learning_rate": 5.076751946607342e-05,
+      "loss": 2.4353,
+      "step": 3360
+    },
+    {
+      "epoch": 44.82943143812709,
+      "grad_norm": 0.7519358992576599,
+      "learning_rate": 5.072302558398221e-05,
+      "loss": 2.5155,
+      "step": 3361
+    },
+    {
+      "epoch": 44.84280936454849,
+      "grad_norm": 0.6849926114082336,
+      "learning_rate": 5.0678531701890996e-05,
+      "loss": 2.6557,
+      "step": 3362
+    },
+    {
+      "epoch": 44.8561872909699,
+      "grad_norm": 0.6906036138534546,
+      "learning_rate": 5.0634037819799786e-05,
+      "loss": 2.5513,
+      "step": 3363
+    },
+    {
+      "epoch": 44.869565217391305,
+      "grad_norm": 0.712207019329071,
+      "learning_rate": 5.0589543937708575e-05,
+      "loss": 2.5477,
+      "step": 3364
+    },
+    {
+      "epoch": 44.88294314381271,
+      "grad_norm": 0.7856250405311584,
+      "learning_rate": 5.054505005561736e-05,
+      "loss": 2.5645,
+      "step": 3365
+    },
+    {
+      "epoch": 44.896321070234116,
+      "grad_norm": 0.6578722596168518,
+      "learning_rate": 5.0500556173526146e-05,
+      "loss": 2.5097,
+      "step": 3366
+    },
+    {
+      "epoch": 44.90969899665552,
+      "grad_norm": 0.6745465397834778,
+      "learning_rate": 5.045606229143492e-05,
+      "loss": 2.4405,
+      "step": 3367
+    },
+    {
+      "epoch": 44.92307692307692,
+      "grad_norm": 0.6686838865280151,
+      "learning_rate": 5.041156840934371e-05,
+      "loss": 2.4774,
+      "step": 3368
+    },
+    {
+      "epoch": 44.936454849498325,
+      "grad_norm": 0.6725925803184509,
+      "learning_rate": 5.03670745272525e-05,
+      "loss": 2.5365,
+      "step": 3369
+    },
+    {
+      "epoch": 44.94983277591973,
+      "grad_norm": 0.7038573026657104,
+      "learning_rate": 5.032258064516129e-05,
+      "loss": 2.5415,
+      "step": 3370
+    },
+    {
+      "epoch": 44.96321070234114,
+      "grad_norm": 0.670062780380249,
+      "learning_rate": 5.027808676307008e-05,
+      "loss": 2.5289,
+      "step": 3371
+    },
+    {
+      "epoch": 44.97658862876254,
+      "grad_norm": 0.6854854226112366,
+      "learning_rate": 5.023359288097886e-05,
+      "loss": 2.5487,
+      "step": 3372
+    },
+    {
+      "epoch": 44.98996655518395,
+      "grad_norm": 0.840345025062561,
+      "learning_rate": 5.018909899888765e-05,
+      "loss": 2.6067,
+      "step": 3373
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 0.790497362613678,
+      "learning_rate": 5.014460511679644e-05,
+      "loss": 2.3824,
+      "step": 3374
+    },
+    {
+      "epoch": 45.013377926421406,
+      "grad_norm": 0.6614347100257874,
+      "learning_rate": 5.010011123470523e-05,
+      "loss": 2.3392,
+      "step": 3375
+    },
+    {
+      "epoch": 45.02675585284281,
+      "grad_norm": 0.6427727341651917,
+      "learning_rate": 5.005561735261402e-05,
+      "loss": 2.526,
+      "step": 3376
+    },
+    {
+      "epoch": 45.04013377926422,
+      "grad_norm": 0.6234596967697144,
+      "learning_rate": 5.00111234705228e-05,
+      "loss": 2.4324,
+      "step": 3377
+    },
+    {
+      "epoch": 45.05351170568562,
+      "grad_norm": 0.6480961441993713,
+      "learning_rate": 4.996662958843159e-05,
+      "loss": 2.4503,
+      "step": 3378
+    },
+    {
+      "epoch": 45.06688963210702,
+      "grad_norm": 0.6448456048965454,
+      "learning_rate": 4.992213570634038e-05,
+      "loss": 2.5241,
+      "step": 3379
+    },
+    {
+      "epoch": 45.080267558528426,
+      "grad_norm": 0.6678739786148071,
+      "learning_rate": 4.987764182424917e-05,
+      "loss": 2.5715,
+      "step": 3380
+    },
+    {
+      "epoch": 45.09364548494983,
+      "grad_norm": 0.6756695508956909,
+      "learning_rate": 4.983314794215796e-05,
+      "loss": 2.5005,
+      "step": 3381
+    },
+    {
+      "epoch": 45.10702341137124,
+      "grad_norm": 0.6389010548591614,
+      "learning_rate": 4.978865406006674e-05,
+      "loss": 2.3691,
+      "step": 3382
+    },
+    {
+      "epoch": 45.12040133779264,
+      "grad_norm": 0.66241854429245,
+      "learning_rate": 4.974416017797553e-05,
+      "loss": 2.251,
+      "step": 3383
+    },
+    {
+      "epoch": 45.13377926421405,
+      "grad_norm": 0.7921980619430542,
+      "learning_rate": 4.969966629588432e-05,
+      "loss": 2.2895,
+      "step": 3384
+    },
+    {
+      "epoch": 45.147157190635454,
+      "grad_norm": 0.6602982878684998,
+      "learning_rate": 4.9655172413793107e-05,
+      "loss": 2.4911,
+      "step": 3385
+    },
+    {
+      "epoch": 45.16053511705686,
+      "grad_norm": 0.6390572190284729,
+      "learning_rate": 4.9610678531701896e-05,
+      "loss": 2.2301,
+      "step": 3386
+    },
+    {
+      "epoch": 45.17391304347826,
+      "grad_norm": 0.6479654312133789,
+      "learning_rate": 4.956618464961068e-05,
+      "loss": 2.416,
+      "step": 3387
+    },
+    {
+      "epoch": 45.187290969899664,
+      "grad_norm": 0.6543484330177307,
+      "learning_rate": 4.952169076751947e-05,
+      "loss": 2.6481,
+      "step": 3388
+    },
+    {
+      "epoch": 45.20066889632107,
+      "grad_norm": 0.6472519636154175,
+      "learning_rate": 4.9477196885428256e-05,
+      "loss": 2.2218,
+      "step": 3389
+    },
+    {
+      "epoch": 45.214046822742475,
+      "grad_norm": 0.6654759049415588,
+      "learning_rate": 4.9432703003337046e-05,
+      "loss": 2.559,
+      "step": 3390
+    },
+    {
+      "epoch": 45.22742474916388,
+      "grad_norm": 0.833139955997467,
+      "learning_rate": 4.9388209121245835e-05,
+      "loss": 2.4773,
+      "step": 3391
+    },
+    {
+      "epoch": 45.240802675585286,
+      "grad_norm": 0.7028539180755615,
+      "learning_rate": 4.934371523915462e-05,
+      "loss": 2.5329,
+      "step": 3392
+    },
+    {
+      "epoch": 45.25418060200669,
+      "grad_norm": 0.6716699600219727,
+      "learning_rate": 4.9299221357063406e-05,
+      "loss": 2.6069,
+      "step": 3393
+    },
+    {
+      "epoch": 45.2675585284281,
+      "grad_norm": 0.8206474184989929,
+      "learning_rate": 4.9254727474972196e-05,
+      "loss": 2.5227,
+      "step": 3394
+    },
+    {
+      "epoch": 45.280936454849495,
+      "grad_norm": 0.6916563510894775,
+      "learning_rate": 4.9210233592880985e-05,
+      "loss": 2.5564,
+      "step": 3395
+    },
+    {
+      "epoch": 45.2943143812709,
+      "grad_norm": 0.67179936170578,
+      "learning_rate": 4.9165739710789774e-05,
+      "loss": 2.5468,
+      "step": 3396
+    },
+    {
+      "epoch": 45.30769230769231,
+      "grad_norm": 0.7964248657226562,
+      "learning_rate": 4.9121245828698556e-05,
+      "loss": 2.6126,
+      "step": 3397
+    },
+    {
+      "epoch": 45.32107023411371,
+      "grad_norm": 0.6877651810646057,
+      "learning_rate": 4.907675194660734e-05,
+      "loss": 2.6272,
+      "step": 3398
+    },
+    {
+      "epoch": 45.33444816053512,
+      "grad_norm": 0.7093799710273743,
+      "learning_rate": 4.903225806451613e-05,
+      "loss": 2.4046,
+      "step": 3399
+    },
+    {
+      "epoch": 45.34782608695652,
+      "grad_norm": 0.772832453250885,
+      "learning_rate": 4.898776418242492e-05,
+      "loss": 2.5126,
+      "step": 3400
+    },
+    {
+      "epoch": 45.36120401337793,
+      "grad_norm": 0.7250183820724487,
+      "learning_rate": 4.8943270300333706e-05,
+      "loss": 2.4396,
+      "step": 3401
+    },
+    {
+      "epoch": 45.374581939799334,
+      "grad_norm": 0.6946377754211426,
+      "learning_rate": 4.8898776418242495e-05,
+      "loss": 2.6412,
+      "step": 3402
+    },
+    {
+      "epoch": 45.38795986622073,
+      "grad_norm": 0.6398693919181824,
+      "learning_rate": 4.885428253615128e-05,
+      "loss": 2.4037,
+      "step": 3403
+    },
+    {
+      "epoch": 45.40133779264214,
+      "grad_norm": 0.7096537947654724,
+      "learning_rate": 4.880978865406007e-05,
+      "loss": 2.5131,
+      "step": 3404
+    },
+    {
+      "epoch": 45.414715719063544,
+      "grad_norm": 0.6392034888267517,
+      "learning_rate": 4.8765294771968856e-05,
+      "loss": 2.4144,
+      "step": 3405
+    },
+    {
+      "epoch": 45.42809364548495,
+      "grad_norm": 0.647911787033081,
+      "learning_rate": 4.8720800889877645e-05,
+      "loss": 2.4081,
+      "step": 3406
+    },
+    {
+      "epoch": 45.441471571906355,
+      "grad_norm": 0.8484491109848022,
+      "learning_rate": 4.8676307007786434e-05,
+      "loss": 2.2981,
+      "step": 3407
+    },
+    {
+      "epoch": 45.45484949832776,
+      "grad_norm": 0.6788907051086426,
+      "learning_rate": 4.863181312569522e-05,
+      "loss": 2.3547,
+      "step": 3408
+    },
+    {
+      "epoch": 45.468227424749166,
+      "grad_norm": 0.7060588002204895,
+      "learning_rate": 4.8587319243604006e-05,
+      "loss": 2.506,
+      "step": 3409
+    },
+    {
+      "epoch": 45.48160535117057,
+      "grad_norm": 0.7509225606918335,
+      "learning_rate": 4.8542825361512795e-05,
+      "loss": 2.4514,
+      "step": 3410
+    },
+    {
+      "epoch": 45.49498327759197,
+      "grad_norm": 0.6925180554389954,
+      "learning_rate": 4.8498331479421584e-05,
+      "loss": 2.5358,
+      "step": 3411
+    },
+    {
+      "epoch": 45.508361204013376,
+      "grad_norm": 0.6838158965110779,
+      "learning_rate": 4.8453837597330373e-05,
+      "loss": 2.2683,
+      "step": 3412
+    },
+    {
+      "epoch": 45.52173913043478,
+      "grad_norm": 0.6462475657463074,
+      "learning_rate": 4.8409343715239156e-05,
+      "loss": 2.3029,
+      "step": 3413
+    },
+    {
+      "epoch": 45.53511705685619,
+      "grad_norm": 0.6882359981536865,
+      "learning_rate": 4.8364849833147945e-05,
+      "loss": 2.6943,
+      "step": 3414
+    },
+    {
+      "epoch": 45.54849498327759,
+      "grad_norm": 0.6654660105705261,
+      "learning_rate": 4.8320355951056734e-05,
+      "loss": 2.2613,
+      "step": 3415
+    },
+    {
+      "epoch": 45.561872909699,
+      "grad_norm": 0.8237322568893433,
+      "learning_rate": 4.827586206896552e-05,
+      "loss": 2.2294,
+      "step": 3416
+    },
+    {
+      "epoch": 45.575250836120404,
+      "grad_norm": 0.7393941283226013,
+      "learning_rate": 4.823136818687431e-05,
+      "loss": 2.5682,
+      "step": 3417
+    },
+    {
+      "epoch": 45.58862876254181,
+      "grad_norm": 0.7490308880805969,
+      "learning_rate": 4.8186874304783095e-05,
+      "loss": 2.5922,
+      "step": 3418
+    },
+    {
+      "epoch": 45.60200668896321,
+      "grad_norm": 0.7629905343055725,
+      "learning_rate": 4.814238042269188e-05,
+      "loss": 2.6547,
+      "step": 3419
+    },
+    {
+      "epoch": 45.61538461538461,
+      "grad_norm": 0.7757886648178101,
+      "learning_rate": 4.8097886540600666e-05,
+      "loss": 2.4786,
+      "step": 3420
+    },
+    {
+      "epoch": 45.62876254180602,
+      "grad_norm": 0.6690565943717957,
+      "learning_rate": 4.8053392658509456e-05,
+      "loss": 2.4899,
+      "step": 3421
+    },
+    {
+      "epoch": 45.642140468227424,
+      "grad_norm": 0.674462616443634,
+      "learning_rate": 4.8008898776418245e-05,
+      "loss": 2.3315,
+      "step": 3422
+    },
+    {
+      "epoch": 45.65551839464883,
+      "grad_norm": 0.7295622229576111,
+      "learning_rate": 4.7964404894327034e-05,
+      "loss": 2.6889,
+      "step": 3423
+    },
+    {
+      "epoch": 45.668896321070235,
+      "grad_norm": 0.7250505685806274,
+      "learning_rate": 4.7919911012235816e-05,
+      "loss": 2.2723,
+      "step": 3424
+    },
+    {
+      "epoch": 45.68227424749164,
+      "grad_norm": 1.0119596719741821,
+      "learning_rate": 4.7875417130144605e-05,
+      "loss": 2.4533,
+      "step": 3425
+    },
+    {
+      "epoch": 45.69565217391305,
+      "grad_norm": 0.6516237854957581,
+      "learning_rate": 4.7830923248053395e-05,
+      "loss": 2.3416,
+      "step": 3426
+    },
+    {
+      "epoch": 45.709030100334445,
+      "grad_norm": 0.6897451281547546,
+      "learning_rate": 4.7786429365962184e-05,
+      "loss": 2.532,
+      "step": 3427
+    },
+    {
+      "epoch": 45.72240802675585,
+      "grad_norm": 0.6455211043357849,
+      "learning_rate": 4.774193548387097e-05,
+      "loss": 2.2748,
+      "step": 3428
+    },
+    {
+      "epoch": 45.735785953177256,
+      "grad_norm": 0.6655951142311096,
+      "learning_rate": 4.7697441601779755e-05,
+      "loss": 2.2649,
+      "step": 3429
+    },
+    {
+      "epoch": 45.74916387959866,
+      "grad_norm": 0.8087528944015503,
+      "learning_rate": 4.7652947719688545e-05,
+      "loss": 2.4608,
+      "step": 3430
+    },
+    {
+      "epoch": 45.76254180602007,
+      "grad_norm": 0.661077618598938,
+      "learning_rate": 4.7608453837597334e-05,
+      "loss": 2.3469,
+      "step": 3431
+    },
+    {
+      "epoch": 45.77591973244147,
+      "grad_norm": 0.7252074480056763,
+      "learning_rate": 4.756395995550612e-05,
+      "loss": 2.6173,
+      "step": 3432
+    },
+    {
+      "epoch": 45.78929765886288,
+      "grad_norm": 0.7233737111091614,
+      "learning_rate": 4.751946607341491e-05,
+      "loss": 2.4967,
+      "step": 3433
+    },
+    {
+      "epoch": 45.802675585284284,
+      "grad_norm": 0.6680254340171814,
+      "learning_rate": 4.7474972191323694e-05,
+      "loss": 2.4175,
+      "step": 3434
+    },
+    {
+      "epoch": 45.81605351170568,
+      "grad_norm": 0.6761989593505859,
+      "learning_rate": 4.7430478309232484e-05,
+      "loss": 2.5149,
+      "step": 3435
+    },
+    {
+      "epoch": 45.82943143812709,
+      "grad_norm": 0.6871892213821411,
+      "learning_rate": 4.738598442714127e-05,
+      "loss": 2.4105,
+      "step": 3436
+    },
+    {
+      "epoch": 45.84280936454849,
+      "grad_norm": 0.7473261952400208,
+      "learning_rate": 4.734149054505006e-05,
+      "loss": 2.5763,
+      "step": 3437
+    },
+    {
+      "epoch": 45.8561872909699,
+      "grad_norm": 0.703223705291748,
+      "learning_rate": 4.729699666295885e-05,
+      "loss": 2.4285,
+      "step": 3438
+    },
+    {
+      "epoch": 45.869565217391305,
+      "grad_norm": 0.7379417419433594,
+      "learning_rate": 4.7252502780867633e-05,
+      "loss": 2.5296,
+      "step": 3439
+    },
+    {
+      "epoch": 45.88294314381271,
+      "grad_norm": 0.8234548568725586,
+      "learning_rate": 4.7208008898776416e-05,
+      "loss": 2.4177,
+      "step": 3440
+    },
+    {
+      "epoch": 45.896321070234116,
+      "grad_norm": 0.6252479553222656,
+      "learning_rate": 4.7163515016685205e-05,
+      "loss": 2.174,
+      "step": 3441
+    },
+    {
+      "epoch": 45.90969899665552,
+      "grad_norm": 0.6623761057853699,
+      "learning_rate": 4.7119021134593994e-05,
+      "loss": 2.4613,
+      "step": 3442
+    },
+    {
+      "epoch": 45.92307692307692,
+      "grad_norm": 0.7237697839736938,
+      "learning_rate": 4.707452725250278e-05,
+      "loss": 2.5461,
+      "step": 3443
+    },
+    {
+      "epoch": 45.936454849498325,
+      "grad_norm": 0.7058880925178528,
+      "learning_rate": 4.703003337041157e-05,
+      "loss": 2.3805,
+      "step": 3444
+    },
+    {
+      "epoch": 45.94983277591973,
+      "grad_norm": 0.7062870860099792,
+      "learning_rate": 4.6985539488320355e-05,
+      "loss": 2.6266,
+      "step": 3445
+    },
+    {
+      "epoch": 45.96321070234114,
+      "grad_norm": 0.6760918498039246,
+      "learning_rate": 4.6941045606229144e-05,
+      "loss": 2.5284,
+      "step": 3446
+    },
+    {
+      "epoch": 45.97658862876254,
+      "grad_norm": 0.6973199844360352,
+      "learning_rate": 4.689655172413793e-05,
+      "loss": 2.3388,
+      "step": 3447
+    },
+    {
+      "epoch": 45.98996655518395,
+      "grad_norm": 0.6965627670288086,
+      "learning_rate": 4.685205784204672e-05,
+      "loss": 2.4212,
+      "step": 3448
+    },
+    {
+      "epoch": 46.0,
+      "grad_norm": 0.8428727388381958,
+      "learning_rate": 4.680756395995551e-05,
+      "loss": 2.5794,
+      "step": 3449
+    },
+    {
+      "epoch": 46.013377926421406,
+      "grad_norm": 0.6538412570953369,
+      "learning_rate": 4.6763070077864294e-05,
+      "loss": 2.3321,
+      "step": 3450
+    },
+    {
+      "epoch": 46.02675585284281,
+      "grad_norm": 0.6650435328483582,
+      "learning_rate": 4.671857619577308e-05,
+      "loss": 2.3299,
+      "step": 3451
+    },
+    {
+      "epoch": 46.04013377926422,
+      "grad_norm": 0.6518568396568298,
+      "learning_rate": 4.667408231368187e-05,
+      "loss": 2.3671,
+      "step": 3452
+    },
+    {
+      "epoch": 46.05351170568562,
+      "grad_norm": 0.6615071296691895,
+      "learning_rate": 4.662958843159066e-05,
+      "loss": 2.2856,
+      "step": 3453
+    },
+    {
+      "epoch": 46.06688963210702,
+      "grad_norm": 0.7280906438827515,
+      "learning_rate": 4.658509454949945e-05,
+      "loss": 2.6337,
+      "step": 3454
+    },
+    {
+      "epoch": 46.080267558528426,
+      "grad_norm": 0.6038973927497864,
+      "learning_rate": 4.654060066740823e-05,
+      "loss": 2.0686,
+      "step": 3455
+    },
+    {
+      "epoch": 46.09364548494983,
+      "grad_norm": 0.7241243720054626,
+      "learning_rate": 4.649610678531702e-05,
+      "loss": 2.4142,
+      "step": 3456
+    },
+    {
+      "epoch": 46.10702341137124,
+      "grad_norm": 0.716877281665802,
+      "learning_rate": 4.645161290322581e-05,
+      "loss": 2.6186,
+      "step": 3457
+    },
+    {
+      "epoch": 46.12040133779264,
+      "grad_norm": 0.7181987166404724,
+      "learning_rate": 4.64071190211346e-05,
+      "loss": 2.4412,
+      "step": 3458
+    },
+    {
+      "epoch": 46.13377926421405,
+      "grad_norm": 0.699880838394165,
+      "learning_rate": 4.636262513904339e-05,
+      "loss": 2.5438,
+      "step": 3459
+    },
+    {
+      "epoch": 46.147157190635454,
+      "grad_norm": 0.7127376198768616,
+      "learning_rate": 4.6318131256952165e-05,
+      "loss": 2.3073,
+      "step": 3460
+    },
+    {
+      "epoch": 46.16053511705686,
+      "grad_norm": 0.7011304497718811,
+      "learning_rate": 4.6273637374860954e-05,
+      "loss": 2.4327,
+      "step": 3461
+    },
+    {
+      "epoch": 46.17391304347826,
+      "grad_norm": 0.7037985920906067,
+      "learning_rate": 4.6229143492769744e-05,
+      "loss": 2.444,
+      "step": 3462
+    },
+    {
+      "epoch": 46.187290969899664,
+      "grad_norm": 0.6529151201248169,
+      "learning_rate": 4.618464961067853e-05,
+      "loss": 2.3799,
+      "step": 3463
+    },
+    {
+      "epoch": 46.20066889632107,
+      "grad_norm": 0.7062216997146606,
+      "learning_rate": 4.614015572858732e-05,
+      "loss": 2.4849,
+      "step": 3464
+    },
+    {
+      "epoch": 46.214046822742475,
+      "grad_norm": 0.7128186821937561,
+      "learning_rate": 4.6095661846496104e-05,
+      "loss": 2.3768,
+      "step": 3465
+    },
+    {
+      "epoch": 46.22742474916388,
+      "grad_norm": 0.629882276058197,
+      "learning_rate": 4.6051167964404894e-05,
+      "loss": 2.4646,
+      "step": 3466
+    },
+    {
+      "epoch": 46.240802675585286,
+      "grad_norm": 0.6844696998596191,
+      "learning_rate": 4.600667408231368e-05,
+      "loss": 2.6883,
+      "step": 3467
+    },
+    {
+      "epoch": 46.25418060200669,
+      "grad_norm": 0.6740692853927612,
+      "learning_rate": 4.596218020022247e-05,
+      "loss": 2.4321,
+      "step": 3468
+    },
+    {
+      "epoch": 46.2675585284281,
+      "grad_norm": 0.6562219262123108,
+      "learning_rate": 4.591768631813126e-05,
+      "loss": 2.6083,
+      "step": 3469
+    },
+    {
+      "epoch": 46.280936454849495,
+      "grad_norm": 0.7180445194244385,
+      "learning_rate": 4.5873192436040043e-05,
+      "loss": 2.4661,
+      "step": 3470
+    },
+    {
+      "epoch": 46.2943143812709,
+      "grad_norm": 0.683229923248291,
+      "learning_rate": 4.582869855394883e-05,
+      "loss": 2.3699,
+      "step": 3471
+    },
+    {
+      "epoch": 46.30769230769231,
+      "grad_norm": 0.7104426622390747,
+      "learning_rate": 4.578420467185762e-05,
+      "loss": 2.4708,
+      "step": 3472
+    },
+    {
+      "epoch": 46.32107023411371,
+      "grad_norm": 0.6398786306381226,
+      "learning_rate": 4.573971078976641e-05,
+      "loss": 2.3402,
+      "step": 3473
+    },
+    {
+      "epoch": 46.33444816053512,
+      "grad_norm": 0.6786426901817322,
+      "learning_rate": 4.56952169076752e-05,
+      "loss": 2.4108,
+      "step": 3474
+    },
+    {
+      "epoch": 46.34782608695652,
+      "grad_norm": 0.718450129032135,
+      "learning_rate": 4.565072302558398e-05,
+      "loss": 2.5797,
+      "step": 3475
+    },
+    {
+      "epoch": 46.36120401337793,
+      "grad_norm": 0.6450575590133667,
+      "learning_rate": 4.560622914349277e-05,
+      "loss": 2.0375,
+      "step": 3476
+    },
+    {
+      "epoch": 46.374581939799334,
+      "grad_norm": 0.6493234038352966,
+      "learning_rate": 4.556173526140156e-05,
+      "loss": 2.4572,
+      "step": 3477
+    },
+    {
+      "epoch": 46.38795986622073,
+      "grad_norm": 0.5922126173973083,
+      "learning_rate": 4.551724137931035e-05,
+      "loss": 2.2419,
+      "step": 3478
+    },
+    {
+      "epoch": 46.40133779264214,
+      "grad_norm": 0.6579440832138062,
+      "learning_rate": 4.547274749721914e-05,
+      "loss": 2.2189,
+      "step": 3479
+    },
+    {
+      "epoch": 46.414715719063544,
+      "grad_norm": 0.6737288236618042,
+      "learning_rate": 4.542825361512792e-05,
+      "loss": 2.4462,
+      "step": 3480
+    },
+    {
+      "epoch": 46.42809364548495,
+      "grad_norm": 0.7465442419052124,
+      "learning_rate": 4.5383759733036704e-05,
+      "loss": 2.2322,
+      "step": 3481
+    },
+    {
+      "epoch": 46.441471571906355,
+      "grad_norm": 0.8456260561943054,
+      "learning_rate": 4.533926585094549e-05,
+      "loss": 2.7195,
+      "step": 3482
+    },
+    {
+      "epoch": 46.45484949832776,
+      "grad_norm": 0.6909487843513489,
+      "learning_rate": 4.529477196885428e-05,
+      "loss": 2.3794,
+      "step": 3483
+    },
+    {
+      "epoch": 46.468227424749166,
+      "grad_norm": 0.6962025761604309,
+      "learning_rate": 4.525027808676307e-05,
+      "loss": 2.4683,
+      "step": 3484
+    },
+    {
+      "epoch": 46.48160535117057,
+      "grad_norm": 0.6593672037124634,
+      "learning_rate": 4.520578420467186e-05,
+      "loss": 2.5137,
+      "step": 3485
+    },
+    {
+      "epoch": 46.49498327759197,
+      "grad_norm": 0.8271158933639526,
+      "learning_rate": 4.516129032258064e-05,
+      "loss": 2.4657,
+      "step": 3486
+    },
+    {
+      "epoch": 46.508361204013376,
+      "grad_norm": 0.7557095289230347,
+      "learning_rate": 4.511679644048943e-05,
+      "loss": 2.5075,
+      "step": 3487
+    },
+    {
+      "epoch": 46.52173913043478,
+      "grad_norm": 0.7621429562568665,
+      "learning_rate": 4.507230255839822e-05,
+      "loss": 2.4163,
+      "step": 3488
+    },
+    {
+      "epoch": 46.53511705685619,
+      "grad_norm": 0.6624872088432312,
+      "learning_rate": 4.502780867630701e-05,
+      "loss": 2.5192,
+      "step": 3489
+    },
+    {
+      "epoch": 46.54849498327759,
+      "grad_norm": 0.7335086464881897,
+      "learning_rate": 4.49833147942158e-05,
+      "loss": 2.4538,
+      "step": 3490
+    },
+    {
+      "epoch": 46.561872909699,
+      "grad_norm": 0.7712231278419495,
+      "learning_rate": 4.493882091212458e-05,
+      "loss": 2.4444,
+      "step": 3491
+    },
+    {
+      "epoch": 46.575250836120404,
+      "grad_norm": 0.6709562540054321,
+      "learning_rate": 4.489432703003337e-05,
+      "loss": 2.3849,
+      "step": 3492
+    },
+    {
+      "epoch": 46.58862876254181,
+      "grad_norm": 0.704768180847168,
+      "learning_rate": 4.484983314794216e-05,
+      "loss": 2.5614,
+      "step": 3493
+    },
+    {
+      "epoch": 46.60200668896321,
+      "grad_norm": 0.6898999214172363,
+      "learning_rate": 4.480533926585095e-05,
+      "loss": 2.4023,
+      "step": 3494
+    },
+    {
+      "epoch": 46.61538461538461,
+      "grad_norm": 0.6936464309692383,
+      "learning_rate": 4.476084538375974e-05,
+      "loss": 2.4075,
+      "step": 3495
+    },
+    {
+      "epoch": 46.62876254180602,
+      "grad_norm": 0.7783868908882141,
+      "learning_rate": 4.471635150166852e-05,
+      "loss": 2.4102,
+      "step": 3496
+    },
+    {
+      "epoch": 46.642140468227424,
+      "grad_norm": 1.0688285827636719,
+      "learning_rate": 4.467185761957731e-05,
+      "loss": 2.2541,
+      "step": 3497
+    },
+    {
+      "epoch": 46.65551839464883,
+      "grad_norm": 0.6693544983863831,
+      "learning_rate": 4.46273637374861e-05,
+      "loss": 2.3223,
+      "step": 3498
+    },
+    {
+      "epoch": 46.668896321070235,
+      "grad_norm": 0.6926013231277466,
+      "learning_rate": 4.458286985539489e-05,
+      "loss": 2.3858,
+      "step": 3499
+    },
+    {
+      "epoch": 46.68227424749164,
+      "grad_norm": 0.6820552945137024,
+      "learning_rate": 4.453837597330368e-05,
+      "loss": 2.6187,
+      "step": 3500
+    },
+    {
+      "epoch": 46.68227424749164,
+      "grad_norm": 1.874079704284668,
+      "learning_rate": 4.449388209121246e-05,
+      "loss": 2.7609,
+      "step": 3501
+    },
+    {
+      "epoch": 46.69565217391305,
+      "grad_norm": 1.8859320878982544,
+      "learning_rate": 4.444938820912124e-05,
+      "loss": 2.6921,
+      "step": 3502
+    },
+    {
+      "epoch": 46.709030100334445,
+      "grad_norm": 2.095224618911743,
+      "learning_rate": 4.440489432703003e-05,
+      "loss": 2.641,
+      "step": 3503
+    },
+    {
+      "epoch": 46.72240802675585,
+      "grad_norm": 1.8568474054336548,
+      "learning_rate": 4.436040044493882e-05,
+      "loss": 2.9348,
+      "step": 3504
+    },
+    {
+      "epoch": 46.735785953177256,
+      "grad_norm": 2.0055668354034424,
+      "learning_rate": 4.431590656284761e-05,
+      "loss": 2.7498,
+      "step": 3505
+    },
+    {
+      "epoch": 46.74916387959866,
+      "grad_norm": 1.6685417890548706,
+      "learning_rate": 4.42714126807564e-05,
+      "loss": 2.8001,
+      "step": 3506
+    },
+    {
+      "epoch": 46.76254180602007,
+      "grad_norm": 1.9633355140686035,
+      "learning_rate": 4.422691879866518e-05,
+      "loss": 2.9822,
+      "step": 3507
+    },
+    {
+      "epoch": 46.77591973244147,
+      "grad_norm": 1.688116192817688,
+      "learning_rate": 4.418242491657397e-05,
+      "loss": 2.8157,
+      "step": 3508
+    },
+    {
+      "epoch": 46.78929765886288,
+      "grad_norm": 1.7179639339447021,
+      "learning_rate": 4.413793103448276e-05,
+      "loss": 2.7049,
+      "step": 3509
+    },
+    {
+      "epoch": 46.802675585284284,
+      "grad_norm": 1.546454906463623,
+      "learning_rate": 4.409343715239155e-05,
+      "loss": 2.6513,
+      "step": 3510
+    },
+    {
+      "epoch": 46.81605351170568,
+      "grad_norm": 1.8116494417190552,
+      "learning_rate": 4.404894327030034e-05,
+      "loss": 2.933,
+      "step": 3511
+    },
+    {
+      "epoch": 46.82943143812709,
+      "grad_norm": 1.8387575149536133,
+      "learning_rate": 4.400444938820912e-05,
+      "loss": 2.864,
+      "step": 3512
+    },
+    {
+      "epoch": 46.84280936454849,
+      "grad_norm": 1.7248547077178955,
+      "learning_rate": 4.395995550611791e-05,
+      "loss": 2.928,
+      "step": 3513
+    },
+    {
+      "epoch": 46.8561872909699,
+      "grad_norm": 1.5455210208892822,
+      "learning_rate": 4.39154616240267e-05,
+      "loss": 2.8224,
+      "step": 3514
+    },
+    {
+      "epoch": 46.869565217391305,
+      "grad_norm": 1.7810604572296143,
+      "learning_rate": 4.387096774193549e-05,
+      "loss": 2.726,
+      "step": 3515
+    },
+    {
+      "epoch": 46.88294314381271,
+      "grad_norm": 1.8111591339111328,
+      "learning_rate": 4.382647385984428e-05,
+      "loss": 2.7562,
+      "step": 3516
+    },
+    {
+      "epoch": 46.896321070234116,
+      "grad_norm": 1.5729373693466187,
+      "learning_rate": 4.378197997775306e-05,
+      "loss": 2.7545,
+      "step": 3517
+    },
+    {
+      "epoch": 46.90969899665552,
+      "grad_norm": 1.7884842157363892,
+      "learning_rate": 4.373748609566185e-05,
+      "loss": 2.8122,
+      "step": 3518
+    },
+    {
+      "epoch": 46.92307692307692,
+      "grad_norm": 1.8262654542922974,
+      "learning_rate": 4.369299221357064e-05,
+      "loss": 2.814,
+      "step": 3519
+    },
+    {
+      "epoch": 46.936454849498325,
+      "grad_norm": 1.5744242668151855,
+      "learning_rate": 4.364849833147943e-05,
+      "loss": 2.7071,
+      "step": 3520
+    },
+    {
+      "epoch": 46.94983277591973,
+      "grad_norm": 1.506430983543396,
+      "learning_rate": 4.3604004449388216e-05,
+      "loss": 2.8561,
+      "step": 3521
+    },
+    {
+      "epoch": 46.96321070234114,
+      "grad_norm": 1.4957503080368042,
+      "learning_rate": 4.3559510567297e-05,
+      "loss": 2.3908,
+      "step": 3522
+    },
+    {
+      "epoch": 46.97658862876254,
+      "grad_norm": 1.5997592210769653,
+      "learning_rate": 4.351501668520578e-05,
+      "loss": 2.6824,
+      "step": 3523
+    },
+    {
+      "epoch": 46.98996655518395,
+      "grad_norm": 1.5681554079055786,
+      "learning_rate": 4.347052280311457e-05,
+      "loss": 2.6847,
+      "step": 3524
+    },
+    {
+      "epoch": 47.013377926421406,
+      "grad_norm": 2.1808457374572754,
+      "learning_rate": 4.342602892102336e-05,
+      "loss": 5.1443,
+      "step": 3525
+    },
+    {
+      "epoch": 47.02675585284281,
+      "grad_norm": 1.6048583984375,
+      "learning_rate": 4.338153503893215e-05,
+      "loss": 2.8358,
+      "step": 3526
+    },
+    {
+      "epoch": 47.04013377926422,
+      "grad_norm": 1.60443115234375,
+      "learning_rate": 4.333704115684094e-05,
+      "loss": 2.8681,
+      "step": 3527
+    },
+    {
+      "epoch": 47.05351170568562,
+      "grad_norm": 1.571325421333313,
+      "learning_rate": 4.329254727474972e-05,
+      "loss": 2.7476,
+      "step": 3528
+    },
+    {
+      "epoch": 47.06688963210702,
+      "grad_norm": 1.3896905183792114,
+      "learning_rate": 4.324805339265851e-05,
+      "loss": 2.9488,
+      "step": 3529
+    },
+    {
+      "epoch": 47.080267558528426,
+      "grad_norm": 1.6123690605163574,
+      "learning_rate": 4.32035595105673e-05,
+      "loss": 2.6438,
+      "step": 3530
+    },
+    {
+      "epoch": 47.09364548494983,
+      "grad_norm": 1.3029465675354004,
+      "learning_rate": 4.315906562847609e-05,
+      "loss": 2.4713,
+      "step": 3531
+    },
+    {
+      "epoch": 47.10702341137124,
+      "grad_norm": 1.6188567876815796,
+      "learning_rate": 4.311457174638488e-05,
+      "loss": 3.0083,
+      "step": 3532
+    },
+    {
+      "epoch": 47.12040133779264,
+      "grad_norm": 1.6200324296951294,
+      "learning_rate": 4.307007786429366e-05,
+      "loss": 2.7221,
+      "step": 3533
+    },
+    {
+      "epoch": 47.13377926421405,
+      "grad_norm": 1.5296354293823242,
+      "learning_rate": 4.302558398220245e-05,
+      "loss": 2.9623,
+      "step": 3534
+    },
+    {
+      "epoch": 47.147157190635454,
+      "grad_norm": 1.4072962999343872,
+      "learning_rate": 4.298109010011124e-05,
+      "loss": 2.7184,
+      "step": 3535
+    },
+    {
+      "epoch": 47.16053511705686,
+      "grad_norm": 1.2409498691558838,
+      "learning_rate": 4.293659621802003e-05,
+      "loss": 2.842,
+      "step": 3536
+    },
+    {
+      "epoch": 47.17391304347826,
+      "grad_norm": 1.6017128229141235,
+      "learning_rate": 4.2892102335928816e-05,
+      "loss": 2.7034,
+      "step": 3537
+    },
+    {
+      "epoch": 47.187290969899664,
+      "grad_norm": 1.248794436454773,
+      "learning_rate": 4.28476084538376e-05,
+      "loss": 2.4749,
+      "step": 3538
+    },
+    {
+      "epoch": 47.20066889632107,
+      "grad_norm": 1.3597090244293213,
+      "learning_rate": 4.280311457174639e-05,
+      "loss": 2.7913,
+      "step": 3539
+    },
+    {
+      "epoch": 47.214046822742475,
+      "grad_norm": 1.3427294492721558,
+      "learning_rate": 4.275862068965518e-05,
+      "loss": 2.6795,
+      "step": 3540
+    },
+    {
+      "epoch": 47.22742474916388,
+      "grad_norm": 1.4270846843719482,
+      "learning_rate": 4.2714126807563966e-05,
+      "loss": 2.5893,
+      "step": 3541
+    },
+    {
+      "epoch": 47.240802675585286,
+      "grad_norm": 1.5535361766815186,
+      "learning_rate": 4.2669632925472755e-05,
+      "loss": 2.5322,
+      "step": 3542
+    },
+    {
+      "epoch": 47.25418060200669,
+      "grad_norm": 1.269716739654541,
+      "learning_rate": 4.262513904338154e-05,
+      "loss": 2.7256,
+      "step": 3543
+    },
+    {
+      "epoch": 47.2675585284281,
+      "grad_norm": 1.465528130531311,
+      "learning_rate": 4.258064516129032e-05,
+      "loss": 2.7631,
+      "step": 3544
+    },
+    {
+      "epoch": 47.280936454849495,
+      "grad_norm": 1.4620834589004517,
+      "learning_rate": 4.253615127919911e-05,
+      "loss": 2.7169,
+      "step": 3545
+    },
+    {
+      "epoch": 47.2943143812709,
+      "grad_norm": 1.4662063121795654,
+      "learning_rate": 4.24916573971079e-05,
+      "loss": 2.8878,
+      "step": 3546
+    },
+    {
+      "epoch": 47.30769230769231,
+      "grad_norm": 1.6685733795166016,
+      "learning_rate": 4.244716351501669e-05,
+      "loss": 2.7765,
+      "step": 3547
+    },
+    {
+      "epoch": 47.32107023411371,
+      "grad_norm": 1.516026258468628,
+      "learning_rate": 4.2402669632925476e-05,
+      "loss": 2.8278,
+      "step": 3548
+    },
+    {
+      "epoch": 47.33444816053512,
+      "grad_norm": 1.5270005464553833,
+      "learning_rate": 4.235817575083426e-05,
+      "loss": 2.7561,
+      "step": 3549
+    },
+    {
+      "epoch": 47.34782608695652,
+      "grad_norm": 1.5817084312438965,
+      "learning_rate": 4.231368186874305e-05,
+      "loss": 2.8527,
+      "step": 3550
+    },
+    {
+      "epoch": 47.36120401337793,
+      "grad_norm": 1.3843607902526855,
+      "learning_rate": 4.226918798665184e-05,
+      "loss": 2.819,
+      "step": 3551
+    },
+    {
+      "epoch": 47.374581939799334,
+      "grad_norm": 1.4046587944030762,
+      "learning_rate": 4.2224694104560626e-05,
+      "loss": 2.5891,
+      "step": 3552
+    },
+    {
+      "epoch": 47.38795986622073,
+      "grad_norm": 1.660375952720642,
+      "learning_rate": 4.2180200222469415e-05,
+      "loss": 2.8561,
+      "step": 3553
+    },
+    {
+      "epoch": 47.40133779264214,
+      "grad_norm": 1.3785984516143799,
+      "learning_rate": 4.21357063403782e-05,
+      "loss": 2.6979,
+      "step": 3554
+    },
+    {
+      "epoch": 47.414715719063544,
+      "grad_norm": 1.667795181274414,
+      "learning_rate": 4.209121245828699e-05,
+      "loss": 2.6532,
+      "step": 3555
+    },
+    {
+      "epoch": 47.42809364548495,
+      "grad_norm": 1.4713093042373657,
+      "learning_rate": 4.2046718576195776e-05,
+      "loss": 2.8053,
+      "step": 3556
+    },
+    {
+      "epoch": 47.441471571906355,
+      "grad_norm": 1.7168391942977905,
+      "learning_rate": 4.2002224694104565e-05,
+      "loss": 2.6546,
+      "step": 3557
+    },
+    {
+      "epoch": 47.45484949832776,
+      "grad_norm": 1.476688265800476,
+      "learning_rate": 4.1957730812013354e-05,
+      "loss": 2.604,
+      "step": 3558
+    },
+    {
+      "epoch": 47.468227424749166,
+      "grad_norm": 1.4487282037734985,
+      "learning_rate": 4.191323692992214e-05,
+      "loss": 2.5379,
+      "step": 3559
+    },
+    {
+      "epoch": 47.48160535117057,
+      "grad_norm": 1.1151924133300781,
+      "learning_rate": 4.1868743047830926e-05,
+      "loss": 2.5306,
+      "step": 3560
+    },
+    {
+      "epoch": 47.49498327759197,
+      "grad_norm": 1.3186345100402832,
+      "learning_rate": 4.1824249165739715e-05,
+      "loss": 2.6516,
+      "step": 3561
+    },
+    {
+      "epoch": 47.508361204013376,
+      "grad_norm": 1.4641391038894653,
+      "learning_rate": 4.1779755283648504e-05,
+      "loss": 2.8421,
+      "step": 3562
+    },
+    {
+      "epoch": 47.52173913043478,
+      "grad_norm": 1.3929246664047241,
+      "learning_rate": 4.173526140155729e-05,
+      "loss": 2.6351,
+      "step": 3563
+    },
+    {
+      "epoch": 47.53511705685619,
+      "grad_norm": 1.8099298477172852,
+      "learning_rate": 4.1690767519466076e-05,
+      "loss": 2.6873,
+      "step": 3564
+    },
+    {
+      "epoch": 47.54849498327759,
+      "grad_norm": 1.5447300672531128,
+      "learning_rate": 4.164627363737486e-05,
+      "loss": 2.8718,
+      "step": 3565
+    },
+    {
+      "epoch": 47.561872909699,
+      "grad_norm": 1.5041502714157104,
+      "learning_rate": 4.160177975528365e-05,
+      "loss": 2.6741,
+      "step": 3566
+    },
+    {
+      "epoch": 47.575250836120404,
+      "grad_norm": 1.713394284248352,
+      "learning_rate": 4.155728587319244e-05,
+      "loss": 2.9518,
+      "step": 3567
+    },
+    {
+      "epoch": 47.58862876254181,
+      "grad_norm": 1.3753432035446167,
+      "learning_rate": 4.1512791991101226e-05,
+      "loss": 2.9874,
+      "step": 3568
+    },
+    {
+      "epoch": 47.60200668896321,
+      "grad_norm": 1.901676893234253,
+      "learning_rate": 4.1468298109010015e-05,
+      "loss": 2.7524,
+      "step": 3569
+    },
+    {
+      "epoch": 47.61538461538461,
+      "grad_norm": 1.8974645137786865,
+      "learning_rate": 4.14238042269188e-05,
+      "loss": 2.744,
+      "step": 3570
+    },
+    {
+      "epoch": 47.62876254180602,
+      "grad_norm": 1.278735637664795,
+      "learning_rate": 4.1379310344827587e-05,
+      "loss": 2.6875,
+      "step": 3571
+    },
+    {
+      "epoch": 47.642140468227424,
+      "grad_norm": 1.7426670789718628,
+      "learning_rate": 4.1334816462736376e-05,
+      "loss": 2.7088,
+      "step": 3572
+    },
+    {
+      "epoch": 47.65551839464883,
+      "grad_norm": 1.2926254272460938,
+      "learning_rate": 4.1290322580645165e-05,
+      "loss": 2.536,
+      "step": 3573
+    },
+    {
+      "epoch": 47.668896321070235,
+      "grad_norm": 1.4211947917938232,
+      "learning_rate": 4.1245828698553954e-05,
+      "loss": 2.7895,
+      "step": 3574
+    },
+    {
+      "epoch": 47.68227424749164,
+      "grad_norm": 1.4811222553253174,
+      "learning_rate": 4.1201334816462736e-05,
+      "loss": 2.8819,
+      "step": 3575
+    },
+    {
+      "epoch": 47.69565217391305,
+      "grad_norm": 1.441867470741272,
+      "learning_rate": 4.1156840934371526e-05,
+      "loss": 2.9332,
+      "step": 3576
+    },
+    {
+      "epoch": 47.709030100334445,
+      "grad_norm": 1.3309211730957031,
+      "learning_rate": 4.1112347052280315e-05,
+      "loss": 2.5375,
+      "step": 3577
+    },
+    {
+      "epoch": 47.72240802675585,
+      "grad_norm": 1.1746070384979248,
+      "learning_rate": 4.1067853170189104e-05,
+      "loss": 2.8976,
+      "step": 3578
+    },
+    {
+      "epoch": 47.735785953177256,
+      "grad_norm": 1.572267770767212,
+      "learning_rate": 4.102335928809789e-05,
+      "loss": 2.7175,
+      "step": 3579
+    },
+    {
+      "epoch": 47.74916387959866,
+      "grad_norm": 1.6516088247299194,
+      "learning_rate": 4.0978865406006675e-05,
+      "loss": 2.7369,
+      "step": 3580
+    },
+    {
+      "epoch": 47.76254180602007,
+      "grad_norm": 1.2800843715667725,
+      "learning_rate": 4.0934371523915465e-05,
+      "loss": 2.6551,
+      "step": 3581
+    },
+    {
+      "epoch": 47.77591973244147,
+      "grad_norm": 1.115822434425354,
+      "learning_rate": 4.0889877641824254e-05,
+      "loss": 2.8808,
+      "step": 3582
+    },
+    {
+      "epoch": 47.78929765886288,
+      "grad_norm": 1.5338307619094849,
+      "learning_rate": 4.084538375973304e-05,
+      "loss": 2.6126,
+      "step": 3583
+    },
+    {
+      "epoch": 47.802675585284284,
+      "grad_norm": 1.226580262184143,
+      "learning_rate": 4.0800889877641825e-05,
+      "loss": 2.6029,
+      "step": 3584
+    },
+    {
+      "epoch": 47.81605351170568,
+      "grad_norm": 1.2477607727050781,
+      "learning_rate": 4.0756395995550615e-05,
+      "loss": 2.7603,
+      "step": 3585
+    },
+    {
+      "epoch": 47.82943143812709,
+      "grad_norm": 1.5089812278747559,
+      "learning_rate": 4.07119021134594e-05,
+      "loss": 2.9377,
+      "step": 3586
+    },
+    {
+      "epoch": 47.84280936454849,
+      "grad_norm": 1.3173625469207764,
+      "learning_rate": 4.0667408231368186e-05,
+      "loss": 2.7398,
+      "step": 3587
+    },
+    {
+      "epoch": 47.8561872909699,
+      "grad_norm": 1.217244267463684,
+      "learning_rate": 4.0622914349276975e-05,
+      "loss": 2.6319,
+      "step": 3588
+    },
+    {
+      "epoch": 47.869565217391305,
+      "grad_norm": 1.5038195848464966,
+      "learning_rate": 4.0578420467185764e-05,
+      "loss": 2.6373,
+      "step": 3589
+    },
+    {
+      "epoch": 47.88294314381271,
+      "grad_norm": 1.4669184684753418,
+      "learning_rate": 4.0533926585094554e-05,
+      "loss": 2.7596,
+      "step": 3590
+    },
+    {
+      "epoch": 47.896321070234116,
+      "grad_norm": 1.1874650716781616,
+      "learning_rate": 4.0489432703003336e-05,
+      "loss": 2.8438,
+      "step": 3591
+    },
+    {
+      "epoch": 47.90969899665552,
+      "grad_norm": 1.3088831901550293,
+      "learning_rate": 4.0444938820912125e-05,
+      "loss": 2.6426,
+      "step": 3592
+    },
+    {
+      "epoch": 47.92307692307692,
+      "grad_norm": 1.2181847095489502,
+      "learning_rate": 4.0400444938820914e-05,
+      "loss": 2.577,
+      "step": 3593
+    },
+    {
+      "epoch": 47.936454849498325,
+      "grad_norm": 1.4106916189193726,
+      "learning_rate": 4.0355951056729703e-05,
+      "loss": 2.8785,
+      "step": 3594
+    },
+    {
+      "epoch": 47.94983277591973,
+      "grad_norm": 1.3725144863128662,
+      "learning_rate": 4.031145717463849e-05,
+      "loss": 2.6663,
+      "step": 3595
+    },
+    {
+      "epoch": 47.96321070234114,
+      "grad_norm": 1.6895320415496826,
+      "learning_rate": 4.0266963292547275e-05,
+      "loss": 2.8136,
+      "step": 3596
+    },
+    {
+      "epoch": 47.97658862876254,
+      "grad_norm": 1.2098865509033203,
+      "learning_rate": 4.0222469410456064e-05,
+      "loss": 2.6721,
+      "step": 3597
+    },
+    {
+      "epoch": 47.98996655518395,
+      "grad_norm": 1.6445807218551636,
+      "learning_rate": 4.017797552836485e-05,
+      "loss": 2.8354,
+      "step": 3598
+    },
+    {
+      "epoch": 48.0,
+      "grad_norm": 1.6511502265930176,
+      "learning_rate": 4.013348164627364e-05,
+      "loss": 2.7395,
+      "step": 3599
+    },
+    {
+      "epoch": 48.013377926421406,
+      "grad_norm": 0.930826723575592,
+      "learning_rate": 4.008898776418243e-05,
+      "loss": 2.6881,
+      "step": 3600
+    },
+    {
+      "epoch": 48.02675585284281,
+      "grad_norm": 1.0226924419403076,
+      "learning_rate": 4.0044493882091214e-05,
+      "loss": 2.4951,
+      "step": 3601
+    },
+    {
+      "epoch": 48.04013377926422,
+      "grad_norm": 0.8309192061424255,
+      "learning_rate": 4e-05,
+      "loss": 2.7274,
+      "step": 3602
+    },
+    {
+      "epoch": 48.05351170568562,
+      "grad_norm": 0.8343853950500488,
+      "learning_rate": 3.995550611790879e-05,
+      "loss": 2.7413,
+      "step": 3603
+    },
+    {
+      "epoch": 48.06688963210702,
+      "grad_norm": 0.8467966914176941,
+      "learning_rate": 3.991101223581758e-05,
+      "loss": 2.4937,
+      "step": 3604
+    },
+    {
+      "epoch": 48.080267558528426,
+      "grad_norm": 0.8166081309318542,
+      "learning_rate": 3.9866518353726364e-05,
+      "loss": 2.6457,
+      "step": 3605
+    },
+    {
+      "epoch": 48.09364548494983,
+      "grad_norm": 0.9212111234664917,
+      "learning_rate": 3.982202447163515e-05,
+      "loss": 2.6507,
+      "step": 3606
+    },
+    {
+      "epoch": 48.10702341137124,
+      "grad_norm": 0.8982736468315125,
+      "learning_rate": 3.9777530589543936e-05,
+      "loss": 2.6162,
+      "step": 3607
+    },
+    {
+      "epoch": 48.12040133779264,
+      "grad_norm": 1.1569843292236328,
+      "learning_rate": 3.9733036707452725e-05,
+      "loss": 2.8513,
+      "step": 3608
+    },
+    {
+      "epoch": 48.13377926421405,
+      "grad_norm": 1.0270570516586304,
+      "learning_rate": 3.9688542825361514e-05,
+      "loss": 2.8707,
+      "step": 3609
+    },
+    {
+      "epoch": 48.147157190635454,
+      "grad_norm": 0.9687206745147705,
+      "learning_rate": 3.96440489432703e-05,
+      "loss": 2.7854,
+      "step": 3610
+    },
+    {
+      "epoch": 48.16053511705686,
+      "grad_norm": 0.9363980293273926,
+      "learning_rate": 3.959955506117909e-05,
+      "loss": 2.7802,
+      "step": 3611
+    },
+    {
+      "epoch": 48.17391304347826,
+      "grad_norm": 0.8202338218688965,
+      "learning_rate": 3.9555061179087875e-05,
+      "loss": 2.7608,
+      "step": 3612
+    },
+    {
+      "epoch": 48.187290969899664,
+      "grad_norm": 0.9861024618148804,
+      "learning_rate": 3.9510567296996664e-05,
+      "loss": 2.5243,
+      "step": 3613
+    },
+    {
+      "epoch": 48.20066889632107,
+      "grad_norm": 0.8582180738449097,
+      "learning_rate": 3.946607341490545e-05,
+      "loss": 2.7575,
+      "step": 3614
+    },
+    {
+      "epoch": 48.214046822742475,
+      "grad_norm": 0.8339632749557495,
+      "learning_rate": 3.942157953281424e-05,
+      "loss": 2.6449,
+      "step": 3615
+    },
+    {
+      "epoch": 48.22742474916388,
+      "grad_norm": 0.9281173944473267,
+      "learning_rate": 3.937708565072303e-05,
+      "loss": 2.6076,
+      "step": 3616
+    },
+    {
+      "epoch": 48.240802675585286,
+      "grad_norm": 1.1367732286453247,
+      "learning_rate": 3.9332591768631814e-05,
+      "loss": 2.3871,
+      "step": 3617
+    },
+    {
+      "epoch": 48.25418060200669,
+      "grad_norm": 0.8201642632484436,
+      "learning_rate": 3.92880978865406e-05,
+      "loss": 2.5026,
+      "step": 3618
+    },
+    {
+      "epoch": 48.2675585284281,
+      "grad_norm": 0.9126654267311096,
+      "learning_rate": 3.924360400444939e-05,
+      "loss": 2.8502,
+      "step": 3619
+    },
+    {
+      "epoch": 48.280936454849495,
+      "grad_norm": 1.1465530395507812,
+      "learning_rate": 3.919911012235818e-05,
+      "loss": 2.8908,
+      "step": 3620
+    },
+    {
+      "epoch": 48.2943143812709,
+      "grad_norm": 0.8558413982391357,
+      "learning_rate": 3.915461624026697e-05,
+      "loss": 2.7329,
+      "step": 3621
+    },
+    {
+      "epoch": 48.30769230769231,
+      "grad_norm": 0.8293155431747437,
+      "learning_rate": 3.911012235817575e-05,
+      "loss": 2.7223,
+      "step": 3622
+    },
+    {
+      "epoch": 48.32107023411371,
+      "grad_norm": 0.9223231077194214,
+      "learning_rate": 3.906562847608454e-05,
+      "loss": 2.7142,
+      "step": 3623
+    },
+    {
+      "epoch": 48.33444816053512,
+      "grad_norm": 0.8491235971450806,
+      "learning_rate": 3.902113459399333e-05,
+      "loss": 2.6362,
+      "step": 3624
+    },
+    {
+      "epoch": 48.34782608695652,
+      "grad_norm": 0.8782801032066345,
+      "learning_rate": 3.8976640711902113e-05,
+      "loss": 2.7339,
+      "step": 3625
+    },
+    {
+      "epoch": 48.36120401337793,
+      "grad_norm": 0.8530482053756714,
+      "learning_rate": 3.89321468298109e-05,
+      "loss": 2.8429,
+      "step": 3626
+    },
+    {
+      "epoch": 48.374581939799334,
+      "grad_norm": 1.0677621364593506,
+      "learning_rate": 3.8887652947719685e-05,
+      "loss": 2.6479,
+      "step": 3627
+    },
+    {
+      "epoch": 48.38795986622073,
+      "grad_norm": 0.9075163006782532,
+      "learning_rate": 3.8843159065628474e-05,
+      "loss": 2.4386,
+      "step": 3628
+    },
+    {
+      "epoch": 48.40133779264214,
+      "grad_norm": 0.8299190402030945,
+      "learning_rate": 3.879866518353726e-05,
+      "loss": 2.5828,
+      "step": 3629
+    },
+    {
+      "epoch": 48.414715719063544,
+      "grad_norm": 0.877830982208252,
+      "learning_rate": 3.875417130144605e-05,
+      "loss": 2.7312,
+      "step": 3630
+    },
+    {
+      "epoch": 48.42809364548495,
+      "grad_norm": 1.0116523504257202,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 2.6551,
+      "step": 3631
+    },
+    {
+      "epoch": 48.441471571906355,
+      "grad_norm": 0.9170435667037964,
+      "learning_rate": 3.8665183537263624e-05,
+      "loss": 2.6219,
+      "step": 3632
+    },
+    {
+      "epoch": 48.45484949832776,
+      "grad_norm": 1.041170358657837,
+      "learning_rate": 3.862068965517241e-05,
+      "loss": 2.6453,
+      "step": 3633
+    },
+    {
+      "epoch": 48.468227424749166,
+      "grad_norm": 1.0471312999725342,
+      "learning_rate": 3.85761957730812e-05,
+      "loss": 2.7593,
+      "step": 3634
+    },
+    {
+      "epoch": 48.48160535117057,
+      "grad_norm": 0.962745726108551,
+      "learning_rate": 3.853170189098999e-05,
+      "loss": 2.7035,
+      "step": 3635
+    },
+    {
+      "epoch": 48.49498327759197,
+      "grad_norm": 1.091069221496582,
+      "learning_rate": 3.848720800889878e-05,
+      "loss": 2.55,
+      "step": 3636
+    },
+    {
+      "epoch": 48.508361204013376,
+      "grad_norm": 0.932161808013916,
+      "learning_rate": 3.844271412680756e-05,
+      "loss": 2.6504,
+      "step": 3637
+    },
+    {
+      "epoch": 48.52173913043478,
+      "grad_norm": 0.9401114583015442,
+      "learning_rate": 3.839822024471635e-05,
+      "loss": 2.6304,
+      "step": 3638
+    },
+    {
+      "epoch": 48.53511705685619,
+      "grad_norm": 0.8302827477455139,
+      "learning_rate": 3.835372636262514e-05,
+      "loss": 2.6115,
+      "step": 3639
+    },
+    {
+      "epoch": 48.54849498327759,
+      "grad_norm": 0.9349223971366882,
+      "learning_rate": 3.830923248053393e-05,
+      "loss": 2.4827,
+      "step": 3640
+    },
+    {
+      "epoch": 48.561872909699,
+      "grad_norm": 1.053091287612915,
+      "learning_rate": 3.826473859844272e-05,
+      "loss": 2.515,
+      "step": 3641
+    },
+    {
+      "epoch": 48.575250836120404,
+      "grad_norm": 0.9354698657989502,
+      "learning_rate": 3.82202447163515e-05,
+      "loss": 2.8267,
+      "step": 3642
+    },
+    {
+      "epoch": 48.58862876254181,
+      "grad_norm": 0.9621760845184326,
+      "learning_rate": 3.817575083426029e-05,
+      "loss": 2.7924,
+      "step": 3643
+    },
+    {
+      "epoch": 48.60200668896321,
+      "grad_norm": 1.0033005475997925,
+      "learning_rate": 3.813125695216908e-05,
+      "loss": 2.7694,
+      "step": 3644
+    },
+    {
+      "epoch": 48.61538461538461,
+      "grad_norm": 0.9117392301559448,
+      "learning_rate": 3.808676307007787e-05,
+      "loss": 2.8041,
+      "step": 3645
+    },
+    {
+      "epoch": 48.62876254180602,
+      "grad_norm": 1.1089485883712769,
+      "learning_rate": 3.804226918798665e-05,
+      "loss": 2.8602,
+      "step": 3646
+    },
+    {
+      "epoch": 48.642140468227424,
+      "grad_norm": 1.098798155784607,
+      "learning_rate": 3.799777530589544e-05,
+      "loss": 2.6619,
+      "step": 3647
+    },
+    {
+      "epoch": 48.65551839464883,
+      "grad_norm": 0.9343369603157043,
+      "learning_rate": 3.7953281423804224e-05,
+      "loss": 2.4415,
+      "step": 3648
+    },
+    {
+      "epoch": 48.668896321070235,
+      "grad_norm": 1.1379038095474243,
+      "learning_rate": 3.790878754171301e-05,
+      "loss": 2.6495,
+      "step": 3649
+    },
+    {
+      "epoch": 48.68227424749164,
+      "grad_norm": 0.8679940700531006,
+      "learning_rate": 3.78642936596218e-05,
+      "loss": 2.6208,
+      "step": 3650
+    },
+    {
+      "epoch": 48.69565217391305,
+      "grad_norm": 0.8417056202888489,
+      "learning_rate": 3.781979977753059e-05,
+      "loss": 2.7265,
+      "step": 3651
+    },
+    {
+      "epoch": 48.709030100334445,
+      "grad_norm": 1.0414758920669556,
+      "learning_rate": 3.777530589543938e-05,
+      "loss": 2.876,
+      "step": 3652
+    },
+    {
+      "epoch": 48.72240802675585,
+      "grad_norm": 0.9891830682754517,
+      "learning_rate": 3.773081201334816e-05,
+      "loss": 2.5307,
+      "step": 3653
+    },
+    {
+      "epoch": 48.735785953177256,
+      "grad_norm": 0.8791108727455139,
+      "learning_rate": 3.768631813125695e-05,
+      "loss": 2.6067,
+      "step": 3654
+    },
+    {
+      "epoch": 48.74916387959866,
+      "grad_norm": 1.0120600461959839,
+      "learning_rate": 3.764182424916574e-05,
+      "loss": 2.6719,
+      "step": 3655
+    },
+    {
+      "epoch": 48.76254180602007,
+      "grad_norm": 0.9853880405426025,
+      "learning_rate": 3.759733036707453e-05,
+      "loss": 2.6551,
+      "step": 3656
+    },
+    {
+      "epoch": 48.77591973244147,
+      "grad_norm": 0.8162937164306641,
+      "learning_rate": 3.755283648498332e-05,
+      "loss": 2.7002,
+      "step": 3657
+    },
+    {
+      "epoch": 48.78929765886288,
+      "grad_norm": 0.8841816782951355,
+      "learning_rate": 3.75083426028921e-05,
+      "loss": 2.6199,
+      "step": 3658
+    },
+    {
+      "epoch": 48.802675585284284,
+      "grad_norm": 0.8989855051040649,
+      "learning_rate": 3.746384872080089e-05,
+      "loss": 2.8484,
+      "step": 3659
+    },
+    {
+      "epoch": 48.81605351170568,
+      "grad_norm": 0.9171043634414673,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 2.7621,
+      "step": 3660
+    },
+    {
+      "epoch": 48.82943143812709,
+      "grad_norm": 0.9595106840133667,
+      "learning_rate": 3.737486095661847e-05,
+      "loss": 2.8166,
+      "step": 3661
+    },
+    {
+      "epoch": 48.84280936454849,
+      "grad_norm": 1.1387951374053955,
+      "learning_rate": 3.733036707452726e-05,
+      "loss": 2.6994,
+      "step": 3662
+    },
+    {
+      "epoch": 48.8561872909699,
+      "grad_norm": 0.9676811695098877,
+      "learning_rate": 3.728587319243604e-05,
+      "loss": 2.7282,
+      "step": 3663
+    },
+    {
+      "epoch": 48.869565217391305,
+      "grad_norm": 1.1474648714065552,
+      "learning_rate": 3.724137931034483e-05,
+      "loss": 2.6101,
+      "step": 3664
+    },
+    {
+      "epoch": 48.88294314381271,
+      "grad_norm": 0.8521884679794312,
+      "learning_rate": 3.719688542825362e-05,
+      "loss": 2.3969,
+      "step": 3665
+    },
+    {
+      "epoch": 48.896321070234116,
+      "grad_norm": 0.8633421063423157,
+      "learning_rate": 3.71523915461624e-05,
+      "loss": 2.7976,
+      "step": 3666
+    },
+    {
+      "epoch": 48.90969899665552,
+      "grad_norm": 0.9763745665550232,
+      "learning_rate": 3.710789766407119e-05,
+      "loss": 2.6065,
+      "step": 3667
+    },
+    {
+      "epoch": 48.92307692307692,
+      "grad_norm": 0.7422575354576111,
+      "learning_rate": 3.706340378197998e-05,
+      "loss": 2.6796,
+      "step": 3668
+    },
+    {
+      "epoch": 48.936454849498325,
+      "grad_norm": 0.933746874332428,
+      "learning_rate": 3.701890989988876e-05,
+      "loss": 2.6876,
+      "step": 3669
+    },
+    {
+      "epoch": 48.94983277591973,
+      "grad_norm": 0.8498068451881409,
+      "learning_rate": 3.697441601779755e-05,
+      "loss": 2.7027,
+      "step": 3670
+    },
+    {
+      "epoch": 48.96321070234114,
+      "grad_norm": 0.9815676212310791,
+      "learning_rate": 3.692992213570634e-05,
+      "loss": 2.5987,
+      "step": 3671
+    },
+    {
+      "epoch": 48.97658862876254,
+      "grad_norm": 0.9350228309631348,
+      "learning_rate": 3.688542825361513e-05,
+      "loss": 2.6082,
+      "step": 3672
+    },
+    {
+      "epoch": 48.98996655518395,
+      "grad_norm": 0.8007590770721436,
+      "learning_rate": 3.684093437152392e-05,
+      "loss": 2.3251,
+      "step": 3673
+    },
+    {
+      "epoch": 49.0,
+      "grad_norm": 1.2886775732040405,
+      "learning_rate": 3.67964404894327e-05,
+      "loss": 2.5578,
+      "step": 3674
+    },
+    {
+      "epoch": 49.013377926421406,
+      "grad_norm": 0.8050103187561035,
+      "learning_rate": 3.675194660734149e-05,
+      "loss": 2.6582,
+      "step": 3675
+    },
+    {
+      "epoch": 49.02675585284281,
+      "grad_norm": 0.7852005958557129,
+      "learning_rate": 3.670745272525028e-05,
+      "loss": 2.7088,
+      "step": 3676
+    },
+    {
+      "epoch": 49.04013377926422,
+      "grad_norm": 0.8211432695388794,
+      "learning_rate": 3.666295884315907e-05,
+      "loss": 2.3469,
+      "step": 3677
+    },
+    {
+      "epoch": 49.05351170568562,
+      "grad_norm": 0.8142633438110352,
+      "learning_rate": 3.661846496106786e-05,
+      "loss": 2.5533,
+      "step": 3678
+    },
+    {
+      "epoch": 49.06688963210702,
+      "grad_norm": 0.7460059523582458,
+      "learning_rate": 3.657397107897664e-05,
+      "loss": 2.7169,
+      "step": 3679
+    },
+    {
+      "epoch": 49.080267558528426,
+      "grad_norm": 0.8187868595123291,
+      "learning_rate": 3.652947719688543e-05,
+      "loss": 2.6557,
+      "step": 3680
+    },
+    {
+      "epoch": 49.09364548494983,
+      "grad_norm": 0.6976158022880554,
+      "learning_rate": 3.648498331479422e-05,
+      "loss": 2.6892,
+      "step": 3681
+    },
+    {
+      "epoch": 49.10702341137124,
+      "grad_norm": 0.8464581370353699,
+      "learning_rate": 3.644048943270301e-05,
+      "loss": 2.6912,
+      "step": 3682
+    },
+    {
+      "epoch": 49.12040133779264,
+      "grad_norm": 0.7306657433509827,
+      "learning_rate": 3.63959955506118e-05,
+      "loss": 2.4801,
+      "step": 3683
+    },
+    {
+      "epoch": 49.13377926421405,
+      "grad_norm": 0.7567654252052307,
+      "learning_rate": 3.635150166852058e-05,
+      "loss": 2.6735,
+      "step": 3684
+    },
+    {
+      "epoch": 49.147157190635454,
+      "grad_norm": 0.804836630821228,
+      "learning_rate": 3.630700778642937e-05,
+      "loss": 2.5206,
+      "step": 3685
+    },
+    {
+      "epoch": 49.16053511705686,
+      "grad_norm": 0.7669990658760071,
+      "learning_rate": 3.626251390433816e-05,
+      "loss": 2.6387,
+      "step": 3686
+    },
+    {
+      "epoch": 49.17391304347826,
+      "grad_norm": 0.7316834926605225,
+      "learning_rate": 3.621802002224694e-05,
+      "loss": 2.5791,
+      "step": 3687
+    },
+    {
+      "epoch": 49.187290969899664,
+      "grad_norm": 0.6893475651741028,
+      "learning_rate": 3.617352614015573e-05,
+      "loss": 2.5242,
+      "step": 3688
+    },
+    {
+      "epoch": 49.20066889632107,
+      "grad_norm": 0.7602696418762207,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 2.6918,
+      "step": 3689
+    },
+    {
+      "epoch": 49.214046822742475,
+      "grad_norm": 0.7876319289207458,
+      "learning_rate": 3.60845383759733e-05,
+      "loss": 2.5738,
+      "step": 3690
+    },
+    {
+      "epoch": 49.22742474916388,
+      "grad_norm": 0.9337881207466125,
+      "learning_rate": 3.604004449388209e-05,
+      "loss": 2.6884,
+      "step": 3691
+    },
+    {
+      "epoch": 49.240802675585286,
+      "grad_norm": 0.7203758955001831,
+      "learning_rate": 3.599555061179088e-05,
+      "loss": 2.4134,
+      "step": 3692
+    },
+    {
+      "epoch": 49.25418060200669,
+      "grad_norm": 0.8985403776168823,
+      "learning_rate": 3.595105672969967e-05,
+      "loss": 2.4833,
+      "step": 3693
+    },
+    {
+      "epoch": 49.2675585284281,
+      "grad_norm": 0.7891301512718201,
+      "learning_rate": 3.590656284760846e-05,
+      "loss": 2.8944,
+      "step": 3694
+    },
+    {
+      "epoch": 49.280936454849495,
+      "grad_norm": 0.7096188068389893,
+      "learning_rate": 3.586206896551724e-05,
+      "loss": 2.4957,
+      "step": 3695
+    },
+    {
+      "epoch": 49.2943143812709,
+      "grad_norm": 0.7646763920783997,
+      "learning_rate": 3.581757508342603e-05,
+      "loss": 2.2093,
+      "step": 3696
+    },
+    {
+      "epoch": 49.30769230769231,
+      "grad_norm": 0.8256480693817139,
+      "learning_rate": 3.577308120133482e-05,
+      "loss": 2.6597,
+      "step": 3697
+    },
+    {
+      "epoch": 49.32107023411371,
+      "grad_norm": 0.7693411707878113,
+      "learning_rate": 3.572858731924361e-05,
+      "loss": 2.5376,
+      "step": 3698
+    },
+    {
+      "epoch": 49.33444816053512,
+      "grad_norm": 0.7902242541313171,
+      "learning_rate": 3.5684093437152397e-05,
+      "loss": 2.887,
+      "step": 3699
+    },
+    {
+      "epoch": 49.34782608695652,
+      "grad_norm": 0.7505605220794678,
+      "learning_rate": 3.563959955506118e-05,
+      "loss": 2.7285,
+      "step": 3700
+    },
+    {
+      "epoch": 49.36120401337793,
+      "grad_norm": 0.9204686284065247,
+      "learning_rate": 3.559510567296997e-05,
+      "loss": 2.7478,
+      "step": 3701
+    },
+    {
+      "epoch": 49.374581939799334,
+      "grad_norm": 0.6998848915100098,
+      "learning_rate": 3.555061179087876e-05,
+      "loss": 2.405,
+      "step": 3702
+    },
+    {
+      "epoch": 49.38795986622073,
+      "grad_norm": 0.7633227109909058,
+      "learning_rate": 3.5506117908787546e-05,
+      "loss": 2.4304,
+      "step": 3703
+    },
+    {
+      "epoch": 49.40133779264214,
+      "grad_norm": 0.8731907606124878,
+      "learning_rate": 3.5461624026696336e-05,
+      "loss": 2.6215,
+      "step": 3704
+    },
+    {
+      "epoch": 49.414715719063544,
+      "grad_norm": 0.7742621302604675,
+      "learning_rate": 3.541713014460512e-05,
+      "loss": 2.6656,
+      "step": 3705
+    },
+    {
+      "epoch": 49.42809364548495,
+      "grad_norm": 0.7795779705047607,
+      "learning_rate": 3.537263626251391e-05,
+      "loss": 2.6521,
+      "step": 3706
+    },
+    {
+      "epoch": 49.441471571906355,
+      "grad_norm": 0.8041906356811523,
+      "learning_rate": 3.5328142380422696e-05,
+      "loss": 2.6883,
+      "step": 3707
+    },
+    {
+      "epoch": 49.45484949832776,
+      "grad_norm": 0.8720248937606812,
+      "learning_rate": 3.528364849833148e-05,
+      "loss": 2.295,
+      "step": 3708
+    },
+    {
+      "epoch": 49.468227424749166,
+      "grad_norm": 0.8696540594100952,
+      "learning_rate": 3.523915461624027e-05,
+      "loss": 2.6877,
+      "step": 3709
+    },
+    {
+      "epoch": 49.48160535117057,
+      "grad_norm": 0.9745573401451111,
+      "learning_rate": 3.519466073414906e-05,
+      "loss": 2.5912,
+      "step": 3710
+    },
+    {
+      "epoch": 49.49498327759197,
+      "grad_norm": 1.1198558807373047,
+      "learning_rate": 3.515016685205784e-05,
+      "loss": 2.7686,
+      "step": 3711
+    },
+    {
+      "epoch": 49.508361204013376,
+      "grad_norm": 0.759647011756897,
+      "learning_rate": 3.510567296996663e-05,
+      "loss": 2.6488,
+      "step": 3712
+    },
+    {
+      "epoch": 49.52173913043478,
+      "grad_norm": 0.7952780723571777,
+      "learning_rate": 3.506117908787542e-05,
+      "loss": 2.7659,
+      "step": 3713
+    },
+    {
+      "epoch": 49.53511705685619,
+      "grad_norm": 0.8217593431472778,
+      "learning_rate": 3.501668520578421e-05,
+      "loss": 2.7766,
+      "step": 3714
+    },
+    {
+      "epoch": 49.54849498327759,
+      "grad_norm": 0.7650991678237915,
+      "learning_rate": 3.4972191323692996e-05,
+      "loss": 2.5743,
+      "step": 3715
+    },
+    {
+      "epoch": 49.561872909699,
+      "grad_norm": 0.726296067237854,
+      "learning_rate": 3.492769744160178e-05,
+      "loss": 2.6984,
+      "step": 3716
+    },
+    {
+      "epoch": 49.575250836120404,
+      "grad_norm": 0.7600445747375488,
+      "learning_rate": 3.488320355951057e-05,
+      "loss": 2.454,
+      "step": 3717
+    },
+    {
+      "epoch": 49.58862876254181,
+      "grad_norm": 0.8778021335601807,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 2.7389,
+      "step": 3718
+    },
+    {
+      "epoch": 49.60200668896321,
+      "grad_norm": 0.7271983623504639,
+      "learning_rate": 3.4794215795328146e-05,
+      "loss": 2.4673,
+      "step": 3719
+    },
+    {
+      "epoch": 49.61538461538461,
+      "grad_norm": 0.7534515261650085,
+      "learning_rate": 3.4749721913236935e-05,
+      "loss": 2.7838,
+      "step": 3720
+    },
+    {
+      "epoch": 49.62876254180602,
+      "grad_norm": 0.9334620237350464,
+      "learning_rate": 3.470522803114572e-05,
+      "loss": 2.7135,
+      "step": 3721
+    },
+    {
+      "epoch": 49.642140468227424,
+      "grad_norm": 0.7726050019264221,
+      "learning_rate": 3.466073414905451e-05,
+      "loss": 2.5617,
+      "step": 3722
+    },
+    {
+      "epoch": 49.65551839464883,
+      "grad_norm": 0.8581987023353577,
+      "learning_rate": 3.4616240266963296e-05,
+      "loss": 2.7521,
+      "step": 3723
+    },
+    {
+      "epoch": 49.668896321070235,
+      "grad_norm": 0.8535495400428772,
+      "learning_rate": 3.4571746384872085e-05,
+      "loss": 2.4661,
+      "step": 3724
+    },
+    {
+      "epoch": 49.68227424749164,
+      "grad_norm": 0.691425621509552,
+      "learning_rate": 3.4527252502780874e-05,
+      "loss": 2.4683,
+      "step": 3725
+    },
+    {
+      "epoch": 49.69565217391305,
+      "grad_norm": 0.7674373984336853,
+      "learning_rate": 3.4482758620689657e-05,
+      "loss": 2.6441,
+      "step": 3726
+    },
+    {
+      "epoch": 49.709030100334445,
+      "grad_norm": 0.7647296190261841,
+      "learning_rate": 3.4438264738598446e-05,
+      "loss": 2.6286,
+      "step": 3727
+    },
+    {
+      "epoch": 49.72240802675585,
+      "grad_norm": 0.7668997645378113,
+      "learning_rate": 3.439377085650723e-05,
+      "loss": 2.762,
+      "step": 3728
+    },
+    {
+      "epoch": 49.735785953177256,
+      "grad_norm": 0.8111737370491028,
+      "learning_rate": 3.434927697441602e-05,
+      "loss": 2.6294,
+      "step": 3729
+    },
+    {
+      "epoch": 49.74916387959866,
+      "grad_norm": 0.7185457348823547,
+      "learning_rate": 3.4304783092324806e-05,
+      "loss": 2.5763,
+      "step": 3730
+    },
+    {
+      "epoch": 49.76254180602007,
+      "grad_norm": 0.8142096996307373,
+      "learning_rate": 3.4260289210233596e-05,
+      "loss": 2.8209,
+      "step": 3731
+    },
+    {
+      "epoch": 49.77591973244147,
+      "grad_norm": 0.7466010451316833,
+      "learning_rate": 3.421579532814238e-05,
+      "loss": 2.7076,
+      "step": 3732
+    },
+    {
+      "epoch": 49.78929765886288,
+      "grad_norm": 0.8199285268783569,
+      "learning_rate": 3.417130144605117e-05,
+      "loss": 2.3427,
+      "step": 3733
+    },
+    {
+      "epoch": 49.802675585284284,
+      "grad_norm": 0.8280915021896362,
+      "learning_rate": 3.4126807563959956e-05,
+      "loss": 2.5477,
+      "step": 3734
+    },
+    {
+      "epoch": 49.81605351170568,
+      "grad_norm": 0.7132146954536438,
+      "learning_rate": 3.4082313681868746e-05,
+      "loss": 2.6196,
+      "step": 3735
+    },
+    {
+      "epoch": 49.82943143812709,
+      "grad_norm": 0.7380895018577576,
+      "learning_rate": 3.4037819799777535e-05,
+      "loss": 2.8091,
+      "step": 3736
+    },
+    {
+      "epoch": 49.84280936454849,
+      "grad_norm": 0.7611865401268005,
+      "learning_rate": 3.399332591768632e-05,
+      "loss": 2.7362,
+      "step": 3737
+    },
+    {
+      "epoch": 49.8561872909699,
+      "grad_norm": 0.7730408906936646,
+      "learning_rate": 3.3948832035595106e-05,
+      "loss": 2.7238,
+      "step": 3738
+    },
+    {
+      "epoch": 49.869565217391305,
+      "grad_norm": 0.7552332282066345,
+      "learning_rate": 3.3904338153503895e-05,
+      "loss": 2.7884,
+      "step": 3739
+    },
+    {
+      "epoch": 49.88294314381271,
+      "grad_norm": 0.8960534334182739,
+      "learning_rate": 3.3859844271412685e-05,
+      "loss": 2.7067,
+      "step": 3740
+    },
+    {
+      "epoch": 49.896321070234116,
+      "grad_norm": 0.8455086350440979,
+      "learning_rate": 3.3815350389321474e-05,
+      "loss": 2.5756,
+      "step": 3741
+    },
+    {
+      "epoch": 49.90969899665552,
+      "grad_norm": 0.8840497732162476,
+      "learning_rate": 3.3770856507230256e-05,
+      "loss": 2.6779,
+      "step": 3742
+    },
+    {
+      "epoch": 49.92307692307692,
+      "grad_norm": 0.7964041829109192,
+      "learning_rate": 3.3726362625139045e-05,
+      "loss": 2.7642,
+      "step": 3743
+    },
+    {
+      "epoch": 49.936454849498325,
+      "grad_norm": 0.7956447601318359,
+      "learning_rate": 3.3681868743047834e-05,
+      "loss": 2.5019,
+      "step": 3744
+    },
+    {
+      "epoch": 49.94983277591973,
+      "grad_norm": 0.743416965007782,
+      "learning_rate": 3.3637374860956624e-05,
+      "loss": 2.5265,
+      "step": 3745
+    },
+    {
+      "epoch": 49.96321070234114,
+      "grad_norm": 0.8189771771430969,
+      "learning_rate": 3.359288097886541e-05,
+      "loss": 2.671,
+      "step": 3746
+    },
+    {
+      "epoch": 49.97658862876254,
+      "grad_norm": 0.778721034526825,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 2.6753,
+      "step": 3747
+    },
+    {
+      "epoch": 49.98996655518395,
+      "grad_norm": 0.7933366894721985,
+      "learning_rate": 3.3503893214682984e-05,
+      "loss": 2.6491,
+      "step": 3748
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 1.0016274452209473,
+      "learning_rate": 3.345939933259177e-05,
+      "loss": 2.5347,
+      "step": 3749
+    },
+    {
+      "epoch": 50.013377926421406,
+      "grad_norm": 0.7115795016288757,
+      "learning_rate": 3.3414905450500556e-05,
+      "loss": 2.4834,
+      "step": 3750
+    },
+    {
+      "epoch": 50.02675585284281,
+      "grad_norm": 0.6772646307945251,
+      "learning_rate": 3.3370411568409345e-05,
+      "loss": 2.7454,
+      "step": 3751
+    },
+    {
+      "epoch": 50.04013377926422,
+      "grad_norm": 0.7457217574119568,
+      "learning_rate": 3.3325917686318134e-05,
+      "loss": 2.6569,
+      "step": 3752
+    },
+    {
+      "epoch": 50.05351170568562,
+      "grad_norm": 0.6905792951583862,
+      "learning_rate": 3.328142380422692e-05,
+      "loss": 2.5026,
+      "step": 3753
+    },
+    {
+      "epoch": 50.06688963210702,
+      "grad_norm": 0.6815123558044434,
+      "learning_rate": 3.3236929922135706e-05,
+      "loss": 2.6324,
+      "step": 3754
+    },
+    {
+      "epoch": 50.080267558528426,
+      "grad_norm": 0.7720118761062622,
+      "learning_rate": 3.3192436040044495e-05,
+      "loss": 2.6267,
+      "step": 3755
+    },
+    {
+      "epoch": 50.09364548494983,
+      "grad_norm": 0.7029479146003723,
+      "learning_rate": 3.3147942157953284e-05,
+      "loss": 2.3862,
+      "step": 3756
+    },
+    {
+      "epoch": 50.10702341137124,
+      "grad_norm": 0.6969912648200989,
+      "learning_rate": 3.310344827586207e-05,
+      "loss": 2.6517,
+      "step": 3757
+    },
+    {
+      "epoch": 50.12040133779264,
+      "grad_norm": 0.71588134765625,
+      "learning_rate": 3.3058954393770856e-05,
+      "loss": 2.5768,
+      "step": 3758
+    },
+    {
+      "epoch": 50.13377926421405,
+      "grad_norm": 0.7532247304916382,
+      "learning_rate": 3.3014460511679645e-05,
+      "loss": 2.7685,
+      "step": 3759
+    },
+    {
+      "epoch": 50.147157190635454,
+      "grad_norm": 0.7348672747612,
+      "learning_rate": 3.2969966629588434e-05,
+      "loss": 2.653,
+      "step": 3760
+    },
+    {
+      "epoch": 50.16053511705686,
+      "grad_norm": 0.7328438758850098,
+      "learning_rate": 3.292547274749722e-05,
+      "loss": 2.4988,
+      "step": 3761
+    },
+    {
+      "epoch": 50.17391304347826,
+      "grad_norm": 0.7021594047546387,
+      "learning_rate": 3.288097886540601e-05,
+      "loss": 2.4435,
+      "step": 3762
+    },
+    {
+      "epoch": 50.187290969899664,
+      "grad_norm": 0.6721189022064209,
+      "learning_rate": 3.2836484983314795e-05,
+      "loss": 2.5429,
+      "step": 3763
+    },
+    {
+      "epoch": 50.20066889632107,
+      "grad_norm": 0.7436168193817139,
+      "learning_rate": 3.2791991101223584e-05,
+      "loss": 2.9638,
+      "step": 3764
+    },
+    {
+      "epoch": 50.214046822742475,
+      "grad_norm": 0.7094192504882812,
+      "learning_rate": 3.274749721913237e-05,
+      "loss": 2.6454,
+      "step": 3765
+    },
+    {
+      "epoch": 50.22742474916388,
+      "grad_norm": 0.7569435238838196,
+      "learning_rate": 3.270300333704116e-05,
+      "loss": 2.5448,
+      "step": 3766
+    },
+    {
+      "epoch": 50.240802675585286,
+      "grad_norm": 0.692587673664093,
+      "learning_rate": 3.265850945494995e-05,
+      "loss": 2.5065,
+      "step": 3767
+    },
+    {
+      "epoch": 50.25418060200669,
+      "grad_norm": 0.6578419208526611,
+      "learning_rate": 3.2614015572858734e-05,
+      "loss": 2.3608,
+      "step": 3768
+    },
+    {
+      "epoch": 50.2675585284281,
+      "grad_norm": 0.7179403305053711,
+      "learning_rate": 3.256952169076752e-05,
+      "loss": 2.5417,
+      "step": 3769
+    },
+    {
+      "epoch": 50.280936454849495,
+      "grad_norm": 0.7056894898414612,
+      "learning_rate": 3.2525027808676305e-05,
+      "loss": 2.5924,
+      "step": 3770
+    },
+    {
+      "epoch": 50.2943143812709,
+      "grad_norm": 0.6727268695831299,
+      "learning_rate": 3.2480533926585095e-05,
+      "loss": 2.7164,
+      "step": 3771
+    },
+    {
+      "epoch": 50.30769230769231,
+      "grad_norm": 0.7499269247055054,
+      "learning_rate": 3.2436040044493884e-05,
+      "loss": 2.6172,
+      "step": 3772
+    },
+    {
+      "epoch": 50.32107023411371,
+      "grad_norm": 0.735407829284668,
+      "learning_rate": 3.239154616240267e-05,
+      "loss": 2.5634,
+      "step": 3773
+    },
+    {
+      "epoch": 50.33444816053512,
+      "grad_norm": 0.8683627247810364,
+      "learning_rate": 3.2347052280311455e-05,
+      "loss": 2.6344,
+      "step": 3774
+    },
+    {
+      "epoch": 50.34782608695652,
+      "grad_norm": 0.6927014589309692,
+      "learning_rate": 3.2302558398220244e-05,
+      "loss": 2.4513,
+      "step": 3775
+    },
+    {
+      "epoch": 50.36120401337793,
+      "grad_norm": 0.7480311989784241,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 2.5682,
+      "step": 3776
+    },
+    {
+      "epoch": 50.374581939799334,
+      "grad_norm": 0.6959646344184875,
+      "learning_rate": 3.221357063403782e-05,
+      "loss": 2.4993,
+      "step": 3777
+    },
+    {
+      "epoch": 50.38795986622073,
+      "grad_norm": 0.6775467395782471,
+      "learning_rate": 3.2169076751946605e-05,
+      "loss": 2.4055,
+      "step": 3778
+    },
+    {
+      "epoch": 50.40133779264214,
+      "grad_norm": 0.74509197473526,
+      "learning_rate": 3.2124582869855394e-05,
+      "loss": 2.4928,
+      "step": 3779
+    },
+    {
+      "epoch": 50.414715719063544,
+      "grad_norm": 0.7157049775123596,
+      "learning_rate": 3.2080088987764183e-05,
+      "loss": 2.7255,
+      "step": 3780
+    },
+    {
+      "epoch": 50.42809364548495,
+      "grad_norm": 0.8020923733711243,
+      "learning_rate": 3.203559510567297e-05,
+      "loss": 2.6627,
+      "step": 3781
+    },
+    {
+      "epoch": 50.441471571906355,
+      "grad_norm": 0.7195694446563721,
+      "learning_rate": 3.199110122358176e-05,
+      "loss": 2.5675,
+      "step": 3782
+    },
+    {
+      "epoch": 50.45484949832776,
+      "grad_norm": 0.6762703657150269,
+      "learning_rate": 3.1946607341490544e-05,
+      "loss": 2.5307,
+      "step": 3783
+    },
+    {
+      "epoch": 50.468227424749166,
+      "grad_norm": 0.697120726108551,
+      "learning_rate": 3.190211345939933e-05,
+      "loss": 2.7164,
+      "step": 3784
+    },
+    {
+      "epoch": 50.48160535117057,
+      "grad_norm": 0.7056689262390137,
+      "learning_rate": 3.185761957730812e-05,
+      "loss": 2.5651,
+      "step": 3785
+    },
+    {
+      "epoch": 50.49498327759197,
+      "grad_norm": 0.7277424335479736,
+      "learning_rate": 3.181312569521691e-05,
+      "loss": 2.482,
+      "step": 3786
+    },
+    {
+      "epoch": 50.508361204013376,
+      "grad_norm": 0.7807835340499878,
+      "learning_rate": 3.17686318131257e-05,
+      "loss": 2.5597,
+      "step": 3787
+    },
+    {
+      "epoch": 50.52173913043478,
+      "grad_norm": 0.7482343912124634,
+      "learning_rate": 3.172413793103448e-05,
+      "loss": 2.4444,
+      "step": 3788
+    },
+    {
+      "epoch": 50.53511705685619,
+      "grad_norm": 0.6976211071014404,
+      "learning_rate": 3.167964404894327e-05,
+      "loss": 2.6283,
+      "step": 3789
+    },
+    {
+      "epoch": 50.54849498327759,
+      "grad_norm": 0.6839179396629333,
+      "learning_rate": 3.1635150166852055e-05,
+      "loss": 2.712,
+      "step": 3790
+    },
+    {
+      "epoch": 50.561872909699,
+      "grad_norm": 0.7053922414779663,
+      "learning_rate": 3.1590656284760844e-05,
+      "loss": 2.5473,
+      "step": 3791
+    },
+    {
+      "epoch": 50.575250836120404,
+      "grad_norm": 0.6922830939292908,
+      "learning_rate": 3.154616240266963e-05,
+      "loss": 2.6293,
+      "step": 3792
+    },
+    {
+      "epoch": 50.58862876254181,
+      "grad_norm": 0.8351788520812988,
+      "learning_rate": 3.150166852057842e-05,
+      "loss": 2.807,
+      "step": 3793
+    },
+    {
+      "epoch": 50.60200668896321,
+      "grad_norm": 0.7981114387512207,
+      "learning_rate": 3.1457174638487205e-05,
+      "loss": 2.3676,
+      "step": 3794
+    },
+    {
+      "epoch": 50.61538461538461,
+      "grad_norm": 0.7550584077835083,
+      "learning_rate": 3.1412680756395994e-05,
+      "loss": 2.7333,
+      "step": 3795
+    },
+    {
+      "epoch": 50.62876254180602,
+      "grad_norm": 0.7444553971290588,
+      "learning_rate": 3.136818687430478e-05,
+      "loss": 2.5449,
+      "step": 3796
+    },
+    {
+      "epoch": 50.642140468227424,
+      "grad_norm": 0.7490617632865906,
+      "learning_rate": 3.132369299221357e-05,
+      "loss": 2.523,
+      "step": 3797
+    },
+    {
+      "epoch": 50.65551839464883,
+      "grad_norm": 0.7713622450828552,
+      "learning_rate": 3.127919911012236e-05,
+      "loss": 2.2943,
+      "step": 3798
+    },
+    {
+      "epoch": 50.668896321070235,
+      "grad_norm": 0.6862428784370422,
+      "learning_rate": 3.1234705228031144e-05,
+      "loss": 2.5255,
+      "step": 3799
+    },
+    {
+      "epoch": 50.68227424749164,
+      "grad_norm": 0.755214273929596,
+      "learning_rate": 3.119021134593993e-05,
+      "loss": 2.5067,
+      "step": 3800
+    },
+    {
+      "epoch": 50.69565217391305,
+      "grad_norm": 0.690545380115509,
+      "learning_rate": 3.114571746384872e-05,
+      "loss": 2.5803,
+      "step": 3801
+    },
+    {
+      "epoch": 50.709030100334445,
+      "grad_norm": 0.7627785801887512,
+      "learning_rate": 3.110122358175751e-05,
+      "loss": 2.5416,
+      "step": 3802
+    },
+    {
+      "epoch": 50.72240802675585,
+      "grad_norm": 0.7653105854988098,
+      "learning_rate": 3.10567296996663e-05,
+      "loss": 2.8272,
+      "step": 3803
+    },
+    {
+      "epoch": 50.735785953177256,
+      "grad_norm": 0.7366088628768921,
+      "learning_rate": 3.101223581757508e-05,
+      "loss": 2.6717,
+      "step": 3804
+    },
+    {
+      "epoch": 50.74916387959866,
+      "grad_norm": 0.7387148141860962,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 2.6479,
+      "step": 3805
+    },
+    {
+      "epoch": 50.76254180602007,
+      "grad_norm": 0.76811283826828,
+      "learning_rate": 3.092324805339266e-05,
+      "loss": 2.6197,
+      "step": 3806
+    },
+    {
+      "epoch": 50.77591973244147,
+      "grad_norm": 0.7530215978622437,
+      "learning_rate": 3.087875417130145e-05,
+      "loss": 2.7431,
+      "step": 3807
+    },
+    {
+      "epoch": 50.78929765886288,
+      "grad_norm": 0.7515009641647339,
+      "learning_rate": 3.083426028921024e-05,
+      "loss": 2.7594,
+      "step": 3808
+    },
+    {
+      "epoch": 50.802675585284284,
+      "grad_norm": 0.6830329298973083,
+      "learning_rate": 3.078976640711902e-05,
+      "loss": 2.6105,
+      "step": 3809
+    },
+    {
+      "epoch": 50.81605351170568,
+      "grad_norm": 0.7070818543434143,
+      "learning_rate": 3.074527252502781e-05,
+      "loss": 2.6362,
+      "step": 3810
+    },
+    {
+      "epoch": 50.82943143812709,
+      "grad_norm": 0.7041717171669006,
+      "learning_rate": 3.0700778642936593e-05,
+      "loss": 2.5358,
+      "step": 3811
+    },
+    {
+      "epoch": 50.84280936454849,
+      "grad_norm": 0.7511767148971558,
+      "learning_rate": 3.065628476084538e-05,
+      "loss": 2.482,
+      "step": 3812
+    },
+    {
+      "epoch": 50.8561872909699,
+      "grad_norm": 0.7591724395751953,
+      "learning_rate": 3.061179087875417e-05,
+      "loss": 2.7301,
+      "step": 3813
+    },
+    {
+      "epoch": 50.869565217391305,
+      "grad_norm": 0.7175691723823547,
+      "learning_rate": 3.056729699666296e-05,
+      "loss": 2.689,
+      "step": 3814
+    },
+    {
+      "epoch": 50.88294314381271,
+      "grad_norm": 0.7458793520927429,
+      "learning_rate": 3.052280311457174e-05,
+      "loss": 2.5394,
+      "step": 3815
+    },
+    {
+      "epoch": 50.896321070234116,
+      "grad_norm": 0.7417124509811401,
+      "learning_rate": 3.0478309232480532e-05,
+      "loss": 2.7544,
+      "step": 3816
+    },
+    {
+      "epoch": 50.90969899665552,
+      "grad_norm": 0.7595741748809814,
+      "learning_rate": 3.043381535038932e-05,
+      "loss": 2.5443,
+      "step": 3817
+    },
+    {
+      "epoch": 50.92307692307692,
+      "grad_norm": 0.6968060731887817,
+      "learning_rate": 3.038932146829811e-05,
+      "loss": 2.8318,
+      "step": 3818
+    },
+    {
+      "epoch": 50.936454849498325,
+      "grad_norm": 0.8147920966148376,
+      "learning_rate": 3.0344827586206897e-05,
+      "loss": 2.51,
+      "step": 3819
+    },
+    {
+      "epoch": 50.94983277591973,
+      "grad_norm": 0.7169308066368103,
+      "learning_rate": 3.0300333704115686e-05,
+      "loss": 2.7062,
+      "step": 3820
+    },
+    {
+      "epoch": 50.96321070234114,
+      "grad_norm": 0.6849644184112549,
+      "learning_rate": 3.025583982202447e-05,
+      "loss": 2.5129,
+      "step": 3821
+    },
+    {
+      "epoch": 50.97658862876254,
+      "grad_norm": 0.6856783628463745,
+      "learning_rate": 3.021134593993326e-05,
+      "loss": 2.606,
+      "step": 3822
+    },
+    {
+      "epoch": 50.98996655518395,
+      "grad_norm": 0.6814429759979248,
+      "learning_rate": 3.016685205784205e-05,
+      "loss": 2.606,
+      "step": 3823
+    },
+    {
+      "epoch": 51.0,
+      "grad_norm": 0.8198628425598145,
+      "learning_rate": 3.0122358175750836e-05,
+      "loss": 2.1563,
+      "step": 3824
+    },
+    {
+      "epoch": 51.013377926421406,
+      "grad_norm": 0.6644238829612732,
+      "learning_rate": 3.0077864293659625e-05,
+      "loss": 2.4632,
+      "step": 3825
+    },
+    {
+      "epoch": 51.02675585284281,
+      "grad_norm": 0.6920905113220215,
+      "learning_rate": 3.003337041156841e-05,
+      "loss": 2.436,
+      "step": 3826
+    },
+    {
+      "epoch": 51.04013377926422,
+      "grad_norm": 0.6688025593757629,
+      "learning_rate": 2.99888765294772e-05,
+      "loss": 2.6136,
+      "step": 3827
+    },
+    {
+      "epoch": 51.05351170568562,
+      "grad_norm": 0.6641514897346497,
+      "learning_rate": 2.994438264738599e-05,
+      "loss": 2.5133,
+      "step": 3828
+    },
+    {
+      "epoch": 51.06688963210702,
+      "grad_norm": 0.6722247004508972,
+      "learning_rate": 2.9899888765294775e-05,
+      "loss": 2.5963,
+      "step": 3829
+    },
+    {
+      "epoch": 51.080267558528426,
+      "grad_norm": 0.698442280292511,
+      "learning_rate": 2.9855394883203564e-05,
+      "loss": 2.5602,
+      "step": 3830
+    },
+    {
+      "epoch": 51.09364548494983,
+      "grad_norm": 0.6918148398399353,
+      "learning_rate": 2.9810901001112346e-05,
+      "loss": 2.6382,
+      "step": 3831
+    },
+    {
+      "epoch": 51.10702341137124,
+      "grad_norm": 0.7268350720405579,
+      "learning_rate": 2.9766407119021132e-05,
+      "loss": 2.6523,
+      "step": 3832
+    },
+    {
+      "epoch": 51.12040133779264,
+      "grad_norm": 0.708554744720459,
+      "learning_rate": 2.972191323692992e-05,
+      "loss": 2.6575,
+      "step": 3833
+    },
+    {
+      "epoch": 51.13377926421405,
+      "grad_norm": 0.6651294231414795,
+      "learning_rate": 2.967741935483871e-05,
+      "loss": 2.6154,
+      "step": 3834
+    },
+    {
+      "epoch": 51.147157190635454,
+      "grad_norm": 0.6972305774688721,
+      "learning_rate": 2.9632925472747496e-05,
+      "loss": 2.3533,
+      "step": 3835
+    },
+    {
+      "epoch": 51.16053511705686,
+      "grad_norm": 0.6912644505500793,
+      "learning_rate": 2.9588431590656285e-05,
+      "loss": 2.6753,
+      "step": 3836
+    },
+    {
+      "epoch": 51.17391304347826,
+      "grad_norm": 0.7234140038490295,
+      "learning_rate": 2.954393770856507e-05,
+      "loss": 2.7457,
+      "step": 3837
+    },
+    {
+      "epoch": 51.187290969899664,
+      "grad_norm": 0.7113907337188721,
+      "learning_rate": 2.949944382647386e-05,
+      "loss": 2.5897,
+      "step": 3838
+    },
+    {
+      "epoch": 51.20066889632107,
+      "grad_norm": 0.6739325523376465,
+      "learning_rate": 2.945494994438265e-05,
+      "loss": 2.4148,
+      "step": 3839
+    },
+    {
+      "epoch": 51.214046822742475,
+      "grad_norm": 0.6763201951980591,
+      "learning_rate": 2.9410456062291435e-05,
+      "loss": 2.4693,
+      "step": 3840
+    },
+    {
+      "epoch": 51.22742474916388,
+      "grad_norm": 0.6864936947822571,
+      "learning_rate": 2.9365962180200224e-05,
+      "loss": 2.5916,
+      "step": 3841
+    },
+    {
+      "epoch": 51.240802675585286,
+      "grad_norm": 0.6833862066268921,
+      "learning_rate": 2.932146829810901e-05,
+      "loss": 2.5388,
+      "step": 3842
+    },
+    {
+      "epoch": 51.25418060200669,
+      "grad_norm": 0.6838318705558777,
+      "learning_rate": 2.92769744160178e-05,
+      "loss": 2.7633,
+      "step": 3843
+    },
+    {
+      "epoch": 51.2675585284281,
+      "grad_norm": 0.729987382888794,
+      "learning_rate": 2.923248053392659e-05,
+      "loss": 2.7342,
+      "step": 3844
+    },
+    {
+      "epoch": 51.280936454849495,
+      "grad_norm": 0.7263936400413513,
+      "learning_rate": 2.9187986651835374e-05,
+      "loss": 2.668,
+      "step": 3845
+    },
+    {
+      "epoch": 51.2943143812709,
+      "grad_norm": 0.6780832409858704,
+      "learning_rate": 2.9143492769744163e-05,
+      "loss": 2.5854,
+      "step": 3846
+    },
+    {
+      "epoch": 51.30769230769231,
+      "grad_norm": 0.6763190031051636,
+      "learning_rate": 2.909899888765295e-05,
+      "loss": 2.2402,
+      "step": 3847
+    },
+    {
+      "epoch": 51.32107023411371,
+      "grad_norm": 0.725577712059021,
+      "learning_rate": 2.905450500556174e-05,
+      "loss": 2.4074,
+      "step": 3848
+    },
+    {
+      "epoch": 51.33444816053512,
+      "grad_norm": 0.720762312412262,
+      "learning_rate": 2.9010011123470528e-05,
+      "loss": 2.6617,
+      "step": 3849
+    },
+    {
+      "epoch": 51.34782608695652,
+      "grad_norm": 0.7184842228889465,
+      "learning_rate": 2.8965517241379313e-05,
+      "loss": 2.7605,
+      "step": 3850
+    },
+    {
+      "epoch": 51.36120401337793,
+      "grad_norm": 0.7217162847518921,
+      "learning_rate": 2.8921023359288102e-05,
+      "loss": 2.735,
+      "step": 3851
+    },
+    {
+      "epoch": 51.374581939799334,
+      "grad_norm": 0.7379841208457947,
+      "learning_rate": 2.8876529477196885e-05,
+      "loss": 2.4665,
+      "step": 3852
+    },
+    {
+      "epoch": 51.38795986622073,
+      "grad_norm": 0.7272505760192871,
+      "learning_rate": 2.883203559510567e-05,
+      "loss": 2.5609,
+      "step": 3853
+    },
+    {
+      "epoch": 51.40133779264214,
+      "grad_norm": 0.7013726830482483,
+      "learning_rate": 2.878754171301446e-05,
+      "loss": 2.7946,
+      "step": 3854
+    },
+    {
+      "epoch": 51.414715719063544,
+      "grad_norm": 0.7304871082305908,
+      "learning_rate": 2.874304783092325e-05,
+      "loss": 2.3282,
+      "step": 3855
+    },
+    {
+      "epoch": 51.42809364548495,
+      "grad_norm": 0.6831701993942261,
+      "learning_rate": 2.8698553948832035e-05,
+      "loss": 2.6357,
+      "step": 3856
+    },
+    {
+      "epoch": 51.441471571906355,
+      "grad_norm": 0.7182192206382751,
+      "learning_rate": 2.8654060066740824e-05,
+      "loss": 2.319,
+      "step": 3857
+    },
+    {
+      "epoch": 51.45484949832776,
+      "grad_norm": 0.7064784169197083,
+      "learning_rate": 2.860956618464961e-05,
+      "loss": 2.6095,
+      "step": 3858
+    },
+    {
+      "epoch": 51.468227424749166,
+      "grad_norm": 0.687251627445221,
+      "learning_rate": 2.85650723025584e-05,
+      "loss": 2.4059,
+      "step": 3859
+    },
+    {
+      "epoch": 51.48160535117057,
+      "grad_norm": 0.672291100025177,
+      "learning_rate": 2.8520578420467188e-05,
+      "loss": 2.4958,
+      "step": 3860
+    },
+    {
+      "epoch": 51.49498327759197,
+      "grad_norm": 0.7099835276603699,
+      "learning_rate": 2.8476084538375974e-05,
+      "loss": 2.6736,
+      "step": 3861
+    },
+    {
+      "epoch": 51.508361204013376,
+      "grad_norm": 0.7420845031738281,
+      "learning_rate": 2.8431590656284763e-05,
+      "loss": 2.6133,
+      "step": 3862
+    },
+    {
+      "epoch": 51.52173913043478,
+      "grad_norm": 0.6821379065513611,
+      "learning_rate": 2.838709677419355e-05,
+      "loss": 2.4062,
+      "step": 3863
+    },
+    {
+      "epoch": 51.53511705685619,
+      "grad_norm": 0.7185118794441223,
+      "learning_rate": 2.8342602892102338e-05,
+      "loss": 2.7629,
+      "step": 3864
+    },
+    {
+      "epoch": 51.54849498327759,
+      "grad_norm": 0.7404976487159729,
+      "learning_rate": 2.8298109010011127e-05,
+      "loss": 2.6438,
+      "step": 3865
+    },
+    {
+      "epoch": 51.561872909699,
+      "grad_norm": 0.7032206654548645,
+      "learning_rate": 2.8253615127919913e-05,
+      "loss": 2.7265,
+      "step": 3866
+    },
+    {
+      "epoch": 51.575250836120404,
+      "grad_norm": 0.6758597493171692,
+      "learning_rate": 2.8209121245828702e-05,
+      "loss": 2.6224,
+      "step": 3867
+    },
+    {
+      "epoch": 51.58862876254181,
+      "grad_norm": 0.6467934250831604,
+      "learning_rate": 2.8164627363737488e-05,
+      "loss": 2.4554,
+      "step": 3868
+    },
+    {
+      "epoch": 51.60200668896321,
+      "grad_norm": 0.6494783759117126,
+      "learning_rate": 2.8120133481646277e-05,
+      "loss": 2.4571,
+      "step": 3869
+    },
+    {
+      "epoch": 51.61538461538461,
+      "grad_norm": 0.6202491521835327,
+      "learning_rate": 2.8075639599555066e-05,
+      "loss": 2.1063,
+      "step": 3870
+    },
+    {
+      "epoch": 51.62876254180602,
+      "grad_norm": 0.6956090331077576,
+      "learning_rate": 2.8031145717463852e-05,
+      "loss": 2.719,
+      "step": 3871
+    },
+    {
+      "epoch": 51.642140468227424,
+      "grad_norm": 0.6946972012519836,
+      "learning_rate": 2.798665183537264e-05,
+      "loss": 2.5629,
+      "step": 3872
+    },
+    {
+      "epoch": 51.65551839464883,
+      "grad_norm": 0.6977154612541199,
+      "learning_rate": 2.7942157953281423e-05,
+      "loss": 2.6202,
+      "step": 3873
+    },
+    {
+      "epoch": 51.668896321070235,
+      "grad_norm": 0.7147237062454224,
+      "learning_rate": 2.789766407119021e-05,
+      "loss": 2.4834,
+      "step": 3874
+    },
+    {
+      "epoch": 51.68227424749164,
+      "grad_norm": 0.644244372844696,
+      "learning_rate": 2.7853170189099e-05,
+      "loss": 2.4317,
+      "step": 3875
+    },
+    {
+      "epoch": 51.69565217391305,
+      "grad_norm": 0.6681378483772278,
+      "learning_rate": 2.7808676307007788e-05,
+      "loss": 2.459,
+      "step": 3876
+    },
+    {
+      "epoch": 51.709030100334445,
+      "grad_norm": 0.6902260780334473,
+      "learning_rate": 2.7764182424916573e-05,
+      "loss": 2.5687,
+      "step": 3877
+    },
+    {
+      "epoch": 51.72240802675585,
+      "grad_norm": 0.6847380995750427,
+      "learning_rate": 2.7719688542825363e-05,
+      "loss": 2.7226,
+      "step": 3878
+    },
+    {
+      "epoch": 51.735785953177256,
+      "grad_norm": 0.664297342300415,
+      "learning_rate": 2.7675194660734148e-05,
+      "loss": 2.5813,
+      "step": 3879
+    },
+    {
+      "epoch": 51.74916387959866,
+      "grad_norm": 0.6731799840927124,
+      "learning_rate": 2.7630700778642937e-05,
+      "loss": 2.4737,
+      "step": 3880
+    },
+    {
+      "epoch": 51.76254180602007,
+      "grad_norm": 0.642483115196228,
+      "learning_rate": 2.7586206896551727e-05,
+      "loss": 2.4885,
+      "step": 3881
+    },
+    {
+      "epoch": 51.77591973244147,
+      "grad_norm": 0.7031340599060059,
+      "learning_rate": 2.7541713014460512e-05,
+      "loss": 2.574,
+      "step": 3882
+    },
+    {
+      "epoch": 51.78929765886288,
+      "grad_norm": 0.6814936995506287,
+      "learning_rate": 2.74972191323693e-05,
+      "loss": 2.6551,
+      "step": 3883
+    },
+    {
+      "epoch": 51.802675585284284,
+      "grad_norm": 0.6501394510269165,
+      "learning_rate": 2.7452725250278087e-05,
+      "loss": 2.458,
+      "step": 3884
+    },
+    {
+      "epoch": 51.81605351170568,
+      "grad_norm": 0.6935567855834961,
+      "learning_rate": 2.7408231368186877e-05,
+      "loss": 2.621,
+      "step": 3885
+    },
+    {
+      "epoch": 51.82943143812709,
+      "grad_norm": 0.6672533750534058,
+      "learning_rate": 2.7363737486095666e-05,
+      "loss": 2.4009,
+      "step": 3886
+    },
+    {
+      "epoch": 51.84280936454849,
+      "grad_norm": 0.6834401488304138,
+      "learning_rate": 2.731924360400445e-05,
+      "loss": 2.6903,
+      "step": 3887
+    },
+    {
+      "epoch": 51.8561872909699,
+      "grad_norm": 0.7198247909545898,
+      "learning_rate": 2.727474972191324e-05,
+      "loss": 2.7527,
+      "step": 3888
+    },
+    {
+      "epoch": 51.869565217391305,
+      "grad_norm": 0.6896756887435913,
+      "learning_rate": 2.7230255839822026e-05,
+      "loss": 2.6941,
+      "step": 3889
+    },
+    {
+      "epoch": 51.88294314381271,
+      "grad_norm": 0.6827479600906372,
+      "learning_rate": 2.7185761957730816e-05,
+      "loss": 2.6794,
+      "step": 3890
+    },
+    {
+      "epoch": 51.896321070234116,
+      "grad_norm": 0.7218233942985535,
+      "learning_rate": 2.7141268075639605e-05,
+      "loss": 2.5273,
+      "step": 3891
+    },
+    {
+      "epoch": 51.90969899665552,
+      "grad_norm": 0.6801145076751709,
+      "learning_rate": 2.709677419354839e-05,
+      "loss": 2.4922,
+      "step": 3892
+    },
+    {
+      "epoch": 51.92307692307692,
+      "grad_norm": 0.6611994504928589,
+      "learning_rate": 2.7052280311457173e-05,
+      "loss": 2.5052,
+      "step": 3893
+    },
+    {
+      "epoch": 51.936454849498325,
+      "grad_norm": 0.6727268099784851,
+      "learning_rate": 2.7007786429365962e-05,
+      "loss": 2.6148,
+      "step": 3894
+    },
+    {
+      "epoch": 51.94983277591973,
+      "grad_norm": 0.7804195284843445,
+      "learning_rate": 2.6963292547274748e-05,
+      "loss": 2.5588,
+      "step": 3895
+    },
+    {
+      "epoch": 51.96321070234114,
+      "grad_norm": 0.6668366193771362,
+      "learning_rate": 2.6918798665183537e-05,
+      "loss": 2.5478,
+      "step": 3896
+    },
+    {
+      "epoch": 51.97658862876254,
+      "grad_norm": 0.6974259614944458,
+      "learning_rate": 2.6874304783092323e-05,
+      "loss": 2.7024,
+      "step": 3897
+    },
+    {
+      "epoch": 51.98996655518395,
+      "grad_norm": 0.6899526715278625,
+      "learning_rate": 2.6829810901001112e-05,
+      "loss": 2.5015,
+      "step": 3898
+    },
+    {
+      "epoch": 52.0,
+      "grad_norm": 0.824705183506012,
+      "learning_rate": 2.67853170189099e-05,
+      "loss": 2.668,
+      "step": 3899
+    },
+    {
+      "epoch": 52.013377926421406,
+      "grad_norm": 0.6349574327468872,
+      "learning_rate": 2.6740823136818687e-05,
+      "loss": 2.3712,
+      "step": 3900
+    },
+    {
+      "epoch": 52.02675585284281,
+      "grad_norm": 0.6901255249977112,
+      "learning_rate": 2.6696329254727476e-05,
+      "loss": 2.4776,
+      "step": 3901
+    },
+    {
+      "epoch": 52.04013377926422,
+      "grad_norm": 0.6744613647460938,
+      "learning_rate": 2.6651835372636262e-05,
+      "loss": 2.5294,
+      "step": 3902
+    },
+    {
+      "epoch": 52.05351170568562,
+      "grad_norm": 0.6527228951454163,
+      "learning_rate": 2.660734149054505e-05,
+      "loss": 2.5578,
+      "step": 3903
+    },
+    {
+      "epoch": 52.06688963210702,
+      "grad_norm": 0.6679868698120117,
+      "learning_rate": 2.656284760845384e-05,
+      "loss": 2.5264,
+      "step": 3904
+    },
+    {
+      "epoch": 52.080267558528426,
+      "grad_norm": 0.7045819759368896,
+      "learning_rate": 2.6518353726362626e-05,
+      "loss": 2.8701,
+      "step": 3905
+    },
+    {
+      "epoch": 52.09364548494983,
+      "grad_norm": 0.6314506530761719,
+      "learning_rate": 2.6473859844271415e-05,
+      "loss": 2.394,
+      "step": 3906
+    },
+    {
+      "epoch": 52.10702341137124,
+      "grad_norm": 0.681894838809967,
+      "learning_rate": 2.64293659621802e-05,
+      "loss": 2.2975,
+      "step": 3907
+    },
+    {
+      "epoch": 52.12040133779264,
+      "grad_norm": 0.6718135476112366,
+      "learning_rate": 2.638487208008899e-05,
+      "loss": 2.4091,
+      "step": 3908
+    },
+    {
+      "epoch": 52.13377926421405,
+      "grad_norm": 0.6556568145751953,
+      "learning_rate": 2.634037819799778e-05,
+      "loss": 2.652,
+      "step": 3909
+    },
+    {
+      "epoch": 52.147157190635454,
+      "grad_norm": 0.6845338940620422,
+      "learning_rate": 2.6295884315906565e-05,
+      "loss": 2.6104,
+      "step": 3910
+    },
+    {
+      "epoch": 52.16053511705686,
+      "grad_norm": 0.6780291199684143,
+      "learning_rate": 2.6251390433815354e-05,
+      "loss": 2.6291,
+      "step": 3911
+    },
+    {
+      "epoch": 52.17391304347826,
+      "grad_norm": 0.6501689553260803,
+      "learning_rate": 2.620689655172414e-05,
+      "loss": 2.4703,
+      "step": 3912
+    },
+    {
+      "epoch": 52.187290969899664,
+      "grad_norm": 0.6705746650695801,
+      "learning_rate": 2.616240266963293e-05,
+      "loss": 2.5792,
+      "step": 3913
+    },
+    {
+      "epoch": 52.20066889632107,
+      "grad_norm": 0.7104156017303467,
+      "learning_rate": 2.611790878754171e-05,
+      "loss": 2.6099,
+      "step": 3914
+    },
+    {
+      "epoch": 52.214046822742475,
+      "grad_norm": 0.6317254304885864,
+      "learning_rate": 2.60734149054505e-05,
+      "loss": 2.2911,
+      "step": 3915
+    },
+    {
+      "epoch": 52.22742474916388,
+      "grad_norm": 0.7213663458824158,
+      "learning_rate": 2.6028921023359286e-05,
+      "loss": 2.5813,
+      "step": 3916
+    },
+    {
+      "epoch": 52.240802675585286,
+      "grad_norm": 0.6910906434059143,
+      "learning_rate": 2.5984427141268076e-05,
+      "loss": 2.599,
+      "step": 3917
+    },
+    {
+      "epoch": 52.25418060200669,
+      "grad_norm": 0.7245139479637146,
+      "learning_rate": 2.593993325917686e-05,
+      "loss": 2.6023,
+      "step": 3918
+    },
+    {
+      "epoch": 52.2675585284281,
+      "grad_norm": 0.6673336029052734,
+      "learning_rate": 2.589543937708565e-05,
+      "loss": 2.5356,
+      "step": 3919
+    },
+    {
+      "epoch": 52.280936454849495,
+      "grad_norm": 0.6572391390800476,
+      "learning_rate": 2.585094549499444e-05,
+      "loss": 2.492,
+      "step": 3920
+    },
+    {
+      "epoch": 52.2943143812709,
+      "grad_norm": 0.6730433106422424,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 2.559,
+      "step": 3921
+    },
+    {
+      "epoch": 52.30769230769231,
+      "grad_norm": 0.699166476726532,
+      "learning_rate": 2.5761957730812015e-05,
+      "loss": 2.6597,
+      "step": 3922
+    },
+    {
+      "epoch": 52.32107023411371,
+      "grad_norm": 0.6620599031448364,
+      "learning_rate": 2.57174638487208e-05,
+      "loss": 2.5504,
+      "step": 3923
+    },
+    {
+      "epoch": 52.33444816053512,
+      "grad_norm": 0.7208895683288574,
+      "learning_rate": 2.567296996662959e-05,
+      "loss": 2.568,
+      "step": 3924
+    },
+    {
+      "epoch": 52.34782608695652,
+      "grad_norm": 0.7013632655143738,
+      "learning_rate": 2.562847608453838e-05,
+      "loss": 2.7051,
+      "step": 3925
+    },
+    {
+      "epoch": 52.36120401337793,
+      "grad_norm": 0.7046389579772949,
+      "learning_rate": 2.5583982202447165e-05,
+      "loss": 2.6715,
+      "step": 3926
+    },
+    {
+      "epoch": 52.374581939799334,
+      "grad_norm": 0.6850835084915161,
+      "learning_rate": 2.5539488320355954e-05,
+      "loss": 2.5796,
+      "step": 3927
+    },
+    {
+      "epoch": 52.38795986622073,
+      "grad_norm": 0.6935270428657532,
+      "learning_rate": 2.549499443826474e-05,
+      "loss": 2.3779,
+      "step": 3928
+    },
+    {
+      "epoch": 52.40133779264214,
+      "grad_norm": 0.659723699092865,
+      "learning_rate": 2.545050055617353e-05,
+      "loss": 2.3896,
+      "step": 3929
+    },
+    {
+      "epoch": 52.414715719063544,
+      "grad_norm": 0.6537664532661438,
+      "learning_rate": 2.5406006674082318e-05,
+      "loss": 2.3389,
+      "step": 3930
+    },
+    {
+      "epoch": 52.42809364548495,
+      "grad_norm": 0.6706110239028931,
+      "learning_rate": 2.5361512791991104e-05,
+      "loss": 2.5741,
+      "step": 3931
+    },
+    {
+      "epoch": 52.441471571906355,
+      "grad_norm": 0.6661157011985779,
+      "learning_rate": 2.5317018909899893e-05,
+      "loss": 2.4866,
+      "step": 3932
+    },
+    {
+      "epoch": 52.45484949832776,
+      "grad_norm": 0.6940733194351196,
+      "learning_rate": 2.527252502780868e-05,
+      "loss": 2.576,
+      "step": 3933
+    },
+    {
+      "epoch": 52.468227424749166,
+      "grad_norm": 0.6871435642242432,
+      "learning_rate": 2.522803114571746e-05,
+      "loss": 2.5895,
+      "step": 3934
+    },
+    {
+      "epoch": 52.48160535117057,
+      "grad_norm": 0.6974420547485352,
+      "learning_rate": 2.518353726362625e-05,
+      "loss": 2.6832,
+      "step": 3935
+    },
+    {
+      "epoch": 52.49498327759197,
+      "grad_norm": 0.6962250471115112,
+      "learning_rate": 2.513904338153504e-05,
+      "loss": 2.3006,
+      "step": 3936
+    },
+    {
+      "epoch": 52.508361204013376,
+      "grad_norm": 0.6863418817520142,
+      "learning_rate": 2.5094549499443825e-05,
+      "loss": 2.6124,
+      "step": 3937
+    },
+    {
+      "epoch": 52.52173913043478,
+      "grad_norm": 0.6994590163230896,
+      "learning_rate": 2.5050055617352614e-05,
+      "loss": 2.5151,
+      "step": 3938
+    },
+    {
+      "epoch": 52.53511705685619,
+      "grad_norm": 0.6534810662269592,
+      "learning_rate": 2.50055617352614e-05,
+      "loss": 2.6399,
+      "step": 3939
+    },
+    {
+      "epoch": 52.54849498327759,
+      "grad_norm": 0.6296973824501038,
+      "learning_rate": 2.496106785317019e-05,
+      "loss": 2.3892,
+      "step": 3940
+    },
+    {
+      "epoch": 52.561872909699,
+      "grad_norm": 0.6417976021766663,
+      "learning_rate": 2.491657397107898e-05,
+      "loss": 2.3044,
+      "step": 3941
+    },
+    {
+      "epoch": 52.575250836120404,
+      "grad_norm": 0.6997331976890564,
+      "learning_rate": 2.4872080088987764e-05,
+      "loss": 2.5604,
+      "step": 3942
+    },
+    {
+      "epoch": 52.58862876254181,
+      "grad_norm": 0.6694924235343933,
+      "learning_rate": 2.4827586206896553e-05,
+      "loss": 2.4752,
+      "step": 3943
+    },
+    {
+      "epoch": 52.60200668896321,
+      "grad_norm": 0.6598913669586182,
+      "learning_rate": 2.478309232480534e-05,
+      "loss": 2.5687,
+      "step": 3944
+    },
+    {
+      "epoch": 52.61538461538461,
+      "grad_norm": 0.7147283554077148,
+      "learning_rate": 2.4738598442714128e-05,
+      "loss": 2.5372,
+      "step": 3945
+    },
+    {
+      "epoch": 52.62876254180602,
+      "grad_norm": 0.6867512464523315,
+      "learning_rate": 2.4694104560622917e-05,
+      "loss": 2.6,
+      "step": 3946
+    },
+    {
+      "epoch": 52.642140468227424,
+      "grad_norm": 0.6952562928199768,
+      "learning_rate": 2.4649610678531703e-05,
+      "loss": 2.3294,
+      "step": 3947
+    },
+    {
+      "epoch": 52.65551839464883,
+      "grad_norm": 0.6790039539337158,
+      "learning_rate": 2.4605116796440492e-05,
+      "loss": 2.2903,
+      "step": 3948
+    },
+    {
+      "epoch": 52.668896321070235,
+      "grad_norm": 0.6904420852661133,
+      "learning_rate": 2.4560622914349278e-05,
+      "loss": 2.5636,
+      "step": 3949
+    },
+    {
+      "epoch": 52.68227424749164,
+      "grad_norm": 0.6746246218681335,
+      "learning_rate": 2.4516129032258064e-05,
+      "loss": 2.5318,
+      "step": 3950
+    },
+    {
+      "epoch": 52.69565217391305,
+      "grad_norm": 0.7066764235496521,
+      "learning_rate": 2.4471635150166853e-05,
+      "loss": 2.4879,
+      "step": 3951
+    },
+    {
+      "epoch": 52.709030100334445,
+      "grad_norm": 0.667173445224762,
+      "learning_rate": 2.442714126807564e-05,
+      "loss": 2.6719,
+      "step": 3952
+    },
+    {
+      "epoch": 52.72240802675585,
+      "grad_norm": 0.651569128036499,
+      "learning_rate": 2.4382647385984428e-05,
+      "loss": 2.5145,
+      "step": 3953
+    },
+    {
+      "epoch": 52.735785953177256,
+      "grad_norm": 0.6653866767883301,
+      "learning_rate": 2.4338153503893217e-05,
+      "loss": 2.647,
+      "step": 3954
+    },
+    {
+      "epoch": 52.74916387959866,
+      "grad_norm": 0.6874131560325623,
+      "learning_rate": 2.4293659621802003e-05,
+      "loss": 2.6872,
+      "step": 3955
+    },
+    {
+      "epoch": 52.76254180602007,
+      "grad_norm": 0.7200673222541809,
+      "learning_rate": 2.4249165739710792e-05,
+      "loss": 2.538,
+      "step": 3956
+    },
+    {
+      "epoch": 52.77591973244147,
+      "grad_norm": 0.6785191893577576,
+      "learning_rate": 2.4204671857619578e-05,
+      "loss": 2.7182,
+      "step": 3957
+    },
+    {
+      "epoch": 52.78929765886288,
+      "grad_norm": 0.6519473195075989,
+      "learning_rate": 2.4160177975528367e-05,
+      "loss": 2.3929,
+      "step": 3958
+    },
+    {
+      "epoch": 52.802675585284284,
+      "grad_norm": 0.7064288258552551,
+      "learning_rate": 2.4115684093437156e-05,
+      "loss": 2.6197,
+      "step": 3959
+    },
+    {
+      "epoch": 52.81605351170568,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 2.407119021134594e-05,
+      "loss": 2.5221,
+      "step": 3960
+    },
+    {
+      "epoch": 52.82943143812709,
+      "grad_norm": 0.7092622518539429,
+      "learning_rate": 2.4026696329254728e-05,
+      "loss": 2.6919,
+      "step": 3961
+    },
+    {
+      "epoch": 52.84280936454849,
+      "grad_norm": 0.6915587782859802,
+      "learning_rate": 2.3982202447163517e-05,
+      "loss": 2.627,
+      "step": 3962
+    },
+    {
+      "epoch": 52.8561872909699,
+      "grad_norm": 0.6888433694839478,
+      "learning_rate": 2.3937708565072303e-05,
+      "loss": 2.7742,
+      "step": 3963
+    },
+    {
+      "epoch": 52.869565217391305,
+      "grad_norm": 0.6663147211074829,
+      "learning_rate": 2.3893214682981092e-05,
+      "loss": 2.2218,
+      "step": 3964
+    },
+    {
+      "epoch": 52.88294314381271,
+      "grad_norm": 0.6786331534385681,
+      "learning_rate": 2.3848720800889878e-05,
+      "loss": 2.6369,
+      "step": 3965
+    },
+    {
+      "epoch": 52.896321070234116,
+      "grad_norm": 0.6943885087966919,
+      "learning_rate": 2.3804226918798667e-05,
+      "loss": 2.464,
+      "step": 3966
+    },
+    {
+      "epoch": 52.90969899665552,
+      "grad_norm": 0.7094883322715759,
+      "learning_rate": 2.3759733036707456e-05,
+      "loss": 2.801,
+      "step": 3967
+    },
+    {
+      "epoch": 52.92307692307692,
+      "grad_norm": 0.6632527709007263,
+      "learning_rate": 2.3715239154616242e-05,
+      "loss": 2.5922,
+      "step": 3968
+    },
+    {
+      "epoch": 52.936454849498325,
+      "grad_norm": 0.7414517998695374,
+      "learning_rate": 2.367074527252503e-05,
+      "loss": 2.9419,
+      "step": 3969
+    },
+    {
+      "epoch": 52.94983277591973,
+      "grad_norm": 0.6803048849105835,
+      "learning_rate": 2.3626251390433817e-05,
+      "loss": 2.5849,
+      "step": 3970
+    },
+    {
+      "epoch": 52.96321070234114,
+      "grad_norm": 0.6940471529960632,
+      "learning_rate": 2.3581757508342603e-05,
+      "loss": 2.7496,
+      "step": 3971
+    },
+    {
+      "epoch": 52.97658862876254,
+      "grad_norm": 0.6914702653884888,
+      "learning_rate": 2.353726362625139e-05,
+      "loss": 2.5185,
+      "step": 3972
+    },
+    {
+      "epoch": 52.98996655518395,
+      "grad_norm": 0.6751214265823364,
+      "learning_rate": 2.3492769744160177e-05,
+      "loss": 2.4305,
+      "step": 3973
+    },
+    {
+      "epoch": 53.0,
+      "grad_norm": 0.831278920173645,
+      "learning_rate": 2.3448275862068967e-05,
+      "loss": 2.3934,
+      "step": 3974
+    },
+    {
+      "epoch": 53.013377926421406,
+      "grad_norm": 0.662552535533905,
+      "learning_rate": 2.3403781979977756e-05,
+      "loss": 2.388,
+      "step": 3975
+    },
+    {
+      "epoch": 53.02675585284281,
+      "grad_norm": 0.7117273211479187,
+      "learning_rate": 2.335928809788654e-05,
+      "loss": 2.6864,
+      "step": 3976
+    },
+    {
+      "epoch": 53.04013377926422,
+      "grad_norm": 0.623512327671051,
+      "learning_rate": 2.331479421579533e-05,
+      "loss": 2.3419,
+      "step": 3977
+    },
+    {
+      "epoch": 53.05351170568562,
+      "grad_norm": 0.6845181584358215,
+      "learning_rate": 2.3270300333704117e-05,
+      "loss": 2.5291,
+      "step": 3978
+    },
+    {
+      "epoch": 53.06688963210702,
+      "grad_norm": 0.6794723272323608,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 2.6106,
+      "step": 3979
+    },
+    {
+      "epoch": 53.080267558528426,
+      "grad_norm": 0.6423758864402771,
+      "learning_rate": 2.3181312569521695e-05,
+      "loss": 2.5044,
+      "step": 3980
+    },
+    {
+      "epoch": 53.09364548494983,
+      "grad_norm": 0.6895222663879395,
+      "learning_rate": 2.3136818687430477e-05,
+      "loss": 2.5586,
+      "step": 3981
+    },
+    {
+      "epoch": 53.10702341137124,
+      "grad_norm": 0.711919903755188,
+      "learning_rate": 2.3092324805339266e-05,
+      "loss": 2.6076,
+      "step": 3982
+    },
+    {
+      "epoch": 53.12040133779264,
+      "grad_norm": 0.6487022042274475,
+      "learning_rate": 2.3047830923248052e-05,
+      "loss": 2.4783,
+      "step": 3983
+    },
+    {
+      "epoch": 53.13377926421405,
+      "grad_norm": 0.664364218711853,
+      "learning_rate": 2.300333704115684e-05,
+      "loss": 2.4453,
+      "step": 3984
+    },
+    {
+      "epoch": 53.147157190635454,
+      "grad_norm": 0.672069251537323,
+      "learning_rate": 2.295884315906563e-05,
+      "loss": 2.5366,
+      "step": 3985
+    },
+    {
+      "epoch": 53.16053511705686,
+      "grad_norm": 0.724062979221344,
+      "learning_rate": 2.2914349276974416e-05,
+      "loss": 2.6009,
+      "step": 3986
+    },
+    {
+      "epoch": 53.17391304347826,
+      "grad_norm": 0.660426914691925,
+      "learning_rate": 2.2869855394883205e-05,
+      "loss": 2.4559,
+      "step": 3987
+    },
+    {
+      "epoch": 53.187290969899664,
+      "grad_norm": 0.6827124357223511,
+      "learning_rate": 2.282536151279199e-05,
+      "loss": 2.2968,
+      "step": 3988
+    },
+    {
+      "epoch": 53.20066889632107,
+      "grad_norm": 0.6487954258918762,
+      "learning_rate": 2.278086763070078e-05,
+      "loss": 2.4694,
+      "step": 3989
+    },
+    {
+      "epoch": 53.214046822742475,
+      "grad_norm": 0.7343469262123108,
+      "learning_rate": 2.273637374860957e-05,
+      "loss": 2.584,
+      "step": 3990
+    },
+    {
+      "epoch": 53.22742474916388,
+      "grad_norm": 0.6878166198730469,
+      "learning_rate": 2.2691879866518352e-05,
+      "loss": 2.6855,
+      "step": 3991
+    },
+    {
+      "epoch": 53.240802675585286,
+      "grad_norm": 0.69071364402771,
+      "learning_rate": 2.264738598442714e-05,
+      "loss": 2.579,
+      "step": 3992
+    },
+    {
+      "epoch": 53.25418060200669,
+      "grad_norm": 0.6821252107620239,
+      "learning_rate": 2.260289210233593e-05,
+      "loss": 2.6543,
+      "step": 3993
+    },
+    {
+      "epoch": 53.2675585284281,
+      "grad_norm": 0.6615633964538574,
+      "learning_rate": 2.2558398220244716e-05,
+      "loss": 2.3427,
+      "step": 3994
+    },
+    {
+      "epoch": 53.280936454849495,
+      "grad_norm": 0.641324520111084,
+      "learning_rate": 2.2513904338153505e-05,
+      "loss": 2.479,
+      "step": 3995
+    },
+    {
+      "epoch": 53.2943143812709,
+      "grad_norm": 0.7013201117515564,
+      "learning_rate": 2.246941045606229e-05,
+      "loss": 2.6016,
+      "step": 3996
+    },
+    {
+      "epoch": 53.30769230769231,
+      "grad_norm": 0.7007108330726624,
+      "learning_rate": 2.242491657397108e-05,
+      "loss": 2.6184,
+      "step": 3997
+    },
+    {
+      "epoch": 53.32107023411371,
+      "grad_norm": 0.6686177849769592,
+      "learning_rate": 2.238042269187987e-05,
+      "loss": 2.6286,
+      "step": 3998
+    },
+    {
+      "epoch": 53.33444816053512,
+      "grad_norm": 0.6843096017837524,
+      "learning_rate": 2.2335928809788655e-05,
+      "loss": 2.4684,
+      "step": 3999
+    },
+    {
+      "epoch": 53.34782608695652,
+      "grad_norm": 0.66323322057724,
+      "learning_rate": 2.2291434927697444e-05,
+      "loss": 2.4277,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 60,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2380878140440832e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}