diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 28.19383259911894,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.007048458149779736,
+      "grad_norm": 10.376355533994332,
+      "learning_rate": 9.433962264150943e-07,
+      "loss": 0.9269,
+      "step": 1
+    },
+    {
+      "epoch": 0.014096916299559472,
+      "grad_norm": 10.62478530312751,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 0.9428,
+      "step": 2
+    },
+    {
+      "epoch": 0.021145374449339206,
+      "grad_norm": 10.04196037555283,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.9398,
+      "step": 3
+    },
+    {
+      "epoch": 0.028193832599118944,
+      "grad_norm": 10.300193060087278,
+      "learning_rate": 3.7735849056603773e-06,
+      "loss": 0.8982,
+      "step": 4
+    },
+    {
+      "epoch": 0.03524229074889868,
+      "grad_norm": 8.713019926971189,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.907,
+      "step": 5
+    },
+    {
+      "epoch": 0.04229074889867841,
+      "grad_norm": 5.574211384916085,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.8436,
+      "step": 6
+    },
+    {
+      "epoch": 0.04933920704845815,
+      "grad_norm": 5.1209720272304065,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.8553,
+      "step": 7
+    },
+    {
+      "epoch": 0.05638766519823789,
+      "grad_norm": 1.9289096235595913,
+      "learning_rate": 7.547169811320755e-06,
+      "loss": 0.7665,
+      "step": 8
+    },
+    {
+      "epoch": 0.06343612334801763,
+      "grad_norm": 1.6021043210586539,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.7413,
+      "step": 9
+    },
+    {
+      "epoch": 0.07048458149779736,
+      "grad_norm": 0.9226010595974447,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.7562,
+      "step": 10
+    },
+    {
+      "epoch": 0.07753303964757709,
+      "grad_norm": 0.656331750439419,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.7089,
+      "step": 11
+    },
+    {
+      "epoch": 0.08458149779735682,
+      "grad_norm": 0.6758156489180911,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.7195,
+      "step": 12
+    },
+    {
+      "epoch": 0.09162995594713656,
+      "grad_norm": 0.6291551280893382,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.7197,
+      "step": 13
+    },
+    {
+      "epoch": 0.0986784140969163,
+      "grad_norm": 0.5459665752714382,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.706,
+      "step": 14
+    },
+    {
+      "epoch": 0.10572687224669604,
+      "grad_norm": 0.5940481697170278,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.7369,
+      "step": 15
+    },
+    {
+      "epoch": 0.11277533039647578,
+      "grad_norm": 0.4858384536269203,
+      "learning_rate": 1.509433962264151e-05,
+      "loss": 0.6852,
+      "step": 16
+    },
+    {
+      "epoch": 0.1198237885462555,
+      "grad_norm": 0.4408356838565912,
+      "learning_rate": 1.6037735849056604e-05,
+      "loss": 0.6706,
+      "step": 17
+    },
+    {
+      "epoch": 0.12687224669603525,
+      "grad_norm": 0.3977947341971036,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.6618,
+      "step": 18
+    },
+    {
+      "epoch": 0.13392070484581498,
+      "grad_norm": 0.338446839739168,
+      "learning_rate": 1.7924528301886792e-05,
+      "loss": 0.6472,
+      "step": 19
+    },
+    {
+      "epoch": 0.14096916299559473,
+      "grad_norm": 0.3329072502860991,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.6166,
+      "step": 20
+    },
+    {
+      "epoch": 0.14801762114537445,
+      "grad_norm": 0.3491696264909759,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.6392,
+      "step": 21
+    },
+    {
+      "epoch": 0.15506607929515417,
+      "grad_norm": 0.26467989173806666,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.6037,
+      "step": 22
+    },
+    {
+      "epoch": 0.16211453744493393,
+      "grad_norm": 0.3267510775597588,
+      "learning_rate": 2.1698113207547172e-05,
+      "loss": 0.6112,
+      "step": 23
+    },
+    {
+      "epoch": 0.16916299559471365,
+      "grad_norm": 0.25127583034295475,
+      "learning_rate": 2.2641509433962265e-05,
+      "loss": 0.6014,
+      "step": 24
+    },
+    {
+      "epoch": 0.1762114537444934,
+      "grad_norm": 0.2417544924753575,
+      "learning_rate": 2.358490566037736e-05,
+      "loss": 0.5704,
+      "step": 25
+    },
+    {
+      "epoch": 0.18325991189427313,
+      "grad_norm": 0.2557044912994469,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 0.5755,
+      "step": 26
+    },
+    {
+      "epoch": 0.19030837004405288,
+      "grad_norm": 0.2672662997795438,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.5739,
+      "step": 27
+    },
+    {
+      "epoch": 0.1973568281938326,
+      "grad_norm": 0.25772584187211406,
+      "learning_rate": 2.641509433962264e-05,
+      "loss": 0.5464,
+      "step": 28
+    },
+    {
+      "epoch": 0.20440528634361232,
+      "grad_norm": 0.26260430781608146,
+      "learning_rate": 2.7358490566037738e-05,
+      "loss": 0.5416,
+      "step": 29
+    },
+    {
+      "epoch": 0.21145374449339208,
+      "grad_norm": 0.23403640804662373,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.5252,
+      "step": 30
+    },
+    {
+      "epoch": 0.2185022026431718,
+      "grad_norm": 0.2783582936088169,
+      "learning_rate": 2.9245283018867926e-05,
+      "loss": 0.5395,
+      "step": 31
+    },
+    {
+      "epoch": 0.22555066079295155,
+      "grad_norm": 0.304992925248862,
+      "learning_rate": 3.018867924528302e-05,
+      "loss": 0.52,
+      "step": 32
+    },
+    {
+      "epoch": 0.23259911894273128,
+      "grad_norm": 0.19345490536029789,
+      "learning_rate": 3.113207547169811e-05,
+      "loss": 0.5019,
+      "step": 33
+    },
+    {
+      "epoch": 0.239647577092511,
+      "grad_norm": 0.2168924901990453,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.5066,
+      "step": 34
+    },
+    {
+      "epoch": 0.24669603524229075,
+      "grad_norm": 0.2124323055104462,
+      "learning_rate": 3.30188679245283e-05,
+      "loss": 0.5059,
+      "step": 35
+    },
+    {
+      "epoch": 0.2537444933920705,
+      "grad_norm": 0.12496598272690891,
+      "learning_rate": 3.39622641509434e-05,
+      "loss": 0.4727,
+      "step": 36
+    },
+    {
+      "epoch": 0.2607929515418502,
+      "grad_norm": 0.12115432406389982,
+      "learning_rate": 3.490566037735849e-05,
+      "loss": 0.4845,
+      "step": 37
+    },
+    {
+      "epoch": 0.26784140969162995,
+      "grad_norm": 0.12106408877828782,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.5023,
+      "step": 38
+    },
+    {
+      "epoch": 0.2748898678414097,
+      "grad_norm": 0.12615454375311133,
+      "learning_rate": 3.679245283018868e-05,
+      "loss": 0.4919,
+      "step": 39
+    },
+    {
+      "epoch": 0.28193832599118945,
+      "grad_norm": 0.11954583018400289,
+      "learning_rate": 3.7735849056603776e-05,
+      "loss": 0.4853,
+      "step": 40
+    },
+    {
+      "epoch": 0.2889867841409692,
+      "grad_norm": 0.10729651012304586,
+      "learning_rate": 3.867924528301887e-05,
+      "loss": 0.4741,
+      "step": 41
+    },
+    {
+      "epoch": 0.2960352422907489,
+      "grad_norm": 0.11582934234037676,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.4921,
+      "step": 42
+    },
+    {
+      "epoch": 0.3030837004405286,
+      "grad_norm": 0.13310683217077324,
+      "learning_rate": 4.0566037735849064e-05,
+      "loss": 0.473,
+      "step": 43
+    },
+    {
+      "epoch": 0.31013215859030835,
+      "grad_norm": 0.10444189074434215,
+      "learning_rate": 4.150943396226415e-05,
+      "loss": 0.4874,
+      "step": 44
+    },
+    {
+      "epoch": 0.31718061674008813,
+      "grad_norm": 0.094402909325761,
+      "learning_rate": 4.245283018867925e-05,
+      "loss": 0.4827,
+      "step": 45
+    },
+    {
+      "epoch": 0.32422907488986785,
+      "grad_norm": 0.10731652747494196,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.4917,
+      "step": 46
+    },
+    {
+      "epoch": 0.3312775330396476,
+      "grad_norm": 0.11519686559079342,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.4687,
+      "step": 47
+    },
+    {
+      "epoch": 0.3383259911894273,
+      "grad_norm": 0.1011606580739382,
+      "learning_rate": 4.528301886792453e-05,
+      "loss": 0.4761,
+      "step": 48
+    },
+    {
+      "epoch": 0.345374449339207,
+      "grad_norm": 0.10327995394194993,
+      "learning_rate": 4.6226415094339625e-05,
+      "loss": 0.4546,
+      "step": 49
+    },
+    {
+      "epoch": 0.3524229074889868,
+      "grad_norm": 0.10038474115943215,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.4651,
+      "step": 50
+    },
+    {
+      "epoch": 0.3594713656387665,
+      "grad_norm": 0.10648724148237948,
+      "learning_rate": 4.811320754716982e-05,
+      "loss": 0.4673,
+      "step": 51
+    },
+    {
+      "epoch": 0.36651982378854625,
+      "grad_norm": 0.10005404914407207,
+      "learning_rate": 4.9056603773584906e-05,
+      "loss": 0.4656,
+      "step": 52
+    },
+    {
+      "epoch": 0.373568281938326,
+      "grad_norm": 0.10160965353709094,
+      "learning_rate": 5e-05,
+      "loss": 0.4633,
+      "step": 53
+    },
+    {
+      "epoch": 0.38061674008810575,
+      "grad_norm": 0.11127831690687728,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.4718,
+      "step": 54
+    },
+    {
+      "epoch": 0.3876651982378855,
+      "grad_norm": 0.11175059823779314,
+      "learning_rate": 5.188679245283019e-05,
+      "loss": 0.4552,
+      "step": 55
+    },
+    {
+      "epoch": 0.3947136563876652,
+      "grad_norm": 0.11580156437808585,
+      "learning_rate": 5.283018867924528e-05,
+      "loss": 0.4886,
+      "step": 56
+    },
+    {
+      "epoch": 0.4017621145374449,
+      "grad_norm": 0.12382315469226232,
+      "learning_rate": 5.377358490566038e-05,
+      "loss": 0.4882,
+      "step": 57
+    },
+    {
+      "epoch": 0.40881057268722465,
+      "grad_norm": 0.11578442675304722,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.4633,
+      "step": 58
+    },
+    {
+      "epoch": 0.41585903083700443,
+      "grad_norm": 0.12156514306686417,
+      "learning_rate": 5.5660377358490564e-05,
+      "loss": 0.4522,
+      "step": 59
+    },
+    {
+      "epoch": 0.42290748898678415,
+      "grad_norm": 0.13000194667343104,
+      "learning_rate": 5.660377358490566e-05,
+      "loss": 0.4758,
+      "step": 60
+    },
+    {
+      "epoch": 0.4299559471365639,
+      "grad_norm": 0.11843557329493942,
+      "learning_rate": 5.7547169811320756e-05,
+      "loss": 0.4599,
+      "step": 61
+    },
+    {
+      "epoch": 0.4370044052863436,
+      "grad_norm": 0.11443637519005759,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.4637,
+      "step": 62
+    },
+    {
+      "epoch": 0.4440528634361233,
+      "grad_norm": 0.11929372861301925,
+      "learning_rate": 5.943396226415094e-05,
+      "loss": 0.4661,
+      "step": 63
+    },
+    {
+      "epoch": 0.4511013215859031,
+      "grad_norm": 0.12700506549300355,
+      "learning_rate": 6.037735849056604e-05,
+      "loss": 0.4495,
+      "step": 64
+    },
+    {
+      "epoch": 0.4581497797356828,
+      "grad_norm": 0.12655765485728668,
+      "learning_rate": 6.132075471698113e-05,
+      "loss": 0.4463,
+      "step": 65
+    },
+    {
+      "epoch": 0.46519823788546255,
+      "grad_norm": 0.11595831082573389,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.4479,
+      "step": 66
+    },
+    {
+      "epoch": 0.4722466960352423,
+      "grad_norm": 0.12944259828464236,
+      "learning_rate": 6.320754716981132e-05,
+      "loss": 0.4622,
+      "step": 67
+    },
+    {
+      "epoch": 0.479295154185022,
+      "grad_norm": 0.1257922742866891,
+      "learning_rate": 6.415094339622641e-05,
+      "loss": 0.4516,
+      "step": 68
+    },
+    {
+      "epoch": 0.4863436123348018,
+      "grad_norm": 0.12430481584821111,
+      "learning_rate": 6.50943396226415e-05,
+      "loss": 0.4499,
+      "step": 69
+    },
+    {
+      "epoch": 0.4933920704845815,
+      "grad_norm": 0.12179427795348166,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.458,
+      "step": 70
+    },
+    {
+      "epoch": 0.5004405286343613,
+      "grad_norm": 0.13591498604965535,
+      "learning_rate": 6.69811320754717e-05,
+      "loss": 0.4398,
+      "step": 71
+    },
+    {
+      "epoch": 0.507488986784141,
+      "grad_norm": 0.11836461696811945,
+      "learning_rate": 6.79245283018868e-05,
+      "loss": 0.4388,
+      "step": 72
+    },
+    {
+      "epoch": 0.5145374449339207,
+      "grad_norm": 0.14655293057809665,
+      "learning_rate": 6.886792452830189e-05,
+      "loss": 0.4462,
+      "step": 73
+    },
+    {
+      "epoch": 0.5215859030837005,
+      "grad_norm": 0.13708967218938622,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.4511,
+      "step": 74
+    },
+    {
+      "epoch": 0.5286343612334802,
+      "grad_norm": 0.11945060428755523,
+      "learning_rate": 7.075471698113208e-05,
+      "loss": 0.449,
+      "step": 75
+    },
+    {
+      "epoch": 0.5356828193832599,
+      "grad_norm": 0.12771541926682178,
+      "learning_rate": 7.169811320754717e-05,
+      "loss": 0.4386,
+      "step": 76
+    },
+    {
+      "epoch": 0.5427312775330396,
+      "grad_norm": 0.12480959145159178,
+      "learning_rate": 7.264150943396226e-05,
+      "loss": 0.4361,
+      "step": 77
+    },
+    {
+      "epoch": 0.5497797356828193,
+      "grad_norm": 0.12168847176852518,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.4442,
+      "step": 78
+    },
+    {
+      "epoch": 0.5568281938325991,
+      "grad_norm": 0.12066725687608243,
+      "learning_rate": 7.452830188679245e-05,
+      "loss": 0.4397,
+      "step": 79
+    },
+    {
+      "epoch": 0.5638766519823789,
+      "grad_norm": 0.12222965678103333,
+      "learning_rate": 7.547169811320755e-05,
+      "loss": 0.4455,
+      "step": 80
+    },
+    {
+      "epoch": 0.5709251101321586,
+      "grad_norm": 0.11942181371835363,
+      "learning_rate": 7.641509433962265e-05,
+      "loss": 0.44,
+      "step": 81
+    },
+    {
+      "epoch": 0.5779735682819384,
+      "grad_norm": 0.1424339130015169,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.4394,
+      "step": 82
+    },
+    {
+      "epoch": 0.5850220264317181,
+      "grad_norm": 0.12884811893333864,
+      "learning_rate": 7.830188679245283e-05,
+      "loss": 0.4354,
+      "step": 83
+    },
+    {
+      "epoch": 0.5920704845814978,
+      "grad_norm": 0.1599761797342499,
+      "learning_rate": 7.924528301886794e-05,
+      "loss": 0.4346,
+      "step": 84
+    },
+    {
+      "epoch": 0.5991189427312775,
+      "grad_norm": 0.1345019736198752,
+      "learning_rate": 8.018867924528302e-05,
+      "loss": 0.4228,
+      "step": 85
+    },
+    {
+      "epoch": 0.6061674008810573,
+      "grad_norm": 0.1307231063751821,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.4172,
+      "step": 86
+    },
+    {
+      "epoch": 0.613215859030837,
+      "grad_norm": 0.13956773844485437,
+      "learning_rate": 8.207547169811322e-05,
+      "loss": 0.4314,
+      "step": 87
+    },
+    {
+      "epoch": 0.6202643171806167,
+      "grad_norm": 0.16113452433023726,
+      "learning_rate": 8.30188679245283e-05,
+      "loss": 0.4245,
+      "step": 88
+    },
+    {
+      "epoch": 0.6273127753303964,
+      "grad_norm": 0.14981202603419433,
+      "learning_rate": 8.396226415094341e-05,
+      "loss": 0.4292,
+      "step": 89
+    },
+    {
+      "epoch": 0.6343612334801763,
+      "grad_norm": 0.15954035758904872,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.4379,
+      "step": 90
+    },
+    {
+      "epoch": 0.641409691629956,
+      "grad_norm": 0.1526363497973273,
+      "learning_rate": 8.584905660377359e-05,
+      "loss": 0.4292,
+      "step": 91
+    },
+    {
+      "epoch": 0.6484581497797357,
+      "grad_norm": 0.16353511278070604,
+      "learning_rate": 8.679245283018869e-05,
+      "loss": 0.4319,
+      "step": 92
+    },
+    {
+      "epoch": 0.6555066079295154,
+      "grad_norm": 0.1343484225464562,
+      "learning_rate": 8.773584905660378e-05,
+      "loss": 0.4236,
+      "step": 93
+    },
+    {
+      "epoch": 0.6625550660792952,
+      "grad_norm": 0.1544345307577561,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.4219,
+      "step": 94
+    },
+    {
+      "epoch": 0.6696035242290749,
+      "grad_norm": 0.16529298661433087,
+      "learning_rate": 8.962264150943397e-05,
+      "loss": 0.4285,
+      "step": 95
+    },
+    {
+      "epoch": 0.6766519823788546,
+      "grad_norm": 0.1521904587248344,
+      "learning_rate": 9.056603773584906e-05,
+      "loss": 0.4231,
+      "step": 96
+    },
+    {
+      "epoch": 0.6837004405286343,
+      "grad_norm": 0.15084481828818738,
+      "learning_rate": 9.150943396226416e-05,
+      "loss": 0.4296,
+      "step": 97
+    },
+    {
+      "epoch": 0.690748898678414,
+      "grad_norm": 0.17464219830644734,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.4215,
+      "step": 98
+    },
+    {
+      "epoch": 0.6977973568281939,
+      "grad_norm": 0.13313486780304176,
+      "learning_rate": 9.339622641509434e-05,
+      "loss": 0.4178,
+      "step": 99
+    },
+    {
+      "epoch": 0.7048458149779736,
+      "grad_norm": 0.20036813062472839,
+      "learning_rate": 9.433962264150944e-05,
+      "loss": 0.4317,
+      "step": 100
+    },
+    {
+      "epoch": 0.7118942731277533,
+      "grad_norm": 0.3231493050203587,
+      "learning_rate": 9.528301886792453e-05,
+      "loss": 0.4234,
+      "step": 101
+    },
+    {
+      "epoch": 0.718942731277533,
+      "grad_norm": 0.25414170188994517,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.4314,
+      "step": 102
+    },
+    {
+      "epoch": 0.7259911894273128,
+      "grad_norm": 0.15893218430285017,
+      "learning_rate": 9.716981132075472e-05,
+      "loss": 0.4088,
+      "step": 103
+    },
+    {
+      "epoch": 0.7330396475770925,
+      "grad_norm": 0.17297630338335662,
+      "learning_rate": 9.811320754716981e-05,
+      "loss": 0.4096,
+      "step": 104
+    },
+    {
+      "epoch": 0.7400881057268722,
+      "grad_norm": 0.1778908274631403,
+      "learning_rate": 9.905660377358492e-05,
+      "loss": 0.409,
+      "step": 105
+    },
+    {
+      "epoch": 0.747136563876652,
+      "grad_norm": 0.15201007166453892,
+      "learning_rate": 0.0001,
+      "loss": 0.4172,
+      "step": 106
+    },
+    {
+      "epoch": 0.7541850220264317,
+      "grad_norm": 0.18855896413423692,
+      "learning_rate": 0.00010094339622641511,
+      "loss": 0.4344,
+      "step": 107
+    },
+    {
+      "epoch": 0.7612334801762115,
+      "grad_norm": 0.20688658145786876,
+      "learning_rate": 0.0001018867924528302,
+      "loss": 0.4191,
+      "step": 108
+    },
+    {
+      "epoch": 0.7682819383259912,
+      "grad_norm": 0.20814857190559022,
+      "learning_rate": 0.0001028301886792453,
+      "loss": 0.4164,
+      "step": 109
+    },
+    {
+      "epoch": 0.775330396475771,
+      "grad_norm": 0.16890233302893676,
+      "learning_rate": 0.00010377358490566037,
+      "loss": 0.4131,
+      "step": 110
+    },
+    {
+      "epoch": 0.7823788546255507,
+      "grad_norm": 0.23281920040705448,
+      "learning_rate": 0.00010471698113207549,
+      "loss": 0.4153,
+      "step": 111
+    },
+    {
+      "epoch": 0.7894273127753304,
+      "grad_norm": 0.1673341479623689,
+      "learning_rate": 0.00010566037735849057,
+      "loss": 0.4131,
+      "step": 112
+    },
+    {
+      "epoch": 0.7964757709251101,
+      "grad_norm": 0.2133000194676935,
+      "learning_rate": 0.00010660377358490567,
+      "loss": 0.4115,
+      "step": 113
+    },
+    {
+      "epoch": 0.8035242290748899,
+      "grad_norm": 0.2698971116698262,
+      "learning_rate": 0.00010754716981132076,
+      "loss": 0.4127,
+      "step": 114
+    },
+    {
+      "epoch": 0.8105726872246696,
+      "grad_norm": 0.15160885960067624,
+      "learning_rate": 0.00010849056603773586,
+      "loss": 0.4205,
+      "step": 115
+    },
+    {
+      "epoch": 0.8176211453744493,
+      "grad_norm": 0.2594815468068123,
+      "learning_rate": 0.00010943396226415095,
+      "loss": 0.4111,
+      "step": 116
+    },
+    {
+      "epoch": 0.824669603524229,
+      "grad_norm": 0.21808893449959363,
+      "learning_rate": 0.00011037735849056605,
+      "loss": 0.4107,
+      "step": 117
+    },
+    {
+      "epoch": 0.8317180616740089,
+      "grad_norm": 0.1481411878267263,
+      "learning_rate": 0.00011132075471698113,
+      "loss": 0.3991,
+      "step": 118
+    },
+    {
+      "epoch": 0.8387665198237886,
+      "grad_norm": 0.19539130849106487,
+      "learning_rate": 0.00011226415094339624,
+      "loss": 0.4141,
+      "step": 119
+    },
+    {
+      "epoch": 0.8458149779735683,
+      "grad_norm": 0.2719447884569931,
+      "learning_rate": 0.00011320754716981132,
+      "loss": 0.4158,
+      "step": 120
+    },
+    {
+      "epoch": 0.852863436123348,
+      "grad_norm": 0.1861897688923645,
+      "learning_rate": 0.00011415094339622642,
+      "loss": 0.4141,
+      "step": 121
+    },
+    {
+      "epoch": 0.8599118942731278,
+      "grad_norm": 0.18137601372494408,
+      "learning_rate": 0.00011509433962264151,
+      "loss": 0.419,
+      "step": 122
+    },
+    {
+      "epoch": 0.8669603524229075,
+      "grad_norm": 0.17018304076099408,
+      "learning_rate": 0.00011603773584905662,
+      "loss": 0.3931,
+      "step": 123
+    },
+    {
+      "epoch": 0.8740088105726872,
+      "grad_norm": 0.15448687558435517,
+      "learning_rate": 0.0001169811320754717,
+      "loss": 0.406,
+      "step": 124
+    },
+    {
+      "epoch": 0.8810572687224669,
+      "grad_norm": 0.1603014095733474,
+      "learning_rate": 0.00011792452830188681,
+      "loss": 0.3899,
+      "step": 125
+    },
+    {
+      "epoch": 0.8881057268722466,
+      "grad_norm": 0.19978005726529252,
+      "learning_rate": 0.00011886792452830188,
+      "loss": 0.4122,
+      "step": 126
+    },
+    {
+      "epoch": 0.8951541850220265,
+      "grad_norm": 0.29277294473264837,
+      "learning_rate": 0.000119811320754717,
+      "loss": 0.4128,
+      "step": 127
+    },
+    {
+      "epoch": 0.9022026431718062,
+      "grad_norm": 0.2944542550974979,
+      "learning_rate": 0.00012075471698113207,
+      "loss": 0.4066,
+      "step": 128
+    },
+    {
+      "epoch": 0.9092511013215859,
+      "grad_norm": 0.16486407568846756,
+      "learning_rate": 0.00012169811320754718,
+      "loss": 0.4172,
+      "step": 129
+    },
+    {
+      "epoch": 0.9162995594713657,
+      "grad_norm": 0.1948207568018318,
+      "learning_rate": 0.00012264150943396227,
+      "loss": 0.4096,
+      "step": 130
+    },
+    {
+      "epoch": 0.9233480176211454,
+      "grad_norm": 0.16514685291379108,
+      "learning_rate": 0.00012358490566037738,
+      "loss": 0.3945,
+      "step": 131
+    },
+    {
+      "epoch": 0.9303964757709251,
+      "grad_norm": 0.2639282918079393,
+      "learning_rate": 0.00012452830188679244,
+      "loss": 0.41,
+      "step": 132
+    },
+    {
+      "epoch": 0.9374449339207048,
+      "grad_norm": 0.28621356600867337,
+      "learning_rate": 0.00012547169811320756,
+      "loss": 0.409,
+      "step": 133
+    },
+    {
+      "epoch": 0.9444933920704845,
+      "grad_norm": 0.20446846483241798,
+      "learning_rate": 0.00012641509433962265,
+      "loss": 0.4074,
+      "step": 134
+    },
+    {
+      "epoch": 0.9515418502202643,
+      "grad_norm": 0.2039626057374481,
+      "learning_rate": 0.00012735849056603774,
+      "loss": 0.4024,
+      "step": 135
+    },
+    {
+      "epoch": 0.958590308370044,
+      "grad_norm": 0.16218719452039437,
+      "learning_rate": 0.00012830188679245283,
+      "loss": 0.4195,
+      "step": 136
+    },
+    {
+      "epoch": 0.9656387665198238,
+      "grad_norm": 0.2299628940847805,
+      "learning_rate": 0.00012924528301886794,
+      "loss": 0.4082,
+      "step": 137
+    },
+    {
+      "epoch": 0.9726872246696036,
+      "grad_norm": 0.19230688116227387,
+      "learning_rate": 0.000130188679245283,
+      "loss": 0.4085,
+      "step": 138
+    },
+    {
+      "epoch": 0.9797356828193833,
+      "grad_norm": 0.20247846928896215,
+      "learning_rate": 0.00013113207547169812,
+      "loss": 0.4185,
+      "step": 139
+    },
+    {
+      "epoch": 0.986784140969163,
+      "grad_norm": 0.18379665210208151,
+      "learning_rate": 0.0001320754716981132,
+      "loss": 0.3995,
+      "step": 140
+    },
+    {
+      "epoch": 0.9938325991189427,
+      "grad_norm": 0.18220627997660008,
+      "learning_rate": 0.0001330188679245283,
+      "loss": 0.4022,
+      "step": 141
+    },
+    {
+      "epoch": 1.0008810572687226,
+      "grad_norm": 0.24543974361593682,
+      "learning_rate": 0.0001339622641509434,
+      "loss": 0.4105,
+      "step": 142
+    },
+    {
+      "epoch": 1.0079295154185022,
+      "grad_norm": 0.14521115911857854,
+      "learning_rate": 0.0001349056603773585,
+      "loss": 0.3838,
+      "step": 143
+    },
+    {
+      "epoch": 1.014977973568282,
+      "grad_norm": 0.2316348450979017,
+      "learning_rate": 0.0001358490566037736,
+      "loss": 0.4066,
+      "step": 144
+    },
+    {
+      "epoch": 1.0220264317180616,
+      "grad_norm": 0.18472797212680125,
+      "learning_rate": 0.00013679245283018868,
+      "loss": 0.3912,
+      "step": 145
+    },
+    {
+      "epoch": 1.0290748898678415,
+      "grad_norm": 0.17861033071994495,
+      "learning_rate": 0.00013773584905660377,
+      "loss": 0.4153,
+      "step": 146
+    },
+    {
+      "epoch": 1.036123348017621,
+      "grad_norm": 0.21164655503598104,
+      "learning_rate": 0.0001386792452830189,
+      "loss": 0.3968,
+      "step": 147
+    },
+    {
+      "epoch": 1.043171806167401,
+      "grad_norm": 0.21837043786988286,
+      "learning_rate": 0.00013962264150943395,
+      "loss": 0.4003,
+      "step": 148
+    },
+    {
+      "epoch": 1.0502202643171805,
+      "grad_norm": 0.16866427716682722,
+      "learning_rate": 0.00014056603773584907,
+      "loss": 0.4059,
+      "step": 149
+    },
+    {
+      "epoch": 1.0572687224669604,
+      "grad_norm": 0.19726528767767187,
+      "learning_rate": 0.00014150943396226416,
+      "loss": 0.3855,
+      "step": 150
+    },
+    {
+      "epoch": 1.0643171806167402,
+      "grad_norm": 0.1650573728566638,
+      "learning_rate": 0.00014245283018867925,
+      "loss": 0.3963,
+      "step": 151
+    },
+    {
+      "epoch": 1.0713656387665198,
+      "grad_norm": 0.20897047326219031,
+      "learning_rate": 0.00014339622641509434,
+      "loss": 0.4018,
+      "step": 152
+    },
+    {
+      "epoch": 1.0784140969162996,
+      "grad_norm": 0.16600661861987795,
+      "learning_rate": 0.00014433962264150945,
+      "loss": 0.3942,
+      "step": 153
+    },
+    {
+      "epoch": 1.0854625550660792,
+      "grad_norm": 0.20248169593435789,
+      "learning_rate": 0.00014528301886792451,
+      "loss": 0.3949,
+      "step": 154
+    },
+    {
+      "epoch": 1.092511013215859,
+      "grad_norm": 0.1899157963129669,
+      "learning_rate": 0.00014622641509433963,
+      "loss": 0.4051,
+      "step": 155
+    },
+    {
+      "epoch": 1.0995594713656387,
+      "grad_norm": 0.1543256013373606,
+      "learning_rate": 0.00014716981132075472,
+      "loss": 0.4004,
+      "step": 156
+    },
+    {
+      "epoch": 1.1066079295154185,
+      "grad_norm": 0.30266512337497614,
+      "learning_rate": 0.0001481132075471698,
+      "loss": 0.4044,
+      "step": 157
+    },
+    {
+      "epoch": 1.1136563876651981,
+      "grad_norm": 0.2738849864528508,
+      "learning_rate": 0.0001490566037735849,
+      "loss": 0.4131,
+      "step": 158
+    },
+    {
+      "epoch": 1.120704845814978,
+      "grad_norm": 0.1754753402585768,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.4009,
+      "step": 159
+    },
+    {
+      "epoch": 1.1277533039647576,
+      "grad_norm": 0.19687618885887317,
+      "learning_rate": 0.0001509433962264151,
+      "loss": 0.3946,
+      "step": 160
+    },
+    {
+      "epoch": 1.1348017621145374,
+      "grad_norm": 0.14789784385079102,
+      "learning_rate": 0.0001518867924528302,
+      "loss": 0.3937,
+      "step": 161
+    },
+    {
+      "epoch": 1.1418502202643173,
+      "grad_norm": 0.203561013109258,
+      "learning_rate": 0.0001528301886792453,
+      "loss": 0.3904,
+      "step": 162
+    },
+    {
+      "epoch": 1.1488986784140969,
+      "grad_norm": 0.15395151538514965,
+      "learning_rate": 0.00015377358490566037,
+      "loss": 0.3832,
+      "step": 163
+    },
+    {
+      "epoch": 1.1559471365638767,
+      "grad_norm": 0.2112637633592059,
+      "learning_rate": 0.0001547169811320755,
+      "loss": 0.3838,
+      "step": 164
+    },
+    {
+      "epoch": 1.1629955947136563,
+      "grad_norm": 0.2236508693979234,
+      "learning_rate": 0.00015566037735849058,
+      "loss": 0.4179,
+      "step": 165
+    },
+    {
+      "epoch": 1.1700440528634362,
+      "grad_norm": 0.23627481898736172,
+      "learning_rate": 0.00015660377358490567,
+      "loss": 0.3985,
+      "step": 166
+    },
+    {
+      "epoch": 1.1770925110132158,
+      "grad_norm": 0.15554675692289846,
+      "learning_rate": 0.00015754716981132075,
+      "loss": 0.4,
+      "step": 167
+    },
+    {
+      "epoch": 1.1841409691629956,
+      "grad_norm": 0.2043200534456285,
+      "learning_rate": 0.00015849056603773587,
+      "loss": 0.393,
+      "step": 168
+    },
+    {
+      "epoch": 1.1911894273127754,
+      "grad_norm": 0.16760325280272487,
+      "learning_rate": 0.00015943396226415096,
+      "loss": 0.3973,
+      "step": 169
+    },
+    {
+      "epoch": 1.198237885462555,
+      "grad_norm": 0.18277149986409216,
+      "learning_rate": 0.00016037735849056605,
+      "loss": 0.3941,
+      "step": 170
+    },
+    {
+      "epoch": 1.2052863436123349,
+      "grad_norm": 0.16165318328749534,
+      "learning_rate": 0.00016132075471698114,
+      "loss": 0.3905,
+      "step": 171
+    },
+    {
+      "epoch": 1.2123348017621145,
+      "grad_norm": 0.23214987320406266,
+      "learning_rate": 0.00016226415094339625,
+      "loss": 0.3973,
+      "step": 172
+    },
+    {
+      "epoch": 1.2193832599118943,
+      "grad_norm": 0.16811013675496983,
+      "learning_rate": 0.00016320754716981132,
+      "loss": 0.3911,
+      "step": 173
+    },
+    {
+      "epoch": 1.226431718061674,
+      "grad_norm": 0.2771921701826626,
+      "learning_rate": 0.00016415094339622643,
+      "loss": 0.4069,
+      "step": 174
+    },
+    {
+      "epoch": 1.2334801762114538,
+      "grad_norm": 0.14987971358800123,
+      "learning_rate": 0.00016509433962264152,
+      "loss": 0.3859,
+      "step": 175
+    },
+    {
+      "epoch": 1.2405286343612334,
+      "grad_norm": 0.1758883867425635,
+      "learning_rate": 0.0001660377358490566,
+      "loss": 0.3963,
+      "step": 176
+    },
+    {
+      "epoch": 1.2475770925110132,
+      "grad_norm": 0.17008131919834954,
+      "learning_rate": 0.0001669811320754717,
+      "loss": 0.4074,
+      "step": 177
+    },
+    {
+      "epoch": 1.2546255506607928,
+      "grad_norm": 0.14267006241171531,
+      "learning_rate": 0.00016792452830188682,
+      "loss": 0.403,
+      "step": 178
+    },
+    {
+      "epoch": 1.2616740088105727,
+      "grad_norm": 0.20615648819756205,
+      "learning_rate": 0.00016886792452830188,
+      "loss": 0.3982,
+      "step": 179
+    },
+    {
+      "epoch": 1.2687224669603525,
+      "grad_norm": 0.1606220256534235,
+      "learning_rate": 0.000169811320754717,
+      "loss": 0.3936,
+      "step": 180
+    },
+    {
+      "epoch": 1.2757709251101321,
+      "grad_norm": 0.1709175186303027,
+      "learning_rate": 0.00017075471698113208,
+      "loss": 0.3909,
+      "step": 181
+    },
+    {
+      "epoch": 1.282819383259912,
+      "grad_norm": 0.15534871577878678,
+      "learning_rate": 0.00017169811320754717,
+      "loss": 0.3999,
+      "step": 182
+    },
+    {
+      "epoch": 1.2898678414096916,
+      "grad_norm": 0.20339817097013557,
+      "learning_rate": 0.00017264150943396226,
+      "loss": 0.3826,
+      "step": 183
+    },
+    {
+      "epoch": 1.2969162995594714,
+      "grad_norm": 0.2075431342016748,
+      "learning_rate": 0.00017358490566037738,
+      "loss": 0.3954,
+      "step": 184
+    },
+    {
+      "epoch": 1.303964757709251,
+      "grad_norm": 0.168856577206171,
+      "learning_rate": 0.00017452830188679247,
+      "loss": 0.4115,
+      "step": 185
+    },
+    {
+      "epoch": 1.3110132158590309,
+      "grad_norm": 0.2462376786143235,
+      "learning_rate": 0.00017547169811320756,
+      "loss": 0.3949,
+      "step": 186
+    },
+    {
+      "epoch": 1.3180616740088107,
+      "grad_norm": 0.1545592377630597,
+      "learning_rate": 0.00017641509433962265,
+      "loss": 0.3827,
+      "step": 187
+    },
+    {
+      "epoch": 1.3251101321585903,
+      "grad_norm": 0.19905695967349515,
+      "learning_rate": 0.00017735849056603776,
+      "loss": 0.3911,
+      "step": 188
+    },
+    {
+      "epoch": 1.33215859030837,
+      "grad_norm": 0.1955300150323613,
+      "learning_rate": 0.00017830188679245282,
+      "loss": 0.3947,
+      "step": 189
+    },
+    {
+      "epoch": 1.3392070484581498,
+      "grad_norm": 0.2229628021702979,
+      "learning_rate": 0.00017924528301886794,
+      "loss": 0.398,
+      "step": 190
+    },
+    {
+      "epoch": 1.3462555066079296,
+      "grad_norm": 0.19614606570075574,
+      "learning_rate": 0.00018018867924528303,
+      "loss": 0.4074,
+      "step": 191
+    },
+    {
+      "epoch": 1.3533039647577092,
+      "grad_norm": 0.1598604540282328,
+      "learning_rate": 0.00018113207547169812,
+      "loss": 0.3992,
+      "step": 192
+    },
+    {
+      "epoch": 1.360352422907489,
+      "grad_norm": 0.20387991518496754,
+      "learning_rate": 0.0001820754716981132,
+      "loss": 0.3929,
+      "step": 193
+    },
+    {
+      "epoch": 1.3674008810572686,
+      "grad_norm": 0.12454308014231463,
+      "learning_rate": 0.00018301886792452832,
+      "loss": 0.3992,
+      "step": 194
+    },
+    {
+      "epoch": 1.3744493392070485,
+      "grad_norm": 0.21484213485198708,
+      "learning_rate": 0.00018396226415094339,
+      "loss": 0.4061,
+      "step": 195
+    },
+    {
+      "epoch": 1.381497797356828,
+      "grad_norm": 0.15401308336611058,
+      "learning_rate": 0.0001849056603773585,
+      "loss": 0.3884,
+      "step": 196
+    },
+    {
+      "epoch": 1.388546255506608,
+      "grad_norm": 0.19815138618847297,
+      "learning_rate": 0.0001858490566037736,
+      "loss": 0.3946,
+      "step": 197
+    },
+    {
+      "epoch": 1.3955947136563878,
+      "grad_norm": 0.14461642895473947,
+      "learning_rate": 0.00018679245283018868,
+      "loss": 0.3904,
+      "step": 198
+    },
+    {
+      "epoch": 1.4026431718061674,
+      "grad_norm": 0.212484633696722,
+      "learning_rate": 0.00018773584905660377,
+      "loss": 0.4076,
+      "step": 199
+    },
+    {
+      "epoch": 1.4096916299559472,
+      "grad_norm": 0.14496075359581817,
+      "learning_rate": 0.00018867924528301889,
+      "loss": 0.3981,
+      "step": 200
+    },
+    {
+      "epoch": 1.4167400881057268,
+      "grad_norm": 0.1661533751884942,
+      "learning_rate": 0.00018962264150943397,
+      "loss": 0.3958,
+      "step": 201
+    },
+    {
+      "epoch": 1.4237885462555067,
+      "grad_norm": 0.12111638013838515,
+      "learning_rate": 0.00019056603773584906,
+      "loss": 0.3919,
+      "step": 202
+    },
+    {
+      "epoch": 1.4308370044052863,
+      "grad_norm": 0.169786199430474,
+      "learning_rate": 0.00019150943396226415,
+      "loss": 0.3914,
+      "step": 203
+    },
+    {
+      "epoch": 1.437885462555066,
+      "grad_norm": 0.12166795407993976,
+      "learning_rate": 0.00019245283018867927,
+      "loss": 0.3897,
+      "step": 204
+    },
+    {
+      "epoch": 1.444933920704846,
+      "grad_norm": 0.1428170754466983,
+      "learning_rate": 0.00019339622641509433,
+      "loss": 0.3899,
+      "step": 205
+    },
+    {
+      "epoch": 1.4519823788546256,
+      "grad_norm": 0.1325940097004292,
+      "learning_rate": 0.00019433962264150945,
+      "loss": 0.3811,
+      "step": 206
+    },
+    {
+      "epoch": 1.4590308370044052,
+      "grad_norm": 0.1764342055840309,
+      "learning_rate": 0.00019528301886792454,
+      "loss": 0.3842,
+      "step": 207
+    },
+    {
+      "epoch": 1.466079295154185,
+      "grad_norm": 0.16525555816033247,
+      "learning_rate": 0.00019622641509433963,
+      "loss": 0.3866,
+      "step": 208
+    },
+    {
+      "epoch": 1.4731277533039648,
+      "grad_norm": 0.16609219008188364,
+      "learning_rate": 0.00019716981132075472,
+      "loss": 0.3906,
+      "step": 209
+    },
+    {
+      "epoch": 1.4801762114537445,
+      "grad_norm": 0.15372596133679914,
+      "learning_rate": 0.00019811320754716983,
+      "loss": 0.3946,
+      "step": 210
+    },
+    {
+      "epoch": 1.4872246696035243,
+      "grad_norm": 0.15509792124270474,
+      "learning_rate": 0.0001990566037735849,
+      "loss": 0.3937,
+      "step": 211
+    },
+    {
+      "epoch": 1.494273127753304,
+      "grad_norm": 0.1634015929646829,
+      "learning_rate": 0.0002,
+      "loss": 0.3897,
+      "step": 212
+    },
+    {
+      "epoch": 1.5013215859030837,
+      "grad_norm": 0.1469888354740858,
+      "learning_rate": 0.00019999998944613464,
+      "loss": 0.3912,
+      "step": 213
+    },
+    {
+      "epoch": 1.5083700440528633,
+      "grad_norm": 0.1593563642564238,
+      "learning_rate": 0.00019999995778454083,
+      "loss": 0.3882,
+      "step": 214
+    },
+    {
+      "epoch": 1.5154185022026432,
+      "grad_norm": 0.18383440350747568,
+      "learning_rate": 0.0001999999050152252,
+      "loss": 0.3941,
+      "step": 215
+    },
+    {
+      "epoch": 1.522466960352423,
+      "grad_norm": 0.14625376925978495,
+      "learning_rate": 0.0001999998311381989,
+      "loss": 0.3831,
+      "step": 216
+    },
+    {
+      "epoch": 1.5295154185022026,
+      "grad_norm": 0.20348472021175565,
+      "learning_rate": 0.00019999973615347753,
+      "loss": 0.4028,
+      "step": 217
+    },
+    {
+      "epoch": 1.5365638766519822,
+      "grad_norm": 0.1459851542946225,
+      "learning_rate": 0.00019999962006108114,
+      "loss": 0.368,
+      "step": 218
+    },
+    {
+      "epoch": 1.543612334801762,
+      "grad_norm": 0.17403122336092575,
+      "learning_rate": 0.00019999948286103427,
+      "loss": 0.3918,
+      "step": 219
+    },
+    {
+      "epoch": 1.550660792951542,
+      "grad_norm": 0.1737338137830641,
+      "learning_rate": 0.00019999932455336584,
+      "loss": 0.3803,
+      "step": 220
+    },
+    {
+      "epoch": 1.5577092511013215,
+      "grad_norm": 0.17521659492131283,
+      "learning_rate": 0.00019999914513810926,
+      "loss": 0.3969,
+      "step": 221
+    },
+    {
+      "epoch": 1.5647577092511014,
+      "grad_norm": 0.2100149939943712,
+      "learning_rate": 0.0001999989446153024,
+      "loss": 0.4004,
+      "step": 222
+    },
+    {
+      "epoch": 1.5718061674008812,
+      "grad_norm": 0.2702616436100345,
+      "learning_rate": 0.00019999872298498767,
+      "loss": 0.3889,
+      "step": 223
+    },
+    {
+      "epoch": 1.5788546255506608,
+      "grad_norm": 0.1704787448711956,
+      "learning_rate": 0.00019999848024721171,
+      "loss": 0.3985,
+      "step": 224
+    },
+    {
+      "epoch": 1.5859030837004404,
+      "grad_norm": 0.16527742261422113,
+      "learning_rate": 0.00019999821640202586,
+      "loss": 0.385,
+      "step": 225
+    },
+    {
+      "epoch": 1.5929515418502203,
+      "grad_norm": 0.18481177152889494,
+      "learning_rate": 0.00019999793144948575,
+      "loss": 0.3847,
+      "step": 226
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.14358265590229097,
+      "learning_rate": 0.00019999762538965157,
+      "loss": 0.3852,
+      "step": 227
+    },
+    {
+      "epoch": 1.6070484581497797,
+      "grad_norm": 0.14902660767746098,
+      "learning_rate": 0.0001999972982225879,
+      "loss": 0.3909,
+      "step": 228
+    },
+    {
+      "epoch": 1.6140969162995593,
+      "grad_norm": 0.14539663756523594,
+      "learning_rate": 0.00019999694994836381,
+      "loss": 0.3718,
+      "step": 229
+    },
+    {
+      "epoch": 1.6211453744493394,
+      "grad_norm": 0.13311415074054977,
+      "learning_rate": 0.00019999658056705282,
+      "loss": 0.3976,
+      "step": 230
+    },
+    {
+      "epoch": 1.628193832599119,
+      "grad_norm": 0.14697957435968456,
+      "learning_rate": 0.00019999619007873286,
+      "loss": 0.3827,
+      "step": 231
+    },
+    {
+      "epoch": 1.6352422907488986,
+      "grad_norm": 0.14103522649491224,
+      "learning_rate": 0.0001999957784834864,
+      "loss": 0.3772,
+      "step": 232
+    },
+    {
+      "epoch": 1.6422907488986784,
+      "grad_norm": 0.1275211092321242,
+      "learning_rate": 0.00019999534578140032,
+      "loss": 0.3773,
+      "step": 233
+    },
+    {
+      "epoch": 1.6493392070484583,
+      "grad_norm": 0.12772061453421965,
+      "learning_rate": 0.0001999948919725659,
+      "loss": 0.3706,
+      "step": 234
+    },
+    {
+      "epoch": 1.6563876651982379,
+      "grad_norm": 0.13820533988710054,
+      "learning_rate": 0.00019999441705707897,
+      "loss": 0.3796,
+      "step": 235
+    },
+    {
+      "epoch": 1.6634361233480175,
+      "grad_norm": 0.20976957095877102,
+      "learning_rate": 0.00019999392103503975,
+      "loss": 0.3932,
+      "step": 236
+    },
+    {
+      "epoch": 1.6704845814977973,
+      "grad_norm": 0.1519248881793249,
+      "learning_rate": 0.00019999340390655296,
+      "loss": 0.3956,
+      "step": 237
+    },
+    {
+      "epoch": 1.6775330396475772,
+      "grad_norm": 0.127234979495567,
+      "learning_rate": 0.00019999286567172776,
+      "loss": 0.3764,
+      "step": 238
+    },
+    {
+      "epoch": 1.6845814977973568,
+      "grad_norm": 0.13481818150332225,
+      "learning_rate": 0.00019999230633067775,
+      "loss": 0.3789,
+      "step": 239
+    },
+    {
+      "epoch": 1.6916299559471366,
+      "grad_norm": 0.14740157149728003,
+      "learning_rate": 0.00019999172588352096,
+      "loss": 0.3781,
+      "step": 240
+    },
+    {
+      "epoch": 1.6986784140969164,
+      "grad_norm": 0.15443158886894584,
+      "learning_rate": 0.00019999112433037998,
+      "loss": 0.3847,
+      "step": 241
+    },
+    {
+      "epoch": 1.705726872246696,
+      "grad_norm": 0.14061006940640633,
+      "learning_rate": 0.00019999050167138172,
+      "loss": 0.3758,
+      "step": 242
+    },
+    {
+      "epoch": 1.7127753303964757,
+      "grad_norm": 0.15946709261370404,
+      "learning_rate": 0.00019998985790665763,
+      "loss": 0.4021,
+      "step": 243
+    },
+    {
+      "epoch": 1.7198237885462555,
+      "grad_norm": 0.16309363803698249,
+      "learning_rate": 0.0001999891930363436,
+      "loss": 0.3845,
+      "step": 244
+    },
+    {
+      "epoch": 1.7268722466960353,
+      "grad_norm": 0.22086798322876178,
+      "learning_rate": 0.00019998850706057998,
+      "loss": 0.3874,
+      "step": 245
+    },
+    {
+      "epoch": 1.733920704845815,
+      "grad_norm": 0.16785321471919676,
+      "learning_rate": 0.00019998779997951153,
+      "loss": 0.379,
+      "step": 246
+    },
+    {
+      "epoch": 1.7409691629955946,
+      "grad_norm": 0.22567933843828242,
+      "learning_rate": 0.00019998707179328756,
+      "loss": 0.378,
+      "step": 247
+    },
+    {
+      "epoch": 1.7480176211453744,
+      "grad_norm": 0.1796406086483522,
+      "learning_rate": 0.00019998632250206168,
+      "loss": 0.3842,
+      "step": 248
+    },
+    {
+      "epoch": 1.7550660792951542,
+      "grad_norm": 0.24656850844056646,
+      "learning_rate": 0.00019998555210599213,
+      "loss": 0.388,
+      "step": 249
+    },
+    {
+      "epoch": 1.7621145374449338,
+      "grad_norm": 0.1673244982371339,
+      "learning_rate": 0.0001999847606052415,
+      "loss": 0.3906,
+      "step": 250
+    },
+    {
+      "epoch": 1.7691629955947137,
+      "grad_norm": 0.1742916312332658,
+      "learning_rate": 0.00019998394799997682,
+      "loss": 0.3797,
+      "step": 251
+    },
+    {
+      "epoch": 1.7762114537444935,
+      "grad_norm": 0.15631548388378366,
+      "learning_rate": 0.00019998311429036966,
+      "loss": 0.3891,
+      "step": 252
+    },
+    {
+      "epoch": 1.7832599118942731,
+      "grad_norm": 0.169651216142054,
+      "learning_rate": 0.000199982259476596,
+      "loss": 0.3732,
+      "step": 253
+    },
+    {
+      "epoch": 1.7903083700440527,
+      "grad_norm": 0.15115105786779337,
+      "learning_rate": 0.00019998138355883623,
+      "loss": 0.3724,
+      "step": 254
+    },
+    {
+      "epoch": 1.7973568281938326,
+      "grad_norm": 0.20421334953882098,
+      "learning_rate": 0.00019998048653727528,
+      "loss": 0.3861,
+      "step": 255
+    },
+    {
+      "epoch": 1.8044052863436124,
+      "grad_norm": 0.17868950740380884,
+      "learning_rate": 0.00019997956841210247,
+      "loss": 0.3918,
+      "step": 256
+    },
+    {
+      "epoch": 1.811453744493392,
+      "grad_norm": 0.15508471832327436,
+      "learning_rate": 0.00019997862918351157,
+      "loss": 0.3904,
+      "step": 257
+    },
+    {
+      "epoch": 1.8185022026431719,
+      "grad_norm": 0.1473398057786542,
+      "learning_rate": 0.00019997766885170086,
+      "loss": 0.3782,
+      "step": 258
+    },
+    {
+      "epoch": 1.8255506607929517,
+      "grad_norm": 0.12982622026400994,
+      "learning_rate": 0.00019997668741687306,
+      "loss": 0.3902,
+      "step": 259
+    },
+    {
+      "epoch": 1.8325991189427313,
+      "grad_norm": 0.13683573451548087,
+      "learning_rate": 0.0001999756848792353,
+      "loss": 0.3986,
+      "step": 260
+    },
+    {
+      "epoch": 1.839647577092511,
+      "grad_norm": 0.149242985169114,
+      "learning_rate": 0.00019997466123899922,
+      "loss": 0.389,
+      "step": 261
+    },
+    {
+      "epoch": 1.8466960352422908,
+      "grad_norm": 0.12268970339381816,
+      "learning_rate": 0.00019997361649638085,
+      "loss": 0.3737,
+      "step": 262
+    },
+    {
+      "epoch": 1.8537444933920706,
+      "grad_norm": 0.1647488091707197,
+      "learning_rate": 0.0001999725506516007,
+      "loss": 0.3818,
+      "step": 263
+    },
+    {
+      "epoch": 1.8607929515418502,
+      "grad_norm": 0.15270217537365696,
+      "learning_rate": 0.00019997146370488383,
+      "loss": 0.3842,
+      "step": 264
+    },
+    {
+      "epoch": 1.8678414096916298,
+      "grad_norm": 0.1339066320598215,
+      "learning_rate": 0.0001999703556564596,
+      "loss": 0.3716,
+      "step": 265
+    },
+    {
+      "epoch": 1.8748898678414097,
+      "grad_norm": 0.12385908215521177,
+      "learning_rate": 0.0001999692265065619,
+      "loss": 0.3862,
+      "step": 266
+    },
+    {
+      "epoch": 1.8819383259911895,
+      "grad_norm": 0.13291822276942483,
+      "learning_rate": 0.00019996807625542908,
+      "loss": 0.3815,
+      "step": 267
+    },
+    {
+      "epoch": 1.888986784140969,
+      "grad_norm": 0.12877054655912915,
+      "learning_rate": 0.00019996690490330394,
+      "loss": 0.3879,
+      "step": 268
+    },
+    {
+      "epoch": 1.896035242290749,
+      "grad_norm": 0.14412459033702943,
+      "learning_rate": 0.00019996571245043372,
+      "loss": 0.373,
+      "step": 269
+    },
+    {
+      "epoch": 1.9030837004405288,
+      "grad_norm": 0.12234934104012078,
+      "learning_rate": 0.0001999644988970701,
+      "loss": 0.3667,
+      "step": 270
+    },
+    {
+      "epoch": 1.9101321585903084,
+      "grad_norm": 0.17335843591552944,
+      "learning_rate": 0.00019996326424346922,
+      "loss": 0.3701,
+      "step": 271
+    },
+    {
+      "epoch": 1.917180616740088,
+      "grad_norm": 0.20004776540069402,
+      "learning_rate": 0.00019996200848989177,
+      "loss": 0.3824,
+      "step": 272
+    },
+    {
+      "epoch": 1.9242290748898678,
+      "grad_norm": 0.1943228053331175,
+      "learning_rate": 0.00019996073163660273,
+      "loss": 0.3692,
+      "step": 273
+    },
+    {
+      "epoch": 1.9312775330396477,
+      "grad_norm": 0.19427616187401034,
+      "learning_rate": 0.00019995943368387168,
+      "loss": 0.3746,
+      "step": 274
+    },
+    {
+      "epoch": 1.9383259911894273,
+      "grad_norm": 0.1834446514913428,
+      "learning_rate": 0.0001999581146319725,
+      "loss": 0.3856,
+      "step": 275
+    },
+    {
+      "epoch": 1.945374449339207,
+      "grad_norm": 0.14094973359864504,
+      "learning_rate": 0.00019995677448118366,
+      "loss": 0.3881,
+      "step": 276
+    },
+    {
+      "epoch": 1.952422907488987,
+      "grad_norm": 0.1285117891160721,
+      "learning_rate": 0.00019995541323178807,
+      "loss": 0.3735,
+      "step": 277
+    },
+    {
+      "epoch": 1.9594713656387666,
+      "grad_norm": 0.13927346067205204,
+      "learning_rate": 0.000199954030884073,
+      "loss": 0.3806,
+      "step": 278
+    },
+    {
+      "epoch": 1.9665198237885462,
+      "grad_norm": 0.15042177818582794,
+      "learning_rate": 0.00019995262743833028,
+      "loss": 0.3803,
+      "step": 279
+    },
+    {
+      "epoch": 1.973568281938326,
+      "grad_norm": 0.159457813726019,
+      "learning_rate": 0.0001999512028948561,
+      "loss": 0.3788,
+      "step": 280
+    },
+    {
+      "epoch": 1.9806167400881058,
+      "grad_norm": 0.13997778728125015,
+      "learning_rate": 0.00019994975725395121,
+      "loss": 0.3757,
+      "step": 281
+    },
+    {
+      "epoch": 1.9876651982378855,
+      "grad_norm": 0.13592059010400467,
+      "learning_rate": 0.0001999482905159207,
+      "loss": 0.3745,
+      "step": 282
+    },
+    {
+      "epoch": 1.994713656387665,
+      "grad_norm": 0.14707154698465957,
+      "learning_rate": 0.00019994680268107417,
+      "loss": 0.3739,
+      "step": 283
+    },
+    {
+      "epoch": 2.001762114537445,
+      "grad_norm": 0.14206265212653796,
+      "learning_rate": 0.0001999452937497257,
+      "loss": 0.3626,
+      "step": 284
+    },
+    {
+      "epoch": 2.0088105726872247,
+      "grad_norm": 0.1458083925977779,
+      "learning_rate": 0.00019994376372219376,
+      "loss": 0.3738,
+      "step": 285
+    },
+    {
+      "epoch": 2.0158590308370044,
+      "grad_norm": 0.19999382996316467,
+      "learning_rate": 0.0001999422125988013,
+      "loss": 0.3782,
+      "step": 286
+    },
+    {
+      "epoch": 2.022907488986784,
+      "grad_norm": 0.1383255570071201,
+      "learning_rate": 0.00019994064037987574,
+      "loss": 0.3653,
+      "step": 287
+    },
+    {
+      "epoch": 2.029955947136564,
+      "grad_norm": 0.18600400282271545,
+      "learning_rate": 0.00019993904706574895,
+      "loss": 0.3787,
+      "step": 288
+    },
+    {
+      "epoch": 2.0370044052863436,
+      "grad_norm": 0.15694842931370023,
+      "learning_rate": 0.00019993743265675723,
+      "loss": 0.3791,
+      "step": 289
+    },
+    {
+      "epoch": 2.0440528634361232,
+      "grad_norm": 0.2103262597884862,
+      "learning_rate": 0.00019993579715324135,
+      "loss": 0.3803,
+      "step": 290
+    },
+    {
+      "epoch": 2.051101321585903,
+      "grad_norm": 0.19071699768795614,
+      "learning_rate": 0.00019993414055554653,
+      "loss": 0.3719,
+      "step": 291
+    },
+    {
+      "epoch": 2.058149779735683,
+      "grad_norm": 0.17181912825143084,
+      "learning_rate": 0.00019993246286402244,
+      "loss": 0.3685,
+      "step": 292
+    },
+    {
+      "epoch": 2.0651982378854625,
+      "grad_norm": 0.1521934193571064,
+      "learning_rate": 0.0001999307640790232,
+      "loss": 0.3886,
+      "step": 293
+    },
+    {
+      "epoch": 2.072246696035242,
+      "grad_norm": 0.15654354773224607,
+      "learning_rate": 0.00019992904420090737,
+      "loss": 0.3614,
+      "step": 294
+    },
+    {
+      "epoch": 2.079295154185022,
+      "grad_norm": 0.12407551093334107,
+      "learning_rate": 0.00019992730323003797,
+      "loss": 0.3618,
+      "step": 295
+    },
+    {
+      "epoch": 2.086343612334802,
+      "grad_norm": 0.13616612377920087,
+      "learning_rate": 0.00019992554116678255,
+      "loss": 0.3772,
+      "step": 296
+    },
+    {
+      "epoch": 2.0933920704845814,
+      "grad_norm": 0.128780072364095,
+      "learning_rate": 0.00019992375801151295,
+      "loss": 0.3804,
+      "step": 297
+    },
+    {
+      "epoch": 2.100440528634361,
+      "grad_norm": 0.15148633739390183,
+      "learning_rate": 0.0001999219537646056,
+      "loss": 0.3671,
+      "step": 298
+    },
+    {
+      "epoch": 2.107488986784141,
+      "grad_norm": 0.14673445707464852,
+      "learning_rate": 0.00019992012842644135,
+      "loss": 0.3624,
+      "step": 299
+    },
+    {
+      "epoch": 2.1145374449339207,
+      "grad_norm": 0.17890694329195725,
+      "learning_rate": 0.00019991828199740548,
+      "loss": 0.3707,
+      "step": 300
+    },
+    {
+      "epoch": 2.1215859030837003,
+      "grad_norm": 0.17881297268545335,
+      "learning_rate": 0.00019991641447788768,
+      "loss": 0.3829,
+      "step": 301
+    },
+    {
+      "epoch": 2.1286343612334804,
+      "grad_norm": 0.14131640775887777,
+      "learning_rate": 0.0001999145258682822,
+      "loss": 0.387,
+      "step": 302
+    },
+    {
+      "epoch": 2.13568281938326,
+      "grad_norm": 0.1687908014083778,
+      "learning_rate": 0.00019991261616898767,
+      "loss": 0.3745,
+      "step": 303
+    },
+    {
+      "epoch": 2.1427312775330396,
+      "grad_norm": 0.13233162325149653,
+      "learning_rate": 0.00019991068538040713,
+      "loss": 0.3674,
+      "step": 304
+    },
+    {
+      "epoch": 2.149779735682819,
+      "grad_norm": 0.16639479987452302,
+      "learning_rate": 0.00019990873350294824,
+      "loss": 0.3825,
+      "step": 305
+    },
+    {
+      "epoch": 2.1568281938325993,
+      "grad_norm": 0.14711540951844376,
+      "learning_rate": 0.0001999067605370229,
+      "loss": 0.3719,
+      "step": 306
+    },
+    {
+      "epoch": 2.163876651982379,
+      "grad_norm": 0.1697164190158014,
+      "learning_rate": 0.00019990476648304754,
+      "loss": 0.3692,
+      "step": 307
+    },
+    {
+      "epoch": 2.1709251101321585,
+      "grad_norm": 0.17009838555854048,
+      "learning_rate": 0.00019990275134144315,
+      "loss": 0.3667,
+      "step": 308
+    },
+    {
+      "epoch": 2.1779735682819386,
+      "grad_norm": 0.1672770193090266,
+      "learning_rate": 0.00019990071511263505,
+      "loss": 0.3704,
+      "step": 309
+    },
+    {
+      "epoch": 2.185022026431718,
+      "grad_norm": 0.16413911905821238,
+      "learning_rate": 0.000199898657797053,
+      "loss": 0.3762,
+      "step": 310
+    },
+    {
+      "epoch": 2.192070484581498,
+      "grad_norm": 0.1370438154348255,
+      "learning_rate": 0.0001998965793951313,
+      "loss": 0.3709,
+      "step": 311
+    },
+    {
+      "epoch": 2.1991189427312774,
+      "grad_norm": 0.1416559418009047,
+      "learning_rate": 0.0001998944799073086,
+      "loss": 0.3764,
+      "step": 312
+    },
+    {
+      "epoch": 2.2061674008810574,
+      "grad_norm": 0.1418624381118658,
+      "learning_rate": 0.0001998923593340281,
+      "loss": 0.378,
+      "step": 313
+    },
+    {
+      "epoch": 2.213215859030837,
+      "grad_norm": 0.14314674527049984,
+      "learning_rate": 0.00019989021767573745,
+      "loss": 0.3624,
+      "step": 314
+    },
+    {
+      "epoch": 2.2202643171806167,
+      "grad_norm": 0.15578213052321677,
+      "learning_rate": 0.00019988805493288857,
+      "loss": 0.3712,
+      "step": 315
+    },
+    {
+      "epoch": 2.2273127753303963,
+      "grad_norm": 0.1558602177799957,
+      "learning_rate": 0.00019988587110593808,
+      "loss": 0.3515,
+      "step": 316
+    },
+    {
+      "epoch": 2.2343612334801763,
+      "grad_norm": 0.16653014339886094,
+      "learning_rate": 0.00019988366619534692,
+      "loss": 0.3746,
+      "step": 317
+    },
+    {
+      "epoch": 2.241409691629956,
+      "grad_norm": 0.1847716631483791,
+      "learning_rate": 0.00019988144020158046,
+      "loss": 0.3614,
+      "step": 318
+    },
+    {
+      "epoch": 2.2484581497797356,
+      "grad_norm": 0.13429777446303995,
+      "learning_rate": 0.00019987919312510857,
+      "loss": 0.3725,
+      "step": 319
+    },
+    {
+      "epoch": 2.255506607929515,
+      "grad_norm": 0.14665971737842493,
+      "learning_rate": 0.00019987692496640556,
+      "loss": 0.3699,
+      "step": 320
+    },
+    {
+      "epoch": 2.2625550660792952,
+      "grad_norm": 0.14402712422523692,
+      "learning_rate": 0.00019987463572595019,
+      "loss": 0.3599,
+      "step": 321
+    },
+    {
+      "epoch": 2.269603524229075,
+      "grad_norm": 0.1629851658503471,
+      "learning_rate": 0.00019987232540422567,
+      "loss": 0.374,
+      "step": 322
+    },
+    {
+      "epoch": 2.2766519823788545,
+      "grad_norm": 0.19763827461608496,
+      "learning_rate": 0.00019986999400171965,
+      "loss": 0.3729,
+      "step": 323
+    },
+    {
+      "epoch": 2.2837004405286345,
+      "grad_norm": 0.14216402024325037,
+      "learning_rate": 0.00019986764151892422,
+      "loss": 0.3723,
+      "step": 324
+    },
+    {
+      "epoch": 2.290748898678414,
+      "grad_norm": 0.1860647096542648,
+      "learning_rate": 0.00019986526795633596,
+      "loss": 0.3755,
+      "step": 325
+    },
+    {
+      "epoch": 2.2977973568281937,
+      "grad_norm": 0.12336388766240439,
+      "learning_rate": 0.0001998628733144559,
+      "loss": 0.373,
+      "step": 326
+    },
+    {
+      "epoch": 2.3048458149779734,
+      "grad_norm": 0.15123535871227883,
+      "learning_rate": 0.00019986045759378947,
+      "loss": 0.3646,
+      "step": 327
+    },
+    {
+      "epoch": 2.3118942731277534,
+      "grad_norm": 0.1480602677694806,
+      "learning_rate": 0.00019985802079484653,
+      "loss": 0.3724,
+      "step": 328
+    },
+    {
+      "epoch": 2.318942731277533,
+      "grad_norm": 0.1302076105114072,
+      "learning_rate": 0.00019985556291814147,
+      "loss": 0.3617,
+      "step": 329
+    },
+    {
+      "epoch": 2.3259911894273126,
+      "grad_norm": 0.17147867579888978,
+      "learning_rate": 0.0001998530839641931,
+      "loss": 0.3674,
+      "step": 330
+    },
+    {
+      "epoch": 2.3330396475770927,
+      "grad_norm": 0.16466984473422716,
+      "learning_rate": 0.00019985058393352467,
+      "loss": 0.364,
+      "step": 331
+    },
+    {
+      "epoch": 2.3400881057268723,
+      "grad_norm": 0.2039429046437469,
+      "learning_rate": 0.00019984806282666385,
+      "loss": 0.3861,
+      "step": 332
+    },
+    {
+      "epoch": 2.347136563876652,
+      "grad_norm": 0.17762803794241866,
+      "learning_rate": 0.00019984552064414282,
+      "loss": 0.3751,
+      "step": 333
+    },
+    {
+      "epoch": 2.3541850220264315,
+      "grad_norm": 0.12724857202905812,
+      "learning_rate": 0.00019984295738649822,
+      "loss": 0.3732,
+      "step": 334
+    },
+    {
+      "epoch": 2.3612334801762116,
+      "grad_norm": 0.16088579961662378,
+      "learning_rate": 0.00019984037305427097,
+      "loss": 0.3568,
+      "step": 335
+    },
+    {
+      "epoch": 2.368281938325991,
+      "grad_norm": 0.15167660059660987,
+      "learning_rate": 0.00019983776764800668,
+      "loss": 0.372,
+      "step": 336
+    },
+    {
+      "epoch": 2.375330396475771,
+      "grad_norm": 0.12971740469555612,
+      "learning_rate": 0.0001998351411682552,
+      "loss": 0.3744,
+      "step": 337
+    },
+    {
+      "epoch": 2.382378854625551,
+      "grad_norm": 0.16532908477398073,
+      "learning_rate": 0.00019983249361557105,
+      "loss": 0.3731,
+      "step": 338
+    },
+    {
+      "epoch": 2.3894273127753305,
+      "grad_norm": 0.15932597599071294,
+      "learning_rate": 0.00019982982499051292,
+      "loss": 0.3689,
+      "step": 339
+    },
+    {
+      "epoch": 2.39647577092511,
+      "grad_norm": 0.16734778982359053,
+      "learning_rate": 0.00019982713529364418,
+      "loss": 0.368,
+      "step": 340
+    },
+    {
+      "epoch": 2.4035242290748897,
+      "grad_norm": 0.14735013248531414,
+      "learning_rate": 0.00019982442452553257,
+      "loss": 0.3689,
+      "step": 341
+    },
+    {
+      "epoch": 2.4105726872246698,
+      "grad_norm": 0.1497084703167721,
+      "learning_rate": 0.00019982169268675023,
+      "loss": 0.378,
+      "step": 342
+    },
+    {
+      "epoch": 2.4176211453744494,
+      "grad_norm": 0.16086744408421122,
+      "learning_rate": 0.00019981893977787384,
+      "loss": 0.3706,
+      "step": 343
+    },
+    {
+      "epoch": 2.424669603524229,
+      "grad_norm": 0.14036856518106583,
+      "learning_rate": 0.00019981616579948442,
+      "loss": 0.3502,
+      "step": 344
+    },
+    {
+      "epoch": 2.431718061674009,
+      "grad_norm": 0.13680289920639105,
+      "learning_rate": 0.00019981337075216754,
+      "loss": 0.3653,
+      "step": 345
+    },
+    {
+      "epoch": 2.4387665198237887,
+      "grad_norm": 0.14851965862526328,
+      "learning_rate": 0.0001998105546365131,
+      "loss": 0.3619,
+      "step": 346
+    },
+    {
+      "epoch": 2.4458149779735683,
+      "grad_norm": 0.14004249039549074,
+      "learning_rate": 0.00019980771745311561,
+      "loss": 0.3673,
+      "step": 347
+    },
+    {
+      "epoch": 2.452863436123348,
+      "grad_norm": 0.1386965973430117,
+      "learning_rate": 0.00019980485920257388,
+      "loss": 0.3758,
+      "step": 348
+    },
+    {
+      "epoch": 2.4599118942731275,
+      "grad_norm": 0.12702388422739236,
+      "learning_rate": 0.00019980197988549124,
+      "loss": 0.3773,
+      "step": 349
+    },
+    {
+      "epoch": 2.4669603524229076,
+      "grad_norm": 0.16113471913603836,
+      "learning_rate": 0.00019979907950247545,
+      "loss": 0.3636,
+      "step": 350
+    },
+    {
+      "epoch": 2.474008810572687,
+      "grad_norm": 0.1464062510719453,
+      "learning_rate": 0.00019979615805413873,
+      "loss": 0.3759,
+      "step": 351
+    },
+    {
+      "epoch": 2.481057268722467,
+      "grad_norm": 0.15208102594609738,
+      "learning_rate": 0.00019979321554109767,
+      "loss": 0.3744,
+      "step": 352
+    },
+    {
+      "epoch": 2.488105726872247,
+      "grad_norm": 0.13775739347512653,
+      "learning_rate": 0.0001997902519639734,
+      "loss": 0.3567,
+      "step": 353
+    },
+    {
+      "epoch": 2.4951541850220265,
+      "grad_norm": 0.15371809154242025,
+      "learning_rate": 0.0001997872673233915,
+      "loss": 0.3728,
+      "step": 354
+    },
+    {
+      "epoch": 2.502202643171806,
+      "grad_norm": 0.16027145524137118,
+      "learning_rate": 0.00019978426161998194,
+      "loss": 0.3603,
+      "step": 355
+    },
+    {
+      "epoch": 2.5092511013215857,
+      "grad_norm": 0.1927208274628228,
+      "learning_rate": 0.00019978123485437914,
+      "loss": 0.3666,
+      "step": 356
+    },
+    {
+      "epoch": 2.5162995594713657,
+      "grad_norm": 0.15831114022245443,
+      "learning_rate": 0.00019977818702722198,
+      "loss": 0.3611,
+      "step": 357
+    },
+    {
+      "epoch": 2.5233480176211454,
+      "grad_norm": 0.13694606291481892,
+      "learning_rate": 0.0001997751181391538,
+      "loss": 0.3754,
+      "step": 358
+    },
+    {
+      "epoch": 2.530396475770925,
+      "grad_norm": 0.14547375519189973,
+      "learning_rate": 0.00019977202819082237,
+      "loss": 0.3566,
+      "step": 359
+    },
+    {
+      "epoch": 2.537444933920705,
+      "grad_norm": 0.12290738375700438,
+      "learning_rate": 0.0001997689171828799,
+      "loss": 0.3581,
+      "step": 360
+    },
+    {
+      "epoch": 2.5444933920704846,
+      "grad_norm": 0.14227329827690405,
+      "learning_rate": 0.00019976578511598307,
+      "loss": 0.3724,
+      "step": 361
+    },
+    {
+      "epoch": 2.5515418502202643,
+      "grad_norm": 0.14829225641207405,
+      "learning_rate": 0.00019976263199079298,
+      "loss": 0.3696,
+      "step": 362
+    },
+    {
+      "epoch": 2.558590308370044,
+      "grad_norm": 0.13678801452212694,
+      "learning_rate": 0.0001997594578079752,
+      "loss": 0.3747,
+      "step": 363
+    },
+    {
+      "epoch": 2.565638766519824,
+      "grad_norm": 0.14635177238468183,
+      "learning_rate": 0.00019975626256819968,
+      "loss": 0.3681,
+      "step": 364
+    },
+    {
+      "epoch": 2.5726872246696035,
+      "grad_norm": 0.15289589496252276,
+      "learning_rate": 0.00019975304627214087,
+      "loss": 0.3741,
+      "step": 365
+    },
+    {
+      "epoch": 2.579735682819383,
+      "grad_norm": 0.13779300265165564,
+      "learning_rate": 0.0001997498089204777,
+      "loss": 0.3492,
+      "step": 366
+    },
+    {
+      "epoch": 2.586784140969163,
+      "grad_norm": 0.16531592760758387,
+      "learning_rate": 0.00019974655051389353,
+      "loss": 0.3724,
+      "step": 367
+    },
+    {
+      "epoch": 2.593832599118943,
+      "grad_norm": 0.16868337212968942,
+      "learning_rate": 0.00019974327105307604,
+      "loss": 0.3681,
+      "step": 368
+    },
+    {
+      "epoch": 2.6008810572687224,
+      "grad_norm": 0.14725911846216463,
+      "learning_rate": 0.0001997399705387175,
+      "loss": 0.3646,
+      "step": 369
+    },
+    {
+      "epoch": 2.607929515418502,
+      "grad_norm": 0.19930030576896282,
+      "learning_rate": 0.00019973664897151458,
+      "loss": 0.3618,
+      "step": 370
+    },
+    {
+      "epoch": 2.614977973568282,
+      "grad_norm": 0.1445263035887974,
+      "learning_rate": 0.00019973330635216839,
+      "loss": 0.379,
+      "step": 371
+    },
+    {
+      "epoch": 2.6220264317180617,
+      "grad_norm": 0.18462489138006302,
+      "learning_rate": 0.00019972994268138442,
+      "loss": 0.36,
+      "step": 372
+    },
+    {
+      "epoch": 2.6290748898678413,
+      "grad_norm": 0.13496753767715913,
+      "learning_rate": 0.00019972655795987273,
+      "loss": 0.3588,
+      "step": 373
+    },
+    {
+      "epoch": 2.6361233480176214,
+      "grad_norm": 0.1621299260004981,
+      "learning_rate": 0.00019972315218834777,
+      "loss": 0.3643,
+      "step": 374
+    },
+    {
+      "epoch": 2.643171806167401,
+      "grad_norm": 0.13358352349442537,
+      "learning_rate": 0.0001997197253675284,
+      "loss": 0.3554,
+      "step": 375
+    },
+    {
+      "epoch": 2.6502202643171806,
+      "grad_norm": 0.17948708212924663,
+      "learning_rate": 0.0001997162774981379,
+      "loss": 0.3725,
+      "step": 376
+    },
+    {
+      "epoch": 2.65726872246696,
+      "grad_norm": 0.14221872382605433,
+      "learning_rate": 0.00019971280858090408,
+      "loss": 0.3631,
+      "step": 377
+    },
+    {
+      "epoch": 2.66431718061674,
+      "grad_norm": 0.1838647462304091,
+      "learning_rate": 0.00019970931861655914,
+      "loss": 0.3838,
+      "step": 378
+    },
+    {
+      "epoch": 2.67136563876652,
+      "grad_norm": 0.20035477924776496,
+      "learning_rate": 0.00019970580760583973,
+      "loss": 0.3692,
+      "step": 379
+    },
+    {
+      "epoch": 2.6784140969162995,
+      "grad_norm": 0.14317141252203067,
+      "learning_rate": 0.000199702275549487,
+      "loss": 0.3664,
+      "step": 380
+    },
+    {
+      "epoch": 2.6854625550660796,
+      "grad_norm": 0.15068592532717978,
+      "learning_rate": 0.00019969872244824638,
+      "loss": 0.3662,
+      "step": 381
+    },
+    {
+      "epoch": 2.692511013215859,
+      "grad_norm": 0.14865341206297536,
+      "learning_rate": 0.00019969514830286796,
+      "loss": 0.3633,
+      "step": 382
+    },
+    {
+      "epoch": 2.699559471365639,
+      "grad_norm": 0.13270346210755282,
+      "learning_rate": 0.00019969155311410605,
+      "loss": 0.3537,
+      "step": 383
+    },
+    {
+      "epoch": 2.7066079295154184,
+      "grad_norm": 0.17920618920355513,
+      "learning_rate": 0.00019968793688271959,
+      "loss": 0.3629,
+      "step": 384
+    },
+    {
+      "epoch": 2.713656387665198,
+      "grad_norm": 0.119411238474704,
+      "learning_rate": 0.00019968429960947188,
+      "loss": 0.3621,
+      "step": 385
+    },
+    {
+      "epoch": 2.720704845814978,
+      "grad_norm": 0.16520770126940176,
+      "learning_rate": 0.00019968064129513065,
+      "loss": 0.3669,
+      "step": 386
+    },
+    {
+      "epoch": 2.7277533039647577,
+      "grad_norm": 0.14358661499025444,
+      "learning_rate": 0.00019967696194046807,
+      "loss": 0.366,
+      "step": 387
+    },
+    {
+      "epoch": 2.7348017621145373,
+      "grad_norm": 0.14683505818710937,
+      "learning_rate": 0.0001996732615462608,
+      "loss": 0.355,
+      "step": 388
+    },
+    {
+      "epoch": 2.7418502202643174,
+      "grad_norm": 0.13927985034246912,
+      "learning_rate": 0.00019966954011328987,
+      "loss": 0.3655,
+      "step": 389
+    },
+    {
+      "epoch": 2.748898678414097,
+      "grad_norm": 0.15145939731020616,
+      "learning_rate": 0.00019966579764234084,
+      "loss": 0.3626,
+      "step": 390
+    },
+    {
+      "epoch": 2.7559471365638766,
+      "grad_norm": 0.131309938986961,
+      "learning_rate": 0.0001996620341342036,
+      "loss": 0.3585,
+      "step": 391
+    },
+    {
+      "epoch": 2.762995594713656,
+      "grad_norm": 0.15789813878076026,
+      "learning_rate": 0.0001996582495896726,
+      "loss": 0.3677,
+      "step": 392
+    },
+    {
+      "epoch": 2.7700440528634362,
+      "grad_norm": 0.1623570247433147,
+      "learning_rate": 0.00019965444400954664,
+      "loss": 0.3681,
+      "step": 393
+    },
+    {
+      "epoch": 2.777092511013216,
+      "grad_norm": 0.1246233487436172,
+      "learning_rate": 0.00019965061739462902,
+      "loss": 0.3614,
+      "step": 394
+    },
+    {
+      "epoch": 2.7841409691629955,
+      "grad_norm": 0.18498933866563294,
+      "learning_rate": 0.00019964676974572744,
+      "loss": 0.3587,
+      "step": 395
+    },
+    {
+      "epoch": 2.7911894273127755,
+      "grad_norm": 0.11126474394977104,
+      "learning_rate": 0.000199642901063654,
+      "loss": 0.3621,
+      "step": 396
+    },
+    {
+      "epoch": 2.798237885462555,
+      "grad_norm": 0.13736702668188772,
+      "learning_rate": 0.00019963901134922537,
+      "loss": 0.3691,
+      "step": 397
+    },
+    {
+      "epoch": 2.8052863436123348,
+      "grad_norm": 0.15159789880352756,
+      "learning_rate": 0.00019963510060326255,
+      "loss": 0.3491,
+      "step": 398
+    },
+    {
+      "epoch": 2.8123348017621144,
+      "grad_norm": 0.14288441697316226,
+      "learning_rate": 0.00019963116882659097,
+      "loss": 0.3634,
+      "step": 399
+    },
+    {
+      "epoch": 2.8193832599118944,
+      "grad_norm": 0.13452866791029736,
+      "learning_rate": 0.00019962721602004058,
+      "loss": 0.355,
+      "step": 400
+    },
+    {
+      "epoch": 2.826431718061674,
+      "grad_norm": 0.15050391455015735,
+      "learning_rate": 0.0001996232421844457,
+      "loss": 0.3661,
+      "step": 401
+    },
+    {
+      "epoch": 2.8334801762114536,
+      "grad_norm": 0.14757815422482062,
+      "learning_rate": 0.0001996192473206452,
+      "loss": 0.3644,
+      "step": 402
+    },
+    {
+      "epoch": 2.8405286343612337,
+      "grad_norm": 0.154745811981344,
+      "learning_rate": 0.00019961523142948218,
+      "loss": 0.3792,
+      "step": 403
+    },
+    {
+      "epoch": 2.8475770925110133,
+      "grad_norm": 0.12634997485728813,
+      "learning_rate": 0.00019961119451180438,
+      "loss": 0.3702,
+      "step": 404
+    },
+    {
+      "epoch": 2.854625550660793,
+      "grad_norm": 0.12674159823100195,
+      "learning_rate": 0.00019960713656846392,
+      "loss": 0.3643,
+      "step": 405
+    },
+    {
+      "epoch": 2.8616740088105725,
+      "grad_norm": 0.13321745357124373,
+      "learning_rate": 0.00019960305760031728,
+      "loss": 0.3609,
+      "step": 406
+    },
+    {
+      "epoch": 2.868722466960352,
+      "grad_norm": 0.1856585797004788,
+      "learning_rate": 0.00019959895760822546,
+      "loss": 0.3587,
+      "step": 407
+    },
+    {
+      "epoch": 2.875770925110132,
+      "grad_norm": 0.19251215548118117,
+      "learning_rate": 0.00019959483659305388,
+      "loss": 0.3636,
+      "step": 408
+    },
+    {
+      "epoch": 2.882819383259912,
+      "grad_norm": 0.14414809011347715,
+      "learning_rate": 0.00019959069455567242,
+      "loss": 0.36,
+      "step": 409
+    },
+    {
+      "epoch": 2.889867841409692,
+      "grad_norm": 0.16208342836647502,
+      "learning_rate": 0.00019958653149695533,
+      "loss": 0.3644,
+      "step": 410
+    },
+    {
+      "epoch": 2.8969162995594715,
+      "grad_norm": 0.1289389488956927,
+      "learning_rate": 0.00019958234741778133,
+      "loss": 0.363,
+      "step": 411
+    },
+    {
+      "epoch": 2.903964757709251,
+      "grad_norm": 0.1664107375748134,
+      "learning_rate": 0.00019957814231903363,
+      "loss": 0.3736,
+      "step": 412
+    },
+    {
+      "epoch": 2.9110132158590307,
+      "grad_norm": 0.14196672130335375,
+      "learning_rate": 0.00019957391620159978,
+      "loss": 0.3669,
+      "step": 413
+    },
+    {
+      "epoch": 2.9180616740088103,
+      "grad_norm": 0.11733123048825823,
+      "learning_rate": 0.00019956966906637186,
+      "loss": 0.3724,
+      "step": 414
+    },
+    {
+      "epoch": 2.9251101321585904,
+      "grad_norm": 0.1738242080067085,
+      "learning_rate": 0.00019956540091424632,
+      "loss": 0.3705,
+      "step": 415
+    },
+    {
+      "epoch": 2.93215859030837,
+      "grad_norm": 0.13792271084556862,
+      "learning_rate": 0.00019956111174612405,
+      "loss": 0.3598,
+      "step": 416
+    },
+    {
+      "epoch": 2.9392070484581496,
+      "grad_norm": 0.14811611430978458,
+      "learning_rate": 0.00019955680156291049,
+      "loss": 0.3702,
+      "step": 417
+    },
+    {
+      "epoch": 2.9462555066079297,
+      "grad_norm": 0.14635389921873676,
+      "learning_rate": 0.0001995524703655153,
+      "loss": 0.3684,
+      "step": 418
+    },
+    {
+      "epoch": 2.9533039647577093,
+      "grad_norm": 0.16620892726461883,
+      "learning_rate": 0.00019954811815485276,
+      "loss": 0.3643,
+      "step": 419
+    },
+    {
+      "epoch": 2.960352422907489,
+      "grad_norm": 0.1452385610335707,
+      "learning_rate": 0.00019954374493184152,
+      "loss": 0.3647,
+      "step": 420
+    },
+    {
+      "epoch": 2.9674008810572685,
+      "grad_norm": 0.15616710859837676,
+      "learning_rate": 0.00019953935069740467,
+      "loss": 0.3626,
+      "step": 421
+    },
+    {
+      "epoch": 2.9744493392070486,
+      "grad_norm": 0.15385010094990623,
+      "learning_rate": 0.0001995349354524697,
+      "loss": 0.3651,
+      "step": 422
+    },
+    {
+      "epoch": 2.981497797356828,
+      "grad_norm": 0.14672314015267904,
+      "learning_rate": 0.00019953049919796862,
+      "loss": 0.3629,
+      "step": 423
+    },
+    {
+      "epoch": 2.988546255506608,
+      "grad_norm": 0.14747656507885049,
+      "learning_rate": 0.0001995260419348378,
+      "loss": 0.3742,
+      "step": 424
+    },
+    {
+      "epoch": 2.995594713656388,
+      "grad_norm": 0.13616922150083577,
+      "learning_rate": 0.00019952156366401805,
+      "loss": 0.3746,
+      "step": 425
+    },
+    {
+      "epoch": 3.0026431718061675,
+      "grad_norm": 0.16150824573711983,
+      "learning_rate": 0.00019951706438645466,
+      "loss": 0.3564,
+      "step": 426
+    },
+    {
+      "epoch": 3.009691629955947,
+      "grad_norm": 0.13567099580505082,
+      "learning_rate": 0.00019951254410309726,
+      "loss": 0.3591,
+      "step": 427
+    },
+    {
+      "epoch": 3.0167400881057267,
+      "grad_norm": 0.16992307852142083,
+      "learning_rate": 0.00019950800281490007,
+      "loss": 0.3426,
+      "step": 428
+    },
+    {
+      "epoch": 3.0237885462555067,
+      "grad_norm": 0.15511652477067847,
+      "learning_rate": 0.00019950344052282163,
+      "loss": 0.3574,
+      "step": 429
+    },
+    {
+      "epoch": 3.0308370044052864,
+      "grad_norm": 0.14930126093681725,
+      "learning_rate": 0.0001994988572278249,
+      "loss": 0.3648,
+      "step": 430
+    },
+    {
+      "epoch": 3.037885462555066,
+      "grad_norm": 0.2165849988064566,
+      "learning_rate": 0.00019949425293087734,
+      "loss": 0.3619,
+      "step": 431
+    },
+    {
+      "epoch": 3.044933920704846,
+      "grad_norm": 0.1899620544213891,
+      "learning_rate": 0.00019948962763295075,
+      "loss": 0.3681,
+      "step": 432
+    },
+    {
+      "epoch": 3.0519823788546256,
+      "grad_norm": 0.13159653659951254,
+      "learning_rate": 0.00019948498133502153,
+      "loss": 0.3536,
+      "step": 433
+    },
+    {
+      "epoch": 3.0590308370044053,
+      "grad_norm": 0.19597794004426933,
+      "learning_rate": 0.00019948031403807035,
+      "loss": 0.3627,
+      "step": 434
+    },
+    {
+      "epoch": 3.066079295154185,
+      "grad_norm": 0.14971347912074287,
+      "learning_rate": 0.00019947562574308236,
+      "loss": 0.3538,
+      "step": 435
+    },
+    {
+      "epoch": 3.073127753303965,
+      "grad_norm": 0.140686188036909,
+      "learning_rate": 0.0001994709164510472,
+      "loss": 0.353,
+      "step": 436
+    },
+    {
+      "epoch": 3.0801762114537445,
+      "grad_norm": 0.14339908561727557,
+      "learning_rate": 0.0001994661861629588,
+      "loss": 0.347,
+      "step": 437
+    },
+    {
+      "epoch": 3.087224669603524,
+      "grad_norm": 0.14362148707028793,
+      "learning_rate": 0.00019946143487981573,
+      "loss": 0.3484,
+      "step": 438
+    },
+    {
+      "epoch": 3.0942731277533038,
+      "grad_norm": 0.13364001643131998,
+      "learning_rate": 0.0001994566626026208,
+      "loss": 0.3523,
+      "step": 439
+    },
+    {
+      "epoch": 3.101321585903084,
+      "grad_norm": 0.1443885582688635,
+      "learning_rate": 0.00019945186933238138,
+      "loss": 0.3514,
+      "step": 440
+    },
+    {
+      "epoch": 3.1083700440528634,
+      "grad_norm": 0.15318075363324601,
+      "learning_rate": 0.00019944705507010917,
+      "loss": 0.3565,
+      "step": 441
+    },
+    {
+      "epoch": 3.115418502202643,
+      "grad_norm": 0.1625834589431934,
+      "learning_rate": 0.00019944221981682038,
+      "loss": 0.3371,
+      "step": 442
+    },
+    {
+      "epoch": 3.122466960352423,
+      "grad_norm": 0.1606588716202299,
+      "learning_rate": 0.00019943736357353565,
+      "loss": 0.3471,
+      "step": 443
+    },
+    {
+      "epoch": 3.1295154185022027,
+      "grad_norm": 0.14798971331334146,
+      "learning_rate": 0.00019943248634127998,
+      "loss": 0.3554,
+      "step": 444
+    },
+    {
+      "epoch": 3.1365638766519823,
+      "grad_norm": 0.1458319286521924,
+      "learning_rate": 0.00019942758812108285,
+      "loss": 0.3583,
+      "step": 445
+    },
+    {
+      "epoch": 3.143612334801762,
+      "grad_norm": 0.139537089301728,
+      "learning_rate": 0.00019942266891397815,
+      "loss": 0.3518,
+      "step": 446
+    },
+    {
+      "epoch": 3.150660792951542,
+      "grad_norm": 0.12840960291674033,
+      "learning_rate": 0.00019941772872100427,
+      "loss": 0.3413,
+      "step": 447
+    },
+    {
+      "epoch": 3.1577092511013216,
+      "grad_norm": 0.13204170272966584,
+      "learning_rate": 0.0001994127675432039,
+      "loss": 0.3403,
+      "step": 448
+    },
+    {
+      "epoch": 3.1647577092511012,
+      "grad_norm": 0.16465297730916656,
+      "learning_rate": 0.0001994077853816243,
+      "loss": 0.3469,
+      "step": 449
+    },
+    {
+      "epoch": 3.171806167400881,
+      "grad_norm": 0.13889579453293738,
+      "learning_rate": 0.00019940278223731705,
+      "loss": 0.3518,
+      "step": 450
+    },
+    {
+      "epoch": 3.178854625550661,
+      "grad_norm": 0.17255206650567684,
+      "learning_rate": 0.0001993977581113382,
+      "loss": 0.3543,
+      "step": 451
+    },
+    {
+      "epoch": 3.1859030837004405,
+      "grad_norm": 0.15565396165802473,
+      "learning_rate": 0.00019939271300474824,
+      "loss": 0.3486,
+      "step": 452
+    },
+    {
+      "epoch": 3.19295154185022,
+      "grad_norm": 0.14324447684884412,
+      "learning_rate": 0.00019938764691861208,
+      "loss": 0.3534,
+      "step": 453
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.18273223822042278,
+      "learning_rate": 0.00019938255985399905,
+      "loss": 0.3553,
+      "step": 454
+    },
+    {
+      "epoch": 3.20704845814978,
+      "grad_norm": 0.16089834515122986,
+      "learning_rate": 0.0001993774518119829,
+      "loss": 0.3616,
+      "step": 455
+    },
+    {
+      "epoch": 3.2140969162995594,
+      "grad_norm": 0.14205639215415486,
+      "learning_rate": 0.00019937232279364187,
+      "loss": 0.3632,
+      "step": 456
+    },
+    {
+      "epoch": 3.221145374449339,
+      "grad_norm": 0.17610102094770114,
+      "learning_rate": 0.00019936717280005851,
+      "loss": 0.3639,
+      "step": 457
+    },
+    {
+      "epoch": 3.228193832599119,
+      "grad_norm": 0.13667171744767187,
+      "learning_rate": 0.00019936200183231993,
+      "loss": 0.3558,
+      "step": 458
+    },
+    {
+      "epoch": 3.2352422907488987,
+      "grad_norm": 0.16274450608372715,
+      "learning_rate": 0.00019935680989151757,
+      "loss": 0.3499,
+      "step": 459
+    },
+    {
+      "epoch": 3.2422907488986783,
+      "grad_norm": 0.13949642632810816,
+      "learning_rate": 0.00019935159697874732,
+      "loss": 0.3479,
+      "step": 460
+    },
+    {
+      "epoch": 3.2493392070484584,
+      "grad_norm": 0.14639548253982834,
+      "learning_rate": 0.00019934636309510955,
+      "loss": 0.353,
+      "step": 461
+    },
+    {
+      "epoch": 3.256387665198238,
+      "grad_norm": 0.1528241994625873,
+      "learning_rate": 0.00019934110824170899,
+      "loss": 0.3452,
+      "step": 462
+    },
+    {
+      "epoch": 3.2634361233480176,
+      "grad_norm": 0.13687280744455566,
+      "learning_rate": 0.0001993358324196548,
+      "loss": 0.3488,
+      "step": 463
+    },
+    {
+      "epoch": 3.270484581497797,
+      "grad_norm": 0.1702058552874099,
+      "learning_rate": 0.0001993305356300606,
+      "loss": 0.3692,
+      "step": 464
+    },
+    {
+      "epoch": 3.2775330396475773,
+      "grad_norm": 0.17949089217044084,
+      "learning_rate": 0.00019932521787404444,
+      "loss": 0.3583,
+      "step": 465
+    },
+    {
+      "epoch": 3.284581497797357,
+      "grad_norm": 0.17936296718451436,
+      "learning_rate": 0.00019931987915272877,
+      "loss": 0.3538,
+      "step": 466
+    },
+    {
+      "epoch": 3.2916299559471365,
+      "grad_norm": 0.1505766110796783,
+      "learning_rate": 0.00019931451946724046,
+      "loss": 0.3464,
+      "step": 467
+    },
+    {
+      "epoch": 3.2986784140969165,
+      "grad_norm": 0.1626252813581082,
+      "learning_rate": 0.00019930913881871084,
+      "loss": 0.3396,
+      "step": 468
+    },
+    {
+      "epoch": 3.305726872246696,
+      "grad_norm": 0.15379385909534216,
+      "learning_rate": 0.0001993037372082756,
+      "loss": 0.3496,
+      "step": 469
+    },
+    {
+      "epoch": 3.3127753303964758,
+      "grad_norm": 0.14404632327171263,
+      "learning_rate": 0.00019929831463707495,
+      "loss": 0.3502,
+      "step": 470
+    },
+    {
+      "epoch": 3.3198237885462554,
+      "grad_norm": 0.1684009242093903,
+      "learning_rate": 0.00019929287110625348,
+      "loss": 0.3455,
+      "step": 471
+    },
+    {
+      "epoch": 3.3268722466960354,
+      "grad_norm": 0.12360723910711703,
+      "learning_rate": 0.0001992874066169601,
+      "loss": 0.3565,
+      "step": 472
+    },
+    {
+      "epoch": 3.333920704845815,
+      "grad_norm": 0.22630738917002552,
+      "learning_rate": 0.00019928192117034835,
+      "loss": 0.349,
+      "step": 473
+    },
+    {
+      "epoch": 3.3409691629955947,
+      "grad_norm": 0.14116487527080118,
+      "learning_rate": 0.00019927641476757605,
+      "loss": 0.3503,
+      "step": 474
+    },
+    {
+      "epoch": 3.3480176211453743,
+      "grad_norm": 0.20919971605349494,
+      "learning_rate": 0.0001992708874098054,
+      "loss": 0.3502,
+      "step": 475
+    },
+    {
+      "epoch": 3.3550660792951543,
+      "grad_norm": 0.1405924616358806,
+      "learning_rate": 0.0001992653390982032,
+      "loss": 0.3619,
+      "step": 476
+    },
+    {
+      "epoch": 3.362114537444934,
+      "grad_norm": 0.24000015389422735,
+      "learning_rate": 0.00019925976983394054,
+      "loss": 0.3614,
+      "step": 477
+    },
+    {
+      "epoch": 3.3691629955947135,
+      "grad_norm": 0.129259584310445,
+      "learning_rate": 0.00019925417961819296,
+      "loss": 0.3352,
+      "step": 478
+    },
+    {
+      "epoch": 3.376211453744493,
+      "grad_norm": 0.15123890294187634,
+      "learning_rate": 0.00019924856845214044,
+      "loss": 0.3503,
+      "step": 479
+    },
+    {
+      "epoch": 3.383259911894273,
+      "grad_norm": 0.16983115015766367,
+      "learning_rate": 0.00019924293633696733,
+      "loss": 0.3492,
+      "step": 480
+    },
+    {
+      "epoch": 3.390308370044053,
+      "grad_norm": 0.13228294353981884,
+      "learning_rate": 0.0001992372832738625,
+      "loss": 0.3707,
+      "step": 481
+    },
+    {
+      "epoch": 3.3973568281938324,
+      "grad_norm": 0.16291849656906188,
+      "learning_rate": 0.00019923160926401914,
+      "loss": 0.3522,
+      "step": 482
+    },
+    {
+      "epoch": 3.4044052863436125,
+      "grad_norm": 0.15848061921047854,
+      "learning_rate": 0.0001992259143086349,
+      "loss": 0.3486,
+      "step": 483
+    },
+    {
+      "epoch": 3.411453744493392,
+      "grad_norm": 0.13533863967498982,
+      "learning_rate": 0.00019922019840891192,
+      "loss": 0.3679,
+      "step": 484
+    },
+    {
+      "epoch": 3.4185022026431717,
+      "grad_norm": 0.1291174870780889,
+      "learning_rate": 0.00019921446156605664,
+      "loss": 0.3564,
+      "step": 485
+    },
+    {
+      "epoch": 3.4255506607929513,
+      "grad_norm": 0.14239605824314558,
+      "learning_rate": 0.00019920870378127992,
+      "loss": 0.3502,
+      "step": 486
+    },
+    {
+      "epoch": 3.4325991189427314,
+      "grad_norm": 0.1374533528464369,
+      "learning_rate": 0.00019920292505579726,
+      "loss": 0.3587,
+      "step": 487
+    },
+    {
+      "epoch": 3.439647577092511,
+      "grad_norm": 0.13752208981671687,
+      "learning_rate": 0.00019919712539082828,
+      "loss": 0.3577,
+      "step": 488
+    },
+    {
+      "epoch": 3.4466960352422906,
+      "grad_norm": 0.14119602949711194,
+      "learning_rate": 0.0001991913047875972,
+      "loss": 0.354,
+      "step": 489
+    },
+    {
+      "epoch": 3.4537444933920707,
+      "grad_norm": 0.12136716236298661,
+      "learning_rate": 0.00019918546324733264,
+      "loss": 0.3521,
+      "step": 490
+    },
+    {
+      "epoch": 3.4607929515418503,
+      "grad_norm": 0.1758741605370076,
+      "learning_rate": 0.00019917960077126757,
+      "loss": 0.356,
+      "step": 491
+    },
+    {
+      "epoch": 3.46784140969163,
+      "grad_norm": 0.1314237955816173,
+      "learning_rate": 0.00019917371736063948,
+      "loss": 0.3394,
+      "step": 492
+    },
+    {
+      "epoch": 3.4748898678414095,
+      "grad_norm": 0.14274095708246945,
+      "learning_rate": 0.00019916781301669017,
+      "loss": 0.3598,
+      "step": 493
+    },
+    {
+      "epoch": 3.4819383259911896,
+      "grad_norm": 0.1391854135273465,
+      "learning_rate": 0.00019916188774066594,
+      "loss": 0.3464,
+      "step": 494
+    },
+    {
+      "epoch": 3.488986784140969,
+      "grad_norm": 0.14770463685908689,
+      "learning_rate": 0.0001991559415338175,
+      "loss": 0.3503,
+      "step": 495
+    },
+    {
+      "epoch": 3.496035242290749,
+      "grad_norm": 0.13862086505922713,
+      "learning_rate": 0.00019914997439739995,
+      "loss": 0.3358,
+      "step": 496
+    },
+    {
+      "epoch": 3.503083700440529,
+      "grad_norm": 0.13855900777553032,
+      "learning_rate": 0.00019914398633267278,
+      "loss": 0.3465,
+      "step": 497
+    },
+    {
+      "epoch": 3.5101321585903085,
+      "grad_norm": 0.12569576936370264,
+      "learning_rate": 0.00019913797734089997,
+      "loss": 0.3393,
+      "step": 498
+    },
+    {
+      "epoch": 3.517180616740088,
+      "grad_norm": 0.11989343762451682,
+      "learning_rate": 0.00019913194742334986,
+      "loss": 0.3506,
+      "step": 499
+    },
+    {
+      "epoch": 3.5242290748898677,
+      "grad_norm": 0.1296946176729515,
+      "learning_rate": 0.00019912589658129527,
+      "loss": 0.3548,
+      "step": 500
+    },
+    {
+      "epoch": 3.5312775330396473,
+      "grad_norm": 0.13922785207087238,
+      "learning_rate": 0.00019911982481601336,
+      "loss": 0.3669,
+      "step": 501
+    },
+    {
+      "epoch": 3.5383259911894274,
+      "grad_norm": 0.12270073212237259,
+      "learning_rate": 0.0001991137321287857,
+      "loss": 0.3542,
+      "step": 502
+    },
+    {
+      "epoch": 3.545374449339207,
+      "grad_norm": 0.12588454514475536,
+      "learning_rate": 0.00019910761852089844,
+      "loss": 0.3548,
+      "step": 503
+    },
+    {
+      "epoch": 3.552422907488987,
+      "grad_norm": 0.1699412899807686,
+      "learning_rate": 0.0001991014839936419,
+      "loss": 0.363,
+      "step": 504
+    },
+    {
+      "epoch": 3.5594713656387666,
+      "grad_norm": 0.14602483719163561,
+      "learning_rate": 0.00019909532854831102,
+      "loss": 0.3449,
+      "step": 505
+    },
+    {
+      "epoch": 3.5665198237885463,
+      "grad_norm": 0.17179888689875858,
+      "learning_rate": 0.00019908915218620502,
+      "loss": 0.3681,
+      "step": 506
+    },
+    {
+      "epoch": 3.573568281938326,
+      "grad_norm": 0.1659790245901169,
+      "learning_rate": 0.00019908295490862762,
+      "loss": 0.3496,
+      "step": 507
+    },
+    {
+      "epoch": 3.5806167400881055,
+      "grad_norm": 0.13697368707022226,
+      "learning_rate": 0.00019907673671688695,
+      "loss": 0.3451,
+      "step": 508
+    },
+    {
+      "epoch": 3.5876651982378855,
+      "grad_norm": 0.1834744935935674,
+      "learning_rate": 0.00019907049761229548,
+      "loss": 0.3518,
+      "step": 509
+    },
+    {
+      "epoch": 3.594713656387665,
+      "grad_norm": 0.12028171037806093,
+      "learning_rate": 0.00019906423759617016,
+      "loss": 0.3496,
+      "step": 510
+    },
+    {
+      "epoch": 3.6017621145374448,
+      "grad_norm": 0.15902915171851317,
+      "learning_rate": 0.00019905795666983234,
+      "loss": 0.3573,
+      "step": 511
+    },
+    {
+      "epoch": 3.608810572687225,
+      "grad_norm": 0.13560741138413204,
+      "learning_rate": 0.0001990516548346078,
+      "loss": 0.3544,
+      "step": 512
+    },
+    {
+      "epoch": 3.6158590308370044,
+      "grad_norm": 0.13489578383832385,
+      "learning_rate": 0.00019904533209182668,
+      "loss": 0.3598,
+      "step": 513
+    },
+    {
+      "epoch": 3.622907488986784,
+      "grad_norm": 0.13169763401669024,
+      "learning_rate": 0.00019903898844282356,
+      "loss": 0.3601,
+      "step": 514
+    },
+    {
+      "epoch": 3.6299559471365637,
+      "grad_norm": 0.14953022987467862,
+      "learning_rate": 0.00019903262388893748,
+      "loss": 0.351,
+      "step": 515
+    },
+    {
+      "epoch": 3.6370044052863437,
+      "grad_norm": 0.12910876130007612,
+      "learning_rate": 0.00019902623843151185,
+      "loss": 0.363,
+      "step": 516
+    },
+    {
+      "epoch": 3.6440528634361233,
+      "grad_norm": 0.11968591704293234,
+      "learning_rate": 0.00019901983207189447,
+      "loss": 0.3622,
+      "step": 517
+    },
+    {
+      "epoch": 3.651101321585903,
+      "grad_norm": 0.12606179549567414,
+      "learning_rate": 0.0001990134048114376,
+      "loss": 0.354,
+      "step": 518
+    },
+    {
+      "epoch": 3.658149779735683,
+      "grad_norm": 0.11757025415129234,
+      "learning_rate": 0.00019900695665149787,
+      "loss": 0.3513,
+      "step": 519
+    },
+    {
+      "epoch": 3.6651982378854626,
+      "grad_norm": 0.12736766478000686,
+      "learning_rate": 0.00019900048759343632,
+      "loss": 0.3518,
+      "step": 520
+    },
+    {
+      "epoch": 3.6722466960352422,
+      "grad_norm": 0.12693223983659754,
+      "learning_rate": 0.00019899399763861848,
+      "loss": 0.3461,
+      "step": 521
+    },
+    {
+      "epoch": 3.679295154185022,
+      "grad_norm": 0.14875195918095674,
+      "learning_rate": 0.00019898748678841418,
+      "loss": 0.3614,
+      "step": 522
+    },
+    {
+      "epoch": 3.686343612334802,
+      "grad_norm": 0.14217386654496456,
+      "learning_rate": 0.00019898095504419775,
+      "loss": 0.3509,
+      "step": 523
+    },
+    {
+      "epoch": 3.6933920704845815,
+      "grad_norm": 0.12912305277499767,
+      "learning_rate": 0.00019897440240734788,
+      "loss": 0.3389,
+      "step": 524
+    },
+    {
+      "epoch": 3.700440528634361,
+      "grad_norm": 0.12804200179338152,
+      "learning_rate": 0.00019896782887924766,
+      "loss": 0.3579,
+      "step": 525
+    },
+    {
+      "epoch": 3.707488986784141,
+      "grad_norm": 0.14050896818939093,
+      "learning_rate": 0.00019896123446128466,
+      "loss": 0.3379,
+      "step": 526
+    },
+    {
+      "epoch": 3.714537444933921,
+      "grad_norm": 0.15080844076700628,
+      "learning_rate": 0.00019895461915485075,
+      "loss": 0.355,
+      "step": 527
+    },
+    {
+      "epoch": 3.7215859030837004,
+      "grad_norm": 0.1505437747249761,
+      "learning_rate": 0.00019894798296134236,
+      "loss": 0.3522,
+      "step": 528
+    },
+    {
+      "epoch": 3.72863436123348,
+      "grad_norm": 0.14312222207611736,
+      "learning_rate": 0.00019894132588216016,
+      "loss": 0.3485,
+      "step": 529
+    },
+    {
+      "epoch": 3.73568281938326,
+      "grad_norm": 0.17757466167620187,
+      "learning_rate": 0.0001989346479187093,
+      "loss": 0.3548,
+      "step": 530
+    },
+    {
+      "epoch": 3.7427312775330397,
+      "grad_norm": 0.23616044477961684,
+      "learning_rate": 0.00019892794907239942,
+      "loss": 0.3595,
+      "step": 531
+    },
+    {
+      "epoch": 3.7497797356828193,
+      "grad_norm": 0.1662925044047658,
+      "learning_rate": 0.00019892122934464442,
+      "loss": 0.3565,
+      "step": 532
+    },
+    {
+      "epoch": 3.7568281938325994,
+      "grad_norm": 0.19221313757150735,
+      "learning_rate": 0.00019891448873686276,
+      "loss": 0.3649,
+      "step": 533
+    },
+    {
+      "epoch": 3.763876651982379,
+      "grad_norm": 0.1627404411641412,
+      "learning_rate": 0.00019890772725047714,
+      "loss": 0.3387,
+      "step": 534
+    },
+    {
+      "epoch": 3.7709251101321586,
+      "grad_norm": 0.16075726209592892,
+      "learning_rate": 0.00019890094488691484,
+      "loss": 0.362,
+      "step": 535
+    },
+    {
+      "epoch": 3.777973568281938,
+      "grad_norm": 0.15330025157004837,
+      "learning_rate": 0.0001988941416476074,
+      "loss": 0.3419,
+      "step": 536
+    },
+    {
+      "epoch": 3.785022026431718,
+      "grad_norm": 0.17144739593584854,
+      "learning_rate": 0.00019888731753399088,
+      "loss": 0.3597,
+      "step": 537
+    },
+    {
+      "epoch": 3.792070484581498,
+      "grad_norm": 0.14878191307233052,
+      "learning_rate": 0.00019888047254750565,
+      "loss": 0.3424,
+      "step": 538
+    },
+    {
+      "epoch": 3.7991189427312775,
+      "grad_norm": 0.431999724838745,
+      "learning_rate": 0.00019887360668959656,
+      "loss": 0.3516,
+      "step": 539
+    },
+    {
+      "epoch": 3.8061674008810575,
+      "grad_norm": 0.13746572785145442,
+      "learning_rate": 0.00019886671996171282,
+      "loss": 0.3521,
+      "step": 540
+    },
+    {
+      "epoch": 3.813215859030837,
+      "grad_norm": 0.1599228003590644,
+      "learning_rate": 0.00019885981236530807,
+      "loss": 0.3464,
+      "step": 541
+    },
+    {
+      "epoch": 3.8202643171806168,
+      "grad_norm": 0.16192538140303794,
+      "learning_rate": 0.00019885288390184032,
+      "loss": 0.3482,
+      "step": 542
+    },
+    {
+      "epoch": 3.8273127753303964,
+      "grad_norm": 0.1497361691115734,
+      "learning_rate": 0.0001988459345727721,
+      "loss": 0.3515,
+      "step": 543
+    },
+    {
+      "epoch": 3.834361233480176,
+      "grad_norm": 0.1876021820566995,
+      "learning_rate": 0.00019883896437957013,
+      "loss": 0.3572,
+      "step": 544
+    },
+    {
+      "epoch": 3.841409691629956,
+      "grad_norm": 0.15225680645744552,
+      "learning_rate": 0.00019883197332370577,
+      "loss": 0.3448,
+      "step": 545
+    },
+    {
+      "epoch": 3.8484581497797357,
+      "grad_norm": 0.1873566853756614,
+      "learning_rate": 0.00019882496140665462,
+      "loss": 0.3475,
+      "step": 546
+    },
+    {
+      "epoch": 3.8555066079295153,
+      "grad_norm": 0.11783665954572549,
+      "learning_rate": 0.00019881792862989675,
+      "loss": 0.3369,
+      "step": 547
+    },
+    {
+      "epoch": 3.8625550660792953,
+      "grad_norm": 0.14634929139344885,
+      "learning_rate": 0.00019881087499491659,
+      "loss": 0.3333,
+      "step": 548
+    },
+    {
+      "epoch": 3.869603524229075,
+      "grad_norm": 0.13682127033051972,
+      "learning_rate": 0.00019880380050320303,
+      "loss": 0.3538,
+      "step": 549
+    },
+    {
+      "epoch": 3.8766519823788546,
+      "grad_norm": 0.14563699429298682,
+      "learning_rate": 0.00019879670515624936,
+      "loss": 0.3547,
+      "step": 550
+    },
+    {
+      "epoch": 3.883700440528634,
+      "grad_norm": 0.13877917134526394,
+      "learning_rate": 0.0001987895889555532,
+      "loss": 0.349,
+      "step": 551
+    },
+    {
+      "epoch": 3.8907488986784142,
+      "grad_norm": 0.12926038033653905,
+      "learning_rate": 0.00019878245190261661,
+      "loss": 0.3369,
+      "step": 552
+    },
+    {
+      "epoch": 3.897797356828194,
+      "grad_norm": 0.16148343151709782,
+      "learning_rate": 0.0001987752939989461,
+      "loss": 0.3545,
+      "step": 553
+    },
+    {
+      "epoch": 3.9048458149779735,
+      "grad_norm": 0.1368034584794499,
+      "learning_rate": 0.00019876811524605255,
+      "loss": 0.3489,
+      "step": 554
+    },
+    {
+      "epoch": 3.9118942731277535,
+      "grad_norm": 0.17224748477181898,
+      "learning_rate": 0.00019876091564545117,
+      "loss": 0.3458,
+      "step": 555
+    },
+    {
+      "epoch": 3.918942731277533,
+      "grad_norm": 0.11554618178874848,
+      "learning_rate": 0.00019875369519866168,
+      "loss": 0.3405,
+      "step": 556
+    },
+    {
+      "epoch": 3.9259911894273127,
+      "grad_norm": 0.13804484718117238,
+      "learning_rate": 0.00019874645390720815,
+      "loss": 0.3417,
+      "step": 557
+    },
+    {
+      "epoch": 3.9330396475770923,
+      "grad_norm": 0.1375259840883439,
+      "learning_rate": 0.00019873919177261905,
+      "loss": 0.3644,
+      "step": 558
+    },
+    {
+      "epoch": 3.9400881057268724,
+      "grad_norm": 0.16016309993548597,
+      "learning_rate": 0.00019873190879642728,
+      "loss": 0.3605,
+      "step": 559
+    },
+    {
+      "epoch": 3.947136563876652,
+      "grad_norm": 0.15139039321173892,
+      "learning_rate": 0.00019872460498017002,
+      "loss": 0.3558,
+      "step": 560
+    },
+    {
+      "epoch": 3.9541850220264316,
+      "grad_norm": 0.14308943659481185,
+      "learning_rate": 0.00019871728032538904,
+      "loss": 0.3514,
+      "step": 561
+    },
+    {
+      "epoch": 3.9612334801762117,
+      "grad_norm": 0.16176696484423955,
+      "learning_rate": 0.00019870993483363033,
+      "loss": 0.3449,
+      "step": 562
+    },
+    {
+      "epoch": 3.9682819383259913,
+      "grad_norm": 0.13364286032748776,
+      "learning_rate": 0.0001987025685064444,
+      "loss": 0.3517,
+      "step": 563
+    },
+    {
+      "epoch": 3.975330396475771,
+      "grad_norm": 0.14585469946335297,
+      "learning_rate": 0.00019869518134538612,
+      "loss": 0.3489,
+      "step": 564
+    },
+    {
+      "epoch": 3.9823788546255505,
+      "grad_norm": 0.1617030530915135,
+      "learning_rate": 0.00019868777335201474,
+      "loss": 0.352,
+      "step": 565
+    },
+    {
+      "epoch": 3.98942731277533,
+      "grad_norm": 0.16645695807563093,
+      "learning_rate": 0.00019868034452789391,
+      "loss": 0.3563,
+      "step": 566
+    },
+    {
+      "epoch": 3.99647577092511,
+      "grad_norm": 0.1310547089730212,
+      "learning_rate": 0.00019867289487459172,
+      "loss": 0.3437,
+      "step": 567
+    },
+    {
+      "epoch": 4.00352422907489,
+      "grad_norm": 0.15986521507689078,
+      "learning_rate": 0.00019866542439368057,
+      "loss": 0.3555,
+      "step": 568
+    },
+    {
+      "epoch": 4.01057268722467,
+      "grad_norm": 0.13749654457904367,
+      "learning_rate": 0.00019865793308673736,
+      "loss": 0.3456,
+      "step": 569
+    },
+    {
+      "epoch": 4.0176211453744495,
+      "grad_norm": 0.17678445635111953,
+      "learning_rate": 0.0001986504209553433,
+      "loss": 0.3383,
+      "step": 570
+    },
+    {
+      "epoch": 4.024669603524229,
+      "grad_norm": 0.13239936293959725,
+      "learning_rate": 0.00019864288800108403,
+      "loss": 0.3405,
+      "step": 571
+    },
+    {
+      "epoch": 4.031718061674009,
+      "grad_norm": 0.15356082006225058,
+      "learning_rate": 0.0001986353342255496,
+      "loss": 0.3284,
+      "step": 572
+    },
+    {
+      "epoch": 4.038766519823788,
+      "grad_norm": 0.14365656670189292,
+      "learning_rate": 0.00019862775963033443,
+      "loss": 0.3349,
+      "step": 573
+    },
+    {
+      "epoch": 4.045814977973568,
+      "grad_norm": 0.13905031563562378,
+      "learning_rate": 0.00019862016421703737,
+      "loss": 0.3485,
+      "step": 574
+    },
+    {
+      "epoch": 4.052863436123348,
+      "grad_norm": 0.12806573463189697,
+      "learning_rate": 0.00019861254798726164,
+      "loss": 0.3474,
+      "step": 575
+    },
+    {
+      "epoch": 4.059911894273128,
+      "grad_norm": 0.17530740538836045,
+      "learning_rate": 0.0001986049109426148,
+      "loss": 0.3387,
+      "step": 576
+    },
+    {
+      "epoch": 4.066960352422908,
+      "grad_norm": 0.1276855890391535,
+      "learning_rate": 0.0001985972530847089,
+      "loss": 0.3278,
+      "step": 577
+    },
+    {
+      "epoch": 4.074008810572687,
+      "grad_norm": 0.15408017568408242,
+      "learning_rate": 0.00019858957441516034,
+      "loss": 0.3381,
+      "step": 578
+    },
+    {
+      "epoch": 4.081057268722467,
+      "grad_norm": 0.132634801797325,
+      "learning_rate": 0.00019858187493558994,
+      "loss": 0.3444,
+      "step": 579
+    },
+    {
+      "epoch": 4.0881057268722465,
+      "grad_norm": 0.1433695355175549,
+      "learning_rate": 0.00019857415464762278,
+      "loss": 0.3442,
+      "step": 580
+    },
+    {
+      "epoch": 4.095154185022026,
+      "grad_norm": 0.13237600513523282,
+      "learning_rate": 0.00019856641355288856,
+      "loss": 0.3405,
+      "step": 581
+    },
+    {
+      "epoch": 4.102202643171806,
+      "grad_norm": 0.12768964298145086,
+      "learning_rate": 0.00019855865165302122,
+      "loss": 0.3345,
+      "step": 582
+    },
+    {
+      "epoch": 4.109251101321586,
+      "grad_norm": 0.13347865013200205,
+      "learning_rate": 0.00019855086894965903,
+      "loss": 0.3418,
+      "step": 583
+    },
+    {
+      "epoch": 4.116299559471366,
+      "grad_norm": 0.13009768143073688,
+      "learning_rate": 0.00019854306544444487,
+      "loss": 0.3417,
+      "step": 584
+    },
+    {
+      "epoch": 4.1233480176211454,
+      "grad_norm": 0.16926072391146016,
+      "learning_rate": 0.0001985352411390258,
+      "loss": 0.3437,
+      "step": 585
+    },
+    {
+      "epoch": 4.130396475770925,
+      "grad_norm": 0.1420329345177529,
+      "learning_rate": 0.00019852739603505338,
+      "loss": 0.3413,
+      "step": 586
+    },
+    {
+      "epoch": 4.137444933920705,
+      "grad_norm": 0.14142229163906694,
+      "learning_rate": 0.00019851953013418351,
+      "loss": 0.3448,
+      "step": 587
+    },
+    {
+      "epoch": 4.144493392070484,
+      "grad_norm": 0.1505358994130672,
+      "learning_rate": 0.00019851164343807657,
+      "loss": 0.3255,
+      "step": 588
+    },
+    {
+      "epoch": 4.151541850220264,
+      "grad_norm": 0.1489756404885626,
+      "learning_rate": 0.00019850373594839716,
+      "loss": 0.3364,
+      "step": 589
+    },
+    {
+      "epoch": 4.158590308370044,
+      "grad_norm": 0.14557176940100597,
+      "learning_rate": 0.00019849580766681448,
+      "loss": 0.3392,
+      "step": 590
+    },
+    {
+      "epoch": 4.165638766519824,
+      "grad_norm": 0.14896168525169645,
+      "learning_rate": 0.00019848785859500194,
+      "loss": 0.3429,
+      "step": 591
+    },
+    {
+      "epoch": 4.172687224669604,
+      "grad_norm": 0.15680829292724766,
+      "learning_rate": 0.00019847988873463742,
+      "loss": 0.3424,
+      "step": 592
+    },
+    {
+      "epoch": 4.179735682819383,
+      "grad_norm": 0.13342187463647795,
+      "learning_rate": 0.0001984718980874032,
+      "loss": 0.339,
+      "step": 593
+    },
+    {
+      "epoch": 4.186784140969163,
+      "grad_norm": 0.13832211605710407,
+      "learning_rate": 0.0001984638866549859,
+      "loss": 0.3408,
+      "step": 594
+    },
+    {
+      "epoch": 4.1938325991189425,
+      "grad_norm": 0.1266829211570311,
+      "learning_rate": 0.00019845585443907656,
+      "loss": 0.3383,
+      "step": 595
+    },
+    {
+      "epoch": 4.200881057268722,
+      "grad_norm": 0.13481837029908125,
+      "learning_rate": 0.00019844780144137058,
+      "loss": 0.3274,
+      "step": 596
+    },
+    {
+      "epoch": 4.207929515418503,
+      "grad_norm": 0.13717166945254827,
+      "learning_rate": 0.0001984397276635678,
+      "loss": 0.3405,
+      "step": 597
+    },
+    {
+      "epoch": 4.214977973568282,
+      "grad_norm": 0.18778684405144833,
+      "learning_rate": 0.00019843163310737238,
+      "loss": 0.3335,
+      "step": 598
+    },
+    {
+      "epoch": 4.222026431718062,
+      "grad_norm": 0.23945087314668545,
+      "learning_rate": 0.00019842351777449294,
+      "loss": 0.3533,
+      "step": 599
+    },
+    {
+      "epoch": 4.229074889867841,
+      "grad_norm": 0.16759233170316265,
+      "learning_rate": 0.0001984153816666424,
+      "loss": 0.344,
+      "step": 600
+    },
+    {
+      "epoch": 4.236123348017621,
+      "grad_norm": 0.15612415207801758,
+      "learning_rate": 0.00019840722478553814,
+      "loss": 0.3441,
+      "step": 601
+    },
+    {
+      "epoch": 4.243171806167401,
+      "grad_norm": 0.14424576253025784,
+      "learning_rate": 0.00019839904713290184,
+      "loss": 0.3344,
+      "step": 602
+    },
+    {
+      "epoch": 4.25022026431718,
+      "grad_norm": 0.16122575097999894,
+      "learning_rate": 0.0001983908487104597,
+      "loss": 0.3387,
+      "step": 603
+    },
+    {
+      "epoch": 4.257268722466961,
+      "grad_norm": 0.1684476009751477,
+      "learning_rate": 0.0001983826295199421,
+      "loss": 0.3472,
+      "step": 604
+    },
+    {
+      "epoch": 4.26431718061674,
+      "grad_norm": 0.1431727617666872,
+      "learning_rate": 0.00019837438956308404,
+      "loss": 0.3342,
+      "step": 605
+    },
+    {
+      "epoch": 4.27136563876652,
+      "grad_norm": 0.17136033819912236,
+      "learning_rate": 0.00019836612884162477,
+      "loss": 0.3472,
+      "step": 606
+    },
+    {
+      "epoch": 4.2784140969163,
+      "grad_norm": 0.14862328338982608,
+      "learning_rate": 0.00019835784735730788,
+      "loss": 0.3475,
+      "step": 607
+    },
+    {
+      "epoch": 4.285462555066079,
+      "grad_norm": 0.1434755378565175,
+      "learning_rate": 0.00019834954511188145,
+      "loss": 0.3379,
+      "step": 608
+    },
+    {
+      "epoch": 4.292511013215859,
+      "grad_norm": 0.13112029355376145,
+      "learning_rate": 0.0001983412221070979,
+      "loss": 0.3328,
+      "step": 609
+    },
+    {
+      "epoch": 4.299559471365638,
+      "grad_norm": 0.12399074023472105,
+      "learning_rate": 0.000198332878344714,
+      "loss": 0.3454,
+      "step": 610
+    },
+    {
+      "epoch": 4.306607929515419,
+      "grad_norm": 0.1287793838995882,
+      "learning_rate": 0.0001983245138264909,
+      "loss": 0.3335,
+      "step": 611
+    },
+    {
+      "epoch": 4.3136563876651985,
+      "grad_norm": 0.12803612362585884,
+      "learning_rate": 0.00019831612855419426,
+      "loss": 0.3338,
+      "step": 612
+    },
+    {
+      "epoch": 4.320704845814978,
+      "grad_norm": 0.14219328337203224,
+      "learning_rate": 0.00019830772252959393,
+      "loss": 0.3416,
+      "step": 613
+    },
+    {
+      "epoch": 4.327753303964758,
+      "grad_norm": 0.13383529104556638,
+      "learning_rate": 0.00019829929575446427,
+      "loss": 0.3328,
+      "step": 614
+    },
+    {
+      "epoch": 4.334801762114537,
+      "grad_norm": 0.15451024915637282,
+      "learning_rate": 0.000198290848230584,
+      "loss": 0.3409,
+      "step": 615
+    },
+    {
+      "epoch": 4.341850220264317,
+      "grad_norm": 0.15021952161769098,
+      "learning_rate": 0.00019828237995973616,
+      "loss": 0.3312,
+      "step": 616
+    },
+    {
+      "epoch": 4.348898678414097,
+      "grad_norm": 0.13439980884795236,
+      "learning_rate": 0.00019827389094370822,
+      "loss": 0.3274,
+      "step": 617
+    },
+    {
+      "epoch": 4.355947136563877,
+      "grad_norm": 0.12179827685620843,
+      "learning_rate": 0.000198265381184292,
+      "loss": 0.3468,
+      "step": 618
+    },
+    {
+      "epoch": 4.362995594713657,
+      "grad_norm": 0.13298175644517146,
+      "learning_rate": 0.00019825685068328375,
+      "loss": 0.3359,
+      "step": 619
+    },
+    {
+      "epoch": 4.370044052863436,
+      "grad_norm": 0.17362833434202146,
+      "learning_rate": 0.00019824829944248406,
+      "loss": 0.3386,
+      "step": 620
+    },
+    {
+      "epoch": 4.377092511013216,
+      "grad_norm": 0.12945482571771094,
+      "learning_rate": 0.00019823972746369793,
+      "loss": 0.3499,
+      "step": 621
+    },
+    {
+      "epoch": 4.384140969162996,
+      "grad_norm": 0.15689971725918941,
+      "learning_rate": 0.00019823113474873462,
+      "loss": 0.3376,
+      "step": 622
+    },
+    {
+      "epoch": 4.391189427312775,
+      "grad_norm": 0.12522564475422812,
+      "learning_rate": 0.00019822252129940795,
+      "loss": 0.3447,
+      "step": 623
+    },
+    {
+      "epoch": 4.398237885462555,
+      "grad_norm": 0.16499331122761735,
+      "learning_rate": 0.000198213887117536,
+      "loss": 0.3433,
+      "step": 624
+    },
+    {
+      "epoch": 4.405286343612334,
+      "grad_norm": 0.14079073159682415,
+      "learning_rate": 0.0001982052322049412,
+      "loss": 0.3386,
+      "step": 625
+    },
+    {
+      "epoch": 4.412334801762115,
+      "grad_norm": 0.13965094808537842,
+      "learning_rate": 0.00019819655656345048,
+      "loss": 0.3338,
+      "step": 626
+    },
+    {
+      "epoch": 4.4193832599118945,
+      "grad_norm": 0.1528507960425957,
+      "learning_rate": 0.00019818786019489503,
+      "loss": 0.3363,
+      "step": 627
+    },
+    {
+      "epoch": 4.426431718061674,
+      "grad_norm": 0.13725273570394939,
+      "learning_rate": 0.00019817914310111046,
+      "loss": 0.3434,
+      "step": 628
+    },
+    {
+      "epoch": 4.433480176211454,
+      "grad_norm": 0.14039191740865162,
+      "learning_rate": 0.00019817040528393674,
+      "loss": 0.3382,
+      "step": 629
+    },
+    {
+      "epoch": 4.440528634361233,
+      "grad_norm": 0.12383051238259699,
+      "learning_rate": 0.00019816164674521825,
+      "loss": 0.3485,
+      "step": 630
+    },
+    {
+      "epoch": 4.447577092511013,
+      "grad_norm": 0.1567242542602749,
+      "learning_rate": 0.00019815286748680371,
+      "loss": 0.3453,
+      "step": 631
+    },
+    {
+      "epoch": 4.454625550660793,
+      "grad_norm": 0.1325732941625189,
+      "learning_rate": 0.0001981440675105462,
+      "loss": 0.3419,
+      "step": 632
+    },
+    {
+      "epoch": 4.461674008810573,
+      "grad_norm": 0.17616159878266133,
+      "learning_rate": 0.0001981352468183032,
+      "loss": 0.3317,
+      "step": 633
+    },
+    {
+      "epoch": 4.468722466960353,
+      "grad_norm": 0.13117354963154138,
+      "learning_rate": 0.00019812640541193663,
+      "loss": 0.3353,
+      "step": 634
+    },
+    {
+      "epoch": 4.475770925110132,
+      "grad_norm": 0.13464180581277416,
+      "learning_rate": 0.00019811754329331261,
+      "loss": 0.3309,
+      "step": 635
+    },
+    {
+      "epoch": 4.482819383259912,
+      "grad_norm": 0.12883199530207956,
+      "learning_rate": 0.00019810866046430177,
+      "loss": 0.3416,
+      "step": 636
+    },
+    {
+      "epoch": 4.4898678414096915,
+      "grad_norm": 0.14800882211741093,
+      "learning_rate": 0.00019809975692677912,
+      "loss": 0.3464,
+      "step": 637
+    },
+    {
+      "epoch": 4.496916299559471,
+      "grad_norm": 0.1297185728257677,
+      "learning_rate": 0.00019809083268262388,
+      "loss": 0.3479,
+      "step": 638
+    },
+    {
+      "epoch": 4.503964757709251,
+      "grad_norm": 0.13842097611621798,
+      "learning_rate": 0.00019808188773371988,
+      "loss": 0.3495,
+      "step": 639
+    },
+    {
+      "epoch": 4.51101321585903,
+      "grad_norm": 0.1494631975025016,
+      "learning_rate": 0.00019807292208195513,
+      "loss": 0.3472,
+      "step": 640
+    },
+    {
+      "epoch": 4.518061674008811,
+      "grad_norm": 0.1286629092849426,
+      "learning_rate": 0.0001980639357292221,
+      "loss": 0.3292,
+      "step": 641
+    },
+    {
+      "epoch": 4.5251101321585905,
+      "grad_norm": 0.17073986607442304,
+      "learning_rate": 0.0001980549286774176,
+      "loss": 0.3518,
+      "step": 642
+    },
+    {
+      "epoch": 4.53215859030837,
+      "grad_norm": 0.16181723405077744,
+      "learning_rate": 0.00019804590092844276,
+      "loss": 0.3329,
+      "step": 643
+    },
+    {
+      "epoch": 4.53920704845815,
+      "grad_norm": 0.13025819649171552,
+      "learning_rate": 0.00019803685248420325,
+      "loss": 0.324,
+      "step": 644
+    },
+    {
+      "epoch": 4.546255506607929,
+      "grad_norm": 0.1619999420748522,
+      "learning_rate": 0.00019802778334660889,
+      "loss": 0.3536,
+      "step": 645
+    },
+    {
+      "epoch": 4.553303964757709,
+      "grad_norm": 0.13299552560663988,
+      "learning_rate": 0.000198018693517574,
+      "loss": 0.3365,
+      "step": 646
+    },
+    {
+      "epoch": 4.5603524229074885,
+      "grad_norm": 0.12853095610352794,
+      "learning_rate": 0.00019800958299901722,
+      "loss": 0.3346,
+      "step": 647
+    },
+    {
+      "epoch": 4.567400881057269,
+      "grad_norm": 0.15229705465940974,
+      "learning_rate": 0.00019800045179286163,
+      "loss": 0.3329,
+      "step": 648
+    },
+    {
+      "epoch": 4.574449339207049,
+      "grad_norm": 0.12617924319663512,
+      "learning_rate": 0.00019799129990103454,
+      "loss": 0.3317,
+      "step": 649
+    },
+    {
+      "epoch": 4.581497797356828,
+      "grad_norm": 0.16013986809249456,
+      "learning_rate": 0.0001979821273254678,
+      "loss": 0.335,
+      "step": 650
+    },
+    {
+      "epoch": 4.588546255506608,
+      "grad_norm": 0.14530811625566117,
+      "learning_rate": 0.00019797293406809745,
+      "loss": 0.3331,
+      "step": 651
+    },
+    {
+      "epoch": 4.5955947136563875,
+      "grad_norm": 0.17337495053125943,
+      "learning_rate": 0.00019796372013086399,
+      "loss": 0.329,
+      "step": 652
+    },
+    {
+      "epoch": 4.602643171806167,
+      "grad_norm": 0.12473912113519772,
+      "learning_rate": 0.00019795448551571235,
+      "loss": 0.3367,
+      "step": 653
+    },
+    {
+      "epoch": 4.609691629955947,
+      "grad_norm": 0.15490332780485752,
+      "learning_rate": 0.00019794523022459166,
+      "loss": 0.3362,
+      "step": 654
+    },
+    {
+      "epoch": 4.616740088105727,
+      "grad_norm": 0.12012201505741764,
+      "learning_rate": 0.00019793595425945552,
+      "loss": 0.345,
+      "step": 655
+    },
+    {
+      "epoch": 4.623788546255507,
+      "grad_norm": 0.15926244442906665,
+      "learning_rate": 0.0001979266576222619,
+      "loss": 0.3488,
+      "step": 656
+    },
+    {
+      "epoch": 4.6308370044052865,
+      "grad_norm": 0.14018830100320284,
+      "learning_rate": 0.0001979173403149731,
+      "loss": 0.356,
+      "step": 657
+    },
+    {
+      "epoch": 4.637885462555066,
+      "grad_norm": 0.14184535866186337,
+      "learning_rate": 0.00019790800233955577,
+      "loss": 0.3247,
+      "step": 658
+    },
+    {
+      "epoch": 4.644933920704846,
+      "grad_norm": 0.1545107614113569,
+      "learning_rate": 0.000197898643697981,
+      "loss": 0.3302,
+      "step": 659
+    },
+    {
+      "epoch": 4.651982378854625,
+      "grad_norm": 0.14273978244099167,
+      "learning_rate": 0.0001978892643922241,
+      "loss": 0.3406,
+      "step": 660
+    },
+    {
+      "epoch": 4.659030837004405,
+      "grad_norm": 0.17105006952166502,
+      "learning_rate": 0.00019787986442426495,
+      "loss": 0.3441,
+      "step": 661
+    },
+    {
+      "epoch": 4.666079295154185,
+      "grad_norm": 0.13868457403482162,
+      "learning_rate": 0.00019787044379608753,
+      "loss": 0.3376,
+      "step": 662
+    },
+    {
+      "epoch": 4.673127753303965,
+      "grad_norm": 0.14605218707951037,
+      "learning_rate": 0.00019786100250968042,
+      "loss": 0.3361,
+      "step": 663
+    },
+    {
+      "epoch": 4.680176211453745,
+      "grad_norm": 0.1429198624184254,
+      "learning_rate": 0.00019785154056703642,
+      "loss": 0.3418,
+      "step": 664
+    },
+    {
+      "epoch": 4.687224669603524,
+      "grad_norm": 0.14013455586967552,
+      "learning_rate": 0.00019784205797015276,
+      "loss": 0.3303,
+      "step": 665
+    },
+    {
+      "epoch": 4.694273127753304,
+      "grad_norm": 0.16321131128960198,
+      "learning_rate": 0.000197832554721031,
+      "loss": 0.3365,
+      "step": 666
+    },
+    {
+      "epoch": 4.7013215859030835,
+      "grad_norm": 0.14041778393057627,
+      "learning_rate": 0.00019782303082167704,
+      "loss": 0.34,
+      "step": 667
+    },
+    {
+      "epoch": 4.708370044052863,
+      "grad_norm": 0.13098484154411447,
+      "learning_rate": 0.00019781348627410112,
+      "loss": 0.3358,
+      "step": 668
+    },
+    {
+      "epoch": 4.715418502202644,
+      "grad_norm": 0.1548274250082925,
+      "learning_rate": 0.00019780392108031795,
+      "loss": 0.3333,
+      "step": 669
+    },
+    {
+      "epoch": 4.722466960352423,
+      "grad_norm": 0.13590722277477593,
+      "learning_rate": 0.0001977943352423465,
+      "loss": 0.3318,
+      "step": 670
+    },
+    {
+      "epoch": 4.729515418502203,
+      "grad_norm": 0.12711873800086235,
+      "learning_rate": 0.0001977847287622101,
+      "loss": 0.3369,
+      "step": 671
+    },
+    {
+      "epoch": 4.736563876651982,
+      "grad_norm": 0.15026034684383405,
+      "learning_rate": 0.0001977751016419365,
+      "loss": 0.3438,
+      "step": 672
+    },
+    {
+      "epoch": 4.743612334801762,
+      "grad_norm": 0.14090632129167766,
+      "learning_rate": 0.00019776545388355775,
+      "loss": 0.3487,
+      "step": 673
+    },
+    {
+      "epoch": 4.750660792951542,
+      "grad_norm": 0.13382021925315754,
+      "learning_rate": 0.00019775578548911026,
+      "loss": 0.3327,
+      "step": 674
+    },
+    {
+      "epoch": 4.757709251101321,
+      "grad_norm": 0.14278364948992606,
+      "learning_rate": 0.0001977460964606348,
+      "loss": 0.3415,
+      "step": 675
+    },
+    {
+      "epoch": 4.764757709251102,
+      "grad_norm": 0.1362825986516914,
+      "learning_rate": 0.00019773638680017653,
+      "loss": 0.3361,
+      "step": 676
+    },
+    {
+      "epoch": 4.771806167400881,
+      "grad_norm": 0.14166725051075815,
+      "learning_rate": 0.00019772665650978494,
+      "loss": 0.3379,
+      "step": 677
+    },
+    {
+      "epoch": 4.778854625550661,
+      "grad_norm": 0.1645969489908101,
+      "learning_rate": 0.00019771690559151385,
+      "loss": 0.3313,
+      "step": 678
+    },
+    {
+      "epoch": 4.785903083700441,
+      "grad_norm": 0.1288819623257621,
+      "learning_rate": 0.0001977071340474215,
+      "loss": 0.3409,
+      "step": 679
+    },
+    {
+      "epoch": 4.79295154185022,
+      "grad_norm": 0.1426011541637681,
+      "learning_rate": 0.00019769734187957038,
+      "loss": 0.3457,
+      "step": 680
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.13554080364751472,
+      "learning_rate": 0.00019768752909002743,
+      "loss": 0.3516,
+      "step": 681
+    },
+    {
+      "epoch": 4.807048458149779,
+      "grad_norm": 0.1326703693917846,
+      "learning_rate": 0.0001976776956808639,
+      "loss": 0.3336,
+      "step": 682
+    },
+    {
+      "epoch": 4.81409691629956,
+      "grad_norm": 0.14681270422896903,
+      "learning_rate": 0.00019766784165415544,
+      "loss": 0.3408,
+      "step": 683
+    },
+    {
+      "epoch": 4.8211453744493395,
+      "grad_norm": 0.12548240573171893,
+      "learning_rate": 0.00019765796701198195,
+      "loss": 0.3372,
+      "step": 684
+    },
+    {
+      "epoch": 4.828193832599119,
+      "grad_norm": 0.14709870597970065,
+      "learning_rate": 0.00019764807175642776,
+      "loss": 0.3396,
+      "step": 685
+    },
+    {
+      "epoch": 4.835242290748899,
+      "grad_norm": 0.12057095312517414,
+      "learning_rate": 0.00019763815588958155,
+      "loss": 0.349,
+      "step": 686
+    },
+    {
+      "epoch": 4.842290748898678,
+      "grad_norm": 0.150211279260299,
+      "learning_rate": 0.00019762821941353634,
+      "loss": 0.3465,
+      "step": 687
+    },
+    {
+      "epoch": 4.849339207048458,
+      "grad_norm": 0.1254852858525492,
+      "learning_rate": 0.00019761826233038944,
+      "loss": 0.3384,
+      "step": 688
+    },
+    {
+      "epoch": 4.856387665198238,
+      "grad_norm": 0.1604867151044819,
+      "learning_rate": 0.00019760828464224266,
+      "loss": 0.3342,
+      "step": 689
+    },
+    {
+      "epoch": 4.863436123348018,
+      "grad_norm": 0.14176365744894867,
+      "learning_rate": 0.000197598286351202,
+      "loss": 0.3506,
+      "step": 690
+    },
+    {
+      "epoch": 4.870484581497798,
+      "grad_norm": 0.1772406041803444,
+      "learning_rate": 0.00019758826745937784,
+      "loss": 0.3388,
+      "step": 691
+    },
+    {
+      "epoch": 4.877533039647577,
+      "grad_norm": 0.1720790157181186,
+      "learning_rate": 0.00019757822796888502,
+      "loss": 0.3341,
+      "step": 692
+    },
+    {
+      "epoch": 4.884581497797357,
+      "grad_norm": 0.1976619045339488,
+      "learning_rate": 0.00019756816788184259,
+      "loss": 0.3348,
+      "step": 693
+    },
+    {
+      "epoch": 4.891629955947137,
+      "grad_norm": 0.14088360130961083,
+      "learning_rate": 0.00019755808720037403,
+      "loss": 0.3382,
+      "step": 694
+    },
+    {
+      "epoch": 4.898678414096916,
+      "grad_norm": 0.1780036252742257,
+      "learning_rate": 0.00019754798592660713,
+      "loss": 0.3427,
+      "step": 695
+    },
+    {
+      "epoch": 4.905726872246696,
+      "grad_norm": 0.13226995405907965,
+      "learning_rate": 0.00019753786406267407,
+      "loss": 0.3302,
+      "step": 696
+    },
+    {
+      "epoch": 4.912775330396475,
+      "grad_norm": 0.15402247928088628,
+      "learning_rate": 0.00019752772161071128,
+      "loss": 0.3324,
+      "step": 697
+    },
+    {
+      "epoch": 4.919823788546255,
+      "grad_norm": 0.14133667521778387,
+      "learning_rate": 0.0001975175585728597,
+      "loss": 0.3422,
+      "step": 698
+    },
+    {
+      "epoch": 4.9268722466960355,
+      "grad_norm": 0.1455325283895691,
+      "learning_rate": 0.00019750737495126446,
+      "loss": 0.3464,
+      "step": 699
+    },
+    {
+      "epoch": 4.933920704845815,
+      "grad_norm": 0.11927209203666213,
+      "learning_rate": 0.00019749717074807506,
+      "loss": 0.3222,
+      "step": 700
+    },
+    {
+      "epoch": 4.940969162995595,
+      "grad_norm": 0.1619236524213507,
+      "learning_rate": 0.00019748694596544543,
+      "loss": 0.3392,
+      "step": 701
+    },
+    {
+      "epoch": 4.948017621145374,
+      "grad_norm": 0.14509487532897372,
+      "learning_rate": 0.00019747670060553377,
+      "loss": 0.3459,
+      "step": 702
+    },
+    {
+      "epoch": 4.955066079295154,
+      "grad_norm": 0.14608003508299283,
+      "learning_rate": 0.00019746643467050264,
+      "loss": 0.338,
+      "step": 703
+    },
+    {
+      "epoch": 4.962114537444934,
+      "grad_norm": 0.14029793781681796,
+      "learning_rate": 0.00019745614816251897,
+      "loss": 0.3443,
+      "step": 704
+    },
+    {
+      "epoch": 4.969162995594713,
+      "grad_norm": 0.124041620654676,
+      "learning_rate": 0.00019744584108375395,
+      "loss": 0.3425,
+      "step": 705
+    },
+    {
+      "epoch": 4.976211453744494,
+      "grad_norm": 0.14429319391747017,
+      "learning_rate": 0.00019743551343638324,
+      "loss": 0.3384,
+      "step": 706
+    },
+    {
+      "epoch": 4.983259911894273,
+      "grad_norm": 0.12910998301466087,
+      "learning_rate": 0.0001974251652225867,
+      "loss": 0.3441,
+      "step": 707
+    },
+    {
+      "epoch": 4.990308370044053,
+      "grad_norm": 0.13665256844215032,
+      "learning_rate": 0.0001974147964445487,
+      "loss": 0.348,
+      "step": 708
+    },
+    {
+      "epoch": 4.9973568281938325,
+      "grad_norm": 0.12703919402450153,
+      "learning_rate": 0.00019740440710445777,
+      "loss": 0.3306,
+      "step": 709
+    },
+    {
+      "epoch": 5.004405286343612,
+      "grad_norm": 0.12584612328126202,
+      "learning_rate": 0.00019739399720450688,
+      "loss": 0.3336,
+      "step": 710
+    },
+    {
+      "epoch": 5.011453744493392,
+      "grad_norm": 0.1404289803295744,
+      "learning_rate": 0.00019738356674689332,
+      "loss": 0.3378,
+      "step": 711
+    },
+    {
+      "epoch": 5.018502202643171,
+      "grad_norm": 0.14670093158716982,
+      "learning_rate": 0.00019737311573381872,
+      "loss": 0.3328,
+      "step": 712
+    },
+    {
+      "epoch": 5.025550660792952,
+      "grad_norm": 0.20976333615597262,
+      "learning_rate": 0.00019736264416748912,
+      "loss": 0.3253,
+      "step": 713
+    },
+    {
+      "epoch": 5.0325991189427315,
+      "grad_norm": 0.17240034260070167,
+      "learning_rate": 0.00019735215205011477,
+      "loss": 0.3303,
+      "step": 714
+    },
+    {
+      "epoch": 5.039647577092511,
+      "grad_norm": 0.15250362631066305,
+      "learning_rate": 0.0001973416393839103,
+      "loss": 0.328,
+      "step": 715
+    },
+    {
+      "epoch": 5.046696035242291,
+      "grad_norm": 0.1530901163739243,
+      "learning_rate": 0.00019733110617109475,
+      "loss": 0.3243,
+      "step": 716
+    },
+    {
+      "epoch": 5.05374449339207,
+      "grad_norm": 0.14727087346921036,
+      "learning_rate": 0.00019732055241389142,
+      "loss": 0.3299,
+      "step": 717
+    },
+    {
+      "epoch": 5.06079295154185,
+      "grad_norm": 0.1319860856000918,
+      "learning_rate": 0.00019730997811452793,
+      "loss": 0.3093,
+      "step": 718
+    },
+    {
+      "epoch": 5.0678414096916296,
+      "grad_norm": 0.15098989324243975,
+      "learning_rate": 0.00019729938327523635,
+      "loss": 0.3174,
+      "step": 719
+    },
+    {
+      "epoch": 5.07488986784141,
+      "grad_norm": 0.14795030269485393,
+      "learning_rate": 0.00019728876789825293,
+      "loss": 0.3305,
+      "step": 720
+    },
+    {
+      "epoch": 5.08193832599119,
+      "grad_norm": 0.13395763128573818,
+      "learning_rate": 0.00019727813198581838,
+      "loss": 0.3265,
+      "step": 721
+    },
+    {
+      "epoch": 5.088986784140969,
+      "grad_norm": 0.13868191966506418,
+      "learning_rate": 0.00019726747554017768,
+      "loss": 0.3202,
+      "step": 722
+    },
+    {
+      "epoch": 5.096035242290749,
+      "grad_norm": 0.1374772007862514,
+      "learning_rate": 0.0001972567985635802,
+      "loss": 0.3196,
+      "step": 723
+    },
+    {
+      "epoch": 5.1030837004405285,
+      "grad_norm": 0.14761350135529958,
+      "learning_rate": 0.00019724610105827957,
+      "loss": 0.3342,
+      "step": 724
+    },
+    {
+      "epoch": 5.110132158590308,
+      "grad_norm": 0.14486927261156501,
+      "learning_rate": 0.0001972353830265338,
+      "loss": 0.3272,
+      "step": 725
+    },
+    {
+      "epoch": 5.117180616740088,
+      "grad_norm": 0.19430016086652288,
+      "learning_rate": 0.00019722464447060527,
+      "loss": 0.3294,
+      "step": 726
+    },
+    {
+      "epoch": 5.124229074889868,
+      "grad_norm": 0.20051518375777566,
+      "learning_rate": 0.00019721388539276056,
+      "loss": 0.3195,
+      "step": 727
+    },
+    {
+      "epoch": 5.131277533039648,
+      "grad_norm": 0.13660793339582702,
+      "learning_rate": 0.0001972031057952707,
+      "loss": 0.3316,
+      "step": 728
+    },
+    {
+      "epoch": 5.1383259911894275,
+      "grad_norm": 0.12881494575045607,
+      "learning_rate": 0.00019719230568041108,
+      "loss": 0.3266,
+      "step": 729
+    },
+    {
+      "epoch": 5.145374449339207,
+      "grad_norm": 0.15010953870453786,
+      "learning_rate": 0.00019718148505046124,
+      "loss": 0.3243,
+      "step": 730
+    },
+    {
+      "epoch": 5.152422907488987,
+      "grad_norm": 0.14639707186476988,
+      "learning_rate": 0.00019717064390770527,
+      "loss": 0.3122,
+      "step": 731
+    },
+    {
+      "epoch": 5.159471365638766,
+      "grad_norm": 0.2066192777318282,
+      "learning_rate": 0.00019715978225443148,
+      "loss": 0.33,
+      "step": 732
+    },
+    {
+      "epoch": 5.166519823788546,
+      "grad_norm": 0.1899217989872675,
+      "learning_rate": 0.00019714890009293247,
+      "loss": 0.3162,
+      "step": 733
+    },
+    {
+      "epoch": 5.173568281938326,
+      "grad_norm": 0.14214241643911005,
+      "learning_rate": 0.00019713799742550526,
+      "loss": 0.316,
+      "step": 734
+    },
+    {
+      "epoch": 5.180616740088106,
+      "grad_norm": 0.20187115987424822,
+      "learning_rate": 0.00019712707425445112,
+      "loss": 0.3141,
+      "step": 735
+    },
+    {
+      "epoch": 5.187665198237886,
+      "grad_norm": 0.1878724540233295,
+      "learning_rate": 0.00019711613058207573,
+      "loss": 0.3392,
+      "step": 736
+    },
+    {
+      "epoch": 5.194713656387665,
+      "grad_norm": 0.1545197101686749,
+      "learning_rate": 0.00019710516641068896,
+      "loss": 0.3319,
+      "step": 737
+    },
+    {
+      "epoch": 5.201762114537445,
+      "grad_norm": 0.19355163146151327,
+      "learning_rate": 0.0001970941817426052,
+      "loss": 0.3345,
+      "step": 738
+    },
+    {
+      "epoch": 5.2088105726872245,
+      "grad_norm": 0.1546431732568414,
+      "learning_rate": 0.00019708317658014303,
+      "loss": 0.332,
+      "step": 739
+    },
+    {
+      "epoch": 5.215859030837004,
+      "grad_norm": 0.15652222542802227,
+      "learning_rate": 0.0001970721509256254,
+      "loss": 0.3279,
+      "step": 740
+    },
+    {
+      "epoch": 5.222907488986785,
+      "grad_norm": 0.13857047863749453,
+      "learning_rate": 0.0001970611047813795,
+      "loss": 0.331,
+      "step": 741
+    },
+    {
+      "epoch": 5.229955947136564,
+      "grad_norm": 0.1417642592153288,
+      "learning_rate": 0.00019705003814973705,
+      "loss": 0.3366,
+      "step": 742
+    },
+    {
+      "epoch": 5.237004405286344,
+      "grad_norm": 0.15419950392330384,
+      "learning_rate": 0.00019703895103303384,
+      "loss": 0.3335,
+      "step": 743
+    },
+    {
+      "epoch": 5.244052863436123,
+      "grad_norm": 0.15611964867183278,
+      "learning_rate": 0.00019702784343361018,
+      "loss": 0.3302,
+      "step": 744
+    },
+    {
+      "epoch": 5.251101321585903,
+      "grad_norm": 0.15916916337841622,
+      "learning_rate": 0.00019701671535381064,
+      "loss": 0.3515,
+      "step": 745
+    },
+    {
+      "epoch": 5.258149779735683,
+      "grad_norm": 0.1409997150543297,
+      "learning_rate": 0.00019700556679598408,
+      "loss": 0.3318,
+      "step": 746
+    },
+    {
+      "epoch": 5.265198237885462,
+      "grad_norm": 0.13833725852228163,
+      "learning_rate": 0.00019699439776248368,
+      "loss": 0.3302,
+      "step": 747
+    },
+    {
+      "epoch": 5.272246696035242,
+      "grad_norm": 0.1376349288483988,
+      "learning_rate": 0.00019698320825566702,
+      "loss": 0.3262,
+      "step": 748
+    },
+    {
+      "epoch": 5.279295154185022,
+      "grad_norm": 0.1504135493908035,
+      "learning_rate": 0.0001969719982778959,
+      "loss": 0.3231,
+      "step": 749
+    },
+    {
+      "epoch": 5.286343612334802,
+      "grad_norm": 0.13971245656238435,
+      "learning_rate": 0.00019696076783153654,
+      "loss": 0.3374,
+      "step": 750
+    },
+    {
+      "epoch": 5.293392070484582,
+      "grad_norm": 0.14534162666530778,
+      "learning_rate": 0.00019694951691895943,
+      "loss": 0.3124,
+      "step": 751
+    },
+    {
+      "epoch": 5.300440528634361,
+      "grad_norm": 0.12824194417996532,
+      "learning_rate": 0.00019693824554253934,
+      "loss": 0.3256,
+      "step": 752
+    },
+    {
+      "epoch": 5.307488986784141,
+      "grad_norm": 0.14766814976589468,
+      "learning_rate": 0.00019692695370465541,
+      "loss": 0.3345,
+      "step": 753
+    },
+    {
+      "epoch": 5.31453744493392,
+      "grad_norm": 0.15341490332192545,
+      "learning_rate": 0.00019691564140769115,
+      "loss": 0.3387,
+      "step": 754
+    },
+    {
+      "epoch": 5.3215859030837,
+      "grad_norm": 0.15724301225823387,
+      "learning_rate": 0.00019690430865403427,
+      "loss": 0.3391,
+      "step": 755
+    },
+    {
+      "epoch": 5.3286343612334806,
+      "grad_norm": 0.16221148252125667,
+      "learning_rate": 0.00019689295544607685,
+      "loss": 0.3301,
+      "step": 756
+    },
+    {
+      "epoch": 5.33568281938326,
+      "grad_norm": 0.14737071735819507,
+      "learning_rate": 0.00019688158178621538,
+      "loss": 0.3276,
+      "step": 757
+    },
+    {
+      "epoch": 5.34273127753304,
+      "grad_norm": 0.1557993590750549,
+      "learning_rate": 0.0001968701876768505,
+      "loss": 0.3204,
+      "step": 758
+    },
+    {
+      "epoch": 5.349779735682819,
+      "grad_norm": 0.1879406649161191,
+      "learning_rate": 0.00019685877312038724,
+      "loss": 0.3314,
+      "step": 759
+    },
+    {
+      "epoch": 5.356828193832599,
+      "grad_norm": 0.162943039541616,
+      "learning_rate": 0.000196847338119235,
+      "loss": 0.345,
+      "step": 760
+    },
+    {
+      "epoch": 5.363876651982379,
+      "grad_norm": 0.15371765871144677,
+      "learning_rate": 0.0001968358826758074,
+      "loss": 0.327,
+      "step": 761
+    },
+    {
+      "epoch": 5.370925110132158,
+      "grad_norm": 0.1371125445938455,
+      "learning_rate": 0.0001968244067925225,
+      "loss": 0.3122,
+      "step": 762
+    },
+    {
+      "epoch": 5.377973568281939,
+      "grad_norm": 0.12764907109969378,
+      "learning_rate": 0.00019681291047180254,
+      "loss": 0.3152,
+      "step": 763
+    },
+    {
+      "epoch": 5.385022026431718,
+      "grad_norm": 0.1593864059815221,
+      "learning_rate": 0.00019680139371607415,
+      "loss": 0.3398,
+      "step": 764
+    },
+    {
+      "epoch": 5.392070484581498,
+      "grad_norm": 0.16207053933311322,
+      "learning_rate": 0.00019678985652776826,
+      "loss": 0.3343,
+      "step": 765
+    },
+    {
+      "epoch": 5.399118942731278,
+      "grad_norm": 0.15165542064412843,
+      "learning_rate": 0.00019677829890932006,
+      "loss": 0.3193,
+      "step": 766
+    },
+    {
+      "epoch": 5.406167400881057,
+      "grad_norm": 0.17092744261288387,
+      "learning_rate": 0.0001967667208631692,
+      "loss": 0.3145,
+      "step": 767
+    },
+    {
+      "epoch": 5.413215859030837,
+      "grad_norm": 0.16299532446233722,
+      "learning_rate": 0.00019675512239175944,
+      "loss": 0.3435,
+      "step": 768
+    },
+    {
+      "epoch": 5.420264317180616,
+      "grad_norm": 0.1744010885402697,
+      "learning_rate": 0.000196743503497539,
+      "loss": 0.3331,
+      "step": 769
+    },
+    {
+      "epoch": 5.427312775330396,
+      "grad_norm": 0.1562290309114667,
+      "learning_rate": 0.0001967318641829604,
+      "loss": 0.3194,
+      "step": 770
+    },
+    {
+      "epoch": 5.4343612334801765,
+      "grad_norm": 0.14587504278870783,
+      "learning_rate": 0.00019672020445048036,
+      "loss": 0.3246,
+      "step": 771
+    },
+    {
+      "epoch": 5.441409691629956,
+      "grad_norm": 0.15078609282567426,
+      "learning_rate": 0.00019670852430256005,
+      "loss": 0.3284,
+      "step": 772
+    },
+    {
+      "epoch": 5.448458149779736,
+      "grad_norm": 0.14541490331269302,
+      "learning_rate": 0.00019669682374166486,
+      "loss": 0.3233,
+      "step": 773
+    },
+    {
+      "epoch": 5.455506607929515,
+      "grad_norm": 0.16775832282734046,
+      "learning_rate": 0.00019668510277026448,
+      "loss": 0.3527,
+      "step": 774
+    },
+    {
+      "epoch": 5.462555066079295,
+      "grad_norm": 0.13840066340404278,
+      "learning_rate": 0.00019667336139083302,
+      "loss": 0.3188,
+      "step": 775
+    },
+    {
+      "epoch": 5.469603524229075,
+      "grad_norm": 0.1550792715974197,
+      "learning_rate": 0.00019666159960584874,
+      "loss": 0.3236,
+      "step": 776
+    },
+    {
+      "epoch": 5.476651982378854,
+      "grad_norm": 0.13392938330774307,
+      "learning_rate": 0.00019664981741779428,
+      "loss": 0.326,
+      "step": 777
+    },
+    {
+      "epoch": 5.483700440528635,
+      "grad_norm": 0.1707054905828015,
+      "learning_rate": 0.00019663801482915668,
+      "loss": 0.3223,
+      "step": 778
+    },
+    {
+      "epoch": 5.490748898678414,
+      "grad_norm": 0.13181247256077133,
+      "learning_rate": 0.00019662619184242715,
+      "loss": 0.3363,
+      "step": 779
+    },
+    {
+      "epoch": 5.497797356828194,
+      "grad_norm": 0.17404495191074987,
+      "learning_rate": 0.00019661434846010125,
+      "loss": 0.3379,
+      "step": 780
+    },
+    {
+      "epoch": 5.5048458149779735,
+      "grad_norm": 0.14313729432491926,
+      "learning_rate": 0.00019660248468467883,
+      "loss": 0.3312,
+      "step": 781
+    },
+    {
+      "epoch": 5.511894273127753,
+      "grad_norm": 0.1415873499359228,
+      "learning_rate": 0.00019659060051866408,
+      "loss": 0.3296,
+      "step": 782
+    },
+    {
+      "epoch": 5.518942731277533,
+      "grad_norm": 0.12599195755602036,
+      "learning_rate": 0.0001965786959645655,
+      "loss": 0.3269,
+      "step": 783
+    },
+    {
+      "epoch": 5.525991189427312,
+      "grad_norm": 0.15762543848768396,
+      "learning_rate": 0.00019656677102489588,
+      "loss": 0.3337,
+      "step": 784
+    },
+    {
+      "epoch": 5.533039647577093,
+      "grad_norm": 0.1339156389660052,
+      "learning_rate": 0.00019655482570217223,
+      "loss": 0.3455,
+      "step": 785
+    },
+    {
+      "epoch": 5.5400881057268725,
+      "grad_norm": 0.1501696787849796,
+      "learning_rate": 0.000196542859998916,
+      "loss": 0.3273,
+      "step": 786
+    },
+    {
+      "epoch": 5.547136563876652,
+      "grad_norm": 0.1505070431282537,
+      "learning_rate": 0.0001965308739176529,
+      "loss": 0.3364,
+      "step": 787
+    },
+    {
+      "epoch": 5.554185022026432,
+      "grad_norm": 0.1498293689556713,
+      "learning_rate": 0.00019651886746091283,
+      "loss": 0.3219,
+      "step": 788
+    },
+    {
+      "epoch": 5.561233480176211,
+      "grad_norm": 0.14615948904173343,
+      "learning_rate": 0.00019650684063123017,
+      "loss": 0.3194,
+      "step": 789
+    },
+    {
+      "epoch": 5.568281938325991,
+      "grad_norm": 0.13836339916705567,
+      "learning_rate": 0.00019649479343114345,
+      "loss": 0.3236,
+      "step": 790
+    },
+    {
+      "epoch": 5.5753303964757706,
+      "grad_norm": 0.15304106446835333,
+      "learning_rate": 0.0001964827258631956,
+      "loss": 0.3195,
+      "step": 791
+    },
+    {
+      "epoch": 5.582378854625551,
+      "grad_norm": 0.1445623570362839,
+      "learning_rate": 0.00019647063792993378,
+      "loss": 0.3304,
+      "step": 792
+    },
+    {
+      "epoch": 5.589427312775331,
+      "grad_norm": 0.15586979535427828,
+      "learning_rate": 0.0001964585296339095,
+      "loss": 0.3302,
+      "step": 793
+    },
+    {
+      "epoch": 5.59647577092511,
+      "grad_norm": 0.15636262942649398,
+      "learning_rate": 0.00019644640097767857,
+      "loss": 0.3148,
+      "step": 794
+    },
+    {
+      "epoch": 5.60352422907489,
+      "grad_norm": 0.16019374518373833,
+      "learning_rate": 0.000196434251963801,
+      "loss": 0.33,
+      "step": 795
+    },
+    {
+      "epoch": 5.6105726872246695,
+      "grad_norm": 0.14459124867581855,
+      "learning_rate": 0.00019642208259484126,
+      "loss": 0.3309,
+      "step": 796
+    },
+    {
+      "epoch": 5.617621145374449,
+      "grad_norm": 0.1595273984967937,
+      "learning_rate": 0.00019640989287336792,
+      "loss": 0.3284,
+      "step": 797
+    },
+    {
+      "epoch": 5.624669603524229,
+      "grad_norm": 0.12921981838712596,
+      "learning_rate": 0.00019639768280195407,
+      "loss": 0.3201,
+      "step": 798
+    },
+    {
+      "epoch": 5.631718061674009,
+      "grad_norm": 0.15783836598555495,
+      "learning_rate": 0.00019638545238317688,
+      "loss": 0.3441,
+      "step": 799
+    },
+    {
+      "epoch": 5.638766519823789,
+      "grad_norm": 0.1369397572681452,
+      "learning_rate": 0.00019637320161961799,
+      "loss": 0.3235,
+      "step": 800
+    },
+    {
+      "epoch": 5.6458149779735685,
+      "grad_norm": 0.1317245498303914,
+      "learning_rate": 0.00019636093051386322,
+      "loss": 0.3258,
+      "step": 801
+    },
+    {
+      "epoch": 5.652863436123348,
+      "grad_norm": 0.1440967084197898,
+      "learning_rate": 0.00019634863906850273,
+      "loss": 0.3268,
+      "step": 802
+    },
+    {
+      "epoch": 5.659911894273128,
+      "grad_norm": 0.13320356045076537,
+      "learning_rate": 0.00019633632728613098,
+      "loss": 0.34,
+      "step": 803
+    },
+    {
+      "epoch": 5.666960352422907,
+      "grad_norm": 0.11868163910858952,
+      "learning_rate": 0.00019632399516934665,
+      "loss": 0.336,
+      "step": 804
+    },
+    {
+      "epoch": 5.674008810572687,
+      "grad_norm": 0.16436377785022072,
+      "learning_rate": 0.00019631164272075285,
+      "loss": 0.3375,
+      "step": 805
+    },
+    {
+      "epoch": 5.681057268722467,
+      "grad_norm": 0.13463239123190576,
+      "learning_rate": 0.00019629926994295686,
+      "loss": 0.3308,
+      "step": 806
+    },
+    {
+      "epoch": 5.688105726872247,
+      "grad_norm": 0.16106046964476028,
+      "learning_rate": 0.0001962868768385703,
+      "loss": 0.3258,
+      "step": 807
+    },
+    {
+      "epoch": 5.695154185022027,
+      "grad_norm": 0.16220140727069607,
+      "learning_rate": 0.00019627446341020902,
+      "loss": 0.3287,
+      "step": 808
+    },
+    {
+      "epoch": 5.702202643171806,
+      "grad_norm": 0.18453166219799824,
+      "learning_rate": 0.00019626202966049335,
+      "loss": 0.3301,
+      "step": 809
+    },
+    {
+      "epoch": 5.709251101321586,
+      "grad_norm": 0.14951672593669751,
+      "learning_rate": 0.00019624957559204761,
+      "loss": 0.3235,
+      "step": 810
+    },
+    {
+      "epoch": 5.7162995594713655,
+      "grad_norm": 0.1436353658341361,
+      "learning_rate": 0.00019623710120750067,
+      "loss": 0.322,
+      "step": 811
+    },
+    {
+      "epoch": 5.723348017621145,
+      "grad_norm": 0.1624133280834532,
+      "learning_rate": 0.00019622460650948556,
+      "loss": 0.3337,
+      "step": 812
+    },
+    {
+      "epoch": 5.730396475770926,
+      "grad_norm": 0.14405097611785686,
+      "learning_rate": 0.00019621209150063966,
+      "loss": 0.3273,
+      "step": 813
+    },
+    {
+      "epoch": 5.737444933920705,
+      "grad_norm": 0.1423057254696731,
+      "learning_rate": 0.00019619955618360455,
+      "loss": 0.3264,
+      "step": 814
+    },
+    {
+      "epoch": 5.744493392070485,
+      "grad_norm": 0.1279329873572671,
+      "learning_rate": 0.0001961870005610262,
+      "loss": 0.3211,
+      "step": 815
+    },
+    {
+      "epoch": 5.751541850220264,
+      "grad_norm": 0.1545284533575573,
+      "learning_rate": 0.00019617442463555474,
+      "loss": 0.3185,
+      "step": 816
+    },
+    {
+      "epoch": 5.758590308370044,
+      "grad_norm": 0.1450146181861979,
+      "learning_rate": 0.00019616182840984477,
+      "loss": 0.3337,
+      "step": 817
+    },
+    {
+      "epoch": 5.765638766519824,
+      "grad_norm": 0.1393006893190091,
+      "learning_rate": 0.00019614921188655502,
+      "loss": 0.3318,
+      "step": 818
+    },
+    {
+      "epoch": 5.772687224669603,
+      "grad_norm": 0.15165136473034435,
+      "learning_rate": 0.00019613657506834854,
+      "loss": 0.3387,
+      "step": 819
+    },
+    {
+      "epoch": 5.779735682819383,
+      "grad_norm": 0.16018859058529386,
+      "learning_rate": 0.00019612391795789265,
+      "loss": 0.3203,
+      "step": 820
+    },
+    {
+      "epoch": 5.786784140969163,
+      "grad_norm": 0.13586610767932603,
+      "learning_rate": 0.00019611124055785904,
+      "loss": 0.3185,
+      "step": 821
+    },
+    {
+      "epoch": 5.793832599118943,
+      "grad_norm": 0.1968055921054932,
+      "learning_rate": 0.00019609854287092358,
+      "loss": 0.3334,
+      "step": 822
+    },
+    {
+      "epoch": 5.800881057268723,
+      "grad_norm": 0.1872901270840429,
+      "learning_rate": 0.00019608582489976647,
+      "loss": 0.3288,
+      "step": 823
+    },
+    {
+      "epoch": 5.807929515418502,
+      "grad_norm": 0.13189889142365424,
+      "learning_rate": 0.00019607308664707223,
+      "loss": 0.3351,
+      "step": 824
+    },
+    {
+      "epoch": 5.814977973568282,
+      "grad_norm": 0.1472368614965851,
+      "learning_rate": 0.00019606032811552955,
+      "loss": 0.3269,
+      "step": 825
+    },
+    {
+      "epoch": 5.8220264317180614,
+      "grad_norm": 0.13790915807323068,
+      "learning_rate": 0.00019604754930783148,
+      "loss": 0.3309,
+      "step": 826
+    },
+    {
+      "epoch": 5.829074889867841,
+      "grad_norm": 0.13056521303559765,
+      "learning_rate": 0.00019603475022667537,
+      "loss": 0.3232,
+      "step": 827
+    },
+    {
+      "epoch": 5.836123348017621,
+      "grad_norm": 0.1429086669083775,
+      "learning_rate": 0.0001960219308747628,
+      "loss": 0.3266,
+      "step": 828
+    },
+    {
+      "epoch": 5.843171806167401,
+      "grad_norm": 0.14652252629752327,
+      "learning_rate": 0.00019600909125479962,
+      "loss": 0.3307,
+      "step": 829
+    },
+    {
+      "epoch": 5.850220264317181,
+      "grad_norm": 0.13132319116728822,
+      "learning_rate": 0.000195996231369496,
+      "loss": 0.3236,
+      "step": 830
+    },
+    {
+      "epoch": 5.85726872246696,
+      "grad_norm": 0.1287768425174828,
+      "learning_rate": 0.0001959833512215664,
+      "loss": 0.319,
+      "step": 831
+    },
+    {
+      "epoch": 5.86431718061674,
+      "grad_norm": 0.14420511242763145,
+      "learning_rate": 0.00019597045081372948,
+      "loss": 0.3302,
+      "step": 832
+    },
+    {
+      "epoch": 5.87136563876652,
+      "grad_norm": 0.13168633395007776,
+      "learning_rate": 0.00019595753014870824,
+      "loss": 0.3361,
+      "step": 833
+    },
+    {
+      "epoch": 5.878414096916299,
+      "grad_norm": 0.15887104716958395,
+      "learning_rate": 0.00019594458922922995,
+      "loss": 0.3307,
+      "step": 834
+    },
+    {
+      "epoch": 5.885462555066079,
+      "grad_norm": 0.15515368802013588,
+      "learning_rate": 0.00019593162805802612,
+      "loss": 0.3379,
+      "step": 835
+    },
+    {
+      "epoch": 5.892511013215859,
+      "grad_norm": 0.15936393463844134,
+      "learning_rate": 0.0001959186466378326,
+      "loss": 0.3399,
+      "step": 836
+    },
+    {
+      "epoch": 5.899559471365639,
+      "grad_norm": 0.11938949607859671,
+      "learning_rate": 0.00019590564497138946,
+      "loss": 0.3274,
+      "step": 837
+    },
+    {
+      "epoch": 5.906607929515419,
+      "grad_norm": 0.13025094411249094,
+      "learning_rate": 0.000195892623061441,
+      "loss": 0.3323,
+      "step": 838
+    },
+    {
+      "epoch": 5.913656387665198,
+      "grad_norm": 0.1268545419632158,
+      "learning_rate": 0.00019587958091073592,
+      "loss": 0.3206,
+      "step": 839
+    },
+    {
+      "epoch": 5.920704845814978,
+      "grad_norm": 0.1452848716031609,
+      "learning_rate": 0.0001958665185220271,
+      "loss": 0.3329,
+      "step": 840
+    },
+    {
+      "epoch": 5.927753303964757,
+      "grad_norm": 0.13802333946192213,
+      "learning_rate": 0.0001958534358980717,
+      "loss": 0.3291,
+      "step": 841
+    },
+    {
+      "epoch": 5.934801762114537,
+      "grad_norm": 0.1292106101319078,
+      "learning_rate": 0.0001958403330416312,
+      "loss": 0.3267,
+      "step": 842
+    },
+    {
+      "epoch": 5.9418502202643175,
+      "grad_norm": 0.1352212272884195,
+      "learning_rate": 0.00019582720995547126,
+      "loss": 0.3233,
+      "step": 843
+    },
+    {
+      "epoch": 5.948898678414097,
+      "grad_norm": 0.12582683513719625,
+      "learning_rate": 0.00019581406664236192,
+      "loss": 0.3195,
+      "step": 844
+    },
+    {
+      "epoch": 5.955947136563877,
+      "grad_norm": 0.1291834914817452,
+      "learning_rate": 0.0001958009031050774,
+      "loss": 0.3183,
+      "step": 845
+    },
+    {
+      "epoch": 5.962995594713656,
+      "grad_norm": 0.12873988638087833,
+      "learning_rate": 0.00019578771934639623,
+      "loss": 0.3201,
+      "step": 846
+    },
+    {
+      "epoch": 5.970044052863436,
+      "grad_norm": 0.1369634758993936,
+      "learning_rate": 0.00019577451536910122,
+      "loss": 0.3389,
+      "step": 847
+    },
+    {
+      "epoch": 5.977092511013216,
+      "grad_norm": 0.17265188390604072,
+      "learning_rate": 0.0001957612911759794,
+      "loss": 0.3273,
+      "step": 848
+    },
+    {
+      "epoch": 5.984140969162995,
+      "grad_norm": 0.14409312773089605,
+      "learning_rate": 0.00019574804676982216,
+      "loss": 0.3312,
+      "step": 849
+    },
+    {
+      "epoch": 5.991189427312776,
+      "grad_norm": 0.1280206046885042,
+      "learning_rate": 0.00019573478215342501,
+      "loss": 0.3328,
+      "step": 850
+    },
+    {
+      "epoch": 5.998237885462555,
+      "grad_norm": 0.17131422837178842,
+      "learning_rate": 0.00019572149732958788,
+      "loss": 0.3152,
+      "step": 851
+    },
+    {
+      "epoch": 6.005286343612335,
+      "grad_norm": 0.12324848606313275,
+      "learning_rate": 0.00019570819230111484,
+      "loss": 0.3167,
+      "step": 852
+    },
+    {
+      "epoch": 6.0123348017621145,
+      "grad_norm": 0.18320342129596673,
+      "learning_rate": 0.0001956948670708143,
+      "loss": 0.3213,
+      "step": 853
+    },
+    {
+      "epoch": 6.019383259911894,
+      "grad_norm": 0.1889704204706042,
+      "learning_rate": 0.00019568152164149894,
+      "loss": 0.3135,
+      "step": 854
+    },
+    {
+      "epoch": 6.026431718061674,
+      "grad_norm": 0.16212801711535174,
+      "learning_rate": 0.00019566815601598564,
+      "loss": 0.3195,
+      "step": 855
+    },
+    {
+      "epoch": 6.033480176211453,
+      "grad_norm": 0.2147223658514147,
+      "learning_rate": 0.00019565477019709558,
+      "loss": 0.3149,
+      "step": 856
+    },
+    {
+      "epoch": 6.040528634361234,
+      "grad_norm": 0.1696278458460191,
+      "learning_rate": 0.00019564136418765422,
+      "loss": 0.307,
+      "step": 857
+    },
+    {
+      "epoch": 6.0475770925110135,
+      "grad_norm": 0.1694705038164429,
+      "learning_rate": 0.00019562793799049127,
+      "loss": 0.3257,
+      "step": 858
+    },
+    {
+      "epoch": 6.054625550660793,
+      "grad_norm": 0.1506461524259322,
+      "learning_rate": 0.0001956144916084407,
+      "loss": 0.3155,
+      "step": 859
+    },
+    {
+      "epoch": 6.061674008810573,
+      "grad_norm": 0.17849054780667192,
+      "learning_rate": 0.0001956010250443407,
+      "loss": 0.3068,
+      "step": 860
+    },
+    {
+      "epoch": 6.068722466960352,
+      "grad_norm": 0.1595245441916914,
+      "learning_rate": 0.0001955875383010338,
+      "loss": 0.3156,
+      "step": 861
+    },
+    {
+      "epoch": 6.075770925110132,
+      "grad_norm": 0.1595044212561119,
+      "learning_rate": 0.00019557403138136672,
+      "loss": 0.309,
+      "step": 862
+    },
+    {
+      "epoch": 6.082819383259912,
+      "grad_norm": 0.15111799359822103,
+      "learning_rate": 0.00019556050428819045,
+      "loss": 0.301,
+      "step": 863
+    },
+    {
+      "epoch": 6.089867841409692,
+      "grad_norm": 0.1445614835498229,
+      "learning_rate": 0.0001955469570243603,
+      "loss": 0.3041,
+      "step": 864
+    },
+    {
+      "epoch": 6.096916299559472,
+      "grad_norm": 0.13844512774674872,
+      "learning_rate": 0.00019553338959273574,
+      "loss": 0.312,
+      "step": 865
+    },
+    {
+      "epoch": 6.103964757709251,
+      "grad_norm": 0.1429856463198849,
+      "learning_rate": 0.00019551980199618058,
+      "loss": 0.3185,
+      "step": 866
+    },
+    {
+      "epoch": 6.111013215859031,
+      "grad_norm": 0.14933757187201963,
+      "learning_rate": 0.00019550619423756285,
+      "loss": 0.3176,
+      "step": 867
+    },
+    {
+      "epoch": 6.1180616740088105,
+      "grad_norm": 0.14766935086552824,
+      "learning_rate": 0.0001954925663197548,
+      "loss": 0.31,
+      "step": 868
+    },
+    {
+      "epoch": 6.12511013215859,
+      "grad_norm": 0.17148685954124204,
+      "learning_rate": 0.00019547891824563306,
+      "loss": 0.3328,
+      "step": 869
+    },
+    {
+      "epoch": 6.13215859030837,
+      "grad_norm": 0.14394508279118876,
+      "learning_rate": 0.0001954652500180783,
+      "loss": 0.3245,
+      "step": 870
+    },
+    {
+      "epoch": 6.139207048458149,
+      "grad_norm": 0.13076614437193865,
+      "learning_rate": 0.00019545156163997568,
+      "loss": 0.3052,
+      "step": 871
+    },
+    {
+      "epoch": 6.14625550660793,
+      "grad_norm": 0.13919187109582987,
+      "learning_rate": 0.00019543785311421446,
+      "loss": 0.3158,
+      "step": 872
+    },
+    {
+      "epoch": 6.1533039647577095,
+      "grad_norm": 0.18238314125214034,
+      "learning_rate": 0.00019542412444368824,
+      "loss": 0.3238,
+      "step": 873
+    },
+    {
+      "epoch": 6.160352422907489,
+      "grad_norm": 0.17774337945355378,
+      "learning_rate": 0.0001954103756312948,
+      "loss": 0.3176,
+      "step": 874
+    },
+    {
+      "epoch": 6.167400881057269,
+      "grad_norm": 0.15604819569974143,
+      "learning_rate": 0.0001953966066799362,
+      "loss": 0.3293,
+      "step": 875
+    },
+    {
+      "epoch": 6.174449339207048,
+      "grad_norm": 0.1506277828895116,
+      "learning_rate": 0.00019538281759251872,
+      "loss": 0.3137,
+      "step": 876
+    },
+    {
+      "epoch": 6.181497797356828,
+      "grad_norm": 0.14714630009192883,
+      "learning_rate": 0.000195369008371953,
+      "loss": 0.3094,
+      "step": 877
+    },
+    {
+      "epoch": 6.1885462555066075,
+      "grad_norm": 0.16409478404210975,
+      "learning_rate": 0.0001953551790211538,
+      "loss": 0.3001,
+      "step": 878
+    },
+    {
+      "epoch": 6.195594713656388,
+      "grad_norm": 0.16752511180261462,
+      "learning_rate": 0.00019534132954304018,
+      "loss": 0.3108,
+      "step": 879
+    },
+    {
+      "epoch": 6.202643171806168,
+      "grad_norm": 0.16897870320318456,
+      "learning_rate": 0.00019532745994053552,
+      "loss": 0.3321,
+      "step": 880
+    },
+    {
+      "epoch": 6.209691629955947,
+      "grad_norm": 0.15398795174668503,
+      "learning_rate": 0.0001953135702165673,
+      "loss": 0.3239,
+      "step": 881
+    },
+    {
+      "epoch": 6.216740088105727,
+      "grad_norm": 0.17605468161295842,
+      "learning_rate": 0.0001952996603740673,
+      "loss": 0.3224,
+      "step": 882
+    },
+    {
+      "epoch": 6.2237885462555065,
+      "grad_norm": 0.13503391752285376,
+      "learning_rate": 0.00019528573041597167,
+      "loss": 0.2996,
+      "step": 883
+    },
+    {
+      "epoch": 6.230837004405286,
+      "grad_norm": 0.15660400138447034,
+      "learning_rate": 0.00019527178034522063,
+      "loss": 0.3238,
+      "step": 884
+    },
+    {
+      "epoch": 6.237885462555066,
+      "grad_norm": 0.15183029044709395,
+      "learning_rate": 0.0001952578101647588,
+      "loss": 0.3126,
+      "step": 885
+    },
+    {
+      "epoch": 6.244933920704846,
+      "grad_norm": 0.16115821740185127,
+      "learning_rate": 0.0001952438198775349,
+      "loss": 0.334,
+      "step": 886
+    },
+    {
+      "epoch": 6.251982378854626,
+      "grad_norm": 0.14043291110708045,
+      "learning_rate": 0.00019522980948650197,
+      "loss": 0.3191,
+      "step": 887
+    },
+    {
+      "epoch": 6.259030837004405,
+      "grad_norm": 0.1475256341388436,
+      "learning_rate": 0.00019521577899461731,
+      "loss": 0.3092,
+      "step": 888
+    },
+    {
+      "epoch": 6.266079295154185,
+      "grad_norm": 0.1661065501199881,
+      "learning_rate": 0.00019520172840484243,
+      "loss": 0.3155,
+      "step": 889
+    },
+    {
+      "epoch": 6.273127753303965,
+      "grad_norm": 0.14551914254964643,
+      "learning_rate": 0.00019518765772014307,
+      "loss": 0.3084,
+      "step": 890
+    },
+    {
+      "epoch": 6.280176211453744,
+      "grad_norm": 0.14786447514512704,
+      "learning_rate": 0.00019517356694348926,
+      "loss": 0.3162,
+      "step": 891
+    },
+    {
+      "epoch": 6.287224669603524,
+      "grad_norm": 0.15670417105213896,
+      "learning_rate": 0.00019515945607785526,
+      "loss": 0.3268,
+      "step": 892
+    },
+    {
+      "epoch": 6.2942731277533035,
+      "grad_norm": 0.2005959189526323,
+      "learning_rate": 0.00019514532512621952,
+      "loss": 0.323,
+      "step": 893
+    },
+    {
+      "epoch": 6.301321585903084,
+      "grad_norm": 0.15041958965084026,
+      "learning_rate": 0.00019513117409156476,
+      "loss": 0.3349,
+      "step": 894
+    },
+    {
+      "epoch": 6.308370044052864,
+      "grad_norm": 0.1889891009706391,
+      "learning_rate": 0.00019511700297687794,
+      "loss": 0.3158,
+      "step": 895
+    },
+    {
+      "epoch": 6.315418502202643,
+      "grad_norm": 0.17386606192690154,
+      "learning_rate": 0.00019510281178515028,
+      "loss": 0.3091,
+      "step": 896
+    },
+    {
+      "epoch": 6.322466960352423,
+      "grad_norm": 0.1610872286861319,
+      "learning_rate": 0.00019508860051937723,
+      "loss": 0.3222,
+      "step": 897
+    },
+    {
+      "epoch": 6.3295154185022025,
+      "grad_norm": 0.14566159846028726,
+      "learning_rate": 0.00019507436918255844,
+      "loss": 0.3059,
+      "step": 898
+    },
+    {
+      "epoch": 6.336563876651982,
+      "grad_norm": 0.15876805816075687,
+      "learning_rate": 0.00019506011777769782,
+      "loss": 0.3278,
+      "step": 899
+    },
+    {
+      "epoch": 6.343612334801762,
+      "grad_norm": 0.15217026881098306,
+      "learning_rate": 0.00019504584630780354,
+      "loss": 0.3223,
+      "step": 900
+    },
+    {
+      "epoch": 6.350660792951542,
+      "grad_norm": 0.15551353649509586,
+      "learning_rate": 0.00019503155477588796,
+      "loss": 0.3233,
+      "step": 901
+    },
+    {
+      "epoch": 6.357709251101322,
+      "grad_norm": 0.15443236558090437,
+      "learning_rate": 0.00019501724318496773,
+      "loss": 0.3095,
+      "step": 902
+    },
+    {
+      "epoch": 6.364757709251101,
+      "grad_norm": 0.16149559437415245,
+      "learning_rate": 0.00019500291153806365,
+      "loss": 0.318,
+      "step": 903
+    },
+    {
+      "epoch": 6.371806167400881,
+      "grad_norm": 0.15879569101639507,
+      "learning_rate": 0.00019498855983820086,
+      "loss": 0.3154,
+      "step": 904
+    },
+    {
+      "epoch": 6.378854625550661,
+      "grad_norm": 0.13803124718187174,
+      "learning_rate": 0.00019497418808840867,
+      "loss": 0.3205,
+      "step": 905
+    },
+    {
+      "epoch": 6.38590308370044,
+      "grad_norm": 0.15973635061332583,
+      "learning_rate": 0.00019495979629172058,
+      "loss": 0.3129,
+      "step": 906
+    },
+    {
+      "epoch": 6.39295154185022,
+      "grad_norm": 0.1366976344684056,
+      "learning_rate": 0.00019494538445117443,
+      "loss": 0.3138,
+      "step": 907
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.14473048767216917,
+      "learning_rate": 0.0001949309525698122,
+      "loss": 0.3304,
+      "step": 908
+    },
+    {
+      "epoch": 6.40704845814978,
+      "grad_norm": 0.13383794155844772,
+      "learning_rate": 0.00019491650065068018,
+      "loss": 0.3132,
+      "step": 909
+    },
+    {
+      "epoch": 6.41409691629956,
+      "grad_norm": 0.15112612789338106,
+      "learning_rate": 0.00019490202869682874,
+      "loss": 0.3173,
+      "step": 910
+    },
+    {
+      "epoch": 6.421145374449339,
+      "grad_norm": 0.1470498463002044,
+      "learning_rate": 0.0001948875367113127,
+      "loss": 0.3289,
+      "step": 911
+    },
+    {
+      "epoch": 6.428193832599119,
+      "grad_norm": 0.14813404625065882,
+      "learning_rate": 0.0001948730246971909,
+      "loss": 0.3156,
+      "step": 912
+    },
+    {
+      "epoch": 6.435242290748898,
+      "grad_norm": 0.13322998784103862,
+      "learning_rate": 0.00019485849265752655,
+      "loss": 0.3101,
+      "step": 913
+    },
+    {
+      "epoch": 6.442290748898678,
+      "grad_norm": 0.17228176733472247,
+      "learning_rate": 0.000194843940595387,
+      "loss": 0.33,
+      "step": 914
+    },
+    {
+      "epoch": 6.4493392070484585,
+      "grad_norm": 0.15271655999332195,
+      "learning_rate": 0.0001948293685138439,
+      "loss": 0.3166,
+      "step": 915
+    },
+    {
+      "epoch": 6.456387665198238,
+      "grad_norm": 0.14934074322132146,
+      "learning_rate": 0.00019481477641597304,
+      "loss": 0.3178,
+      "step": 916
+    },
+    {
+      "epoch": 6.463436123348018,
+      "grad_norm": 0.1773548237333988,
+      "learning_rate": 0.00019480016430485452,
+      "loss": 0.3149,
+      "step": 917
+    },
+    {
+      "epoch": 6.470484581497797,
+      "grad_norm": 0.18953415966028456,
+      "learning_rate": 0.00019478553218357257,
+      "loss": 0.3205,
+      "step": 918
+    },
+    {
+      "epoch": 6.477533039647577,
+      "grad_norm": 0.14770734752978573,
+      "learning_rate": 0.0001947708800552158,
+      "loss": 0.3227,
+      "step": 919
+    },
+    {
+      "epoch": 6.484581497797357,
+      "grad_norm": 0.18753235603049032,
+      "learning_rate": 0.00019475620792287684,
+      "loss": 0.3305,
+      "step": 920
+    },
+    {
+      "epoch": 6.491629955947136,
+      "grad_norm": 0.13418519034532114,
+      "learning_rate": 0.00019474151578965267,
+      "loss": 0.3088,
+      "step": 921
+    },
+    {
+      "epoch": 6.498678414096917,
+      "grad_norm": 0.17572843101944904,
+      "learning_rate": 0.00019472680365864448,
+      "loss": 0.3261,
+      "step": 922
+    },
+    {
+      "epoch": 6.505726872246696,
+      "grad_norm": 0.13660875296128916,
+      "learning_rate": 0.0001947120715329577,
+      "loss": 0.3149,
+      "step": 923
+    },
+    {
+      "epoch": 6.512775330396476,
+      "grad_norm": 0.16064223616604323,
+      "learning_rate": 0.0001946973194157019,
+      "loss": 0.3139,
+      "step": 924
+    },
+    {
+      "epoch": 6.5198237885462555,
+      "grad_norm": 0.1598936737246455,
+      "learning_rate": 0.00019468254730999092,
+      "loss": 0.317,
+      "step": 925
+    },
+    {
+      "epoch": 6.526872246696035,
+      "grad_norm": 0.15746658252460033,
+      "learning_rate": 0.00019466775521894281,
+      "loss": 0.3086,
+      "step": 926
+    },
+    {
+      "epoch": 6.533920704845815,
+      "grad_norm": 0.15601302523772403,
+      "learning_rate": 0.00019465294314567987,
+      "loss": 0.3054,
+      "step": 927
+    },
+    {
+      "epoch": 6.540969162995594,
+      "grad_norm": 0.15452464270121813,
+      "learning_rate": 0.0001946381110933286,
+      "loss": 0.3136,
+      "step": 928
+    },
+    {
+      "epoch": 6.548017621145375,
+      "grad_norm": 0.13235982622572431,
+      "learning_rate": 0.00019462325906501968,
+      "loss": 0.3148,
+      "step": 929
+    },
+    {
+      "epoch": 6.5550660792951545,
+      "grad_norm": 0.15161970877073777,
+      "learning_rate": 0.00019460838706388805,
+      "loss": 0.3181,
+      "step": 930
+    },
+    {
+      "epoch": 6.562114537444934,
+      "grad_norm": 0.1440912815713729,
+      "learning_rate": 0.0001945934950930729,
+      "loss": 0.3303,
+      "step": 931
+    },
+    {
+      "epoch": 6.569162995594714,
+      "grad_norm": 0.14342367511506512,
+      "learning_rate": 0.00019457858315571747,
+      "loss": 0.3252,
+      "step": 932
+    },
+    {
+      "epoch": 6.576211453744493,
+      "grad_norm": 0.14926999412571484,
+      "learning_rate": 0.00019456365125496941,
+      "loss": 0.3133,
+      "step": 933
+    },
+    {
+      "epoch": 6.583259911894273,
+      "grad_norm": 0.1601119355787362,
+      "learning_rate": 0.0001945486993939805,
+      "loss": 0.3255,
+      "step": 934
+    },
+    {
+      "epoch": 6.590308370044053,
+      "grad_norm": 0.14874366639638584,
+      "learning_rate": 0.00019453372757590677,
+      "loss": 0.3139,
+      "step": 935
+    },
+    {
+      "epoch": 6.597356828193833,
+      "grad_norm": 0.1467796146140031,
+      "learning_rate": 0.00019451873580390833,
+      "loss": 0.3109,
+      "step": 936
+    },
+    {
+      "epoch": 6.604405286343613,
+      "grad_norm": 0.1494883971967105,
+      "learning_rate": 0.00019450372408114973,
+      "loss": 0.3195,
+      "step": 937
+    },
+    {
+      "epoch": 6.611453744493392,
+      "grad_norm": 0.16317729033797593,
+      "learning_rate": 0.0001944886924107995,
+      "loss": 0.3239,
+      "step": 938
+    },
+    {
+      "epoch": 6.618502202643172,
+      "grad_norm": 0.13914103309285333,
+      "learning_rate": 0.00019447364079603054,
+      "loss": 0.3256,
+      "step": 939
+    },
+    {
+      "epoch": 6.6255506607929515,
+      "grad_norm": 0.13737927411071565,
+      "learning_rate": 0.0001944585692400199,
+      "loss": 0.3219,
+      "step": 940
+    },
+    {
+      "epoch": 6.632599118942731,
+      "grad_norm": 0.1347213453633559,
+      "learning_rate": 0.00019444347774594882,
+      "loss": 0.3087,
+      "step": 941
+    },
+    {
+      "epoch": 6.639647577092511,
+      "grad_norm": 0.15105015484800466,
+      "learning_rate": 0.0001944283663170028,
+      "loss": 0.3178,
+      "step": 942
+    },
+    {
+      "epoch": 6.646696035242291,
+      "grad_norm": 0.12896506221070841,
+      "learning_rate": 0.00019441323495637147,
+      "loss": 0.315,
+      "step": 943
+    },
+    {
+      "epoch": 6.653744493392071,
+      "grad_norm": 0.14532308102759597,
+      "learning_rate": 0.00019439808366724875,
+      "loss": 0.3116,
+      "step": 944
+    },
+    {
+      "epoch": 6.6607929515418505,
+      "grad_norm": 0.13785404912842328,
+      "learning_rate": 0.00019438291245283277,
+      "loss": 0.3157,
+      "step": 945
+    },
+    {
+      "epoch": 6.66784140969163,
+      "grad_norm": 0.1545579855916239,
+      "learning_rate": 0.00019436772131632582,
+      "loss": 0.3125,
+      "step": 946
+    },
+    {
+      "epoch": 6.67488986784141,
+      "grad_norm": 0.13555486287053575,
+      "learning_rate": 0.00019435251026093435,
+      "loss": 0.3211,
+      "step": 947
+    },
+    {
+      "epoch": 6.681938325991189,
+      "grad_norm": 0.14002027638059844,
+      "learning_rate": 0.00019433727928986908,
+      "loss": 0.3266,
+      "step": 948
+    },
+    {
+      "epoch": 6.688986784140969,
+      "grad_norm": 0.13666956033756056,
+      "learning_rate": 0.00019432202840634496,
+      "loss": 0.3145,
+      "step": 949
+    },
+    {
+      "epoch": 6.6960352422907485,
+      "grad_norm": 0.13885922881815294,
+      "learning_rate": 0.00019430675761358108,
+      "loss": 0.317,
+      "step": 950
+    },
+    {
+      "epoch": 6.703083700440528,
+      "grad_norm": 0.13592672912664797,
+      "learning_rate": 0.00019429146691480077,
+      "loss": 0.3139,
+      "step": 951
+    },
+    {
+      "epoch": 6.710132158590309,
+      "grad_norm": 0.1393269907453392,
+      "learning_rate": 0.00019427615631323152,
+      "loss": 0.3134,
+      "step": 952
+    },
+    {
+      "epoch": 6.717180616740088,
+      "grad_norm": 0.15030521949067105,
+      "learning_rate": 0.0001942608258121051,
+      "loss": 0.3148,
+      "step": 953
+    },
+    {
+      "epoch": 6.724229074889868,
+      "grad_norm": 0.16462363916009798,
+      "learning_rate": 0.00019424547541465738,
+      "loss": 0.3035,
+      "step": 954
+    },
+    {
+      "epoch": 6.7312775330396475,
+      "grad_norm": 0.17147131731432907,
+      "learning_rate": 0.00019423010512412852,
+      "loss": 0.3133,
+      "step": 955
+    },
+    {
+      "epoch": 6.738325991189427,
+      "grad_norm": 0.13933541186224974,
+      "learning_rate": 0.0001942147149437628,
+      "loss": 0.3132,
+      "step": 956
+    },
+    {
+      "epoch": 6.745374449339207,
+      "grad_norm": 0.1551076544818111,
+      "learning_rate": 0.0001941993048768088,
+      "loss": 0.3226,
+      "step": 957
+    },
+    {
+      "epoch": 6.752422907488986,
+      "grad_norm": 0.1311816921903428,
+      "learning_rate": 0.00019418387492651918,
+      "loss": 0.3163,
+      "step": 958
+    },
+    {
+      "epoch": 6.759471365638767,
+      "grad_norm": 0.1542871318204159,
+      "learning_rate": 0.0001941684250961509,
+      "loss": 0.308,
+      "step": 959
+    },
+    {
+      "epoch": 6.766519823788546,
+      "grad_norm": 0.1350532166384727,
+      "learning_rate": 0.000194152955388965,
+      "loss": 0.3119,
+      "step": 960
+    },
+    {
+      "epoch": 6.773568281938326,
+      "grad_norm": 0.19304295669784854,
+      "learning_rate": 0.00019413746580822683,
+      "loss": 0.3249,
+      "step": 961
+    },
+    {
+      "epoch": 6.780616740088106,
+      "grad_norm": 0.13201553157846563,
+      "learning_rate": 0.00019412195635720586,
+      "loss": 0.3212,
+      "step": 962
+    },
+    {
+      "epoch": 6.787665198237885,
+      "grad_norm": 0.1645843397020744,
+      "learning_rate": 0.00019410642703917582,
+      "loss": 0.3124,
+      "step": 963
+    },
+    {
+      "epoch": 6.794713656387665,
+      "grad_norm": 0.1553839127562425,
+      "learning_rate": 0.0001940908778574146,
+      "loss": 0.3209,
+      "step": 964
+    },
+    {
+      "epoch": 6.8017621145374445,
+      "grad_norm": 0.1359490576278085,
+      "learning_rate": 0.00019407530881520427,
+      "loss": 0.3182,
+      "step": 965
+    },
+    {
+      "epoch": 6.808810572687225,
+      "grad_norm": 0.14809751370370733,
+      "learning_rate": 0.00019405971991583108,
+      "loss": 0.3122,
+      "step": 966
+    },
+    {
+      "epoch": 6.815859030837005,
+      "grad_norm": 0.14452086852366333,
+      "learning_rate": 0.00019404411116258547,
+      "loss": 0.3281,
+      "step": 967
+    },
+    {
+      "epoch": 6.822907488986784,
+      "grad_norm": 0.15282853586416678,
+      "learning_rate": 0.00019402848255876216,
+      "loss": 0.3208,
+      "step": 968
+    },
+    {
+      "epoch": 6.829955947136564,
+      "grad_norm": 0.13610470168484334,
+      "learning_rate": 0.00019401283410765995,
+      "loss": 0.3185,
+      "step": 969
+    },
+    {
+      "epoch": 6.8370044052863435,
+      "grad_norm": 0.137561185076078,
+      "learning_rate": 0.0001939971658125819,
+      "loss": 0.3216,
+      "step": 970
+    },
+    {
+      "epoch": 6.844052863436123,
+      "grad_norm": 0.1448658166693122,
+      "learning_rate": 0.0001939814776768352,
+      "loss": 0.3187,
+      "step": 971
+    },
+    {
+      "epoch": 6.851101321585903,
+      "grad_norm": 0.13719776985249119,
+      "learning_rate": 0.0001939657697037313,
+      "loss": 0.3184,
+      "step": 972
+    },
+    {
+      "epoch": 6.858149779735683,
+      "grad_norm": 0.15421446050574245,
+      "learning_rate": 0.00019395004189658575,
+      "loss": 0.3118,
+      "step": 973
+    },
+    {
+      "epoch": 6.865198237885463,
+      "grad_norm": 0.13917724986456687,
+      "learning_rate": 0.00019393429425871837,
+      "loss": 0.3267,
+      "step": 974
+    },
+    {
+      "epoch": 6.872246696035242,
+      "grad_norm": 0.16534124696572006,
+      "learning_rate": 0.00019391852679345313,
+      "loss": 0.3383,
+      "step": 975
+    },
+    {
+      "epoch": 6.879295154185022,
+      "grad_norm": 0.13534773915530138,
+      "learning_rate": 0.00019390273950411815,
+      "loss": 0.3281,
+      "step": 976
+    },
+    {
+      "epoch": 6.886343612334802,
+      "grad_norm": 0.15074453994500028,
+      "learning_rate": 0.00019388693239404578,
+      "loss": 0.3078,
+      "step": 977
+    },
+    {
+      "epoch": 6.893392070484581,
+      "grad_norm": 0.132758870815729,
+      "learning_rate": 0.00019387110546657254,
+      "loss": 0.3132,
+      "step": 978
+    },
+    {
+      "epoch": 6.900440528634361,
+      "grad_norm": 0.15081755527443394,
+      "learning_rate": 0.0001938552587250392,
+      "loss": 0.3251,
+      "step": 979
+    },
+    {
+      "epoch": 6.907488986784141,
+      "grad_norm": 0.14412634591956322,
+      "learning_rate": 0.0001938393921727905,
+      "loss": 0.3261,
+      "step": 980
+    },
+    {
+      "epoch": 6.914537444933921,
+      "grad_norm": 0.13080522961711954,
+      "learning_rate": 0.0001938235058131757,
+      "loss": 0.3197,
+      "step": 981
+    },
+    {
+      "epoch": 6.921585903083701,
+      "grad_norm": 0.12895032933751638,
+      "learning_rate": 0.00019380759964954792,
+      "loss": 0.3099,
+      "step": 982
+    },
+    {
+      "epoch": 6.92863436123348,
+      "grad_norm": 0.13046621424588684,
+      "learning_rate": 0.0001937916736852646,
+      "loss": 0.3208,
+      "step": 983
+    },
+    {
+      "epoch": 6.93568281938326,
+      "grad_norm": 0.13833020506437305,
+      "learning_rate": 0.00019377572792368737,
+      "loss": 0.3078,
+      "step": 984
+    },
+    {
+      "epoch": 6.942731277533039,
+      "grad_norm": 0.1346542405406993,
+      "learning_rate": 0.00019375976236818204,
+      "loss": 0.3064,
+      "step": 985
+    },
+    {
+      "epoch": 6.949779735682819,
+      "grad_norm": 0.14628547320516333,
+      "learning_rate": 0.00019374377702211854,
+      "loss": 0.307,
+      "step": 986
+    },
+    {
+      "epoch": 6.9568281938325995,
+      "grad_norm": 0.13794006564378306,
+      "learning_rate": 0.00019372777188887103,
+      "loss": 0.3172,
+      "step": 987
+    },
+    {
+      "epoch": 6.963876651982379,
+      "grad_norm": 0.16115867522222926,
+      "learning_rate": 0.00019371174697181784,
+      "loss": 0.3058,
+      "step": 988
+    },
+    {
+      "epoch": 6.970925110132159,
+      "grad_norm": 0.14254710623346892,
+      "learning_rate": 0.00019369570227434148,
+      "loss": 0.313,
+      "step": 989
+    },
+    {
+      "epoch": 6.977973568281938,
+      "grad_norm": 0.14046791130711894,
+      "learning_rate": 0.0001936796377998286,
+      "loss": 0.3126,
+      "step": 990
+    },
+    {
+      "epoch": 6.985022026431718,
+      "grad_norm": 0.13045998591155666,
+      "learning_rate": 0.00019366355355166998,
+      "loss": 0.3258,
+      "step": 991
+    },
+    {
+      "epoch": 6.992070484581498,
+      "grad_norm": 0.13700237131521137,
+      "learning_rate": 0.00019364744953326074,
+      "loss": 0.3165,
+      "step": 992
+    },
+    {
+      "epoch": 6.999118942731277,
+      "grad_norm": 0.13732527408630524,
+      "learning_rate": 0.00019363132574800007,
+      "loss": 0.315,
+      "step": 993
+    },
+    {
+      "epoch": 7.006167400881057,
+      "grad_norm": 0.1417838407498301,
+      "learning_rate": 0.00019361518219929128,
+      "loss": 0.3025,
+      "step": 994
+    },
+    {
+      "epoch": 7.013215859030837,
+      "grad_norm": 0.14337218664348045,
+      "learning_rate": 0.0001935990188905419,
+      "loss": 0.302,
+      "step": 995
+    },
+    {
+      "epoch": 7.020264317180617,
+      "grad_norm": 0.18696789850965476,
+      "learning_rate": 0.0001935828358251637,
+      "loss": 0.3116,
+      "step": 996
+    },
+    {
+      "epoch": 7.0273127753303966,
+      "grad_norm": 0.1894055978721307,
+      "learning_rate": 0.0001935666330065725,
+      "loss": 0.305,
+      "step": 997
+    },
+    {
+      "epoch": 7.034361233480176,
+      "grad_norm": 0.14755523764698983,
+      "learning_rate": 0.0001935504104381884,
+      "loss": 0.2903,
+      "step": 998
+    },
+    {
+      "epoch": 7.041409691629956,
+      "grad_norm": 0.16818771354258152,
+      "learning_rate": 0.0001935341681234356,
+      "loss": 0.3054,
+      "step": 999
+    },
+    {
+      "epoch": 7.048458149779735,
+      "grad_norm": 0.15915163480915404,
+      "learning_rate": 0.00019351790606574244,
+      "loss": 0.3084,
+      "step": 1000
+    },
+    {
+      "epoch": 7.055506607929515,
+      "grad_norm": 0.16282624705012202,
+      "learning_rate": 0.0001935016242685415,
+      "loss": 0.3108,
+      "step": 1001
+    },
+    {
+      "epoch": 7.0625550660792955,
+      "grad_norm": 0.1757996241944758,
+      "learning_rate": 0.0001934853227352695,
+      "loss": 0.3184,
+      "step": 1002
+    },
+    {
+      "epoch": 7.069603524229075,
+      "grad_norm": 0.15339968827331316,
+      "learning_rate": 0.00019346900146936734,
+      "loss": 0.3051,
+      "step": 1003
+    },
+    {
+      "epoch": 7.076651982378855,
+      "grad_norm": 0.15364922714350698,
+      "learning_rate": 0.00019345266047428005,
+      "loss": 0.305,
+      "step": 1004
+    },
+    {
+      "epoch": 7.083700440528634,
+      "grad_norm": 0.16941690398711098,
+      "learning_rate": 0.00019343629975345685,
+      "loss": 0.3135,
+      "step": 1005
+    },
+    {
+      "epoch": 7.090748898678414,
+      "grad_norm": 0.15993408747556356,
+      "learning_rate": 0.0001934199193103511,
+      "loss": 0.2928,
+      "step": 1006
+    },
+    {
+      "epoch": 7.097797356828194,
+      "grad_norm": 0.14901345379415887,
+      "learning_rate": 0.00019340351914842038,
+      "loss": 0.2958,
+      "step": 1007
+    },
+    {
+      "epoch": 7.104845814977973,
+      "grad_norm": 0.15351077361002124,
+      "learning_rate": 0.00019338709927112636,
+      "loss": 0.3016,
+      "step": 1008
+    },
+    {
+      "epoch": 7.111894273127754,
+      "grad_norm": 0.15748376583488616,
+      "learning_rate": 0.00019337065968193492,
+      "loss": 0.3093,
+      "step": 1009
+    },
+    {
+      "epoch": 7.118942731277533,
+      "grad_norm": 0.16267084981851268,
+      "learning_rate": 0.00019335420038431604,
+      "loss": 0.2906,
+      "step": 1010
+    },
+    {
+      "epoch": 7.125991189427313,
+      "grad_norm": 0.16393693552213337,
+      "learning_rate": 0.00019333772138174394,
+      "loss": 0.307,
+      "step": 1011
+    },
+    {
+      "epoch": 7.1330396475770925,
+      "grad_norm": 0.1730368470777147,
+      "learning_rate": 0.000193321222677697,
+      "loss": 0.3027,
+      "step": 1012
+    },
+    {
+      "epoch": 7.140088105726872,
+      "grad_norm": 0.18186844601869073,
+      "learning_rate": 0.00019330470427565764,
+      "loss": 0.3085,
+      "step": 1013
+    },
+    {
+      "epoch": 7.147136563876652,
+      "grad_norm": 0.1572760514114645,
+      "learning_rate": 0.0001932881661791126,
+      "loss": 0.3042,
+      "step": 1014
+    },
+    {
+      "epoch": 7.154185022026431,
+      "grad_norm": 0.15783456372391352,
+      "learning_rate": 0.00019327160839155262,
+      "loss": 0.2963,
+      "step": 1015
+    },
+    {
+      "epoch": 7.161233480176212,
+      "grad_norm": 0.15116782654720268,
+      "learning_rate": 0.00019325503091647276,
+      "loss": 0.2937,
+      "step": 1016
+    },
+    {
+      "epoch": 7.1682819383259915,
+      "grad_norm": 0.13859648093064306,
+      "learning_rate": 0.00019323843375737204,
+      "loss": 0.2943,
+      "step": 1017
+    },
+    {
+      "epoch": 7.175330396475771,
+      "grad_norm": 0.15782177613804518,
+      "learning_rate": 0.00019322181691775386,
+      "loss": 0.3062,
+      "step": 1018
+    },
+    {
+      "epoch": 7.182378854625551,
+      "grad_norm": 0.1512018513616645,
+      "learning_rate": 0.00019320518040112557,
+      "loss": 0.3124,
+      "step": 1019
+    },
+    {
+      "epoch": 7.18942731277533,
+      "grad_norm": 0.14873330692070014,
+      "learning_rate": 0.00019318852421099884,
+      "loss": 0.3078,
+      "step": 1020
+    },
+    {
+      "epoch": 7.19647577092511,
+      "grad_norm": 0.13807489061419395,
+      "learning_rate": 0.00019317184835088936,
+      "loss": 0.2922,
+      "step": 1021
+    },
+    {
+      "epoch": 7.2035242290748895,
+      "grad_norm": 0.1647566739373716,
+      "learning_rate": 0.000193155152824317,
+      "loss": 0.3133,
+      "step": 1022
+    },
+    {
+      "epoch": 7.210572687224669,
+      "grad_norm": 0.14659910297849604,
+      "learning_rate": 0.00019313843763480587,
+      "loss": 0.2939,
+      "step": 1023
+    },
+    {
+      "epoch": 7.21762114537445,
+      "grad_norm": 0.16033289728206584,
+      "learning_rate": 0.0001931217027858841,
+      "loss": 0.2955,
+      "step": 1024
+    },
+    {
+      "epoch": 7.224669603524229,
+      "grad_norm": 0.15350638711341955,
+      "learning_rate": 0.0001931049482810841,
+      "loss": 0.2969,
+      "step": 1025
+    },
+    {
+      "epoch": 7.231718061674009,
+      "grad_norm": 0.16922610488001197,
+      "learning_rate": 0.00019308817412394235,
+      "loss": 0.3171,
+      "step": 1026
+    },
+    {
+      "epoch": 7.2387665198237885,
+      "grad_norm": 0.13765321125091098,
+      "learning_rate": 0.00019307138031799946,
+      "loss": 0.3108,
+      "step": 1027
+    },
+    {
+      "epoch": 7.245814977973568,
+      "grad_norm": 0.1509422575127806,
+      "learning_rate": 0.00019305456686680028,
+      "loss": 0.3053,
+      "step": 1028
+    },
+    {
+      "epoch": 7.252863436123348,
+      "grad_norm": 0.14815330109567088,
+      "learning_rate": 0.00019303773377389369,
+      "loss": 0.3137,
+      "step": 1029
+    },
+    {
+      "epoch": 7.259911894273127,
+      "grad_norm": 0.14042321695353452,
+      "learning_rate": 0.0001930208810428328,
+      "loss": 0.2988,
+      "step": 1030
+    },
+    {
+      "epoch": 7.266960352422908,
+      "grad_norm": 0.1547851645702065,
+      "learning_rate": 0.00019300400867717482,
+      "loss": 0.3204,
+      "step": 1031
+    },
+    {
+      "epoch": 7.2740088105726874,
+      "grad_norm": 0.19508225328285367,
+      "learning_rate": 0.00019298711668048118,
+      "loss": 0.296,
+      "step": 1032
+    },
+    {
+      "epoch": 7.281057268722467,
+      "grad_norm": 0.17040788056543954,
+      "learning_rate": 0.00019297020505631732,
+      "loss": 0.3043,
+      "step": 1033
+    },
+    {
+      "epoch": 7.288105726872247,
+      "grad_norm": 0.16698801773159896,
+      "learning_rate": 0.00019295327380825295,
+      "loss": 0.2936,
+      "step": 1034
+    },
+    {
+      "epoch": 7.295154185022026,
+      "grad_norm": 0.15272289619668272,
+      "learning_rate": 0.0001929363229398619,
+      "loss": 0.309,
+      "step": 1035
+    },
+    {
+      "epoch": 7.302202643171806,
+      "grad_norm": 0.17739751943340112,
+      "learning_rate": 0.000192919352454722,
+      "loss": 0.2997,
+      "step": 1036
+    },
+    {
+      "epoch": 7.3092511013215855,
+      "grad_norm": 0.13929458746098441,
+      "learning_rate": 0.00019290236235641547,
+      "loss": 0.2993,
+      "step": 1037
+    },
+    {
+      "epoch": 7.316299559471366,
+      "grad_norm": 0.16351917012996167,
+      "learning_rate": 0.00019288535264852841,
+      "loss": 0.2988,
+      "step": 1038
+    },
+    {
+      "epoch": 7.323348017621146,
+      "grad_norm": 0.15680932454544366,
+      "learning_rate": 0.0001928683233346513,
+      "loss": 0.3078,
+      "step": 1039
+    },
+    {
+      "epoch": 7.330396475770925,
+      "grad_norm": 0.14820770480582574,
+      "learning_rate": 0.00019285127441837858,
+      "loss": 0.3119,
+      "step": 1040
+    },
+    {
+      "epoch": 7.337444933920705,
+      "grad_norm": 0.15631391142337175,
+      "learning_rate": 0.0001928342059033089,
+      "loss": 0.3077,
+      "step": 1041
+    },
+    {
+      "epoch": 7.3444933920704845,
+      "grad_norm": 0.13488544371139477,
+      "learning_rate": 0.00019281711779304498,
+      "loss": 0.3016,
+      "step": 1042
+    },
+    {
+      "epoch": 7.351541850220264,
+      "grad_norm": 0.19436358332309314,
+      "learning_rate": 0.00019280001009119383,
+      "loss": 0.2974,
+      "step": 1043
+    },
+    {
+      "epoch": 7.358590308370044,
+      "grad_norm": 0.16783755678613235,
+      "learning_rate": 0.00019278288280136647,
+      "loss": 0.3017,
+      "step": 1044
+    },
+    {
+      "epoch": 7.365638766519824,
+      "grad_norm": 0.18065166529523588,
+      "learning_rate": 0.00019276573592717804,
+      "loss": 0.308,
+      "step": 1045
+    },
+    {
+      "epoch": 7.372687224669604,
+      "grad_norm": 0.16433715980254396,
+      "learning_rate": 0.00019274856947224786,
+      "loss": 0.3113,
+      "step": 1046
+    },
+    {
+      "epoch": 7.379735682819383,
+      "grad_norm": 0.16335354500517024,
+      "learning_rate": 0.00019273138344019945,
+      "loss": 0.2973,
+      "step": 1047
+    },
+    {
+      "epoch": 7.386784140969163,
+      "grad_norm": 0.16400370169409786,
+      "learning_rate": 0.0001927141778346603,
+      "loss": 0.3177,
+      "step": 1048
+    },
+    {
+      "epoch": 7.393832599118943,
+      "grad_norm": 0.16479769758336177,
+      "learning_rate": 0.00019269695265926217,
+      "loss": 0.3049,
+      "step": 1049
+    },
+    {
+      "epoch": 7.400881057268722,
+      "grad_norm": 0.19352280354716384,
+      "learning_rate": 0.00019267970791764092,
+      "loss": 0.3075,
+      "step": 1050
+    },
+    {
+      "epoch": 7.407929515418502,
+      "grad_norm": 0.1397994182256079,
+      "learning_rate": 0.00019266244361343648,
+      "loss": 0.3019,
+      "step": 1051
+    },
+    {
+      "epoch": 7.414977973568282,
+      "grad_norm": 0.17796281776283704,
+      "learning_rate": 0.000192645159750293,
+      "loss": 0.3141,
+      "step": 1052
+    },
+    {
+      "epoch": 7.422026431718062,
+      "grad_norm": 0.1397652053386815,
+      "learning_rate": 0.00019262785633185865,
+      "loss": 0.2956,
+      "step": 1053
+    },
+    {
+      "epoch": 7.429074889867842,
+      "grad_norm": 0.18086211177014752,
+      "learning_rate": 0.00019261053336178585,
+      "loss": 0.3111,
+      "step": 1054
+    },
+    {
+      "epoch": 7.436123348017621,
+      "grad_norm": 0.17520435141856247,
+      "learning_rate": 0.00019259319084373105,
+      "loss": 0.312,
+      "step": 1055
+    },
+    {
+      "epoch": 7.443171806167401,
+      "grad_norm": 0.15373096396857386,
+      "learning_rate": 0.00019257582878135488,
+      "loss": 0.3098,
+      "step": 1056
+    },
+    {
+      "epoch": 7.45022026431718,
+      "grad_norm": 0.159658991969229,
+      "learning_rate": 0.00019255844717832205,
+      "loss": 0.3049,
+      "step": 1057
+    },
+    {
+      "epoch": 7.45726872246696,
+      "grad_norm": 0.14736293643633158,
+      "learning_rate": 0.00019254104603830147,
+      "loss": 0.3086,
+      "step": 1058
+    },
+    {
+      "epoch": 7.4643171806167405,
+      "grad_norm": 0.15921734601028645,
+      "learning_rate": 0.00019252362536496608,
+      "loss": 0.3053,
+      "step": 1059
+    },
+    {
+      "epoch": 7.47136563876652,
+      "grad_norm": 0.1449374591689634,
+      "learning_rate": 0.000192506185161993,
+      "loss": 0.2982,
+      "step": 1060
+    },
+    {
+      "epoch": 7.4784140969163,
+      "grad_norm": 0.1607299799856251,
+      "learning_rate": 0.0001924887254330635,
+      "loss": 0.3072,
+      "step": 1061
+    },
+    {
+      "epoch": 7.485462555066079,
+      "grad_norm": 0.1573529260614065,
+      "learning_rate": 0.00019247124618186284,
+      "loss": 0.3094,
+      "step": 1062
+    },
+    {
+      "epoch": 7.492511013215859,
+      "grad_norm": 0.15073275730353036,
+      "learning_rate": 0.0001924537474120806,
+      "loss": 0.3153,
+      "step": 1063
+    },
+    {
+      "epoch": 7.499559471365639,
+      "grad_norm": 0.1464479320674903,
+      "learning_rate": 0.00019243622912741034,
+      "loss": 0.2902,
+      "step": 1064
+    },
+    {
+      "epoch": 7.506607929515418,
+      "grad_norm": 0.15896741993523578,
+      "learning_rate": 0.00019241869133154973,
+      "loss": 0.3155,
+      "step": 1065
+    },
+    {
+      "epoch": 7.513656387665199,
+      "grad_norm": 0.17172407000980963,
+      "learning_rate": 0.00019240113402820062,
+      "loss": 0.3008,
+      "step": 1066
+    },
+    {
+      "epoch": 7.520704845814978,
+      "grad_norm": 0.1469873817302848,
+      "learning_rate": 0.000192383557221069,
+      "loss": 0.3155,
+      "step": 1067
+    },
+    {
+      "epoch": 7.527753303964758,
+      "grad_norm": 0.14300027338274313,
+      "learning_rate": 0.00019236596091386487,
+      "loss": 0.2929,
+      "step": 1068
+    },
+    {
+      "epoch": 7.534801762114538,
+      "grad_norm": 0.1488414484068325,
+      "learning_rate": 0.00019234834511030246,
+      "loss": 0.3038,
+      "step": 1069
+    },
+    {
+      "epoch": 7.541850220264317,
+      "grad_norm": 0.15472238416997072,
+      "learning_rate": 0.00019233070981410007,
+      "loss": 0.3024,
+      "step": 1070
+    },
+    {
+      "epoch": 7.548898678414097,
+      "grad_norm": 0.13920820502653874,
+      "learning_rate": 0.00019231305502898004,
+      "loss": 0.3039,
+      "step": 1071
+    },
+    {
+      "epoch": 7.555947136563876,
+      "grad_norm": 0.14759184274571802,
+      "learning_rate": 0.000192295380758669,
+      "loss": 0.2997,
+      "step": 1072
+    },
+    {
+      "epoch": 7.562995594713656,
+      "grad_norm": 0.14699568960750686,
+      "learning_rate": 0.0001922776870068975,
+      "loss": 0.2974,
+      "step": 1073
+    },
+    {
+      "epoch": 7.570044052863436,
+      "grad_norm": 0.15383747356621272,
+      "learning_rate": 0.00019225997377740034,
+      "loss": 0.301,
+      "step": 1074
+    },
+    {
+      "epoch": 7.577092511013216,
+      "grad_norm": 0.1326974427192935,
+      "learning_rate": 0.00019224224107391636,
+      "loss": 0.2952,
+      "step": 1075
+    },
+    {
+      "epoch": 7.584140969162996,
+      "grad_norm": 0.13928476118647165,
+      "learning_rate": 0.00019222448890018852,
+      "loss": 0.307,
+      "step": 1076
+    },
+    {
+      "epoch": 7.591189427312775,
+      "grad_norm": 0.14658563658504514,
+      "learning_rate": 0.00019220671725996393,
+      "loss": 0.3108,
+      "step": 1077
+    },
+    {
+      "epoch": 7.598237885462555,
+      "grad_norm": 0.15615175933158482,
+      "learning_rate": 0.00019218892615699375,
+      "loss": 0.3175,
+      "step": 1078
+    },
+    {
+      "epoch": 7.605286343612335,
+      "grad_norm": 0.13772235798125493,
+      "learning_rate": 0.00019217111559503334,
+      "loss": 0.3154,
+      "step": 1079
+    },
+    {
+      "epoch": 7.612334801762114,
+      "grad_norm": 0.14356911598046657,
+      "learning_rate": 0.00019215328557784203,
+      "loss": 0.3106,
+      "step": 1080
+    },
+    {
+      "epoch": 7.619383259911894,
+      "grad_norm": 0.15948028949014487,
+      "learning_rate": 0.00019213543610918334,
+      "loss": 0.3204,
+      "step": 1081
+    },
+    {
+      "epoch": 7.626431718061674,
+      "grad_norm": 0.16812526178797865,
+      "learning_rate": 0.0001921175671928249,
+      "loss": 0.31,
+      "step": 1082
+    },
+    {
+      "epoch": 7.633480176211454,
+      "grad_norm": 0.15254346520001782,
+      "learning_rate": 0.00019209967883253849,
+      "loss": 0.3142,
+      "step": 1083
+    },
+    {
+      "epoch": 7.6405286343612335,
+      "grad_norm": 0.15111206810889002,
+      "learning_rate": 0.00019208177103209985,
+      "loss": 0.3029,
+      "step": 1084
+    },
+    {
+      "epoch": 7.647577092511013,
+      "grad_norm": 0.15556659394605438,
+      "learning_rate": 0.00019206384379528898,
+      "loss": 0.3042,
+      "step": 1085
+    },
+    {
+      "epoch": 7.654625550660793,
+      "grad_norm": 0.14636102517984412,
+      "learning_rate": 0.00019204589712588986,
+      "loss": 0.3169,
+      "step": 1086
+    },
+    {
+      "epoch": 7.661674008810572,
+      "grad_norm": 0.13432575269306288,
+      "learning_rate": 0.00019202793102769068,
+      "loss": 0.2973,
+      "step": 1087
+    },
+    {
+      "epoch": 7.668722466960352,
+      "grad_norm": 0.14806585220400162,
+      "learning_rate": 0.0001920099455044836,
+      "loss": 0.3022,
+      "step": 1088
+    },
+    {
+      "epoch": 7.6757709251101325,
+      "grad_norm": 0.16951604409303025,
+      "learning_rate": 0.00019199194056006503,
+      "loss": 0.2922,
+      "step": 1089
+    },
+    {
+      "epoch": 7.682819383259912,
+      "grad_norm": 0.15640833861257467,
+      "learning_rate": 0.00019197391619823535,
+      "loss": 0.3057,
+      "step": 1090
+    },
+    {
+      "epoch": 7.689867841409692,
+      "grad_norm": 0.1565711759438039,
+      "learning_rate": 0.00019195587242279912,
+      "loss": 0.3062,
+      "step": 1091
+    },
+    {
+      "epoch": 7.696916299559471,
+      "grad_norm": 0.1448541921724594,
+      "learning_rate": 0.000191937809237565,
+      "loss": 0.3064,
+      "step": 1092
+    },
+    {
+      "epoch": 7.703964757709251,
+      "grad_norm": 0.16009758818403713,
+      "learning_rate": 0.00019191972664634564,
+      "loss": 0.3072,
+      "step": 1093
+    },
+    {
+      "epoch": 7.7110132158590305,
+      "grad_norm": 0.12985689726093916,
+      "learning_rate": 0.00019190162465295793,
+      "loss": 0.3028,
+      "step": 1094
+    },
+    {
+      "epoch": 7.71806167400881,
+      "grad_norm": 0.1503994830500829,
+      "learning_rate": 0.00019188350326122277,
+      "loss": 0.3123,
+      "step": 1095
+    },
+    {
+      "epoch": 7.725110132158591,
+      "grad_norm": 0.15903366293644788,
+      "learning_rate": 0.00019186536247496518,
+      "loss": 0.3115,
+      "step": 1096
+    },
+    {
+      "epoch": 7.73215859030837,
+      "grad_norm": 0.14819532281092013,
+      "learning_rate": 0.00019184720229801429,
+      "loss": 0.3047,
+      "step": 1097
+    },
+    {
+      "epoch": 7.73920704845815,
+      "grad_norm": 0.1370811823231416,
+      "learning_rate": 0.00019182902273420321,
+      "loss": 0.3061,
+      "step": 1098
+    },
+    {
+      "epoch": 7.7462555066079295,
+      "grad_norm": 0.1372551937734758,
+      "learning_rate": 0.00019181082378736934,
+      "loss": 0.3068,
+      "step": 1099
+    },
+    {
+      "epoch": 7.753303964757709,
+      "grad_norm": 0.136029711137383,
+      "learning_rate": 0.00019179260546135405,
+      "loss": 0.3045,
+      "step": 1100
+    },
+    {
+      "epoch": 7.760352422907489,
+      "grad_norm": 0.13602268766701336,
+      "learning_rate": 0.00019177436776000276,
+      "loss": 0.3047,
+      "step": 1101
+    },
+    {
+      "epoch": 7.767400881057268,
+      "grad_norm": 0.14862882588923146,
+      "learning_rate": 0.00019175611068716507,
+      "loss": 0.3113,
+      "step": 1102
+    },
+    {
+      "epoch": 7.774449339207049,
+      "grad_norm": 0.1477779837212756,
+      "learning_rate": 0.00019173783424669463,
+      "loss": 0.305,
+      "step": 1103
+    },
+    {
+      "epoch": 7.7814977973568284,
+      "grad_norm": 0.14190682478807232,
+      "learning_rate": 0.00019171953844244915,
+      "loss": 0.3029,
+      "step": 1104
+    },
+    {
+      "epoch": 7.788546255506608,
+      "grad_norm": 0.13152899918307398,
+      "learning_rate": 0.0001917012232782905,
+      "loss": 0.3091,
+      "step": 1105
+    },
+    {
+      "epoch": 7.795594713656388,
+      "grad_norm": 0.14069023178914009,
+      "learning_rate": 0.0001916828887580846,
+      "loss": 0.2998,
+      "step": 1106
+    },
+    {
+      "epoch": 7.802643171806167,
+      "grad_norm": 0.14934607686817783,
+      "learning_rate": 0.00019166453488570143,
+      "loss": 0.3037,
+      "step": 1107
+    },
+    {
+      "epoch": 7.809691629955947,
+      "grad_norm": 0.13792757004479855,
+      "learning_rate": 0.0001916461616650151,
+      "loss": 0.2968,
+      "step": 1108
+    },
+    {
+      "epoch": 7.8167400881057265,
+      "grad_norm": 0.17011225491092916,
+      "learning_rate": 0.00019162776909990373,
+      "loss": 0.3019,
+      "step": 1109
+    },
+    {
+      "epoch": 7.823788546255507,
+      "grad_norm": 0.15850798575650374,
+      "learning_rate": 0.00019160935719424963,
+      "loss": 0.3123,
+      "step": 1110
+    },
+    {
+      "epoch": 7.830837004405287,
+      "grad_norm": 0.1403432537423809,
+      "learning_rate": 0.0001915909259519391,
+      "loss": 0.3094,
+      "step": 1111
+    },
+    {
+      "epoch": 7.837885462555066,
+      "grad_norm": 0.15659373041720187,
+      "learning_rate": 0.00019157247537686256,
+      "loss": 0.3141,
+      "step": 1112
+    },
+    {
+      "epoch": 7.844933920704846,
+      "grad_norm": 0.13110113551750477,
+      "learning_rate": 0.00019155400547291452,
+      "loss": 0.2949,
+      "step": 1113
+    },
+    {
+      "epoch": 7.8519823788546255,
+      "grad_norm": 0.16491558542805748,
+      "learning_rate": 0.00019153551624399355,
+      "loss": 0.3074,
+      "step": 1114
+    },
+    {
+      "epoch": 7.859030837004405,
+      "grad_norm": 0.1552186887690726,
+      "learning_rate": 0.00019151700769400233,
+      "loss": 0.3195,
+      "step": 1115
+    },
+    {
+      "epoch": 7.866079295154185,
+      "grad_norm": 0.17185532941833265,
+      "learning_rate": 0.00019149847982684754,
+      "loss": 0.2998,
+      "step": 1116
+    },
+    {
+      "epoch": 7.873127753303965,
+      "grad_norm": 0.1559579673697212,
+      "learning_rate": 0.00019147993264644004,
+      "loss": 0.3002,
+      "step": 1117
+    },
+    {
+      "epoch": 7.880176211453745,
+      "grad_norm": 0.1467797771931789,
+      "learning_rate": 0.0001914613661566947,
+      "loss": 0.2846,
+      "step": 1118
+    },
+    {
+      "epoch": 7.887224669603524,
+      "grad_norm": 0.1319164463674855,
+      "learning_rate": 0.0001914427803615305,
+      "loss": 0.307,
+      "step": 1119
+    },
+    {
+      "epoch": 7.894273127753304,
+      "grad_norm": 0.13229194747748024,
+      "learning_rate": 0.00019142417526487045,
+      "loss": 0.3113,
+      "step": 1120
+    },
+    {
+      "epoch": 7.901321585903084,
+      "grad_norm": 0.12935408082997663,
+      "learning_rate": 0.00019140555087064168,
+      "loss": 0.299,
+      "step": 1121
+    },
+    {
+      "epoch": 7.908370044052863,
+      "grad_norm": 0.14407848519117364,
+      "learning_rate": 0.00019138690718277542,
+      "loss": 0.2977,
+      "step": 1122
+    },
+    {
+      "epoch": 7.915418502202643,
+      "grad_norm": 0.149652567599669,
+      "learning_rate": 0.00019136824420520684,
+      "loss": 0.3057,
+      "step": 1123
+    },
+    {
+      "epoch": 7.922466960352423,
+      "grad_norm": 0.14070895579387907,
+      "learning_rate": 0.00019134956194187536,
+      "loss": 0.3066,
+      "step": 1124
+    },
+    {
+      "epoch": 7.929515418502203,
+      "grad_norm": 0.14906330610897545,
+      "learning_rate": 0.00019133086039672433,
+      "loss": 0.3075,
+      "step": 1125
+    },
+    {
+      "epoch": 7.936563876651983,
+      "grad_norm": 0.14513254781582668,
+      "learning_rate": 0.00019131213957370124,
+      "loss": 0.3141,
+      "step": 1126
+    },
+    {
+      "epoch": 7.943612334801762,
+      "grad_norm": 0.15503579904219605,
+      "learning_rate": 0.00019129339947675762,
+      "loss": 0.3197,
+      "step": 1127
+    },
+    {
+      "epoch": 7.950660792951542,
+      "grad_norm": 0.1375071440188523,
+      "learning_rate": 0.0001912746401098491,
+      "loss": 0.2999,
+      "step": 1128
+    },
+    {
+      "epoch": 7.957709251101321,
+      "grad_norm": 0.1598492836000749,
+      "learning_rate": 0.0001912558614769353,
+      "loss": 0.3055,
+      "step": 1129
+    },
+    {
+      "epoch": 7.964757709251101,
+      "grad_norm": 0.14748092690094594,
+      "learning_rate": 0.00019123706358198,
+      "loss": 0.3046,
+      "step": 1130
+    },
+    {
+      "epoch": 7.9718061674008815,
+      "grad_norm": 0.14688900641389283,
+      "learning_rate": 0.00019121824642895105,
+      "loss": 0.3021,
+      "step": 1131
+    },
+    {
+      "epoch": 7.978854625550661,
+      "grad_norm": 0.16814158508279636,
+      "learning_rate": 0.00019119941002182026,
+      "loss": 0.3247,
+      "step": 1132
+    },
+    {
+      "epoch": 7.985903083700441,
+      "grad_norm": 0.14938931395561364,
+      "learning_rate": 0.0001911805543645636,
+      "loss": 0.3044,
+      "step": 1133
+    },
+    {
+      "epoch": 7.99295154185022,
+      "grad_norm": 0.17108530326442545,
+      "learning_rate": 0.00019116167946116108,
+      "loss": 0.3161,
+      "step": 1134
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.14285203138806365,
+      "learning_rate": 0.00019114278531559675,
+      "loss": 0.3051,
+      "step": 1135
+    },
+    {
+      "epoch": 8.00704845814978,
+      "grad_norm": 0.14458975854133896,
+      "learning_rate": 0.00019112387193185874,
+      "loss": 0.2913,
+      "step": 1136
+    },
+    {
+      "epoch": 8.01409691629956,
+      "grad_norm": 0.1496490913284144,
+      "learning_rate": 0.0001911049393139392,
+      "loss": 0.2794,
+      "step": 1137
+    },
+    {
+      "epoch": 8.02114537444934,
+      "grad_norm": 0.1796519852491093,
+      "learning_rate": 0.00019108598746583443,
+      "loss": 0.2845,
+      "step": 1138
+    },
+    {
+      "epoch": 8.028193832599118,
+      "grad_norm": 0.22540100900621413,
+      "learning_rate": 0.00019106701639154473,
+      "loss": 0.2957,
+      "step": 1139
+    },
+    {
+      "epoch": 8.035242290748899,
+      "grad_norm": 0.15931722336378135,
+      "learning_rate": 0.00019104802609507443,
+      "loss": 0.2791,
+      "step": 1140
+    },
+    {
+      "epoch": 8.042290748898678,
+      "grad_norm": 0.17301512612897738,
+      "learning_rate": 0.00019102901658043196,
+      "loss": 0.2974,
+      "step": 1141
+    },
+    {
+      "epoch": 8.049339207048458,
+      "grad_norm": 0.16152179510468373,
+      "learning_rate": 0.0001910099878516298,
+      "loss": 0.2855,
+      "step": 1142
+    },
+    {
+      "epoch": 8.056387665198239,
+      "grad_norm": 0.19601359997355577,
+      "learning_rate": 0.00019099093991268448,
+      "loss": 0.2915,
+      "step": 1143
+    },
+    {
+      "epoch": 8.063436123348017,
+      "grad_norm": 0.2165739320956518,
+      "learning_rate": 0.00019097187276761663,
+      "loss": 0.2996,
+      "step": 1144
+    },
+    {
+      "epoch": 8.070484581497798,
+      "grad_norm": 0.19082318678060822,
+      "learning_rate": 0.00019095278642045082,
+      "loss": 0.2928,
+      "step": 1145
+    },
+    {
+      "epoch": 8.077533039647577,
+      "grad_norm": 0.2656479552056437,
+      "learning_rate": 0.00019093368087521578,
+      "loss": 0.2973,
+      "step": 1146
+    },
+    {
+      "epoch": 8.084581497797357,
+      "grad_norm": 0.16657108544193358,
+      "learning_rate": 0.0001909145561359443,
+      "loss": 0.294,
+      "step": 1147
+    },
+    {
+      "epoch": 8.091629955947136,
+      "grad_norm": 0.18810932286706547,
+      "learning_rate": 0.0001908954122066731,
+      "loss": 0.2847,
+      "step": 1148
+    },
+    {
+      "epoch": 8.098678414096916,
+      "grad_norm": 0.14710050454739676,
+      "learning_rate": 0.0001908762490914431,
+      "loss": 0.278,
+      "step": 1149
+    },
+    {
+      "epoch": 8.105726872246697,
+      "grad_norm": 0.1753579108507199,
+      "learning_rate": 0.00019085706679429913,
+      "loss": 0.2996,
+      "step": 1150
+    },
+    {
+      "epoch": 8.112775330396476,
+      "grad_norm": 0.17409846578823127,
+      "learning_rate": 0.00019083786531929014,
+      "loss": 0.3006,
+      "step": 1151
+    },
+    {
+      "epoch": 8.119823788546256,
+      "grad_norm": 0.191836600561761,
+      "learning_rate": 0.0001908186446704692,
+      "loss": 0.3016,
+      "step": 1152
+    },
+    {
+      "epoch": 8.126872246696035,
+      "grad_norm": 0.18884213005685602,
+      "learning_rate": 0.00019079940485189334,
+      "loss": 0.2837,
+      "step": 1153
+    },
+    {
+      "epoch": 8.133920704845815,
+      "grad_norm": 0.15897255869936042,
+      "learning_rate": 0.00019078014586762358,
+      "loss": 0.2953,
+      "step": 1154
+    },
+    {
+      "epoch": 8.140969162995594,
+      "grad_norm": 0.17570760903403315,
+      "learning_rate": 0.0001907608677217251,
+      "loss": 0.2952,
+      "step": 1155
+    },
+    {
+      "epoch": 8.148017621145375,
+      "grad_norm": 0.16900058977289112,
+      "learning_rate": 0.0001907415704182671,
+      "loss": 0.2917,
+      "step": 1156
+    },
+    {
+      "epoch": 8.155066079295155,
+      "grad_norm": 0.15891200048781803,
+      "learning_rate": 0.0001907222539613227,
+      "loss": 0.2929,
+      "step": 1157
+    },
+    {
+      "epoch": 8.162114537444934,
+      "grad_norm": 0.17713314111655895,
+      "learning_rate": 0.0001907029183549693,
+      "loss": 0.2845,
+      "step": 1158
+    },
+    {
+      "epoch": 8.169162995594714,
+      "grad_norm": 0.17945385432591904,
+      "learning_rate": 0.00019068356360328815,
+      "loss": 0.2838,
+      "step": 1159
+    },
+    {
+      "epoch": 8.176211453744493,
+      "grad_norm": 0.19890133944503668,
+      "learning_rate": 0.00019066418971036458,
+      "loss": 0.29,
+      "step": 1160
+    },
+    {
+      "epoch": 8.183259911894273,
+      "grad_norm": 0.20605651162622454,
+      "learning_rate": 0.000190644796680288,
+      "loss": 0.2971,
+      "step": 1161
+    },
+    {
+      "epoch": 8.190308370044052,
+      "grad_norm": 0.18021636769422641,
+      "learning_rate": 0.00019062538451715183,
+      "loss": 0.2908,
+      "step": 1162
+    },
+    {
+      "epoch": 8.197356828193833,
+      "grad_norm": 0.1908062832512802,
+      "learning_rate": 0.00019060595322505357,
+      "loss": 0.2863,
+      "step": 1163
+    },
+    {
+      "epoch": 8.204405286343611,
+      "grad_norm": 0.17198946165254694,
+      "learning_rate": 0.00019058650280809466,
+      "loss": 0.2797,
+      "step": 1164
+    },
+    {
+      "epoch": 8.211453744493392,
+      "grad_norm": 0.1665696293608029,
+      "learning_rate": 0.0001905670332703807,
+      "loss": 0.2901,
+      "step": 1165
+    },
+    {
+      "epoch": 8.218502202643172,
+      "grad_norm": 0.20184645127876658,
+      "learning_rate": 0.00019054754461602126,
+      "loss": 0.2993,
+      "step": 1166
+    },
+    {
+      "epoch": 8.225550660792951,
+      "grad_norm": 0.16318585361516466,
+      "learning_rate": 0.00019052803684912992,
+      "loss": 0.2923,
+      "step": 1167
+    },
+    {
+      "epoch": 8.232599118942732,
+      "grad_norm": 0.17183748059644469,
+      "learning_rate": 0.00019050850997382433,
+      "loss": 0.2838,
+      "step": 1168
+    },
+    {
+      "epoch": 8.23964757709251,
+      "grad_norm": 0.1731543001804214,
+      "learning_rate": 0.0001904889639942262,
+      "loss": 0.292,
+      "step": 1169
+    },
+    {
+      "epoch": 8.246696035242291,
+      "grad_norm": 0.16444291229240793,
+      "learning_rate": 0.00019046939891446125,
+      "loss": 0.2874,
+      "step": 1170
+    },
+    {
+      "epoch": 8.25374449339207,
+      "grad_norm": 0.16213298859666656,
+      "learning_rate": 0.0001904498147386592,
+      "loss": 0.2985,
+      "step": 1171
+    },
+    {
+      "epoch": 8.26079295154185,
+      "grad_norm": 0.1762022636335075,
+      "learning_rate": 0.00019043021147095382,
+      "loss": 0.2923,
+      "step": 1172
+    },
+    {
+      "epoch": 8.26784140969163,
+      "grad_norm": 0.15505673939015158,
+      "learning_rate": 0.00019041058911548292,
+      "loss": 0.2717,
+      "step": 1173
+    },
+    {
+      "epoch": 8.27488986784141,
+      "grad_norm": 0.17652417263302417,
+      "learning_rate": 0.00019039094767638832,
+      "loss": 0.2978,
+      "step": 1174
+    },
+    {
+      "epoch": 8.28193832599119,
+      "grad_norm": 0.1737745916418488,
+      "learning_rate": 0.00019037128715781591,
+      "loss": 0.2913,
+      "step": 1175
+    },
+    {
+      "epoch": 8.288986784140969,
+      "grad_norm": 0.17652880427012427,
+      "learning_rate": 0.0001903516075639156,
+      "loss": 0.3027,
+      "step": 1176
+    },
+    {
+      "epoch": 8.296035242290749,
+      "grad_norm": 0.16975885893881465,
+      "learning_rate": 0.00019033190889884124,
+      "loss": 0.2859,
+      "step": 1177
+    },
+    {
+      "epoch": 8.303083700440528,
+      "grad_norm": 0.16551629958814065,
+      "learning_rate": 0.00019031219116675082,
+      "loss": 0.2985,
+      "step": 1178
+    },
+    {
+      "epoch": 8.310132158590308,
+      "grad_norm": 0.16883559107589188,
+      "learning_rate": 0.00019029245437180632,
+      "loss": 0.2767,
+      "step": 1179
+    },
+    {
+      "epoch": 8.317180616740089,
+      "grad_norm": 0.17280085337674272,
+      "learning_rate": 0.00019027269851817365,
+      "loss": 0.293,
+      "step": 1180
+    },
+    {
+      "epoch": 8.324229074889868,
+      "grad_norm": 0.17307831501350887,
+      "learning_rate": 0.0001902529236100229,
+      "loss": 0.2947,
+      "step": 1181
+    },
+    {
+      "epoch": 8.331277533039648,
+      "grad_norm": 0.15394133885883451,
+      "learning_rate": 0.00019023312965152806,
+      "loss": 0.2975,
+      "step": 1182
+    },
+    {
+      "epoch": 8.338325991189427,
+      "grad_norm": 0.17177453474494886,
+      "learning_rate": 0.00019021331664686723,
+      "loss": 0.3046,
+      "step": 1183
+    },
+    {
+      "epoch": 8.345374449339207,
+      "grad_norm": 0.15390513552975152,
+      "learning_rate": 0.00019019348460022246,
+      "loss": 0.2933,
+      "step": 1184
+    },
+    {
+      "epoch": 8.352422907488986,
+      "grad_norm": 0.1630618511451408,
+      "learning_rate": 0.00019017363351577982,
+      "loss": 0.2808,
+      "step": 1185
+    },
+    {
+      "epoch": 8.359471365638766,
+      "grad_norm": 0.17294466788001964,
+      "learning_rate": 0.00019015376339772945,
+      "loss": 0.2902,
+      "step": 1186
+    },
+    {
+      "epoch": 8.366519823788547,
+      "grad_norm": 0.14408506934534027,
+      "learning_rate": 0.0001901338742502655,
+      "loss": 0.2797,
+      "step": 1187
+    },
+    {
+      "epoch": 8.373568281938326,
+      "grad_norm": 0.16773287205514742,
+      "learning_rate": 0.00019011396607758607,
+      "loss": 0.3015,
+      "step": 1188
+    },
+    {
+      "epoch": 8.380616740088106,
+      "grad_norm": 0.15121091168940193,
+      "learning_rate": 0.00019009403888389337,
+      "loss": 0.2852,
+      "step": 1189
+    },
+    {
+      "epoch": 8.387665198237885,
+      "grad_norm": 0.17331107761491704,
+      "learning_rate": 0.00019007409267339355,
+      "loss": 0.3007,
+      "step": 1190
+    },
+    {
+      "epoch": 8.394713656387665,
+      "grad_norm": 0.15352006209893657,
+      "learning_rate": 0.0001900541274502968,
+      "loss": 0.29,
+      "step": 1191
+    },
+    {
+      "epoch": 8.401762114537444,
+      "grad_norm": 0.15394952015573476,
+      "learning_rate": 0.00019003414321881734,
+      "loss": 0.2992,
+      "step": 1192
+    },
+    {
+      "epoch": 8.408810572687225,
+      "grad_norm": 0.15882313579187696,
+      "learning_rate": 0.00019001413998317342,
+      "loss": 0.2943,
+      "step": 1193
+    },
+    {
+      "epoch": 8.415859030837005,
+      "grad_norm": 0.1539219172385755,
+      "learning_rate": 0.0001899941177475872,
+      "loss": 0.2947,
+      "step": 1194
+    },
+    {
+      "epoch": 8.422907488986784,
+      "grad_norm": 0.15384603763453242,
+      "learning_rate": 0.00018997407651628493,
+      "loss": 0.2923,
+      "step": 1195
+    },
+    {
+      "epoch": 8.429955947136564,
+      "grad_norm": 0.17128793216015006,
+      "learning_rate": 0.00018995401629349692,
+      "loss": 0.2875,
+      "step": 1196
+    },
+    {
+      "epoch": 8.437004405286343,
+      "grad_norm": 0.17152794992299814,
+      "learning_rate": 0.00018993393708345737,
+      "loss": 0.2978,
+      "step": 1197
+    },
+    {
+      "epoch": 8.444052863436124,
+      "grad_norm": 0.18305957031174386,
+      "learning_rate": 0.00018991383889040457,
+      "loss": 0.2926,
+      "step": 1198
+    },
+    {
+      "epoch": 8.451101321585902,
+      "grad_norm": 0.1645014158269071,
+      "learning_rate": 0.00018989372171858077,
+      "loss": 0.2907,
+      "step": 1199
+    },
+    {
+      "epoch": 8.458149779735683,
+      "grad_norm": 0.16541973989689457,
+      "learning_rate": 0.00018987358557223232,
+      "loss": 0.2876,
+      "step": 1200
+    },
+    {
+      "epoch": 8.465198237885463,
+      "grad_norm": 0.16417550464997968,
+      "learning_rate": 0.0001898534304556094,
+      "loss": 0.2882,
+      "step": 1201
+    },
+    {
+      "epoch": 8.472246696035242,
+      "grad_norm": 0.17671594312236205,
+      "learning_rate": 0.00018983325637296635,
+      "loss": 0.285,
+      "step": 1202
+    },
+    {
+      "epoch": 8.479295154185023,
+      "grad_norm": 0.16203753042223626,
+      "learning_rate": 0.0001898130633285615,
+      "loss": 0.2863,
+      "step": 1203
+    },
+    {
+      "epoch": 8.486343612334801,
+      "grad_norm": 0.17680043451001232,
+      "learning_rate": 0.00018979285132665705,
+      "loss": 0.2956,
+      "step": 1204
+    },
+    {
+      "epoch": 8.493392070484582,
+      "grad_norm": 0.15200490903465064,
+      "learning_rate": 0.00018977262037151937,
+      "loss": 0.2926,
+      "step": 1205
+    },
+    {
+      "epoch": 8.50044052863436,
+      "grad_norm": 0.17037941305111676,
+      "learning_rate": 0.00018975237046741871,
+      "loss": 0.2882,
+      "step": 1206
+    },
+    {
+      "epoch": 8.507488986784141,
+      "grad_norm": 0.17390751467746957,
+      "learning_rate": 0.00018973210161862944,
+      "loss": 0.311,
+      "step": 1207
+    },
+    {
+      "epoch": 8.514537444933922,
+      "grad_norm": 0.1420279686799837,
+      "learning_rate": 0.00018971181382942975,
+      "loss": 0.3072,
+      "step": 1208
+    },
+    {
+      "epoch": 8.5215859030837,
+      "grad_norm": 0.16387746842968884,
+      "learning_rate": 0.000189691507104102,
+      "loss": 0.304,
+      "step": 1209
+    },
+    {
+      "epoch": 8.52863436123348,
+      "grad_norm": 0.14909576919943188,
+      "learning_rate": 0.00018967118144693246,
+      "loss": 0.2892,
+      "step": 1210
+    },
+    {
+      "epoch": 8.53568281938326,
+      "grad_norm": 0.1741622729190691,
+      "learning_rate": 0.00018965083686221142,
+      "loss": 0.3061,
+      "step": 1211
+    },
+    {
+      "epoch": 8.54273127753304,
+      "grad_norm": 0.1652450004664769,
+      "learning_rate": 0.00018963047335423316,
+      "loss": 0.2928,
+      "step": 1212
+    },
+    {
+      "epoch": 8.549779735682819,
+      "grad_norm": 0.15755834876283295,
+      "learning_rate": 0.000189610090927296,
+      "loss": 0.2849,
+      "step": 1213
+    },
+    {
+      "epoch": 8.5568281938326,
+      "grad_norm": 0.15476237422639996,
+      "learning_rate": 0.00018958968958570206,
+      "loss": 0.2953,
+      "step": 1214
+    },
+    {
+      "epoch": 8.56387665198238,
+      "grad_norm": 0.1625141872266953,
+      "learning_rate": 0.00018956926933375777,
+      "loss": 0.287,
+      "step": 1215
+    },
+    {
+      "epoch": 8.570925110132158,
+      "grad_norm": 0.15465616966233886,
+      "learning_rate": 0.0001895488301757733,
+      "loss": 0.3002,
+      "step": 1216
+    },
+    {
+      "epoch": 8.577973568281939,
+      "grad_norm": 0.1696127726610247,
+      "learning_rate": 0.00018952837211606288,
+      "loss": 0.2956,
+      "step": 1217
+    },
+    {
+      "epoch": 8.585022026431718,
+      "grad_norm": 0.14918937089038764,
+      "learning_rate": 0.0001895078951589448,
+      "loss": 0.2871,
+      "step": 1218
+    },
+    {
+      "epoch": 8.592070484581498,
+      "grad_norm": 0.1694994039018744,
+      "learning_rate": 0.00018948739930874122,
+      "loss": 0.2852,
+      "step": 1219
+    },
+    {
+      "epoch": 8.599118942731277,
+      "grad_norm": 0.17553189599991734,
+      "learning_rate": 0.00018946688456977838,
+      "loss": 0.3095,
+      "step": 1220
+    },
+    {
+      "epoch": 8.606167400881057,
+      "grad_norm": 0.15532045303455277,
+      "learning_rate": 0.00018944635094638647,
+      "loss": 0.2795,
+      "step": 1221
+    },
+    {
+      "epoch": 8.613215859030838,
+      "grad_norm": 0.15788682494301284,
+      "learning_rate": 0.00018942579844289967,
+      "loss": 0.2886,
+      "step": 1222
+    },
+    {
+      "epoch": 8.620264317180617,
+      "grad_norm": 0.14993887084326996,
+      "learning_rate": 0.00018940522706365617,
+      "loss": 0.2859,
+      "step": 1223
+    },
+    {
+      "epoch": 8.627312775330397,
+      "grad_norm": 0.1603725886421236,
+      "learning_rate": 0.0001893846368129981,
+      "loss": 0.3054,
+      "step": 1224
+    },
+    {
+      "epoch": 8.634361233480176,
+      "grad_norm": 0.1623756106137082,
+      "learning_rate": 0.00018936402769527158,
+      "loss": 0.2914,
+      "step": 1225
+    },
+    {
+      "epoch": 8.641409691629956,
+      "grad_norm": 0.16014022217712875,
+      "learning_rate": 0.00018934339971482674,
+      "loss": 0.2901,
+      "step": 1226
+    },
+    {
+      "epoch": 8.648458149779735,
+      "grad_norm": 0.17844424131146655,
+      "learning_rate": 0.00018932275287601773,
+      "loss": 0.2989,
+      "step": 1227
+    },
+    {
+      "epoch": 8.655506607929516,
+      "grad_norm": 0.14886180578884667,
+      "learning_rate": 0.00018930208718320252,
+      "loss": 0.2911,
+      "step": 1228
+    },
+    {
+      "epoch": 8.662555066079296,
+      "grad_norm": 0.14427063356018097,
+      "learning_rate": 0.00018928140264074328,
+      "loss": 0.2949,
+      "step": 1229
+    },
+    {
+      "epoch": 8.669603524229075,
+      "grad_norm": 0.1464896219401621,
+      "learning_rate": 0.000189260699253006,
+      "loss": 0.29,
+      "step": 1230
+    },
+    {
+      "epoch": 8.676651982378855,
+      "grad_norm": 0.1634663113275393,
+      "learning_rate": 0.00018923997702436065,
+      "loss": 0.2996,
+      "step": 1231
+    },
+    {
+      "epoch": 8.683700440528634,
+      "grad_norm": 0.15532934641998125,
+      "learning_rate": 0.00018921923595918128,
+      "loss": 0.3025,
+      "step": 1232
+    },
+    {
+      "epoch": 8.690748898678414,
+      "grad_norm": 0.1718514195781784,
+      "learning_rate": 0.00018919847606184583,
+      "loss": 0.3015,
+      "step": 1233
+    },
+    {
+      "epoch": 8.697797356828193,
+      "grad_norm": 0.14932066360449334,
+      "learning_rate": 0.0001891776973367363,
+      "loss": 0.2949,
+      "step": 1234
+    },
+    {
+      "epoch": 8.704845814977974,
+      "grad_norm": 0.15700403812177682,
+      "learning_rate": 0.00018915689978823855,
+      "loss": 0.2986,
+      "step": 1235
+    },
+    {
+      "epoch": 8.711894273127754,
+      "grad_norm": 0.19127152219197144,
+      "learning_rate": 0.00018913608342074247,
+      "loss": 0.2955,
+      "step": 1236
+    },
+    {
+      "epoch": 8.718942731277533,
+      "grad_norm": 0.14675518688693287,
+      "learning_rate": 0.00018911524823864192,
+      "loss": 0.299,
+      "step": 1237
+    },
+    {
+      "epoch": 8.725991189427313,
+      "grad_norm": 0.170404356333295,
+      "learning_rate": 0.0001890943942463348,
+      "loss": 0.2859,
+      "step": 1238
+    },
+    {
+      "epoch": 8.733039647577092,
+      "grad_norm": 0.15888920993810451,
+      "learning_rate": 0.00018907352144822284,
+      "loss": 0.2846,
+      "step": 1239
+    },
+    {
+      "epoch": 8.740088105726873,
+      "grad_norm": 0.1718448602959601,
+      "learning_rate": 0.0001890526298487118,
+      "loss": 0.2858,
+      "step": 1240
+    },
+    {
+      "epoch": 8.747136563876651,
+      "grad_norm": 0.15891233806966026,
+      "learning_rate": 0.00018903171945221155,
+      "loss": 0.2915,
+      "step": 1241
+    },
+    {
+      "epoch": 8.754185022026432,
+      "grad_norm": 0.15382595258479645,
+      "learning_rate": 0.00018901079026313565,
+      "loss": 0.2864,
+      "step": 1242
+    },
+    {
+      "epoch": 8.761233480176212,
+      "grad_norm": 0.1583906192607324,
+      "learning_rate": 0.00018898984228590186,
+      "loss": 0.3026,
+      "step": 1243
+    },
+    {
+      "epoch": 8.768281938325991,
+      "grad_norm": 0.1620410308838539,
+      "learning_rate": 0.0001889688755249318,
+      "loss": 0.2843,
+      "step": 1244
+    },
+    {
+      "epoch": 8.775330396475772,
+      "grad_norm": 0.15636270922404327,
+      "learning_rate": 0.00018894788998465107,
+      "loss": 0.3071,
+      "step": 1245
+    },
+    {
+      "epoch": 8.78237885462555,
+      "grad_norm": 0.16800018849172427,
+      "learning_rate": 0.00018892688566948926,
+      "loss": 0.2908,
+      "step": 1246
+    },
+    {
+      "epoch": 8.78942731277533,
+      "grad_norm": 0.15551100461192266,
+      "learning_rate": 0.0001889058625838799,
+      "loss": 0.2967,
+      "step": 1247
+    },
+    {
+      "epoch": 8.79647577092511,
+      "grad_norm": 0.1380825268348693,
+      "learning_rate": 0.00018888482073226048,
+      "loss": 0.2917,
+      "step": 1248
+    },
+    {
+      "epoch": 8.80352422907489,
+      "grad_norm": 0.1527348507018593,
+      "learning_rate": 0.00018886376011907245,
+      "loss": 0.2778,
+      "step": 1249
+    },
+    {
+      "epoch": 8.810572687224669,
+      "grad_norm": 0.1517353555899026,
+      "learning_rate": 0.00018884268074876123,
+      "loss": 0.2935,
+      "step": 1250
+    },
+    {
+      "epoch": 8.81762114537445,
+      "grad_norm": 0.1525561993203098,
+      "learning_rate": 0.00018882158262577625,
+      "loss": 0.3005,
+      "step": 1251
+    },
+    {
+      "epoch": 8.82466960352423,
+      "grad_norm": 0.16579763952525547,
+      "learning_rate": 0.00018880046575457074,
+      "loss": 0.301,
+      "step": 1252
+    },
+    {
+      "epoch": 8.831718061674009,
+      "grad_norm": 0.14911413733987358,
+      "learning_rate": 0.0001887793301396021,
+      "loss": 0.296,
+      "step": 1253
+    },
+    {
+      "epoch": 8.838766519823789,
+      "grad_norm": 0.15400517061253016,
+      "learning_rate": 0.0001887581757853315,
+      "loss": 0.2858,
+      "step": 1254
+    },
+    {
+      "epoch": 8.845814977973568,
+      "grad_norm": 0.15782383525374657,
+      "learning_rate": 0.00018873700269622415,
+      "loss": 0.2928,
+      "step": 1255
+    },
+    {
+      "epoch": 8.852863436123348,
+      "grad_norm": 0.15491344937789928,
+      "learning_rate": 0.0001887158108767493,
+      "loss": 0.2979,
+      "step": 1256
+    },
+    {
+      "epoch": 8.859911894273127,
+      "grad_norm": 0.14926914187029205,
+      "learning_rate": 0.00018869460033137992,
+      "loss": 0.2874,
+      "step": 1257
+    },
+    {
+      "epoch": 8.866960352422907,
+      "grad_norm": 0.14993396546583093,
+      "learning_rate": 0.00018867337106459318,
+      "loss": 0.2893,
+      "step": 1258
+    },
+    {
+      "epoch": 8.874008810572688,
+      "grad_norm": 0.14844315331141172,
+      "learning_rate": 0.00018865212308087008,
+      "loss": 0.3017,
+      "step": 1259
+    },
+    {
+      "epoch": 8.881057268722467,
+      "grad_norm": 0.17280247573018523,
+      "learning_rate": 0.00018863085638469557,
+      "loss": 0.3018,
+      "step": 1260
+    },
+    {
+      "epoch": 8.888105726872247,
+      "grad_norm": 0.15338399902013022,
+      "learning_rate": 0.00018860957098055855,
+      "loss": 0.2935,
+      "step": 1261
+    },
+    {
+      "epoch": 8.895154185022026,
+      "grad_norm": 0.1501844357513368,
+      "learning_rate": 0.00018858826687295194,
+      "loss": 0.2988,
+      "step": 1262
+    },
+    {
+      "epoch": 8.902202643171806,
+      "grad_norm": 0.17091771442821388,
+      "learning_rate": 0.0001885669440663725,
+      "loss": 0.2952,
+      "step": 1263
+    },
+    {
+      "epoch": 8.909251101321585,
+      "grad_norm": 0.16181158436445342,
+      "learning_rate": 0.000188545602565321,
+      "loss": 0.2869,
+      "step": 1264
+    },
+    {
+      "epoch": 8.916299559471366,
+      "grad_norm": 0.1698402780273519,
+      "learning_rate": 0.00018852424237430216,
+      "loss": 0.2937,
+      "step": 1265
+    },
+    {
+      "epoch": 8.923348017621146,
+      "grad_norm": 0.1702196591506054,
+      "learning_rate": 0.00018850286349782463,
+      "loss": 0.306,
+      "step": 1266
+    },
+    {
+      "epoch": 8.930396475770925,
+      "grad_norm": 0.18517491891761312,
+      "learning_rate": 0.000188481465940401,
+      "loss": 0.287,
+      "step": 1267
+    },
+    {
+      "epoch": 8.937444933920705,
+      "grad_norm": 0.15348331318009034,
+      "learning_rate": 0.00018846004970654782,
+      "loss": 0.2915,
+      "step": 1268
+    },
+    {
+      "epoch": 8.944493392070484,
+      "grad_norm": 0.16878504853984447,
+      "learning_rate": 0.00018843861480078559,
+      "loss": 0.2869,
+      "step": 1269
+    },
+    {
+      "epoch": 8.951541850220265,
+      "grad_norm": 0.14499481869490471,
+      "learning_rate": 0.00018841716122763866,
+      "loss": 0.2951,
+      "step": 1270
+    },
+    {
+      "epoch": 8.958590308370043,
+      "grad_norm": 0.173954000525515,
+      "learning_rate": 0.00018839568899163545,
+      "loss": 0.2825,
+      "step": 1271
+    },
+    {
+      "epoch": 8.965638766519824,
+      "grad_norm": 0.1550870809285605,
+      "learning_rate": 0.00018837419809730825,
+      "loss": 0.2866,
+      "step": 1272
+    },
+    {
+      "epoch": 8.972687224669604,
+      "grad_norm": 0.14990896926899813,
+      "learning_rate": 0.0001883526885491933,
+      "loss": 0.2972,
+      "step": 1273
+    },
+    {
+      "epoch": 8.979735682819383,
+      "grad_norm": 0.14394056544478376,
+      "learning_rate": 0.0001883311603518308,
+      "loss": 0.2969,
+      "step": 1274
+    },
+    {
+      "epoch": 8.986784140969164,
+      "grad_norm": 0.1493144002013214,
+      "learning_rate": 0.00018830961350976483,
+      "loss": 0.2928,
+      "step": 1275
+    },
+    {
+      "epoch": 8.993832599118942,
+      "grad_norm": 0.15018264718267474,
+      "learning_rate": 0.00018828804802754342,
+      "loss": 0.3083,
+      "step": 1276
+    },
+    {
+      "epoch": 9.000881057268723,
+      "grad_norm": 0.15465519986435503,
+      "learning_rate": 0.00018826646390971864,
+      "loss": 0.3062,
+      "step": 1277
+    },
+    {
+      "epoch": 9.007929515418502,
+      "grad_norm": 0.1470122582799358,
+      "learning_rate": 0.0001882448611608463,
+      "loss": 0.265,
+      "step": 1278
+    },
+    {
+      "epoch": 9.014977973568282,
+      "grad_norm": 0.18229508776128528,
+      "learning_rate": 0.00018822323978548634,
+      "loss": 0.2709,
+      "step": 1279
+    },
+    {
+      "epoch": 9.022026431718063,
+      "grad_norm": 0.1933589571414497,
+      "learning_rate": 0.00018820159978820246,
+      "loss": 0.2616,
+      "step": 1280
+    },
+    {
+      "epoch": 9.029074889867841,
+      "grad_norm": 0.1819913395567821,
+      "learning_rate": 0.00018817994117356247,
+      "loss": 0.2684,
+      "step": 1281
+    },
+    {
+      "epoch": 9.036123348017622,
+      "grad_norm": 0.17744894190776747,
+      "learning_rate": 0.00018815826394613793,
+      "loss": 0.2656,
+      "step": 1282
+    },
+    {
+      "epoch": 9.0431718061674,
+      "grad_norm": 0.1729436641968048,
+      "learning_rate": 0.00018813656811050443,
+      "loss": 0.257,
+      "step": 1283
+    },
+    {
+      "epoch": 9.050220264317181,
+      "grad_norm": 0.17945239119766074,
+      "learning_rate": 0.00018811485367124152,
+      "loss": 0.284,
+      "step": 1284
+    },
+    {
+      "epoch": 9.05726872246696,
+      "grad_norm": 0.20757409890360362,
+      "learning_rate": 0.0001880931206329325,
+      "loss": 0.2664,
+      "step": 1285
+    },
+    {
+      "epoch": 9.06431718061674,
+      "grad_norm": 0.18737556393606236,
+      "learning_rate": 0.0001880713690001649,
+      "loss": 0.2601,
+      "step": 1286
+    },
+    {
+      "epoch": 9.07136563876652,
+      "grad_norm": 0.2409964642080102,
+      "learning_rate": 0.00018804959877752984,
+      "loss": 0.2769,
+      "step": 1287
+    },
+    {
+      "epoch": 9.0784140969163,
+      "grad_norm": 0.18239291559099494,
+      "learning_rate": 0.0001880278099696226,
+      "loss": 0.2646,
+      "step": 1288
+    },
+    {
+      "epoch": 9.08546255506608,
+      "grad_norm": 0.22110682379668878,
+      "learning_rate": 0.00018800600258104226,
+      "loss": 0.2604,
+      "step": 1289
+    },
+    {
+      "epoch": 9.092511013215859,
+      "grad_norm": 0.16652129619870198,
+      "learning_rate": 0.0001879841766163919,
+      "loss": 0.2643,
+      "step": 1290
+    },
+    {
+      "epoch": 9.09955947136564,
+      "grad_norm": 0.1989412911771759,
+      "learning_rate": 0.0001879623320802785,
+      "loss": 0.2739,
+      "step": 1291
+    },
+    {
+      "epoch": 9.106607929515418,
+      "grad_norm": 0.1654749101602937,
+      "learning_rate": 0.0001879404689773129,
+      "loss": 0.2693,
+      "step": 1292
+    },
+    {
+      "epoch": 9.113656387665198,
+      "grad_norm": 0.20807708217265497,
+      "learning_rate": 0.00018791858731210991,
+      "loss": 0.2769,
+      "step": 1293
+    },
+    {
+      "epoch": 9.120704845814977,
+      "grad_norm": 0.1925322131076615,
+      "learning_rate": 0.0001878966870892883,
+      "loss": 0.2759,
+      "step": 1294
+    },
+    {
+      "epoch": 9.127753303964758,
+      "grad_norm": 0.18850641093591552,
+      "learning_rate": 0.00018787476831347066,
+      "loss": 0.2817,
+      "step": 1295
+    },
+    {
+      "epoch": 9.134801762114538,
+      "grad_norm": 0.19080052519910193,
+      "learning_rate": 0.00018785283098928356,
+      "loss": 0.2778,
+      "step": 1296
+    },
+    {
+      "epoch": 9.141850220264317,
+      "grad_norm": 0.1821426884387287,
+      "learning_rate": 0.00018783087512135748,
+      "loss": 0.2894,
+      "step": 1297
+    },
+    {
+      "epoch": 9.148898678414097,
+      "grad_norm": 0.17024348832995365,
+      "learning_rate": 0.0001878089007143268,
+      "loss": 0.2711,
+      "step": 1298
+    },
+    {
+      "epoch": 9.155947136563876,
+      "grad_norm": 0.17994325119425794,
+      "learning_rate": 0.00018778690777282981,
+      "loss": 0.2722,
+      "step": 1299
+    },
+    {
+      "epoch": 9.162995594713657,
+      "grad_norm": 0.2040442001259944,
+      "learning_rate": 0.00018776489630150875,
+      "loss": 0.2712,
+      "step": 1300
+    },
+    {
+      "epoch": 9.170044052863435,
+      "grad_norm": 0.16797463029217274,
+      "learning_rate": 0.00018774286630500974,
+      "loss": 0.2647,
+      "step": 1301
+    },
+    {
+      "epoch": 9.177092511013216,
+      "grad_norm": 0.18246697163003053,
+      "learning_rate": 0.00018772081778798273,
+      "loss": 0.2734,
+      "step": 1302
+    },
+    {
+      "epoch": 9.184140969162996,
+      "grad_norm": 0.168212869914223,
+      "learning_rate": 0.00018769875075508178,
+      "loss": 0.2885,
+      "step": 1303
+    },
+    {
+      "epoch": 9.191189427312775,
+      "grad_norm": 0.18769283794012778,
+      "learning_rate": 0.00018767666521096466,
+      "loss": 0.2694,
+      "step": 1304
+    },
+    {
+      "epoch": 9.198237885462555,
+      "grad_norm": 0.18757783535882805,
+      "learning_rate": 0.00018765456116029316,
+      "loss": 0.2838,
+      "step": 1305
+    },
+    {
+      "epoch": 9.205286343612334,
+      "grad_norm": 0.20386639524600508,
+      "learning_rate": 0.00018763243860773295,
+      "loss": 0.2677,
+      "step": 1306
+    },
+    {
+      "epoch": 9.212334801762115,
+      "grad_norm": 0.1773657872754707,
+      "learning_rate": 0.00018761029755795357,
+      "loss": 0.2649,
+      "step": 1307
+    },
+    {
+      "epoch": 9.219383259911893,
+      "grad_norm": 0.17299128070373568,
+      "learning_rate": 0.0001875881380156285,
+      "loss": 0.2681,
+      "step": 1308
+    },
+    {
+      "epoch": 9.226431718061674,
+      "grad_norm": 0.1916855019215153,
+      "learning_rate": 0.00018756595998543514,
+      "loss": 0.2762,
+      "step": 1309
+    },
+    {
+      "epoch": 9.233480176211454,
+      "grad_norm": 0.17861133794459577,
+      "learning_rate": 0.00018754376347205476,
+      "loss": 0.2771,
+      "step": 1310
+    },
+    {
+      "epoch": 9.240528634361233,
+      "grad_norm": 0.17716203466264982,
+      "learning_rate": 0.00018752154848017253,
+      "loss": 0.2755,
+      "step": 1311
+    },
+    {
+      "epoch": 9.247577092511014,
+      "grad_norm": 0.17659609396555928,
+      "learning_rate": 0.00018749931501447752,
+      "loss": 0.2837,
+      "step": 1312
+    },
+    {
+      "epoch": 9.254625550660792,
+      "grad_norm": 0.18158110718383652,
+      "learning_rate": 0.00018747706307966272,
+      "loss": 0.2717,
+      "step": 1313
+    },
+    {
+      "epoch": 9.261674008810573,
+      "grad_norm": 0.18175231388040394,
+      "learning_rate": 0.00018745479268042505,
+      "loss": 0.2686,
+      "step": 1314
+    },
+    {
+      "epoch": 9.268722466960352,
+      "grad_norm": 0.18732217312646599,
+      "learning_rate": 0.00018743250382146522,
+      "loss": 0.2637,
+      "step": 1315
+    },
+    {
+      "epoch": 9.275770925110132,
+      "grad_norm": 0.19751954530599536,
+      "learning_rate": 0.00018741019650748796,
+      "loss": 0.2721,
+      "step": 1316
+    },
+    {
+      "epoch": 9.282819383259913,
+      "grad_norm": 0.18711847089722808,
+      "learning_rate": 0.00018738787074320179,
+      "loss": 0.2837,
+      "step": 1317
+    },
+    {
+      "epoch": 9.289867841409691,
+      "grad_norm": 0.1820814352109305,
+      "learning_rate": 0.0001873655265333192,
+      "loss": 0.2803,
+      "step": 1318
+    },
+    {
+      "epoch": 9.296916299559472,
+      "grad_norm": 0.17580932719018189,
+      "learning_rate": 0.0001873431638825565,
+      "loss": 0.2718,
+      "step": 1319
+    },
+    {
+      "epoch": 9.30396475770925,
+      "grad_norm": 0.20438582259270865,
+      "learning_rate": 0.00018732078279563402,
+      "loss": 0.2726,
+      "step": 1320
+    },
+    {
+      "epoch": 9.311013215859031,
+      "grad_norm": 0.18776116709398274,
+      "learning_rate": 0.00018729838327727585,
+      "loss": 0.2853,
+      "step": 1321
+    },
+    {
+      "epoch": 9.31806167400881,
+      "grad_norm": 0.21108523299037704,
+      "learning_rate": 0.00018727596533220999,
+      "loss": 0.2757,
+      "step": 1322
+    },
+    {
+      "epoch": 9.32511013215859,
+      "grad_norm": 0.20050198147183618,
+      "learning_rate": 0.00018725352896516842,
+      "loss": 0.2891,
+      "step": 1323
+    },
+    {
+      "epoch": 9.33215859030837,
+      "grad_norm": 0.17596313990090565,
+      "learning_rate": 0.00018723107418088693,
+      "loss": 0.2744,
+      "step": 1324
+    },
+    {
+      "epoch": 9.33920704845815,
+      "grad_norm": 0.18387542347910132,
+      "learning_rate": 0.0001872086009841052,
+      "loss": 0.2808,
+      "step": 1325
+    },
+    {
+      "epoch": 9.34625550660793,
+      "grad_norm": 0.19898644972765173,
+      "learning_rate": 0.0001871861093795668,
+      "loss": 0.2782,
+      "step": 1326
+    },
+    {
+      "epoch": 9.353303964757709,
+      "grad_norm": 0.1730698531432664,
+      "learning_rate": 0.00018716359937201926,
+      "loss": 0.2712,
+      "step": 1327
+    },
+    {
+      "epoch": 9.36035242290749,
+      "grad_norm": 0.21429778722004567,
+      "learning_rate": 0.00018714107096621388,
+      "loss": 0.276,
+      "step": 1328
+    },
+    {
+      "epoch": 9.367400881057268,
+      "grad_norm": 0.1796452927526479,
+      "learning_rate": 0.00018711852416690588,
+      "loss": 0.2767,
+      "step": 1329
+    },
+    {
+      "epoch": 9.374449339207048,
+      "grad_norm": 0.1677945570129597,
+      "learning_rate": 0.00018709595897885439,
+      "loss": 0.2594,
+      "step": 1330
+    },
+    {
+      "epoch": 9.381497797356829,
+      "grad_norm": 0.18339479255128058,
+      "learning_rate": 0.00018707337540682247,
+      "loss": 0.2918,
+      "step": 1331
+    },
+    {
+      "epoch": 9.388546255506608,
+      "grad_norm": 0.16744955745704268,
+      "learning_rate": 0.00018705077345557692,
+      "loss": 0.2769,
+      "step": 1332
+    },
+    {
+      "epoch": 9.395594713656388,
+      "grad_norm": 0.20449075291716287,
+      "learning_rate": 0.00018702815312988852,
+      "loss": 0.2796,
+      "step": 1333
+    },
+    {
+      "epoch": 9.402643171806167,
+      "grad_norm": 0.17378146911535522,
+      "learning_rate": 0.00018700551443453194,
+      "loss": 0.2813,
+      "step": 1334
+    },
+    {
+      "epoch": 9.409691629955947,
+      "grad_norm": 0.2002319725010115,
+      "learning_rate": 0.00018698285737428564,
+      "loss": 0.2849,
+      "step": 1335
+    },
+    {
+      "epoch": 9.416740088105726,
+      "grad_norm": 0.20876400348476212,
+      "learning_rate": 0.00018696018195393208,
+      "loss": 0.286,
+      "step": 1336
+    },
+    {
+      "epoch": 9.423788546255507,
+      "grad_norm": 0.17393749582577722,
+      "learning_rate": 0.00018693748817825748,
+      "loss": 0.2769,
+      "step": 1337
+    },
+    {
+      "epoch": 9.430837004405287,
+      "grad_norm": 0.20659579001382086,
+      "learning_rate": 0.00018691477605205199,
+      "loss": 0.2949,
+      "step": 1338
+    },
+    {
+      "epoch": 9.437885462555066,
+      "grad_norm": 0.17743207370993735,
+      "learning_rate": 0.00018689204558010962,
+      "loss": 0.263,
+      "step": 1339
+    },
+    {
+      "epoch": 9.444933920704846,
+      "grad_norm": 0.17764192680623223,
+      "learning_rate": 0.00018686929676722827,
+      "loss": 0.2798,
+      "step": 1340
+    },
+    {
+      "epoch": 9.451982378854625,
+      "grad_norm": 0.18686140096922982,
+      "learning_rate": 0.00018684652961820967,
+      "loss": 0.281,
+      "step": 1341
+    },
+    {
+      "epoch": 9.459030837004406,
+      "grad_norm": 0.19764524506425166,
+      "learning_rate": 0.00018682374413785948,
+      "loss": 0.2804,
+      "step": 1342
+    },
+    {
+      "epoch": 9.466079295154184,
+      "grad_norm": 0.17545723690117004,
+      "learning_rate": 0.00018680094033098716,
+      "loss": 0.2756,
+      "step": 1343
+    },
+    {
+      "epoch": 9.473127753303965,
+      "grad_norm": 0.17559864123592062,
+      "learning_rate": 0.00018677811820240613,
+      "loss": 0.2761,
+      "step": 1344
+    },
+    {
+      "epoch": 9.480176211453745,
+      "grad_norm": 0.17043733658768856,
+      "learning_rate": 0.0001867552777569336,
+      "loss": 0.2735,
+      "step": 1345
+    },
+    {
+      "epoch": 9.487224669603524,
+      "grad_norm": 0.17350040930169788,
+      "learning_rate": 0.00018673241899939066,
+      "loss": 0.2732,
+      "step": 1346
+    },
+    {
+      "epoch": 9.494273127753305,
+      "grad_norm": 0.18750120471523704,
+      "learning_rate": 0.00018670954193460227,
+      "loss": 0.2791,
+      "step": 1347
+    },
+    {
+      "epoch": 9.501321585903083,
+      "grad_norm": 0.18838193444066226,
+      "learning_rate": 0.0001866866465673973,
+      "loss": 0.2854,
+      "step": 1348
+    },
+    {
+      "epoch": 9.508370044052864,
+      "grad_norm": 0.16818323130668522,
+      "learning_rate": 0.00018666373290260842,
+      "loss": 0.2753,
+      "step": 1349
+    },
+    {
+      "epoch": 9.515418502202643,
+      "grad_norm": 0.17371864087482794,
+      "learning_rate": 0.00018664080094507214,
+      "loss": 0.2717,
+      "step": 1350
+    },
+    {
+      "epoch": 9.522466960352423,
+      "grad_norm": 0.17155312175649262,
+      "learning_rate": 0.0001866178506996289,
+      "loss": 0.2798,
+      "step": 1351
+    },
+    {
+      "epoch": 9.529515418502204,
+      "grad_norm": 0.18990812256471848,
+      "learning_rate": 0.000186594882171123,
+      "loss": 0.2763,
+      "step": 1352
+    },
+    {
+      "epoch": 9.536563876651982,
+      "grad_norm": 0.17344228655767144,
+      "learning_rate": 0.00018657189536440261,
+      "loss": 0.2825,
+      "step": 1353
+    },
+    {
+      "epoch": 9.543612334801763,
+      "grad_norm": 0.15508273670551834,
+      "learning_rate": 0.00018654889028431963,
+      "loss": 0.2779,
+      "step": 1354
+    },
+    {
+      "epoch": 9.550660792951541,
+      "grad_norm": 0.1731636587632561,
+      "learning_rate": 0.00018652586693572998,
+      "loss": 0.2945,
+      "step": 1355
+    },
+    {
+      "epoch": 9.557709251101322,
+      "grad_norm": 0.1777181503784259,
+      "learning_rate": 0.00018650282532349332,
+      "loss": 0.2816,
+      "step": 1356
+    },
+    {
+      "epoch": 9.5647577092511,
+      "grad_norm": 0.17460643238016962,
+      "learning_rate": 0.00018647976545247324,
+      "loss": 0.2733,
+      "step": 1357
+    },
+    {
+      "epoch": 9.571806167400881,
+      "grad_norm": 0.18528039626893397,
+      "learning_rate": 0.00018645668732753722,
+      "loss": 0.2732,
+      "step": 1358
+    },
+    {
+      "epoch": 9.578854625550662,
+      "grad_norm": 0.18257989362570212,
+      "learning_rate": 0.00018643359095355636,
+      "loss": 0.2852,
+      "step": 1359
+    },
+    {
+      "epoch": 9.58590308370044,
+      "grad_norm": 0.17215551965115505,
+      "learning_rate": 0.00018641047633540593,
+      "loss": 0.2746,
+      "step": 1360
+    },
+    {
+      "epoch": 9.592951541850221,
+      "grad_norm": 0.1566309123727862,
+      "learning_rate": 0.00018638734347796484,
+      "loss": 0.281,
+      "step": 1361
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.15558432043530307,
+      "learning_rate": 0.00018636419238611594,
+      "loss": 0.2731,
+      "step": 1362
+    },
+    {
+      "epoch": 9.60704845814978,
+      "grad_norm": 0.16345549182491828,
+      "learning_rate": 0.00018634102306474586,
+      "loss": 0.2805,
+      "step": 1363
+    },
+    {
+      "epoch": 9.614096916299559,
+      "grad_norm": 0.18068947015584771,
+      "learning_rate": 0.00018631783551874514,
+      "loss": 0.2702,
+      "step": 1364
+    },
+    {
+      "epoch": 9.62114537444934,
+      "grad_norm": 0.1858356530402995,
+      "learning_rate": 0.00018629462975300817,
+      "loss": 0.2722,
+      "step": 1365
+    },
+    {
+      "epoch": 9.62819383259912,
+      "grad_norm": 0.17962398890989362,
+      "learning_rate": 0.00018627140577243313,
+      "loss": 0.2734,
+      "step": 1366
+    },
+    {
+      "epoch": 9.635242290748899,
+      "grad_norm": 0.16768200684781703,
+      "learning_rate": 0.00018624816358192207,
+      "loss": 0.2798,
+      "step": 1367
+    },
+    {
+      "epoch": 9.642290748898679,
+      "grad_norm": 0.17110182742453892,
+      "learning_rate": 0.00018622490318638088,
+      "loss": 0.2807,
+      "step": 1368
+    },
+    {
+      "epoch": 9.649339207048458,
+      "grad_norm": 0.17668042245692558,
+      "learning_rate": 0.00018620162459071936,
+      "loss": 0.2782,
+      "step": 1369
+    },
+    {
+      "epoch": 9.656387665198238,
+      "grad_norm": 0.16934014098204345,
+      "learning_rate": 0.00018617832779985104,
+      "loss": 0.2721,
+      "step": 1370
+    },
+    {
+      "epoch": 9.663436123348017,
+      "grad_norm": 0.19284503998129418,
+      "learning_rate": 0.00018615501281869334,
+      "loss": 0.2873,
+      "step": 1371
+    },
+    {
+      "epoch": 9.670484581497798,
+      "grad_norm": 0.17604781387625207,
+      "learning_rate": 0.00018613167965216758,
+      "loss": 0.2793,
+      "step": 1372
+    },
+    {
+      "epoch": 9.677533039647576,
+      "grad_norm": 0.18264996849446652,
+      "learning_rate": 0.0001861083283051988,
+      "loss": 0.2839,
+      "step": 1373
+    },
+    {
+      "epoch": 9.684581497797357,
+      "grad_norm": 0.1670713794013828,
+      "learning_rate": 0.00018608495878271597,
+      "loss": 0.285,
+      "step": 1374
+    },
+    {
+      "epoch": 9.691629955947137,
+      "grad_norm": 0.19174139178082278,
+      "learning_rate": 0.00018606157108965187,
+      "loss": 0.2854,
+      "step": 1375
+    },
+    {
+      "epoch": 9.698678414096916,
+      "grad_norm": 0.1759865077723311,
+      "learning_rate": 0.00018603816523094307,
+      "loss": 0.2929,
+      "step": 1376
+    },
+    {
+      "epoch": 9.705726872246697,
+      "grad_norm": 0.1776268952975606,
+      "learning_rate": 0.00018601474121153006,
+      "loss": 0.2734,
+      "step": 1377
+    },
+    {
+      "epoch": 9.712775330396475,
+      "grad_norm": 0.17458018154925972,
+      "learning_rate": 0.00018599129903635713,
+      "loss": 0.2908,
+      "step": 1378
+    },
+    {
+      "epoch": 9.719823788546256,
+      "grad_norm": 0.1710326891417709,
+      "learning_rate": 0.00018596783871037236,
+      "loss": 0.2743,
+      "step": 1379
+    },
+    {
+      "epoch": 9.726872246696034,
+      "grad_norm": 0.17874023432233954,
+      "learning_rate": 0.00018594436023852765,
+      "loss": 0.2752,
+      "step": 1380
+    },
+    {
+      "epoch": 9.733920704845815,
+      "grad_norm": 0.16913625711267316,
+      "learning_rate": 0.0001859208636257789,
+      "loss": 0.2821,
+      "step": 1381
+    },
+    {
+      "epoch": 9.740969162995595,
+      "grad_norm": 0.1677126656337971,
+      "learning_rate": 0.00018589734887708556,
+      "loss": 0.2739,
+      "step": 1382
+    },
+    {
+      "epoch": 9.748017621145374,
+      "grad_norm": 0.1765541694549315,
+      "learning_rate": 0.0001858738159974112,
+      "loss": 0.2846,
+      "step": 1383
+    },
+    {
+      "epoch": 9.755066079295155,
+      "grad_norm": 0.16981247267656407,
+      "learning_rate": 0.00018585026499172295,
+      "loss": 0.2911,
+      "step": 1384
+    },
+    {
+      "epoch": 9.762114537444933,
+      "grad_norm": 0.1545902208156347,
+      "learning_rate": 0.000185826695864992,
+      "loss": 0.2804,
+      "step": 1385
+    },
+    {
+      "epoch": 9.769162995594714,
+      "grad_norm": 0.1768094761206055,
+      "learning_rate": 0.0001858031086221932,
+      "loss": 0.2785,
+      "step": 1386
+    },
+    {
+      "epoch": 9.776211453744493,
+      "grad_norm": 0.15799508371667556,
+      "learning_rate": 0.00018577950326830527,
+      "loss": 0.2715,
+      "step": 1387
+    },
+    {
+      "epoch": 9.783259911894273,
+      "grad_norm": 0.1830518656367563,
+      "learning_rate": 0.0001857558798083108,
+      "loss": 0.2762,
+      "step": 1388
+    },
+    {
+      "epoch": 9.790308370044054,
+      "grad_norm": 0.1800085451594197,
+      "learning_rate": 0.00018573223824719614,
+      "loss": 0.2869,
+      "step": 1389
+    },
+    {
+      "epoch": 9.797356828193832,
+      "grad_norm": 0.19729122281716985,
+      "learning_rate": 0.0001857085785899515,
+      "loss": 0.2802,
+      "step": 1390
+    },
+    {
+      "epoch": 9.804405286343613,
+      "grad_norm": 0.19252904892630163,
+      "learning_rate": 0.00018568490084157094,
+      "loss": 0.2794,
+      "step": 1391
+    },
+    {
+      "epoch": 9.811453744493392,
+      "grad_norm": 0.1762277806379642,
+      "learning_rate": 0.0001856612050070522,
+      "loss": 0.2855,
+      "step": 1392
+    },
+    {
+      "epoch": 9.818502202643172,
+      "grad_norm": 0.1560085357938526,
+      "learning_rate": 0.00018563749109139704,
+      "loss": 0.2729,
+      "step": 1393
+    },
+    {
+      "epoch": 9.82555066079295,
+      "grad_norm": 0.17912347013126634,
+      "learning_rate": 0.00018561375909961086,
+      "loss": 0.2896,
+      "step": 1394
+    },
+    {
+      "epoch": 9.832599118942731,
+      "grad_norm": 0.1857692954895102,
+      "learning_rate": 0.0001855900090367029,
+      "loss": 0.2825,
+      "step": 1395
+    },
+    {
+      "epoch": 9.839647577092512,
+      "grad_norm": 0.1962152756234081,
+      "learning_rate": 0.00018556624090768637,
+      "loss": 0.2835,
+      "step": 1396
+    },
+    {
+      "epoch": 9.84669603524229,
+      "grad_norm": 0.17692535296216638,
+      "learning_rate": 0.00018554245471757815,
+      "loss": 0.28,
+      "step": 1397
+    },
+    {
+      "epoch": 9.853744493392071,
+      "grad_norm": 0.19359684066349275,
+      "learning_rate": 0.0001855186504713989,
+      "loss": 0.284,
+      "step": 1398
+    },
+    {
+      "epoch": 9.86079295154185,
+      "grad_norm": 0.1926175624756351,
+      "learning_rate": 0.00018549482817417325,
+      "loss": 0.2872,
+      "step": 1399
+    },
+    {
+      "epoch": 9.86784140969163,
+      "grad_norm": 0.20838842370754507,
+      "learning_rate": 0.00018547098783092947,
+      "loss": 0.292,
+      "step": 1400
+    },
+    {
+      "epoch": 9.874889867841409,
+      "grad_norm": 0.1813275704410424,
+      "learning_rate": 0.00018544712944669974,
+      "loss": 0.2901,
+      "step": 1401
+    },
+    {
+      "epoch": 9.88193832599119,
+      "grad_norm": 0.16835911021905758,
+      "learning_rate": 0.00018542325302652003,
+      "loss": 0.2838,
+      "step": 1402
+    },
+    {
+      "epoch": 9.88898678414097,
+      "grad_norm": 0.1643274463769745,
+      "learning_rate": 0.00018539935857543014,
+      "loss": 0.2804,
+      "step": 1403
+    },
+    {
+      "epoch": 9.896035242290749,
+      "grad_norm": 0.1763223948332538,
+      "learning_rate": 0.00018537544609847358,
+      "loss": 0.278,
+      "step": 1404
+    },
+    {
+      "epoch": 9.90308370044053,
+      "grad_norm": 0.17615079696145874,
+      "learning_rate": 0.0001853515156006978,
+      "loss": 0.2877,
+      "step": 1405
+    },
+    {
+      "epoch": 9.910132158590308,
+      "grad_norm": 0.1825082968833505,
+      "learning_rate": 0.00018532756708715392,
+      "loss": 0.284,
+      "step": 1406
+    },
+    {
+      "epoch": 9.917180616740088,
+      "grad_norm": 0.1739490546640844,
+      "learning_rate": 0.00018530360056289696,
+      "loss": 0.2892,
+      "step": 1407
+    },
+    {
+      "epoch": 9.924229074889867,
+      "grad_norm": 0.16026475278141217,
+      "learning_rate": 0.00018527961603298572,
+      "loss": 0.2794,
+      "step": 1408
+    },
+    {
+      "epoch": 9.931277533039648,
+      "grad_norm": 0.17221000339699646,
+      "learning_rate": 0.00018525561350248278,
+      "loss": 0.2648,
+      "step": 1409
+    },
+    {
+      "epoch": 9.938325991189426,
+      "grad_norm": 0.17288406421487834,
+      "learning_rate": 0.00018523159297645454,
+      "loss": 0.2833,
+      "step": 1410
+    },
+    {
+      "epoch": 9.945374449339207,
+      "grad_norm": 0.16956867418934612,
+      "learning_rate": 0.00018520755445997113,
+      "loss": 0.2749,
+      "step": 1411
+    },
+    {
+      "epoch": 9.952422907488987,
+      "grad_norm": 0.18700497313431425,
+      "learning_rate": 0.00018518349795810664,
+      "loss": 0.2842,
+      "step": 1412
+    },
+    {
+      "epoch": 9.959471365638766,
+      "grad_norm": 0.1741553319191229,
+      "learning_rate": 0.00018515942347593876,
+      "loss": 0.2788,
+      "step": 1413
+    },
+    {
+      "epoch": 9.966519823788547,
+      "grad_norm": 0.15778746322013418,
+      "learning_rate": 0.00018513533101854911,
+      "loss": 0.2846,
+      "step": 1414
+    },
+    {
+      "epoch": 9.973568281938325,
+      "grad_norm": 0.16677913012642048,
+      "learning_rate": 0.00018511122059102302,
+      "loss": 0.2813,
+      "step": 1415
+    },
+    {
+      "epoch": 9.980616740088106,
+      "grad_norm": 0.15596576433614737,
+      "learning_rate": 0.00018508709219844973,
+      "loss": 0.2678,
+      "step": 1416
+    },
+    {
+      "epoch": 9.987665198237885,
+      "grad_norm": 0.1541996809051296,
+      "learning_rate": 0.00018506294584592214,
+      "loss": 0.2732,
+      "step": 1417
+    },
+    {
+      "epoch": 9.994713656387665,
+      "grad_norm": 0.16235672510172813,
+      "learning_rate": 0.000185038781538537,
+      "loss": 0.2784,
+      "step": 1418
+    },
+    {
+      "epoch": 10.001762114537446,
+      "grad_norm": 0.1572440756241025,
+      "learning_rate": 0.00018501459928139483,
+      "loss": 0.2609,
+      "step": 1419
+    },
+    {
+      "epoch": 10.008810572687224,
+      "grad_norm": 0.1605193567746416,
+      "learning_rate": 0.00018499039907960002,
+      "loss": 0.2481,
+      "step": 1420
+    },
+    {
+      "epoch": 10.015859030837005,
+      "grad_norm": 0.17736479002242192,
+      "learning_rate": 0.00018496618093826063,
+      "loss": 0.2429,
+      "step": 1421
+    },
+    {
+      "epoch": 10.022907488986784,
+      "grad_norm": 0.23460814001902044,
+      "learning_rate": 0.0001849419448624886,
+      "loss": 0.2511,
+      "step": 1422
+    },
+    {
+      "epoch": 10.029955947136564,
+      "grad_norm": 0.19782485049002768,
+      "learning_rate": 0.00018491769085739957,
+      "loss": 0.2475,
+      "step": 1423
+    },
+    {
+      "epoch": 10.037004405286343,
+      "grad_norm": 0.18280619388011285,
+      "learning_rate": 0.00018489341892811303,
+      "loss": 0.2442,
+      "step": 1424
+    },
+    {
+      "epoch": 10.044052863436123,
+      "grad_norm": 0.19055378808251022,
+      "learning_rate": 0.00018486912907975223,
+      "loss": 0.2505,
+      "step": 1425
+    },
+    {
+      "epoch": 10.051101321585904,
+      "grad_norm": 0.1807366863268995,
+      "learning_rate": 0.00018484482131744418,
+      "loss": 0.2358,
+      "step": 1426
+    },
+    {
+      "epoch": 10.058149779735682,
+      "grad_norm": 0.22153167694261638,
+      "learning_rate": 0.00018482049564631976,
+      "loss": 0.2517,
+      "step": 1427
+    },
+    {
+      "epoch": 10.065198237885463,
+      "grad_norm": 0.20101816614620852,
+      "learning_rate": 0.00018479615207151354,
+      "loss": 0.2368,
+      "step": 1428
+    },
+    {
+      "epoch": 10.072246696035242,
+      "grad_norm": 0.20384298500283968,
+      "learning_rate": 0.00018477179059816386,
+      "loss": 0.2465,
+      "step": 1429
+    },
+    {
+      "epoch": 10.079295154185022,
+      "grad_norm": 0.18253882854325276,
+      "learning_rate": 0.00018474741123141293,
+      "loss": 0.241,
+      "step": 1430
+    },
+    {
+      "epoch": 10.086343612334801,
+      "grad_norm": 0.1991412387844603,
+      "learning_rate": 0.00018472301397640665,
+      "loss": 0.2446,
+      "step": 1431
+    },
+    {
+      "epoch": 10.093392070484581,
+      "grad_norm": 0.2114993065003741,
+      "learning_rate": 0.00018469859883829472,
+      "loss": 0.2505,
+      "step": 1432
+    },
+    {
+      "epoch": 10.100440528634362,
+      "grad_norm": 0.22970860094508844,
+      "learning_rate": 0.00018467416582223066,
+      "loss": 0.2551,
+      "step": 1433
+    },
+    {
+      "epoch": 10.10748898678414,
+      "grad_norm": 0.19862950373708987,
+      "learning_rate": 0.00018464971493337167,
+      "loss": 0.2513,
+      "step": 1434
+    },
+    {
+      "epoch": 10.114537444933921,
+      "grad_norm": 0.1965349396301641,
+      "learning_rate": 0.00018462524617687882,
+      "loss": 0.2581,
+      "step": 1435
+    },
+    {
+      "epoch": 10.1215859030837,
+      "grad_norm": 0.18358470208474437,
+      "learning_rate": 0.0001846007595579169,
+      "loss": 0.2497,
+      "step": 1436
+    },
+    {
+      "epoch": 10.12863436123348,
+      "grad_norm": 0.19784172266508132,
+      "learning_rate": 0.00018457625508165446,
+      "loss": 0.2559,
+      "step": 1437
+    },
+    {
+      "epoch": 10.135682819383259,
+      "grad_norm": 0.21720445444100125,
+      "learning_rate": 0.00018455173275326386,
+      "loss": 0.2521,
+      "step": 1438
+    },
+    {
+      "epoch": 10.14273127753304,
+      "grad_norm": 0.20055970195046452,
+      "learning_rate": 0.0001845271925779212,
+      "loss": 0.2535,
+      "step": 1439
+    },
+    {
+      "epoch": 10.14977973568282,
+      "grad_norm": 0.1944032895628798,
+      "learning_rate": 0.00018450263456080637,
+      "loss": 0.2411,
+      "step": 1440
+    },
+    {
+      "epoch": 10.156828193832599,
+      "grad_norm": 0.21559245343878278,
+      "learning_rate": 0.00018447805870710297,
+      "loss": 0.2589,
+      "step": 1441
+    },
+    {
+      "epoch": 10.16387665198238,
+      "grad_norm": 0.20978330724342123,
+      "learning_rate": 0.00018445346502199845,
+      "loss": 0.2554,
+      "step": 1442
+    },
+    {
+      "epoch": 10.170925110132158,
+      "grad_norm": 0.21462424722990742,
+      "learning_rate": 0.00018442885351068394,
+      "loss": 0.2562,
+      "step": 1443
+    },
+    {
+      "epoch": 10.177973568281939,
+      "grad_norm": 0.19648965171296096,
+      "learning_rate": 0.00018440422417835445,
+      "loss": 0.2615,
+      "step": 1444
+    },
+    {
+      "epoch": 10.185022026431717,
+      "grad_norm": 0.18754243563797052,
+      "learning_rate": 0.00018437957703020853,
+      "loss": 0.25,
+      "step": 1445
+    },
+    {
+      "epoch": 10.192070484581498,
+      "grad_norm": 0.1929744177693042,
+      "learning_rate": 0.00018435491207144877,
+      "loss": 0.2582,
+      "step": 1446
+    },
+    {
+      "epoch": 10.199118942731278,
+      "grad_norm": 0.1791616672580515,
+      "learning_rate": 0.00018433022930728133,
+      "loss": 0.2492,
+      "step": 1447
+    },
+    {
+      "epoch": 10.206167400881057,
+      "grad_norm": 0.19927795295535738,
+      "learning_rate": 0.00018430552874291615,
+      "loss": 0.2461,
+      "step": 1448
+    },
+    {
+      "epoch": 10.213215859030838,
+      "grad_norm": 0.19462749511538677,
+      "learning_rate": 0.00018428081038356702,
+      "loss": 0.2616,
+      "step": 1449
+    },
+    {
+      "epoch": 10.220264317180616,
+      "grad_norm": 0.1955659305318928,
+      "learning_rate": 0.00018425607423445137,
+      "loss": 0.2528,
+      "step": 1450
+    },
+    {
+      "epoch": 10.227312775330397,
+      "grad_norm": 0.19380847408059285,
+      "learning_rate": 0.00018423132030079048,
+      "loss": 0.2512,
+      "step": 1451
+    },
+    {
+      "epoch": 10.234361233480175,
+      "grad_norm": 0.21073134241701627,
+      "learning_rate": 0.0001842065485878093,
+      "loss": 0.2595,
+      "step": 1452
+    },
+    {
+      "epoch": 10.241409691629956,
+      "grad_norm": 0.1914103939723457,
+      "learning_rate": 0.00018418175910073667,
+      "loss": 0.2506,
+      "step": 1453
+    },
+    {
+      "epoch": 10.248458149779736,
+      "grad_norm": 0.1881863750828962,
+      "learning_rate": 0.000184156951844805,
+      "loss": 0.2539,
+      "step": 1454
+    },
+    {
+      "epoch": 10.255506607929515,
+      "grad_norm": 0.1947614579676636,
+      "learning_rate": 0.00018413212682525054,
+      "loss": 0.2654,
+      "step": 1455
+    },
+    {
+      "epoch": 10.262555066079296,
+      "grad_norm": 0.1912722648466952,
+      "learning_rate": 0.00018410728404731332,
+      "loss": 0.2611,
+      "step": 1456
+    },
+    {
+      "epoch": 10.269603524229074,
+      "grad_norm": 0.19631458089378984,
+      "learning_rate": 0.00018408242351623707,
+      "loss": 0.2616,
+      "step": 1457
+    },
+    {
+      "epoch": 10.276651982378855,
+      "grad_norm": 0.19011269288168006,
+      "learning_rate": 0.0001840575452372693,
+      "loss": 0.2538,
+      "step": 1458
+    },
+    {
+      "epoch": 10.283700440528634,
+      "grad_norm": 0.2009508147941738,
+      "learning_rate": 0.00018403264921566125,
+      "loss": 0.2589,
+      "step": 1459
+    },
+    {
+      "epoch": 10.290748898678414,
+      "grad_norm": 0.19613812872403352,
+      "learning_rate": 0.00018400773545666787,
+      "loss": 0.2551,
+      "step": 1460
+    },
+    {
+      "epoch": 10.297797356828195,
+      "grad_norm": 0.19503092282788523,
+      "learning_rate": 0.00018398280396554793,
+      "loss": 0.2536,
+      "step": 1461
+    },
+    {
+      "epoch": 10.304845814977973,
+      "grad_norm": 0.20106158824740514,
+      "learning_rate": 0.0001839578547475639,
+      "loss": 0.2549,
+      "step": 1462
+    },
+    {
+      "epoch": 10.311894273127754,
+      "grad_norm": 0.21161537382286397,
+      "learning_rate": 0.00018393288780798196,
+      "loss": 0.2561,
+      "step": 1463
+    },
+    {
+      "epoch": 10.318942731277533,
+      "grad_norm": 0.19704947562586247,
+      "learning_rate": 0.00018390790315207212,
+      "loss": 0.2544,
+      "step": 1464
+    },
+    {
+      "epoch": 10.325991189427313,
+      "grad_norm": 0.19010669543596057,
+      "learning_rate": 0.000183882900785108,
+      "loss": 0.264,
+      "step": 1465
+    },
+    {
+      "epoch": 10.333039647577092,
+      "grad_norm": 0.18887457868199442,
+      "learning_rate": 0.0001838578807123671,
+      "loss": 0.2519,
+      "step": 1466
+    },
+    {
+      "epoch": 10.340088105726872,
+      "grad_norm": 0.18474415248039663,
+      "learning_rate": 0.00018383284293913054,
+      "loss": 0.2556,
+      "step": 1467
+    },
+    {
+      "epoch": 10.347136563876653,
+      "grad_norm": 0.20285725950321193,
+      "learning_rate": 0.00018380778747068325,
+      "loss": 0.2577,
+      "step": 1468
+    },
+    {
+      "epoch": 10.354185022026432,
+      "grad_norm": 0.19512351073745562,
+      "learning_rate": 0.0001837827143123139,
+      "loss": 0.254,
+      "step": 1469
+    },
+    {
+      "epoch": 10.361233480176212,
+      "grad_norm": 0.21475633905605976,
+      "learning_rate": 0.0001837576234693148,
+      "loss": 0.2672,
+      "step": 1470
+    },
+    {
+      "epoch": 10.36828193832599,
+      "grad_norm": 0.19388117546121214,
+      "learning_rate": 0.0001837325149469821,
+      "loss": 0.2614,
+      "step": 1471
+    },
+    {
+      "epoch": 10.375330396475771,
+      "grad_norm": 0.1798154447540228,
+      "learning_rate": 0.00018370738875061565,
+      "loss": 0.2648,
+      "step": 1472
+    },
+    {
+      "epoch": 10.38237885462555,
+      "grad_norm": 0.19846148654096077,
+      "learning_rate": 0.00018368224488551896,
+      "loss": 0.2673,
+      "step": 1473
+    },
+    {
+      "epoch": 10.38942731277533,
+      "grad_norm": 0.1958248262391291,
+      "learning_rate": 0.00018365708335699938,
+      "loss": 0.2546,
+      "step": 1474
+    },
+    {
+      "epoch": 10.396475770925111,
+      "grad_norm": 0.1995149904319366,
+      "learning_rate": 0.00018363190417036793,
+      "loss": 0.2563,
+      "step": 1475
+    },
+    {
+      "epoch": 10.40352422907489,
+      "grad_norm": 0.21028702570528518,
+      "learning_rate": 0.00018360670733093937,
+      "loss": 0.2547,
+      "step": 1476
+    },
+    {
+      "epoch": 10.41057268722467,
+      "grad_norm": 0.21558726900557598,
+      "learning_rate": 0.00018358149284403218,
+      "loss": 0.2647,
+      "step": 1477
+    },
+    {
+      "epoch": 10.417621145374449,
+      "grad_norm": 0.22174969286720808,
+      "learning_rate": 0.00018355626071496852,
+      "loss": 0.2582,
+      "step": 1478
+    },
+    {
+      "epoch": 10.42466960352423,
+      "grad_norm": 0.19140324299950284,
+      "learning_rate": 0.0001835310109490744,
+      "loss": 0.2596,
+      "step": 1479
+    },
+    {
+      "epoch": 10.431718061674008,
+      "grad_norm": 0.20903043389623258,
+      "learning_rate": 0.00018350574355167941,
+      "loss": 0.2708,
+      "step": 1480
+    },
+    {
+      "epoch": 10.438766519823789,
+      "grad_norm": 0.20404835099428364,
+      "learning_rate": 0.00018348045852811695,
+      "loss": 0.2536,
+      "step": 1481
+    },
+    {
+      "epoch": 10.44581497797357,
+      "grad_norm": 0.19801299864271982,
+      "learning_rate": 0.0001834551558837241,
+      "loss": 0.257,
+      "step": 1482
+    },
+    {
+      "epoch": 10.452863436123348,
+      "grad_norm": 0.20253870005901578,
+      "learning_rate": 0.0001834298356238417,
+      "loss": 0.2697,
+      "step": 1483
+    },
+    {
+      "epoch": 10.459911894273128,
+      "grad_norm": 0.1673495865353548,
+      "learning_rate": 0.00018340449775381426,
+      "loss": 0.2556,
+      "step": 1484
+    },
+    {
+      "epoch": 10.466960352422907,
+      "grad_norm": 0.18572428099074936,
+      "learning_rate": 0.00018337914227899003,
+      "loss": 0.2525,
+      "step": 1485
+    },
+    {
+      "epoch": 10.474008810572688,
+      "grad_norm": 0.19094898598195814,
+      "learning_rate": 0.00018335376920472097,
+      "loss": 0.2636,
+      "step": 1486
+    },
+    {
+      "epoch": 10.481057268722466,
+      "grad_norm": 0.1896091598650693,
+      "learning_rate": 0.0001833283785363628,
+      "loss": 0.2625,
+      "step": 1487
+    },
+    {
+      "epoch": 10.488105726872247,
+      "grad_norm": 0.19149971235556065,
+      "learning_rate": 0.00018330297027927488,
+      "loss": 0.2622,
+      "step": 1488
+    },
+    {
+      "epoch": 10.495154185022026,
+      "grad_norm": 0.18954341875121347,
+      "learning_rate": 0.0001832775444388203,
+      "loss": 0.2539,
+      "step": 1489
+    },
+    {
+      "epoch": 10.502202643171806,
+      "grad_norm": 0.1812071680319654,
+      "learning_rate": 0.00018325210102036591,
+      "loss": 0.2529,
+      "step": 1490
+    },
+    {
+      "epoch": 10.509251101321587,
+      "grad_norm": 0.19735127721030615,
+      "learning_rate": 0.00018322664002928221,
+      "loss": 0.2513,
+      "step": 1491
+    },
+    {
+      "epoch": 10.516299559471365,
+      "grad_norm": 0.19141989096630277,
+      "learning_rate": 0.0001832011614709435,
+      "loss": 0.2637,
+      "step": 1492
+    },
+    {
+      "epoch": 10.523348017621146,
+      "grad_norm": 0.20251320998979844,
+      "learning_rate": 0.00018317566535072766,
+      "loss": 0.2508,
+      "step": 1493
+    },
+    {
+      "epoch": 10.530396475770925,
+      "grad_norm": 0.18215079827061223,
+      "learning_rate": 0.00018315015167401634,
+      "loss": 0.246,
+      "step": 1494
+    },
+    {
+      "epoch": 10.537444933920705,
+      "grad_norm": 0.1915137272483673,
+      "learning_rate": 0.00018312462044619497,
+      "loss": 0.2492,
+      "step": 1495
+    },
+    {
+      "epoch": 10.544493392070484,
+      "grad_norm": 0.18538872666075554,
+      "learning_rate": 0.00018309907167265253,
+      "loss": 0.2539,
+      "step": 1496
+    },
+    {
+      "epoch": 10.551541850220264,
+      "grad_norm": 0.21054700109274507,
+      "learning_rate": 0.0001830735053587818,
+      "loss": 0.2711,
+      "step": 1497
+    },
+    {
+      "epoch": 10.558590308370045,
+      "grad_norm": 0.2111602682809098,
+      "learning_rate": 0.0001830479215099793,
+      "loss": 0.2625,
+      "step": 1498
+    },
+    {
+      "epoch": 10.565638766519823,
+      "grad_norm": 0.19340621520018822,
+      "learning_rate": 0.00018302232013164518,
+      "loss": 0.2616,
+      "step": 1499
+    },
+    {
+      "epoch": 10.572687224669604,
+      "grad_norm": 0.18555551615127025,
+      "learning_rate": 0.00018299670122918328,
+      "loss": 0.2605,
+      "step": 1500
+    },
+    {
+      "epoch": 10.579735682819383,
+      "grad_norm": 0.19746895588214858,
+      "learning_rate": 0.00018297106480800118,
+      "loss": 0.2658,
+      "step": 1501
+    },
+    {
+      "epoch": 10.586784140969163,
+      "grad_norm": 0.23087709401047432,
+      "learning_rate": 0.00018294541087351016,
+      "loss": 0.2634,
+      "step": 1502
+    },
+    {
+      "epoch": 10.593832599118942,
+      "grad_norm": 0.19969101912822104,
+      "learning_rate": 0.00018291973943112516,
+      "loss": 0.2556,
+      "step": 1503
+    },
+    {
+      "epoch": 10.600881057268722,
+      "grad_norm": 0.19770158084223027,
+      "learning_rate": 0.0001828940504862649,
+      "loss": 0.2561,
+      "step": 1504
+    },
+    {
+      "epoch": 10.607929515418503,
+      "grad_norm": 0.20384270473612504,
+      "learning_rate": 0.00018286834404435167,
+      "loss": 0.2666,
+      "step": 1505
+    },
+    {
+      "epoch": 10.614977973568282,
+      "grad_norm": 0.1991116046509892,
+      "learning_rate": 0.00018284262011081153,
+      "loss": 0.2721,
+      "step": 1506
+    },
+    {
+      "epoch": 10.622026431718062,
+      "grad_norm": 0.2106445599058913,
+      "learning_rate": 0.00018281687869107423,
+      "loss": 0.2632,
+      "step": 1507
+    },
+    {
+      "epoch": 10.62907488986784,
+      "grad_norm": 0.21512317826159352,
+      "learning_rate": 0.00018279111979057318,
+      "loss": 0.2584,
+      "step": 1508
+    },
+    {
+      "epoch": 10.636123348017621,
+      "grad_norm": 0.20575567567673986,
+      "learning_rate": 0.00018276534341474552,
+      "loss": 0.2623,
+      "step": 1509
+    },
+    {
+      "epoch": 10.6431718061674,
+      "grad_norm": 0.1892565754625666,
+      "learning_rate": 0.00018273954956903206,
+      "loss": 0.2561,
+      "step": 1510
+    },
+    {
+      "epoch": 10.65022026431718,
+      "grad_norm": 0.1952218302360461,
+      "learning_rate": 0.00018271373825887727,
+      "loss": 0.2563,
+      "step": 1511
+    },
+    {
+      "epoch": 10.657268722466961,
+      "grad_norm": 0.18845313964260738,
+      "learning_rate": 0.0001826879094897294,
+      "loss": 0.2632,
+      "step": 1512
+    },
+    {
+      "epoch": 10.66431718061674,
+      "grad_norm": 0.18292755284652956,
+      "learning_rate": 0.0001826620632670402,
+      "loss": 0.2602,
+      "step": 1513
+    },
+    {
+      "epoch": 10.67136563876652,
+      "grad_norm": 0.19125769298436326,
+      "learning_rate": 0.00018263619959626531,
+      "loss": 0.2614,
+      "step": 1514
+    },
+    {
+      "epoch": 10.678414096916299,
+      "grad_norm": 0.20176237800581837,
+      "learning_rate": 0.00018261031848286398,
+      "loss": 0.2665,
+      "step": 1515
+    },
+    {
+      "epoch": 10.68546255506608,
+      "grad_norm": 0.20042484138222413,
+      "learning_rate": 0.00018258441993229907,
+      "loss": 0.2634,
+      "step": 1516
+    },
+    {
+      "epoch": 10.692511013215858,
+      "grad_norm": 0.18194671032966003,
+      "learning_rate": 0.00018255850395003717,
+      "loss": 0.2654,
+      "step": 1517
+    },
+    {
+      "epoch": 10.699559471365639,
+      "grad_norm": 0.21235055553306892,
+      "learning_rate": 0.0001825325705415486,
+      "loss": 0.2663,
+      "step": 1518
+    },
+    {
+      "epoch": 10.70660792951542,
+      "grad_norm": 0.2005253888910637,
+      "learning_rate": 0.00018250661971230727,
+      "loss": 0.2729,
+      "step": 1519
+    },
+    {
+      "epoch": 10.713656387665198,
+      "grad_norm": 0.19916218154573778,
+      "learning_rate": 0.00018248065146779085,
+      "loss": 0.2662,
+      "step": 1520
+    },
+    {
+      "epoch": 10.720704845814979,
+      "grad_norm": 0.17447019462742103,
+      "learning_rate": 0.00018245466581348066,
+      "loss": 0.269,
+      "step": 1521
+    },
+    {
+      "epoch": 10.727753303964757,
+      "grad_norm": 0.18936897479447884,
+      "learning_rate": 0.00018242866275486162,
+      "loss": 0.2651,
+      "step": 1522
+    },
+    {
+      "epoch": 10.734801762114538,
+      "grad_norm": 0.19335323591575762,
+      "learning_rate": 0.0001824026422974224,
+      "loss": 0.2608,
+      "step": 1523
+    },
+    {
+      "epoch": 10.741850220264316,
+      "grad_norm": 0.21207143854777566,
+      "learning_rate": 0.0001823766044466554,
+      "loss": 0.2731,
+      "step": 1524
+    },
+    {
+      "epoch": 10.748898678414097,
+      "grad_norm": 0.18373016085063376,
+      "learning_rate": 0.00018235054920805652,
+      "loss": 0.2544,
+      "step": 1525
+    },
+    {
+      "epoch": 10.755947136563877,
+      "grad_norm": 0.19242156629004245,
+      "learning_rate": 0.0001823244765871255,
+      "loss": 0.2639,
+      "step": 1526
+    },
+    {
+      "epoch": 10.762995594713656,
+      "grad_norm": 0.1779994185165384,
+      "learning_rate": 0.00018229838658936564,
+      "loss": 0.2628,
+      "step": 1527
+    },
+    {
+      "epoch": 10.770044052863437,
+      "grad_norm": 0.17796504826003035,
+      "learning_rate": 0.00018227227922028397,
+      "loss": 0.2615,
+      "step": 1528
+    },
+    {
+      "epoch": 10.777092511013215,
+      "grad_norm": 0.19505614286571973,
+      "learning_rate": 0.00018224615448539114,
+      "loss": 0.2625,
+      "step": 1529
+    },
+    {
+      "epoch": 10.784140969162996,
+      "grad_norm": 0.19113792630329404,
+      "learning_rate": 0.00018222001239020155,
+      "loss": 0.2768,
+      "step": 1530
+    },
+    {
+      "epoch": 10.791189427312775,
+      "grad_norm": 0.16976282177609114,
+      "learning_rate": 0.00018219385294023313,
+      "loss": 0.2528,
+      "step": 1531
+    },
+    {
+      "epoch": 10.798237885462555,
+      "grad_norm": 0.17184565844787153,
+      "learning_rate": 0.00018216767614100756,
+      "loss": 0.263,
+      "step": 1532
+    },
+    {
+      "epoch": 10.805286343612334,
+      "grad_norm": 0.1781897250170396,
+      "learning_rate": 0.0001821414819980502,
+      "loss": 0.2677,
+      "step": 1533
+    },
+    {
+      "epoch": 10.812334801762114,
+      "grad_norm": 0.18941994346347468,
+      "learning_rate": 0.00018211527051689,
+      "loss": 0.2662,
+      "step": 1534
+    },
+    {
+      "epoch": 10.819383259911895,
+      "grad_norm": 0.193113756853752,
+      "learning_rate": 0.00018208904170305962,
+      "loss": 0.2593,
+      "step": 1535
+    },
+    {
+      "epoch": 10.826431718061674,
+      "grad_norm": 0.19305938219986302,
+      "learning_rate": 0.0001820627955620954,
+      "loss": 0.2753,
+      "step": 1536
+    },
+    {
+      "epoch": 10.833480176211454,
+      "grad_norm": 0.1746428922039724,
+      "learning_rate": 0.00018203653209953729,
+      "loss": 0.2588,
+      "step": 1537
+    },
+    {
+      "epoch": 10.840528634361233,
+      "grad_norm": 0.18382465746985302,
+      "learning_rate": 0.00018201025132092889,
+      "loss": 0.2804,
+      "step": 1538
+    },
+    {
+      "epoch": 10.847577092511013,
+      "grad_norm": 0.18908275424311943,
+      "learning_rate": 0.00018198395323181746,
+      "loss": 0.2625,
+      "step": 1539
+    },
+    {
+      "epoch": 10.854625550660792,
+      "grad_norm": 0.1817131945477787,
+      "learning_rate": 0.000181957637837754,
+      "loss": 0.2637,
+      "step": 1540
+    },
+    {
+      "epoch": 10.861674008810573,
+      "grad_norm": 0.2051893650371467,
+      "learning_rate": 0.00018193130514429303,
+      "loss": 0.2586,
+      "step": 1541
+    },
+    {
+      "epoch": 10.868722466960353,
+      "grad_norm": 0.19568111871979957,
+      "learning_rate": 0.00018190495515699282,
+      "loss": 0.2695,
+      "step": 1542
+    },
+    {
+      "epoch": 10.875770925110132,
+      "grad_norm": 0.1997610458008817,
+      "learning_rate": 0.00018187858788141522,
+      "loss": 0.2661,
+      "step": 1543
+    },
+    {
+      "epoch": 10.882819383259912,
+      "grad_norm": 0.1855445610876759,
+      "learning_rate": 0.00018185220332312578,
+      "loss": 0.2631,
+      "step": 1544
+    },
+    {
+      "epoch": 10.889867841409691,
+      "grad_norm": 0.18013174896787426,
+      "learning_rate": 0.0001818258014876937,
+      "loss": 0.2628,
+      "step": 1545
+    },
+    {
+      "epoch": 10.896916299559471,
+      "grad_norm": 0.19366648444029178,
+      "learning_rate": 0.0001817993823806918,
+      "loss": 0.2625,
+      "step": 1546
+    },
+    {
+      "epoch": 10.90396475770925,
+      "grad_norm": 0.19737976451466804,
+      "learning_rate": 0.0001817729460076965,
+      "loss": 0.2634,
+      "step": 1547
+    },
+    {
+      "epoch": 10.91101321585903,
+      "grad_norm": 0.2066817206528808,
+      "learning_rate": 0.00018174649237428798,
+      "loss": 0.2652,
+      "step": 1548
+    },
+    {
+      "epoch": 10.918061674008811,
+      "grad_norm": 0.18912946279380172,
+      "learning_rate": 0.00018172002148605004,
+      "loss": 0.2608,
+      "step": 1549
+    },
+    {
+      "epoch": 10.92511013215859,
+      "grad_norm": 0.202387754005307,
+      "learning_rate": 0.00018169353334856995,
+      "loss": 0.263,
+      "step": 1550
+    },
+    {
+      "epoch": 10.93215859030837,
+      "grad_norm": 0.18974775846564637,
+      "learning_rate": 0.00018166702796743888,
+      "loss": 0.2662,
+      "step": 1551
+    },
+    {
+      "epoch": 10.93920704845815,
+      "grad_norm": 0.16973802667646512,
+      "learning_rate": 0.00018164050534825145,
+      "loss": 0.2463,
+      "step": 1552
+    },
+    {
+      "epoch": 10.94625550660793,
+      "grad_norm": 0.19239790353658795,
+      "learning_rate": 0.00018161396549660604,
+      "loss": 0.2609,
+      "step": 1553
+    },
+    {
+      "epoch": 10.953303964757708,
+      "grad_norm": 0.19192320826644269,
+      "learning_rate": 0.00018158740841810455,
+      "loss": 0.2644,
+      "step": 1554
+    },
+    {
+      "epoch": 10.960352422907489,
+      "grad_norm": 0.19301123100438028,
+      "learning_rate": 0.0001815608341183526,
+      "loss": 0.2621,
+      "step": 1555
+    },
+    {
+      "epoch": 10.96740088105727,
+      "grad_norm": 0.184225353026429,
+      "learning_rate": 0.0001815342426029594,
+      "loss": 0.2677,
+      "step": 1556
+    },
+    {
+      "epoch": 10.974449339207048,
+      "grad_norm": 0.1870332523556953,
+      "learning_rate": 0.00018150763387753786,
+      "loss": 0.2686,
+      "step": 1557
+    },
+    {
+      "epoch": 10.981497797356829,
+      "grad_norm": 0.1669517633981154,
+      "learning_rate": 0.00018148100794770447,
+      "loss": 0.2519,
+      "step": 1558
+    },
+    {
+      "epoch": 10.988546255506607,
+      "grad_norm": 0.17849320818061842,
+      "learning_rate": 0.00018145436481907933,
+      "loss": 0.2632,
+      "step": 1559
+    },
+    {
+      "epoch": 10.995594713656388,
+      "grad_norm": 0.18889075137673184,
+      "learning_rate": 0.00018142770449728617,
+      "loss": 0.2736,
+      "step": 1560
+    },
+    {
+      "epoch": 11.002643171806167,
+      "grad_norm": 0.1874833027139948,
+      "learning_rate": 0.00018140102698795246,
+      "loss": 0.2425,
+      "step": 1561
+    },
+    {
+      "epoch": 11.009691629955947,
+      "grad_norm": 0.18829291126154352,
+      "learning_rate": 0.00018137433229670918,
+      "loss": 0.2334,
+      "step": 1562
+    },
+    {
+      "epoch": 11.016740088105728,
+      "grad_norm": 0.24824094302232944,
+      "learning_rate": 0.00018134762042919098,
+      "loss": 0.225,
+      "step": 1563
+    },
+    {
+      "epoch": 11.023788546255506,
+      "grad_norm": 0.26839063374519684,
+      "learning_rate": 0.00018132089139103613,
+      "loss": 0.2241,
+      "step": 1564
+    },
+    {
+      "epoch": 11.030837004405287,
+      "grad_norm": 0.21991701497872873,
+      "learning_rate": 0.0001812941451878865,
+      "loss": 0.2233,
+      "step": 1565
+    },
+    {
+      "epoch": 11.037885462555066,
+      "grad_norm": 0.22351264822384895,
+      "learning_rate": 0.00018126738182538762,
+      "loss": 0.2256,
+      "step": 1566
+    },
+    {
+      "epoch": 11.044933920704846,
+      "grad_norm": 0.20509961248185313,
+      "learning_rate": 0.0001812406013091886,
+      "loss": 0.2157,
+      "step": 1567
+    },
+    {
+      "epoch": 11.051982378854625,
+      "grad_norm": 0.225617212498463,
+      "learning_rate": 0.00018121380364494226,
+      "loss": 0.2214,
+      "step": 1568
+    },
+    {
+      "epoch": 11.059030837004405,
+      "grad_norm": 0.27365139427213625,
+      "learning_rate": 0.00018118698883830495,
+      "loss": 0.2309,
+      "step": 1569
+    },
+    {
+      "epoch": 11.066079295154186,
+      "grad_norm": 0.23543244294309038,
+      "learning_rate": 0.0001811601568949367,
+      "loss": 0.2214,
+      "step": 1570
+    },
+    {
+      "epoch": 11.073127753303964,
+      "grad_norm": 0.21833416474816053,
+      "learning_rate": 0.00018113330782050104,
+      "loss": 0.2233,
+      "step": 1571
+    },
+    {
+      "epoch": 11.080176211453745,
+      "grad_norm": 0.21519976339093624,
+      "learning_rate": 0.00018110644162066527,
+      "loss": 0.2225,
+      "step": 1572
+    },
+    {
+      "epoch": 11.087224669603524,
+      "grad_norm": 0.22944235804062751,
+      "learning_rate": 0.0001810795583011002,
+      "loss": 0.2296,
+      "step": 1573
+    },
+    {
+      "epoch": 11.094273127753304,
+      "grad_norm": 0.2155932399141418,
+      "learning_rate": 0.00018105265786748031,
+      "loss": 0.227,
+      "step": 1574
+    },
+    {
+      "epoch": 11.101321585903083,
+      "grad_norm": 0.2266491955999857,
+      "learning_rate": 0.0001810257403254837,
+      "loss": 0.2281,
+      "step": 1575
+    },
+    {
+      "epoch": 11.108370044052863,
+      "grad_norm": 0.21940115746821542,
+      "learning_rate": 0.000180998805680792,
+      "loss": 0.2301,
+      "step": 1576
+    },
+    {
+      "epoch": 11.115418502202644,
+      "grad_norm": 0.23955036867366822,
+      "learning_rate": 0.00018097185393909049,
+      "loss": 0.227,
+      "step": 1577
+    },
+    {
+      "epoch": 11.122466960352423,
+      "grad_norm": 0.23470836708109272,
+      "learning_rate": 0.00018094488510606813,
+      "loss": 0.2315,
+      "step": 1578
+    },
+    {
+      "epoch": 11.129515418502203,
+      "grad_norm": 0.2278474695470004,
+      "learning_rate": 0.00018091789918741737,
+      "loss": 0.2248,
+      "step": 1579
+    },
+    {
+      "epoch": 11.136563876651982,
+      "grad_norm": 0.20395435096130718,
+      "learning_rate": 0.0001808908961888344,
+      "loss": 0.2233,
+      "step": 1580
+    },
+    {
+      "epoch": 11.143612334801762,
+      "grad_norm": 0.2104404880658542,
+      "learning_rate": 0.00018086387611601884,
+      "loss": 0.2208,
+      "step": 1581
+    },
+    {
+      "epoch": 11.150660792951541,
+      "grad_norm": 0.20998679891928754,
+      "learning_rate": 0.0001808368389746741,
+      "loss": 0.2141,
+      "step": 1582
+    },
+    {
+      "epoch": 11.157709251101322,
+      "grad_norm": 0.19599534229551088,
+      "learning_rate": 0.00018080978477050708,
+      "loss": 0.2191,
+      "step": 1583
+    },
+    {
+      "epoch": 11.164757709251102,
+      "grad_norm": 0.22625804154150436,
+      "learning_rate": 0.0001807827135092283,
+      "loss": 0.2288,
+      "step": 1584
+    },
+    {
+      "epoch": 11.17180616740088,
+      "grad_norm": 0.2150307001179582,
+      "learning_rate": 0.0001807556251965519,
+      "loss": 0.237,
+      "step": 1585
+    },
+    {
+      "epoch": 11.178854625550661,
+      "grad_norm": 0.22971456195270062,
+      "learning_rate": 0.0001807285198381956,
+      "loss": 0.236,
+      "step": 1586
+    },
+    {
+      "epoch": 11.18590308370044,
+      "grad_norm": 0.22833305204101098,
+      "learning_rate": 0.0001807013974398807,
+      "loss": 0.2293,
+      "step": 1587
+    },
+    {
+      "epoch": 11.19295154185022,
+      "grad_norm": 0.2137118182468764,
+      "learning_rate": 0.00018067425800733219,
+      "loss": 0.2199,
+      "step": 1588
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 0.2213406840385332,
+      "learning_rate": 0.00018064710154627853,
+      "loss": 0.2279,
+      "step": 1589
+    },
+    {
+      "epoch": 11.20704845814978,
+      "grad_norm": 0.21510641251634877,
+      "learning_rate": 0.00018061992806245184,
+      "loss": 0.2252,
+      "step": 1590
+    },
+    {
+      "epoch": 11.21409691629956,
+      "grad_norm": 0.22354367337657075,
+      "learning_rate": 0.00018059273756158785,
+      "loss": 0.231,
+      "step": 1591
+    },
+    {
+      "epoch": 11.221145374449339,
+      "grad_norm": 0.21520199687806443,
+      "learning_rate": 0.00018056553004942586,
+      "loss": 0.2308,
+      "step": 1592
+    },
+    {
+      "epoch": 11.22819383259912,
+      "grad_norm": 0.2187102371443824,
+      "learning_rate": 0.0001805383055317087,
+      "loss": 0.2385,
+      "step": 1593
+    },
+    {
+      "epoch": 11.235242290748898,
+      "grad_norm": 0.22418589932270347,
+      "learning_rate": 0.00018051106401418293,
+      "loss": 0.2303,
+      "step": 1594
+    },
+    {
+      "epoch": 11.242290748898679,
+      "grad_norm": 0.20659548603323102,
+      "learning_rate": 0.00018048380550259857,
+      "loss": 0.2371,
+      "step": 1595
+    },
+    {
+      "epoch": 11.249339207048457,
+      "grad_norm": 0.22003497453545326,
+      "learning_rate": 0.0001804565300027093,
+      "loss": 0.2349,
+      "step": 1596
+    },
+    {
+      "epoch": 11.256387665198238,
+      "grad_norm": 0.20594098444350606,
+      "learning_rate": 0.0001804292375202723,
+      "loss": 0.2253,
+      "step": 1597
+    },
+    {
+      "epoch": 11.263436123348018,
+      "grad_norm": 0.2256052501949267,
+      "learning_rate": 0.00018040192806104846,
+      "loss": 0.2254,
+      "step": 1598
+    },
+    {
+      "epoch": 11.270484581497797,
+      "grad_norm": 0.22282754783658837,
+      "learning_rate": 0.00018037460163080216,
+      "loss": 0.2368,
+      "step": 1599
+    },
+    {
+      "epoch": 11.277533039647578,
+      "grad_norm": 0.21194743123575877,
+      "learning_rate": 0.00018034725823530136,
+      "loss": 0.2351,
+      "step": 1600
+    },
+    {
+      "epoch": 11.284581497797356,
+      "grad_norm": 0.1982435220624061,
+      "learning_rate": 0.0001803198978803177,
+      "loss": 0.2284,
+      "step": 1601
+    },
+    {
+      "epoch": 11.291629955947137,
+      "grad_norm": 0.21672279285497653,
+      "learning_rate": 0.00018029252057162628,
+      "loss": 0.2273,
+      "step": 1602
+    },
+    {
+      "epoch": 11.298678414096916,
+      "grad_norm": 0.2129984169084898,
+      "learning_rate": 0.00018026512631500583,
+      "loss": 0.2332,
+      "step": 1603
+    },
+    {
+      "epoch": 11.305726872246696,
+      "grad_norm": 0.22304779154438623,
+      "learning_rate": 0.00018023771511623866,
+      "loss": 0.237,
+      "step": 1604
+    },
+    {
+      "epoch": 11.312775330396477,
+      "grad_norm": 0.20257881035201722,
+      "learning_rate": 0.00018021028698111065,
+      "loss": 0.2412,
+      "step": 1605
+    },
+    {
+      "epoch": 11.319823788546255,
+      "grad_norm": 0.1906510490336891,
+      "learning_rate": 0.00018018284191541128,
+      "loss": 0.236,
+      "step": 1606
+    },
+    {
+      "epoch": 11.326872246696036,
+      "grad_norm": 0.19959253897510795,
+      "learning_rate": 0.00018015537992493354,
+      "loss": 0.2256,
+      "step": 1607
+    },
+    {
+      "epoch": 11.333920704845815,
+      "grad_norm": 0.19824831346173358,
+      "learning_rate": 0.0001801279010154741,
+      "loss": 0.2266,
+      "step": 1608
+    },
+    {
+      "epoch": 11.340969162995595,
+      "grad_norm": 0.24584562120112313,
+      "learning_rate": 0.00018010040519283308,
+      "loss": 0.2429,
+      "step": 1609
+    },
+    {
+      "epoch": 11.348017621145374,
+      "grad_norm": 0.205232128575824,
+      "learning_rate": 0.00018007289246281421,
+      "loss": 0.2311,
+      "step": 1610
+    },
+    {
+      "epoch": 11.355066079295154,
+      "grad_norm": 0.2069155928083295,
+      "learning_rate": 0.00018004536283122485,
+      "loss": 0.2504,
+      "step": 1611
+    },
+    {
+      "epoch": 11.362114537444935,
+      "grad_norm": 0.19657776595119875,
+      "learning_rate": 0.00018001781630387585,
+      "loss": 0.2336,
+      "step": 1612
+    },
+    {
+      "epoch": 11.369162995594714,
+      "grad_norm": 0.20333709418601442,
+      "learning_rate": 0.00017999025288658167,
+      "loss": 0.2433,
+      "step": 1613
+    },
+    {
+      "epoch": 11.376211453744494,
+      "grad_norm": 0.22123350460922434,
+      "learning_rate": 0.00017996267258516032,
+      "loss": 0.2338,
+      "step": 1614
+    },
+    {
+      "epoch": 11.383259911894273,
+      "grad_norm": 0.20657805544823388,
+      "learning_rate": 0.00017993507540543335,
+      "loss": 0.2214,
+      "step": 1615
+    },
+    {
+      "epoch": 11.390308370044053,
+      "grad_norm": 0.2315886233690114,
+      "learning_rate": 0.00017990746135322592,
+      "loss": 0.2354,
+      "step": 1616
+    },
+    {
+      "epoch": 11.397356828193832,
+      "grad_norm": 0.2183283406037868,
+      "learning_rate": 0.00017987983043436678,
+      "loss": 0.2356,
+      "step": 1617
+    },
+    {
+      "epoch": 11.404405286343613,
+      "grad_norm": 0.21136208155096683,
+      "learning_rate": 0.00017985218265468813,
+      "loss": 0.2351,
+      "step": 1618
+    },
+    {
+      "epoch": 11.411453744493391,
+      "grad_norm": 0.22187218148345228,
+      "learning_rate": 0.00017982451802002575,
+      "loss": 0.2388,
+      "step": 1619
+    },
+    {
+      "epoch": 11.418502202643172,
+      "grad_norm": 0.22321767593404554,
+      "learning_rate": 0.00017979683653621912,
+      "loss": 0.2324,
+      "step": 1620
+    },
+    {
+      "epoch": 11.425550660792952,
+      "grad_norm": 0.22188955531590954,
+      "learning_rate": 0.0001797691382091111,
+      "loss": 0.2397,
+      "step": 1621
+    },
+    {
+      "epoch": 11.432599118942731,
+      "grad_norm": 0.20440139204934116,
+      "learning_rate": 0.0001797414230445482,
+      "loss": 0.2492,
+      "step": 1622
+    },
+    {
+      "epoch": 11.439647577092511,
+      "grad_norm": 0.20991472152372564,
+      "learning_rate": 0.00017971369104838046,
+      "loss": 0.2291,
+      "step": 1623
+    },
+    {
+      "epoch": 11.44669603524229,
+      "grad_norm": 0.22566725532460175,
+      "learning_rate": 0.0001796859422264615,
+      "loss": 0.234,
+      "step": 1624
+    },
+    {
+      "epoch": 11.45374449339207,
+      "grad_norm": 0.23255967715453157,
+      "learning_rate": 0.00017965817658464837,
+      "loss": 0.2278,
+      "step": 1625
+    },
+    {
+      "epoch": 11.46079295154185,
+      "grad_norm": 0.2230130315329347,
+      "learning_rate": 0.00017963039412880193,
+      "loss": 0.2456,
+      "step": 1626
+    },
+    {
+      "epoch": 11.46784140969163,
+      "grad_norm": 0.21493300424071515,
+      "learning_rate": 0.0001796025948647863,
+      "loss": 0.2362,
+      "step": 1627
+    },
+    {
+      "epoch": 11.47488986784141,
+      "grad_norm": 0.19809527878496422,
+      "learning_rate": 0.00017957477879846926,
+      "loss": 0.2281,
+      "step": 1628
+    },
+    {
+      "epoch": 11.48193832599119,
+      "grad_norm": 0.2100829275877845,
+      "learning_rate": 0.00017954694593572227,
+      "loss": 0.2373,
+      "step": 1629
+    },
+    {
+      "epoch": 11.48898678414097,
+      "grad_norm": 0.2142567846651513,
+      "learning_rate": 0.0001795190962824201,
+      "loss": 0.226,
+      "step": 1630
+    },
+    {
+      "epoch": 11.496035242290748,
+      "grad_norm": 0.21792975007316268,
+      "learning_rate": 0.00017949122984444124,
+      "loss": 0.2342,
+      "step": 1631
+    },
+    {
+      "epoch": 11.503083700440529,
+      "grad_norm": 0.21397912178219833,
+      "learning_rate": 0.00017946334662766766,
+      "loss": 0.2355,
+      "step": 1632
+    },
+    {
+      "epoch": 11.510132158590308,
+      "grad_norm": 0.2235428604314411,
+      "learning_rate": 0.00017943544663798485,
+      "loss": 0.2449,
+      "step": 1633
+    },
+    {
+      "epoch": 11.517180616740088,
+      "grad_norm": 0.21522952469817416,
+      "learning_rate": 0.0001794075298812819,
+      "loss": 0.2377,
+      "step": 1634
+    },
+    {
+      "epoch": 11.524229074889869,
+      "grad_norm": 0.20931109163699887,
+      "learning_rate": 0.00017937959636345134,
+      "loss": 0.2437,
+      "step": 1635
+    },
+    {
+      "epoch": 11.531277533039647,
+      "grad_norm": 0.22067617331038447,
+      "learning_rate": 0.00017935164609038937,
+      "loss": 0.2452,
+      "step": 1636
+    },
+    {
+      "epoch": 11.538325991189428,
+      "grad_norm": 0.2076273337700216,
+      "learning_rate": 0.0001793236790679956,
+      "loss": 0.2347,
+      "step": 1637
+    },
+    {
+      "epoch": 11.545374449339207,
+      "grad_norm": 0.2069526971216862,
+      "learning_rate": 0.00017929569530217328,
+      "loss": 0.2436,
+      "step": 1638
+    },
+    {
+      "epoch": 11.552422907488987,
+      "grad_norm": 0.2142494088121796,
+      "learning_rate": 0.00017926769479882916,
+      "loss": 0.2483,
+      "step": 1639
+    },
+    {
+      "epoch": 11.559471365638766,
+      "grad_norm": 0.20374535909817174,
+      "learning_rate": 0.00017923967756387345,
+      "loss": 0.2362,
+      "step": 1640
+    },
+    {
+      "epoch": 11.566519823788546,
+      "grad_norm": 0.22002448626053708,
+      "learning_rate": 0.00017921164360322,
+      "loss": 0.2478,
+      "step": 1641
+    },
+    {
+      "epoch": 11.573568281938327,
+      "grad_norm": 0.22172878845940025,
+      "learning_rate": 0.00017918359292278611,
+      "loss": 0.2411,
+      "step": 1642
+    },
+    {
+      "epoch": 11.580616740088105,
+      "grad_norm": 0.20388396019647667,
+      "learning_rate": 0.00017915552552849267,
+      "loss": 0.2285,
+      "step": 1643
+    },
+    {
+      "epoch": 11.587665198237886,
+      "grad_norm": 0.2204401005523915,
+      "learning_rate": 0.00017912744142626407,
+      "loss": 0.2399,
+      "step": 1644
+    },
+    {
+      "epoch": 11.594713656387665,
+      "grad_norm": 0.21050161170342524,
+      "learning_rate": 0.00017909934062202822,
+      "loss": 0.2322,
+      "step": 1645
+    },
+    {
+      "epoch": 11.601762114537445,
+      "grad_norm": 0.20213903235618189,
+      "learning_rate": 0.00017907122312171657,
+      "loss": 0.2368,
+      "step": 1646
+    },
+    {
+      "epoch": 11.608810572687224,
+      "grad_norm": 0.21232122664399808,
+      "learning_rate": 0.00017904308893126406,
+      "loss": 0.2389,
+      "step": 1647
+    },
+    {
+      "epoch": 11.615859030837004,
+      "grad_norm": 0.19697916690068382,
+      "learning_rate": 0.00017901493805660921,
+      "loss": 0.2506,
+      "step": 1648
+    },
+    {
+      "epoch": 11.622907488986785,
+      "grad_norm": 0.20217642865761729,
+      "learning_rate": 0.00017898677050369398,
+      "loss": 0.2471,
+      "step": 1649
+    },
+    {
+      "epoch": 11.629955947136564,
+      "grad_norm": 0.1957607670557837,
+      "learning_rate": 0.00017895858627846396,
+      "loss": 0.2378,
+      "step": 1650
+    },
+    {
+      "epoch": 11.637004405286344,
+      "grad_norm": 0.2127270065396973,
+      "learning_rate": 0.00017893038538686817,
+      "loss": 0.2422,
+      "step": 1651
+    },
+    {
+      "epoch": 11.644052863436123,
+      "grad_norm": 0.21453174573244893,
+      "learning_rate": 0.0001789021678348592,
+      "loss": 0.2411,
+      "step": 1652
+    },
+    {
+      "epoch": 11.651101321585903,
+      "grad_norm": 0.2114071058890688,
+      "learning_rate": 0.00017887393362839312,
+      "loss": 0.2426,
+      "step": 1653
+    },
+    {
+      "epoch": 11.658149779735682,
+      "grad_norm": 0.20738036529013681,
+      "learning_rate": 0.0001788456827734295,
+      "loss": 0.2288,
+      "step": 1654
+    },
+    {
+      "epoch": 11.665198237885463,
+      "grad_norm": 0.19256023128050523,
+      "learning_rate": 0.0001788174152759315,
+      "loss": 0.2319,
+      "step": 1655
+    },
+    {
+      "epoch": 11.672246696035243,
+      "grad_norm": 0.20243562185214764,
+      "learning_rate": 0.00017878913114186573,
+      "loss": 0.2364,
+      "step": 1656
+    },
+    {
+      "epoch": 11.679295154185022,
+      "grad_norm": 0.19481711765241577,
+      "learning_rate": 0.00017876083037720236,
+      "loss": 0.236,
+      "step": 1657
+    },
+    {
+      "epoch": 11.686343612334802,
+      "grad_norm": 0.21788975541251265,
+      "learning_rate": 0.00017873251298791497,
+      "loss": 0.2522,
+      "step": 1658
+    },
+    {
+      "epoch": 11.693392070484581,
+      "grad_norm": 0.20630387960308952,
+      "learning_rate": 0.00017870417897998078,
+      "loss": 0.2318,
+      "step": 1659
+    },
+    {
+      "epoch": 11.700440528634362,
+      "grad_norm": 0.2151184066874302,
+      "learning_rate": 0.00017867582835938043,
+      "loss": 0.2422,
+      "step": 1660
+    },
+    {
+      "epoch": 11.70748898678414,
+      "grad_norm": 0.20617661848791916,
+      "learning_rate": 0.00017864746113209812,
+      "loss": 0.2458,
+      "step": 1661
+    },
+    {
+      "epoch": 11.71453744493392,
+      "grad_norm": 0.1986868896196073,
+      "learning_rate": 0.00017861907730412147,
+      "loss": 0.2407,
+      "step": 1662
+    },
+    {
+      "epoch": 11.7215859030837,
+      "grad_norm": 0.19591549772232944,
+      "learning_rate": 0.00017859067688144172,
+      "loss": 0.2384,
+      "step": 1663
+    },
+    {
+      "epoch": 11.72863436123348,
+      "grad_norm": 0.19269434577307887,
+      "learning_rate": 0.0001785622598700535,
+      "loss": 0.2354,
+      "step": 1664
+    },
+    {
+      "epoch": 11.73568281938326,
+      "grad_norm": 0.21670721766573403,
+      "learning_rate": 0.00017853382627595508,
+      "loss": 0.2491,
+      "step": 1665
+    },
+    {
+      "epoch": 11.74273127753304,
+      "grad_norm": 0.20543882986473397,
+      "learning_rate": 0.00017850537610514806,
+      "loss": 0.243,
+      "step": 1666
+    },
+    {
+      "epoch": 11.74977973568282,
+      "grad_norm": 0.21099864322951248,
+      "learning_rate": 0.00017847690936363764,
+      "loss": 0.2448,
+      "step": 1667
+    },
+    {
+      "epoch": 11.756828193832598,
+      "grad_norm": 0.21272694757754024,
+      "learning_rate": 0.00017844842605743258,
+      "loss": 0.2503,
+      "step": 1668
+    },
+    {
+      "epoch": 11.763876651982379,
+      "grad_norm": 0.20355887030052394,
+      "learning_rate": 0.00017841992619254497,
+      "loss": 0.2347,
+      "step": 1669
+    },
+    {
+      "epoch": 11.770925110132158,
+      "grad_norm": 0.20469532235106494,
+      "learning_rate": 0.00017839140977499052,
+      "loss": 0.2325,
+      "step": 1670
+    },
+    {
+      "epoch": 11.777973568281938,
+      "grad_norm": 0.21785373149103726,
+      "learning_rate": 0.0001783628768107884,
+      "loss": 0.2424,
+      "step": 1671
+    },
+    {
+      "epoch": 11.785022026431719,
+      "grad_norm": 0.2295437047810133,
+      "learning_rate": 0.00017833432730596127,
+      "loss": 0.2501,
+      "step": 1672
+    },
+    {
+      "epoch": 11.792070484581497,
+      "grad_norm": 0.2086586562617365,
+      "learning_rate": 0.00017830576126653526,
+      "loss": 0.2415,
+      "step": 1673
+    },
+    {
+      "epoch": 11.799118942731278,
+      "grad_norm": 0.2069282437350981,
+      "learning_rate": 0.00017827717869854003,
+      "loss": 0.2506,
+      "step": 1674
+    },
+    {
+      "epoch": 11.806167400881057,
+      "grad_norm": 0.20159473179231774,
+      "learning_rate": 0.00017824857960800877,
+      "loss": 0.2548,
+      "step": 1675
+    },
+    {
+      "epoch": 11.813215859030837,
+      "grad_norm": 0.2168419211325142,
+      "learning_rate": 0.00017821996400097797,
+      "loss": 0.2384,
+      "step": 1676
+    },
+    {
+      "epoch": 11.820264317180616,
+      "grad_norm": 0.20575434131804493,
+      "learning_rate": 0.00017819133188348786,
+      "loss": 0.2449,
+      "step": 1677
+    },
+    {
+      "epoch": 11.827312775330396,
+      "grad_norm": 0.1968942779033645,
+      "learning_rate": 0.00017816268326158196,
+      "loss": 0.2369,
+      "step": 1678
+    },
+    {
+      "epoch": 11.834361233480177,
+      "grad_norm": 0.20919297711698268,
+      "learning_rate": 0.00017813401814130737,
+      "loss": 0.2419,
+      "step": 1679
+    },
+    {
+      "epoch": 11.841409691629956,
+      "grad_norm": 0.2209938743608623,
+      "learning_rate": 0.00017810533652871463,
+      "loss": 0.2427,
+      "step": 1680
+    },
+    {
+      "epoch": 11.848458149779736,
+      "grad_norm": 0.21942908129063252,
+      "learning_rate": 0.00017807663842985777,
+      "loss": 0.2455,
+      "step": 1681
+    },
+    {
+      "epoch": 11.855506607929515,
+      "grad_norm": 0.22300643536996714,
+      "learning_rate": 0.00017804792385079437,
+      "loss": 0.2477,
+      "step": 1682
+    },
+    {
+      "epoch": 11.862555066079295,
+      "grad_norm": 0.22507190468962854,
+      "learning_rate": 0.00017801919279758534,
+      "loss": 0.2425,
+      "step": 1683
+    },
+    {
+      "epoch": 11.869603524229074,
+      "grad_norm": 0.23639318618256497,
+      "learning_rate": 0.00017799044527629521,
+      "loss": 0.2441,
+      "step": 1684
+    },
+    {
+      "epoch": 11.876651982378855,
+      "grad_norm": 0.2141228335363765,
+      "learning_rate": 0.0001779616812929919,
+      "loss": 0.2533,
+      "step": 1685
+    },
+    {
+      "epoch": 11.883700440528635,
+      "grad_norm": 0.2376398794243427,
+      "learning_rate": 0.00017793290085374688,
+      "loss": 0.2393,
+      "step": 1686
+    },
+    {
+      "epoch": 11.890748898678414,
+      "grad_norm": 0.22838672429248144,
+      "learning_rate": 0.00017790410396463497,
+      "loss": 0.2458,
+      "step": 1687
+    },
+    {
+      "epoch": 11.897797356828194,
+      "grad_norm": 0.20609484732710437,
+      "learning_rate": 0.00017787529063173463,
+      "loss": 0.2304,
+      "step": 1688
+    },
+    {
+      "epoch": 11.904845814977973,
+      "grad_norm": 0.21291190647705918,
+      "learning_rate": 0.0001778464608611276,
+      "loss": 0.2365,
+      "step": 1689
+    },
+    {
+      "epoch": 11.911894273127754,
+      "grad_norm": 0.19744865985980387,
+      "learning_rate": 0.0001778176146588993,
+      "loss": 0.2332,
+      "step": 1690
+    },
+    {
+      "epoch": 11.918942731277532,
+      "grad_norm": 0.18518984242727227,
+      "learning_rate": 0.00017778875203113844,
+      "loss": 0.2388,
+      "step": 1691
+    },
+    {
+      "epoch": 11.925991189427313,
+      "grad_norm": 0.21072800896684699,
+      "learning_rate": 0.00017775987298393727,
+      "loss": 0.2426,
+      "step": 1692
+    },
+    {
+      "epoch": 11.933039647577093,
+      "grad_norm": 0.22779358615283163,
+      "learning_rate": 0.00017773097752339152,
+      "loss": 0.2428,
+      "step": 1693
+    },
+    {
+      "epoch": 11.940088105726872,
+      "grad_norm": 0.223123572503929,
+      "learning_rate": 0.00017770206565560033,
+      "loss": 0.2532,
+      "step": 1694
+    },
+    {
+      "epoch": 11.947136563876652,
+      "grad_norm": 0.19248729515452886,
+      "learning_rate": 0.00017767313738666638,
+      "loss": 0.2444,
+      "step": 1695
+    },
+    {
+      "epoch": 11.954185022026431,
+      "grad_norm": 0.2038959456668687,
+      "learning_rate": 0.00017764419272269576,
+      "loss": 0.2417,
+      "step": 1696
+    },
+    {
+      "epoch": 11.961233480176212,
+      "grad_norm": 0.1791936032068716,
+      "learning_rate": 0.00017761523166979806,
+      "loss": 0.2386,
+      "step": 1697
+    },
+    {
+      "epoch": 11.96828193832599,
+      "grad_norm": 0.211115589268615,
+      "learning_rate": 0.0001775862542340862,
+      "loss": 0.2453,
+      "step": 1698
+    },
+    {
+      "epoch": 11.975330396475771,
+      "grad_norm": 0.19788564433943673,
+      "learning_rate": 0.0001775572604216768,
+      "loss": 0.237,
+      "step": 1699
+    },
+    {
+      "epoch": 11.982378854625551,
+      "grad_norm": 0.23542378106559061,
+      "learning_rate": 0.0001775282502386897,
+      "loss": 0.2423,
+      "step": 1700
+    },
+    {
+      "epoch": 11.98942731277533,
+      "grad_norm": 0.204146222463213,
+      "learning_rate": 0.00017749922369124832,
+      "loss": 0.2344,
+      "step": 1701
+    },
+    {
+      "epoch": 11.99647577092511,
+      "grad_norm": 0.21672382211641997,
+      "learning_rate": 0.0001774701807854795,
+      "loss": 0.254,
+      "step": 1702
+    },
+    {
+      "epoch": 12.00352422907489,
+      "grad_norm": 0.23406445810727103,
+      "learning_rate": 0.00017744112152751353,
+      "loss": 0.2267,
+      "step": 1703
+    },
+    {
+      "epoch": 12.01057268722467,
+      "grad_norm": 0.20960880983029562,
+      "learning_rate": 0.00017741204592348417,
+      "loss": 0.1976,
+      "step": 1704
+    },
+    {
+      "epoch": 12.017621145374449,
+      "grad_norm": 0.31261991313650805,
+      "learning_rate": 0.00017738295397952864,
+      "loss": 0.2037,
+      "step": 1705
+    },
+    {
+      "epoch": 12.024669603524229,
+      "grad_norm": 0.28368221539959504,
+      "learning_rate": 0.00017735384570178754,
+      "loss": 0.2121,
+      "step": 1706
+    },
+    {
+      "epoch": 12.03171806167401,
+      "grad_norm": 0.24632929881858362,
+      "learning_rate": 0.00017732472109640503,
+      "loss": 0.2055,
+      "step": 1707
+    },
+    {
+      "epoch": 12.038766519823788,
+      "grad_norm": 0.2510858471747751,
+      "learning_rate": 0.0001772955801695286,
+      "loss": 0.19,
+      "step": 1708
+    },
+    {
+      "epoch": 12.045814977973569,
+      "grad_norm": 0.26511794927042776,
+      "learning_rate": 0.00017726642292730928,
+      "loss": 0.2043,
+      "step": 1709
+    },
+    {
+      "epoch": 12.052863436123348,
+      "grad_norm": 0.2534135058057462,
+      "learning_rate": 0.00017723724937590145,
+      "loss": 0.2008,
+      "step": 1710
+    },
+    {
+      "epoch": 12.059911894273128,
+      "grad_norm": 0.26063905687325234,
+      "learning_rate": 0.00017720805952146305,
+      "loss": 0.2103,
+      "step": 1711
+    },
+    {
+      "epoch": 12.066960352422907,
+      "grad_norm": 0.2664324807969897,
+      "learning_rate": 0.0001771788533701553,
+      "loss": 0.1969,
+      "step": 1712
+    },
+    {
+      "epoch": 12.074008810572687,
+      "grad_norm": 0.23529671101246427,
+      "learning_rate": 0.00017714963092814307,
+      "loss": 0.196,
+      "step": 1713
+    },
+    {
+      "epoch": 12.081057268722468,
+      "grad_norm": 0.23536259647199118,
+      "learning_rate": 0.0001771203922015945,
+      "loss": 0.2039,
+      "step": 1714
+    },
+    {
+      "epoch": 12.088105726872246,
+      "grad_norm": 0.2278588273283411,
+      "learning_rate": 0.0001770911371966812,
+      "loss": 0.1997,
+      "step": 1715
+    },
+    {
+      "epoch": 12.095154185022027,
+      "grad_norm": 0.21186729484965994,
+      "learning_rate": 0.00017706186591957826,
+      "loss": 0.208,
+      "step": 1716
+    },
+    {
+      "epoch": 12.102202643171806,
+      "grad_norm": 0.25211703298424004,
+      "learning_rate": 0.00017703257837646424,
+      "loss": 0.2016,
+      "step": 1717
+    },
+    {
+      "epoch": 12.109251101321586,
+      "grad_norm": 0.25543202076509997,
+      "learning_rate": 0.00017700327457352094,
+      "loss": 0.2097,
+      "step": 1718
+    },
+    {
+      "epoch": 12.116299559471365,
+      "grad_norm": 0.24663549316844408,
+      "learning_rate": 0.00017697395451693386,
+      "loss": 0.2062,
+      "step": 1719
+    },
+    {
+      "epoch": 12.123348017621145,
+      "grad_norm": 0.24146837319197825,
+      "learning_rate": 0.00017694461821289172,
+      "loss": 0.2,
+      "step": 1720
+    },
+    {
+      "epoch": 12.130396475770926,
+      "grad_norm": 0.2369856225027672,
+      "learning_rate": 0.0001769152656675868,
+      "loss": 0.1939,
+      "step": 1721
+    },
+    {
+      "epoch": 12.137444933920705,
+      "grad_norm": 0.24514977133161617,
+      "learning_rate": 0.0001768858968872147,
+      "loss": 0.1974,
+      "step": 1722
+    },
+    {
+      "epoch": 12.144493392070485,
+      "grad_norm": 0.262868756138796,
+      "learning_rate": 0.00017685651187797453,
+      "loss": 0.2066,
+      "step": 1723
+    },
+    {
+      "epoch": 12.151541850220264,
+      "grad_norm": 0.23027402736523342,
+      "learning_rate": 0.0001768271106460688,
+      "loss": 0.1931,
+      "step": 1724
+    },
+    {
+      "epoch": 12.158590308370044,
+      "grad_norm": 0.2319278656827074,
+      "learning_rate": 0.00017679769319770344,
+      "loss": 0.2101,
+      "step": 1725
+    },
+    {
+      "epoch": 12.165638766519823,
+      "grad_norm": 0.22899547906333437,
+      "learning_rate": 0.00017676825953908784,
+      "loss": 0.2019,
+      "step": 1726
+    },
+    {
+      "epoch": 12.172687224669604,
+      "grad_norm": 0.23051708177344615,
+      "learning_rate": 0.00017673880967643472,
+      "loss": 0.2054,
+      "step": 1727
+    },
+    {
+      "epoch": 12.179735682819384,
+      "grad_norm": 0.23349792655805618,
+      "learning_rate": 0.00017670934361596033,
+      "loss": 0.2065,
+      "step": 1728
+    },
+    {
+      "epoch": 12.186784140969163,
+      "grad_norm": 0.2557285932189355,
+      "learning_rate": 0.00017667986136388422,
+      "loss": 0.2008,
+      "step": 1729
+    },
+    {
+      "epoch": 12.193832599118943,
+      "grad_norm": 0.23847313934316985,
+      "learning_rate": 0.0001766503629264295,
+      "loss": 0.2002,
+      "step": 1730
+    },
+    {
+      "epoch": 12.200881057268722,
+      "grad_norm": 0.23279986808985215,
+      "learning_rate": 0.0001766208483098226,
+      "loss": 0.2059,
+      "step": 1731
+    },
+    {
+      "epoch": 12.207929515418503,
+      "grad_norm": 0.23127258247822485,
+      "learning_rate": 0.00017659131752029335,
+      "loss": 0.2019,
+      "step": 1732
+    },
+    {
+      "epoch": 12.214977973568281,
+      "grad_norm": 0.23557069580211115,
+      "learning_rate": 0.00017656177056407505,
+      "loss": 0.1977,
+      "step": 1733
+    },
+    {
+      "epoch": 12.222026431718062,
+      "grad_norm": 0.2512854978664044,
+      "learning_rate": 0.00017653220744740438,
+      "loss": 0.2096,
+      "step": 1734
+    },
+    {
+      "epoch": 12.229074889867842,
+      "grad_norm": 0.2456494433649835,
+      "learning_rate": 0.0001765026281765215,
+      "loss": 0.205,
+      "step": 1735
+    },
+    {
+      "epoch": 12.236123348017621,
+      "grad_norm": 0.25807929721606715,
+      "learning_rate": 0.00017647303275766983,
+      "loss": 0.2121,
+      "step": 1736
+    },
+    {
+      "epoch": 12.243171806167402,
+      "grad_norm": 0.2450005115804829,
+      "learning_rate": 0.00017644342119709636,
+      "loss": 0.1966,
+      "step": 1737
+    },
+    {
+      "epoch": 12.25022026431718,
+      "grad_norm": 0.23269794193607468,
+      "learning_rate": 0.00017641379350105143,
+      "loss": 0.2022,
+      "step": 1738
+    },
+    {
+      "epoch": 12.25726872246696,
+      "grad_norm": 0.22478975766256443,
+      "learning_rate": 0.00017638414967578871,
+      "loss": 0.2172,
+      "step": 1739
+    },
+    {
+      "epoch": 12.26431718061674,
+      "grad_norm": 0.24642328691062998,
+      "learning_rate": 0.0001763544897275654,
+      "loss": 0.1971,
+      "step": 1740
+    },
+    {
+      "epoch": 12.27136563876652,
+      "grad_norm": 0.24057214903488203,
+      "learning_rate": 0.00017632481366264196,
+      "loss": 0.2039,
+      "step": 1741
+    },
+    {
+      "epoch": 12.278414096916299,
+      "grad_norm": 0.23871994132089155,
+      "learning_rate": 0.00017629512148728245,
+      "loss": 0.2109,
+      "step": 1742
+    },
+    {
+      "epoch": 12.28546255506608,
+      "grad_norm": 0.24855095365018515,
+      "learning_rate": 0.0001762654132077541,
+      "loss": 0.2016,
+      "step": 1743
+    },
+    {
+      "epoch": 12.29251101321586,
+      "grad_norm": 0.24347772869830703,
+      "learning_rate": 0.00017623568883032772,
+      "loss": 0.2108,
+      "step": 1744
+    },
+    {
+      "epoch": 12.299559471365638,
+      "grad_norm": 0.23231574774470662,
+      "learning_rate": 0.00017620594836127747,
+      "loss": 0.2013,
+      "step": 1745
+    },
+    {
+      "epoch": 12.306607929515419,
+      "grad_norm": 0.21513613118243557,
+      "learning_rate": 0.00017617619180688085,
+      "loss": 0.1958,
+      "step": 1746
+    },
+    {
+      "epoch": 12.313656387665198,
+      "grad_norm": 0.22004128617190058,
+      "learning_rate": 0.00017614641917341878,
+      "loss": 0.2073,
+      "step": 1747
+    },
+    {
+      "epoch": 12.320704845814978,
+      "grad_norm": 0.2380932640128837,
+      "learning_rate": 0.00017611663046717562,
+      "loss": 0.195,
+      "step": 1748
+    },
+    {
+      "epoch": 12.327753303964757,
+      "grad_norm": 0.2550235179382209,
+      "learning_rate": 0.00017608682569443906,
+      "loss": 0.2146,
+      "step": 1749
+    },
+    {
+      "epoch": 12.334801762114537,
+      "grad_norm": 0.24149710013111816,
+      "learning_rate": 0.00017605700486150023,
+      "loss": 0.2091,
+      "step": 1750
+    },
+    {
+      "epoch": 12.341850220264318,
+      "grad_norm": 0.23187015645058,
+      "learning_rate": 0.00017602716797465365,
+      "loss": 0.2112,
+      "step": 1751
+    },
+    {
+      "epoch": 12.348898678414097,
+      "grad_norm": 0.22095060474108758,
+      "learning_rate": 0.00017599731504019717,
+      "loss": 0.2028,
+      "step": 1752
+    },
+    {
+      "epoch": 12.355947136563877,
+      "grad_norm": 0.22890745733622495,
+      "learning_rate": 0.0001759674460644321,
+      "loss": 0.2031,
+      "step": 1753
+    },
+    {
+      "epoch": 12.362995594713656,
+      "grad_norm": 0.23251978970779053,
+      "learning_rate": 0.00017593756105366306,
+      "loss": 0.2086,
+      "step": 1754
+    },
+    {
+      "epoch": 12.370044052863436,
+      "grad_norm": 0.24186618403252858,
+      "learning_rate": 0.00017590766001419813,
+      "loss": 0.2137,
+      "step": 1755
+    },
+    {
+      "epoch": 12.377092511013215,
+      "grad_norm": 0.2492409499239499,
+      "learning_rate": 0.00017587774295234877,
+      "loss": 0.2165,
+      "step": 1756
+    },
+    {
+      "epoch": 12.384140969162996,
+      "grad_norm": 0.26259468489343585,
+      "learning_rate": 0.00017584780987442975,
+      "loss": 0.2024,
+      "step": 1757
+    },
+    {
+      "epoch": 12.391189427312776,
+      "grad_norm": 0.22537159596782566,
+      "learning_rate": 0.00017581786078675927,
+      "loss": 0.2149,
+      "step": 1758
+    },
+    {
+      "epoch": 12.398237885462555,
+      "grad_norm": 0.25222517617155377,
+      "learning_rate": 0.0001757878956956589,
+      "loss": 0.2059,
+      "step": 1759
+    },
+    {
+      "epoch": 12.405286343612335,
+      "grad_norm": 0.23250235123148308,
+      "learning_rate": 0.0001757579146074536,
+      "loss": 0.2143,
+      "step": 1760
+    },
+    {
+      "epoch": 12.412334801762114,
+      "grad_norm": 0.2226055200738205,
+      "learning_rate": 0.00017572791752847172,
+      "loss": 0.2152,
+      "step": 1761
+    },
+    {
+      "epoch": 12.419383259911895,
+      "grad_norm": 0.23900631452270898,
+      "learning_rate": 0.00017569790446504487,
+      "loss": 0.2102,
+      "step": 1762
+    },
+    {
+      "epoch": 12.426431718061673,
+      "grad_norm": 0.2305553795511685,
+      "learning_rate": 0.00017566787542350824,
+      "loss": 0.1995,
+      "step": 1763
+    },
+    {
+      "epoch": 12.433480176211454,
+      "grad_norm": 0.2349954353886335,
+      "learning_rate": 0.00017563783041020024,
+      "loss": 0.211,
+      "step": 1764
+    },
+    {
+      "epoch": 12.440528634361234,
+      "grad_norm": 0.24187650234246552,
+      "learning_rate": 0.00017560776943146268,
+      "loss": 0.2188,
+      "step": 1765
+    },
+    {
+      "epoch": 12.447577092511013,
+      "grad_norm": 0.2120366753691329,
+      "learning_rate": 0.00017557769249364072,
+      "loss": 0.2017,
+      "step": 1766
+    },
+    {
+      "epoch": 12.454625550660793,
+      "grad_norm": 0.23498314008817794,
+      "learning_rate": 0.000175547599603083,
+      "loss": 0.2112,
+      "step": 1767
+    },
+    {
+      "epoch": 12.461674008810572,
+      "grad_norm": 0.218399852014103,
+      "learning_rate": 0.00017551749076614136,
+      "loss": 0.2083,
+      "step": 1768
+    },
+    {
+      "epoch": 12.468722466960353,
+      "grad_norm": 0.20929457534881674,
+      "learning_rate": 0.00017548736598917115,
+      "loss": 0.2055,
+      "step": 1769
+    },
+    {
+      "epoch": 12.475770925110131,
+      "grad_norm": 0.22673108136320663,
+      "learning_rate": 0.000175457225278531,
+      "loss": 0.2107,
+      "step": 1770
+    },
+    {
+      "epoch": 12.482819383259912,
+      "grad_norm": 0.23854452691004077,
+      "learning_rate": 0.00017542706864058296,
+      "loss": 0.2151,
+      "step": 1771
+    },
+    {
+      "epoch": 12.489867841409692,
+      "grad_norm": 0.22903883170207368,
+      "learning_rate": 0.00017539689608169238,
+      "loss": 0.206,
+      "step": 1772
+    },
+    {
+      "epoch": 12.496916299559471,
+      "grad_norm": 0.22947550640312875,
+      "learning_rate": 0.000175366707608228,
+      "loss": 0.2101,
+      "step": 1773
+    },
+    {
+      "epoch": 12.503964757709252,
+      "grad_norm": 0.21090460260611288,
+      "learning_rate": 0.00017533650322656193,
+      "loss": 0.2111,
+      "step": 1774
+    },
+    {
+      "epoch": 12.51101321585903,
+      "grad_norm": 0.2148168194204494,
+      "learning_rate": 0.00017530628294306965,
+      "loss": 0.2153,
+      "step": 1775
+    },
+    {
+      "epoch": 12.51806167400881,
+      "grad_norm": 0.24149036434398846,
+      "learning_rate": 0.00017527604676412996,
+      "loss": 0.2107,
+      "step": 1776
+    },
+    {
+      "epoch": 12.52511013215859,
+      "grad_norm": 0.21313999805317224,
+      "learning_rate": 0.00017524579469612502,
+      "loss": 0.2037,
+      "step": 1777
+    },
+    {
+      "epoch": 12.53215859030837,
+      "grad_norm": 0.24683090807373734,
+      "learning_rate": 0.00017521552674544037,
+      "loss": 0.2135,
+      "step": 1778
+    },
+    {
+      "epoch": 12.53920704845815,
+      "grad_norm": 0.23628771613081279,
+      "learning_rate": 0.0001751852429184649,
+      "loss": 0.2172,
+      "step": 1779
+    },
+    {
+      "epoch": 12.54625550660793,
+      "grad_norm": 0.2421058971072757,
+      "learning_rate": 0.0001751549432215908,
+      "loss": 0.2223,
+      "step": 1780
+    },
+    {
+      "epoch": 12.55330396475771,
+      "grad_norm": 0.2350578655601712,
+      "learning_rate": 0.0001751246276612137,
+      "loss": 0.2254,
+      "step": 1781
+    },
+    {
+      "epoch": 12.560352422907489,
+      "grad_norm": 0.22992684825561,
+      "learning_rate": 0.00017509429624373247,
+      "loss": 0.2138,
+      "step": 1782
+    },
+    {
+      "epoch": 12.567400881057269,
+      "grad_norm": 0.21565665948060975,
+      "learning_rate": 0.00017506394897554941,
+      "loss": 0.2071,
+      "step": 1783
+    },
+    {
+      "epoch": 12.574449339207048,
+      "grad_norm": 0.23041085548976162,
+      "learning_rate": 0.00017503358586307014,
+      "loss": 0.2147,
+      "step": 1784
+    },
+    {
+      "epoch": 12.581497797356828,
+      "grad_norm": 0.23679784344169866,
+      "learning_rate": 0.00017500320691270365,
+      "loss": 0.2101,
+      "step": 1785
+    },
+    {
+      "epoch": 12.588546255506607,
+      "grad_norm": 0.26906327763742277,
+      "learning_rate": 0.0001749728121308622,
+      "loss": 0.2127,
+      "step": 1786
+    },
+    {
+      "epoch": 12.595594713656387,
+      "grad_norm": 0.22955993002218555,
+      "learning_rate": 0.00017494240152396147,
+      "loss": 0.2056,
+      "step": 1787
+    },
+    {
+      "epoch": 12.602643171806168,
+      "grad_norm": 0.2252065554609689,
+      "learning_rate": 0.00017491197509842044,
+      "loss": 0.2101,
+      "step": 1788
+    },
+    {
+      "epoch": 12.609691629955947,
+      "grad_norm": 0.21120921935331993,
+      "learning_rate": 0.00017488153286066143,
+      "loss": 0.217,
+      "step": 1789
+    },
+    {
+      "epoch": 12.616740088105727,
+      "grad_norm": 0.2374346094913998,
+      "learning_rate": 0.00017485107481711012,
+      "loss": 0.2251,
+      "step": 1790
+    },
+    {
+      "epoch": 12.623788546255506,
+      "grad_norm": 0.2346802973952338,
+      "learning_rate": 0.0001748206009741955,
+      "loss": 0.2188,
+      "step": 1791
+    },
+    {
+      "epoch": 12.630837004405286,
+      "grad_norm": 0.239747480104051,
+      "learning_rate": 0.00017479011133834992,
+      "loss": 0.2178,
+      "step": 1792
+    },
+    {
+      "epoch": 12.637885462555065,
+      "grad_norm": 0.24791559713597697,
+      "learning_rate": 0.00017475960591600904,
+      "loss": 0.2156,
+      "step": 1793
+    },
+    {
+      "epoch": 12.644933920704846,
+      "grad_norm": 0.2424220331998214,
+      "learning_rate": 0.00017472908471361187,
+      "loss": 0.2089,
+      "step": 1794
+    },
+    {
+      "epoch": 12.651982378854626,
+      "grad_norm": 0.23496482761426796,
+      "learning_rate": 0.00017469854773760073,
+      "loss": 0.2256,
+      "step": 1795
+    },
+    {
+      "epoch": 12.659030837004405,
+      "grad_norm": 0.2306342689872748,
+      "learning_rate": 0.0001746679949944213,
+      "loss": 0.2168,
+      "step": 1796
+    },
+    {
+      "epoch": 12.666079295154185,
+      "grad_norm": 0.23021236371592937,
+      "learning_rate": 0.00017463742649052253,
+      "loss": 0.2163,
+      "step": 1797
+    },
+    {
+      "epoch": 12.673127753303964,
+      "grad_norm": 0.222941107919682,
+      "learning_rate": 0.0001746068422323568,
+      "loss": 0.2112,
+      "step": 1798
+    },
+    {
+      "epoch": 12.680176211453745,
+      "grad_norm": 0.23266327712179496,
+      "learning_rate": 0.0001745762422263797,
+      "loss": 0.1997,
+      "step": 1799
+    },
+    {
+      "epoch": 12.687224669603523,
+      "grad_norm": 0.22858122917664914,
+      "learning_rate": 0.00017454562647905024,
+      "loss": 0.2129,
+      "step": 1800
+    },
+    {
+      "epoch": 12.694273127753304,
+      "grad_norm": 0.24357712874979803,
+      "learning_rate": 0.00017451499499683069,
+      "loss": 0.2218,
+      "step": 1801
+    },
+    {
+      "epoch": 12.701321585903084,
+      "grad_norm": 0.2360516833397959,
+      "learning_rate": 0.00017448434778618665,
+      "loss": 0.2169,
+      "step": 1802
+    },
+    {
+      "epoch": 12.708370044052863,
+      "grad_norm": 0.22591015359606756,
+      "learning_rate": 0.00017445368485358707,
+      "loss": 0.2049,
+      "step": 1803
+    },
+    {
+      "epoch": 12.715418502202644,
+      "grad_norm": 0.22109916233307814,
+      "learning_rate": 0.00017442300620550415,
+      "loss": 0.2043,
+      "step": 1804
+    },
+    {
+      "epoch": 12.722466960352422,
+      "grad_norm": 0.22551211424422124,
+      "learning_rate": 0.0001743923118484135,
+      "loss": 0.2224,
+      "step": 1805
+    },
+    {
+      "epoch": 12.729515418502203,
+      "grad_norm": 0.22834537750708098,
+      "learning_rate": 0.00017436160178879403,
+      "loss": 0.2141,
+      "step": 1806
+    },
+    {
+      "epoch": 12.736563876651982,
+      "grad_norm": 0.24543408422659477,
+      "learning_rate": 0.00017433087603312792,
+      "loss": 0.2229,
+      "step": 1807
+    },
+    {
+      "epoch": 12.743612334801762,
+      "grad_norm": 0.2232645301814986,
+      "learning_rate": 0.00017430013458790062,
+      "loss": 0.219,
+      "step": 1808
+    },
+    {
+      "epoch": 12.750660792951543,
+      "grad_norm": 0.23140819660227588,
+      "learning_rate": 0.00017426937745960102,
+      "loss": 0.2218,
+      "step": 1809
+    },
+    {
+      "epoch": 12.757709251101321,
+      "grad_norm": 0.22303667806802427,
+      "learning_rate": 0.0001742386046547212,
+      "loss": 0.2214,
+      "step": 1810
+    },
+    {
+      "epoch": 12.764757709251102,
+      "grad_norm": 0.23175381026863076,
+      "learning_rate": 0.00017420781617975665,
+      "loss": 0.2129,
+      "step": 1811
+    },
+    {
+      "epoch": 12.77180616740088,
+      "grad_norm": 0.23654694348508581,
+      "learning_rate": 0.0001741770120412061,
+      "loss": 0.2181,
+      "step": 1812
+    },
+    {
+      "epoch": 12.778854625550661,
+      "grad_norm": 0.25420221109572005,
+      "learning_rate": 0.00017414619224557162,
+      "loss": 0.2083,
+      "step": 1813
+    },
+    {
+      "epoch": 12.78590308370044,
+      "grad_norm": 0.22267693714454226,
+      "learning_rate": 0.00017411535679935847,
+      "loss": 0.2209,
+      "step": 1814
+    },
+    {
+      "epoch": 12.79295154185022,
+      "grad_norm": 0.23685873293149293,
+      "learning_rate": 0.0001740845057090755,
+      "loss": 0.2166,
+      "step": 1815
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 0.22683797620182605,
+      "learning_rate": 0.0001740536389812345,
+      "loss": 0.2144,
+      "step": 1816
+    },
+    {
+      "epoch": 12.80704845814978,
+      "grad_norm": 0.21367794202831397,
+      "learning_rate": 0.00017402275662235081,
+      "loss": 0.2203,
+      "step": 1817
+    },
+    {
+      "epoch": 12.81409691629956,
+      "grad_norm": 0.2256738097888267,
+      "learning_rate": 0.00017399185863894298,
+      "loss": 0.2079,
+      "step": 1818
+    },
+    {
+      "epoch": 12.821145374449339,
+      "grad_norm": 0.2316250646138615,
+      "learning_rate": 0.00017396094503753292,
+      "loss": 0.2079,
+      "step": 1819
+    },
+    {
+      "epoch": 12.82819383259912,
+      "grad_norm": 0.216364278303892,
+      "learning_rate": 0.00017393001582464575,
+      "loss": 0.2102,
+      "step": 1820
+    },
+    {
+      "epoch": 12.835242290748898,
+      "grad_norm": 0.2338592956848604,
+      "learning_rate": 0.0001738990710068099,
+      "loss": 0.2118,
+      "step": 1821
+    },
+    {
+      "epoch": 12.842290748898678,
+      "grad_norm": 0.21475222070684222,
+      "learning_rate": 0.00017386811059055715,
+      "loss": 0.2132,
+      "step": 1822
+    },
+    {
+      "epoch": 12.849339207048459,
+      "grad_norm": 0.22453168355939154,
+      "learning_rate": 0.00017383713458242255,
+      "loss": 0.2137,
+      "step": 1823
+    },
+    {
+      "epoch": 12.856387665198238,
+      "grad_norm": 0.2474907080327533,
+      "learning_rate": 0.00017380614298894442,
+      "loss": 0.223,
+      "step": 1824
+    },
+    {
+      "epoch": 12.863436123348018,
+      "grad_norm": 0.21803102538773167,
+      "learning_rate": 0.00017377513581666435,
+      "loss": 0.2192,
+      "step": 1825
+    },
+    {
+      "epoch": 12.870484581497797,
+      "grad_norm": 0.21072197326345632,
+      "learning_rate": 0.0001737441130721273,
+      "loss": 0.2108,
+      "step": 1826
+    },
+    {
+      "epoch": 12.877533039647577,
+      "grad_norm": 0.20876598337678617,
+      "learning_rate": 0.00017371307476188148,
+      "loss": 0.2163,
+      "step": 1827
+    },
+    {
+      "epoch": 12.884581497797356,
+      "grad_norm": 0.23485953412020996,
+      "learning_rate": 0.00017368202089247832,
+      "loss": 0.2154,
+      "step": 1828
+    },
+    {
+      "epoch": 12.891629955947137,
+      "grad_norm": 0.2256325621021266,
+      "learning_rate": 0.00017365095147047258,
+      "loss": 0.2195,
+      "step": 1829
+    },
+    {
+      "epoch": 12.898678414096917,
+      "grad_norm": 0.24848847650936767,
+      "learning_rate": 0.00017361986650242238,
+      "loss": 0.2257,
+      "step": 1830
+    },
+    {
+      "epoch": 12.905726872246696,
+      "grad_norm": 0.2145263433912583,
+      "learning_rate": 0.000173588765994889,
+      "loss": 0.2158,
+      "step": 1831
+    },
+    {
+      "epoch": 12.912775330396476,
+      "grad_norm": 0.20616310582799316,
+      "learning_rate": 0.00017355764995443708,
+      "loss": 0.2111,
+      "step": 1832
+    },
+    {
+      "epoch": 12.919823788546255,
+      "grad_norm": 0.20574122999193398,
+      "learning_rate": 0.00017352651838763446,
+      "loss": 0.2155,
+      "step": 1833
+    },
+    {
+      "epoch": 12.926872246696036,
+      "grad_norm": 0.21415171031856645,
+      "learning_rate": 0.00017349537130105237,
+      "loss": 0.2164,
+      "step": 1834
+    },
+    {
+      "epoch": 12.933920704845814,
+      "grad_norm": 0.23166408932438964,
+      "learning_rate": 0.0001734642087012652,
+      "loss": 0.2198,
+      "step": 1835
+    },
+    {
+      "epoch": 12.940969162995595,
+      "grad_norm": 0.26541218105262676,
+      "learning_rate": 0.0001734330305948507,
+      "loss": 0.2227,
+      "step": 1836
+    },
+    {
+      "epoch": 12.948017621145375,
+      "grad_norm": 0.22863684572594975,
+      "learning_rate": 0.0001734018369883898,
+      "loss": 0.2134,
+      "step": 1837
+    },
+    {
+      "epoch": 12.955066079295154,
+      "grad_norm": 0.21651143052167737,
+      "learning_rate": 0.0001733706278884669,
+      "loss": 0.2209,
+      "step": 1838
+    },
+    {
+      "epoch": 12.962114537444934,
+      "grad_norm": 0.21826394859779105,
+      "learning_rate": 0.0001733394033016694,
+      "loss": 0.2077,
+      "step": 1839
+    },
+    {
+      "epoch": 12.969162995594713,
+      "grad_norm": 0.23215811247514745,
+      "learning_rate": 0.00017330816323458815,
+      "loss": 0.215,
+      "step": 1840
+    },
+    {
+      "epoch": 12.976211453744494,
+      "grad_norm": 0.21532182690590193,
+      "learning_rate": 0.0001732769076938172,
+      "loss": 0.2181,
+      "step": 1841
+    },
+    {
+      "epoch": 12.983259911894272,
+      "grad_norm": 0.22767260177560011,
+      "learning_rate": 0.00017324563668595392,
+      "loss": 0.2141,
+      "step": 1842
+    },
+    {
+      "epoch": 12.990308370044053,
+      "grad_norm": 0.22037439120875787,
+      "learning_rate": 0.0001732143502175989,
+      "loss": 0.2161,
+      "step": 1843
+    },
+    {
+      "epoch": 12.997356828193833,
+      "grad_norm": 0.23333828512576893,
+      "learning_rate": 0.000173183048295356,
+      "loss": 0.2125,
+      "step": 1844
+    },
+    {
+      "epoch": 13.004405286343612,
+      "grad_norm": 0.21952810555916177,
+      "learning_rate": 0.00017315173092583233,
+      "loss": 0.1845,
+      "step": 1845
+    },
+    {
+      "epoch": 13.011453744493393,
+      "grad_norm": 0.22289584510020632,
+      "learning_rate": 0.00017312039811563833,
+      "loss": 0.1727,
+      "step": 1846
+    },
+    {
+      "epoch": 13.018502202643171,
+      "grad_norm": 0.24753663803885537,
+      "learning_rate": 0.00017308904987138753,
+      "loss": 0.1737,
+      "step": 1847
+    },
+    {
+      "epoch": 13.025550660792952,
+      "grad_norm": 0.30758152840465786,
+      "learning_rate": 0.00017305768619969695,
+      "loss": 0.1708,
+      "step": 1848
+    },
+    {
+      "epoch": 13.03259911894273,
+      "grad_norm": 0.2608043366298787,
+      "learning_rate": 0.0001730263071071867,
+      "loss": 0.1701,
+      "step": 1849
+    },
+    {
+      "epoch": 13.039647577092511,
+      "grad_norm": 0.25171764142138814,
+      "learning_rate": 0.0001729949126004802,
+      "loss": 0.1678,
+      "step": 1850
+    },
+    {
+      "epoch": 13.046696035242292,
+      "grad_norm": 0.25403386291831187,
+      "learning_rate": 0.00017296350268620412,
+      "loss": 0.1642,
+      "step": 1851
+    },
+    {
+      "epoch": 13.05374449339207,
+      "grad_norm": 0.25748610175962,
+      "learning_rate": 0.00017293207737098835,
+      "loss": 0.1635,
+      "step": 1852
+    },
+    {
+      "epoch": 13.06079295154185,
+      "grad_norm": 0.25310667080096255,
+      "learning_rate": 0.0001729006366614661,
+      "loss": 0.1732,
+      "step": 1853
+    },
+    {
+      "epoch": 13.06784140969163,
+      "grad_norm": 0.2687833423712349,
+      "learning_rate": 0.0001728691805642738,
+      "loss": 0.1757,
+      "step": 1854
+    },
+    {
+      "epoch": 13.07488986784141,
+      "grad_norm": 0.2442375482955018,
+      "learning_rate": 0.0001728377090860511,
+      "loss": 0.1661,
+      "step": 1855
+    },
+    {
+      "epoch": 13.081938325991189,
+      "grad_norm": 0.24450702177789727,
+      "learning_rate": 0.00017280622223344086,
+      "loss": 0.1685,
+      "step": 1856
+    },
+    {
+      "epoch": 13.08898678414097,
+      "grad_norm": 0.2394595600922681,
+      "learning_rate": 0.00017277472001308938,
+      "loss": 0.1614,
+      "step": 1857
+    },
+    {
+      "epoch": 13.09603524229075,
+      "grad_norm": 0.2514059471291986,
+      "learning_rate": 0.0001727432024316459,
+      "loss": 0.1697,
+      "step": 1858
+    },
+    {
+      "epoch": 13.103083700440529,
+      "grad_norm": 0.2557866160020716,
+      "learning_rate": 0.00017271166949576317,
+      "loss": 0.1766,
+      "step": 1859
+    },
+    {
+      "epoch": 13.110132158590309,
+      "grad_norm": 0.2683168063032112,
+      "learning_rate": 0.00017268012121209703,
+      "loss": 0.1639,
+      "step": 1860
+    },
+    {
+      "epoch": 13.117180616740088,
+      "grad_norm": 0.24555903525303963,
+      "learning_rate": 0.00017264855758730664,
+      "loss": 0.1729,
+      "step": 1861
+    },
+    {
+      "epoch": 13.124229074889868,
+      "grad_norm": 0.26417046723915516,
+      "learning_rate": 0.00017261697862805435,
+      "loss": 0.1794,
+      "step": 1862
+    },
+    {
+      "epoch": 13.131277533039647,
+      "grad_norm": 0.24493977470995737,
+      "learning_rate": 0.00017258538434100577,
+      "loss": 0.1772,
+      "step": 1863
+    },
+    {
+      "epoch": 13.138325991189427,
+      "grad_norm": 0.25588607083269155,
+      "learning_rate": 0.00017255377473282971,
+      "loss": 0.1705,
+      "step": 1864
+    },
+    {
+      "epoch": 13.145374449339206,
+      "grad_norm": 0.24767205069822415,
+      "learning_rate": 0.0001725221498101983,
+      "loss": 0.1636,
+      "step": 1865
+    },
+    {
+      "epoch": 13.152422907488987,
+      "grad_norm": 0.26692602859064013,
+      "learning_rate": 0.00017249050957978676,
+      "loss": 0.1613,
+      "step": 1866
+    },
+    {
+      "epoch": 13.159471365638767,
+      "grad_norm": 0.25085346300823014,
+      "learning_rate": 0.0001724588540482737,
+      "loss": 0.1647,
+      "step": 1867
+    },
+    {
+      "epoch": 13.166519823788546,
+      "grad_norm": 0.2544261403246702,
+      "learning_rate": 0.00017242718322234084,
+      "loss": 0.1696,
+      "step": 1868
+    },
+    {
+      "epoch": 13.173568281938326,
+      "grad_norm": 0.25646010156868165,
+      "learning_rate": 0.0001723954971086732,
+      "loss": 0.1761,
+      "step": 1869
+    },
+    {
+      "epoch": 13.180616740088105,
+      "grad_norm": 0.25350433090313884,
+      "learning_rate": 0.00017236379571395895,
+      "loss": 0.1746,
+      "step": 1870
+    },
+    {
+      "epoch": 13.187665198237886,
+      "grad_norm": 0.2580881600694854,
+      "learning_rate": 0.0001723320790448896,
+      "loss": 0.174,
+      "step": 1871
+    },
+    {
+      "epoch": 13.194713656387664,
+      "grad_norm": 0.2401486263028686,
+      "learning_rate": 0.00017230034710815977,
+      "loss": 0.18,
+      "step": 1872
+    },
+    {
+      "epoch": 13.201762114537445,
+      "grad_norm": 0.27640596357084096,
+      "learning_rate": 0.0001722685999104674,
+      "loss": 0.1724,
+      "step": 1873
+    },
+    {
+      "epoch": 13.208810572687225,
+      "grad_norm": 0.23996367560837828,
+      "learning_rate": 0.00017223683745851357,
+      "loss": 0.1724,
+      "step": 1874
+    },
+    {
+      "epoch": 13.215859030837004,
+      "grad_norm": 0.2376875128811146,
+      "learning_rate": 0.00017220505975900258,
+      "loss": 0.1766,
+      "step": 1875
+    },
+    {
+      "epoch": 13.222907488986785,
+      "grad_norm": 0.24865776974471118,
+      "learning_rate": 0.00017217326681864207,
+      "loss": 0.1725,
+      "step": 1876
+    },
+    {
+      "epoch": 13.229955947136563,
+      "grad_norm": 0.22603208988160914,
+      "learning_rate": 0.00017214145864414272,
+      "loss": 0.1711,
+      "step": 1877
+    },
+    {
+      "epoch": 13.237004405286344,
+      "grad_norm": 0.25619014116095185,
+      "learning_rate": 0.00017210963524221854,
+      "loss": 0.172,
+      "step": 1878
+    },
+    {
+      "epoch": 13.244052863436123,
+      "grad_norm": 0.25395992049523364,
+      "learning_rate": 0.00017207779661958678,
+      "loss": 0.175,
+      "step": 1879
+    },
+    {
+      "epoch": 13.251101321585903,
+      "grad_norm": 0.24913072021254887,
+      "learning_rate": 0.0001720459427829678,
+      "loss": 0.1789,
+      "step": 1880
+    },
+    {
+      "epoch": 13.258149779735684,
+      "grad_norm": 0.2459066406539119,
+      "learning_rate": 0.00017201407373908521,
+      "loss": 0.1828,
+      "step": 1881
+    },
+    {
+      "epoch": 13.265198237885462,
+      "grad_norm": 0.2406452643820562,
+      "learning_rate": 0.00017198218949466586,
+      "loss": 0.1728,
+      "step": 1882
+    },
+    {
+      "epoch": 13.272246696035243,
+      "grad_norm": 0.27508218030379084,
+      "learning_rate": 0.0001719502900564398,
+      "loss": 0.1844,
+      "step": 1883
+    },
+    {
+      "epoch": 13.279295154185021,
+      "grad_norm": 0.2548025184337696,
+      "learning_rate": 0.0001719183754311403,
+      "loss": 0.1753,
+      "step": 1884
+    },
+    {
+      "epoch": 13.286343612334802,
+      "grad_norm": 0.24361665687005532,
+      "learning_rate": 0.00017188644562550373,
+      "loss": 0.1731,
+      "step": 1885
+    },
+    {
+      "epoch": 13.29339207048458,
+      "grad_norm": 0.2519544769518712,
+      "learning_rate": 0.00017185450064626984,
+      "loss": 0.1789,
+      "step": 1886
+    },
+    {
+      "epoch": 13.300440528634361,
+      "grad_norm": 0.2548805788182817,
+      "learning_rate": 0.00017182254050018145,
+      "loss": 0.1786,
+      "step": 1887
+    },
+    {
+      "epoch": 13.307488986784142,
+      "grad_norm": 0.2537624338744124,
+      "learning_rate": 0.0001717905651939846,
+      "loss": 0.167,
+      "step": 1888
+    },
+    {
+      "epoch": 13.31453744493392,
+      "grad_norm": 0.2620578039861077,
+      "learning_rate": 0.00017175857473442863,
+      "loss": 0.177,
+      "step": 1889
+    },
+    {
+      "epoch": 13.321585903083701,
+      "grad_norm": 0.25812490816763434,
+      "learning_rate": 0.0001717265691282659,
+      "loss": 0.1775,
+      "step": 1890
+    },
+    {
+      "epoch": 13.32863436123348,
+      "grad_norm": 0.25982386050106054,
+      "learning_rate": 0.0001716945483822521,
+      "loss": 0.1841,
+      "step": 1891
+    },
+    {
+      "epoch": 13.33568281938326,
+      "grad_norm": 0.24235781154419087,
+      "learning_rate": 0.00017166251250314612,
+      "loss": 0.1843,
+      "step": 1892
+    },
+    {
+      "epoch": 13.342731277533039,
+      "grad_norm": 0.2736216585064902,
+      "learning_rate": 0.00017163046149770996,
+      "loss": 0.1869,
+      "step": 1893
+    },
+    {
+      "epoch": 13.34977973568282,
+      "grad_norm": 0.25085166589625196,
+      "learning_rate": 0.0001715983953727089,
+      "loss": 0.1807,
+      "step": 1894
+    },
+    {
+      "epoch": 13.3568281938326,
+      "grad_norm": 0.24488244962060554,
+      "learning_rate": 0.00017156631413491136,
+      "loss": 0.1807,
+      "step": 1895
+    },
+    {
+      "epoch": 13.363876651982379,
+      "grad_norm": 0.26013568141209803,
+      "learning_rate": 0.00017153421779108894,
+      "loss": 0.18,
+      "step": 1896
+    },
+    {
+      "epoch": 13.37092511013216,
+      "grad_norm": 0.2455538405095825,
+      "learning_rate": 0.0001715021063480165,
+      "loss": 0.183,
+      "step": 1897
+    },
+    {
+      "epoch": 13.377973568281938,
+      "grad_norm": 0.271902842149491,
+      "learning_rate": 0.00017146997981247196,
+      "loss": 0.1807,
+      "step": 1898
+    },
+    {
+      "epoch": 13.385022026431718,
+      "grad_norm": 0.2519853106291974,
+      "learning_rate": 0.00017143783819123653,
+      "loss": 0.1801,
+      "step": 1899
+    },
+    {
+      "epoch": 13.392070484581497,
+      "grad_norm": 0.2508025582969974,
+      "learning_rate": 0.0001714056814910946,
+      "loss": 0.1768,
+      "step": 1900
+    },
+    {
+      "epoch": 13.399118942731278,
+      "grad_norm": 0.27464929455144504,
+      "learning_rate": 0.00017137350971883374,
+      "loss": 0.1701,
+      "step": 1901
+    },
+    {
+      "epoch": 13.406167400881058,
+      "grad_norm": 0.2622256248327,
+      "learning_rate": 0.00017134132288124465,
+      "loss": 0.1872,
+      "step": 1902
+    },
+    {
+      "epoch": 13.413215859030837,
+      "grad_norm": 0.2465850749916401,
+      "learning_rate": 0.00017130912098512125,
+      "loss": 0.1806,
+      "step": 1903
+    },
+    {
+      "epoch": 13.420264317180617,
+      "grad_norm": 0.2580856524149855,
+      "learning_rate": 0.00017127690403726054,
+      "loss": 0.1754,
+      "step": 1904
+    },
+    {
+      "epoch": 13.427312775330396,
+      "grad_norm": 0.26352030921833347,
+      "learning_rate": 0.00017124467204446297,
+      "loss": 0.1896,
+      "step": 1905
+    },
+    {
+      "epoch": 13.434361233480177,
+      "grad_norm": 0.25531833209166416,
+      "learning_rate": 0.00017121242501353184,
+      "loss": 0.1795,
+      "step": 1906
+    },
+    {
+      "epoch": 13.441409691629955,
+      "grad_norm": 0.27490961373323425,
+      "learning_rate": 0.00017118016295127378,
+      "loss": 0.1888,
+      "step": 1907
+    },
+    {
+      "epoch": 13.448458149779736,
+      "grad_norm": 0.2351651767437859,
+      "learning_rate": 0.00017114788586449867,
+      "loss": 0.1746,
+      "step": 1908
+    },
+    {
+      "epoch": 13.455506607929516,
+      "grad_norm": 0.25073264846105514,
+      "learning_rate": 0.00017111559376001935,
+      "loss": 0.1851,
+      "step": 1909
+    },
+    {
+      "epoch": 13.462555066079295,
+      "grad_norm": 0.23690828354845034,
+      "learning_rate": 0.00017108328664465204,
+      "loss": 0.1837,
+      "step": 1910
+    },
+    {
+      "epoch": 13.469603524229075,
+      "grad_norm": 0.24682158676280033,
+      "learning_rate": 0.00017105096452521597,
+      "loss": 0.1828,
+      "step": 1911
+    },
+    {
+      "epoch": 13.476651982378854,
+      "grad_norm": 0.25482594031942296,
+      "learning_rate": 0.0001710186274085337,
+      "loss": 0.1936,
+      "step": 1912
+    },
+    {
+      "epoch": 13.483700440528635,
+      "grad_norm": 0.26242295224030066,
+      "learning_rate": 0.00017098627530143077,
+      "loss": 0.1761,
+      "step": 1913
+    },
+    {
+      "epoch": 13.490748898678413,
+      "grad_norm": 0.25235969463697316,
+      "learning_rate": 0.000170953908210736,
+      "loss": 0.1846,
+      "step": 1914
+    },
+    {
+      "epoch": 13.497797356828194,
+      "grad_norm": 0.24226724616227538,
+      "learning_rate": 0.00017092152614328135,
+      "loss": 0.1753,
+      "step": 1915
+    },
+    {
+      "epoch": 13.504845814977973,
+      "grad_norm": 0.26863914296113295,
+      "learning_rate": 0.000170889129105902,
+      "loss": 0.1958,
+      "step": 1916
+    },
+    {
+      "epoch": 13.511894273127753,
+      "grad_norm": 0.24137093350349406,
+      "learning_rate": 0.00017085671710543614,
+      "loss": 0.1897,
+      "step": 1917
+    },
+    {
+      "epoch": 13.518942731277534,
+      "grad_norm": 0.26280365383000504,
+      "learning_rate": 0.00017082429014872523,
+      "loss": 0.1917,
+      "step": 1918
+    },
+    {
+      "epoch": 13.525991189427312,
+      "grad_norm": 0.2397741167921101,
+      "learning_rate": 0.0001707918482426139,
+      "loss": 0.185,
+      "step": 1919
+    },
+    {
+      "epoch": 13.533039647577093,
+      "grad_norm": 0.26204577056651385,
+      "learning_rate": 0.00017075939139394987,
+      "loss": 0.1892,
+      "step": 1920
+    },
+    {
+      "epoch": 13.540088105726872,
+      "grad_norm": 0.24389298316617403,
+      "learning_rate": 0.00017072691960958407,
+      "loss": 0.1826,
+      "step": 1921
+    },
+    {
+      "epoch": 13.547136563876652,
+      "grad_norm": 0.2599660040493307,
+      "learning_rate": 0.0001706944328963705,
+      "loss": 0.1781,
+      "step": 1922
+    },
+    {
+      "epoch": 13.55418502202643,
+      "grad_norm": 0.2369380265850061,
+      "learning_rate": 0.00017066193126116645,
+      "loss": 0.1887,
+      "step": 1923
+    },
+    {
+      "epoch": 13.561233480176211,
+      "grad_norm": 0.24358046861467034,
+      "learning_rate": 0.0001706294147108322,
+      "loss": 0.1706,
+      "step": 1924
+    },
+    {
+      "epoch": 13.568281938325992,
+      "grad_norm": 0.24991325788152605,
+      "learning_rate": 0.00017059688325223138,
+      "loss": 0.1756,
+      "step": 1925
+    },
+    {
+      "epoch": 13.57533039647577,
+      "grad_norm": 0.25271065502359463,
+      "learning_rate": 0.00017056433689223045,
+      "loss": 0.1867,
+      "step": 1926
+    },
+    {
+      "epoch": 13.582378854625551,
+      "grad_norm": 0.272542973049773,
+      "learning_rate": 0.00017053177563769938,
+      "loss": 0.1833,
+      "step": 1927
+    },
+    {
+      "epoch": 13.58942731277533,
+      "grad_norm": 0.2529979514891331,
+      "learning_rate": 0.00017049919949551102,
+      "loss": 0.1827,
+      "step": 1928
+    },
+    {
+      "epoch": 13.59647577092511,
+      "grad_norm": 0.2635036985668748,
+      "learning_rate": 0.00017046660847254148,
+      "loss": 0.1846,
+      "step": 1929
+    },
+    {
+      "epoch": 13.603524229074889,
+      "grad_norm": 0.22906057994016582,
+      "learning_rate": 0.00017043400257566997,
+      "loss": 0.1758,
+      "step": 1930
+    },
+    {
+      "epoch": 13.61057268722467,
+      "grad_norm": 0.24134362771819662,
+      "learning_rate": 0.0001704013818117789,
+      "loss": 0.1811,
+      "step": 1931
+    },
+    {
+      "epoch": 13.61762114537445,
+      "grad_norm": 0.2805185993228023,
+      "learning_rate": 0.0001703687461877537,
+      "loss": 0.1946,
+      "step": 1932
+    },
+    {
+      "epoch": 13.624669603524229,
+      "grad_norm": 0.24202656191712912,
+      "learning_rate": 0.00017033609571048307,
+      "loss": 0.1861,
+      "step": 1933
+    },
+    {
+      "epoch": 13.63171806167401,
+      "grad_norm": 0.24687196088408955,
+      "learning_rate": 0.0001703034303868588,
+      "loss": 0.1875,
+      "step": 1934
+    },
+    {
+      "epoch": 13.638766519823788,
+      "grad_norm": 0.25631132633344667,
+      "learning_rate": 0.00017027075022377572,
+      "loss": 0.1916,
+      "step": 1935
+    },
+    {
+      "epoch": 13.645814977973568,
+      "grad_norm": 0.23014512475904084,
+      "learning_rate": 0.00017023805522813193,
+      "loss": 0.1952,
+      "step": 1936
+    },
+    {
+      "epoch": 13.652863436123347,
+      "grad_norm": 0.261101427910066,
+      "learning_rate": 0.0001702053454068286,
+      "loss": 0.1819,
+      "step": 1937
+    },
+    {
+      "epoch": 13.659911894273128,
+      "grad_norm": 0.23633528575767013,
+      "learning_rate": 0.00017017262076676997,
+      "loss": 0.1809,
+      "step": 1938
+    },
+    {
+      "epoch": 13.666960352422908,
+      "grad_norm": 0.2721258426763472,
+      "learning_rate": 0.00017013988131486357,
+      "loss": 0.1862,
+      "step": 1939
+    },
+    {
+      "epoch": 13.674008810572687,
+      "grad_norm": 0.24770488145896763,
+      "learning_rate": 0.0001701071270580199,
+      "loss": 0.1897,
+      "step": 1940
+    },
+    {
+      "epoch": 13.681057268722467,
+      "grad_norm": 0.2565727960726705,
+      "learning_rate": 0.00017007435800315263,
+      "loss": 0.1789,
+      "step": 1941
+    },
+    {
+      "epoch": 13.688105726872246,
+      "grad_norm": 0.25473735172953466,
+      "learning_rate": 0.0001700415741571786,
+      "loss": 0.1858,
+      "step": 1942
+    },
+    {
+      "epoch": 13.695154185022027,
+      "grad_norm": 0.23902684074447664,
+      "learning_rate": 0.0001700087755270177,
+      "loss": 0.1891,
+      "step": 1943
+    },
+    {
+      "epoch": 13.702202643171805,
+      "grad_norm": 0.2726961770605814,
+      "learning_rate": 0.00016997596211959298,
+      "loss": 0.1907,
+      "step": 1944
+    },
+    {
+      "epoch": 13.709251101321586,
+      "grad_norm": 0.23384311150731832,
+      "learning_rate": 0.0001699431339418306,
+      "loss": 0.1969,
+      "step": 1945
+    },
+    {
+      "epoch": 13.716299559471366,
+      "grad_norm": 0.24600176870482415,
+      "learning_rate": 0.0001699102910006599,
+      "loss": 0.1953,
+      "step": 1946
+    },
+    {
+      "epoch": 13.723348017621145,
+      "grad_norm": 0.24716669612927195,
+      "learning_rate": 0.00016987743330301318,
+      "loss": 0.1857,
+      "step": 1947
+    },
+    {
+      "epoch": 13.730396475770926,
+      "grad_norm": 0.24992097749243347,
+      "learning_rate": 0.00016984456085582605,
+      "loss": 0.1933,
+      "step": 1948
+    },
+    {
+      "epoch": 13.737444933920704,
+      "grad_norm": 0.24100703925437253,
+      "learning_rate": 0.0001698116736660371,
+      "loss": 0.185,
+      "step": 1949
+    },
+    {
+      "epoch": 13.744493392070485,
+      "grad_norm": 0.27241851197629346,
+      "learning_rate": 0.00016977877174058806,
+      "loss": 0.1854,
+      "step": 1950
+    },
+    {
+      "epoch": 13.751541850220264,
+      "grad_norm": 0.24315265287997906,
+      "learning_rate": 0.0001697458550864238,
+      "loss": 0.1958,
+      "step": 1951
+    },
+    {
+      "epoch": 13.758590308370044,
+      "grad_norm": 0.2615876841544895,
+      "learning_rate": 0.00016971292371049227,
+      "loss": 0.1851,
+      "step": 1952
+    },
+    {
+      "epoch": 13.765638766519825,
+      "grad_norm": 0.27153721442740575,
+      "learning_rate": 0.00016967997761974454,
+      "loss": 0.1942,
+      "step": 1953
+    },
+    {
+      "epoch": 13.772687224669603,
+      "grad_norm": 0.24809537345777846,
+      "learning_rate": 0.00016964701682113475,
+      "loss": 0.1854,
+      "step": 1954
+    },
+    {
+      "epoch": 13.779735682819384,
+      "grad_norm": 0.2577862320985244,
+      "learning_rate": 0.00016961404132162023,
+      "loss": 0.1837,
+      "step": 1955
+    },
+    {
+      "epoch": 13.786784140969162,
+      "grad_norm": 0.25412276950329965,
+      "learning_rate": 0.0001695810511281613,
+      "loss": 0.1879,
+      "step": 1956
+    },
+    {
+      "epoch": 13.793832599118943,
+      "grad_norm": 0.2504956296821692,
+      "learning_rate": 0.00016954804624772152,
+      "loss": 0.1937,
+      "step": 1957
+    },
+    {
+      "epoch": 13.800881057268722,
+      "grad_norm": 0.26096398199546783,
+      "learning_rate": 0.00016951502668726738,
+      "loss": 0.1835,
+      "step": 1958
+    },
+    {
+      "epoch": 13.807929515418502,
+      "grad_norm": 0.24088856908024153,
+      "learning_rate": 0.00016948199245376863,
+      "loss": 0.1858,
+      "step": 1959
+    },
+    {
+      "epoch": 13.814977973568283,
+      "grad_norm": 0.23801444769544464,
+      "learning_rate": 0.000169448943554198,
+      "loss": 0.1921,
+      "step": 1960
+    },
+    {
+      "epoch": 13.822026431718061,
+      "grad_norm": 0.2566716009822654,
+      "learning_rate": 0.0001694158799955314,
+      "loss": 0.19,
+      "step": 1961
+    },
+    {
+      "epoch": 13.829074889867842,
+      "grad_norm": 0.25099676074756055,
+      "learning_rate": 0.0001693828017847478,
+      "loss": 0.1949,
+      "step": 1962
+    },
+    {
+      "epoch": 13.83612334801762,
+      "grad_norm": 0.2570138584730345,
+      "learning_rate": 0.00016934970892882922,
+      "loss": 0.1954,
+      "step": 1963
+    },
+    {
+      "epoch": 13.843171806167401,
+      "grad_norm": 0.25464726817836114,
+      "learning_rate": 0.00016931660143476083,
+      "loss": 0.1894,
+      "step": 1964
+    },
+    {
+      "epoch": 13.85022026431718,
+      "grad_norm": 0.25368458750677225,
+      "learning_rate": 0.00016928347930953087,
+      "loss": 0.184,
+      "step": 1965
+    },
+    {
+      "epoch": 13.85726872246696,
+      "grad_norm": 0.22235184910641434,
+      "learning_rate": 0.0001692503425601307,
+      "loss": 0.1836,
+      "step": 1966
+    },
+    {
+      "epoch": 13.864317180616741,
+      "grad_norm": 0.24759771137301775,
+      "learning_rate": 0.00016921719119355468,
+      "loss": 0.1873,
+      "step": 1967
+    },
+    {
+      "epoch": 13.87136563876652,
+      "grad_norm": 0.24149581805431775,
+      "learning_rate": 0.00016918402521680034,
+      "loss": 0.1912,
+      "step": 1968
+    },
+    {
+      "epoch": 13.8784140969163,
+      "grad_norm": 0.24173305865566486,
+      "learning_rate": 0.00016915084463686828,
+      "loss": 0.1896,
+      "step": 1969
+    },
+    {
+      "epoch": 13.885462555066079,
+      "grad_norm": 0.25157496277442293,
+      "learning_rate": 0.00016911764946076215,
+      "loss": 0.1917,
+      "step": 1970
+    },
+    {
+      "epoch": 13.89251101321586,
+      "grad_norm": 0.25816276997750265,
+      "learning_rate": 0.0001690844396954887,
+      "loss": 0.1843,
+      "step": 1971
+    },
+    {
+      "epoch": 13.899559471365638,
+      "grad_norm": 0.23384976820563294,
+      "learning_rate": 0.00016905121534805772,
+      "loss": 0.1949,
+      "step": 1972
+    },
+    {
+      "epoch": 13.906607929515419,
+      "grad_norm": 0.23892706128069213,
+      "learning_rate": 0.0001690179764254822,
+      "loss": 0.1997,
+      "step": 1973
+    },
+    {
+      "epoch": 13.913656387665199,
+      "grad_norm": 0.25773336949528497,
+      "learning_rate": 0.00016898472293477807,
+      "loss": 0.1941,
+      "step": 1974
+    },
+    {
+      "epoch": 13.920704845814978,
+      "grad_norm": 0.23672218590267055,
+      "learning_rate": 0.0001689514548829644,
+      "loss": 0.1923,
+      "step": 1975
+    },
+    {
+      "epoch": 13.927753303964758,
+      "grad_norm": 0.24051055548588446,
+      "learning_rate": 0.00016891817227706327,
+      "loss": 0.1935,
+      "step": 1976
+    },
+    {
+      "epoch": 13.934801762114537,
+      "grad_norm": 0.25712824057343536,
+      "learning_rate": 0.00016888487512409993,
+      "loss": 0.1891,
+      "step": 1977
+    },
+    {
+      "epoch": 13.941850220264318,
+      "grad_norm": 0.26309044710336227,
+      "learning_rate": 0.00016885156343110272,
+      "loss": 0.1909,
+      "step": 1978
+    },
+    {
+      "epoch": 13.948898678414096,
+      "grad_norm": 0.23921962166835822,
+      "learning_rate": 0.00016881823720510283,
+      "loss": 0.2072,
+      "step": 1979
+    },
+    {
+      "epoch": 13.955947136563877,
+      "grad_norm": 0.24166509243330933,
+      "learning_rate": 0.0001687848964531348,
+      "loss": 0.1894,
+      "step": 1980
+    },
+    {
+      "epoch": 13.962995594713657,
+      "grad_norm": 0.2286995117026473,
+      "learning_rate": 0.00016875154118223607,
+      "loss": 0.195,
+      "step": 1981
+    },
+    {
+      "epoch": 13.970044052863436,
+      "grad_norm": 0.23627754027892553,
+      "learning_rate": 0.00016871817139944712,
+      "loss": 0.199,
+      "step": 1982
+    },
+    {
+      "epoch": 13.977092511013216,
+      "grad_norm": 0.24779293299193506,
+      "learning_rate": 0.00016868478711181163,
+      "loss": 0.1985,
+      "step": 1983
+    },
+    {
+      "epoch": 13.984140969162995,
+      "grad_norm": 0.23231353569781177,
+      "learning_rate": 0.00016865138832637623,
+      "loss": 0.1865,
+      "step": 1984
+    },
+    {
+      "epoch": 13.991189427312776,
+      "grad_norm": 0.24989977280504946,
+      "learning_rate": 0.00016861797505019068,
+      "loss": 0.1934,
+      "step": 1985
+    },
+    {
+      "epoch": 13.998237885462554,
+      "grad_norm": 0.24908422577006684,
+      "learning_rate": 0.00016858454729030772,
+      "loss": 0.1888,
+      "step": 1986
+    },
+    {
+      "epoch": 14.005286343612335,
+      "grad_norm": 0.24334471177631506,
+      "learning_rate": 0.0001685511050537832,
+      "loss": 0.1523,
+      "step": 1987
+    },
+    {
+      "epoch": 14.012334801762114,
+      "grad_norm": 0.2214078395209683,
+      "learning_rate": 0.00016851764834767606,
+      "loss": 0.1318,
+      "step": 1988
+    },
+    {
+      "epoch": 14.019383259911894,
+      "grad_norm": 0.29150352766790755,
+      "learning_rate": 0.00016848417717904817,
+      "loss": 0.1415,
+      "step": 1989
+    },
+    {
+      "epoch": 14.026431718061675,
+      "grad_norm": 0.2820428978147202,
+      "learning_rate": 0.00016845069155496464,
+      "loss": 0.1407,
+      "step": 1990
+    },
+    {
+      "epoch": 14.033480176211453,
+      "grad_norm": 0.2793958997342532,
+      "learning_rate": 0.00016841719148249343,
+      "loss": 0.1397,
+      "step": 1991
+    },
+    {
+      "epoch": 14.040528634361234,
+      "grad_norm": 0.25945640241599505,
+      "learning_rate": 0.0001683836769687057,
+      "loss": 0.1336,
+      "step": 1992
+    },
+    {
+      "epoch": 14.047577092511013,
+      "grad_norm": 0.23307956323358936,
+      "learning_rate": 0.00016835014802067558,
+      "loss": 0.1346,
+      "step": 1993
+    },
+    {
+      "epoch": 14.054625550660793,
+      "grad_norm": 0.2755825538281474,
+      "learning_rate": 0.00016831660464548026,
+      "loss": 0.1387,
+      "step": 1994
+    },
+    {
+      "epoch": 14.061674008810572,
+      "grad_norm": 0.2542595215143745,
+      "learning_rate": 0.00016828304685020004,
+      "loss": 0.1471,
+      "step": 1995
+    },
+    {
+      "epoch": 14.068722466960352,
+      "grad_norm": 0.27247993223665573,
+      "learning_rate": 0.00016824947464191814,
+      "loss": 0.144,
+      "step": 1996
+    },
+    {
+      "epoch": 14.075770925110133,
+      "grad_norm": 0.2666435653736045,
+      "learning_rate": 0.00016821588802772093,
+      "loss": 0.1404,
+      "step": 1997
+    },
+    {
+      "epoch": 14.082819383259912,
+      "grad_norm": 0.26760025775929225,
+      "learning_rate": 0.0001681822870146978,
+      "loss": 0.1383,
+      "step": 1998
+    },
+    {
+      "epoch": 14.089867841409692,
+      "grad_norm": 0.25046619881264137,
+      "learning_rate": 0.0001681486716099411,
+      "loss": 0.1332,
+      "step": 1999
+    },
+    {
+      "epoch": 14.09691629955947,
+      "grad_norm": 0.24931371237515917,
+      "learning_rate": 0.00016811504182054633,
+      "loss": 0.1423,
+      "step": 2000
+    },
+    {
+      "epoch": 14.103964757709251,
+      "grad_norm": 0.2394958621048712,
+      "learning_rate": 0.00016808139765361196,
+      "loss": 0.1352,
+      "step": 2001
+    },
+    {
+      "epoch": 14.11101321585903,
+      "grad_norm": 0.25745874850047523,
+      "learning_rate": 0.0001680477391162395,
+      "loss": 0.137,
+      "step": 2002
+    },
+    {
+      "epoch": 14.11806167400881,
+      "grad_norm": 0.2850304764664298,
+      "learning_rate": 0.00016801406621553354,
+      "loss": 0.1422,
+      "step": 2003
+    },
+    {
+      "epoch": 14.125110132158591,
+      "grad_norm": 0.26339941604439415,
+      "learning_rate": 0.00016798037895860164,
+      "loss": 0.1395,
+      "step": 2004
+    },
+    {
+      "epoch": 14.13215859030837,
+      "grad_norm": 0.27033784489322615,
+      "learning_rate": 0.00016794667735255437,
+      "loss": 0.1378,
+      "step": 2005
+    },
+    {
+      "epoch": 14.13920704845815,
+      "grad_norm": 0.2515045325968817,
+      "learning_rate": 0.00016791296140450545,
+      "loss": 0.143,
+      "step": 2006
+    },
+    {
+      "epoch": 14.146255506607929,
+      "grad_norm": 0.24062437503608186,
+      "learning_rate": 0.00016787923112157154,
+      "loss": 0.1383,
+      "step": 2007
+    },
+    {
+      "epoch": 14.15330396475771,
+      "grad_norm": 0.23314167086311985,
+      "learning_rate": 0.0001678454865108723,
+      "loss": 0.1471,
+      "step": 2008
+    },
+    {
+      "epoch": 14.160352422907488,
+      "grad_norm": 0.25100008169414206,
+      "learning_rate": 0.00016781172757953047,
+      "loss": 0.1481,
+      "step": 2009
+    },
+    {
+      "epoch": 14.167400881057269,
+      "grad_norm": 0.26341487035212163,
+      "learning_rate": 0.0001677779543346718,
+      "loss": 0.1505,
+      "step": 2010
+    },
+    {
+      "epoch": 14.17444933920705,
+      "grad_norm": 0.2579504541068952,
+      "learning_rate": 0.00016774416678342501,
+      "loss": 0.1372,
+      "step": 2011
+    },
+    {
+      "epoch": 14.181497797356828,
+      "grad_norm": 0.2593001431921589,
+      "learning_rate": 0.000167710364932922,
+      "loss": 0.1315,
+      "step": 2012
+    },
+    {
+      "epoch": 14.188546255506608,
+      "grad_norm": 0.24264481752496386,
+      "learning_rate": 0.00016767654879029742,
+      "loss": 0.1438,
+      "step": 2013
+    },
+    {
+      "epoch": 14.195594713656387,
+      "grad_norm": 0.2564625038427879,
+      "learning_rate": 0.0001676427183626892,
+      "loss": 0.1445,
+      "step": 2014
+    },
+    {
+      "epoch": 14.202643171806168,
+      "grad_norm": 0.24370335155820755,
+      "learning_rate": 0.00016760887365723813,
+      "loss": 0.144,
+      "step": 2015
+    },
+    {
+      "epoch": 14.209691629955946,
+      "grad_norm": 0.2467198915766038,
+      "learning_rate": 0.00016757501468108813,
+      "loss": 0.1511,
+      "step": 2016
+    },
+    {
+      "epoch": 14.216740088105727,
+      "grad_norm": 0.2591631744296675,
+      "learning_rate": 0.00016754114144138598,
+      "loss": 0.1429,
+      "step": 2017
+    },
+    {
+      "epoch": 14.223788546255507,
+      "grad_norm": 0.2424494644709448,
+      "learning_rate": 0.00016750725394528156,
+      "loss": 0.1498,
+      "step": 2018
+    },
+    {
+      "epoch": 14.230837004405286,
+      "grad_norm": 0.2644773242400212,
+      "learning_rate": 0.00016747335219992774,
+      "loss": 0.1432,
+      "step": 2019
+    },
+    {
+      "epoch": 14.237885462555067,
+      "grad_norm": 0.23648439906765395,
+      "learning_rate": 0.0001674394362124805,
+      "loss": 0.1433,
+      "step": 2020
+    },
+    {
+      "epoch": 14.244933920704845,
+      "grad_norm": 0.2566015353469309,
+      "learning_rate": 0.00016740550599009864,
+      "loss": 0.1479,
+      "step": 2021
+    },
+    {
+      "epoch": 14.251982378854626,
+      "grad_norm": 0.2550565042743828,
+      "learning_rate": 0.00016737156153994408,
+      "loss": 0.141,
+      "step": 2022
+    },
+    {
+      "epoch": 14.259030837004405,
+      "grad_norm": 0.2577209642312449,
+      "learning_rate": 0.00016733760286918176,
+      "loss": 0.141,
+      "step": 2023
+    },
+    {
+      "epoch": 14.266079295154185,
+      "grad_norm": 0.2424142004094277,
+      "learning_rate": 0.00016730362998497954,
+      "loss": 0.1441,
+      "step": 2024
+    },
+    {
+      "epoch": 14.273127753303966,
+      "grad_norm": 0.2560608651881838,
+      "learning_rate": 0.00016726964289450838,
+      "loss": 0.1435,
+      "step": 2025
+    },
+    {
+      "epoch": 14.280176211453744,
+      "grad_norm": 0.2558108638795868,
+      "learning_rate": 0.0001672356416049421,
+      "loss": 0.1445,
+      "step": 2026
+    },
+    {
+      "epoch": 14.287224669603525,
+      "grad_norm": 0.2690474511208858,
+      "learning_rate": 0.0001672016261234577,
+      "loss": 0.1413,
+      "step": 2027
+    },
+    {
+      "epoch": 14.294273127753303,
+      "grad_norm": 0.2611343816436336,
+      "learning_rate": 0.000167167596457235,
+      "loss": 0.1499,
+      "step": 2028
+    },
+    {
+      "epoch": 14.301321585903084,
+      "grad_norm": 0.2603449405383826,
+      "learning_rate": 0.0001671335526134569,
+      "loss": 0.1452,
+      "step": 2029
+    },
+    {
+      "epoch": 14.308370044052863,
+      "grad_norm": 0.25908348608579623,
+      "learning_rate": 0.00016709949459930932,
+      "loss": 0.1415,
+      "step": 2030
+    },
+    {
+      "epoch": 14.315418502202643,
+      "grad_norm": 0.27430682626371367,
+      "learning_rate": 0.00016706542242198112,
+      "loss": 0.144,
+      "step": 2031
+    },
+    {
+      "epoch": 14.322466960352424,
+      "grad_norm": 0.25439612754976026,
+      "learning_rate": 0.00016703133608866414,
+      "loss": 0.1427,
+      "step": 2032
+    },
+    {
+      "epoch": 14.329515418502202,
+      "grad_norm": 0.27269726062966465,
+      "learning_rate": 0.00016699723560655328,
+      "loss": 0.1432,
+      "step": 2033
+    },
+    {
+      "epoch": 14.336563876651983,
+      "grad_norm": 0.27383478220455126,
+      "learning_rate": 0.0001669631209828463,
+      "loss": 0.1497,
+      "step": 2034
+    },
+    {
+      "epoch": 14.343612334801762,
+      "grad_norm": 0.2632831400008714,
+      "learning_rate": 0.00016692899222474412,
+      "loss": 0.1491,
+      "step": 2035
+    },
+    {
+      "epoch": 14.350660792951542,
+      "grad_norm": 0.25688264966017627,
+      "learning_rate": 0.00016689484933945042,
+      "loss": 0.1516,
+      "step": 2036
+    },
+    {
+      "epoch": 14.357709251101321,
+      "grad_norm": 0.2434810312337111,
+      "learning_rate": 0.00016686069233417215,
+      "loss": 0.1409,
+      "step": 2037
+    },
+    {
+      "epoch": 14.364757709251101,
+      "grad_norm": 0.2503083627490826,
+      "learning_rate": 0.00016682652121611895,
+      "loss": 0.1448,
+      "step": 2038
+    },
+    {
+      "epoch": 14.371806167400882,
+      "grad_norm": 0.2486156086106601,
+      "learning_rate": 0.00016679233599250362,
+      "loss": 0.1516,
+      "step": 2039
+    },
+    {
+      "epoch": 14.37885462555066,
+      "grad_norm": 0.28212998446906523,
+      "learning_rate": 0.00016675813667054186,
+      "loss": 0.1427,
+      "step": 2040
+    },
+    {
+      "epoch": 14.385903083700441,
+      "grad_norm": 0.2644240443245315,
+      "learning_rate": 0.0001667239232574524,
+      "loss": 0.1511,
+      "step": 2041
+    },
+    {
+      "epoch": 14.39295154185022,
+      "grad_norm": 0.2570233186800215,
+      "learning_rate": 0.0001666896957604569,
+      "loss": 0.1448,
+      "step": 2042
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 0.267064936242631,
+      "learning_rate": 0.00016665545418678,
+      "loss": 0.1534,
+      "step": 2043
+    },
+    {
+      "epoch": 14.407048458149779,
+      "grad_norm": 0.27039600058127555,
+      "learning_rate": 0.00016662119854364934,
+      "loss": 0.1495,
+      "step": 2044
+    },
+    {
+      "epoch": 14.41409691629956,
+      "grad_norm": 0.25474494504957385,
+      "learning_rate": 0.00016658692883829548,
+      "loss": 0.1517,
+      "step": 2045
+    },
+    {
+      "epoch": 14.421145374449338,
+      "grad_norm": 0.25984327564310106,
+      "learning_rate": 0.00016655264507795198,
+      "loss": 0.1435,
+      "step": 2046
+    },
+    {
+      "epoch": 14.428193832599119,
+      "grad_norm": 0.2707377574552346,
+      "learning_rate": 0.0001665183472698554,
+      "loss": 0.1471,
+      "step": 2047
+    },
+    {
+      "epoch": 14.4352422907489,
+      "grad_norm": 0.2566126961607602,
+      "learning_rate": 0.00016648403542124523,
+      "loss": 0.1554,
+      "step": 2048
+    },
+    {
+      "epoch": 14.442290748898678,
+      "grad_norm": 0.26778707595802,
+      "learning_rate": 0.0001664497095393639,
+      "loss": 0.1524,
+      "step": 2049
+    },
+    {
+      "epoch": 14.449339207048459,
+      "grad_norm": 0.25138541371432754,
+      "learning_rate": 0.00016641536963145676,
+      "loss": 0.1539,
+      "step": 2050
+    },
+    {
+      "epoch": 14.456387665198237,
+      "grad_norm": 0.26305940072913614,
+      "learning_rate": 0.00016638101570477229,
+      "loss": 0.1593,
+      "step": 2051
+    },
+    {
+      "epoch": 14.463436123348018,
+      "grad_norm": 0.25635392552292247,
+      "learning_rate": 0.0001663466477665618,
+      "loss": 0.1487,
+      "step": 2052
+    },
+    {
+      "epoch": 14.470484581497796,
+      "grad_norm": 0.26689260365638856,
+      "learning_rate": 0.00016631226582407952,
+      "loss": 0.1502,
+      "step": 2053
+    },
+    {
+      "epoch": 14.477533039647577,
+      "grad_norm": 0.26476375716722333,
+      "learning_rate": 0.0001662778698845828,
+      "loss": 0.1481,
+      "step": 2054
+    },
+    {
+      "epoch": 14.484581497797357,
+      "grad_norm": 0.2700325846514639,
+      "learning_rate": 0.00016624345995533174,
+      "loss": 0.1486,
+      "step": 2055
+    },
+    {
+      "epoch": 14.491629955947136,
+      "grad_norm": 0.2518065199330781,
+      "learning_rate": 0.00016620903604358956,
+      "loss": 0.1593,
+      "step": 2056
+    },
+    {
+      "epoch": 14.498678414096917,
+      "grad_norm": 0.2642300671188653,
+      "learning_rate": 0.00016617459815662232,
+      "loss": 0.1518,
+      "step": 2057
+    },
+    {
+      "epoch": 14.505726872246695,
+      "grad_norm": 0.27498069243216156,
+      "learning_rate": 0.00016614014630169917,
+      "loss": 0.1576,
+      "step": 2058
+    },
+    {
+      "epoch": 14.512775330396476,
+      "grad_norm": 0.2595741760029623,
+      "learning_rate": 0.00016610568048609195,
+      "loss": 0.157,
+      "step": 2059
+    },
+    {
+      "epoch": 14.519823788546255,
+      "grad_norm": 0.2755235809601913,
+      "learning_rate": 0.0001660712007170758,
+      "loss": 0.1615,
+      "step": 2060
+    },
+    {
+      "epoch": 14.526872246696035,
+      "grad_norm": 0.2576825522351712,
+      "learning_rate": 0.0001660367070019285,
+      "loss": 0.1533,
+      "step": 2061
+    },
+    {
+      "epoch": 14.533920704845816,
+      "grad_norm": 0.2482427935172345,
+      "learning_rate": 0.0001660021993479309,
+      "loss": 0.1465,
+      "step": 2062
+    },
+    {
+      "epoch": 14.540969162995594,
+      "grad_norm": 0.24613079411150873,
+      "learning_rate": 0.0001659676777623668,
+      "loss": 0.1566,
+      "step": 2063
+    },
+    {
+      "epoch": 14.548017621145375,
+      "grad_norm": 0.25124386198821524,
+      "learning_rate": 0.00016593314225252294,
+      "loss": 0.153,
+      "step": 2064
+    },
+    {
+      "epoch": 14.555066079295154,
+      "grad_norm": 0.2775676266225665,
+      "learning_rate": 0.00016589859282568897,
+      "loss": 0.1517,
+      "step": 2065
+    },
+    {
+      "epoch": 14.562114537444934,
+      "grad_norm": 0.264175069049245,
+      "learning_rate": 0.0001658640294891575,
+      "loss": 0.1509,
+      "step": 2066
+    },
+    {
+      "epoch": 14.569162995594713,
+      "grad_norm": 0.26505163047744035,
+      "learning_rate": 0.000165829452250224,
+      "loss": 0.1579,
+      "step": 2067
+    },
+    {
+      "epoch": 14.576211453744493,
+      "grad_norm": 0.2505824087923507,
+      "learning_rate": 0.00016579486111618702,
+      "loss": 0.1547,
+      "step": 2068
+    },
+    {
+      "epoch": 14.583259911894274,
+      "grad_norm": 0.2403471334540065,
+      "learning_rate": 0.00016576025609434794,
+      "loss": 0.1523,
+      "step": 2069
+    },
+    {
+      "epoch": 14.590308370044053,
+      "grad_norm": 0.27177346366673955,
+      "learning_rate": 0.00016572563719201107,
+      "loss": 0.1562,
+      "step": 2070
+    },
+    {
+      "epoch": 14.597356828193833,
+      "grad_norm": 0.28902605826615424,
+      "learning_rate": 0.00016569100441648374,
+      "loss": 0.1542,
+      "step": 2071
+    },
+    {
+      "epoch": 14.604405286343612,
+      "grad_norm": 0.2701603696906795,
+      "learning_rate": 0.00016565635777507603,
+      "loss": 0.1586,
+      "step": 2072
+    },
+    {
+      "epoch": 14.611453744493392,
+      "grad_norm": 0.26964381388708536,
+      "learning_rate": 0.0001656216972751012,
+      "loss": 0.156,
+      "step": 2073
+    },
+    {
+      "epoch": 14.618502202643171,
+      "grad_norm": 0.2548978659731898,
+      "learning_rate": 0.00016558702292387517,
+      "loss": 0.1569,
+      "step": 2074
+    },
+    {
+      "epoch": 14.625550660792952,
+      "grad_norm": 0.24942804552419603,
+      "learning_rate": 0.00016555233472871698,
+      "loss": 0.1584,
+      "step": 2075
+    },
+    {
+      "epoch": 14.632599118942732,
+      "grad_norm": 0.24889024869790968,
+      "learning_rate": 0.00016551763269694852,
+      "loss": 0.163,
+      "step": 2076
+    },
+    {
+      "epoch": 14.63964757709251,
+      "grad_norm": 0.26198084415313355,
+      "learning_rate": 0.00016548291683589456,
+      "loss": 0.1646,
+      "step": 2077
+    },
+    {
+      "epoch": 14.646696035242291,
+      "grad_norm": 0.2699703078136676,
+      "learning_rate": 0.00016544818715288288,
+      "loss": 0.155,
+      "step": 2078
+    },
+    {
+      "epoch": 14.65374449339207,
+      "grad_norm": 0.2696764424881082,
+      "learning_rate": 0.00016541344365524408,
+      "loss": 0.1551,
+      "step": 2079
+    },
+    {
+      "epoch": 14.66079295154185,
+      "grad_norm": 0.26436086433095474,
+      "learning_rate": 0.00016537868635031175,
+      "loss": 0.1562,
+      "step": 2080
+    },
+    {
+      "epoch": 14.66784140969163,
+      "grad_norm": 0.2690508170007313,
+      "learning_rate": 0.00016534391524542237,
+      "loss": 0.1598,
+      "step": 2081
+    },
+    {
+      "epoch": 14.67488986784141,
+      "grad_norm": 0.2524568476572526,
+      "learning_rate": 0.0001653091303479153,
+      "loss": 0.1566,
+      "step": 2082
+    },
+    {
+      "epoch": 14.68193832599119,
+      "grad_norm": 0.2858597945430157,
+      "learning_rate": 0.00016527433166513292,
+      "loss": 0.1601,
+      "step": 2083
+    },
+    {
+      "epoch": 14.688986784140969,
+      "grad_norm": 0.26687196666970525,
+      "learning_rate": 0.00016523951920442034,
+      "loss": 0.1675,
+      "step": 2084
+    },
+    {
+      "epoch": 14.69603524229075,
+      "grad_norm": 0.2720747610750674,
+      "learning_rate": 0.00016520469297312575,
+      "loss": 0.1623,
+      "step": 2085
+    },
+    {
+      "epoch": 14.703083700440528,
+      "grad_norm": 0.2556264385377757,
+      "learning_rate": 0.00016516985297860014,
+      "loss": 0.1596,
+      "step": 2086
+    },
+    {
+      "epoch": 14.710132158590309,
+      "grad_norm": 0.2663310849934029,
+      "learning_rate": 0.0001651349992281975,
+      "loss": 0.1642,
+      "step": 2087
+    },
+    {
+      "epoch": 14.717180616740087,
+      "grad_norm": 0.25860827891680954,
+      "learning_rate": 0.0001651001317292746,
+      "loss": 0.1614,
+      "step": 2088
+    },
+    {
+      "epoch": 14.724229074889868,
+      "grad_norm": 0.23944088626532148,
+      "learning_rate": 0.0001650652504891912,
+      "loss": 0.1571,
+      "step": 2089
+    },
+    {
+      "epoch": 14.731277533039648,
+      "grad_norm": 0.26492266245163776,
+      "learning_rate": 0.00016503035551530996,
+      "loss": 0.1517,
+      "step": 2090
+    },
+    {
+      "epoch": 14.738325991189427,
+      "grad_norm": 0.27530051313030507,
+      "learning_rate": 0.00016499544681499638,
+      "loss": 0.1606,
+      "step": 2091
+    },
+    {
+      "epoch": 14.745374449339208,
+      "grad_norm": 0.2648654633328283,
+      "learning_rate": 0.00016496052439561893,
+      "loss": 0.1525,
+      "step": 2092
+    },
+    {
+      "epoch": 14.752422907488986,
+      "grad_norm": 0.2878562731675984,
+      "learning_rate": 0.0001649255882645489,
+      "loss": 0.1573,
+      "step": 2093
+    },
+    {
+      "epoch": 14.759471365638767,
+      "grad_norm": 0.250780079961198,
+      "learning_rate": 0.0001648906384291606,
+      "loss": 0.1586,
+      "step": 2094
+    },
+    {
+      "epoch": 14.766519823788546,
+      "grad_norm": 0.2592173720930917,
+      "learning_rate": 0.00016485567489683101,
+      "loss": 0.1703,
+      "step": 2095
+    },
+    {
+      "epoch": 14.773568281938326,
+      "grad_norm": 0.2601121740177746,
+      "learning_rate": 0.00016482069767494024,
+      "loss": 0.1599,
+      "step": 2096
+    },
+    {
+      "epoch": 14.780616740088107,
+      "grad_norm": 0.2783954465518634,
+      "learning_rate": 0.00016478570677087116,
+      "loss": 0.1626,
+      "step": 2097
+    },
+    {
+      "epoch": 14.787665198237885,
+      "grad_norm": 0.265250678844388,
+      "learning_rate": 0.00016475070219200956,
+      "loss": 0.1542,
+      "step": 2098
+    },
+    {
+      "epoch": 14.794713656387666,
+      "grad_norm": 0.26353345374830306,
+      "learning_rate": 0.00016471568394574413,
+      "loss": 0.1642,
+      "step": 2099
+    },
+    {
+      "epoch": 14.801762114537445,
+      "grad_norm": 0.2724954603386319,
+      "learning_rate": 0.0001646806520394664,
+      "loss": 0.1648,
+      "step": 2100
+    },
+    {
+      "epoch": 14.808810572687225,
+      "grad_norm": 0.25243499856128,
+      "learning_rate": 0.0001646456064805708,
+      "loss": 0.157,
+      "step": 2101
+    },
+    {
+      "epoch": 14.815859030837004,
+      "grad_norm": 0.26446378608886084,
+      "learning_rate": 0.00016461054727645466,
+      "loss": 0.1654,
+      "step": 2102
+    },
+    {
+      "epoch": 14.822907488986784,
+      "grad_norm": 0.26060487219765055,
+      "learning_rate": 0.00016457547443451823,
+      "loss": 0.1616,
+      "step": 2103
+    },
+    {
+      "epoch": 14.829955947136565,
+      "grad_norm": 0.2537735239186563,
+      "learning_rate": 0.0001645403879621645,
+      "loss": 0.1599,
+      "step": 2104
+    },
+    {
+      "epoch": 14.837004405286343,
+      "grad_norm": 0.25940562268147954,
+      "learning_rate": 0.00016450528786679953,
+      "loss": 0.1614,
+      "step": 2105
+    },
+    {
+      "epoch": 14.844052863436124,
+      "grad_norm": 0.2556281840509093,
+      "learning_rate": 0.0001644701741558321,
+      "loss": 0.1597,
+      "step": 2106
+    },
+    {
+      "epoch": 14.851101321585903,
+      "grad_norm": 0.2577241046938242,
+      "learning_rate": 0.0001644350468366739,
+      "loss": 0.154,
+      "step": 2107
+    },
+    {
+      "epoch": 14.858149779735683,
+      "grad_norm": 0.2495668698078221,
+      "learning_rate": 0.00016439990591673954,
+      "loss": 0.1622,
+      "step": 2108
+    },
+    {
+      "epoch": 14.865198237885462,
+      "grad_norm": 0.27709653064014195,
+      "learning_rate": 0.00016436475140344645,
+      "loss": 0.1626,
+      "step": 2109
+    },
+    {
+      "epoch": 14.872246696035242,
+      "grad_norm": 0.27449838284436195,
+      "learning_rate": 0.00016432958330421497,
+      "loss": 0.162,
+      "step": 2110
+    },
+    {
+      "epoch": 14.879295154185023,
+      "grad_norm": 0.2601239352494764,
+      "learning_rate": 0.0001642944016264683,
+      "loss": 0.1692,
+      "step": 2111
+    },
+    {
+      "epoch": 14.886343612334802,
+      "grad_norm": 0.2917242497476059,
+      "learning_rate": 0.00016425920637763245,
+      "loss": 0.174,
+      "step": 2112
+    },
+    {
+      "epoch": 14.893392070484582,
+      "grad_norm": 0.24923932261388937,
+      "learning_rate": 0.0001642239975651364,
+      "loss": 0.1591,
+      "step": 2113
+    },
+    {
+      "epoch": 14.90044052863436,
+      "grad_norm": 0.27242579998858035,
+      "learning_rate": 0.00016418877519641182,
+      "loss": 0.1629,
+      "step": 2114
+    },
+    {
+      "epoch": 14.907488986784141,
+      "grad_norm": 0.2690075440082295,
+      "learning_rate": 0.00016415353927889349,
+      "loss": 0.1592,
+      "step": 2115
+    },
+    {
+      "epoch": 14.91453744493392,
+      "grad_norm": 0.26846519214109665,
+      "learning_rate": 0.00016411828982001883,
+      "loss": 0.1579,
+      "step": 2116
+    },
+    {
+      "epoch": 14.9215859030837,
+      "grad_norm": 0.2608764896759263,
+      "learning_rate": 0.00016408302682722822,
+      "loss": 0.1639,
+      "step": 2117
+    },
+    {
+      "epoch": 14.928634361233481,
+      "grad_norm": 0.27583367536847175,
+      "learning_rate": 0.00016404775030796488,
+      "loss": 0.1705,
+      "step": 2118
+    },
+    {
+      "epoch": 14.93568281938326,
+      "grad_norm": 0.26729467339859053,
+      "learning_rate": 0.00016401246026967486,
+      "loss": 0.1558,
+      "step": 2119
+    },
+    {
+      "epoch": 14.94273127753304,
+      "grad_norm": 0.2711003683822772,
+      "learning_rate": 0.00016397715671980713,
+      "loss": 0.1645,
+      "step": 2120
+    },
+    {
+      "epoch": 14.949779735682819,
+      "grad_norm": 0.2433426941847409,
+      "learning_rate": 0.00016394183966581344,
+      "loss": 0.1602,
+      "step": 2121
+    },
+    {
+      "epoch": 14.9568281938326,
+      "grad_norm": 0.27059793322344944,
+      "learning_rate": 0.0001639065091151484,
+      "loss": 0.1606,
+      "step": 2122
+    },
+    {
+      "epoch": 14.963876651982378,
+      "grad_norm": 0.25236119694217996,
+      "learning_rate": 0.00016387116507526957,
+      "loss": 0.1633,
+      "step": 2123
+    },
+    {
+      "epoch": 14.970925110132159,
+      "grad_norm": 0.2700113825597542,
+      "learning_rate": 0.0001638358075536372,
+      "loss": 0.1698,
+      "step": 2124
+    },
+    {
+      "epoch": 14.97797356828194,
+      "grad_norm": 0.2631809542165346,
+      "learning_rate": 0.00016380043655771443,
+      "loss": 0.1674,
+      "step": 2125
+    },
+    {
+      "epoch": 14.985022026431718,
+      "grad_norm": 0.26519741889055326,
+      "learning_rate": 0.00016376505209496735,
+      "loss": 0.1586,
+      "step": 2126
+    },
+    {
+      "epoch": 14.992070484581498,
+      "grad_norm": 0.2506079778753953,
+      "learning_rate": 0.0001637296541728648,
+      "loss": 0.1551,
+      "step": 2127
+    },
+    {
+      "epoch": 14.999118942731277,
+      "grad_norm": 0.2505330969709996,
+      "learning_rate": 0.00016369424279887845,
+      "loss": 0.1571,
+      "step": 2128
+    },
+    {
+      "epoch": 15.006167400881058,
+      "grad_norm": 0.2563500957819969,
+      "learning_rate": 0.00016365881798048284,
+      "loss": 0.1212,
+      "step": 2129
+    },
+    {
+      "epoch": 15.013215859030836,
+      "grad_norm": 0.23918607600028585,
+      "learning_rate": 0.00016362337972515535,
+      "loss": 0.116,
+      "step": 2130
+    },
+    {
+      "epoch": 15.020264317180617,
+      "grad_norm": 0.30238732535003854,
+      "learning_rate": 0.00016358792804037627,
+      "loss": 0.1223,
+      "step": 2131
+    },
+    {
+      "epoch": 15.027312775330396,
+      "grad_norm": 0.29814184309280617,
+      "learning_rate": 0.00016355246293362852,
+      "loss": 0.1162,
+      "step": 2132
+    },
+    {
+      "epoch": 15.034361233480176,
+      "grad_norm": 0.2710718829711657,
+      "learning_rate": 0.00016351698441239803,
+      "loss": 0.1122,
+      "step": 2133
+    },
+    {
+      "epoch": 15.041409691629957,
+      "grad_norm": 0.2540838617936457,
+      "learning_rate": 0.00016348149248417352,
+      "loss": 0.1126,
+      "step": 2134
+    },
+    {
+      "epoch": 15.048458149779735,
+      "grad_norm": 0.24615809721170415,
+      "learning_rate": 0.00016344598715644656,
+      "loss": 0.1091,
+      "step": 2135
+    },
+    {
+      "epoch": 15.055506607929516,
+      "grad_norm": 0.25681058240428517,
+      "learning_rate": 0.00016341046843671144,
+      "loss": 0.1123,
+      "step": 2136
+    },
+    {
+      "epoch": 15.062555066079295,
+      "grad_norm": 0.2491586107923605,
+      "learning_rate": 0.00016337493633246543,
+      "loss": 0.1184,
+      "step": 2137
+    },
+    {
+      "epoch": 15.069603524229075,
+      "grad_norm": 0.26436372994716234,
+      "learning_rate": 0.0001633393908512085,
+      "loss": 0.1076,
+      "step": 2138
+    },
+    {
+      "epoch": 15.076651982378854,
+      "grad_norm": 0.2698105407002603,
+      "learning_rate": 0.00016330383200044353,
+      "loss": 0.1129,
+      "step": 2139
+    },
+    {
+      "epoch": 15.083700440528634,
+      "grad_norm": 0.2597418884139394,
+      "learning_rate": 0.0001632682597876762,
+      "loss": 0.1126,
+      "step": 2140
+    },
+    {
+      "epoch": 15.090748898678415,
+      "grad_norm": 0.285710310643828,
+      "learning_rate": 0.00016323267422041493,
+      "loss": 0.1115,
+      "step": 2141
+    },
+    {
+      "epoch": 15.097797356828194,
+      "grad_norm": 0.2640283216384672,
+      "learning_rate": 0.00016319707530617106,
+      "loss": 0.1145,
+      "step": 2142
+    },
+    {
+      "epoch": 15.104845814977974,
+      "grad_norm": 0.2501055949035317,
+      "learning_rate": 0.00016316146305245875,
+      "loss": 0.1133,
+      "step": 2143
+    },
+    {
+      "epoch": 15.111894273127753,
+      "grad_norm": 0.26323269070137434,
+      "learning_rate": 0.0001631258374667949,
+      "loss": 0.1155,
+      "step": 2144
+    },
+    {
+      "epoch": 15.118942731277533,
+      "grad_norm": 0.2518343922112884,
+      "learning_rate": 0.00016309019855669928,
+      "loss": 0.1185,
+      "step": 2145
+    },
+    {
+      "epoch": 15.125991189427312,
+      "grad_norm": 0.2919985470432582,
+      "learning_rate": 0.00016305454632969444,
+      "loss": 0.111,
+      "step": 2146
+    },
+    {
+      "epoch": 15.133039647577093,
+      "grad_norm": 0.24664425106046986,
+      "learning_rate": 0.00016301888079330574,
+      "loss": 0.1136,
+      "step": 2147
+    },
+    {
+      "epoch": 15.140088105726873,
+      "grad_norm": 0.25523133728610936,
+      "learning_rate": 0.0001629832019550614,
+      "loss": 0.1211,
+      "step": 2148
+    },
+    {
+      "epoch": 15.147136563876652,
+      "grad_norm": 0.253872175554803,
+      "learning_rate": 0.0001629475098224924,
+      "loss": 0.1167,
+      "step": 2149
+    },
+    {
+      "epoch": 15.154185022026432,
+      "grad_norm": 0.2413861632247846,
+      "learning_rate": 0.00016291180440313255,
+      "loss": 0.1089,
+      "step": 2150
+    },
+    {
+      "epoch": 15.161233480176211,
+      "grad_norm": 0.27784897814916143,
+      "learning_rate": 0.00016287608570451844,
+      "loss": 0.1164,
+      "step": 2151
+    },
+    {
+      "epoch": 15.168281938325991,
+      "grad_norm": 0.26856487893888686,
+      "learning_rate": 0.00016284035373418946,
+      "loss": 0.1112,
+      "step": 2152
+    },
+    {
+      "epoch": 15.17533039647577,
+      "grad_norm": 0.26032330723267305,
+      "learning_rate": 0.00016280460849968785,
+      "loss": 0.1138,
+      "step": 2153
+    },
+    {
+      "epoch": 15.18237885462555,
+      "grad_norm": 0.26790774625833175,
+      "learning_rate": 0.00016276885000855859,
+      "loss": 0.1124,
+      "step": 2154
+    },
+    {
+      "epoch": 15.189427312775331,
+      "grad_norm": 0.24139792244641572,
+      "learning_rate": 0.00016273307826834954,
+      "loss": 0.1111,
+      "step": 2155
+    },
+    {
+      "epoch": 15.19647577092511,
+      "grad_norm": 0.2464327158356326,
+      "learning_rate": 0.00016269729328661123,
+      "loss": 0.1114,
+      "step": 2156
+    },
+    {
+      "epoch": 15.20352422907489,
+      "grad_norm": 0.26406512080559474,
+      "learning_rate": 0.00016266149507089713,
+      "loss": 0.1158,
+      "step": 2157
+    },
+    {
+      "epoch": 15.21057268722467,
+      "grad_norm": 0.24993948019050852,
+      "learning_rate": 0.00016262568362876336,
+      "loss": 0.1152,
+      "step": 2158
+    },
+    {
+      "epoch": 15.21762114537445,
+      "grad_norm": 0.2571783810056495,
+      "learning_rate": 0.0001625898589677689,
+      "loss": 0.1126,
+      "step": 2159
+    },
+    {
+      "epoch": 15.224669603524228,
+      "grad_norm": 0.2539672158878651,
+      "learning_rate": 0.00016255402109547563,
+      "loss": 0.106,
+      "step": 2160
+    },
+    {
+      "epoch": 15.231718061674009,
+      "grad_norm": 0.25878954637258433,
+      "learning_rate": 0.00016251817001944802,
+      "loss": 0.1142,
+      "step": 2161
+    },
+    {
+      "epoch": 15.23876651982379,
+      "grad_norm": 0.2666027032252859,
+      "learning_rate": 0.0001624823057472534,
+      "loss": 0.1125,
+      "step": 2162
+    },
+    {
+      "epoch": 15.245814977973568,
+      "grad_norm": 0.25875396415647217,
+      "learning_rate": 0.000162446428286462,
+      "loss": 0.1136,
+      "step": 2163
+    },
+    {
+      "epoch": 15.252863436123349,
+      "grad_norm": 0.2504771453835889,
+      "learning_rate": 0.00016241053764464662,
+      "loss": 0.1109,
+      "step": 2164
+    },
+    {
+      "epoch": 15.259911894273127,
+      "grad_norm": 0.25029065461992805,
+      "learning_rate": 0.00016237463382938306,
+      "loss": 0.1158,
+      "step": 2165
+    },
+    {
+      "epoch": 15.266960352422908,
+      "grad_norm": 0.2819667012600368,
+      "learning_rate": 0.00016233871684824975,
+      "loss": 0.1256,
+      "step": 2166
+    },
+    {
+      "epoch": 15.274008810572687,
+      "grad_norm": 0.24941878353846844,
+      "learning_rate": 0.00016230278670882796,
+      "loss": 0.1092,
+      "step": 2167
+    },
+    {
+      "epoch": 15.281057268722467,
+      "grad_norm": 0.26128441850295836,
+      "learning_rate": 0.00016226684341870174,
+      "loss": 0.1149,
+      "step": 2168
+    },
+    {
+      "epoch": 15.288105726872246,
+      "grad_norm": 0.2736430570442749,
+      "learning_rate": 0.0001622308869854579,
+      "loss": 0.1174,
+      "step": 2169
+    },
+    {
+      "epoch": 15.295154185022026,
+      "grad_norm": 0.2645256523605719,
+      "learning_rate": 0.00016219491741668597,
+      "loss": 0.1141,
+      "step": 2170
+    },
+    {
+      "epoch": 15.302202643171807,
+      "grad_norm": 0.2682795056250476,
+      "learning_rate": 0.0001621589347199784,
+      "loss": 0.1144,
+      "step": 2171
+    },
+    {
+      "epoch": 15.309251101321586,
+      "grad_norm": 0.2841410997346254,
+      "learning_rate": 0.00016212293890293026,
+      "loss": 0.1142,
+      "step": 2172
+    },
+    {
+      "epoch": 15.316299559471366,
+      "grad_norm": 0.27320722738325853,
+      "learning_rate": 0.0001620869299731395,
+      "loss": 0.1199,
+      "step": 2173
+    },
+    {
+      "epoch": 15.323348017621145,
+      "grad_norm": 0.27065975431555905,
+      "learning_rate": 0.0001620509079382067,
+      "loss": 0.1196,
+      "step": 2174
+    },
+    {
+      "epoch": 15.330396475770925,
+      "grad_norm": 0.25094074658529264,
+      "learning_rate": 0.00016201487280573536,
+      "loss": 0.1214,
+      "step": 2175
+    },
+    {
+      "epoch": 15.337444933920704,
+      "grad_norm": 0.2553527148185859,
+      "learning_rate": 0.00016197882458333172,
+      "loss": 0.1307,
+      "step": 2176
+    },
+    {
+      "epoch": 15.344493392070484,
+      "grad_norm": 0.2760551777830757,
+      "learning_rate": 0.00016194276327860462,
+      "loss": 0.1211,
+      "step": 2177
+    },
+    {
+      "epoch": 15.351541850220265,
+      "grad_norm": 0.2721108583394494,
+      "learning_rate": 0.00016190668889916592,
+      "loss": 0.1181,
+      "step": 2178
+    },
+    {
+      "epoch": 15.358590308370044,
+      "grad_norm": 0.24981371878982586,
+      "learning_rate": 0.00016187060145263,
+      "loss": 0.1235,
+      "step": 2179
+    },
+    {
+      "epoch": 15.365638766519824,
+      "grad_norm": 0.27428229726399245,
+      "learning_rate": 0.00016183450094661416,
+      "loss": 0.1198,
+      "step": 2180
+    },
+    {
+      "epoch": 15.372687224669603,
+      "grad_norm": 0.2610560206583446,
+      "learning_rate": 0.00016179838738873835,
+      "loss": 0.1139,
+      "step": 2181
+    },
+    {
+      "epoch": 15.379735682819383,
+      "grad_norm": 0.27002768511706804,
+      "learning_rate": 0.00016176226078662534,
+      "loss": 0.1199,
+      "step": 2182
+    },
+    {
+      "epoch": 15.386784140969162,
+      "grad_norm": 0.24640361109992243,
+      "learning_rate": 0.0001617261211479007,
+      "loss": 0.1102,
+      "step": 2183
+    },
+    {
+      "epoch": 15.393832599118943,
+      "grad_norm": 0.26826914187162365,
+      "learning_rate": 0.00016168996848019254,
+      "loss": 0.1199,
+      "step": 2184
+    },
+    {
+      "epoch": 15.400881057268723,
+      "grad_norm": 0.27100274139194896,
+      "learning_rate": 0.00016165380279113204,
+      "loss": 0.114,
+      "step": 2185
+    },
+    {
+      "epoch": 15.407929515418502,
+      "grad_norm": 0.26214399897319235,
+      "learning_rate": 0.00016161762408835285,
+      "loss": 0.1169,
+      "step": 2186
+    },
+    {
+      "epoch": 15.414977973568282,
+      "grad_norm": 0.25614192063221713,
+      "learning_rate": 0.0001615814323794915,
+      "loss": 0.1176,
+      "step": 2187
+    },
+    {
+      "epoch": 15.422026431718061,
+      "grad_norm": 0.2548264952504106,
+      "learning_rate": 0.00016154522767218725,
+      "loss": 0.1196,
+      "step": 2188
+    },
+    {
+      "epoch": 15.429074889867842,
+      "grad_norm": 0.267527242356525,
+      "learning_rate": 0.00016150900997408205,
+      "loss": 0.1196,
+      "step": 2189
+    },
+    {
+      "epoch": 15.43612334801762,
+      "grad_norm": 0.24845605734361528,
+      "learning_rate": 0.0001614727792928207,
+      "loss": 0.1181,
+      "step": 2190
+    },
+    {
+      "epoch": 15.4431718061674,
+      "grad_norm": 0.2668250455480989,
+      "learning_rate": 0.0001614365356360506,
+      "loss": 0.1199,
+      "step": 2191
+    },
+    {
+      "epoch": 15.450220264317181,
+      "grad_norm": 0.273123108793945,
+      "learning_rate": 0.000161400279011422,
+      "loss": 0.1247,
+      "step": 2192
+    },
+    {
+      "epoch": 15.45726872246696,
+      "grad_norm": 0.26055326790432937,
+      "learning_rate": 0.00016136400942658787,
+      "loss": 0.1219,
+      "step": 2193
+    },
+    {
+      "epoch": 15.46431718061674,
+      "grad_norm": 0.25006332784414764,
+      "learning_rate": 0.0001613277268892039,
+      "loss": 0.123,
+      "step": 2194
+    },
+    {
+      "epoch": 15.47136563876652,
+      "grad_norm": 0.2676992481256137,
+      "learning_rate": 0.00016129143140692845,
+      "loss": 0.1223,
+      "step": 2195
+    },
+    {
+      "epoch": 15.4784140969163,
+      "grad_norm": 0.2675434236812215,
+      "learning_rate": 0.00016125512298742272,
+      "loss": 0.1207,
+      "step": 2196
+    },
+    {
+      "epoch": 15.485462555066078,
+      "grad_norm": 0.2525788208714123,
+      "learning_rate": 0.0001612188016383506,
+      "loss": 0.1214,
+      "step": 2197
+    },
+    {
+      "epoch": 15.492511013215859,
+      "grad_norm": 0.25991877448551326,
+      "learning_rate": 0.0001611824673673787,
+      "loss": 0.1147,
+      "step": 2198
+    },
+    {
+      "epoch": 15.49955947136564,
+      "grad_norm": 0.2681915672130508,
+      "learning_rate": 0.0001611461201821763,
+      "loss": 0.1191,
+      "step": 2199
+    },
+    {
+      "epoch": 15.506607929515418,
+      "grad_norm": 0.27296847843778643,
+      "learning_rate": 0.00016110976009041554,
+      "loss": 0.129,
+      "step": 2200
+    },
+    {
+      "epoch": 15.513656387665199,
+      "grad_norm": 0.2515999937373049,
+      "learning_rate": 0.0001610733870997712,
+      "loss": 0.1227,
+      "step": 2201
+    },
+    {
+      "epoch": 15.520704845814977,
+      "grad_norm": 0.26792082115334104,
+      "learning_rate": 0.00016103700121792073,
+      "loss": 0.1242,
+      "step": 2202
+    },
+    {
+      "epoch": 15.527753303964758,
+      "grad_norm": 0.26140962284487645,
+      "learning_rate": 0.00016100060245254446,
+      "loss": 0.1181,
+      "step": 2203
+    },
+    {
+      "epoch": 15.534801762114537,
+      "grad_norm": 0.2608341603100969,
+      "learning_rate": 0.00016096419081132525,
+      "loss": 0.1272,
+      "step": 2204
+    },
+    {
+      "epoch": 15.541850220264317,
+      "grad_norm": 0.2525746781305859,
+      "learning_rate": 0.00016092776630194887,
+      "loss": 0.129,
+      "step": 2205
+    },
+    {
+      "epoch": 15.548898678414098,
+      "grad_norm": 0.26889598553848126,
+      "learning_rate": 0.00016089132893210362,
+      "loss": 0.1254,
+      "step": 2206
+    },
+    {
+      "epoch": 15.555947136563876,
+      "grad_norm": 0.26618128849342987,
+      "learning_rate": 0.00016085487870948063,
+      "loss": 0.1328,
+      "step": 2207
+    },
+    {
+      "epoch": 15.562995594713657,
+      "grad_norm": 0.27105562714622083,
+      "learning_rate": 0.0001608184156417737,
+      "loss": 0.1242,
+      "step": 2208
+    },
+    {
+      "epoch": 15.570044052863436,
+      "grad_norm": 0.25633511679533527,
+      "learning_rate": 0.0001607819397366794,
+      "loss": 0.1193,
+      "step": 2209
+    },
+    {
+      "epoch": 15.577092511013216,
+      "grad_norm": 0.25350028474155634,
+      "learning_rate": 0.00016074545100189694,
+      "loss": 0.1164,
+      "step": 2210
+    },
+    {
+      "epoch": 15.584140969162995,
+      "grad_norm": 0.25552250863731674,
+      "learning_rate": 0.00016070894944512828,
+      "loss": 0.1336,
+      "step": 2211
+    },
+    {
+      "epoch": 15.591189427312775,
+      "grad_norm": 0.2633967475557661,
+      "learning_rate": 0.00016067243507407802,
+      "loss": 0.1271,
+      "step": 2212
+    },
+    {
+      "epoch": 15.598237885462556,
+      "grad_norm": 0.2615733744536034,
+      "learning_rate": 0.0001606359078964536,
+      "loss": 0.1255,
+      "step": 2213
+    },
+    {
+      "epoch": 15.605286343612335,
+      "grad_norm": 0.25929343179923886,
+      "learning_rate": 0.000160599367919965,
+      "loss": 0.1218,
+      "step": 2214
+    },
+    {
+      "epoch": 15.612334801762115,
+      "grad_norm": 0.27463156926971005,
+      "learning_rate": 0.000160562815152325,
+      "loss": 0.133,
+      "step": 2215
+    },
+    {
+      "epoch": 15.619383259911894,
+      "grad_norm": 0.25316307547847106,
+      "learning_rate": 0.0001605262496012491,
+      "loss": 0.1238,
+      "step": 2216
+    },
+    {
+      "epoch": 15.626431718061674,
+      "grad_norm": 0.2726343589550909,
+      "learning_rate": 0.0001604896712744554,
+      "loss": 0.1271,
+      "step": 2217
+    },
+    {
+      "epoch": 15.633480176211453,
+      "grad_norm": 0.265592696493809,
+      "learning_rate": 0.00016045308017966476,
+      "loss": 0.1222,
+      "step": 2218
+    },
+    {
+      "epoch": 15.640528634361234,
+      "grad_norm": 0.2626726902441541,
+      "learning_rate": 0.00016041647632460078,
+      "loss": 0.1303,
+      "step": 2219
+    },
+    {
+      "epoch": 15.647577092511014,
+      "grad_norm": 0.2624261472399329,
+      "learning_rate": 0.00016037985971698965,
+      "loss": 0.1297,
+      "step": 2220
+    },
+    {
+      "epoch": 15.654625550660793,
+      "grad_norm": 0.28717800679443173,
+      "learning_rate": 0.00016034323036456035,
+      "loss": 0.1323,
+      "step": 2221
+    },
+    {
+      "epoch": 15.661674008810573,
+      "grad_norm": 0.27292222475471833,
+      "learning_rate": 0.0001603065882750445,
+      "loss": 0.1288,
+      "step": 2222
+    },
+    {
+      "epoch": 15.668722466960352,
+      "grad_norm": 0.26517466447265947,
+      "learning_rate": 0.00016026993345617636,
+      "loss": 0.1233,
+      "step": 2223
+    },
+    {
+      "epoch": 15.675770925110132,
+      "grad_norm": 0.2584004417924547,
+      "learning_rate": 0.000160233265915693,
+      "loss": 0.1326,
+      "step": 2224
+    },
+    {
+      "epoch": 15.682819383259911,
+      "grad_norm": 0.26845631216271465,
+      "learning_rate": 0.00016019658566133405,
+      "loss": 0.1361,
+      "step": 2225
+    },
+    {
+      "epoch": 15.689867841409692,
+      "grad_norm": 0.26758203671656405,
+      "learning_rate": 0.00016015989270084193,
+      "loss": 0.1309,
+      "step": 2226
+    },
+    {
+      "epoch": 15.696916299559472,
+      "grad_norm": 0.28564601348185187,
+      "learning_rate": 0.00016012318704196164,
+      "loss": 0.1284,
+      "step": 2227
+    },
+    {
+      "epoch": 15.703964757709251,
+      "grad_norm": 0.2590166389207009,
+      "learning_rate": 0.00016008646869244096,
+      "loss": 0.1226,
+      "step": 2228
+    },
+    {
+      "epoch": 15.711013215859031,
+      "grad_norm": 0.26408472467817173,
+      "learning_rate": 0.00016004973766003025,
+      "loss": 0.1304,
+      "step": 2229
+    },
+    {
+      "epoch": 15.71806167400881,
+      "grad_norm": 0.26122528215090773,
+      "learning_rate": 0.00016001299395248264,
+      "loss": 0.1325,
+      "step": 2230
+    },
+    {
+      "epoch": 15.72511013215859,
+      "grad_norm": 0.2690683592303064,
+      "learning_rate": 0.00015997623757755386,
+      "loss": 0.1274,
+      "step": 2231
+    },
+    {
+      "epoch": 15.73215859030837,
+      "grad_norm": 0.269440793139246,
+      "learning_rate": 0.00015993946854300236,
+      "loss": 0.1263,
+      "step": 2232
+    },
+    {
+      "epoch": 15.73920704845815,
+      "grad_norm": 0.2820183978918874,
+      "learning_rate": 0.00015990268685658928,
+      "loss": 0.1381,
+      "step": 2233
+    },
+    {
+      "epoch": 15.74625550660793,
+      "grad_norm": 0.26877143611959303,
+      "learning_rate": 0.00015986589252607836,
+      "loss": 0.1242,
+      "step": 2234
+    },
+    {
+      "epoch": 15.753303964757709,
+      "grad_norm": 0.2806871599798028,
+      "learning_rate": 0.00015982908555923605,
+      "loss": 0.1317,
+      "step": 2235
+    },
+    {
+      "epoch": 15.76035242290749,
+      "grad_norm": 0.27512727625154537,
+      "learning_rate": 0.0001597922659638315,
+      "loss": 0.1312,
+      "step": 2236
+    },
+    {
+      "epoch": 15.767400881057268,
+      "grad_norm": 0.2718020399997642,
+      "learning_rate": 0.00015975543374763645,
+      "loss": 0.127,
+      "step": 2237
+    },
+    {
+      "epoch": 15.774449339207049,
+      "grad_norm": 0.2905026599531269,
+      "learning_rate": 0.00015971858891842536,
+      "loss": 0.1361,
+      "step": 2238
+    },
+    {
+      "epoch": 15.781497797356828,
+      "grad_norm": 0.2626201783925056,
+      "learning_rate": 0.0001596817314839753,
+      "loss": 0.1217,
+      "step": 2239
+    },
+    {
+      "epoch": 15.788546255506608,
+      "grad_norm": 0.2504616376714748,
+      "learning_rate": 0.0001596448614520661,
+      "loss": 0.1327,
+      "step": 2240
+    },
+    {
+      "epoch": 15.795594713656389,
+      "grad_norm": 0.25283602915893144,
+      "learning_rate": 0.0001596079788304802,
+      "loss": 0.1241,
+      "step": 2241
+    },
+    {
+      "epoch": 15.802643171806167,
+      "grad_norm": 0.25547489576938653,
+      "learning_rate": 0.00015957108362700263,
+      "loss": 0.1254,
+      "step": 2242
+    },
+    {
+      "epoch": 15.809691629955948,
+      "grad_norm": 0.2629850054394946,
+      "learning_rate": 0.00015953417584942115,
+      "loss": 0.1268,
+      "step": 2243
+    },
+    {
+      "epoch": 15.816740088105727,
+      "grad_norm": 0.2720134981908641,
+      "learning_rate": 0.00015949725550552611,
+      "loss": 0.132,
+      "step": 2244
+    },
+    {
+      "epoch": 15.823788546255507,
+      "grad_norm": 0.27360449654347785,
+      "learning_rate": 0.00015946032260311064,
+      "loss": 0.1312,
+      "step": 2245
+    },
+    {
+      "epoch": 15.830837004405286,
+      "grad_norm": 0.27132935046109197,
+      "learning_rate": 0.0001594233771499704,
+      "loss": 0.1369,
+      "step": 2246
+    },
+    {
+      "epoch": 15.837885462555066,
+      "grad_norm": 0.26377399711123745,
+      "learning_rate": 0.00015938641915390372,
+      "loss": 0.1324,
+      "step": 2247
+    },
+    {
+      "epoch": 15.844933920704847,
+      "grad_norm": 0.27264012922589503,
+      "learning_rate": 0.0001593494486227116,
+      "loss": 0.131,
+      "step": 2248
+    },
+    {
+      "epoch": 15.851982378854625,
+      "grad_norm": 0.2662187137953817,
+      "learning_rate": 0.00015931246556419773,
+      "loss": 0.1253,
+      "step": 2249
+    },
+    {
+      "epoch": 15.859030837004406,
+      "grad_norm": 0.26926491080949094,
+      "learning_rate": 0.0001592754699861683,
+      "loss": 0.1296,
+      "step": 2250
+    },
+    {
+      "epoch": 15.866079295154185,
+      "grad_norm": 0.281944780490568,
+      "learning_rate": 0.0001592384618964323,
+      "loss": 0.1365,
+      "step": 2251
+    },
+    {
+      "epoch": 15.873127753303965,
+      "grad_norm": 0.29775508329437933,
+      "learning_rate": 0.0001592014413028013,
+      "loss": 0.1337,
+      "step": 2252
+    },
+    {
+      "epoch": 15.880176211453744,
+      "grad_norm": 0.2733968682400423,
+      "learning_rate": 0.0001591644082130895,
+      "loss": 0.1387,
+      "step": 2253
+    },
+    {
+      "epoch": 15.887224669603524,
+      "grad_norm": 0.2733314079239422,
+      "learning_rate": 0.00015912736263511373,
+      "loss": 0.1327,
+      "step": 2254
+    },
+    {
+      "epoch": 15.894273127753303,
+      "grad_norm": 0.26445028739373777,
+      "learning_rate": 0.00015909030457669348,
+      "loss": 0.1278,
+      "step": 2255
+    },
+    {
+      "epoch": 15.901321585903084,
+      "grad_norm": 0.26486996297541765,
+      "learning_rate": 0.00015905323404565084,
+      "loss": 0.1305,
+      "step": 2256
+    },
+    {
+      "epoch": 15.908370044052864,
+      "grad_norm": 0.30345087141053523,
+      "learning_rate": 0.00015901615104981062,
+      "loss": 0.1347,
+      "step": 2257
+    },
+    {
+      "epoch": 15.915418502202643,
+      "grad_norm": 0.25493765694593146,
+      "learning_rate": 0.00015897905559700016,
+      "loss": 0.1378,
+      "step": 2258
+    },
+    {
+      "epoch": 15.922466960352423,
+      "grad_norm": 0.2651337280762559,
+      "learning_rate": 0.00015894194769504945,
+      "loss": 0.1354,
+      "step": 2259
+    },
+    {
+      "epoch": 15.929515418502202,
+      "grad_norm": 0.2826054818822418,
+      "learning_rate": 0.00015890482735179116,
+      "loss": 0.1393,
+      "step": 2260
+    },
+    {
+      "epoch": 15.936563876651983,
+      "grad_norm": 0.27536168474895095,
+      "learning_rate": 0.00015886769457506053,
+      "loss": 0.1458,
+      "step": 2261
+    },
+    {
+      "epoch": 15.943612334801761,
+      "grad_norm": 0.25610966405049995,
+      "learning_rate": 0.00015883054937269547,
+      "loss": 0.1284,
+      "step": 2262
+    },
+    {
+      "epoch": 15.950660792951542,
+      "grad_norm": 0.275777451428029,
+      "learning_rate": 0.00015879339175253648,
+      "loss": 0.1388,
+      "step": 2263
+    },
+    {
+      "epoch": 15.957709251101322,
+      "grad_norm": 0.26733186442295875,
+      "learning_rate": 0.00015875622172242663,
+      "loss": 0.1412,
+      "step": 2264
+    },
+    {
+      "epoch": 15.964757709251101,
+      "grad_norm": 0.2641859750923611,
+      "learning_rate": 0.00015871903929021177,
+      "loss": 0.1364,
+      "step": 2265
+    },
+    {
+      "epoch": 15.971806167400882,
+      "grad_norm": 0.27694394623543894,
+      "learning_rate": 0.0001586818444637402,
+      "loss": 0.1399,
+      "step": 2266
+    },
+    {
+      "epoch": 15.97885462555066,
+      "grad_norm": 0.28425454144988765,
+      "learning_rate": 0.00015864463725086297,
+      "loss": 0.1293,
+      "step": 2267
+    },
+    {
+      "epoch": 15.98590308370044,
+      "grad_norm": 0.28726486786625544,
+      "learning_rate": 0.00015860741765943358,
+      "loss": 0.1407,
+      "step": 2268
+    },
+    {
+      "epoch": 15.99295154185022,
+      "grad_norm": 0.2644365964815448,
+      "learning_rate": 0.00015857018569730835,
+      "loss": 0.132,
+      "step": 2269
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.28965033378939115,
+      "learning_rate": 0.00015853294137234603,
+      "loss": 0.1403,
+      "step": 2270
+    },
+    {
+      "epoch": 16.00704845814978,
+      "grad_norm": 0.24372138371739455,
+      "learning_rate": 0.00015849568469240803,
+      "loss": 0.0845,
+      "step": 2271
+    },
+    {
+      "epoch": 16.01409691629956,
+      "grad_norm": 0.2273948631544325,
+      "learning_rate": 0.00015845841566535846,
+      "loss": 0.0869,
+      "step": 2272
+    },
+    {
+      "epoch": 16.02114537444934,
+      "grad_norm": 0.26610202252835313,
+      "learning_rate": 0.00015842113429906393,
+      "loss": 0.0833,
+      "step": 2273
+    },
+    {
+      "epoch": 16.02819383259912,
+      "grad_norm": 0.2781010534246402,
+      "learning_rate": 0.00015838384060139367,
+      "loss": 0.0892,
+      "step": 2274
+    },
+    {
+      "epoch": 16.035242290748897,
+      "grad_norm": 0.2756683639799539,
+      "learning_rate": 0.00015834653458021963,
+      "loss": 0.0893,
+      "step": 2275
+    },
+    {
+      "epoch": 16.04229074889868,
+      "grad_norm": 0.23875765119319237,
+      "learning_rate": 0.00015830921624341614,
+      "loss": 0.081,
+      "step": 2276
+    },
+    {
+      "epoch": 16.049339207048458,
+      "grad_norm": 0.2710968120046608,
+      "learning_rate": 0.0001582718855988603,
+      "loss": 0.0865,
+      "step": 2277
+    },
+    {
+      "epoch": 16.056387665198237,
+      "grad_norm": 0.25820597833370545,
+      "learning_rate": 0.0001582345426544318,
+      "loss": 0.0816,
+      "step": 2278
+    },
+    {
+      "epoch": 16.06343612334802,
+      "grad_norm": 0.24538074111803795,
+      "learning_rate": 0.00015819718741801283,
+      "loss": 0.0765,
+      "step": 2279
+    },
+    {
+      "epoch": 16.070484581497798,
+      "grad_norm": 0.25415237234825194,
+      "learning_rate": 0.00015815981989748828,
+      "loss": 0.0819,
+      "step": 2280
+    },
+    {
+      "epoch": 16.077533039647577,
+      "grad_norm": 0.27276895364318327,
+      "learning_rate": 0.00015812244010074555,
+      "loss": 0.0895,
+      "step": 2281
+    },
+    {
+      "epoch": 16.084581497797355,
+      "grad_norm": 0.2675501381206662,
+      "learning_rate": 0.00015808504803567467,
+      "loss": 0.0875,
+      "step": 2282
+    },
+    {
+      "epoch": 16.091629955947138,
+      "grad_norm": 0.2506729312354951,
+      "learning_rate": 0.0001580476437101683,
+      "loss": 0.0886,
+      "step": 2283
+    },
+    {
+      "epoch": 16.098678414096916,
+      "grad_norm": 0.23890372526829376,
+      "learning_rate": 0.00015801022713212158,
+      "loss": 0.0887,
+      "step": 2284
+    },
+    {
+      "epoch": 16.105726872246695,
+      "grad_norm": 0.24064797061562682,
+      "learning_rate": 0.00015797279830943238,
+      "loss": 0.0845,
+      "step": 2285
+    },
+    {
+      "epoch": 16.112775330396477,
+      "grad_norm": 0.25063235750290136,
+      "learning_rate": 0.00015793535725000098,
+      "loss": 0.0848,
+      "step": 2286
+    },
+    {
+      "epoch": 16.119823788546256,
+      "grad_norm": 0.25908791457776775,
+      "learning_rate": 0.00015789790396173044,
+      "loss": 0.0863,
+      "step": 2287
+    },
+    {
+      "epoch": 16.126872246696035,
+      "grad_norm": 0.2454283158986973,
+      "learning_rate": 0.00015786043845252622,
+      "loss": 0.0812,
+      "step": 2288
+    },
+    {
+      "epoch": 16.133920704845814,
+      "grad_norm": 0.2571520626976104,
+      "learning_rate": 0.00015782296073029646,
+      "loss": 0.0844,
+      "step": 2289
+    },
+    {
+      "epoch": 16.140969162995596,
+      "grad_norm": 0.2676406397440159,
+      "learning_rate": 0.00015778547080295185,
+      "loss": 0.0916,
+      "step": 2290
+    },
+    {
+      "epoch": 16.148017621145375,
+      "grad_norm": 0.24364450790885683,
+      "learning_rate": 0.0001577479686784057,
+      "loss": 0.0865,
+      "step": 2291
+    },
+    {
+      "epoch": 16.155066079295153,
+      "grad_norm": 0.2526730559486457,
+      "learning_rate": 0.0001577104543645738,
+      "loss": 0.0924,
+      "step": 2292
+    },
+    {
+      "epoch": 16.162114537444936,
+      "grad_norm": 0.24044881672896992,
+      "learning_rate": 0.00015767292786937463,
+      "loss": 0.0885,
+      "step": 2293
+    },
+    {
+      "epoch": 16.169162995594714,
+      "grad_norm": 0.25648958857494947,
+      "learning_rate": 0.00015763538920072914,
+      "loss": 0.0833,
+      "step": 2294
+    },
+    {
+      "epoch": 16.176211453744493,
+      "grad_norm": 0.25696040718556534,
+      "learning_rate": 0.00015759783836656092,
+      "loss": 0.0861,
+      "step": 2295
+    },
+    {
+      "epoch": 16.18325991189427,
+      "grad_norm": 0.2660198292783282,
+      "learning_rate": 0.00015756027537479608,
+      "loss": 0.0892,
+      "step": 2296
+    },
+    {
+      "epoch": 16.190308370044054,
+      "grad_norm": 0.2597362078469247,
+      "learning_rate": 0.00015752270023336332,
+      "loss": 0.0838,
+      "step": 2297
+    },
+    {
+      "epoch": 16.197356828193833,
+      "grad_norm": 0.2651012557447694,
+      "learning_rate": 0.0001574851129501939,
+      "loss": 0.0846,
+      "step": 2298
+    },
+    {
+      "epoch": 16.20440528634361,
+      "grad_norm": 0.24495408424664872,
+      "learning_rate": 0.00015744751353322163,
+      "loss": 0.0895,
+      "step": 2299
+    },
+    {
+      "epoch": 16.211453744493394,
+      "grad_norm": 0.25364552409316526,
+      "learning_rate": 0.00015740990199038287,
+      "loss": 0.0899,
+      "step": 2300
+    },
+    {
+      "epoch": 16.218502202643172,
+      "grad_norm": 0.23143597809589667,
+      "learning_rate": 0.00015737227832961665,
+      "loss": 0.0901,
+      "step": 2301
+    },
+    {
+      "epoch": 16.22555066079295,
+      "grad_norm": 0.25283652879124924,
+      "learning_rate": 0.00015733464255886437,
+      "loss": 0.0939,
+      "step": 2302
+    },
+    {
+      "epoch": 16.23259911894273,
+      "grad_norm": 0.2718401866601472,
+      "learning_rate": 0.0001572969946860702,
+      "loss": 0.0873,
+      "step": 2303
+    },
+    {
+      "epoch": 16.239647577092512,
+      "grad_norm": 0.24204948830255055,
+      "learning_rate": 0.00015725933471918068,
+      "loss": 0.0942,
+      "step": 2304
+    },
+    {
+      "epoch": 16.24669603524229,
+      "grad_norm": 0.25870747270074196,
+      "learning_rate": 0.00015722166266614494,
+      "loss": 0.0843,
+      "step": 2305
+    },
+    {
+      "epoch": 16.25374449339207,
+      "grad_norm": 0.2588707967744293,
+      "learning_rate": 0.00015718397853491478,
+      "loss": 0.0914,
+      "step": 2306
+    },
+    {
+      "epoch": 16.260792951541852,
+      "grad_norm": 0.2447828651160685,
+      "learning_rate": 0.00015714628233344437,
+      "loss": 0.0938,
+      "step": 2307
+    },
+    {
+      "epoch": 16.26784140969163,
+      "grad_norm": 0.24578310661402808,
+      "learning_rate": 0.00015710857406969062,
+      "loss": 0.0939,
+      "step": 2308
+    },
+    {
+      "epoch": 16.27488986784141,
+      "grad_norm": 0.27925832814198226,
+      "learning_rate": 0.00015707085375161282,
+      "loss": 0.0956,
+      "step": 2309
+    },
+    {
+      "epoch": 16.281938325991188,
+      "grad_norm": 0.2476167745148899,
+      "learning_rate": 0.0001570331213871729,
+      "loss": 0.0962,
+      "step": 2310
+    },
+    {
+      "epoch": 16.28898678414097,
+      "grad_norm": 0.27338715037770894,
+      "learning_rate": 0.00015699537698433532,
+      "loss": 0.0929,
+      "step": 2311
+    },
+    {
+      "epoch": 16.29603524229075,
+      "grad_norm": 0.2573514583698475,
+      "learning_rate": 0.00015695762055106704,
+      "loss": 0.0884,
+      "step": 2312
+    },
+    {
+      "epoch": 16.303083700440528,
+      "grad_norm": 0.26073297761572045,
+      "learning_rate": 0.00015691985209533761,
+      "loss": 0.0932,
+      "step": 2313
+    },
+    {
+      "epoch": 16.31013215859031,
+      "grad_norm": 0.26681001659537373,
+      "learning_rate": 0.00015688207162511906,
+      "loss": 0.0913,
+      "step": 2314
+    },
+    {
+      "epoch": 16.31718061674009,
+      "grad_norm": 0.260667189248561,
+      "learning_rate": 0.000156844279148386,
+      "loss": 0.0986,
+      "step": 2315
+    },
+    {
+      "epoch": 16.324229074889868,
+      "grad_norm": 0.2688285869455494,
+      "learning_rate": 0.00015680647467311557,
+      "loss": 0.0908,
+      "step": 2316
+    },
+    {
+      "epoch": 16.331277533039646,
+      "grad_norm": 0.24934611232093878,
+      "learning_rate": 0.0001567686582072875,
+      "loss": 0.094,
+      "step": 2317
+    },
+    {
+      "epoch": 16.33832599118943,
+      "grad_norm": 0.27581886658283905,
+      "learning_rate": 0.00015673082975888386,
+      "loss": 0.0928,
+      "step": 2318
+    },
+    {
+      "epoch": 16.345374449339207,
+      "grad_norm": 0.24999376608915766,
+      "learning_rate": 0.0001566929893358895,
+      "loss": 0.091,
+      "step": 2319
+    },
+    {
+      "epoch": 16.352422907488986,
+      "grad_norm": 0.25867334430434363,
+      "learning_rate": 0.00015665513694629158,
+      "loss": 0.0921,
+      "step": 2320
+    },
+    {
+      "epoch": 16.35947136563877,
+      "grad_norm": 0.26577766504676986,
+      "learning_rate": 0.00015661727259807994,
+      "loss": 0.0878,
+      "step": 2321
+    },
+    {
+      "epoch": 16.366519823788547,
+      "grad_norm": 0.2598603656883206,
+      "learning_rate": 0.00015657939629924687,
+      "loss": 0.0947,
+      "step": 2322
+    },
+    {
+      "epoch": 16.373568281938326,
+      "grad_norm": 0.2533874594650539,
+      "learning_rate": 0.00015654150805778717,
+      "loss": 0.0957,
+      "step": 2323
+    },
+    {
+      "epoch": 16.380616740088104,
+      "grad_norm": 0.25107054885552105,
+      "learning_rate": 0.00015650360788169824,
+      "loss": 0.0987,
+      "step": 2324
+    },
+    {
+      "epoch": 16.387665198237887,
+      "grad_norm": 0.2519961244990702,
+      "learning_rate": 0.0001564656957789799,
+      "loss": 0.0914,
+      "step": 2325
+    },
+    {
+      "epoch": 16.394713656387665,
+      "grad_norm": 0.2545933816201111,
+      "learning_rate": 0.00015642777175763455,
+      "loss": 0.0926,
+      "step": 2326
+    },
+    {
+      "epoch": 16.401762114537444,
+      "grad_norm": 0.2672830114833597,
+      "learning_rate": 0.00015638983582566708,
+      "loss": 0.0945,
+      "step": 2327
+    },
+    {
+      "epoch": 16.408810572687223,
+      "grad_norm": 0.2553591685991786,
+      "learning_rate": 0.00015635188799108495,
+      "loss": 0.0854,
+      "step": 2328
+    },
+    {
+      "epoch": 16.415859030837005,
+      "grad_norm": 0.27417802355843685,
+      "learning_rate": 0.00015631392826189803,
+      "loss": 0.0989,
+      "step": 2329
+    },
+    {
+      "epoch": 16.422907488986784,
+      "grad_norm": 0.26852183133879703,
+      "learning_rate": 0.0001562759566461188,
+      "loss": 0.0932,
+      "step": 2330
+    },
+    {
+      "epoch": 16.429955947136563,
+      "grad_norm": 0.2588506188625754,
+      "learning_rate": 0.00015623797315176218,
+      "loss": 0.0967,
+      "step": 2331
+    },
+    {
+      "epoch": 16.437004405286345,
+      "grad_norm": 0.25396047734542715,
+      "learning_rate": 0.00015619997778684562,
+      "loss": 0.0884,
+      "step": 2332
+    },
+    {
+      "epoch": 16.444052863436124,
+      "grad_norm": 0.26753400174976255,
+      "learning_rate": 0.0001561619705593891,
+      "loss": 0.1009,
+      "step": 2333
+    },
+    {
+      "epoch": 16.451101321585902,
+      "grad_norm": 0.2655329231744339,
+      "learning_rate": 0.00015612395147741507,
+      "loss": 0.0962,
+      "step": 2334
+    },
+    {
+      "epoch": 16.458149779735685,
+      "grad_norm": 0.2522039894571608,
+      "learning_rate": 0.00015608592054894852,
+      "loss": 0.1,
+      "step": 2335
+    },
+    {
+      "epoch": 16.465198237885463,
+      "grad_norm": 0.2914491567632894,
+      "learning_rate": 0.00015604787778201686,
+      "loss": 0.0996,
+      "step": 2336
+    },
+    {
+      "epoch": 16.472246696035242,
+      "grad_norm": 0.2730351808235724,
+      "learning_rate": 0.00015600982318465011,
+      "loss": 0.0958,
+      "step": 2337
+    },
+    {
+      "epoch": 16.47929515418502,
+      "grad_norm": 0.2715776801297249,
+      "learning_rate": 0.00015597175676488072,
+      "loss": 0.0965,
+      "step": 2338
+    },
+    {
+      "epoch": 16.486343612334803,
+      "grad_norm": 0.26378515185522,
+      "learning_rate": 0.00015593367853074364,
+      "loss": 0.0846,
+      "step": 2339
+    },
+    {
+      "epoch": 16.493392070484582,
+      "grad_norm": 0.2568446914280178,
+      "learning_rate": 0.0001558955884902763,
+      "loss": 0.0943,
+      "step": 2340
+    },
+    {
+      "epoch": 16.50044052863436,
+      "grad_norm": 0.2636046985720735,
+      "learning_rate": 0.00015585748665151867,
+      "loss": 0.0952,
+      "step": 2341
+    },
+    {
+      "epoch": 16.50748898678414,
+      "grad_norm": 0.2669039041372428,
+      "learning_rate": 0.00015581937302251318,
+      "loss": 0.0926,
+      "step": 2342
+    },
+    {
+      "epoch": 16.51453744493392,
+      "grad_norm": 0.2603477704689278,
+      "learning_rate": 0.00015578124761130474,
+      "loss": 0.0963,
+      "step": 2343
+    },
+    {
+      "epoch": 16.5215859030837,
+      "grad_norm": 0.2874321453939778,
+      "learning_rate": 0.0001557431104259408,
+      "loss": 0.0932,
+      "step": 2344
+    },
+    {
+      "epoch": 16.52863436123348,
+      "grad_norm": 0.2609839592530272,
+      "learning_rate": 0.0001557049614744712,
+      "loss": 0.0951,
+      "step": 2345
+    },
+    {
+      "epoch": 16.53568281938326,
+      "grad_norm": 0.2598002313332106,
+      "learning_rate": 0.00015566680076494832,
+      "loss": 0.1,
+      "step": 2346
+    },
+    {
+      "epoch": 16.54273127753304,
+      "grad_norm": 0.26162408855866875,
+      "learning_rate": 0.00015562862830542708,
+      "loss": 0.0934,
+      "step": 2347
+    },
+    {
+      "epoch": 16.54977973568282,
+      "grad_norm": 0.2692275374979925,
+      "learning_rate": 0.00015559044410396475,
+      "loss": 0.102,
+      "step": 2348
+    },
+    {
+      "epoch": 16.556828193832597,
+      "grad_norm": 0.24961665437971012,
+      "learning_rate": 0.00015555224816862122,
+      "loss": 0.0965,
+      "step": 2349
+    },
+    {
+      "epoch": 16.56387665198238,
+      "grad_norm": 0.2553098563005212,
+      "learning_rate": 0.0001555140405074587,
+      "loss": 0.0969,
+      "step": 2350
+    },
+    {
+      "epoch": 16.57092511013216,
+      "grad_norm": 0.2625807962921251,
+      "learning_rate": 0.00015547582112854203,
+      "loss": 0.0991,
+      "step": 2351
+    },
+    {
+      "epoch": 16.577973568281937,
+      "grad_norm": 0.2614026294833608,
+      "learning_rate": 0.0001554375900399384,
+      "loss": 0.0964,
+      "step": 2352
+    },
+    {
+      "epoch": 16.58502202643172,
+      "grad_norm": 0.27388169728885864,
+      "learning_rate": 0.00015539934724971758,
+      "loss": 0.0947,
+      "step": 2353
+    },
+    {
+      "epoch": 16.592070484581498,
+      "grad_norm": 0.26946881986869353,
+      "learning_rate": 0.00015536109276595173,
+      "loss": 0.0943,
+      "step": 2354
+    },
+    {
+      "epoch": 16.599118942731277,
+      "grad_norm": 0.2813963901642478,
+      "learning_rate": 0.00015532282659671547,
+      "loss": 0.102,
+      "step": 2355
+    },
+    {
+      "epoch": 16.606167400881056,
+      "grad_norm": 0.27633673993346214,
+      "learning_rate": 0.00015528454875008598,
+      "loss": 0.099,
+      "step": 2356
+    },
+    {
+      "epoch": 16.613215859030838,
+      "grad_norm": 0.2636941218435272,
+      "learning_rate": 0.00015524625923414283,
+      "loss": 0.094,
+      "step": 2357
+    },
+    {
+      "epoch": 16.620264317180617,
+      "grad_norm": 0.25831265440862056,
+      "learning_rate": 0.00015520795805696802,
+      "loss": 0.0935,
+      "step": 2358
+    },
+    {
+      "epoch": 16.627312775330395,
+      "grad_norm": 0.26375605319780937,
+      "learning_rate": 0.0001551696452266461,
+      "loss": 0.0979,
+      "step": 2359
+    },
+    {
+      "epoch": 16.634361233480178,
+      "grad_norm": 0.2858555843941885,
+      "learning_rate": 0.00015513132075126404,
+      "loss": 0.1017,
+      "step": 2360
+    },
+    {
+      "epoch": 16.641409691629956,
+      "grad_norm": 0.27460203139111544,
+      "learning_rate": 0.00015509298463891127,
+      "loss": 0.1056,
+      "step": 2361
+    },
+    {
+      "epoch": 16.648458149779735,
+      "grad_norm": 0.27368565871821143,
+      "learning_rate": 0.00015505463689767965,
+      "loss": 0.1008,
+      "step": 2362
+    },
+    {
+      "epoch": 16.655506607929514,
+      "grad_norm": 0.2621522984927399,
+      "learning_rate": 0.0001550162775356635,
+      "loss": 0.1018,
+      "step": 2363
+    },
+    {
+      "epoch": 16.662555066079296,
+      "grad_norm": 0.2618791021505065,
+      "learning_rate": 0.0001549779065609597,
+      "loss": 0.0974,
+      "step": 2364
+    },
+    {
+      "epoch": 16.669603524229075,
+      "grad_norm": 0.27774598065881817,
+      "learning_rate": 0.00015493952398166742,
+      "loss": 0.102,
+      "step": 2365
+    },
+    {
+      "epoch": 16.676651982378853,
+      "grad_norm": 0.2677761821947088,
+      "learning_rate": 0.00015490112980588835,
+      "loss": 0.1041,
+      "step": 2366
+    },
+    {
+      "epoch": 16.683700440528636,
+      "grad_norm": 0.2697396713906099,
+      "learning_rate": 0.00015486272404172666,
+      "loss": 0.0996,
+      "step": 2367
+    },
+    {
+      "epoch": 16.690748898678414,
+      "grad_norm": 0.2925505594439383,
+      "learning_rate": 0.0001548243066972889,
+      "loss": 0.0994,
+      "step": 2368
+    },
+    {
+      "epoch": 16.697797356828193,
+      "grad_norm": 0.26904199498541176,
+      "learning_rate": 0.00015478587778068413,
+      "loss": 0.1055,
+      "step": 2369
+    },
+    {
+      "epoch": 16.704845814977972,
+      "grad_norm": 0.28513857259238523,
+      "learning_rate": 0.0001547474373000238,
+      "loss": 0.1104,
+      "step": 2370
+    },
+    {
+      "epoch": 16.711894273127754,
+      "grad_norm": 0.2728381416426999,
+      "learning_rate": 0.00015470898526342186,
+      "loss": 0.0945,
+      "step": 2371
+    },
+    {
+      "epoch": 16.718942731277533,
+      "grad_norm": 0.2589230758792151,
+      "learning_rate": 0.00015467052167899463,
+      "loss": 0.095,
+      "step": 2372
+    },
+    {
+      "epoch": 16.72599118942731,
+      "grad_norm": 0.28283062396449243,
+      "learning_rate": 0.0001546320465548609,
+      "loss": 0.1006,
+      "step": 2373
+    },
+    {
+      "epoch": 16.733039647577094,
+      "grad_norm": 0.2821324780171604,
+      "learning_rate": 0.0001545935598991419,
+      "loss": 0.1011,
+      "step": 2374
+    },
+    {
+      "epoch": 16.740088105726873,
+      "grad_norm": 0.2816345197549968,
+      "learning_rate": 0.0001545550617199613,
+      "loss": 0.1023,
+      "step": 2375
+    },
+    {
+      "epoch": 16.74713656387665,
+      "grad_norm": 0.2642507555630299,
+      "learning_rate": 0.00015451655202544518,
+      "loss": 0.1074,
+      "step": 2376
+    },
+    {
+      "epoch": 16.75418502202643,
+      "grad_norm": 0.29133091865321903,
+      "learning_rate": 0.00015447803082372207,
+      "loss": 0.1022,
+      "step": 2377
+    },
+    {
+      "epoch": 16.761233480176212,
+      "grad_norm": 0.2915276035431784,
+      "learning_rate": 0.0001544394981229229,
+      "loss": 0.0997,
+      "step": 2378
+    },
+    {
+      "epoch": 16.76828193832599,
+      "grad_norm": 0.2545774684795472,
+      "learning_rate": 0.00015440095393118113,
+      "loss": 0.0996,
+      "step": 2379
+    },
+    {
+      "epoch": 16.77533039647577,
+      "grad_norm": 0.27835559923224557,
+      "learning_rate": 0.00015436239825663244,
+      "loss": 0.1011,
+      "step": 2380
+    },
+    {
+      "epoch": 16.782378854625552,
+      "grad_norm": 0.27481042739395595,
+      "learning_rate": 0.00015432383110741514,
+      "loss": 0.1046,
+      "step": 2381
+    },
+    {
+      "epoch": 16.78942731277533,
+      "grad_norm": 0.27422213264580986,
+      "learning_rate": 0.00015428525249166984,
+      "loss": 0.098,
+      "step": 2382
+    },
+    {
+      "epoch": 16.79647577092511,
+      "grad_norm": 0.27890890533058815,
+      "learning_rate": 0.00015424666241753966,
+      "loss": 0.1058,
+      "step": 2383
+    },
+    {
+      "epoch": 16.80352422907489,
+      "grad_norm": 0.2704498177470123,
+      "learning_rate": 0.00015420806089317,
+      "loss": 0.1021,
+      "step": 2384
+    },
+    {
+      "epoch": 16.81057268722467,
+      "grad_norm": 0.2753190473527988,
+      "learning_rate": 0.0001541694479267089,
+      "loss": 0.1081,
+      "step": 2385
+    },
+    {
+      "epoch": 16.81762114537445,
+      "grad_norm": 0.29154166242693436,
+      "learning_rate": 0.0001541308235263066,
+      "loss": 0.102,
+      "step": 2386
+    },
+    {
+      "epoch": 16.824669603524228,
+      "grad_norm": 0.27865902918919555,
+      "learning_rate": 0.00015409218770011577,
+      "loss": 0.1074,
+      "step": 2387
+    },
+    {
+      "epoch": 16.83171806167401,
+      "grad_norm": 0.283047850225271,
+      "learning_rate": 0.00015405354045629167,
+      "loss": 0.1065,
+      "step": 2388
+    },
+    {
+      "epoch": 16.83876651982379,
+      "grad_norm": 0.26861556993200136,
+      "learning_rate": 0.00015401488180299182,
+      "loss": 0.1033,
+      "step": 2389
+    },
+    {
+      "epoch": 16.845814977973568,
+      "grad_norm": 0.2634867344948992,
+      "learning_rate": 0.00015397621174837616,
+      "loss": 0.1042,
+      "step": 2390
+    },
+    {
+      "epoch": 16.852863436123346,
+      "grad_norm": 0.28459726851028533,
+      "learning_rate": 0.00015393753030060707,
+      "loss": 0.109,
+      "step": 2391
+    },
+    {
+      "epoch": 16.85991189427313,
+      "grad_norm": 0.28637011968916964,
+      "learning_rate": 0.00015389883746784938,
+      "loss": 0.1087,
+      "step": 2392
+    },
+    {
+      "epoch": 16.866960352422907,
+      "grad_norm": 0.2831332708465193,
+      "learning_rate": 0.0001538601332582702,
+      "loss": 0.1139,
+      "step": 2393
+    },
+    {
+      "epoch": 16.874008810572686,
+      "grad_norm": 0.27076611946906165,
+      "learning_rate": 0.00015382141768003908,
+      "loss": 0.1092,
+      "step": 2394
+    },
+    {
+      "epoch": 16.88105726872247,
+      "grad_norm": 0.26586741880495823,
+      "learning_rate": 0.00015378269074132813,
+      "loss": 0.1014,
+      "step": 2395
+    },
+    {
+      "epoch": 16.888105726872247,
+      "grad_norm": 0.2592564359569745,
+      "learning_rate": 0.0001537439524503116,
+      "loss": 0.1029,
+      "step": 2396
+    },
+    {
+      "epoch": 16.895154185022026,
+      "grad_norm": 0.27248092208830893,
+      "learning_rate": 0.00015370520281516635,
+      "loss": 0.1055,
+      "step": 2397
+    },
+    {
+      "epoch": 16.902202643171805,
+      "grad_norm": 0.26539442695698306,
+      "learning_rate": 0.00015366644184407145,
+      "loss": 0.1009,
+      "step": 2398
+    },
+    {
+      "epoch": 16.909251101321587,
+      "grad_norm": 0.2823707007260589,
+      "learning_rate": 0.0001536276695452086,
+      "loss": 0.0992,
+      "step": 2399
+    },
+    {
+      "epoch": 16.916299559471366,
+      "grad_norm": 0.295996109498354,
+      "learning_rate": 0.00015358888592676163,
+      "loss": 0.1052,
+      "step": 2400
+    },
+    {
+      "epoch": 16.923348017621144,
+      "grad_norm": 0.27769536062193767,
+      "learning_rate": 0.00015355009099691697,
+      "loss": 0.1096,
+      "step": 2401
+    },
+    {
+      "epoch": 16.930396475770927,
+      "grad_norm": 0.2806537546538495,
+      "learning_rate": 0.00015351128476386326,
+      "loss": 0.1099,
+      "step": 2402
+    },
+    {
+      "epoch": 16.937444933920705,
+      "grad_norm": 0.28623311993385386,
+      "learning_rate": 0.0001534724672357917,
+      "loss": 0.1101,
+      "step": 2403
+    },
+    {
+      "epoch": 16.944493392070484,
+      "grad_norm": 0.2800556261254593,
+      "learning_rate": 0.00015343363842089573,
+      "loss": 0.106,
+      "step": 2404
+    },
+    {
+      "epoch": 16.951541850220263,
+      "grad_norm": 0.27010506193560224,
+      "learning_rate": 0.00015339479832737128,
+      "loss": 0.1126,
+      "step": 2405
+    },
+    {
+      "epoch": 16.958590308370045,
+      "grad_norm": 0.2701223105495599,
+      "learning_rate": 0.00015335594696341657,
+      "loss": 0.1034,
+      "step": 2406
+    },
+    {
+      "epoch": 16.965638766519824,
+      "grad_norm": 0.2884995758552436,
+      "learning_rate": 0.00015331708433723225,
+      "loss": 0.1083,
+      "step": 2407
+    },
+    {
+      "epoch": 16.972687224669603,
+      "grad_norm": 0.27877383954299373,
+      "learning_rate": 0.00015327821045702135,
+      "loss": 0.1073,
+      "step": 2408
+    },
+    {
+      "epoch": 16.979735682819385,
+      "grad_norm": 0.27946799205985234,
+      "learning_rate": 0.00015323932533098925,
+      "loss": 0.1065,
+      "step": 2409
+    },
+    {
+      "epoch": 16.986784140969164,
+      "grad_norm": 0.28408929877299355,
+      "learning_rate": 0.00015320042896734375,
+      "loss": 0.1054,
+      "step": 2410
+    },
+    {
+      "epoch": 16.993832599118942,
+      "grad_norm": 0.2764479097878727,
+      "learning_rate": 0.000153161521374295,
+      "loss": 0.1093,
+      "step": 2411
+    },
+    {
+      "epoch": 17.00088105726872,
+      "grad_norm": 0.26786911240526534,
+      "learning_rate": 0.00015312260256005543,
+      "loss": 0.1026,
+      "step": 2412
+    },
+    {
+      "epoch": 17.007929515418503,
+      "grad_norm": 0.2363061536865926,
+      "learning_rate": 0.00015308367253283995,
+      "loss": 0.0653,
+      "step": 2413
+    },
+    {
+      "epoch": 17.014977973568282,
+      "grad_norm": 0.21965365693666977,
+      "learning_rate": 0.00015304473130086589,
+      "loss": 0.0684,
+      "step": 2414
+    },
+    {
+      "epoch": 17.02202643171806,
+      "grad_norm": 0.25072014604287596,
+      "learning_rate": 0.0001530057788723527,
+      "loss": 0.0639,
+      "step": 2415
+    },
+    {
+      "epoch": 17.029074889867843,
+      "grad_norm": 0.273877344933339,
+      "learning_rate": 0.00015296681525552253,
+      "loss": 0.0693,
+      "step": 2416
+    },
+    {
+      "epoch": 17.03612334801762,
+      "grad_norm": 0.25409090939596024,
+      "learning_rate": 0.0001529278404585996,
+      "loss": 0.0674,
+      "step": 2417
+    },
+    {
+      "epoch": 17.0431718061674,
+      "grad_norm": 0.24602285533410878,
+      "learning_rate": 0.00015288885448981062,
+      "loss": 0.0613,
+      "step": 2418
+    },
+    {
+      "epoch": 17.05022026431718,
+      "grad_norm": 0.2412065087791869,
+      "learning_rate": 0.00015284985735738467,
+      "loss": 0.0636,
+      "step": 2419
+    },
+    {
+      "epoch": 17.05726872246696,
+      "grad_norm": 0.236610306007699,
+      "learning_rate": 0.00015281084906955316,
+      "loss": 0.0676,
+      "step": 2420
+    },
+    {
+      "epoch": 17.06431718061674,
+      "grad_norm": 0.22267940205146602,
+      "learning_rate": 0.0001527718296345498,
+      "loss": 0.0656,
+      "step": 2421
+    },
+    {
+      "epoch": 17.07136563876652,
+      "grad_norm": 0.2594778840361566,
+      "learning_rate": 0.00015273279906061082,
+      "loss": 0.0631,
+      "step": 2422
+    },
+    {
+      "epoch": 17.0784140969163,
+      "grad_norm": 0.24159606031841868,
+      "learning_rate": 0.0001526937573559746,
+      "loss": 0.0695,
+      "step": 2423
+    },
+    {
+      "epoch": 17.08546255506608,
+      "grad_norm": 0.25285067814766027,
+      "learning_rate": 0.00015265470452888194,
+      "loss": 0.0683,
+      "step": 2424
+    },
+    {
+      "epoch": 17.09251101321586,
+      "grad_norm": 0.24128938750827938,
+      "learning_rate": 0.00015261564058757605,
+      "loss": 0.0641,
+      "step": 2425
+    },
+    {
+      "epoch": 17.099559471365637,
+      "grad_norm": 0.24396843917803138,
+      "learning_rate": 0.00015257656554030247,
+      "loss": 0.063,
+      "step": 2426
+    },
+    {
+      "epoch": 17.10660792951542,
+      "grad_norm": 0.2329833164626736,
+      "learning_rate": 0.000152537479395309,
+      "loss": 0.0601,
+      "step": 2427
+    },
+    {
+      "epoch": 17.1136563876652,
+      "grad_norm": 0.245516572482844,
+      "learning_rate": 0.00015249838216084585,
+      "loss": 0.0707,
+      "step": 2428
+    },
+    {
+      "epoch": 17.120704845814977,
+      "grad_norm": 0.2305339995779955,
+      "learning_rate": 0.0001524592738451656,
+      "loss": 0.0619,
+      "step": 2429
+    },
+    {
+      "epoch": 17.12775330396476,
+      "grad_norm": 0.23576416802991856,
+      "learning_rate": 0.00015242015445652305,
+      "loss": 0.0683,
+      "step": 2430
+    },
+    {
+      "epoch": 17.134801762114538,
+      "grad_norm": 0.23953018628100134,
+      "learning_rate": 0.0001523810240031755,
+      "loss": 0.0711,
+      "step": 2431
+    },
+    {
+      "epoch": 17.141850220264317,
+      "grad_norm": 0.24453666004625055,
+      "learning_rate": 0.00015234188249338242,
+      "loss": 0.0626,
+      "step": 2432
+    },
+    {
+      "epoch": 17.148898678414096,
+      "grad_norm": 0.262807924280765,
+      "learning_rate": 0.00015230272993540579,
+      "loss": 0.0641,
+      "step": 2433
+    },
+    {
+      "epoch": 17.155947136563878,
+      "grad_norm": 0.23382990363958858,
+      "learning_rate": 0.00015226356633750974,
+      "loss": 0.0633,
+      "step": 2434
+    },
+    {
+      "epoch": 17.162995594713657,
+      "grad_norm": 0.2532749277684863,
+      "learning_rate": 0.0001522243917079608,
+      "loss": 0.0664,
+      "step": 2435
+    },
+    {
+      "epoch": 17.170044052863435,
+      "grad_norm": 0.2515910388251906,
+      "learning_rate": 0.00015218520605502796,
+      "loss": 0.0612,
+      "step": 2436
+    },
+    {
+      "epoch": 17.177092511013218,
+      "grad_norm": 0.2408680863758088,
+      "learning_rate": 0.00015214600938698232,
+      "loss": 0.0656,
+      "step": 2437
+    },
+    {
+      "epoch": 17.184140969162996,
+      "grad_norm": 0.23029310997888716,
+      "learning_rate": 0.00015210680171209748,
+      "loss": 0.0667,
+      "step": 2438
+    },
+    {
+      "epoch": 17.191189427312775,
+      "grad_norm": 0.22703989696767762,
+      "learning_rate": 0.00015206758303864924,
+      "loss": 0.0623,
+      "step": 2439
+    },
+    {
+      "epoch": 17.198237885462554,
+      "grad_norm": 0.2434370711630896,
+      "learning_rate": 0.0001520283533749158,
+      "loss": 0.066,
+      "step": 2440
+    },
+    {
+      "epoch": 17.205286343612336,
+      "grad_norm": 0.25741076245314426,
+      "learning_rate": 0.0001519891127291776,
+      "loss": 0.0673,
+      "step": 2441
+    },
+    {
+      "epoch": 17.212334801762115,
+      "grad_norm": 0.23134554520775377,
+      "learning_rate": 0.0001519498611097175,
+      "loss": 0.0695,
+      "step": 2442
+    },
+    {
+      "epoch": 17.219383259911893,
+      "grad_norm": 0.25918914432893014,
+      "learning_rate": 0.00015191059852482063,
+      "loss": 0.0733,
+      "step": 2443
+    },
+    {
+      "epoch": 17.226431718061676,
+      "grad_norm": 0.24423493431261295,
+      "learning_rate": 0.0001518713249827744,
+      "loss": 0.0608,
+      "step": 2444
+    },
+    {
+      "epoch": 17.233480176211454,
+      "grad_norm": 0.2386778462399181,
+      "learning_rate": 0.00015183204049186859,
+      "loss": 0.0661,
+      "step": 2445
+    },
+    {
+      "epoch": 17.240528634361233,
+      "grad_norm": 0.2457141667978051,
+      "learning_rate": 0.00015179274506039526,
+      "loss": 0.0636,
+      "step": 2446
+    },
+    {
+      "epoch": 17.247577092511012,
+      "grad_norm": 0.25711877103466135,
+      "learning_rate": 0.00015175343869664875,
+      "loss": 0.0712,
+      "step": 2447
+    },
+    {
+      "epoch": 17.254625550660794,
+      "grad_norm": 0.22557163936860009,
+      "learning_rate": 0.00015171412140892575,
+      "loss": 0.0671,
+      "step": 2448
+    },
+    {
+      "epoch": 17.261674008810573,
+      "grad_norm": 0.24746865980309443,
+      "learning_rate": 0.00015167479320552532,
+      "loss": 0.0596,
+      "step": 2449
+    },
+    {
+      "epoch": 17.26872246696035,
+      "grad_norm": 0.2349192276142657,
+      "learning_rate": 0.00015163545409474865,
+      "loss": 0.0692,
+      "step": 2450
+    },
+    {
+      "epoch": 17.275770925110134,
+      "grad_norm": 0.23361968504978045,
+      "learning_rate": 0.0001515961040848994,
+      "loss": 0.0665,
+      "step": 2451
+    },
+    {
+      "epoch": 17.282819383259913,
+      "grad_norm": 0.24646646332841055,
+      "learning_rate": 0.00015155674318428347,
+      "loss": 0.064,
+      "step": 2452
+    },
+    {
+      "epoch": 17.28986784140969,
+      "grad_norm": 0.2304392495861939,
+      "learning_rate": 0.000151517371401209,
+      "loss": 0.0657,
+      "step": 2453
+    },
+    {
+      "epoch": 17.29691629955947,
+      "grad_norm": 0.256461333142206,
+      "learning_rate": 0.00015147798874398652,
+      "loss": 0.0684,
+      "step": 2454
+    },
+    {
+      "epoch": 17.303964757709252,
+      "grad_norm": 0.24800955386943596,
+      "learning_rate": 0.00015143859522092877,
+      "loss": 0.0649,
+      "step": 2455
+    },
+    {
+      "epoch": 17.31101321585903,
+      "grad_norm": 0.26490398604355664,
+      "learning_rate": 0.00015139919084035088,
+      "loss": 0.0694,
+      "step": 2456
+    },
+    {
+      "epoch": 17.31806167400881,
+      "grad_norm": 0.2504799970513583,
+      "learning_rate": 0.0001513597756105702,
+      "loss": 0.0654,
+      "step": 2457
+    },
+    {
+      "epoch": 17.325110132158592,
+      "grad_norm": 0.2468056291651354,
+      "learning_rate": 0.00015132034953990639,
+      "loss": 0.0665,
+      "step": 2458
+    },
+    {
+      "epoch": 17.33215859030837,
+      "grad_norm": 0.2383079441087073,
+      "learning_rate": 0.0001512809126366814,
+      "loss": 0.0681,
+      "step": 2459
+    },
+    {
+      "epoch": 17.33920704845815,
+      "grad_norm": 0.24630649232962443,
+      "learning_rate": 0.00015124146490921948,
+      "loss": 0.0657,
+      "step": 2460
+    },
+    {
+      "epoch": 17.34625550660793,
+      "grad_norm": 0.25110953069273956,
+      "learning_rate": 0.0001512020063658471,
+      "loss": 0.0728,
+      "step": 2461
+    },
+    {
+      "epoch": 17.35330396475771,
+      "grad_norm": 0.23695615384306964,
+      "learning_rate": 0.00015116253701489313,
+      "loss": 0.0671,
+      "step": 2462
+    },
+    {
+      "epoch": 17.36035242290749,
+      "grad_norm": 0.2475553303521458,
+      "learning_rate": 0.00015112305686468863,
+      "loss": 0.0655,
+      "step": 2463
+    },
+    {
+      "epoch": 17.367400881057268,
+      "grad_norm": 0.24980217547111092,
+      "learning_rate": 0.00015108356592356695,
+      "loss": 0.0628,
+      "step": 2464
+    },
+    {
+      "epoch": 17.374449339207047,
+      "grad_norm": 0.24516905137828562,
+      "learning_rate": 0.00015104406419986369,
+      "loss": 0.0713,
+      "step": 2465
+    },
+    {
+      "epoch": 17.38149779735683,
+      "grad_norm": 0.2668397586287365,
+      "learning_rate": 0.00015100455170191686,
+      "loss": 0.0699,
+      "step": 2466
+    },
+    {
+      "epoch": 17.388546255506608,
+      "grad_norm": 0.24201230517475944,
+      "learning_rate": 0.00015096502843806663,
+      "loss": 0.0713,
+      "step": 2467
+    },
+    {
+      "epoch": 17.395594713656386,
+      "grad_norm": 0.2499480238723658,
+      "learning_rate": 0.00015092549441665543,
+      "loss": 0.0683,
+      "step": 2468
+    },
+    {
+      "epoch": 17.40264317180617,
+      "grad_norm": 0.25843048004201297,
+      "learning_rate": 0.00015088594964602798,
+      "loss": 0.0684,
+      "step": 2469
+    },
+    {
+      "epoch": 17.409691629955947,
+      "grad_norm": 0.2582883980688514,
+      "learning_rate": 0.00015084639413453133,
+      "loss": 0.0744,
+      "step": 2470
+    },
+    {
+      "epoch": 17.416740088105726,
+      "grad_norm": 0.2553498789592326,
+      "learning_rate": 0.00015080682789051473,
+      "loss": 0.0749,
+      "step": 2471
+    },
+    {
+      "epoch": 17.423788546255505,
+      "grad_norm": 0.25068654141825036,
+      "learning_rate": 0.00015076725092232974,
+      "loss": 0.0735,
+      "step": 2472
+    },
+    {
+      "epoch": 17.430837004405287,
+      "grad_norm": 0.2575027694561255,
+      "learning_rate": 0.00015072766323833006,
+      "loss": 0.0683,
+      "step": 2473
+    },
+    {
+      "epoch": 17.437885462555066,
+      "grad_norm": 0.24692908884892747,
+      "learning_rate": 0.0001506880648468719,
+      "loss": 0.0667,
+      "step": 2474
+    },
+    {
+      "epoch": 17.444933920704845,
+      "grad_norm": 0.24123019456294653,
+      "learning_rate": 0.0001506484557563135,
+      "loss": 0.0684,
+      "step": 2475
+    },
+    {
+      "epoch": 17.451982378854627,
+      "grad_norm": 0.24405050403997353,
+      "learning_rate": 0.00015060883597501544,
+      "loss": 0.0714,
+      "step": 2476
+    },
+    {
+      "epoch": 17.459030837004406,
+      "grad_norm": 0.23803491637724367,
+      "learning_rate": 0.00015056920551134059,
+      "loss": 0.0688,
+      "step": 2477
+    },
+    {
+      "epoch": 17.466079295154184,
+      "grad_norm": 0.2613230294172996,
+      "learning_rate": 0.00015052956437365395,
+      "loss": 0.0716,
+      "step": 2478
+    },
+    {
+      "epoch": 17.473127753303963,
+      "grad_norm": 0.262308212950135,
+      "learning_rate": 0.000150489912570323,
+      "loss": 0.0716,
+      "step": 2479
+    },
+    {
+      "epoch": 17.480176211453745,
+      "grad_norm": 0.24105969021401522,
+      "learning_rate": 0.00015045025010971725,
+      "loss": 0.0698,
+      "step": 2480
+    },
+    {
+      "epoch": 17.487224669603524,
+      "grad_norm": 0.24442502622313841,
+      "learning_rate": 0.00015041057700020856,
+      "loss": 0.0719,
+      "step": 2481
+    },
+    {
+      "epoch": 17.494273127753303,
+      "grad_norm": 0.2585095631666596,
+      "learning_rate": 0.00015037089325017105,
+      "loss": 0.0678,
+      "step": 2482
+    },
+    {
+      "epoch": 17.501321585903085,
+      "grad_norm": 0.24751557320867068,
+      "learning_rate": 0.00015033119886798102,
+      "loss": 0.0697,
+      "step": 2483
+    },
+    {
+      "epoch": 17.508370044052864,
+      "grad_norm": 0.25356512524101354,
+      "learning_rate": 0.00015029149386201705,
+      "loss": 0.0693,
+      "step": 2484
+    },
+    {
+      "epoch": 17.515418502202643,
+      "grad_norm": 0.23902409394647722,
+      "learning_rate": 0.00015025177824066005,
+      "loss": 0.0718,
+      "step": 2485
+    },
+    {
+      "epoch": 17.52246696035242,
+      "grad_norm": 0.24226174224654298,
+      "learning_rate": 0.00015021205201229298,
+      "loss": 0.0758,
+      "step": 2486
+    },
+    {
+      "epoch": 17.529515418502204,
+      "grad_norm": 0.24299811811054367,
+      "learning_rate": 0.00015017231518530118,
+      "loss": 0.0659,
+      "step": 2487
+    },
+    {
+      "epoch": 17.536563876651982,
+      "grad_norm": 0.24865629177074047,
+      "learning_rate": 0.00015013256776807223,
+      "loss": 0.072,
+      "step": 2488
+    },
+    {
+      "epoch": 17.54361233480176,
+      "grad_norm": 0.2454088391857943,
+      "learning_rate": 0.00015009280976899583,
+      "loss": 0.0696,
+      "step": 2489
+    },
+    {
+      "epoch": 17.550660792951543,
+      "grad_norm": 0.24800280722617657,
+      "learning_rate": 0.00015005304119646405,
+      "loss": 0.0687,
+      "step": 2490
+    },
+    {
+      "epoch": 17.557709251101322,
+      "grad_norm": 0.2444533014400577,
+      "learning_rate": 0.00015001326205887116,
+      "loss": 0.0735,
+      "step": 2491
+    },
+    {
+      "epoch": 17.5647577092511,
+      "grad_norm": 0.2745869223134913,
+      "learning_rate": 0.0001499734723646136,
+      "loss": 0.0711,
+      "step": 2492
+    },
+    {
+      "epoch": 17.57180616740088,
+      "grad_norm": 0.26075063286840683,
+      "learning_rate": 0.00014993367212209004,
+      "loss": 0.0744,
+      "step": 2493
+    },
+    {
+      "epoch": 17.57885462555066,
+      "grad_norm": 0.25964110286569275,
+      "learning_rate": 0.00014989386133970142,
+      "loss": 0.0681,
+      "step": 2494
+    },
+    {
+      "epoch": 17.58590308370044,
+      "grad_norm": 0.24525234682417632,
+      "learning_rate": 0.00014985404002585095,
+      "loss": 0.0671,
+      "step": 2495
+    },
+    {
+      "epoch": 17.59295154185022,
+      "grad_norm": 0.26796396296029434,
+      "learning_rate": 0.00014981420818894393,
+      "loss": 0.0762,
+      "step": 2496
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 0.2561544553349705,
+      "learning_rate": 0.00014977436583738803,
+      "loss": 0.0715,
+      "step": 2497
+    },
+    {
+      "epoch": 17.60704845814978,
+      "grad_norm": 0.25413266135187085,
+      "learning_rate": 0.00014973451297959297,
+      "loss": 0.0785,
+      "step": 2498
+    },
+    {
+      "epoch": 17.61409691629956,
+      "grad_norm": 0.26388245666564303,
+      "learning_rate": 0.00014969464962397088,
+      "loss": 0.0717,
+      "step": 2499
+    },
+    {
+      "epoch": 17.621145374449338,
+      "grad_norm": 0.2528410761999648,
+      "learning_rate": 0.00014965477577893598,
+      "loss": 0.0707,
+      "step": 2500
+    },
+    {
+      "epoch": 17.62819383259912,
+      "grad_norm": 0.2527197799865387,
+      "learning_rate": 0.0001496148914529047,
+      "loss": 0.0685,
+      "step": 2501
+    },
+    {
+      "epoch": 17.6352422907489,
+      "grad_norm": 0.2568276937340587,
+      "learning_rate": 0.00014957499665429577,
+      "loss": 0.0776,
+      "step": 2502
+    },
+    {
+      "epoch": 17.642290748898677,
+      "grad_norm": 0.2564649119191465,
+      "learning_rate": 0.00014953509139153002,
+      "loss": 0.0763,
+      "step": 2503
+    },
+    {
+      "epoch": 17.64933920704846,
+      "grad_norm": 0.2658928411517382,
+      "learning_rate": 0.0001494951756730306,
+      "loss": 0.0716,
+      "step": 2504
+    },
+    {
+      "epoch": 17.65638766519824,
+      "grad_norm": 0.2542917080310929,
+      "learning_rate": 0.00014945524950722274,
+      "loss": 0.072,
+      "step": 2505
+    },
+    {
+      "epoch": 17.663436123348017,
+      "grad_norm": 0.26589032280979036,
+      "learning_rate": 0.000149415312902534,
+      "loss": 0.0767,
+      "step": 2506
+    },
+    {
+      "epoch": 17.670484581497796,
+      "grad_norm": 0.25085655182407096,
+      "learning_rate": 0.00014937536586739412,
+      "loss": 0.0709,
+      "step": 2507
+    },
+    {
+      "epoch": 17.677533039647578,
+      "grad_norm": 0.2631103906829883,
+      "learning_rate": 0.00014933540841023497,
+      "loss": 0.0771,
+      "step": 2508
+    },
+    {
+      "epoch": 17.684581497797357,
+      "grad_norm": 0.26625666192368874,
+      "learning_rate": 0.00014929544053949066,
+      "loss": 0.0758,
+      "step": 2509
+    },
+    {
+      "epoch": 17.691629955947135,
+      "grad_norm": 0.27439638212359213,
+      "learning_rate": 0.0001492554622635975,
+      "loss": 0.0848,
+      "step": 2510
+    },
+    {
+      "epoch": 17.698678414096918,
+      "grad_norm": 0.261414014919438,
+      "learning_rate": 0.00014921547359099402,
+      "loss": 0.0779,
+      "step": 2511
+    },
+    {
+      "epoch": 17.705726872246697,
+      "grad_norm": 0.24003767857048083,
+      "learning_rate": 0.00014917547453012088,
+      "loss": 0.0816,
+      "step": 2512
+    },
+    {
+      "epoch": 17.712775330396475,
+      "grad_norm": 0.26284215232378677,
+      "learning_rate": 0.00014913546508942105,
+      "loss": 0.0685,
+      "step": 2513
+    },
+    {
+      "epoch": 17.719823788546254,
+      "grad_norm": 0.2512130719582773,
+      "learning_rate": 0.00014909544527733952,
+      "loss": 0.0785,
+      "step": 2514
+    },
+    {
+      "epoch": 17.726872246696036,
+      "grad_norm": 0.2586133978878001,
+      "learning_rate": 0.00014905541510232362,
+      "loss": 0.0732,
+      "step": 2515
+    },
+    {
+      "epoch": 17.733920704845815,
+      "grad_norm": 0.25678232873229323,
+      "learning_rate": 0.00014901537457282285,
+      "loss": 0.0752,
+      "step": 2516
+    },
+    {
+      "epoch": 17.740969162995594,
+      "grad_norm": 0.26294144140384085,
+      "learning_rate": 0.00014897532369728876,
+      "loss": 0.0734,
+      "step": 2517
+    },
+    {
+      "epoch": 17.748017621145376,
+      "grad_norm": 0.2520084123566325,
+      "learning_rate": 0.00014893526248417523,
+      "loss": 0.0753,
+      "step": 2518
+    },
+    {
+      "epoch": 17.755066079295155,
+      "grad_norm": 0.27477214430850033,
+      "learning_rate": 0.00014889519094193828,
+      "loss": 0.0792,
+      "step": 2519
+    },
+    {
+      "epoch": 17.762114537444933,
+      "grad_norm": 0.2567250232453732,
+      "learning_rate": 0.0001488551090790361,
+      "loss": 0.0742,
+      "step": 2520
+    },
+    {
+      "epoch": 17.769162995594712,
+      "grad_norm": 0.2664024450458461,
+      "learning_rate": 0.00014881501690392908,
+      "loss": 0.0784,
+      "step": 2521
+    },
+    {
+      "epoch": 17.776211453744494,
+      "grad_norm": 0.2705661909305392,
+      "learning_rate": 0.00014877491442507972,
+      "loss": 0.068,
+      "step": 2522
+    },
+    {
+      "epoch": 17.783259911894273,
+      "grad_norm": 0.2744173043617039,
+      "learning_rate": 0.00014873480165095276,
+      "loss": 0.0795,
+      "step": 2523
+    },
+    {
+      "epoch": 17.790308370044052,
+      "grad_norm": 0.26733589194816926,
+      "learning_rate": 0.00014869467859001512,
+      "loss": 0.0812,
+      "step": 2524
+    },
+    {
+      "epoch": 17.797356828193834,
+      "grad_norm": 0.2740778052984229,
+      "learning_rate": 0.00014865454525073588,
+      "loss": 0.0743,
+      "step": 2525
+    },
+    {
+      "epoch": 17.804405286343613,
+      "grad_norm": 0.257881765509358,
+      "learning_rate": 0.0001486144016415862,
+      "loss": 0.0749,
+      "step": 2526
+    },
+    {
+      "epoch": 17.81145374449339,
+      "grad_norm": 0.27344230073946435,
+      "learning_rate": 0.00014857424777103958,
+      "loss": 0.076,
+      "step": 2527
+    },
+    {
+      "epoch": 17.81850220264317,
+      "grad_norm": 0.28173517079620736,
+      "learning_rate": 0.00014853408364757153,
+      "loss": 0.0737,
+      "step": 2528
+    },
+    {
+      "epoch": 17.825550660792953,
+      "grad_norm": 0.2721204221777276,
+      "learning_rate": 0.00014849390927965982,
+      "loss": 0.0769,
+      "step": 2529
+    },
+    {
+      "epoch": 17.83259911894273,
+      "grad_norm": 0.29180099955566563,
+      "learning_rate": 0.0001484537246757843,
+      "loss": 0.0849,
+      "step": 2530
+    },
+    {
+      "epoch": 17.83964757709251,
+      "grad_norm": 0.2551077300953649,
+      "learning_rate": 0.00014841352984442706,
+      "loss": 0.0776,
+      "step": 2531
+    },
+    {
+      "epoch": 17.846696035242292,
+      "grad_norm": 0.2742064542835756,
+      "learning_rate": 0.00014837332479407234,
+      "loss": 0.0758,
+      "step": 2532
+    },
+    {
+      "epoch": 17.85374449339207,
+      "grad_norm": 0.2717067895462266,
+      "learning_rate": 0.00014833310953320648,
+      "loss": 0.0746,
+      "step": 2533
+    },
+    {
+      "epoch": 17.86079295154185,
+      "grad_norm": 0.2618825127173391,
+      "learning_rate": 0.000148292884070318,
+      "loss": 0.0745,
+      "step": 2534
+    },
+    {
+      "epoch": 17.86784140969163,
+      "grad_norm": 0.2637514969962953,
+      "learning_rate": 0.0001482526484138976,
+      "loss": 0.0808,
+      "step": 2535
+    },
+    {
+      "epoch": 17.87488986784141,
+      "grad_norm": 0.2610650270416784,
+      "learning_rate": 0.0001482124025724381,
+      "loss": 0.0861,
+      "step": 2536
+    },
+    {
+      "epoch": 17.88193832599119,
+      "grad_norm": 0.2769938555668484,
+      "learning_rate": 0.00014817214655443454,
+      "loss": 0.08,
+      "step": 2537
+    },
+    {
+      "epoch": 17.888986784140968,
+      "grad_norm": 0.2823757098649871,
+      "learning_rate": 0.00014813188036838398,
+      "loss": 0.0816,
+      "step": 2538
+    },
+    {
+      "epoch": 17.89603524229075,
+      "grad_norm": 0.2659095361059502,
+      "learning_rate": 0.00014809160402278572,
+      "loss": 0.0796,
+      "step": 2539
+    },
+    {
+      "epoch": 17.90308370044053,
+      "grad_norm": 0.2625684267995063,
+      "learning_rate": 0.00014805131752614124,
+      "loss": 0.0802,
+      "step": 2540
+    },
+    {
+      "epoch": 17.910132158590308,
+      "grad_norm": 0.26999722593786185,
+      "learning_rate": 0.000148011020886954,
+      "loss": 0.0818,
+      "step": 2541
+    },
+    {
+      "epoch": 17.917180616740087,
+      "grad_norm": 0.2577976463369865,
+      "learning_rate": 0.00014797071411372977,
+      "loss": 0.0821,
+      "step": 2542
+    },
+    {
+      "epoch": 17.92422907488987,
+      "grad_norm": 0.29295251046623116,
+      "learning_rate": 0.0001479303972149764,
+      "loss": 0.082,
+      "step": 2543
+    },
+    {
+      "epoch": 17.931277533039648,
+      "grad_norm": 0.2801550851096799,
+      "learning_rate": 0.00014789007019920384,
+      "loss": 0.0778,
+      "step": 2544
+    },
+    {
+      "epoch": 17.938325991189426,
+      "grad_norm": 0.2664571436092863,
+      "learning_rate": 0.00014784973307492428,
+      "loss": 0.0772,
+      "step": 2545
+    },
+    {
+      "epoch": 17.94537444933921,
+      "grad_norm": 0.2517717795310493,
+      "learning_rate": 0.00014780938585065185,
+      "loss": 0.0781,
+      "step": 2546
+    },
+    {
+      "epoch": 17.952422907488987,
+      "grad_norm": 0.2766351640518911,
+      "learning_rate": 0.00014776902853490302,
+      "loss": 0.0786,
+      "step": 2547
+    },
+    {
+      "epoch": 17.959471365638766,
+      "grad_norm": 0.2624710868782872,
+      "learning_rate": 0.0001477286611361963,
+      "loss": 0.0766,
+      "step": 2548
+    },
+    {
+      "epoch": 17.966519823788545,
+      "grad_norm": 0.263719742749758,
+      "learning_rate": 0.00014768828366305232,
+      "loss": 0.0797,
+      "step": 2549
+    },
+    {
+      "epoch": 17.973568281938327,
+      "grad_norm": 0.28555768444732915,
+      "learning_rate": 0.00014764789612399382,
+      "loss": 0.0805,
+      "step": 2550
+    },
+    {
+      "epoch": 17.980616740088106,
+      "grad_norm": 0.2833319961279091,
+      "learning_rate": 0.00014760749852754573,
+      "loss": 0.0848,
+      "step": 2551
+    },
+    {
+      "epoch": 17.987665198237885,
+      "grad_norm": 0.276961379211044,
+      "learning_rate": 0.0001475670908822351,
+      "loss": 0.0784,
+      "step": 2552
+    },
+    {
+      "epoch": 17.994713656387667,
+      "grad_norm": 0.2683461247401077,
+      "learning_rate": 0.00014752667319659096,
+      "loss": 0.0805,
+      "step": 2553
+    },
+    {
+      "epoch": 18.001762114537446,
+      "grad_norm": 0.26741439187134697,
+      "learning_rate": 0.00014748624547914464,
+      "loss": 0.0787,
+      "step": 2554
+    },
+    {
+      "epoch": 18.008810572687224,
+      "grad_norm": 0.20103029898983918,
+      "learning_rate": 0.00014744580773842948,
+      "loss": 0.0445,
+      "step": 2555
+    },
+    {
+      "epoch": 18.015859030837003,
+      "grad_norm": 0.22336967787644701,
+      "learning_rate": 0.00014740535998298104,
+      "loss": 0.0479,
+      "step": 2556
+    },
+    {
+      "epoch": 18.022907488986785,
+      "grad_norm": 0.22072202645988667,
+      "learning_rate": 0.00014736490222133685,
+      "loss": 0.0449,
+      "step": 2557
+    },
+    {
+      "epoch": 18.029955947136564,
+      "grad_norm": 0.22746505153867164,
+      "learning_rate": 0.00014732443446203663,
+      "loss": 0.0515,
+      "step": 2558
+    },
+    {
+      "epoch": 18.037004405286343,
+      "grad_norm": 0.2329879060719801,
+      "learning_rate": 0.00014728395671362226,
+      "loss": 0.0452,
+      "step": 2559
+    },
+    {
+      "epoch": 18.044052863436125,
+      "grad_norm": 0.23418982599278243,
+      "learning_rate": 0.0001472434689846376,
+      "loss": 0.0469,
+      "step": 2560
+    },
+    {
+      "epoch": 18.051101321585904,
+      "grad_norm": 0.22071167926696084,
+      "learning_rate": 0.00014720297128362875,
+      "loss": 0.045,
+      "step": 2561
+    },
+    {
+      "epoch": 18.058149779735682,
+      "grad_norm": 0.20918331989468197,
+      "learning_rate": 0.00014716246361914386,
+      "loss": 0.0452,
+      "step": 2562
+    },
+    {
+      "epoch": 18.06519823788546,
+      "grad_norm": 0.23110618204112704,
+      "learning_rate": 0.00014712194599973311,
+      "loss": 0.0467,
+      "step": 2563
+    },
+    {
+      "epoch": 18.072246696035243,
+      "grad_norm": 0.21899022911208582,
+      "learning_rate": 0.00014708141843394892,
+      "loss": 0.0456,
+      "step": 2564
+    },
+    {
+      "epoch": 18.079295154185022,
+      "grad_norm": 0.22803850680969698,
+      "learning_rate": 0.0001470408809303457,
+      "loss": 0.0509,
+      "step": 2565
+    },
+    {
+      "epoch": 18.0863436123348,
+      "grad_norm": 0.2178695382826286,
+      "learning_rate": 0.00014700033349748,
+      "loss": 0.0488,
+      "step": 2566
+    },
+    {
+      "epoch": 18.093392070484583,
+      "grad_norm": 0.22523835631808237,
+      "learning_rate": 0.00014695977614391053,
+      "loss": 0.0458,
+      "step": 2567
+    },
+    {
+      "epoch": 18.100440528634362,
+      "grad_norm": 0.21536183560884603,
+      "learning_rate": 0.00014691920887819794,
+      "loss": 0.0485,
+      "step": 2568
+    },
+    {
+      "epoch": 18.10748898678414,
+      "grad_norm": 0.2204231613956373,
+      "learning_rate": 0.00014687863170890505,
+      "loss": 0.0472,
+      "step": 2569
+    },
+    {
+      "epoch": 18.11453744493392,
+      "grad_norm": 0.21886747459405245,
+      "learning_rate": 0.00014683804464459685,
+      "loss": 0.0402,
+      "step": 2570
+    },
+    {
+      "epoch": 18.1215859030837,
+      "grad_norm": 0.23233773571677108,
+      "learning_rate": 0.0001467974476938403,
+      "loss": 0.046,
+      "step": 2571
+    },
+    {
+      "epoch": 18.12863436123348,
+      "grad_norm": 0.22638753332366085,
+      "learning_rate": 0.00014675684086520453,
+      "loss": 0.045,
+      "step": 2572
+    },
+    {
+      "epoch": 18.13568281938326,
+      "grad_norm": 0.21935700649995132,
+      "learning_rate": 0.0001467162241672607,
+      "loss": 0.0478,
+      "step": 2573
+    },
+    {
+      "epoch": 18.14273127753304,
+      "grad_norm": 0.2123605752808834,
+      "learning_rate": 0.0001466755976085821,
+      "loss": 0.0441,
+      "step": 2574
+    },
+    {
+      "epoch": 18.14977973568282,
+      "grad_norm": 0.22601124214881313,
+      "learning_rate": 0.000146634961197744,
+      "loss": 0.0426,
+      "step": 2575
+    },
+    {
+      "epoch": 18.1568281938326,
+      "grad_norm": 0.21325403252510713,
+      "learning_rate": 0.00014659431494332392,
+      "loss": 0.0476,
+      "step": 2576
+    },
+    {
+      "epoch": 18.163876651982378,
+      "grad_norm": 0.2237606349607362,
+      "learning_rate": 0.0001465536588539013,
+      "loss": 0.048,
+      "step": 2577
+    },
+    {
+      "epoch": 18.17092511013216,
+      "grad_norm": 0.22474165130775361,
+      "learning_rate": 0.00014651299293805774,
+      "loss": 0.0475,
+      "step": 2578
+    },
+    {
+      "epoch": 18.17797356828194,
+      "grad_norm": 0.24106785385763269,
+      "learning_rate": 0.00014647231720437686,
+      "loss": 0.0468,
+      "step": 2579
+    },
+    {
+      "epoch": 18.185022026431717,
+      "grad_norm": 0.22823383183444085,
+      "learning_rate": 0.00014643163166144443,
+      "loss": 0.0473,
+      "step": 2580
+    },
+    {
+      "epoch": 18.1920704845815,
+      "grad_norm": 0.23155916648301567,
+      "learning_rate": 0.00014639093631784825,
+      "loss": 0.049,
+      "step": 2581
+    },
+    {
+      "epoch": 18.19911894273128,
+      "grad_norm": 0.22528955865400685,
+      "learning_rate": 0.00014635023118217812,
+      "loss": 0.0453,
+      "step": 2582
+    },
+    {
+      "epoch": 18.206167400881057,
+      "grad_norm": 0.22042126536661996,
+      "learning_rate": 0.000146309516263026,
+      "loss": 0.0497,
+      "step": 2583
+    },
+    {
+      "epoch": 18.213215859030836,
+      "grad_norm": 0.22806498971050176,
+      "learning_rate": 0.00014626879156898594,
+      "loss": 0.0501,
+      "step": 2584
+    },
+    {
+      "epoch": 18.220264317180618,
+      "grad_norm": 0.21983773388806846,
+      "learning_rate": 0.00014622805710865392,
+      "loss": 0.0446,
+      "step": 2585
+    },
+    {
+      "epoch": 18.227312775330397,
+      "grad_norm": 0.21244426261750868,
+      "learning_rate": 0.00014618731289062814,
+      "loss": 0.0439,
+      "step": 2586
+    },
+    {
+      "epoch": 18.234361233480175,
+      "grad_norm": 0.2232257122113486,
+      "learning_rate": 0.00014614655892350865,
+      "loss": 0.0497,
+      "step": 2587
+    },
+    {
+      "epoch": 18.241409691629954,
+      "grad_norm": 0.23106398954535132,
+      "learning_rate": 0.00014610579521589782,
+      "loss": 0.0503,
+      "step": 2588
+    },
+    {
+      "epoch": 18.248458149779736,
+      "grad_norm": 0.22323909139326692,
+      "learning_rate": 0.0001460650217763999,
+      "loss": 0.0464,
+      "step": 2589
+    },
+    {
+      "epoch": 18.255506607929515,
+      "grad_norm": 0.2331418174426558,
+      "learning_rate": 0.00014602423861362124,
+      "loss": 0.0459,
+      "step": 2590
+    },
+    {
+      "epoch": 18.262555066079294,
+      "grad_norm": 0.21816682211648408,
+      "learning_rate": 0.00014598344573617022,
+      "loss": 0.0443,
+      "step": 2591
+    },
+    {
+      "epoch": 18.269603524229076,
+      "grad_norm": 0.3062144298922261,
+      "learning_rate": 0.00014594264315265728,
+      "loss": 0.0466,
+      "step": 2592
+    },
+    {
+      "epoch": 18.276651982378855,
+      "grad_norm": 0.23540532938194364,
+      "learning_rate": 0.00014590183087169497,
+      "loss": 0.0466,
+      "step": 2593
+    },
+    {
+      "epoch": 18.283700440528634,
+      "grad_norm": 0.2293183580028914,
+      "learning_rate": 0.00014586100890189783,
+      "loss": 0.0441,
+      "step": 2594
+    },
+    {
+      "epoch": 18.290748898678412,
+      "grad_norm": 0.2377056367736701,
+      "learning_rate": 0.0001458201772518824,
+      "loss": 0.0507,
+      "step": 2595
+    },
+    {
+      "epoch": 18.297797356828195,
+      "grad_norm": 0.2257869141558313,
+      "learning_rate": 0.00014577933593026734,
+      "loss": 0.0473,
+      "step": 2596
+    },
+    {
+      "epoch": 18.304845814977973,
+      "grad_norm": 0.244783343869152,
+      "learning_rate": 0.00014573848494567337,
+      "loss": 0.0436,
+      "step": 2597
+    },
+    {
+      "epoch": 18.311894273127752,
+      "grad_norm": 0.22338288407298026,
+      "learning_rate": 0.00014569762430672312,
+      "loss": 0.0477,
+      "step": 2598
+    },
+    {
+      "epoch": 18.318942731277534,
+      "grad_norm": 0.22949270814835934,
+      "learning_rate": 0.0001456567540220414,
+      "loss": 0.0458,
+      "step": 2599
+    },
+    {
+      "epoch": 18.325991189427313,
+      "grad_norm": 0.24089710069822082,
+      "learning_rate": 0.00014561587410025502,
+      "loss": 0.0462,
+      "step": 2600
+    },
+    {
+      "epoch": 18.333039647577092,
+      "grad_norm": 0.24223372513436628,
+      "learning_rate": 0.00014557498454999273,
+      "loss": 0.0464,
+      "step": 2601
+    },
+    {
+      "epoch": 18.34008810572687,
+      "grad_norm": 0.23169288709295408,
+      "learning_rate": 0.0001455340853798855,
+      "loss": 0.049,
+      "step": 2602
+    },
+    {
+      "epoch": 18.347136563876653,
+      "grad_norm": 0.2417585121776934,
+      "learning_rate": 0.0001454931765985661,
+      "loss": 0.047,
+      "step": 2603
+    },
+    {
+      "epoch": 18.35418502202643,
+      "grad_norm": 0.21742540263194843,
+      "learning_rate": 0.0001454522582146695,
+      "loss": 0.0467,
+      "step": 2604
+    },
+    {
+      "epoch": 18.36123348017621,
+      "grad_norm": 0.2236616643054931,
+      "learning_rate": 0.00014541133023683264,
+      "loss": 0.0411,
+      "step": 2605
+    },
+    {
+      "epoch": 18.368281938325993,
+      "grad_norm": 0.21939947859068681,
+      "learning_rate": 0.00014537039267369445,
+      "loss": 0.0447,
+      "step": 2606
+    },
+    {
+      "epoch": 18.37533039647577,
+      "grad_norm": 0.2280187137203806,
+      "learning_rate": 0.000145329445533896,
+      "loss": 0.0455,
+      "step": 2607
+    },
+    {
+      "epoch": 18.38237885462555,
+      "grad_norm": 0.24689284307787604,
+      "learning_rate": 0.00014528848882608024,
+      "loss": 0.047,
+      "step": 2608
+    },
+    {
+      "epoch": 18.38942731277533,
+      "grad_norm": 0.23061562790279977,
+      "learning_rate": 0.0001452475225588922,
+      "loss": 0.0534,
+      "step": 2609
+    },
+    {
+      "epoch": 18.39647577092511,
+      "grad_norm": 0.22578862564930371,
+      "learning_rate": 0.00014520654674097895,
+      "loss": 0.0501,
+      "step": 2610
+    },
+    {
+      "epoch": 18.40352422907489,
+      "grad_norm": 0.22489192611607267,
+      "learning_rate": 0.00014516556138098957,
+      "loss": 0.0506,
+      "step": 2611
+    },
+    {
+      "epoch": 18.41057268722467,
+      "grad_norm": 0.24338155496930933,
+      "learning_rate": 0.0001451245664875751,
+      "loss": 0.0515,
+      "step": 2612
+    },
+    {
+      "epoch": 18.41762114537445,
+      "grad_norm": 0.23402383236611812,
+      "learning_rate": 0.00014508356206938868,
+      "loss": 0.0502,
+      "step": 2613
+    },
+    {
+      "epoch": 18.42466960352423,
+      "grad_norm": 0.24025850552394298,
+      "learning_rate": 0.00014504254813508535,
+      "loss": 0.0476,
+      "step": 2614
+    },
+    {
+      "epoch": 18.431718061674008,
+      "grad_norm": 0.2405469051991262,
+      "learning_rate": 0.00014500152469332228,
+      "loss": 0.0496,
+      "step": 2615
+    },
+    {
+      "epoch": 18.438766519823787,
+      "grad_norm": 0.24142392706869295,
+      "learning_rate": 0.00014496049175275857,
+      "loss": 0.053,
+      "step": 2616
+    },
+    {
+      "epoch": 18.44581497797357,
+      "grad_norm": 0.2249173782274754,
+      "learning_rate": 0.0001449194493220553,
+      "loss": 0.0474,
+      "step": 2617
+    },
+    {
+      "epoch": 18.452863436123348,
+      "grad_norm": 0.248264922055059,
+      "learning_rate": 0.00014487839740987566,
+      "loss": 0.05,
+      "step": 2618
+    },
+    {
+      "epoch": 18.459911894273127,
+      "grad_norm": 0.22769431429762954,
+      "learning_rate": 0.00014483733602488474,
+      "loss": 0.0444,
+      "step": 2619
+    },
+    {
+      "epoch": 18.46696035242291,
+      "grad_norm": 0.2368981455478066,
+      "learning_rate": 0.0001447962651757497,
+      "loss": 0.0435,
+      "step": 2620
+    },
+    {
+      "epoch": 18.474008810572688,
+      "grad_norm": 0.24602586283325698,
+      "learning_rate": 0.00014475518487113958,
+      "loss": 0.0552,
+      "step": 2621
+    },
+    {
+      "epoch": 18.481057268722466,
+      "grad_norm": 0.23781292494868284,
+      "learning_rate": 0.00014471409511972562,
+      "loss": 0.0511,
+      "step": 2622
+    },
+    {
+      "epoch": 18.488105726872245,
+      "grad_norm": 0.23984258089465063,
+      "learning_rate": 0.00014467299593018084,
+      "loss": 0.0548,
+      "step": 2623
+    },
+    {
+      "epoch": 18.495154185022027,
+      "grad_norm": 0.2428063053183035,
+      "learning_rate": 0.0001446318873111804,
+      "loss": 0.0561,
+      "step": 2624
+    },
+    {
+      "epoch": 18.502202643171806,
+      "grad_norm": 0.23170517030217783,
+      "learning_rate": 0.00014459076927140135,
+      "loss": 0.0552,
+      "step": 2625
+    },
+    {
+      "epoch": 18.509251101321585,
+      "grad_norm": 0.24674040656764457,
+      "learning_rate": 0.00014454964181952282,
+      "loss": 0.0541,
+      "step": 2626
+    },
+    {
+      "epoch": 18.516299559471367,
+      "grad_norm": 0.24190274236400486,
+      "learning_rate": 0.00014450850496422585,
+      "loss": 0.0516,
+      "step": 2627
+    },
+    {
+      "epoch": 18.523348017621146,
+      "grad_norm": 0.23764615881341428,
+      "learning_rate": 0.00014446735871419353,
+      "loss": 0.0518,
+      "step": 2628
+    },
+    {
+      "epoch": 18.530396475770925,
+      "grad_norm": 0.23974634842011763,
+      "learning_rate": 0.00014442620307811085,
+      "loss": 0.05,
+      "step": 2629
+    },
+    {
+      "epoch": 18.537444933920703,
+      "grad_norm": 0.24227523417614813,
+      "learning_rate": 0.0001443850380646649,
+      "loss": 0.0533,
+      "step": 2630
+    },
+    {
+      "epoch": 18.544493392070486,
+      "grad_norm": 0.23580076463499486,
+      "learning_rate": 0.00014434386368254463,
+      "loss": 0.0504,
+      "step": 2631
+    },
+    {
+      "epoch": 18.551541850220264,
+      "grad_norm": 0.24119091009236482,
+      "learning_rate": 0.000144302679940441,
+      "loss": 0.0571,
+      "step": 2632
+    },
+    {
+      "epoch": 18.558590308370043,
+      "grad_norm": 0.2263133392334513,
+      "learning_rate": 0.00014426148684704702,
+      "loss": 0.0488,
+      "step": 2633
+    },
+    {
+      "epoch": 18.565638766519825,
+      "grad_norm": 0.23539528847509114,
+      "learning_rate": 0.00014422028441105758,
+      "loss": 0.0533,
+      "step": 2634
+    },
+    {
+      "epoch": 18.572687224669604,
+      "grad_norm": 0.23879692589565735,
+      "learning_rate": 0.00014417907264116956,
+      "loss": 0.0552,
+      "step": 2635
+    },
+    {
+      "epoch": 18.579735682819383,
+      "grad_norm": 0.24311380605784202,
+      "learning_rate": 0.0001441378515460819,
+      "loss": 0.0508,
+      "step": 2636
+    },
+    {
+      "epoch": 18.58678414096916,
+      "grad_norm": 0.23584965675876896,
+      "learning_rate": 0.00014409662113449537,
+      "loss": 0.0511,
+      "step": 2637
+    },
+    {
+      "epoch": 18.593832599118944,
+      "grad_norm": 0.23464288504408173,
+      "learning_rate": 0.00014405538141511282,
+      "loss": 0.0485,
+      "step": 2638
+    },
+    {
+      "epoch": 18.600881057268722,
+      "grad_norm": 0.2353674416209289,
+      "learning_rate": 0.000144014132396639,
+      "loss": 0.0509,
+      "step": 2639
+    },
+    {
+      "epoch": 18.6079295154185,
+      "grad_norm": 0.24597839592445914,
+      "learning_rate": 0.00014397287408778059,
+      "loss": 0.0536,
+      "step": 2640
+    },
+    {
+      "epoch": 18.614977973568283,
+      "grad_norm": 0.2422417852491735,
+      "learning_rate": 0.0001439316064972464,
+      "loss": 0.0546,
+      "step": 2641
+    },
+    {
+      "epoch": 18.622026431718062,
+      "grad_norm": 0.2582034946503304,
+      "learning_rate": 0.00014389032963374695,
+      "loss": 0.0548,
+      "step": 2642
+    },
+    {
+      "epoch": 18.62907488986784,
+      "grad_norm": 0.24201459443548315,
+      "learning_rate": 0.00014384904350599496,
+      "loss": 0.0537,
+      "step": 2643
+    },
+    {
+      "epoch": 18.63612334801762,
+      "grad_norm": 0.25083747702226794,
+      "learning_rate": 0.00014380774812270495,
+      "loss": 0.0515,
+      "step": 2644
+    },
+    {
+      "epoch": 18.643171806167402,
+      "grad_norm": 0.2391251992152529,
+      "learning_rate": 0.00014376644349259343,
+      "loss": 0.0518,
+      "step": 2645
+    },
+    {
+      "epoch": 18.65022026431718,
+      "grad_norm": 0.25101754537030724,
+      "learning_rate": 0.00014372512962437884,
+      "loss": 0.053,
+      "step": 2646
+    },
+    {
+      "epoch": 18.65726872246696,
+      "grad_norm": 0.23208306789817462,
+      "learning_rate": 0.0001436838065267817,
+      "loss": 0.0497,
+      "step": 2647
+    },
+    {
+      "epoch": 18.66431718061674,
+      "grad_norm": 0.24293162432209048,
+      "learning_rate": 0.0001436424742085243,
+      "loss": 0.054,
+      "step": 2648
+    },
+    {
+      "epoch": 18.67136563876652,
+      "grad_norm": 0.24885260543895568,
+      "learning_rate": 0.00014360113267833092,
+      "loss": 0.0546,
+      "step": 2649
+    },
+    {
+      "epoch": 18.6784140969163,
+      "grad_norm": 0.2365934437347568,
+      "learning_rate": 0.00014355978194492791,
+      "loss": 0.0487,
+      "step": 2650
+    },
+    {
+      "epoch": 18.685462555066078,
+      "grad_norm": 0.2389756571293592,
+      "learning_rate": 0.00014351842201704339,
+      "loss": 0.0545,
+      "step": 2651
+    },
+    {
+      "epoch": 18.69251101321586,
+      "grad_norm": 0.27618801742787435,
+      "learning_rate": 0.0001434770529034076,
+      "loss": 0.0569,
+      "step": 2652
+    },
+    {
+      "epoch": 18.69955947136564,
+      "grad_norm": 0.23688419247482884,
+      "learning_rate": 0.00014343567461275254,
+      "loss": 0.0531,
+      "step": 2653
+    },
+    {
+      "epoch": 18.706607929515418,
+      "grad_norm": 0.24642460176122863,
+      "learning_rate": 0.00014339428715381223,
+      "loss": 0.0563,
+      "step": 2654
+    },
+    {
+      "epoch": 18.7136563876652,
+      "grad_norm": 0.2513237634397049,
+      "learning_rate": 0.00014335289053532267,
+      "loss": 0.0543,
+      "step": 2655
+    },
+    {
+      "epoch": 18.72070484581498,
+      "grad_norm": 0.22982654821761778,
+      "learning_rate": 0.0001433114847660217,
+      "loss": 0.0509,
+      "step": 2656
+    },
+    {
+      "epoch": 18.727753303964757,
+      "grad_norm": 0.23911028663525175,
+      "learning_rate": 0.0001432700698546492,
+      "loss": 0.0528,
+      "step": 2657
+    },
+    {
+      "epoch": 18.734801762114536,
+      "grad_norm": 0.2604316757265864,
+      "learning_rate": 0.00014322864580994683,
+      "loss": 0.0579,
+      "step": 2658
+    },
+    {
+      "epoch": 18.74185022026432,
+      "grad_norm": 0.2378670037664313,
+      "learning_rate": 0.00014318721264065833,
+      "loss": 0.0516,
+      "step": 2659
+    },
+    {
+      "epoch": 18.748898678414097,
+      "grad_norm": 0.2625387983396872,
+      "learning_rate": 0.0001431457703555293,
+      "loss": 0.0558,
+      "step": 2660
+    },
+    {
+      "epoch": 18.755947136563876,
+      "grad_norm": 0.23974376613407575,
+      "learning_rate": 0.00014310431896330725,
+      "loss": 0.0553,
+      "step": 2661
+    },
+    {
+      "epoch": 18.762995594713658,
+      "grad_norm": 0.24188477776352627,
+      "learning_rate": 0.00014306285847274158,
+      "loss": 0.0537,
+      "step": 2662
+    },
+    {
+      "epoch": 18.770044052863437,
+      "grad_norm": 0.2626680609601367,
+      "learning_rate": 0.00014302138889258378,
+      "loss": 0.0586,
+      "step": 2663
+    },
+    {
+      "epoch": 18.777092511013215,
+      "grad_norm": 0.23493273290286684,
+      "learning_rate": 0.00014297991023158704,
+      "loss": 0.0531,
+      "step": 2664
+    },
+    {
+      "epoch": 18.784140969162994,
+      "grad_norm": 0.2464030771389173,
+      "learning_rate": 0.00014293842249850656,
+      "loss": 0.0532,
+      "step": 2665
+    },
+    {
+      "epoch": 18.791189427312776,
+      "grad_norm": 0.25489748074389784,
+      "learning_rate": 0.00014289692570209955,
+      "loss": 0.0613,
+      "step": 2666
+    },
+    {
+      "epoch": 18.798237885462555,
+      "grad_norm": 0.2530544648207399,
+      "learning_rate": 0.00014285541985112492,
+      "loss": 0.0554,
+      "step": 2667
+    },
+    {
+      "epoch": 18.805286343612334,
+      "grad_norm": 0.2586037267359629,
+      "learning_rate": 0.0001428139049543437,
+      "loss": 0.0607,
+      "step": 2668
+    },
+    {
+      "epoch": 18.812334801762116,
+      "grad_norm": 0.24261462853011997,
+      "learning_rate": 0.0001427723810205187,
+      "loss": 0.0581,
+      "step": 2669
+    },
+    {
+      "epoch": 18.819383259911895,
+      "grad_norm": 0.2421116425389537,
+      "learning_rate": 0.00014273084805841474,
+      "loss": 0.0549,
+      "step": 2670
+    },
+    {
+      "epoch": 18.826431718061674,
+      "grad_norm": 0.24001213468616595,
+      "learning_rate": 0.0001426893060767984,
+      "loss": 0.0592,
+      "step": 2671
+    },
+    {
+      "epoch": 18.833480176211452,
+      "grad_norm": 0.2448074912652504,
+      "learning_rate": 0.0001426477550844383,
+      "loss": 0.0534,
+      "step": 2672
+    },
+    {
+      "epoch": 18.840528634361235,
+      "grad_norm": 0.2569706330049187,
+      "learning_rate": 0.00014260619509010492,
+      "loss": 0.0565,
+      "step": 2673
+    },
+    {
+      "epoch": 18.847577092511013,
+      "grad_norm": 0.24922113799700604,
+      "learning_rate": 0.00014256462610257059,
+      "loss": 0.0592,
+      "step": 2674
+    },
+    {
+      "epoch": 18.854625550660792,
+      "grad_norm": 0.2729473410439542,
+      "learning_rate": 0.00014252304813060964,
+      "loss": 0.0645,
+      "step": 2675
+    },
+    {
+      "epoch": 18.861674008810574,
+      "grad_norm": 0.246791874313013,
+      "learning_rate": 0.00014248146118299814,
+      "loss": 0.0563,
+      "step": 2676
+    },
+    {
+      "epoch": 18.868722466960353,
+      "grad_norm": 0.24826750802929692,
+      "learning_rate": 0.0001424398652685143,
+      "loss": 0.0579,
+      "step": 2677
+    },
+    {
+      "epoch": 18.87577092511013,
+      "grad_norm": 0.2479701125007055,
+      "learning_rate": 0.0001423982603959379,
+      "loss": 0.0555,
+      "step": 2678
+    },
+    {
+      "epoch": 18.88281938325991,
+      "grad_norm": 0.2652303874288224,
+      "learning_rate": 0.00014235664657405092,
+      "loss": 0.0609,
+      "step": 2679
+    },
+    {
+      "epoch": 18.889867841409693,
+      "grad_norm": 0.24159124210033106,
+      "learning_rate": 0.00014231502381163705,
+      "loss": 0.0533,
+      "step": 2680
+    },
+    {
+      "epoch": 18.89691629955947,
+      "grad_norm": 0.246854766219499,
+      "learning_rate": 0.00014227339211748187,
+      "loss": 0.0555,
+      "step": 2681
+    },
+    {
+      "epoch": 18.90396475770925,
+      "grad_norm": 0.25156076431611296,
+      "learning_rate": 0.00014223175150037296,
+      "loss": 0.0556,
+      "step": 2682
+    },
+    {
+      "epoch": 18.911013215859032,
+      "grad_norm": 0.26516492857255625,
+      "learning_rate": 0.00014219010196909963,
+      "loss": 0.0578,
+      "step": 2683
+    },
+    {
+      "epoch": 18.91806167400881,
+      "grad_norm": 0.2679562884557641,
+      "learning_rate": 0.0001421484435324532,
+      "loss": 0.0588,
+      "step": 2684
+    },
+    {
+      "epoch": 18.92511013215859,
+      "grad_norm": 0.2581118617523638,
+      "learning_rate": 0.00014210677619922683,
+      "loss": 0.0572,
+      "step": 2685
+    },
+    {
+      "epoch": 18.93215859030837,
+      "grad_norm": 0.2521696422314273,
+      "learning_rate": 0.00014206509997821554,
+      "loss": 0.0629,
+      "step": 2686
+    },
+    {
+      "epoch": 18.93920704845815,
+      "grad_norm": 0.2639294464280062,
+      "learning_rate": 0.0001420234148782162,
+      "loss": 0.0611,
+      "step": 2687
+    },
+    {
+      "epoch": 18.94625550660793,
+      "grad_norm": 0.25080521930741007,
+      "learning_rate": 0.00014198172090802762,
+      "loss": 0.0595,
+      "step": 2688
+    },
+    {
+      "epoch": 18.95330396475771,
+      "grad_norm": 0.25759727551906497,
+      "learning_rate": 0.00014194001807645042,
+      "loss": 0.0564,
+      "step": 2689
+    },
+    {
+      "epoch": 18.96035242290749,
+      "grad_norm": 0.27143191106442116,
+      "learning_rate": 0.0001418983063922872,
+      "loss": 0.0586,
+      "step": 2690
+    },
+    {
+      "epoch": 18.96740088105727,
+      "grad_norm": 0.26588201937198586,
+      "learning_rate": 0.00014185658586434225,
+      "loss": 0.0581,
+      "step": 2691
+    },
+    {
+      "epoch": 18.974449339207048,
+      "grad_norm": 0.27184179228348954,
+      "learning_rate": 0.00014181485650142185,
+      "loss": 0.0604,
+      "step": 2692
+    },
+    {
+      "epoch": 18.981497797356827,
+      "grad_norm": 0.24703793602401464,
+      "learning_rate": 0.00014177311831233423,
+      "loss": 0.0602,
+      "step": 2693
+    },
+    {
+      "epoch": 18.98854625550661,
+      "grad_norm": 0.2658821373491647,
+      "learning_rate": 0.0001417313713058892,
+      "loss": 0.0586,
+      "step": 2694
+    },
+    {
+      "epoch": 18.995594713656388,
+      "grad_norm": 0.24003849300174604,
+      "learning_rate": 0.00014168961549089874,
+      "loss": 0.0536,
+      "step": 2695
+    },
+    {
+      "epoch": 19.002643171806167,
+      "grad_norm": 0.22839003239537578,
+      "learning_rate": 0.0001416478508761765,
+      "loss": 0.0499,
+      "step": 2696
+    },
+    {
+      "epoch": 19.00969162995595,
+      "grad_norm": 0.19848444331390147,
+      "learning_rate": 0.00014160607747053805,
+      "loss": 0.0341,
+      "step": 2697
+    },
+    {
+      "epoch": 19.016740088105728,
+      "grad_norm": 0.19923352065532,
+      "learning_rate": 0.00014156429528280082,
+      "loss": 0.0316,
+      "step": 2698
+    },
+    {
+      "epoch": 19.023788546255506,
+      "grad_norm": 0.19217324384546258,
+      "learning_rate": 0.00014152250432178403,
+      "loss": 0.0325,
+      "step": 2699
+    },
+    {
+      "epoch": 19.030837004405285,
+      "grad_norm": 0.2097818367440442,
+      "learning_rate": 0.0001414807045963089,
+      "loss": 0.0302,
+      "step": 2700
+    },
+    {
+      "epoch": 19.037885462555067,
+      "grad_norm": 0.2090025790262824,
+      "learning_rate": 0.00014143889611519832,
+      "loss": 0.0322,
+      "step": 2701
+    },
+    {
+      "epoch": 19.044933920704846,
+      "grad_norm": 0.2172974607384721,
+      "learning_rate": 0.00014139707888727716,
+      "loss": 0.032,
+      "step": 2702
+    },
+    {
+      "epoch": 19.051982378854625,
+      "grad_norm": 0.21266530436101586,
+      "learning_rate": 0.00014135525292137201,
+      "loss": 0.0346,
+      "step": 2703
+    },
+    {
+      "epoch": 19.059030837004407,
+      "grad_norm": 0.19749959145470108,
+      "learning_rate": 0.00014131341822631153,
+      "loss": 0.0323,
+      "step": 2704
+    },
+    {
+      "epoch": 19.066079295154186,
+      "grad_norm": 0.19967261612273532,
+      "learning_rate": 0.00014127157481092593,
+      "loss": 0.0307,
+      "step": 2705
+    },
+    {
+      "epoch": 19.073127753303964,
+      "grad_norm": 0.217025757401685,
+      "learning_rate": 0.00014122972268404744,
+      "loss": 0.0321,
+      "step": 2706
+    },
+    {
+      "epoch": 19.080176211453743,
+      "grad_norm": 0.18617900100202028,
+      "learning_rate": 0.00014118786185451015,
+      "loss": 0.0303,
+      "step": 2707
+    },
+    {
+      "epoch": 19.087224669603525,
+      "grad_norm": 0.21594312232874013,
+      "learning_rate": 0.00014114599233114986,
+      "loss": 0.0339,
+      "step": 2708
+    },
+    {
+      "epoch": 19.094273127753304,
+      "grad_norm": 0.2108621735798284,
+      "learning_rate": 0.00014110411412280436,
+      "loss": 0.0336,
+      "step": 2709
+    },
+    {
+      "epoch": 19.101321585903083,
+      "grad_norm": 0.1864098423925384,
+      "learning_rate": 0.0001410622272383131,
+      "loss": 0.0289,
+      "step": 2710
+    },
+    {
+      "epoch": 19.10837004405286,
+      "grad_norm": 0.20199738081888988,
+      "learning_rate": 0.00014102033168651753,
+      "loss": 0.033,
+      "step": 2711
+    },
+    {
+      "epoch": 19.115418502202644,
+      "grad_norm": 0.20005674783820693,
+      "learning_rate": 0.00014097842747626076,
+      "loss": 0.0327,
+      "step": 2712
+    },
+    {
+      "epoch": 19.122466960352423,
+      "grad_norm": 0.1907600592136181,
+      "learning_rate": 0.0001409365146163879,
+      "loss": 0.029,
+      "step": 2713
+    },
+    {
+      "epoch": 19.1295154185022,
+      "grad_norm": 0.20065318793395234,
+      "learning_rate": 0.00014089459311574578,
+      "loss": 0.0332,
+      "step": 2714
+    },
+    {
+      "epoch": 19.136563876651984,
+      "grad_norm": 0.1974772804298112,
+      "learning_rate": 0.00014085266298318307,
+      "loss": 0.0294,
+      "step": 2715
+    },
+    {
+      "epoch": 19.143612334801762,
+      "grad_norm": 0.18161276478703017,
+      "learning_rate": 0.00014081072422755027,
+      "loss": 0.027,
+      "step": 2716
+    },
+    {
+      "epoch": 19.15066079295154,
+      "grad_norm": 0.202607211382346,
+      "learning_rate": 0.00014076877685769966,
+      "loss": 0.0306,
+      "step": 2717
+    },
+    {
+      "epoch": 19.15770925110132,
+      "grad_norm": 0.2104212609132182,
+      "learning_rate": 0.00014072682088248547,
+      "loss": 0.0318,
+      "step": 2718
+    },
+    {
+      "epoch": 19.164757709251102,
+      "grad_norm": 0.20539339776247661,
+      "learning_rate": 0.00014068485631076359,
+      "loss": 0.031,
+      "step": 2719
+    },
+    {
+      "epoch": 19.17180616740088,
+      "grad_norm": 0.19114347769910026,
+      "learning_rate": 0.0001406428831513918,
+      "loss": 0.0307,
+      "step": 2720
+    },
+    {
+      "epoch": 19.17885462555066,
+      "grad_norm": 0.2022783054974036,
+      "learning_rate": 0.00014060090141322968,
+      "loss": 0.0308,
+      "step": 2721
+    },
+    {
+      "epoch": 19.185903083700442,
+      "grad_norm": 0.19878635420045992,
+      "learning_rate": 0.0001405589111051386,
+      "loss": 0.032,
+      "step": 2722
+    },
+    {
+      "epoch": 19.19295154185022,
+      "grad_norm": 0.19620599659301177,
+      "learning_rate": 0.0001405169122359818,
+      "loss": 0.0302,
+      "step": 2723
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 0.18759927694872397,
+      "learning_rate": 0.0001404749048146243,
+      "loss": 0.0296,
+      "step": 2724
+    },
+    {
+      "epoch": 19.207048458149778,
+      "grad_norm": 0.19240707242220625,
+      "learning_rate": 0.00014043288884993285,
+      "loss": 0.0309,
+      "step": 2725
+    },
+    {
+      "epoch": 19.21409691629956,
+      "grad_norm": 0.18773779001243515,
+      "learning_rate": 0.00014039086435077612,
+      "loss": 0.0298,
+      "step": 2726
+    },
+    {
+      "epoch": 19.22114537444934,
+      "grad_norm": 0.19917294693789359,
+      "learning_rate": 0.00014034883132602453,
+      "loss": 0.0314,
+      "step": 2727
+    },
+    {
+      "epoch": 19.228193832599118,
+      "grad_norm": 0.19281076200958777,
+      "learning_rate": 0.00014030678978455024,
+      "loss": 0.0326,
+      "step": 2728
+    },
+    {
+      "epoch": 19.2352422907489,
+      "grad_norm": 0.1979648192172092,
+      "learning_rate": 0.00014026473973522733,
+      "loss": 0.0282,
+      "step": 2729
+    },
+    {
+      "epoch": 19.24229074889868,
+      "grad_norm": 0.20186104245811512,
+      "learning_rate": 0.0001402226811869316,
+      "loss": 0.03,
+      "step": 2730
+    },
+    {
+      "epoch": 19.249339207048457,
+      "grad_norm": 0.20138048462099892,
+      "learning_rate": 0.00014018061414854062,
+      "loss": 0.0304,
+      "step": 2731
+    },
+    {
+      "epoch": 19.256387665198236,
+      "grad_norm": 0.1991303391866437,
+      "learning_rate": 0.00014013853862893382,
+      "loss": 0.0288,
+      "step": 2732
+    },
+    {
+      "epoch": 19.26343612334802,
+      "grad_norm": 0.2110053197307814,
+      "learning_rate": 0.00014009645463699234,
+      "loss": 0.0321,
+      "step": 2733
+    },
+    {
+      "epoch": 19.270484581497797,
+      "grad_norm": 0.2079883216265982,
+      "learning_rate": 0.00014005436218159927,
+      "loss": 0.0329,
+      "step": 2734
+    },
+    {
+      "epoch": 19.277533039647576,
+      "grad_norm": 0.19811360635187159,
+      "learning_rate": 0.0001400122612716392,
+      "loss": 0.0307,
+      "step": 2735
+    },
+    {
+      "epoch": 19.284581497797358,
+      "grad_norm": 0.21487948942249396,
+      "learning_rate": 0.00013997015191599883,
+      "loss": 0.0307,
+      "step": 2736
+    },
+    {
+      "epoch": 19.291629955947137,
+      "grad_norm": 0.2021126633221014,
+      "learning_rate": 0.00013992803412356643,
+      "loss": 0.0315,
+      "step": 2737
+    },
+    {
+      "epoch": 19.298678414096916,
+      "grad_norm": 0.202787031469461,
+      "learning_rate": 0.00013988590790323208,
+      "loss": 0.0314,
+      "step": 2738
+    },
+    {
+      "epoch": 19.305726872246694,
+      "grad_norm": 0.20630735213832124,
+      "learning_rate": 0.00013984377326388772,
+      "loss": 0.0317,
+      "step": 2739
+    },
+    {
+      "epoch": 19.312775330396477,
+      "grad_norm": 0.19806390257285345,
+      "learning_rate": 0.00013980163021442698,
+      "loss": 0.0302,
+      "step": 2740
+    },
+    {
+      "epoch": 19.319823788546255,
+      "grad_norm": 0.2175189201972166,
+      "learning_rate": 0.00013975947876374532,
+      "loss": 0.0333,
+      "step": 2741
+    },
+    {
+      "epoch": 19.326872246696034,
+      "grad_norm": 0.19789111089367922,
+      "learning_rate": 0.00013971731892073997,
+      "loss": 0.0331,
+      "step": 2742
+    },
+    {
+      "epoch": 19.333920704845816,
+      "grad_norm": 0.21775345425754628,
+      "learning_rate": 0.0001396751506943099,
+      "loss": 0.033,
+      "step": 2743
+    },
+    {
+      "epoch": 19.340969162995595,
+      "grad_norm": 0.20884391749559103,
+      "learning_rate": 0.00013963297409335583,
+      "loss": 0.0318,
+      "step": 2744
+    },
+    {
+      "epoch": 19.348017621145374,
+      "grad_norm": 0.20170994724585206,
+      "learning_rate": 0.00013959078912678037,
+      "loss": 0.0313,
+      "step": 2745
+    },
+    {
+      "epoch": 19.355066079295153,
+      "grad_norm": 0.20675744432629006,
+      "learning_rate": 0.0001395485958034877,
+      "loss": 0.0313,
+      "step": 2746
+    },
+    {
+      "epoch": 19.362114537444935,
+      "grad_norm": 0.2240866711731941,
+      "learning_rate": 0.00013950639413238394,
+      "loss": 0.0345,
+      "step": 2747
+    },
+    {
+      "epoch": 19.369162995594714,
+      "grad_norm": 0.22751802194296103,
+      "learning_rate": 0.00013946418412237694,
+      "loss": 0.032,
+      "step": 2748
+    },
+    {
+      "epoch": 19.376211453744492,
+      "grad_norm": 0.21738194212007664,
+      "learning_rate": 0.00013942196578237617,
+      "loss": 0.0332,
+      "step": 2749
+    },
+    {
+      "epoch": 19.383259911894275,
+      "grad_norm": 0.2084149834306039,
+      "learning_rate": 0.00013937973912129305,
+      "loss": 0.0311,
+      "step": 2750
+    },
+    {
+      "epoch": 19.390308370044053,
+      "grad_norm": 0.19967504123221805,
+      "learning_rate": 0.00013933750414804063,
+      "loss": 0.0316,
+      "step": 2751
+    },
+    {
+      "epoch": 19.397356828193832,
+      "grad_norm": 0.2085508461057109,
+      "learning_rate": 0.0001392952608715338,
+      "loss": 0.0326,
+      "step": 2752
+    },
+    {
+      "epoch": 19.40440528634361,
+      "grad_norm": 0.21507668641478017,
+      "learning_rate": 0.0001392530093006891,
+      "loss": 0.0329,
+      "step": 2753
+    },
+    {
+      "epoch": 19.411453744493393,
+      "grad_norm": 0.20601975027525954,
+      "learning_rate": 0.00013921074944442488,
+      "loss": 0.0338,
+      "step": 2754
+    },
+    {
+      "epoch": 19.41850220264317,
+      "grad_norm": 0.20964112364526602,
+      "learning_rate": 0.0001391684813116613,
+      "loss": 0.0347,
+      "step": 2755
+    },
+    {
+      "epoch": 19.42555066079295,
+      "grad_norm": 0.2245341563519042,
+      "learning_rate": 0.00013912620491132014,
+      "loss": 0.0355,
+      "step": 2756
+    },
+    {
+      "epoch": 19.432599118942733,
+      "grad_norm": 0.20820797838394864,
+      "learning_rate": 0.00013908392025232502,
+      "loss": 0.0363,
+      "step": 2757
+    },
+    {
+      "epoch": 19.43964757709251,
+      "grad_norm": 0.208294986899641,
+      "learning_rate": 0.00013904162734360123,
+      "loss": 0.0313,
+      "step": 2758
+    },
+    {
+      "epoch": 19.44669603524229,
+      "grad_norm": 0.20874751852588078,
+      "learning_rate": 0.00013899932619407593,
+      "loss": 0.0336,
+      "step": 2759
+    },
+    {
+      "epoch": 19.45374449339207,
+      "grad_norm": 0.22102575177794948,
+      "learning_rate": 0.00013895701681267784,
+      "loss": 0.0371,
+      "step": 2760
+    },
+    {
+      "epoch": 19.46079295154185,
+      "grad_norm": 0.19873809048830088,
+      "learning_rate": 0.00013891469920833754,
+      "loss": 0.0327,
+      "step": 2761
+    },
+    {
+      "epoch": 19.46784140969163,
+      "grad_norm": 0.2031747089263062,
+      "learning_rate": 0.00013887237338998737,
+      "loss": 0.0327,
+      "step": 2762
+    },
+    {
+      "epoch": 19.47488986784141,
+      "grad_norm": 0.20624237001203732,
+      "learning_rate": 0.00013883003936656128,
+      "loss": 0.0339,
+      "step": 2763
+    },
+    {
+      "epoch": 19.48193832599119,
+      "grad_norm": 0.20211648820612693,
+      "learning_rate": 0.00013878769714699505,
+      "loss": 0.0333,
+      "step": 2764
+    },
+    {
+      "epoch": 19.48898678414097,
+      "grad_norm": 0.19932845269505148,
+      "learning_rate": 0.00013874534674022615,
+      "loss": 0.032,
+      "step": 2765
+    },
+    {
+      "epoch": 19.49603524229075,
+      "grad_norm": 0.2095409997840032,
+      "learning_rate": 0.0001387029881551938,
+      "loss": 0.0321,
+      "step": 2766
+    },
+    {
+      "epoch": 19.503083700440527,
+      "grad_norm": 0.21011685142900735,
+      "learning_rate": 0.00013866062140083896,
+      "loss": 0.0345,
+      "step": 2767
+    },
+    {
+      "epoch": 19.51013215859031,
+      "grad_norm": 0.21818841337002925,
+      "learning_rate": 0.00013861824648610425,
+      "loss": 0.037,
+      "step": 2768
+    },
+    {
+      "epoch": 19.517180616740088,
+      "grad_norm": 0.21031048096450586,
+      "learning_rate": 0.000138575863419934,
+      "loss": 0.0343,
+      "step": 2769
+    },
+    {
+      "epoch": 19.524229074889867,
+      "grad_norm": 0.20811553178940337,
+      "learning_rate": 0.0001385334722112745,
+      "loss": 0.033,
+      "step": 2770
+    },
+    {
+      "epoch": 19.53127753303965,
+      "grad_norm": 0.21275708613955463,
+      "learning_rate": 0.00013849107286907338,
+      "loss": 0.0336,
+      "step": 2771
+    },
+    {
+      "epoch": 19.538325991189428,
+      "grad_norm": 0.23363228257767624,
+      "learning_rate": 0.00013844866540228026,
+      "loss": 0.0405,
+      "step": 2772
+    },
+    {
+      "epoch": 19.545374449339207,
+      "grad_norm": 0.2168209617718213,
+      "learning_rate": 0.0001384062498198464,
+      "loss": 0.0351,
+      "step": 2773
+    },
+    {
+      "epoch": 19.552422907488985,
+      "grad_norm": 0.209547860972177,
+      "learning_rate": 0.00013836382613072475,
+      "loss": 0.0382,
+      "step": 2774
+    },
+    {
+      "epoch": 19.559471365638768,
+      "grad_norm": 0.2076345207139478,
+      "learning_rate": 0.00013832139434386998,
+      "loss": 0.0357,
+      "step": 2775
+    },
+    {
+      "epoch": 19.566519823788546,
+      "grad_norm": 0.20834459133422098,
+      "learning_rate": 0.00013827895446823847,
+      "loss": 0.0329,
+      "step": 2776
+    },
+    {
+      "epoch": 19.573568281938325,
+      "grad_norm": 0.22879016662143484,
+      "learning_rate": 0.00013823650651278836,
+      "loss": 0.0381,
+      "step": 2777
+    },
+    {
+      "epoch": 19.580616740088107,
+      "grad_norm": 0.22164221239014398,
+      "learning_rate": 0.00013819405048647943,
+      "loss": 0.0365,
+      "step": 2778
+    },
+    {
+      "epoch": 19.587665198237886,
+      "grad_norm": 0.21410167387912993,
+      "learning_rate": 0.00013815158639827316,
+      "loss": 0.0358,
+      "step": 2779
+    },
+    {
+      "epoch": 19.594713656387665,
+      "grad_norm": 0.21976738260418335,
+      "learning_rate": 0.00013810911425713278,
+      "loss": 0.03,
+      "step": 2780
+    },
+    {
+      "epoch": 19.601762114537443,
+      "grad_norm": 0.22120467540593886,
+      "learning_rate": 0.0001380666340720232,
+      "loss": 0.0405,
+      "step": 2781
+    },
+    {
+      "epoch": 19.608810572687226,
+      "grad_norm": 0.2160738296813782,
+      "learning_rate": 0.000138024145851911,
+      "loss": 0.0364,
+      "step": 2782
+    },
+    {
+      "epoch": 19.615859030837004,
+      "grad_norm": 0.22346997741381555,
+      "learning_rate": 0.00013798164960576446,
+      "loss": 0.0367,
+      "step": 2783
+    },
+    {
+      "epoch": 19.622907488986783,
+      "grad_norm": 0.22041392863586773,
+      "learning_rate": 0.00013793914534255362,
+      "loss": 0.0398,
+      "step": 2784
+    },
+    {
+      "epoch": 19.629955947136565,
+      "grad_norm": 0.2242796963815488,
+      "learning_rate": 0.00013789663307125016,
+      "loss": 0.0374,
+      "step": 2785
+    },
+    {
+      "epoch": 19.637004405286344,
+      "grad_norm": 0.21377337802230825,
+      "learning_rate": 0.00013785411280082746,
+      "loss": 0.037,
+      "step": 2786
+    },
+    {
+      "epoch": 19.644052863436123,
+      "grad_norm": 0.21021228339873063,
+      "learning_rate": 0.00013781158454026054,
+      "loss": 0.0344,
+      "step": 2787
+    },
+    {
+      "epoch": 19.6511013215859,
+      "grad_norm": 0.2137606457481463,
+      "learning_rate": 0.0001377690482985262,
+      "loss": 0.0363,
+      "step": 2788
+    },
+    {
+      "epoch": 19.658149779735684,
+      "grad_norm": 0.211245290434495,
+      "learning_rate": 0.00013772650408460282,
+      "loss": 0.0349,
+      "step": 2789
+    },
+    {
+      "epoch": 19.665198237885463,
+      "grad_norm": 0.21539412559161092,
+      "learning_rate": 0.0001376839519074706,
+      "loss": 0.0362,
+      "step": 2790
+    },
+    {
+      "epoch": 19.67224669603524,
+      "grad_norm": 0.22595444728413439,
+      "learning_rate": 0.00013764139177611123,
+      "loss": 0.0359,
+      "step": 2791
+    },
+    {
+      "epoch": 19.679295154185024,
+      "grad_norm": 0.20333301359141046,
+      "learning_rate": 0.0001375988236995083,
+      "loss": 0.0325,
+      "step": 2792
+    },
+    {
+      "epoch": 19.686343612334802,
+      "grad_norm": 0.21117484509519716,
+      "learning_rate": 0.0001375562476866469,
+      "loss": 0.0331,
+      "step": 2793
+    },
+    {
+      "epoch": 19.69339207048458,
+      "grad_norm": 0.217988988639711,
+      "learning_rate": 0.00013751366374651386,
+      "loss": 0.0344,
+      "step": 2794
+    },
+    {
+      "epoch": 19.70044052863436,
+      "grad_norm": 0.21697628813288294,
+      "learning_rate": 0.0001374710718880977,
+      "loss": 0.0357,
+      "step": 2795
+    },
+    {
+      "epoch": 19.707488986784142,
+      "grad_norm": 0.21462119993797174,
+      "learning_rate": 0.00013742847212038864,
+      "loss": 0.0367,
+      "step": 2796
+    },
+    {
+      "epoch": 19.71453744493392,
+      "grad_norm": 0.2165466097557446,
+      "learning_rate": 0.00013738586445237843,
+      "loss": 0.036,
+      "step": 2797
+    },
+    {
+      "epoch": 19.7215859030837,
+      "grad_norm": 0.2238899729203766,
+      "learning_rate": 0.00013734324889306062,
+      "loss": 0.0388,
+      "step": 2798
+    },
+    {
+      "epoch": 19.728634361233482,
+      "grad_norm": 0.2198212053551803,
+      "learning_rate": 0.0001373006254514304,
+      "loss": 0.0371,
+      "step": 2799
+    },
+    {
+      "epoch": 19.73568281938326,
+      "grad_norm": 0.21663209752633322,
+      "learning_rate": 0.00013725799413648464,
+      "loss": 0.0353,
+      "step": 2800
+    },
+    {
+      "epoch": 19.74273127753304,
+      "grad_norm": 0.22174763480877144,
+      "learning_rate": 0.00013721535495722178,
+      "loss": 0.0385,
+      "step": 2801
+    },
+    {
+      "epoch": 19.749779735682818,
+      "grad_norm": 0.21229730515102857,
+      "learning_rate": 0.000137172707922642,
+      "loss": 0.0386,
+      "step": 2802
+    },
+    {
+      "epoch": 19.7568281938326,
+      "grad_norm": 0.2342244365151114,
+      "learning_rate": 0.0001371300530417472,
+      "loss": 0.0404,
+      "step": 2803
+    },
+    {
+      "epoch": 19.76387665198238,
+      "grad_norm": 0.21698453129014625,
+      "learning_rate": 0.00013708739032354072,
+      "loss": 0.0381,
+      "step": 2804
+    },
+    {
+      "epoch": 19.770925110132158,
+      "grad_norm": 0.2321663786721119,
+      "learning_rate": 0.00013704471977702779,
+      "loss": 0.0375,
+      "step": 2805
+    },
+    {
+      "epoch": 19.77797356828194,
+      "grad_norm": 0.2207189666632061,
+      "learning_rate": 0.00013700204141121517,
+      "loss": 0.0346,
+      "step": 2806
+    },
+    {
+      "epoch": 19.78502202643172,
+      "grad_norm": 0.21774391483577973,
+      "learning_rate": 0.0001369593552351113,
+      "loss": 0.0359,
+      "step": 2807
+    },
+    {
+      "epoch": 19.792070484581497,
+      "grad_norm": 0.21795172756629996,
+      "learning_rate": 0.00013691666125772626,
+      "loss": 0.0372,
+      "step": 2808
+    },
+    {
+      "epoch": 19.799118942731276,
+      "grad_norm": 0.2195000191363589,
+      "learning_rate": 0.00013687395948807176,
+      "loss": 0.0367,
+      "step": 2809
+    },
+    {
+      "epoch": 19.80616740088106,
+      "grad_norm": 0.23391366917872616,
+      "learning_rate": 0.00013683124993516115,
+      "loss": 0.0388,
+      "step": 2810
+    },
+    {
+      "epoch": 19.813215859030837,
+      "grad_norm": 0.21863538426773851,
+      "learning_rate": 0.00013678853260800955,
+      "loss": 0.0376,
+      "step": 2811
+    },
+    {
+      "epoch": 19.820264317180616,
+      "grad_norm": 0.21633352000134629,
+      "learning_rate": 0.00013674580751563356,
+      "loss": 0.037,
+      "step": 2812
+    },
+    {
+      "epoch": 19.827312775330398,
+      "grad_norm": 0.2210304072766595,
+      "learning_rate": 0.00013670307466705144,
+      "loss": 0.0382,
+      "step": 2813
+    },
+    {
+      "epoch": 19.834361233480177,
+      "grad_norm": 0.21966114508755571,
+      "learning_rate": 0.00013666033407128318,
+      "loss": 0.0373,
+      "step": 2814
+    },
+    {
+      "epoch": 19.841409691629956,
+      "grad_norm": 0.2285043453863557,
+      "learning_rate": 0.00013661758573735032,
+      "loss": 0.0404,
+      "step": 2815
+    },
+    {
+      "epoch": 19.848458149779734,
+      "grad_norm": 0.2297052318200462,
+      "learning_rate": 0.00013657482967427607,
+      "loss": 0.0368,
+      "step": 2816
+    },
+    {
+      "epoch": 19.855506607929517,
+      "grad_norm": 0.2321623956220717,
+      "learning_rate": 0.00013653206589108526,
+      "loss": 0.0406,
+      "step": 2817
+    },
+    {
+      "epoch": 19.862555066079295,
+      "grad_norm": 0.22696548355700663,
+      "learning_rate": 0.00013648929439680433,
+      "loss": 0.0367,
+      "step": 2818
+    },
+    {
+      "epoch": 19.869603524229074,
+      "grad_norm": 0.2251947461667959,
+      "learning_rate": 0.00013644651520046142,
+      "loss": 0.0373,
+      "step": 2819
+    },
+    {
+      "epoch": 19.876651982378856,
+      "grad_norm": 0.228096665551841,
+      "learning_rate": 0.00013640372831108625,
+      "loss": 0.0414,
+      "step": 2820
+    },
+    {
+      "epoch": 19.883700440528635,
+      "grad_norm": 0.23159794222540053,
+      "learning_rate": 0.00013636093373771015,
+      "loss": 0.0389,
+      "step": 2821
+    },
+    {
+      "epoch": 19.890748898678414,
+      "grad_norm": 0.2290032369755993,
+      "learning_rate": 0.00013631813148936604,
+      "loss": 0.0408,
+      "step": 2822
+    },
+    {
+      "epoch": 19.897797356828193,
+      "grad_norm": 0.21452615800846858,
+      "learning_rate": 0.00013627532157508857,
+      "loss": 0.037,
+      "step": 2823
+    },
+    {
+      "epoch": 19.904845814977975,
+      "grad_norm": 0.20973691658008436,
+      "learning_rate": 0.00013623250400391385,
+      "loss": 0.0412,
+      "step": 2824
+    },
+    {
+      "epoch": 19.911894273127754,
+      "grad_norm": 0.2338877307940816,
+      "learning_rate": 0.00013618967878487983,
+      "loss": 0.0403,
+      "step": 2825
+    },
+    {
+      "epoch": 19.918942731277532,
+      "grad_norm": 0.220009821026798,
+      "learning_rate": 0.00013614684592702582,
+      "loss": 0.0392,
+      "step": 2826
+    },
+    {
+      "epoch": 19.925991189427315,
+      "grad_norm": 0.23921519822823661,
+      "learning_rate": 0.00013610400543939295,
+      "loss": 0.0392,
+      "step": 2827
+    },
+    {
+      "epoch": 19.933039647577093,
+      "grad_norm": 0.2186679211562813,
+      "learning_rate": 0.00013606115733102382,
+      "loss": 0.0367,
+      "step": 2828
+    },
+    {
+      "epoch": 19.940088105726872,
+      "grad_norm": 0.23666460613470716,
+      "learning_rate": 0.00013601830161096267,
+      "loss": 0.0393,
+      "step": 2829
+    },
+    {
+      "epoch": 19.94713656387665,
+      "grad_norm": 0.2232446032444742,
+      "learning_rate": 0.00013597543828825547,
+      "loss": 0.0415,
+      "step": 2830
+    },
+    {
+      "epoch": 19.954185022026433,
+      "grad_norm": 0.2324811728069831,
+      "learning_rate": 0.0001359325673719496,
+      "loss": 0.0386,
+      "step": 2831
+    },
+    {
+      "epoch": 19.96123348017621,
+      "grad_norm": 0.23728229298363482,
+      "learning_rate": 0.00013588968887109416,
+      "loss": 0.0429,
+      "step": 2832
+    },
+    {
+      "epoch": 19.96828193832599,
+      "grad_norm": 0.22148023508003664,
+      "learning_rate": 0.00013584680279473988,
+      "loss": 0.0387,
+      "step": 2833
+    },
+    {
+      "epoch": 19.97533039647577,
+      "grad_norm": 0.23859941153538497,
+      "learning_rate": 0.00013580390915193897,
+      "loss": 0.0395,
+      "step": 2834
+    },
+    {
+      "epoch": 19.98237885462555,
+      "grad_norm": 0.23431716115401835,
+      "learning_rate": 0.00013576100795174533,
+      "loss": 0.0437,
+      "step": 2835
+    },
+    {
+      "epoch": 19.98942731277533,
+      "grad_norm": 0.22089435402976268,
+      "learning_rate": 0.00013571809920321447,
+      "loss": 0.037,
+      "step": 2836
+    },
+    {
+      "epoch": 19.99647577092511,
+      "grad_norm": 0.21984895448001568,
+      "learning_rate": 0.00013567518291540337,
+      "loss": 0.0418,
+      "step": 2837
+    },
+    {
+      "epoch": 20.00352422907489,
+      "grad_norm": 0.21083698729969524,
+      "learning_rate": 0.00013563225909737076,
+      "loss": 0.0323,
+      "step": 2838
+    },
+    {
+      "epoch": 20.01057268722467,
+      "grad_norm": 0.16256631739958827,
+      "learning_rate": 0.00013558932775817686,
+      "loss": 0.0201,
+      "step": 2839
+    },
+    {
+      "epoch": 20.01762114537445,
+      "grad_norm": 0.17060683063233198,
+      "learning_rate": 0.00013554638890688346,
+      "loss": 0.0223,
+      "step": 2840
+    },
+    {
+      "epoch": 20.024669603524227,
+      "grad_norm": 0.17054386116453807,
+      "learning_rate": 0.00013550344255255404,
+      "loss": 0.0199,
+      "step": 2841
+    },
+    {
+      "epoch": 20.03171806167401,
+      "grad_norm": 0.1806360961405031,
+      "learning_rate": 0.00013546048870425356,
+      "loss": 0.0209,
+      "step": 2842
+    },
+    {
+      "epoch": 20.03876651982379,
+      "grad_norm": 0.17729389102836493,
+      "learning_rate": 0.0001354175273710486,
+      "loss": 0.0228,
+      "step": 2843
+    },
+    {
+      "epoch": 20.045814977973567,
+      "grad_norm": 0.18809856340282124,
+      "learning_rate": 0.00013537455856200742,
+      "loss": 0.0216,
+      "step": 2844
+    },
+    {
+      "epoch": 20.05286343612335,
+      "grad_norm": 0.16852282171202548,
+      "learning_rate": 0.0001353315822861996,
+      "loss": 0.0198,
+      "step": 2845
+    },
+    {
+      "epoch": 20.059911894273128,
+      "grad_norm": 0.17414464254851109,
+      "learning_rate": 0.00013528859855269657,
+      "loss": 0.0193,
+      "step": 2846
+    },
+    {
+      "epoch": 20.066960352422907,
+      "grad_norm": 0.17539535795231115,
+      "learning_rate": 0.00013524560737057117,
+      "loss": 0.0198,
+      "step": 2847
+    },
+    {
+      "epoch": 20.074008810572685,
+      "grad_norm": 0.15695316061068842,
+      "learning_rate": 0.0001352026087488979,
+      "loss": 0.018,
+      "step": 2848
+    },
+    {
+      "epoch": 20.081057268722468,
+      "grad_norm": 0.1609581600382654,
+      "learning_rate": 0.00013515960269675276,
+      "loss": 0.0201,
+      "step": 2849
+    },
+    {
+      "epoch": 20.088105726872246,
+      "grad_norm": 0.18650103295584017,
+      "learning_rate": 0.00013511658922321336,
+      "loss": 0.0226,
+      "step": 2850
+    },
+    {
+      "epoch": 20.095154185022025,
+      "grad_norm": 0.16836758210713412,
+      "learning_rate": 0.00013507356833735888,
+      "loss": 0.0201,
+      "step": 2851
+    },
+    {
+      "epoch": 20.102202643171807,
+      "grad_norm": 0.17356976412499112,
+      "learning_rate": 0.00013503054004827006,
+      "loss": 0.0229,
+      "step": 2852
+    },
+    {
+      "epoch": 20.109251101321586,
+      "grad_norm": 0.16360860905683716,
+      "learning_rate": 0.00013498750436502914,
+      "loss": 0.0201,
+      "step": 2853
+    },
+    {
+      "epoch": 20.116299559471365,
+      "grad_norm": 0.16507809401676823,
+      "learning_rate": 0.00013494446129672002,
+      "loss": 0.0205,
+      "step": 2854
+    },
+    {
+      "epoch": 20.123348017621144,
+      "grad_norm": 0.17422641705647116,
+      "learning_rate": 0.00013490141085242816,
+      "loss": 0.0231,
+      "step": 2855
+    },
+    {
+      "epoch": 20.130396475770926,
+      "grad_norm": 0.172857679705162,
+      "learning_rate": 0.00013485835304124044,
+      "loss": 0.0203,
+      "step": 2856
+    },
+    {
+      "epoch": 20.137444933920705,
+      "grad_norm": 0.17230692890344806,
+      "learning_rate": 0.00013481528787224545,
+      "loss": 0.0209,
+      "step": 2857
+    },
+    {
+      "epoch": 20.144493392070483,
+      "grad_norm": 0.16714483535846972,
+      "learning_rate": 0.00013477221535453327,
+      "loss": 0.0216,
+      "step": 2858
+    },
+    {
+      "epoch": 20.151541850220266,
+      "grad_norm": 0.1560226947557868,
+      "learning_rate": 0.00013472913549719549,
+      "loss": 0.0176,
+      "step": 2859
+    },
+    {
+      "epoch": 20.158590308370044,
+      "grad_norm": 0.16575691073752635,
+      "learning_rate": 0.0001346860483093253,
+      "loss": 0.0209,
+      "step": 2860
+    },
+    {
+      "epoch": 20.165638766519823,
+      "grad_norm": 0.16964782473984164,
+      "learning_rate": 0.00013464295380001743,
+      "loss": 0.02,
+      "step": 2861
+    },
+    {
+      "epoch": 20.172687224669602,
+      "grad_norm": 0.16873788821153823,
+      "learning_rate": 0.0001345998519783682,
+      "loss": 0.0225,
+      "step": 2862
+    },
+    {
+      "epoch": 20.179735682819384,
+      "grad_norm": 0.17194508248328602,
+      "learning_rate": 0.00013455674285347538,
+      "loss": 0.0205,
+      "step": 2863
+    },
+    {
+      "epoch": 20.186784140969163,
+      "grad_norm": 0.1735776735543323,
+      "learning_rate": 0.00013451362643443832,
+      "loss": 0.0208,
+      "step": 2864
+    },
+    {
+      "epoch": 20.19383259911894,
+      "grad_norm": 0.18339949883312687,
+      "learning_rate": 0.00013447050273035792,
+      "loss": 0.0226,
+      "step": 2865
+    },
+    {
+      "epoch": 20.200881057268724,
+      "grad_norm": 0.1709398090841172,
+      "learning_rate": 0.00013442737175033668,
+      "loss": 0.0203,
+      "step": 2866
+    },
+    {
+      "epoch": 20.207929515418503,
+      "grad_norm": 0.16447166434056157,
+      "learning_rate": 0.0001343842335034785,
+      "loss": 0.0194,
+      "step": 2867
+    },
+    {
+      "epoch": 20.21497797356828,
+      "grad_norm": 0.1672001165786165,
+      "learning_rate": 0.00013434108799888886,
+      "loss": 0.0197,
+      "step": 2868
+    },
+    {
+      "epoch": 20.22202643171806,
+      "grad_norm": 0.179765831053164,
+      "learning_rate": 0.00013429793524567492,
+      "loss": 0.0203,
+      "step": 2869
+    },
+    {
+      "epoch": 20.229074889867842,
+      "grad_norm": 0.17497398281553952,
+      "learning_rate": 0.0001342547752529451,
+      "loss": 0.02,
+      "step": 2870
+    },
+    {
+      "epoch": 20.23612334801762,
+      "grad_norm": 0.16856615279370046,
+      "learning_rate": 0.00013421160802980963,
+      "loss": 0.0194,
+      "step": 2871
+    },
+    {
+      "epoch": 20.2431718061674,
+      "grad_norm": 0.1781740357507537,
+      "learning_rate": 0.00013416843358538002,
+      "loss": 0.02,
+      "step": 2872
+    },
+    {
+      "epoch": 20.250220264317182,
+      "grad_norm": 0.17438361344717557,
+      "learning_rate": 0.00013412525192876948,
+      "loss": 0.0211,
+      "step": 2873
+    },
+    {
+      "epoch": 20.25726872246696,
+      "grad_norm": 0.16570195078490177,
+      "learning_rate": 0.00013408206306909267,
+      "loss": 0.0202,
+      "step": 2874
+    },
+    {
+      "epoch": 20.26431718061674,
+      "grad_norm": 0.17434005407619485,
+      "learning_rate": 0.00013403886701546572,
+      "loss": 0.0195,
+      "step": 2875
+    },
+    {
+      "epoch": 20.271365638766518,
+      "grad_norm": 0.16750314668557714,
+      "learning_rate": 0.00013399566377700643,
+      "loss": 0.0194,
+      "step": 2876
+    },
+    {
+      "epoch": 20.2784140969163,
+      "grad_norm": 0.1687832550168169,
+      "learning_rate": 0.00013395245336283396,
+      "loss": 0.0193,
+      "step": 2877
+    },
+    {
+      "epoch": 20.28546255506608,
+      "grad_norm": 0.15183809757973551,
+      "learning_rate": 0.0001339092357820691,
+      "loss": 0.0175,
+      "step": 2878
+    },
+    {
+      "epoch": 20.292511013215858,
+      "grad_norm": 0.18657369320703543,
+      "learning_rate": 0.00013386601104383401,
+      "loss": 0.022,
+      "step": 2879
+    },
+    {
+      "epoch": 20.29955947136564,
+      "grad_norm": 0.17770840869591029,
+      "learning_rate": 0.00013382277915725255,
+      "loss": 0.0212,
+      "step": 2880
+    },
+    {
+      "epoch": 20.30660792951542,
+      "grad_norm": 0.18438436001689157,
+      "learning_rate": 0.0001337795401314499,
+      "loss": 0.0219,
+      "step": 2881
+    },
+    {
+      "epoch": 20.313656387665198,
+      "grad_norm": 0.17171637483234886,
+      "learning_rate": 0.00013373629397555293,
+      "loss": 0.0202,
+      "step": 2882
+    },
+    {
+      "epoch": 20.320704845814976,
+      "grad_norm": 0.18203502261685173,
+      "learning_rate": 0.00013369304069868983,
+      "loss": 0.0215,
+      "step": 2883
+    },
+    {
+      "epoch": 20.32775330396476,
+      "grad_norm": 0.17201385520740115,
+      "learning_rate": 0.00013364978030999049,
+      "loss": 0.0202,
+      "step": 2884
+    },
+    {
+      "epoch": 20.334801762114537,
+      "grad_norm": 0.16489521440383975,
+      "learning_rate": 0.0001336065128185861,
+      "loss": 0.0193,
+      "step": 2885
+    },
+    {
+      "epoch": 20.341850220264316,
+      "grad_norm": 0.1676149510721141,
+      "learning_rate": 0.0001335632382336095,
+      "loss": 0.0193,
+      "step": 2886
+    },
+    {
+      "epoch": 20.3488986784141,
+      "grad_norm": 0.1863487204225832,
+      "learning_rate": 0.00013351995656419495,
+      "loss": 0.022,
+      "step": 2887
+    },
+    {
+      "epoch": 20.355947136563877,
+      "grad_norm": 0.16893584747943682,
+      "learning_rate": 0.00013347666781947823,
+      "loss": 0.0209,
+      "step": 2888
+    },
+    {
+      "epoch": 20.362995594713656,
+      "grad_norm": 0.1660281444096263,
+      "learning_rate": 0.0001334333720085966,
+      "loss": 0.021,
+      "step": 2889
+    },
+    {
+      "epoch": 20.370044052863435,
+      "grad_norm": 0.16057089765182625,
+      "learning_rate": 0.0001333900691406889,
+      "loss": 0.0187,
+      "step": 2890
+    },
+    {
+      "epoch": 20.377092511013217,
+      "grad_norm": 0.18119115362076615,
+      "learning_rate": 0.00013334675922489528,
+      "loss": 0.0221,
+      "step": 2891
+    },
+    {
+      "epoch": 20.384140969162996,
+      "grad_norm": 0.1743597556840215,
+      "learning_rate": 0.00013330344227035752,
+      "loss": 0.0198,
+      "step": 2892
+    },
+    {
+      "epoch": 20.391189427312774,
+      "grad_norm": 0.174205556018695,
+      "learning_rate": 0.00013326011828621888,
+      "loss": 0.0216,
+      "step": 2893
+    },
+    {
+      "epoch": 20.398237885462557,
+      "grad_norm": 0.1871108253541555,
+      "learning_rate": 0.000133216787281624,
+      "loss": 0.0225,
+      "step": 2894
+    },
+    {
+      "epoch": 20.405286343612335,
+      "grad_norm": 0.17302362525674378,
+      "learning_rate": 0.0001331734492657191,
+      "loss": 0.0212,
+      "step": 2895
+    },
+    {
+      "epoch": 20.412334801762114,
+      "grad_norm": 0.17168861635838256,
+      "learning_rate": 0.0001331301042476519,
+      "loss": 0.0219,
+      "step": 2896
+    },
+    {
+      "epoch": 20.419383259911893,
+      "grad_norm": 0.176718970835812,
+      "learning_rate": 0.0001330867522365715,
+      "loss": 0.0184,
+      "step": 2897
+    },
+    {
+      "epoch": 20.426431718061675,
+      "grad_norm": 0.16358498201336866,
+      "learning_rate": 0.00013304339324162853,
+      "loss": 0.0201,
+      "step": 2898
+    },
+    {
+      "epoch": 20.433480176211454,
+      "grad_norm": 0.171044005999453,
+      "learning_rate": 0.0001330000272719751,
+      "loss": 0.0218,
+      "step": 2899
+    },
+    {
+      "epoch": 20.440528634361232,
+      "grad_norm": 0.17675786045924702,
+      "learning_rate": 0.0001329566543367648,
+      "loss": 0.0215,
+      "step": 2900
+    },
+    {
+      "epoch": 20.447577092511015,
+      "grad_norm": 0.17773639652649506,
+      "learning_rate": 0.00013291327444515266,
+      "loss": 0.0217,
+      "step": 2901
+    },
+    {
+      "epoch": 20.454625550660793,
+      "grad_norm": 0.17175353185036427,
+      "learning_rate": 0.00013286988760629516,
+      "loss": 0.0211,
+      "step": 2902
+    },
+    {
+      "epoch": 20.461674008810572,
+      "grad_norm": 0.1825783306961895,
+      "learning_rate": 0.00013282649382935027,
+      "loss": 0.0221,
+      "step": 2903
+    },
+    {
+      "epoch": 20.46872246696035,
+      "grad_norm": 0.17570566452809463,
+      "learning_rate": 0.00013278309312347752,
+      "loss": 0.0215,
+      "step": 2904
+    },
+    {
+      "epoch": 20.475770925110133,
+      "grad_norm": 0.17569668942455016,
+      "learning_rate": 0.00013273968549783774,
+      "loss": 0.0238,
+      "step": 2905
+    },
+    {
+      "epoch": 20.482819383259912,
+      "grad_norm": 0.16858107994439772,
+      "learning_rate": 0.00013269627096159327,
+      "loss": 0.0198,
+      "step": 2906
+    },
+    {
+      "epoch": 20.48986784140969,
+      "grad_norm": 0.1700837444190655,
+      "learning_rate": 0.000132652849523908,
+      "loss": 0.0216,
+      "step": 2907
+    },
+    {
+      "epoch": 20.496916299559473,
+      "grad_norm": 0.17670361797413575,
+      "learning_rate": 0.00013260942119394718,
+      "loss": 0.0215,
+      "step": 2908
+    },
+    {
+      "epoch": 20.50396475770925,
+      "grad_norm": 0.1790263057346337,
+      "learning_rate": 0.00013256598598087755,
+      "loss": 0.0227,
+      "step": 2909
+    },
+    {
+      "epoch": 20.51101321585903,
+      "grad_norm": 0.18735944406236013,
+      "learning_rate": 0.00013252254389386732,
+      "loss": 0.0228,
+      "step": 2910
+    },
+    {
+      "epoch": 20.51806167400881,
+      "grad_norm": 0.17467075322148684,
+      "learning_rate": 0.00013247909494208608,
+      "loss": 0.0212,
+      "step": 2911
+    },
+    {
+      "epoch": 20.52511013215859,
+      "grad_norm": 0.16897237873937204,
+      "learning_rate": 0.00013243563913470493,
+      "loss": 0.021,
+      "step": 2912
+    },
+    {
+      "epoch": 20.53215859030837,
+      "grad_norm": 0.18114171675024507,
+      "learning_rate": 0.00013239217648089647,
+      "loss": 0.0218,
+      "step": 2913
+    },
+    {
+      "epoch": 20.53920704845815,
+      "grad_norm": 0.18523769866055992,
+      "learning_rate": 0.00013234870698983458,
+      "loss": 0.0229,
+      "step": 2914
+    },
+    {
+      "epoch": 20.54625550660793,
+      "grad_norm": 0.1737993173576876,
+      "learning_rate": 0.00013230523067069475,
+      "loss": 0.0201,
+      "step": 2915
+    },
+    {
+      "epoch": 20.55330396475771,
+      "grad_norm": 0.1754794073773065,
+      "learning_rate": 0.0001322617475326538,
+      "loss": 0.0213,
+      "step": 2916
+    },
+    {
+      "epoch": 20.56035242290749,
+      "grad_norm": 0.17643942942478638,
+      "learning_rate": 0.00013221825758489008,
+      "loss": 0.0219,
+      "step": 2917
+    },
+    {
+      "epoch": 20.567400881057267,
+      "grad_norm": 0.1825267803812904,
+      "learning_rate": 0.00013217476083658334,
+      "loss": 0.0204,
+      "step": 2918
+    },
+    {
+      "epoch": 20.57444933920705,
+      "grad_norm": 0.18344734198237933,
+      "learning_rate": 0.0001321312572969147,
+      "loss": 0.0218,
+      "step": 2919
+    },
+    {
+      "epoch": 20.58149779735683,
+      "grad_norm": 0.18597501855399678,
+      "learning_rate": 0.0001320877469750668,
+      "loss": 0.0225,
+      "step": 2920
+    },
+    {
+      "epoch": 20.588546255506607,
+      "grad_norm": 0.17555893933659777,
+      "learning_rate": 0.00013204422988022366,
+      "loss": 0.0216,
+      "step": 2921
+    },
+    {
+      "epoch": 20.59559471365639,
+      "grad_norm": 0.17488832109675292,
+      "learning_rate": 0.00013200070602157075,
+      "loss": 0.0215,
+      "step": 2922
+    },
+    {
+      "epoch": 20.602643171806168,
+      "grad_norm": 0.17689028156599984,
+      "learning_rate": 0.00013195717540829508,
+      "loss": 0.0215,
+      "step": 2923
+    },
+    {
+      "epoch": 20.609691629955947,
+      "grad_norm": 0.1827032495804693,
+      "learning_rate": 0.00013191363804958476,
+      "loss": 0.0234,
+      "step": 2924
+    },
+    {
+      "epoch": 20.616740088105725,
+      "grad_norm": 0.17145360740046084,
+      "learning_rate": 0.00013187009395462974,
+      "loss": 0.0231,
+      "step": 2925
+    },
+    {
+      "epoch": 20.623788546255508,
+      "grad_norm": 0.1691441127509779,
+      "learning_rate": 0.0001318265431326211,
+      "loss": 0.0204,
+      "step": 2926
+    },
+    {
+      "epoch": 20.630837004405286,
+      "grad_norm": 0.17970565203302688,
+      "learning_rate": 0.00013178298559275145,
+      "loss": 0.0225,
+      "step": 2927
+    },
+    {
+      "epoch": 20.637885462555065,
+      "grad_norm": 0.18068659097148218,
+      "learning_rate": 0.00013173942134421475,
+      "loss": 0.0231,
+      "step": 2928
+    },
+    {
+      "epoch": 20.644933920704847,
+      "grad_norm": 0.17221653594800235,
+      "learning_rate": 0.0001316958503962065,
+      "loss": 0.0229,
+      "step": 2929
+    },
+    {
+      "epoch": 20.651982378854626,
+      "grad_norm": 0.1785907872511279,
+      "learning_rate": 0.0001316522727579235,
+      "loss": 0.0207,
+      "step": 2930
+    },
+    {
+      "epoch": 20.659030837004405,
+      "grad_norm": 0.18325274889049364,
+      "learning_rate": 0.000131608688438564,
+      "loss": 0.0248,
+      "step": 2931
+    },
+    {
+      "epoch": 20.666079295154184,
+      "grad_norm": 0.1874484990153694,
+      "learning_rate": 0.00013156509744732765,
+      "loss": 0.0212,
+      "step": 2932
+    },
+    {
+      "epoch": 20.673127753303966,
+      "grad_norm": 0.1844198333754738,
+      "learning_rate": 0.00013152149979341553,
+      "loss": 0.0236,
+      "step": 2933
+    },
+    {
+      "epoch": 20.680176211453745,
+      "grad_norm": 0.1877119528847876,
+      "learning_rate": 0.00013147789548603015,
+      "loss": 0.0228,
+      "step": 2934
+    },
+    {
+      "epoch": 20.687224669603523,
+      "grad_norm": 0.18291136710073644,
+      "learning_rate": 0.00013143428453437535,
+      "loss": 0.0227,
+      "step": 2935
+    },
+    {
+      "epoch": 20.694273127753306,
+      "grad_norm": 0.1851182546632981,
+      "learning_rate": 0.0001313906669476564,
+      "loss": 0.0239,
+      "step": 2936
+    },
+    {
+      "epoch": 20.701321585903084,
+      "grad_norm": 0.18681118504432515,
+      "learning_rate": 0.00013134704273508002,
+      "loss": 0.0243,
+      "step": 2937
+    },
+    {
+      "epoch": 20.708370044052863,
+      "grad_norm": 0.1762707366855808,
+      "learning_rate": 0.00013130341190585425,
+      "loss": 0.0211,
+      "step": 2938
+    },
+    {
+      "epoch": 20.715418502202642,
+      "grad_norm": 0.18129052406569854,
+      "learning_rate": 0.00013125977446918862,
+      "loss": 0.0218,
+      "step": 2939
+    },
+    {
+      "epoch": 20.722466960352424,
+      "grad_norm": 0.21211626919661405,
+      "learning_rate": 0.00013121613043429395,
+      "loss": 0.0231,
+      "step": 2940
+    },
+    {
+      "epoch": 20.729515418502203,
+      "grad_norm": 0.1818148826890189,
+      "learning_rate": 0.00013117247981038257,
+      "loss": 0.0241,
+      "step": 2941
+    },
+    {
+      "epoch": 20.73656387665198,
+      "grad_norm": 0.19019963483683408,
+      "learning_rate": 0.00013112882260666805,
+      "loss": 0.022,
+      "step": 2942
+    },
+    {
+      "epoch": 20.743612334801764,
+      "grad_norm": 0.1938245773197182,
+      "learning_rate": 0.00013108515883236552,
+      "loss": 0.0243,
+      "step": 2943
+    },
+    {
+      "epoch": 20.750660792951543,
+      "grad_norm": 0.1896180712898312,
+      "learning_rate": 0.00013104148849669135,
+      "loss": 0.0229,
+      "step": 2944
+    },
+    {
+      "epoch": 20.75770925110132,
+      "grad_norm": 0.1925980844414913,
+      "learning_rate": 0.0001309978116088634,
+      "loss": 0.0239,
+      "step": 2945
+    },
+    {
+      "epoch": 20.7647577092511,
+      "grad_norm": 0.18594008202521656,
+      "learning_rate": 0.00013095412817810083,
+      "loss": 0.0244,
+      "step": 2946
+    },
+    {
+      "epoch": 20.771806167400882,
+      "grad_norm": 0.19160080735601354,
+      "learning_rate": 0.00013091043821362422,
+      "loss": 0.0242,
+      "step": 2947
+    },
+    {
+      "epoch": 20.77885462555066,
+      "grad_norm": 0.19301420260608096,
+      "learning_rate": 0.00013086674172465562,
+      "loss": 0.0244,
+      "step": 2948
+    },
+    {
+      "epoch": 20.78590308370044,
+      "grad_norm": 0.1849911271978143,
+      "learning_rate": 0.00013082303872041825,
+      "loss": 0.0234,
+      "step": 2949
+    },
+    {
+      "epoch": 20.792951541850222,
+      "grad_norm": 0.18808869713999227,
+      "learning_rate": 0.00013077932921013688,
+      "loss": 0.0248,
+      "step": 2950
+    },
+    {
+      "epoch": 20.8,
+      "grad_norm": 0.1880090093617261,
+      "learning_rate": 0.0001307356132030376,
+      "loss": 0.0215,
+      "step": 2951
+    },
+    {
+      "epoch": 20.80704845814978,
+      "grad_norm": 0.19606052964015644,
+      "learning_rate": 0.00013069189070834782,
+      "loss": 0.0226,
+      "step": 2952
+    },
+    {
+      "epoch": 20.814096916299558,
+      "grad_norm": 0.18988087086395664,
+      "learning_rate": 0.0001306481617352964,
+      "loss": 0.0252,
+      "step": 2953
+    },
+    {
+      "epoch": 20.82114537444934,
+      "grad_norm": 0.1865172850425947,
+      "learning_rate": 0.00013060442629311356,
+      "loss": 0.0224,
+      "step": 2954
+    },
+    {
+      "epoch": 20.82819383259912,
+      "grad_norm": 0.19435930392283224,
+      "learning_rate": 0.00013056068439103085,
+      "loss": 0.0217,
+      "step": 2955
+    },
+    {
+      "epoch": 20.835242290748898,
+      "grad_norm": 0.19790132469600033,
+      "learning_rate": 0.00013051693603828113,
+      "loss": 0.0256,
+      "step": 2956
+    },
+    {
+      "epoch": 20.842290748898677,
+      "grad_norm": 0.1987807094226923,
+      "learning_rate": 0.00013047318124409876,
+      "loss": 0.0252,
+      "step": 2957
+    },
+    {
+      "epoch": 20.84933920704846,
+      "grad_norm": 0.1829937176156822,
+      "learning_rate": 0.00013042942001771933,
+      "loss": 0.0215,
+      "step": 2958
+    },
+    {
+      "epoch": 20.856387665198238,
+      "grad_norm": 0.19126691650284539,
+      "learning_rate": 0.0001303856523683799,
+      "loss": 0.0229,
+      "step": 2959
+    },
+    {
+      "epoch": 20.863436123348016,
+      "grad_norm": 0.18757699434221897,
+      "learning_rate": 0.0001303418783053188,
+      "loss": 0.0227,
+      "step": 2960
+    },
+    {
+      "epoch": 20.8704845814978,
+      "grad_norm": 0.1929756560557066,
+      "learning_rate": 0.00013029809783777566,
+      "loss": 0.0242,
+      "step": 2961
+    },
+    {
+      "epoch": 20.877533039647577,
+      "grad_norm": 0.20730983271132186,
+      "learning_rate": 0.0001302543109749917,
+      "loss": 0.0279,
+      "step": 2962
+    },
+    {
+      "epoch": 20.884581497797356,
+      "grad_norm": 0.1741364517085196,
+      "learning_rate": 0.0001302105177262092,
+      "loss": 0.0258,
+      "step": 2963
+    },
+    {
+      "epoch": 20.89162995594714,
+      "grad_norm": 0.18310889601728558,
+      "learning_rate": 0.000130166718100672,
+      "loss": 0.0258,
+      "step": 2964
+    },
+    {
+      "epoch": 20.898678414096917,
+      "grad_norm": 0.20399429428538787,
+      "learning_rate": 0.0001301229121076252,
+      "loss": 0.0268,
+      "step": 2965
+    },
+    {
+      "epoch": 20.905726872246696,
+      "grad_norm": 0.1926609771814044,
+      "learning_rate": 0.00013007909975631517,
+      "loss": 0.0247,
+      "step": 2966
+    },
+    {
+      "epoch": 20.912775330396475,
+      "grad_norm": 0.1969354849200596,
+      "learning_rate": 0.00013003528105598987,
+      "loss": 0.0255,
+      "step": 2967
+    },
+    {
+      "epoch": 20.919823788546257,
+      "grad_norm": 0.1852654865054112,
+      "learning_rate": 0.00012999145601589823,
+      "loss": 0.0244,
+      "step": 2968
+    },
+    {
+      "epoch": 20.926872246696036,
+      "grad_norm": 0.20377743311338467,
+      "learning_rate": 0.0001299476246452908,
+      "loss": 0.0253,
+      "step": 2969
+    },
+    {
+      "epoch": 20.933920704845814,
+      "grad_norm": 0.20215919464976995,
+      "learning_rate": 0.00012990378695341948,
+      "loss": 0.0269,
+      "step": 2970
+    },
+    {
+      "epoch": 20.940969162995593,
+      "grad_norm": 0.21127580420619546,
+      "learning_rate": 0.00012985994294953732,
+      "loss": 0.027,
+      "step": 2971
+    },
+    {
+      "epoch": 20.948017621145375,
+      "grad_norm": 0.19924959255851463,
+      "learning_rate": 0.0001298160926428988,
+      "loss": 0.0281,
+      "step": 2972
+    },
+    {
+      "epoch": 20.955066079295154,
+      "grad_norm": 0.19290437609017508,
+      "learning_rate": 0.00012977223604275975,
+      "loss": 0.0253,
+      "step": 2973
+    },
+    {
+      "epoch": 20.962114537444933,
+      "grad_norm": 0.19615535683124694,
+      "learning_rate": 0.00012972837315837726,
+      "loss": 0.0258,
+      "step": 2974
+    },
+    {
+      "epoch": 20.969162995594715,
+      "grad_norm": 0.20089830346728438,
+      "learning_rate": 0.00012968450399900987,
+      "loss": 0.0262,
+      "step": 2975
+    },
+    {
+      "epoch": 20.976211453744494,
+      "grad_norm": 0.18797752678160165,
+      "learning_rate": 0.00012964062857391727,
+      "loss": 0.0249,
+      "step": 2976
+    },
+    {
+      "epoch": 20.983259911894272,
+      "grad_norm": 0.21272765909846342,
+      "learning_rate": 0.00012959674689236062,
+      "loss": 0.0266,
+      "step": 2977
+    },
+    {
+      "epoch": 20.99030837004405,
+      "grad_norm": 0.20006119625438543,
+      "learning_rate": 0.00012955285896360238,
+      "loss": 0.026,
+      "step": 2978
+    },
+    {
+      "epoch": 20.997356828193833,
+      "grad_norm": 0.1990024768687335,
+      "learning_rate": 0.00012950896479690618,
+      "loss": 0.0262,
+      "step": 2979
+    },
+    {
+      "epoch": 21.004405286343612,
+      "grad_norm": 0.1695526157070196,
+      "learning_rate": 0.00012946506440153718,
+      "loss": 0.0183,
+      "step": 2980
+    },
+    {
+      "epoch": 21.01145374449339,
+      "grad_norm": 0.13878326024030527,
+      "learning_rate": 0.00012942115778676177,
+      "loss": 0.0142,
+      "step": 2981
+    },
+    {
+      "epoch": 21.018502202643173,
+      "grad_norm": 0.14550155616235105,
+      "learning_rate": 0.00012937724496184756,
+      "loss": 0.0138,
+      "step": 2982
+    },
+    {
+      "epoch": 21.025550660792952,
+      "grad_norm": 0.13377225534861129,
+      "learning_rate": 0.00012933332593606362,
+      "loss": 0.0131,
+      "step": 2983
+    },
+    {
+      "epoch": 21.03259911894273,
+      "grad_norm": 0.15243960715518046,
+      "learning_rate": 0.00012928940071868023,
+      "loss": 0.0138,
+      "step": 2984
+    },
+    {
+      "epoch": 21.03964757709251,
+      "grad_norm": 0.1523918691602627,
+      "learning_rate": 0.000129245469318969,
+      "loss": 0.0141,
+      "step": 2985
+    },
+    {
+      "epoch": 21.04669603524229,
+      "grad_norm": 0.15133789692340635,
+      "learning_rate": 0.00012920153174620285,
+      "loss": 0.0133,
+      "step": 2986
+    },
+    {
+      "epoch": 21.05374449339207,
+      "grad_norm": 0.15801321211805947,
+      "learning_rate": 0.00012915758800965605,
+      "loss": 0.0144,
+      "step": 2987
+    },
+    {
+      "epoch": 21.06079295154185,
+      "grad_norm": 0.1575364455489394,
+      "learning_rate": 0.00012911363811860405,
+      "loss": 0.0144,
+      "step": 2988
+    },
+    {
+      "epoch": 21.06784140969163,
+      "grad_norm": 0.1548126627699904,
+      "learning_rate": 0.00012906968208232378,
+      "loss": 0.0155,
+      "step": 2989
+    },
+    {
+      "epoch": 21.07488986784141,
+      "grad_norm": 0.142732554531002,
+      "learning_rate": 0.00012902571991009321,
+      "loss": 0.0142,
+      "step": 2990
+    },
+    {
+      "epoch": 21.08193832599119,
+      "grad_norm": 0.14500511607934324,
+      "learning_rate": 0.00012898175161119192,
+      "loss": 0.014,
+      "step": 2991
+    },
+    {
+      "epoch": 21.088986784140968,
+      "grad_norm": 0.1675063068447884,
+      "learning_rate": 0.00012893777719490053,
+      "loss": 0.0151,
+      "step": 2992
+    },
+    {
+      "epoch": 21.09603524229075,
+      "grad_norm": 0.15713225385359486,
+      "learning_rate": 0.00012889379667050107,
+      "loss": 0.0133,
+      "step": 2993
+    },
+    {
+      "epoch": 21.10308370044053,
+      "grad_norm": 0.1417689146081461,
+      "learning_rate": 0.00012884981004727676,
+      "loss": 0.0135,
+      "step": 2994
+    },
+    {
+      "epoch": 21.110132158590307,
+      "grad_norm": 0.14656208124124445,
+      "learning_rate": 0.00012880581733451226,
+      "loss": 0.0137,
+      "step": 2995
+    },
+    {
+      "epoch": 21.11718061674009,
+      "grad_norm": 0.14110036813914442,
+      "learning_rate": 0.00012876181854149343,
+      "loss": 0.0143,
+      "step": 2996
+    },
+    {
+      "epoch": 21.12422907488987,
+      "grad_norm": 0.13803822444887878,
+      "learning_rate": 0.0001287178136775074,
+      "loss": 0.0134,
+      "step": 2997
+    },
+    {
+      "epoch": 21.131277533039647,
+      "grad_norm": 0.1464911925680797,
+      "learning_rate": 0.00012867380275184259,
+      "loss": 0.013,
+      "step": 2998
+    },
+    {
+      "epoch": 21.138325991189426,
+      "grad_norm": 0.14469181016624408,
+      "learning_rate": 0.00012862978577378865,
+      "loss": 0.0138,
+      "step": 2999
+    },
+    {
+      "epoch": 21.145374449339208,
+      "grad_norm": 0.1418096397261533,
+      "learning_rate": 0.00012858576275263673,
+      "loss": 0.0134,
+      "step": 3000
+    },
+    {
+      "epoch": 21.152422907488987,
+      "grad_norm": 0.14908972182071095,
+      "learning_rate": 0.00012854173369767892,
+      "loss": 0.0136,
+      "step": 3001
+    },
+    {
+      "epoch": 21.159471365638765,
+      "grad_norm": 0.15414375725432578,
+      "learning_rate": 0.00012849769861820882,
+      "loss": 0.0131,
+      "step": 3002
+    },
+    {
+      "epoch": 21.166519823788548,
+      "grad_norm": 0.14698209640794394,
+      "learning_rate": 0.00012845365752352125,
+      "loss": 0.0142,
+      "step": 3003
+    },
+    {
+      "epoch": 21.173568281938326,
+      "grad_norm": 0.1494886004768104,
+      "learning_rate": 0.00012840961042291226,
+      "loss": 0.0136,
+      "step": 3004
+    },
+    {
+      "epoch": 21.180616740088105,
+      "grad_norm": 0.14918029469341154,
+      "learning_rate": 0.00012836555732567923,
+      "loss": 0.014,
+      "step": 3005
+    },
+    {
+      "epoch": 21.187665198237884,
+      "grad_norm": 0.1483256522799087,
+      "learning_rate": 0.00012832149824112073,
+      "loss": 0.014,
+      "step": 3006
+    },
+    {
+      "epoch": 21.194713656387666,
+      "grad_norm": 0.14023647232482975,
+      "learning_rate": 0.00012827743317853665,
+      "loss": 0.0126,
+      "step": 3007
+    },
+    {
+      "epoch": 21.201762114537445,
+      "grad_norm": 0.1384744616070591,
+      "learning_rate": 0.00012823336214722817,
+      "loss": 0.0133,
+      "step": 3008
+    },
+    {
+      "epoch": 21.208810572687224,
+      "grad_norm": 0.15416631108366616,
+      "learning_rate": 0.0001281892851564976,
+      "loss": 0.014,
+      "step": 3009
+    },
+    {
+      "epoch": 21.215859030837006,
+      "grad_norm": 0.14138933055796665,
+      "learning_rate": 0.0001281452022156486,
+      "loss": 0.0121,
+      "step": 3010
+    },
+    {
+      "epoch": 21.222907488986785,
+      "grad_norm": 0.1469757686197374,
+      "learning_rate": 0.00012810111333398614,
+      "loss": 0.0135,
+      "step": 3011
+    },
+    {
+      "epoch": 21.229955947136563,
+      "grad_norm": 0.13745777431954798,
+      "learning_rate": 0.00012805701852081636,
+      "loss": 0.0133,
+      "step": 3012
+    },
+    {
+      "epoch": 21.237004405286342,
+      "grad_norm": 0.15878505966117729,
+      "learning_rate": 0.00012801291778544665,
+      "loss": 0.0131,
+      "step": 3013
+    },
+    {
+      "epoch": 21.244052863436124,
+      "grad_norm": 0.14222114331522911,
+      "learning_rate": 0.00012796881113718572,
+      "loss": 0.0139,
+      "step": 3014
+    },
+    {
+      "epoch": 21.251101321585903,
+      "grad_norm": 0.1358389977684869,
+      "learning_rate": 0.0001279246985853434,
+      "loss": 0.013,
+      "step": 3015
+    },
+    {
+      "epoch": 21.25814977973568,
+      "grad_norm": 0.15155226747902079,
+      "learning_rate": 0.00012788058013923095,
+      "loss": 0.0133,
+      "step": 3016
+    },
+    {
+      "epoch": 21.265198237885464,
+      "grad_norm": 0.14029312850626588,
+      "learning_rate": 0.0001278364558081607,
+      "loss": 0.013,
+      "step": 3017
+    },
+    {
+      "epoch": 21.272246696035243,
+      "grad_norm": 0.13915890591605026,
+      "learning_rate": 0.0001277923256014463,
+      "loss": 0.0116,
+      "step": 3018
+    },
+    {
+      "epoch": 21.27929515418502,
+      "grad_norm": 0.13286094815514862,
+      "learning_rate": 0.0001277481895284027,
+      "loss": 0.0118,
+      "step": 3019
+    },
+    {
+      "epoch": 21.2863436123348,
+      "grad_norm": 0.14044547100238874,
+      "learning_rate": 0.00012770404759834594,
+      "loss": 0.0131,
+      "step": 3020
+    },
+    {
+      "epoch": 21.293392070484582,
+      "grad_norm": 0.15513217077781072,
+      "learning_rate": 0.00012765989982059344,
+      "loss": 0.0135,
+      "step": 3021
+    },
+    {
+      "epoch": 21.30044052863436,
+      "grad_norm": 0.14644447328263832,
+      "learning_rate": 0.00012761574620446378,
+      "loss": 0.0143,
+      "step": 3022
+    },
+    {
+      "epoch": 21.30748898678414,
+      "grad_norm": 0.14206893890984051,
+      "learning_rate": 0.00012757158675927673,
+      "loss": 0.0136,
+      "step": 3023
+    },
+    {
+      "epoch": 21.314537444933922,
+      "grad_norm": 0.14838832787674117,
+      "learning_rate": 0.00012752742149435344,
+      "loss": 0.0135,
+      "step": 3024
+    },
+    {
+      "epoch": 21.3215859030837,
+      "grad_norm": 0.1521372220600201,
+      "learning_rate": 0.00012748325041901613,
+      "loss": 0.0156,
+      "step": 3025
+    },
+    {
+      "epoch": 21.32863436123348,
+      "grad_norm": 0.13537720931557562,
+      "learning_rate": 0.00012743907354258836,
+      "loss": 0.0126,
+      "step": 3026
+    },
+    {
+      "epoch": 21.33568281938326,
+      "grad_norm": 0.13492200524014938,
+      "learning_rate": 0.0001273948908743948,
+      "loss": 0.0122,
+      "step": 3027
+    },
+    {
+      "epoch": 21.34273127753304,
+      "grad_norm": 0.14651335467307067,
+      "learning_rate": 0.00012735070242376147,
+      "loss": 0.0139,
+      "step": 3028
+    },
+    {
+      "epoch": 21.34977973568282,
+      "grad_norm": 0.15579989880277817,
+      "learning_rate": 0.0001273065082000155,
+      "loss": 0.0125,
+      "step": 3029
+    },
+    {
+      "epoch": 21.356828193832598,
+      "grad_norm": 0.15145944047774293,
+      "learning_rate": 0.00012726230821248535,
+      "loss": 0.0132,
+      "step": 3030
+    },
+    {
+      "epoch": 21.36387665198238,
+      "grad_norm": 0.14286511064286223,
+      "learning_rate": 0.00012721810247050058,
+      "loss": 0.0134,
+      "step": 3031
+    },
+    {
+      "epoch": 21.37092511013216,
+      "grad_norm": 0.14007696056225422,
+      "learning_rate": 0.00012717389098339203,
+      "loss": 0.0127,
+      "step": 3032
+    },
+    {
+      "epoch": 21.377973568281938,
+      "grad_norm": 0.15948231127302478,
+      "learning_rate": 0.00012712967376049176,
+      "loss": 0.0133,
+      "step": 3033
+    },
+    {
+      "epoch": 21.385022026431717,
+      "grad_norm": 0.13682464223610824,
+      "learning_rate": 0.00012708545081113298,
+      "loss": 0.0133,
+      "step": 3034
+    },
+    {
+      "epoch": 21.3920704845815,
+      "grad_norm": 0.1495042858868715,
+      "learning_rate": 0.00012704122214465021,
+      "loss": 0.0135,
+      "step": 3035
+    },
+    {
+      "epoch": 21.399118942731278,
+      "grad_norm": 0.14752377103495978,
+      "learning_rate": 0.0001269969877703791,
+      "loss": 0.0145,
+      "step": 3036
+    },
+    {
+      "epoch": 21.406167400881056,
+      "grad_norm": 0.135787951486112,
+      "learning_rate": 0.0001269527476976565,
+      "loss": 0.0133,
+      "step": 3037
+    },
+    {
+      "epoch": 21.41321585903084,
+      "grad_norm": 0.14494592718325755,
+      "learning_rate": 0.0001269085019358205,
+      "loss": 0.0131,
+      "step": 3038
+    },
+    {
+      "epoch": 21.420264317180617,
+      "grad_norm": 0.14001515434854694,
+      "learning_rate": 0.00012686425049421036,
+      "loss": 0.0119,
+      "step": 3039
+    },
+    {
+      "epoch": 21.427312775330396,
+      "grad_norm": 0.1663373177349362,
+      "learning_rate": 0.00012681999338216656,
+      "loss": 0.0138,
+      "step": 3040
+    },
+    {
+      "epoch": 21.434361233480175,
+      "grad_norm": 0.15060117676998644,
+      "learning_rate": 0.00012677573060903083,
+      "loss": 0.0129,
+      "step": 3041
+    },
+    {
+      "epoch": 21.441409691629957,
+      "grad_norm": 0.14926756785955217,
+      "learning_rate": 0.00012673146218414595,
+      "loss": 0.0136,
+      "step": 3042
+    },
+    {
+      "epoch": 21.448458149779736,
+      "grad_norm": 0.1505179772101389,
+      "learning_rate": 0.000126687188116856,
+      "loss": 0.0127,
+      "step": 3043
+    },
+    {
+      "epoch": 21.455506607929514,
+      "grad_norm": 0.1406829913413487,
+      "learning_rate": 0.0001266429084165063,
+      "loss": 0.0131,
+      "step": 3044
+    },
+    {
+      "epoch": 21.462555066079297,
+      "grad_norm": 0.1493802154524197,
+      "learning_rate": 0.00012659862309244318,
+      "loss": 0.0149,
+      "step": 3045
+    },
+    {
+      "epoch": 21.469603524229075,
+      "grad_norm": 0.1431482595563594,
+      "learning_rate": 0.00012655433215401438,
+      "loss": 0.013,
+      "step": 3046
+    },
+    {
+      "epoch": 21.476651982378854,
+      "grad_norm": 0.1473257142145613,
+      "learning_rate": 0.00012651003561056865,
+      "loss": 0.0142,
+      "step": 3047
+    },
+    {
+      "epoch": 21.483700440528633,
+      "grad_norm": 0.13981185789150277,
+      "learning_rate": 0.00012646573347145596,
+      "loss": 0.0134,
+      "step": 3048
+    },
+    {
+      "epoch": 21.490748898678415,
+      "grad_norm": 0.1655673813774558,
+      "learning_rate": 0.00012642142574602753,
+      "loss": 0.0156,
+      "step": 3049
+    },
+    {
+      "epoch": 21.497797356828194,
+      "grad_norm": 0.1685324868514702,
+      "learning_rate": 0.00012637711244363573,
+      "loss": 0.0146,
+      "step": 3050
+    },
+    {
+      "epoch": 21.504845814977973,
+      "grad_norm": 0.15421572467254127,
+      "learning_rate": 0.00012633279357363402,
+      "loss": 0.0143,
+      "step": 3051
+    },
+    {
+      "epoch": 21.511894273127755,
+      "grad_norm": 0.14358641548674594,
+      "learning_rate": 0.00012628846914537723,
+      "loss": 0.0135,
+      "step": 3052
+    },
+    {
+      "epoch": 21.518942731277534,
+      "grad_norm": 0.15779187890926147,
+      "learning_rate": 0.0001262441391682211,
+      "loss": 0.0147,
+      "step": 3053
+    },
+    {
+      "epoch": 21.525991189427312,
+      "grad_norm": 0.1582174418500736,
+      "learning_rate": 0.00012619980365152276,
+      "loss": 0.015,
+      "step": 3054
+    },
+    {
+      "epoch": 21.53303964757709,
+      "grad_norm": 0.1545720644793578,
+      "learning_rate": 0.00012615546260464044,
+      "loss": 0.0145,
+      "step": 3055
+    },
+    {
+      "epoch": 21.540088105726873,
+      "grad_norm": 0.14391423952531607,
+      "learning_rate": 0.00012611111603693347,
+      "loss": 0.0141,
+      "step": 3056
+    },
+    {
+      "epoch": 21.547136563876652,
+      "grad_norm": 0.16599631986012336,
+      "learning_rate": 0.00012606676395776246,
+      "loss": 0.0157,
+      "step": 3057
+    },
+    {
+      "epoch": 21.55418502202643,
+      "grad_norm": 0.1642613495975359,
+      "learning_rate": 0.00012602240637648912,
+      "loss": 0.0147,
+      "step": 3058
+    },
+    {
+      "epoch": 21.561233480176213,
+      "grad_norm": 0.14592573844167697,
+      "learning_rate": 0.0001259780433024763,
+      "loss": 0.0132,
+      "step": 3059
+    },
+    {
+      "epoch": 21.568281938325992,
+      "grad_norm": 0.1538518615587383,
+      "learning_rate": 0.0001259336747450881,
+      "loss": 0.0136,
+      "step": 3060
+    },
+    {
+      "epoch": 21.57533039647577,
+      "grad_norm": 0.1524455046020259,
+      "learning_rate": 0.0001258893007136896,
+      "loss": 0.0144,
+      "step": 3061
+    },
+    {
+      "epoch": 21.58237885462555,
+      "grad_norm": 0.1616121220284199,
+      "learning_rate": 0.0001258449212176473,
+      "loss": 0.0148,
+      "step": 3062
+    },
+    {
+      "epoch": 21.58942731277533,
+      "grad_norm": 0.14070656783945293,
+      "learning_rate": 0.00012580053626632857,
+      "loss": 0.013,
+      "step": 3063
+    },
+    {
+      "epoch": 21.59647577092511,
+      "grad_norm": 0.1500878375531971,
+      "learning_rate": 0.00012575614586910215,
+      "loss": 0.0144,
+      "step": 3064
+    },
+    {
+      "epoch": 21.60352422907489,
+      "grad_norm": 0.15098707183791646,
+      "learning_rate": 0.00012571175003533782,
+      "loss": 0.014,
+      "step": 3065
+    },
+    {
+      "epoch": 21.61057268722467,
+      "grad_norm": 0.15544211913396008,
+      "learning_rate": 0.0001256673487744065,
+      "loss": 0.0157,
+      "step": 3066
+    },
+    {
+      "epoch": 21.61762114537445,
+      "grad_norm": 0.17184828383382325,
+      "learning_rate": 0.00012562294209568037,
+      "loss": 0.0162,
+      "step": 3067
+    },
+    {
+      "epoch": 21.62466960352423,
+      "grad_norm": 0.15834151357244566,
+      "learning_rate": 0.00012557853000853258,
+      "loss": 0.0154,
+      "step": 3068
+    },
+    {
+      "epoch": 21.631718061674007,
+      "grad_norm": 0.15055810080236162,
+      "learning_rate": 0.0001255341125223376,
+      "loss": 0.0133,
+      "step": 3069
+    },
+    {
+      "epoch": 21.63876651982379,
+      "grad_norm": 0.1511720338390272,
+      "learning_rate": 0.00012548968964647085,
+      "loss": 0.015,
+      "step": 3070
+    },
+    {
+      "epoch": 21.64581497797357,
+      "grad_norm": 0.1474580699684598,
+      "learning_rate": 0.00012544526139030913,
+      "loss": 0.0136,
+      "step": 3071
+    },
+    {
+      "epoch": 21.652863436123347,
+      "grad_norm": 0.14395465628830836,
+      "learning_rate": 0.00012540082776323007,
+      "loss": 0.0134,
+      "step": 3072
+    },
+    {
+      "epoch": 21.65991189427313,
+      "grad_norm": 0.1564803966809426,
+      "learning_rate": 0.00012535638877461276,
+      "loss": 0.0146,
+      "step": 3073
+    },
+    {
+      "epoch": 21.666960352422908,
+      "grad_norm": 0.153247763476535,
+      "learning_rate": 0.00012531194443383717,
+      "loss": 0.0149,
+      "step": 3074
+    },
+    {
+      "epoch": 21.674008810572687,
+      "grad_norm": 0.15825573128937753,
+      "learning_rate": 0.0001252674947502845,
+      "loss": 0.0165,
+      "step": 3075
+    },
+    {
+      "epoch": 21.681057268722466,
+      "grad_norm": 0.15080178586243204,
+      "learning_rate": 0.00012522303973333705,
+      "loss": 0.0146,
+      "step": 3076
+    },
+    {
+      "epoch": 21.688105726872248,
+      "grad_norm": 0.15747656074915345,
+      "learning_rate": 0.00012517857939237835,
+      "loss": 0.015,
+      "step": 3077
+    },
+    {
+      "epoch": 21.695154185022027,
+      "grad_norm": 0.15332637679304817,
+      "learning_rate": 0.0001251341137367929,
+      "loss": 0.0148,
+      "step": 3078
+    },
+    {
+      "epoch": 21.702202643171805,
+      "grad_norm": 0.15507449770997703,
+      "learning_rate": 0.00012508964277596643,
+      "loss": 0.0149,
+      "step": 3079
+    },
+    {
+      "epoch": 21.709251101321584,
+      "grad_norm": 0.1540436506930393,
+      "learning_rate": 0.00012504516651928567,
+      "loss": 0.0157,
+      "step": 3080
+    },
+    {
+      "epoch": 21.716299559471366,
+      "grad_norm": 0.1590823732400862,
+      "learning_rate": 0.00012500068497613864,
+      "loss": 0.0151,
+      "step": 3081
+    },
+    {
+      "epoch": 21.723348017621145,
+      "grad_norm": 0.15345572579989364,
+      "learning_rate": 0.00012495619815591438,
+      "loss": 0.0147,
+      "step": 3082
+    },
+    {
+      "epoch": 21.730396475770924,
+      "grad_norm": 0.17366462888640213,
+      "learning_rate": 0.00012491170606800298,
+      "loss": 0.0161,
+      "step": 3083
+    },
+    {
+      "epoch": 21.737444933920706,
+      "grad_norm": 0.15714583716263988,
+      "learning_rate": 0.00012486720872179573,
+      "loss": 0.0147,
+      "step": 3084
+    },
+    {
+      "epoch": 21.744493392070485,
+      "grad_norm": 0.1660457061331935,
+      "learning_rate": 0.00012482270612668508,
+      "loss": 0.0139,
+      "step": 3085
+    },
+    {
+      "epoch": 21.751541850220264,
+      "grad_norm": 0.15460110713865924,
+      "learning_rate": 0.0001247781982920644,
+      "loss": 0.0146,
+      "step": 3086
+    },
+    {
+      "epoch": 21.758590308370046,
+      "grad_norm": 0.1704896300518509,
+      "learning_rate": 0.0001247336852273284,
+      "loss": 0.0164,
+      "step": 3087
+    },
+    {
+      "epoch": 21.765638766519825,
+      "grad_norm": 0.16296820163208622,
+      "learning_rate": 0.00012468916694187272,
+      "loss": 0.0145,
+      "step": 3088
+    },
+    {
+      "epoch": 21.772687224669603,
+      "grad_norm": 0.1604316039684942,
+      "learning_rate": 0.00012464464344509417,
+      "loss": 0.0159,
+      "step": 3089
+    },
+    {
+      "epoch": 21.779735682819382,
+      "grad_norm": 0.15492245869957197,
+      "learning_rate": 0.00012460011474639063,
+      "loss": 0.0156,
+      "step": 3090
+    },
+    {
+      "epoch": 21.786784140969164,
+      "grad_norm": 0.1599762363331318,
+      "learning_rate": 0.00012455558085516113,
+      "loss": 0.0164,
+      "step": 3091
+    },
+    {
+      "epoch": 21.793832599118943,
+      "grad_norm": 0.15275780548359377,
+      "learning_rate": 0.00012451104178080576,
+      "loss": 0.0171,
+      "step": 3092
+    },
+    {
+      "epoch": 21.80088105726872,
+      "grad_norm": 0.15689072806575272,
+      "learning_rate": 0.00012446649753272568,
+      "loss": 0.0154,
+      "step": 3093
+    },
+    {
+      "epoch": 21.8079295154185,
+      "grad_norm": 0.15033867245329668,
+      "learning_rate": 0.00012442194812032319,
+      "loss": 0.0153,
+      "step": 3094
+    },
+    {
+      "epoch": 21.814977973568283,
+      "grad_norm": 0.1543618723707061,
+      "learning_rate": 0.00012437739355300163,
+      "loss": 0.0146,
+      "step": 3095
+    },
+    {
+      "epoch": 21.82202643171806,
+      "grad_norm": 0.17058452076954764,
+      "learning_rate": 0.00012433283384016554,
+      "loss": 0.017,
+      "step": 3096
+    },
+    {
+      "epoch": 21.82907488986784,
+      "grad_norm": 0.16870108114110963,
+      "learning_rate": 0.00012428826899122037,
+      "loss": 0.0148,
+      "step": 3097
+    },
+    {
+      "epoch": 21.836123348017622,
+      "grad_norm": 0.16397037543915377,
+      "learning_rate": 0.0001242436990155728,
+      "loss": 0.0156,
+      "step": 3098
+    },
+    {
+      "epoch": 21.8431718061674,
+      "grad_norm": 0.1635904730297827,
+      "learning_rate": 0.00012419912392263056,
+      "loss": 0.015,
+      "step": 3099
+    },
+    {
+      "epoch": 21.85022026431718,
+      "grad_norm": 0.16307975421021748,
+      "learning_rate": 0.00012415454372180235,
+      "loss": 0.0155,
+      "step": 3100
+    },
+    {
+      "epoch": 21.857268722466962,
+      "grad_norm": 0.16566255886604556,
+      "learning_rate": 0.00012410995842249812,
+      "loss": 0.0161,
+      "step": 3101
+    },
+    {
+      "epoch": 21.86431718061674,
+      "grad_norm": 0.15302723608877533,
+      "learning_rate": 0.00012406536803412878,
+      "loss": 0.0158,
+      "step": 3102
+    },
+    {
+      "epoch": 21.87136563876652,
+      "grad_norm": 0.1739309416918089,
+      "learning_rate": 0.0001240207725661064,
+      "loss": 0.0162,
+      "step": 3103
+    },
+    {
+      "epoch": 21.8784140969163,
+      "grad_norm": 0.15876740638581646,
+      "learning_rate": 0.000123976172027844,
+      "loss": 0.0163,
+      "step": 3104
+    },
+    {
+      "epoch": 21.88546255506608,
+      "grad_norm": 0.16194972255400228,
+      "learning_rate": 0.0001239315664287558,
+      "loss": 0.0156,
+      "step": 3105
+    },
+    {
+      "epoch": 21.89251101321586,
+      "grad_norm": 0.16229272968193345,
+      "learning_rate": 0.00012388695577825694,
+      "loss": 0.0176,
+      "step": 3106
+    },
+    {
+      "epoch": 21.899559471365638,
+      "grad_norm": 0.1696603782730657,
+      "learning_rate": 0.00012384234008576387,
+      "loss": 0.0161,
+      "step": 3107
+    },
+    {
+      "epoch": 21.906607929515417,
+      "grad_norm": 0.16315760621219288,
+      "learning_rate": 0.00012379771936069382,
+      "loss": 0.016,
+      "step": 3108
+    },
+    {
+      "epoch": 21.9136563876652,
+      "grad_norm": 0.16733509877535896,
+      "learning_rate": 0.00012375309361246524,
+      "loss": 0.017,
+      "step": 3109
+    },
+    {
+      "epoch": 21.920704845814978,
+      "grad_norm": 0.15780128951646352,
+      "learning_rate": 0.00012370846285049764,
+      "loss": 0.0162,
+      "step": 3110
+    },
+    {
+      "epoch": 21.927753303964757,
+      "grad_norm": 0.15591013150773045,
+      "learning_rate": 0.00012366382708421154,
+      "loss": 0.0158,
+      "step": 3111
+    },
+    {
+      "epoch": 21.93480176211454,
+      "grad_norm": 0.17239578326018284,
+      "learning_rate": 0.00012361918632302856,
+      "loss": 0.017,
+      "step": 3112
+    },
+    {
+      "epoch": 21.941850220264318,
+      "grad_norm": 0.16055829825250242,
+      "learning_rate": 0.0001235745405763713,
+      "loss": 0.014,
+      "step": 3113
+    },
+    {
+      "epoch": 21.948898678414096,
+      "grad_norm": 0.16790338054593051,
+      "learning_rate": 0.0001235298898536635,
+      "loss": 0.0183,
+      "step": 3114
+    },
+    {
+      "epoch": 21.955947136563875,
+      "grad_norm": 0.17028339204975917,
+      "learning_rate": 0.00012348523416433,
+      "loss": 0.0162,
+      "step": 3115
+    },
+    {
+      "epoch": 21.962995594713657,
+      "grad_norm": 0.1762340028729618,
+      "learning_rate": 0.0001234405735177964,
+      "loss": 0.016,
+      "step": 3116
+    },
+    {
+      "epoch": 21.970044052863436,
+      "grad_norm": 0.15772487916960184,
+      "learning_rate": 0.00012339590792348975,
+      "loss": 0.0168,
+      "step": 3117
+    },
+    {
+      "epoch": 21.977092511013215,
+      "grad_norm": 0.15549782225223713,
+      "learning_rate": 0.00012335123739083781,
+      "loss": 0.016,
+      "step": 3118
+    },
+    {
+      "epoch": 21.984140969162997,
+      "grad_norm": 0.17415116477464257,
+      "learning_rate": 0.00012330656192926958,
+      "loss": 0.0159,
+      "step": 3119
+    },
+    {
+      "epoch": 21.991189427312776,
+      "grad_norm": 0.15484530343516797,
+      "learning_rate": 0.00012326188154821504,
+      "loss": 0.0152,
+      "step": 3120
+    },
+    {
+      "epoch": 21.998237885462554,
+      "grad_norm": 0.1657890434581184,
+      "learning_rate": 0.00012321719625710514,
+      "loss": 0.0157,
+      "step": 3121
+    },
+    {
+      "epoch": 22.005286343612333,
+      "grad_norm": 0.12556545090483706,
+      "learning_rate": 0.000123172506065372,
+      "loss": 0.0115,
+      "step": 3122
+    },
+    {
+      "epoch": 22.012334801762115,
+      "grad_norm": 0.10893927833035612,
+      "learning_rate": 0.0001231278109824487,
+      "loss": 0.0083,
+      "step": 3123
+    },
+    {
+      "epoch": 22.019383259911894,
+      "grad_norm": 0.12115213925352977,
+      "learning_rate": 0.00012308311101776932,
+      "loss": 0.0093,
+      "step": 3124
+    },
+    {
+      "epoch": 22.026431718061673,
+      "grad_norm": 0.12684402666814842,
+      "learning_rate": 0.00012303840618076902,
+      "loss": 0.0093,
+      "step": 3125
+    },
+    {
+      "epoch": 22.033480176211455,
+      "grad_norm": 0.13261318260024296,
+      "learning_rate": 0.00012299369648088399,
+      "loss": 0.0091,
+      "step": 3126
+    },
+    {
+      "epoch": 22.040528634361234,
+      "grad_norm": 0.12020588838442633,
+      "learning_rate": 0.00012294898192755142,
+      "loss": 0.0094,
+      "step": 3127
+    },
+    {
+      "epoch": 22.047577092511013,
+      "grad_norm": 0.12747698366189075,
+      "learning_rate": 0.00012290426253020958,
+      "loss": 0.0099,
+      "step": 3128
+    },
+    {
+      "epoch": 22.05462555066079,
+      "grad_norm": 0.12319716059818307,
+      "learning_rate": 0.00012285953829829763,
+      "loss": 0.0093,
+      "step": 3129
+    },
+    {
+      "epoch": 22.061674008810574,
+      "grad_norm": 0.134389550556326,
+      "learning_rate": 0.00012281480924125594,
+      "loss": 0.0095,
+      "step": 3130
+    },
+    {
+      "epoch": 22.068722466960352,
+      "grad_norm": 0.1201701508560148,
+      "learning_rate": 0.00012277007536852573,
+      "loss": 0.0094,
+      "step": 3131
+    },
+    {
+      "epoch": 22.07577092511013,
+      "grad_norm": 0.1221996049125982,
+      "learning_rate": 0.00012272533668954933,
+      "loss": 0.0091,
+      "step": 3132
+    },
+    {
+      "epoch": 22.082819383259913,
+      "grad_norm": 0.11179750300576245,
+      "learning_rate": 0.00012268059321377008,
+      "loss": 0.0087,
+      "step": 3133
+    },
+    {
+      "epoch": 22.089867841409692,
+      "grad_norm": 0.12597127554695112,
+      "learning_rate": 0.00012263584495063228,
+      "loss": 0.0092,
+      "step": 3134
+    },
+    {
+      "epoch": 22.09691629955947,
+      "grad_norm": 0.1207435786600922,
+      "learning_rate": 0.00012259109190958126,
+      "loss": 0.0089,
+      "step": 3135
+    },
+    {
+      "epoch": 22.10396475770925,
+      "grad_norm": 0.12343716304732733,
+      "learning_rate": 0.0001225463341000634,
+      "loss": 0.0096,
+      "step": 3136
+    },
+    {
+      "epoch": 22.111013215859032,
+      "grad_norm": 0.10949840638002578,
+      "learning_rate": 0.0001225015715315261,
+      "loss": 0.0082,
+      "step": 3137
+    },
+    {
+      "epoch": 22.11806167400881,
+      "grad_norm": 0.12194476667705394,
+      "learning_rate": 0.00012245680421341763,
+      "loss": 0.0093,
+      "step": 3138
+    },
+    {
+      "epoch": 22.12511013215859,
+      "grad_norm": 0.1286774583555487,
+      "learning_rate": 0.00012241203215518741,
+      "loss": 0.0099,
+      "step": 3139
+    },
+    {
+      "epoch": 22.13215859030837,
+      "grad_norm": 0.13912548016434315,
+      "learning_rate": 0.00012236725536628583,
+      "loss": 0.0095,
+      "step": 3140
+    },
+    {
+      "epoch": 22.13920704845815,
+      "grad_norm": 0.11801650618437876,
+      "learning_rate": 0.00012232247385616417,
+      "loss": 0.0086,
+      "step": 3141
+    },
+    {
+      "epoch": 22.14625550660793,
+      "grad_norm": 0.12220204806994242,
+      "learning_rate": 0.00012227768763427486,
+      "loss": 0.0086,
+      "step": 3142
+    },
+    {
+      "epoch": 22.153303964757708,
+      "grad_norm": 0.12112555795023909,
+      "learning_rate": 0.00012223289671007125,
+      "loss": 0.0088,
+      "step": 3143
+    },
+    {
+      "epoch": 22.16035242290749,
+      "grad_norm": 0.10887822352209625,
+      "learning_rate": 0.00012218810109300768,
+      "loss": 0.0079,
+      "step": 3144
+    },
+    {
+      "epoch": 22.16740088105727,
+      "grad_norm": 0.13963071849631525,
+      "learning_rate": 0.00012214330079253946,
+      "loss": 0.0101,
+      "step": 3145
+    },
+    {
+      "epoch": 22.174449339207047,
+      "grad_norm": 0.12208429425603531,
+      "learning_rate": 0.00012209849581812295,
+      "loss": 0.0098,
+      "step": 3146
+    },
+    {
+      "epoch": 22.18149779735683,
+      "grad_norm": 0.12565264652280825,
+      "learning_rate": 0.00012205368617921542,
+      "loss": 0.0091,
+      "step": 3147
+    },
+    {
+      "epoch": 22.18854625550661,
+      "grad_norm": 0.12980842253585095,
+      "learning_rate": 0.00012200887188527526,
+      "loss": 0.0094,
+      "step": 3148
+    },
+    {
+      "epoch": 22.195594713656387,
+      "grad_norm": 0.12012235374376917,
+      "learning_rate": 0.00012196405294576165,
+      "loss": 0.0091,
+      "step": 3149
+    },
+    {
+      "epoch": 22.202643171806166,
+      "grad_norm": 0.12607720140219322,
+      "learning_rate": 0.00012191922937013489,
+      "loss": 0.0086,
+      "step": 3150
+    },
+    {
+      "epoch": 22.209691629955948,
+      "grad_norm": 0.11376592749967829,
+      "learning_rate": 0.00012187440116785625,
+      "loss": 0.0081,
+      "step": 3151
+    },
+    {
+      "epoch": 22.216740088105727,
+      "grad_norm": 0.11778038442921679,
+      "learning_rate": 0.00012182956834838789,
+      "loss": 0.0096,
+      "step": 3152
+    },
+    {
+      "epoch": 22.223788546255506,
+      "grad_norm": 0.12095738310744059,
+      "learning_rate": 0.00012178473092119302,
+      "loss": 0.0092,
+      "step": 3153
+    },
+    {
+      "epoch": 22.230837004405288,
+      "grad_norm": 0.11411005888065183,
+      "learning_rate": 0.00012173988889573585,
+      "loss": 0.0092,
+      "step": 3154
+    },
+    {
+      "epoch": 22.237885462555067,
+      "grad_norm": 0.11313170196571291,
+      "learning_rate": 0.00012169504228148145,
+      "loss": 0.0084,
+      "step": 3155
+    },
+    {
+      "epoch": 22.244933920704845,
+      "grad_norm": 0.12833389473568435,
+      "learning_rate": 0.00012165019108789597,
+      "loss": 0.0085,
+      "step": 3156
+    },
+    {
+      "epoch": 22.251982378854624,
+      "grad_norm": 0.12025691636131917,
+      "learning_rate": 0.00012160533532444642,
+      "loss": 0.0092,
+      "step": 3157
+    },
+    {
+      "epoch": 22.259030837004406,
+      "grad_norm": 0.11390461061231183,
+      "learning_rate": 0.00012156047500060089,
+      "loss": 0.0082,
+      "step": 3158
+    },
+    {
+      "epoch": 22.266079295154185,
+      "grad_norm": 0.12240247424437081,
+      "learning_rate": 0.00012151561012582835,
+      "loss": 0.0093,
+      "step": 3159
+    },
+    {
+      "epoch": 22.273127753303964,
+      "grad_norm": 0.11996017686623842,
+      "learning_rate": 0.00012147074070959879,
+      "loss": 0.0093,
+      "step": 3160
+    },
+    {
+      "epoch": 22.280176211453746,
+      "grad_norm": 0.11166292501408354,
+      "learning_rate": 0.00012142586676138308,
+      "loss": 0.0082,
+      "step": 3161
+    },
+    {
+      "epoch": 22.287224669603525,
+      "grad_norm": 0.1198122100338361,
+      "learning_rate": 0.0001213809882906531,
+      "loss": 0.0089,
+      "step": 3162
+    },
+    {
+      "epoch": 22.294273127753303,
+      "grad_norm": 0.11973602426659273,
+      "learning_rate": 0.00012133610530688168,
+      "loss": 0.0098,
+      "step": 3163
+    },
+    {
+      "epoch": 22.301321585903082,
+      "grad_norm": 0.12362839952720998,
+      "learning_rate": 0.00012129121781954266,
+      "loss": 0.0098,
+      "step": 3164
+    },
+    {
+      "epoch": 22.308370044052865,
+      "grad_norm": 0.13086157774352913,
+      "learning_rate": 0.00012124632583811068,
+      "loss": 0.0092,
+      "step": 3165
+    },
+    {
+      "epoch": 22.315418502202643,
+      "grad_norm": 0.11768351848155295,
+      "learning_rate": 0.00012120142937206147,
+      "loss": 0.0081,
+      "step": 3166
+    },
+    {
+      "epoch": 22.322466960352422,
+      "grad_norm": 0.11217038463881937,
+      "learning_rate": 0.00012115652843087164,
+      "loss": 0.0086,
+      "step": 3167
+    },
+    {
+      "epoch": 22.329515418502204,
+      "grad_norm": 0.12099893426381512,
+      "learning_rate": 0.00012111162302401875,
+      "loss": 0.0089,
+      "step": 3168
+    },
+    {
+      "epoch": 22.336563876651983,
+      "grad_norm": 0.12009188979027831,
+      "learning_rate": 0.00012106671316098133,
+      "loss": 0.0089,
+      "step": 3169
+    },
+    {
+      "epoch": 22.34361233480176,
+      "grad_norm": 0.12189388504769977,
+      "learning_rate": 0.00012102179885123883,
+      "loss": 0.0086,
+      "step": 3170
+    },
+    {
+      "epoch": 22.35066079295154,
+      "grad_norm": 0.11718149956883044,
+      "learning_rate": 0.00012097688010427164,
+      "loss": 0.0092,
+      "step": 3171
+    },
+    {
+      "epoch": 22.357709251101323,
+      "grad_norm": 0.1279078442110702,
+      "learning_rate": 0.0001209319569295611,
+      "loss": 0.0097,
+      "step": 3172
+    },
+    {
+      "epoch": 22.3647577092511,
+      "grad_norm": 0.10917306749982167,
+      "learning_rate": 0.00012088702933658944,
+      "loss": 0.0084,
+      "step": 3173
+    },
+    {
+      "epoch": 22.37180616740088,
+      "grad_norm": 0.1206783104143742,
+      "learning_rate": 0.00012084209733483989,
+      "loss": 0.0094,
+      "step": 3174
+    },
+    {
+      "epoch": 22.378854625550662,
+      "grad_norm": 0.13234514316124582,
+      "learning_rate": 0.00012079716093379654,
+      "loss": 0.0094,
+      "step": 3175
+    },
+    {
+      "epoch": 22.38590308370044,
+      "grad_norm": 0.11194741541138173,
+      "learning_rate": 0.00012075222014294447,
+      "loss": 0.0087,
+      "step": 3176
+    },
+    {
+      "epoch": 22.39295154185022,
+      "grad_norm": 0.12219652558793025,
+      "learning_rate": 0.00012070727497176966,
+      "loss": 0.0094,
+      "step": 3177
+    },
+    {
+      "epoch": 22.4,
+      "grad_norm": 0.10939942352137828,
+      "learning_rate": 0.00012066232542975902,
+      "loss": 0.0078,
+      "step": 3178
+    },
+    {
+      "epoch": 22.40704845814978,
+      "grad_norm": 0.12462125858583896,
+      "learning_rate": 0.00012061737152640033,
+      "loss": 0.0088,
+      "step": 3179
+    },
+    {
+      "epoch": 22.41409691629956,
+      "grad_norm": 0.1164505510478358,
+      "learning_rate": 0.00012057241327118242,
+      "loss": 0.0092,
+      "step": 3180
+    },
+    {
+      "epoch": 22.42114537444934,
+      "grad_norm": 0.12777230567317724,
+      "learning_rate": 0.00012052745067359494,
+      "loss": 0.0089,
+      "step": 3181
+    },
+    {
+      "epoch": 22.42819383259912,
+      "grad_norm": 0.12515657141431327,
+      "learning_rate": 0.0001204824837431284,
+      "loss": 0.0093,
+      "step": 3182
+    },
+    {
+      "epoch": 22.4352422907489,
+      "grad_norm": 0.12601370721784974,
+      "learning_rate": 0.00012043751248927438,
+      "loss": 0.0095,
+      "step": 3183
+    },
+    {
+      "epoch": 22.442290748898678,
+      "grad_norm": 0.11452575637928086,
+      "learning_rate": 0.00012039253692152523,
+      "loss": 0.0092,
+      "step": 3184
+    },
+    {
+      "epoch": 22.449339207048457,
+      "grad_norm": 0.1163238388393914,
+      "learning_rate": 0.00012034755704937432,
+      "loss": 0.009,
+      "step": 3185
+    },
+    {
+      "epoch": 22.45638766519824,
+      "grad_norm": 0.12854524286488872,
+      "learning_rate": 0.00012030257288231587,
+      "loss": 0.0092,
+      "step": 3186
+    },
+    {
+      "epoch": 22.463436123348018,
+      "grad_norm": 0.12462313865741496,
+      "learning_rate": 0.00012025758442984501,
+      "loss": 0.0088,
+      "step": 3187
+    },
+    {
+      "epoch": 22.470484581497796,
+      "grad_norm": 0.12048254060662165,
+      "learning_rate": 0.00012021259170145775,
+      "loss": 0.0088,
+      "step": 3188
+    },
+    {
+      "epoch": 22.47753303964758,
+      "grad_norm": 0.12376685940429377,
+      "learning_rate": 0.00012016759470665112,
+      "loss": 0.0097,
+      "step": 3189
+    },
+    {
+      "epoch": 22.484581497797357,
+      "grad_norm": 0.11897390553420942,
+      "learning_rate": 0.00012012259345492287,
+      "loss": 0.0085,
+      "step": 3190
+    },
+    {
+      "epoch": 22.491629955947136,
+      "grad_norm": 0.13091026733574526,
+      "learning_rate": 0.00012007758795577178,
+      "loss": 0.0092,
+      "step": 3191
+    },
+    {
+      "epoch": 22.498678414096915,
+      "grad_norm": 0.12318757198190276,
+      "learning_rate": 0.00012003257821869751,
+      "loss": 0.0085,
+      "step": 3192
+    },
+    {
+      "epoch": 22.505726872246697,
+      "grad_norm": 0.12839518646917816,
+      "learning_rate": 0.00011998756425320055,
+      "loss": 0.0094,
+      "step": 3193
+    },
+    {
+      "epoch": 22.512775330396476,
+      "grad_norm": 0.11890514024810639,
+      "learning_rate": 0.00011994254606878237,
+      "loss": 0.0084,
+      "step": 3194
+    },
+    {
+      "epoch": 22.519823788546255,
+      "grad_norm": 0.11747767765480079,
+      "learning_rate": 0.00011989752367494527,
+      "loss": 0.0088,
+      "step": 3195
+    },
+    {
+      "epoch": 22.526872246696037,
+      "grad_norm": 0.12642178345402413,
+      "learning_rate": 0.00011985249708119244,
+      "loss": 0.0082,
+      "step": 3196
+    },
+    {
+      "epoch": 22.533920704845816,
+      "grad_norm": 0.12714196190032975,
+      "learning_rate": 0.00011980746629702801,
+      "loss": 0.0089,
+      "step": 3197
+    },
+    {
+      "epoch": 22.540969162995594,
+      "grad_norm": 0.11881316931922706,
+      "learning_rate": 0.00011976243133195692,
+      "loss": 0.0087,
+      "step": 3198
+    },
+    {
+      "epoch": 22.548017621145373,
+      "grad_norm": 0.11680939266281955,
+      "learning_rate": 0.00011971739219548505,
+      "loss": 0.0092,
+      "step": 3199
+    },
+    {
+      "epoch": 22.555066079295155,
+      "grad_norm": 0.12433479668875466,
+      "learning_rate": 0.00011967234889711914,
+      "loss": 0.009,
+      "step": 3200
+    },
+    {
+      "epoch": 22.562114537444934,
+      "grad_norm": 0.11982045491506788,
+      "learning_rate": 0.00011962730144636678,
+      "loss": 0.0093,
+      "step": 3201
+    },
+    {
+      "epoch": 22.569162995594713,
+      "grad_norm": 0.1422859190823548,
+      "learning_rate": 0.00011958224985273648,
+      "loss": 0.0098,
+      "step": 3202
+    },
+    {
+      "epoch": 22.57621145374449,
+      "grad_norm": 0.13246788254401318,
+      "learning_rate": 0.00011953719412573765,
+      "loss": 0.0099,
+      "step": 3203
+    },
+    {
+      "epoch": 22.583259911894274,
+      "grad_norm": 0.12931386952563398,
+      "learning_rate": 0.00011949213427488044,
+      "loss": 0.01,
+      "step": 3204
+    },
+    {
+      "epoch": 22.590308370044053,
+      "grad_norm": 0.13136555011570075,
+      "learning_rate": 0.00011944707030967607,
+      "loss": 0.0102,
+      "step": 3205
+    },
+    {
+      "epoch": 22.59735682819383,
+      "grad_norm": 0.12983301552747367,
+      "learning_rate": 0.00011940200223963646,
+      "loss": 0.0093,
+      "step": 3206
+    },
+    {
+      "epoch": 22.604405286343614,
+      "grad_norm": 0.1214271312812068,
+      "learning_rate": 0.00011935693007427445,
+      "loss": 0.009,
+      "step": 3207
+    },
+    {
+      "epoch": 22.611453744493392,
+      "grad_norm": 0.12106591949831873,
+      "learning_rate": 0.0001193118538231038,
+      "loss": 0.0089,
+      "step": 3208
+    },
+    {
+      "epoch": 22.61850220264317,
+      "grad_norm": 0.12302418628572957,
+      "learning_rate": 0.00011926677349563899,
+      "loss": 0.0092,
+      "step": 3209
+    },
+    {
+      "epoch": 22.625550660792953,
+      "grad_norm": 0.1381128300102723,
+      "learning_rate": 0.00011922168910139555,
+      "loss": 0.01,
+      "step": 3210
+    },
+    {
+      "epoch": 22.632599118942732,
+      "grad_norm": 0.12568473417196854,
+      "learning_rate": 0.00011917660064988974,
+      "loss": 0.0091,
+      "step": 3211
+    },
+    {
+      "epoch": 22.63964757709251,
+      "grad_norm": 0.1366249135245959,
+      "learning_rate": 0.0001191315081506387,
+      "loss": 0.0119,
+      "step": 3212
+    },
+    {
+      "epoch": 22.64669603524229,
+      "grad_norm": 0.1291258059825842,
+      "learning_rate": 0.00011908641161316046,
+      "loss": 0.0102,
+      "step": 3213
+    },
+    {
+      "epoch": 22.65374449339207,
+      "grad_norm": 0.13050059509851172,
+      "learning_rate": 0.00011904131104697382,
+      "loss": 0.0095,
+      "step": 3214
+    },
+    {
+      "epoch": 22.66079295154185,
+      "grad_norm": 0.1292480017621066,
+      "learning_rate": 0.00011899620646159855,
+      "loss": 0.0089,
+      "step": 3215
+    },
+    {
+      "epoch": 22.66784140969163,
+      "grad_norm": 0.11867953841270383,
+      "learning_rate": 0.00011895109786655514,
+      "loss": 0.0091,
+      "step": 3216
+    },
+    {
+      "epoch": 22.674889867841408,
+      "grad_norm": 0.12162955828632682,
+      "learning_rate": 0.00011890598527136505,
+      "loss": 0.0097,
+      "step": 3217
+    },
+    {
+      "epoch": 22.68193832599119,
+      "grad_norm": 0.13529467461007558,
+      "learning_rate": 0.00011886086868555049,
+      "loss": 0.0098,
+      "step": 3218
+    },
+    {
+      "epoch": 22.68898678414097,
+      "grad_norm": 0.13681009342575054,
+      "learning_rate": 0.00011881574811863458,
+      "loss": 0.0102,
+      "step": 3219
+    },
+    {
+      "epoch": 22.696035242290748,
+      "grad_norm": 0.12367725093698298,
+      "learning_rate": 0.0001187706235801412,
+      "loss": 0.0095,
+      "step": 3220
+    },
+    {
+      "epoch": 22.70308370044053,
+      "grad_norm": 0.1236811749879852,
+      "learning_rate": 0.00011872549507959516,
+      "loss": 0.0096,
+      "step": 3221
+    },
+    {
+      "epoch": 22.71013215859031,
+      "grad_norm": 0.12664837640200519,
+      "learning_rate": 0.00011868036262652204,
+      "loss": 0.0103,
+      "step": 3222
+    },
+    {
+      "epoch": 22.717180616740087,
+      "grad_norm": 0.12849220842039868,
+      "learning_rate": 0.00011863522623044827,
+      "loss": 0.01,
+      "step": 3223
+    },
+    {
+      "epoch": 22.72422907488987,
+      "grad_norm": 0.13224693424770587,
+      "learning_rate": 0.00011859008590090115,
+      "loss": 0.0101,
+      "step": 3224
+    },
+    {
+      "epoch": 22.73127753303965,
+      "grad_norm": 0.12601619246214446,
+      "learning_rate": 0.00011854494164740875,
+      "loss": 0.0094,
+      "step": 3225
+    },
+    {
+      "epoch": 22.738325991189427,
+      "grad_norm": 0.13655798058570306,
+      "learning_rate": 0.00011849979347950002,
+      "loss": 0.0105,
+      "step": 3226
+    },
+    {
+      "epoch": 22.745374449339206,
+      "grad_norm": 0.1358218642840171,
+      "learning_rate": 0.00011845464140670469,
+      "loss": 0.0105,
+      "step": 3227
+    },
+    {
+      "epoch": 22.752422907488988,
+      "grad_norm": 0.11752639384899174,
+      "learning_rate": 0.00011840948543855335,
+      "loss": 0.0088,
+      "step": 3228
+    },
+    {
+      "epoch": 22.759471365638767,
+      "grad_norm": 0.13979996071389184,
+      "learning_rate": 0.00011836432558457737,
+      "loss": 0.0101,
+      "step": 3229
+    },
+    {
+      "epoch": 22.766519823788546,
+      "grad_norm": 0.1296188254888919,
+      "learning_rate": 0.00011831916185430905,
+      "loss": 0.0099,
+      "step": 3230
+    },
+    {
+      "epoch": 22.773568281938324,
+      "grad_norm": 0.12418535508325301,
+      "learning_rate": 0.00011827399425728133,
+      "loss": 0.009,
+      "step": 3231
+    },
+    {
+      "epoch": 22.780616740088107,
+      "grad_norm": 0.12017522631815108,
+      "learning_rate": 0.00011822882280302814,
+      "loss": 0.0102,
+      "step": 3232
+    },
+    {
+      "epoch": 22.787665198237885,
+      "grad_norm": 0.1301619718442125,
+      "learning_rate": 0.00011818364750108415,
+      "loss": 0.0095,
+      "step": 3233
+    },
+    {
+      "epoch": 22.794713656387664,
+      "grad_norm": 0.13589475991389596,
+      "learning_rate": 0.00011813846836098475,
+      "loss": 0.0096,
+      "step": 3234
+    },
+    {
+      "epoch": 22.801762114537446,
+      "grad_norm": 0.12779240480464835,
+      "learning_rate": 0.00011809328539226632,
+      "loss": 0.0093,
+      "step": 3235
+    },
+    {
+      "epoch": 22.808810572687225,
+      "grad_norm": 0.12319631460729075,
+      "learning_rate": 0.00011804809860446595,
+      "loss": 0.0104,
+      "step": 3236
+    },
+    {
+      "epoch": 22.815859030837004,
+      "grad_norm": 0.1385519057968378,
+      "learning_rate": 0.0001180029080071215,
+      "loss": 0.0096,
+      "step": 3237
+    },
+    {
+      "epoch": 22.822907488986782,
+      "grad_norm": 0.12199695152085928,
+      "learning_rate": 0.00011795771360977172,
+      "loss": 0.0098,
+      "step": 3238
+    },
+    {
+      "epoch": 22.829955947136565,
+      "grad_norm": 0.13548903813725438,
+      "learning_rate": 0.00011791251542195614,
+      "loss": 0.0112,
+      "step": 3239
+    },
+    {
+      "epoch": 22.837004405286343,
+      "grad_norm": 0.13236749032681439,
+      "learning_rate": 0.000117867313453215,
+      "loss": 0.0103,
+      "step": 3240
+    },
+    {
+      "epoch": 22.844052863436122,
+      "grad_norm": 0.1297314863667749,
+      "learning_rate": 0.00011782210771308948,
+      "loss": 0.0101,
+      "step": 3241
+    },
+    {
+      "epoch": 22.851101321585904,
+      "grad_norm": 0.139231505707908,
+      "learning_rate": 0.00011777689821112144,
+      "loss": 0.01,
+      "step": 3242
+    },
+    {
+      "epoch": 22.858149779735683,
+      "grad_norm": 0.1253673347341131,
+      "learning_rate": 0.00011773168495685358,
+      "loss": 0.0098,
+      "step": 3243
+    },
+    {
+      "epoch": 22.865198237885462,
+      "grad_norm": 0.13389133660580385,
+      "learning_rate": 0.00011768646795982945,
+      "loss": 0.011,
+      "step": 3244
+    },
+    {
+      "epoch": 22.87224669603524,
+      "grad_norm": 0.12727779153692284,
+      "learning_rate": 0.00011764124722959326,
+      "loss": 0.0106,
+      "step": 3245
+    },
+    {
+      "epoch": 22.879295154185023,
+      "grad_norm": 0.1288372413984455,
+      "learning_rate": 0.00011759602277569012,
+      "loss": 0.0104,
+      "step": 3246
+    },
+    {
+      "epoch": 22.8863436123348,
+      "grad_norm": 0.14486922316573456,
+      "learning_rate": 0.00011755079460766589,
+      "loss": 0.0121,
+      "step": 3247
+    },
+    {
+      "epoch": 22.89339207048458,
+      "grad_norm": 0.1334368500413506,
+      "learning_rate": 0.00011750556273506718,
+      "loss": 0.0104,
+      "step": 3248
+    },
+    {
+      "epoch": 22.900440528634363,
+      "grad_norm": 0.13748820663738756,
+      "learning_rate": 0.00011746032716744144,
+      "loss": 0.0102,
+      "step": 3249
+    },
+    {
+      "epoch": 22.90748898678414,
+      "grad_norm": 0.14201416987864446,
+      "learning_rate": 0.00011741508791433686,
+      "loss": 0.0102,
+      "step": 3250
+    },
+    {
+      "epoch": 22.91453744493392,
+      "grad_norm": 0.13637729674838528,
+      "learning_rate": 0.00011736984498530241,
+      "loss": 0.0099,
+      "step": 3251
+    },
+    {
+      "epoch": 22.9215859030837,
+      "grad_norm": 0.1427268015171895,
+      "learning_rate": 0.00011732459838988787,
+      "loss": 0.0106,
+      "step": 3252
+    },
+    {
+      "epoch": 22.92863436123348,
+      "grad_norm": 0.13254573779952689,
+      "learning_rate": 0.00011727934813764374,
+      "loss": 0.0103,
+      "step": 3253
+    },
+    {
+      "epoch": 22.93568281938326,
+      "grad_norm": 0.14121585956783156,
+      "learning_rate": 0.00011723409423812134,
+      "loss": 0.0107,
+      "step": 3254
+    },
+    {
+      "epoch": 22.94273127753304,
+      "grad_norm": 0.14653026262542473,
+      "learning_rate": 0.00011718883670087275,
+      "loss": 0.0108,
+      "step": 3255
+    },
+    {
+      "epoch": 22.94977973568282,
+      "grad_norm": 0.12299270384899952,
+      "learning_rate": 0.00011714357553545081,
+      "loss": 0.0095,
+      "step": 3256
+    },
+    {
+      "epoch": 22.9568281938326,
+      "grad_norm": 0.13731332604294957,
+      "learning_rate": 0.00011709831075140908,
+      "loss": 0.0103,
+      "step": 3257
+    },
+    {
+      "epoch": 22.96387665198238,
+      "grad_norm": 0.14089378868146513,
+      "learning_rate": 0.00011705304235830198,
+      "loss": 0.0109,
+      "step": 3258
+    },
+    {
+      "epoch": 22.970925110132157,
+      "grad_norm": 0.13719726837757693,
+      "learning_rate": 0.0001170077703656846,
+      "loss": 0.011,
+      "step": 3259
+    },
+    {
+      "epoch": 22.97797356828194,
+      "grad_norm": 0.14920554212130663,
+      "learning_rate": 0.0001169624947831129,
+      "loss": 0.0109,
+      "step": 3260
+    },
+    {
+      "epoch": 22.985022026431718,
+      "grad_norm": 0.13905721380510513,
+      "learning_rate": 0.00011691721562014342,
+      "loss": 0.0116,
+      "step": 3261
+    },
+    {
+      "epoch": 22.992070484581497,
+      "grad_norm": 0.13861888470522635,
+      "learning_rate": 0.00011687193288633365,
+      "loss": 0.011,
+      "step": 3262
+    },
+    {
+      "epoch": 22.99911894273128,
+      "grad_norm": 0.13462388944841694,
+      "learning_rate": 0.00011682664659124174,
+      "loss": 0.0107,
+      "step": 3263
+    },
+    {
+      "epoch": 23.006167400881058,
+      "grad_norm": 0.08728028555381047,
+      "learning_rate": 0.00011678135674442654,
+      "loss": 0.0063,
+      "step": 3264
+    },
+    {
+      "epoch": 23.013215859030836,
+      "grad_norm": 0.09441894359685421,
+      "learning_rate": 0.00011673606335544778,
+      "loss": 0.0062,
+      "step": 3265
+    },
+    {
+      "epoch": 23.020264317180615,
+      "grad_norm": 0.09379155108072945,
+      "learning_rate": 0.00011669076643386583,
+      "loss": 0.006,
+      "step": 3266
+    },
+    {
+      "epoch": 23.027312775330397,
+      "grad_norm": 0.0998758614761668,
+      "learning_rate": 0.00011664546598924184,
+      "loss": 0.0066,
+      "step": 3267
+    },
+    {
+      "epoch": 23.034361233480176,
+      "grad_norm": 0.09988854864289246,
+      "learning_rate": 0.00011660016203113772,
+      "loss": 0.0057,
+      "step": 3268
+    },
+    {
+      "epoch": 23.041409691629955,
+      "grad_norm": 0.1056885551837802,
+      "learning_rate": 0.0001165548545691161,
+      "loss": 0.0065,
+      "step": 3269
+    },
+    {
+      "epoch": 23.048458149779737,
+      "grad_norm": 0.09609480609265346,
+      "learning_rate": 0.00011650954361274033,
+      "loss": 0.006,
+      "step": 3270
+    },
+    {
+      "epoch": 23.055506607929516,
+      "grad_norm": 0.09598063110444033,
+      "learning_rate": 0.00011646422917157462,
+      "loss": 0.0062,
+      "step": 3271
+    },
+    {
+      "epoch": 23.062555066079295,
+      "grad_norm": 0.10444187824162808,
+      "learning_rate": 0.00011641891125518372,
+      "loss": 0.0067,
+      "step": 3272
+    },
+    {
+      "epoch": 23.069603524229073,
+      "grad_norm": 0.09317947455933373,
+      "learning_rate": 0.0001163735898731332,
+      "loss": 0.0057,
+      "step": 3273
+    },
+    {
+      "epoch": 23.076651982378856,
+      "grad_norm": 0.1152119335326372,
+      "learning_rate": 0.00011632826503498951,
+      "loss": 0.0059,
+      "step": 3274
+    },
+    {
+      "epoch": 23.083700440528634,
+      "grad_norm": 0.09799904185350276,
+      "learning_rate": 0.00011628293675031953,
+      "loss": 0.0063,
+      "step": 3275
+    },
+    {
+      "epoch": 23.090748898678413,
+      "grad_norm": 0.09944535820511208,
+      "learning_rate": 0.00011623760502869115,
+      "loss": 0.006,
+      "step": 3276
+    },
+    {
+      "epoch": 23.097797356828195,
+      "grad_norm": 0.10012877734320424,
+      "learning_rate": 0.00011619226987967282,
+      "loss": 0.0062,
+      "step": 3277
+    },
+    {
+      "epoch": 23.104845814977974,
+      "grad_norm": 0.100086180786929,
+      "learning_rate": 0.00011614693131283376,
+      "loss": 0.0063,
+      "step": 3278
+    },
+    {
+      "epoch": 23.111894273127753,
+      "grad_norm": 0.10130625162302448,
+      "learning_rate": 0.00011610158933774394,
+      "loss": 0.0058,
+      "step": 3279
+    },
+    {
+      "epoch": 23.11894273127753,
+      "grad_norm": 0.10918454961514433,
+      "learning_rate": 0.00011605624396397398,
+      "loss": 0.0069,
+      "step": 3280
+    },
+    {
+      "epoch": 23.125991189427314,
+      "grad_norm": 0.10326434243940248,
+      "learning_rate": 0.0001160108952010953,
+      "loss": 0.0058,
+      "step": 3281
+    },
+    {
+      "epoch": 23.133039647577093,
+      "grad_norm": 0.10432691208021426,
+      "learning_rate": 0.00011596554305867995,
+      "loss": 0.0064,
+      "step": 3282
+    },
+    {
+      "epoch": 23.14008810572687,
+      "grad_norm": 0.10387390249316741,
+      "learning_rate": 0.00011592018754630078,
+      "loss": 0.0067,
+      "step": 3283
+    },
+    {
+      "epoch": 23.147136563876654,
+      "grad_norm": 0.10799163966372739,
+      "learning_rate": 0.00011587482867353129,
+      "loss": 0.0064,
+      "step": 3284
+    },
+    {
+      "epoch": 23.154185022026432,
+      "grad_norm": 0.09070653682199865,
+      "learning_rate": 0.00011582946644994574,
+      "loss": 0.0053,
+      "step": 3285
+    },
+    {
+      "epoch": 23.16123348017621,
+      "grad_norm": 0.09593967443668294,
+      "learning_rate": 0.000115784100885119,
+      "loss": 0.006,
+      "step": 3286
+    },
+    {
+      "epoch": 23.16828193832599,
+      "grad_norm": 0.09244923062217304,
+      "learning_rate": 0.00011573873198862677,
+      "loss": 0.0059,
+      "step": 3287
+    },
+    {
+      "epoch": 23.175330396475772,
+      "grad_norm": 0.09743805962501208,
+      "learning_rate": 0.00011569335977004541,
+      "loss": 0.0061,
+      "step": 3288
+    },
+    {
+      "epoch": 23.18237885462555,
+      "grad_norm": 0.09615835718379676,
+      "learning_rate": 0.00011564798423895187,
+      "loss": 0.0057,
+      "step": 3289
+    },
+    {
+      "epoch": 23.18942731277533,
+      "grad_norm": 0.10229208515380087,
+      "learning_rate": 0.00011560260540492398,
+      "loss": 0.0064,
+      "step": 3290
+    },
+    {
+      "epoch": 23.19647577092511,
+      "grad_norm": 0.09956251506585749,
+      "learning_rate": 0.00011555722327754017,
+      "loss": 0.006,
+      "step": 3291
+    },
+    {
+      "epoch": 23.20352422907489,
+      "grad_norm": 0.09923545686960232,
+      "learning_rate": 0.00011551183786637956,
+      "loss": 0.0056,
+      "step": 3292
+    },
+    {
+      "epoch": 23.21057268722467,
+      "grad_norm": 0.1020081774333533,
+      "learning_rate": 0.00011546644918102196,
+      "loss": 0.0059,
+      "step": 3293
+    },
+    {
+      "epoch": 23.217621145374448,
+      "grad_norm": 0.11163529062457896,
+      "learning_rate": 0.00011542105723104795,
+      "loss": 0.0065,
+      "step": 3294
+    },
+    {
+      "epoch": 23.22466960352423,
+      "grad_norm": 0.0957799018424912,
+      "learning_rate": 0.00011537566202603868,
+      "loss": 0.0063,
+      "step": 3295
+    },
+    {
+      "epoch": 23.23171806167401,
+      "grad_norm": 0.09118174183595631,
+      "learning_rate": 0.0001153302635755761,
+      "loss": 0.0057,
+      "step": 3296
+    },
+    {
+      "epoch": 23.238766519823788,
+      "grad_norm": 0.10405959891878593,
+      "learning_rate": 0.00011528486188924277,
+      "loss": 0.0057,
+      "step": 3297
+    },
+    {
+      "epoch": 23.24581497797357,
+      "grad_norm": 0.09656681555077436,
+      "learning_rate": 0.0001152394569766219,
+      "loss": 0.0057,
+      "step": 3298
+    },
+    {
+      "epoch": 23.25286343612335,
+      "grad_norm": 0.11065014517476295,
+      "learning_rate": 0.00011519404884729757,
+      "loss": 0.0061,
+      "step": 3299
+    },
+    {
+      "epoch": 23.259911894273127,
+      "grad_norm": 0.09687745558279186,
+      "learning_rate": 0.00011514863751085425,
+      "loss": 0.0058,
+      "step": 3300
+    },
+    {
+      "epoch": 23.266960352422906,
+      "grad_norm": 0.09767259098950158,
+      "learning_rate": 0.00011510322297687739,
+      "loss": 0.0058,
+      "step": 3301
+    },
+    {
+      "epoch": 23.27400881057269,
+      "grad_norm": 0.09962261443644807,
+      "learning_rate": 0.00011505780525495283,
+      "loss": 0.0055,
+      "step": 3302
+    },
+    {
+      "epoch": 23.281057268722467,
+      "grad_norm": 0.09905982008995727,
+      "learning_rate": 0.00011501238435466732,
+      "loss": 0.0063,
+      "step": 3303
+    },
+    {
+      "epoch": 23.288105726872246,
+      "grad_norm": 0.09760673754180488,
+      "learning_rate": 0.00011496696028560816,
+      "loss": 0.0059,
+      "step": 3304
+    },
+    {
+      "epoch": 23.295154185022028,
+      "grad_norm": 0.08860719417888374,
+      "learning_rate": 0.00011492153305736328,
+      "loss": 0.0059,
+      "step": 3305
+    },
+    {
+      "epoch": 23.302202643171807,
+      "grad_norm": 0.08857981325789181,
+      "learning_rate": 0.00011487610267952142,
+      "loss": 0.0056,
+      "step": 3306
+    },
+    {
+      "epoch": 23.309251101321586,
+      "grad_norm": 0.10418782739809357,
+      "learning_rate": 0.00011483066916167187,
+      "loss": 0.006,
+      "step": 3307
+    },
+    {
+      "epoch": 23.316299559471364,
+      "grad_norm": 0.11550251577938742,
+      "learning_rate": 0.0001147852325134046,
+      "loss": 0.0064,
+      "step": 3308
+    },
+    {
+      "epoch": 23.323348017621147,
+      "grad_norm": 0.10019304364516653,
+      "learning_rate": 0.00011473979274431029,
+      "loss": 0.0062,
+      "step": 3309
+    },
+    {
+      "epoch": 23.330396475770925,
+      "grad_norm": 0.10279895276758022,
+      "learning_rate": 0.0001146943498639802,
+      "loss": 0.0061,
+      "step": 3310
+    },
+    {
+      "epoch": 23.337444933920704,
+      "grad_norm": 0.10416115161588942,
+      "learning_rate": 0.00011464890388200629,
+      "loss": 0.0059,
+      "step": 3311
+    },
+    {
+      "epoch": 23.344493392070486,
+      "grad_norm": 0.0923239988423448,
+      "learning_rate": 0.00011460345480798122,
+      "loss": 0.0055,
+      "step": 3312
+    },
+    {
+      "epoch": 23.351541850220265,
+      "grad_norm": 0.09662780707804812,
+      "learning_rate": 0.00011455800265149825,
+      "loss": 0.0054,
+      "step": 3313
+    },
+    {
+      "epoch": 23.358590308370044,
+      "grad_norm": 0.10168903267968411,
+      "learning_rate": 0.00011451254742215125,
+      "loss": 0.0062,
+      "step": 3314
+    },
+    {
+      "epoch": 23.365638766519822,
+      "grad_norm": 0.09599626830436046,
+      "learning_rate": 0.00011446708912953485,
+      "loss": 0.0058,
+      "step": 3315
+    },
+    {
+      "epoch": 23.372687224669605,
+      "grad_norm": 0.1054382419965715,
+      "learning_rate": 0.0001144216277832442,
+      "loss": 0.0062,
+      "step": 3316
+    },
+    {
+      "epoch": 23.379735682819383,
+      "grad_norm": 0.1054825086121288,
+      "learning_rate": 0.00011437616339287523,
+      "loss": 0.0066,
+      "step": 3317
+    },
+    {
+      "epoch": 23.386784140969162,
+      "grad_norm": 0.09689235855980902,
+      "learning_rate": 0.00011433069596802439,
+      "loss": 0.0063,
+      "step": 3318
+    },
+    {
+      "epoch": 23.393832599118944,
+      "grad_norm": 0.09139538471635422,
+      "learning_rate": 0.00011428522551828884,
+      "loss": 0.0057,
+      "step": 3319
+    },
+    {
+      "epoch": 23.400881057268723,
+      "grad_norm": 0.11003799444989602,
+      "learning_rate": 0.00011423975205326636,
+      "loss": 0.0066,
+      "step": 3320
+    },
+    {
+      "epoch": 23.407929515418502,
+      "grad_norm": 0.10074140846167608,
+      "learning_rate": 0.00011419427558255535,
+      "loss": 0.006,
+      "step": 3321
+    },
+    {
+      "epoch": 23.41497797356828,
+      "grad_norm": 0.10013882736112736,
+      "learning_rate": 0.00011414879611575488,
+      "loss": 0.0059,
+      "step": 3322
+    },
+    {
+      "epoch": 23.422026431718063,
+      "grad_norm": 0.10293560567315413,
+      "learning_rate": 0.00011410331366246463,
+      "loss": 0.0066,
+      "step": 3323
+    },
+    {
+      "epoch": 23.42907488986784,
+      "grad_norm": 0.11401383241535502,
+      "learning_rate": 0.00011405782823228493,
+      "loss": 0.0062,
+      "step": 3324
+    },
+    {
+      "epoch": 23.43612334801762,
+      "grad_norm": 0.09884812454502226,
+      "learning_rate": 0.00011401233983481665,
+      "loss": 0.0058,
+      "step": 3325
+    },
+    {
+      "epoch": 23.443171806167403,
+      "grad_norm": 0.09591704949741377,
+      "learning_rate": 0.00011396684847966148,
+      "loss": 0.006,
+      "step": 3326
+    },
+    {
+      "epoch": 23.45022026431718,
+      "grad_norm": 0.10868449885450858,
+      "learning_rate": 0.00011392135417642146,
+      "loss": 0.0064,
+      "step": 3327
+    },
+    {
+      "epoch": 23.45726872246696,
+      "grad_norm": 0.11160576982577972,
+      "learning_rate": 0.00011387585693469955,
+      "loss": 0.006,
+      "step": 3328
+    },
+    {
+      "epoch": 23.46431718061674,
+      "grad_norm": 0.10664763485140845,
+      "learning_rate": 0.00011383035676409914,
+      "loss": 0.0067,
+      "step": 3329
+    },
+    {
+      "epoch": 23.47136563876652,
+      "grad_norm": 0.1132997754959414,
+      "learning_rate": 0.00011378485367422423,
+      "loss": 0.0064,
+      "step": 3330
+    },
+    {
+      "epoch": 23.4784140969163,
+      "grad_norm": 0.10912433963366099,
+      "learning_rate": 0.00011373934767467954,
+      "loss": 0.0068,
+      "step": 3331
+    },
+    {
+      "epoch": 23.48546255506608,
+      "grad_norm": 0.10405784539631865,
+      "learning_rate": 0.00011369383877507034,
+      "loss": 0.0065,
+      "step": 3332
+    },
+    {
+      "epoch": 23.49251101321586,
+      "grad_norm": 0.10912283193678492,
+      "learning_rate": 0.00011364832698500254,
+      "loss": 0.0067,
+      "step": 3333
+    },
+    {
+      "epoch": 23.49955947136564,
+      "grad_norm": 0.10446690441509345,
+      "learning_rate": 0.00011360281231408263,
+      "loss": 0.0068,
+      "step": 3334
+    },
+    {
+      "epoch": 23.506607929515418,
+      "grad_norm": 0.10707853623543984,
+      "learning_rate": 0.00011355729477191773,
+      "loss": 0.0061,
+      "step": 3335
+    },
+    {
+      "epoch": 23.513656387665197,
+      "grad_norm": 0.10243502853710429,
+      "learning_rate": 0.00011351177436811553,
+      "loss": 0.0064,
+      "step": 3336
+    },
+    {
+      "epoch": 23.52070484581498,
+      "grad_norm": 0.12108794482584985,
+      "learning_rate": 0.00011346625111228442,
+      "loss": 0.0063,
+      "step": 3337
+    },
+    {
+      "epoch": 23.527753303964758,
+      "grad_norm": 0.1029932075583387,
+      "learning_rate": 0.00011342072501403327,
+      "loss": 0.0067,
+      "step": 3338
+    },
+    {
+      "epoch": 23.534801762114537,
+      "grad_norm": 0.11069921763372935,
+      "learning_rate": 0.0001133751960829716,
+      "loss": 0.0065,
+      "step": 3339
+    },
+    {
+      "epoch": 23.541850220264315,
+      "grad_norm": 0.0993146427831075,
+      "learning_rate": 0.0001133296643287096,
+      "loss": 0.0065,
+      "step": 3340
+    },
+    {
+      "epoch": 23.548898678414098,
+      "grad_norm": 0.10540811005705077,
+      "learning_rate": 0.00011328412976085791,
+      "loss": 0.0064,
+      "step": 3341
+    },
+    {
+      "epoch": 23.555947136563876,
+      "grad_norm": 0.10331883385576791,
+      "learning_rate": 0.00011323859238902792,
+      "loss": 0.006,
+      "step": 3342
+    },
+    {
+      "epoch": 23.562995594713655,
+      "grad_norm": 0.0958772406947929,
+      "learning_rate": 0.00011319305222283147,
+      "loss": 0.0056,
+      "step": 3343
+    },
+    {
+      "epoch": 23.570044052863437,
+      "grad_norm": 0.11344359559025792,
+      "learning_rate": 0.00011314750927188111,
+      "loss": 0.0065,
+      "step": 3344
+    },
+    {
+      "epoch": 23.577092511013216,
+      "grad_norm": 0.10170331223166182,
+      "learning_rate": 0.00011310196354578992,
+      "loss": 0.0063,
+      "step": 3345
+    },
+    {
+      "epoch": 23.584140969162995,
+      "grad_norm": 0.0988580445765349,
+      "learning_rate": 0.00011305641505417149,
+      "loss": 0.0062,
+      "step": 3346
+    },
+    {
+      "epoch": 23.591189427312777,
+      "grad_norm": 0.10563181528326043,
+      "learning_rate": 0.00011301086380664017,
+      "loss": 0.0065,
+      "step": 3347
+    },
+    {
+      "epoch": 23.598237885462556,
+      "grad_norm": 0.10181923680379269,
+      "learning_rate": 0.00011296530981281073,
+      "loss": 0.006,
+      "step": 3348
+    },
+    {
+      "epoch": 23.605286343612335,
+      "grad_norm": 0.09545523858463772,
+      "learning_rate": 0.00011291975308229862,
+      "loss": 0.0068,
+      "step": 3349
+    },
+    {
+      "epoch": 23.612334801762113,
+      "grad_norm": 0.11140141252965387,
+      "learning_rate": 0.00011287419362471983,
+      "loss": 0.0066,
+      "step": 3350
+    },
+    {
+      "epoch": 23.619383259911896,
+      "grad_norm": 0.09862089017562133,
+      "learning_rate": 0.00011282863144969091,
+      "loss": 0.0064,
+      "step": 3351
+    },
+    {
+      "epoch": 23.626431718061674,
+      "grad_norm": 0.10818381980024633,
+      "learning_rate": 0.00011278306656682897,
+      "loss": 0.0063,
+      "step": 3352
+    },
+    {
+      "epoch": 23.633480176211453,
+      "grad_norm": 0.09946571210325696,
+      "learning_rate": 0.00011273749898575184,
+      "loss": 0.0062,
+      "step": 3353
+    },
+    {
+      "epoch": 23.64052863436123,
+      "grad_norm": 0.11367910807203581,
+      "learning_rate": 0.0001126919287160777,
+      "loss": 0.0068,
+      "step": 3354
+    },
+    {
+      "epoch": 23.647577092511014,
+      "grad_norm": 0.10788123553656737,
+      "learning_rate": 0.0001126463557674254,
+      "loss": 0.0063,
+      "step": 3355
+    },
+    {
+      "epoch": 23.654625550660793,
+      "grad_norm": 0.10681293638872207,
+      "learning_rate": 0.0001126007801494144,
+      "loss": 0.0064,
+      "step": 3356
+    },
+    {
+      "epoch": 23.66167400881057,
+      "grad_norm": 0.09776996298626715,
+      "learning_rate": 0.00011255520187166464,
+      "loss": 0.0062,
+      "step": 3357
+    },
+    {
+      "epoch": 23.668722466960354,
+      "grad_norm": 0.11408490729506132,
+      "learning_rate": 0.0001125096209437967,
+      "loss": 0.0065,
+      "step": 3358
+    },
+    {
+      "epoch": 23.675770925110132,
+      "grad_norm": 0.10569954039673327,
+      "learning_rate": 0.00011246403737543163,
+      "loss": 0.0068,
+      "step": 3359
+    },
+    {
+      "epoch": 23.68281938325991,
+      "grad_norm": 0.1074901868636517,
+      "learning_rate": 0.00011241845117619114,
+      "loss": 0.0061,
+      "step": 3360
+    },
+    {
+      "epoch": 23.68986784140969,
+      "grad_norm": 0.10876502702312608,
+      "learning_rate": 0.00011237286235569742,
+      "loss": 0.0071,
+      "step": 3361
+    },
+    {
+      "epoch": 23.696916299559472,
+      "grad_norm": 0.11274403984371172,
+      "learning_rate": 0.00011232727092357322,
+      "loss": 0.007,
+      "step": 3362
+    },
+    {
+      "epoch": 23.70396475770925,
+      "grad_norm": 0.0968771488163456,
+      "learning_rate": 0.0001122816768894419,
+      "loss": 0.0065,
+      "step": 3363
+    },
+    {
+      "epoch": 23.71101321585903,
+      "grad_norm": 0.10000394909235526,
+      "learning_rate": 0.00011223608026292725,
+      "loss": 0.006,
+      "step": 3364
+    },
+    {
+      "epoch": 23.718061674008812,
+      "grad_norm": 0.10065502661946764,
+      "learning_rate": 0.00011219048105365375,
+      "loss": 0.0063,
+      "step": 3365
+    },
+    {
+      "epoch": 23.72511013215859,
+      "grad_norm": 0.1122156199160409,
+      "learning_rate": 0.00011214487927124632,
+      "loss": 0.0071,
+      "step": 3366
+    },
+    {
+      "epoch": 23.73215859030837,
+      "grad_norm": 0.10853858645099727,
+      "learning_rate": 0.00011209927492533052,
+      "loss": 0.0064,
+      "step": 3367
+    },
+    {
+      "epoch": 23.739207048458148,
+      "grad_norm": 0.09425726372757337,
+      "learning_rate": 0.0001120536680255323,
+      "loss": 0.0063,
+      "step": 3368
+    },
+    {
+      "epoch": 23.74625550660793,
+      "grad_norm": 0.12308317146229611,
+      "learning_rate": 0.00011200805858147833,
+      "loss": 0.0068,
+      "step": 3369
+    },
+    {
+      "epoch": 23.75330396475771,
+      "grad_norm": 0.11427909571857857,
+      "learning_rate": 0.00011196244660279572,
+      "loss": 0.0071,
+      "step": 3370
+    },
+    {
+      "epoch": 23.760352422907488,
+      "grad_norm": 0.11000944354230346,
+      "learning_rate": 0.00011191683209911202,
+      "loss": 0.0085,
+      "step": 3371
+    },
+    {
+      "epoch": 23.76740088105727,
+      "grad_norm": 0.09913236442405626,
+      "learning_rate": 0.00011187121508005552,
+      "loss": 0.0063,
+      "step": 3372
+    },
+    {
+      "epoch": 23.77444933920705,
+      "grad_norm": 0.10843438391482203,
+      "learning_rate": 0.00011182559555525491,
+      "loss": 0.0073,
+      "step": 3373
+    },
+    {
+      "epoch": 23.781497797356828,
+      "grad_norm": 0.1199299800866117,
+      "learning_rate": 0.00011177997353433946,
+      "loss": 0.0073,
+      "step": 3374
+    },
+    {
+      "epoch": 23.788546255506606,
+      "grad_norm": 0.1125288655613348,
+      "learning_rate": 0.00011173434902693889,
+      "loss": 0.0065,
+      "step": 3375
+    },
+    {
+      "epoch": 23.79559471365639,
+      "grad_norm": 0.10873175199397546,
+      "learning_rate": 0.00011168872204268354,
+      "loss": 0.0072,
+      "step": 3376
+    },
+    {
+      "epoch": 23.802643171806167,
+      "grad_norm": 0.1136570838787563,
+      "learning_rate": 0.00011164309259120422,
+      "loss": 0.0074,
+      "step": 3377
+    },
+    {
+      "epoch": 23.809691629955946,
+      "grad_norm": 0.10358453092253164,
+      "learning_rate": 0.00011159746068213226,
+      "loss": 0.0063,
+      "step": 3378
+    },
+    {
+      "epoch": 23.81674008810573,
+      "grad_norm": 0.11452256332613886,
+      "learning_rate": 0.00011155182632509955,
+      "loss": 0.0067,
+      "step": 3379
+    },
+    {
+      "epoch": 23.823788546255507,
+      "grad_norm": 0.11174380079296857,
+      "learning_rate": 0.00011150618952973842,
+      "loss": 0.0069,
+      "step": 3380
+    },
+    {
+      "epoch": 23.830837004405286,
+      "grad_norm": 0.11110465341249098,
+      "learning_rate": 0.00011146055030568181,
+      "loss": 0.0066,
+      "step": 3381
+    },
+    {
+      "epoch": 23.837885462555064,
+      "grad_norm": 0.10245453258045702,
+      "learning_rate": 0.00011141490866256308,
+      "loss": 0.007,
+      "step": 3382
+    },
+    {
+      "epoch": 23.844933920704847,
+      "grad_norm": 0.12183781647423832,
+      "learning_rate": 0.00011136926461001617,
+      "loss": 0.0072,
+      "step": 3383
+    },
+    {
+      "epoch": 23.851982378854625,
+      "grad_norm": 0.11813262058914442,
+      "learning_rate": 0.00011132361815767554,
+      "loss": 0.0065,
+      "step": 3384
+    },
+    {
+      "epoch": 23.859030837004404,
+      "grad_norm": 0.10844687800252627,
+      "learning_rate": 0.00011127796931517607,
+      "loss": 0.0071,
+      "step": 3385
+    },
+    {
+      "epoch": 23.866079295154186,
+      "grad_norm": 0.12836559046550894,
+      "learning_rate": 0.0001112323180921532,
+      "loss": 0.0076,
+      "step": 3386
+    },
+    {
+      "epoch": 23.873127753303965,
+      "grad_norm": 0.10899902601711874,
+      "learning_rate": 0.00011118666449824286,
+      "loss": 0.0068,
+      "step": 3387
+    },
+    {
+      "epoch": 23.880176211453744,
+      "grad_norm": 0.11309814955828087,
+      "learning_rate": 0.00011114100854308151,
+      "loss": 0.0069,
+      "step": 3388
+    },
+    {
+      "epoch": 23.887224669603523,
+      "grad_norm": 0.10691192077881036,
+      "learning_rate": 0.00011109535023630608,
+      "loss": 0.007,
+      "step": 3389
+    },
+    {
+      "epoch": 23.894273127753305,
+      "grad_norm": 0.10418063512512943,
+      "learning_rate": 0.00011104968958755402,
+      "loss": 0.0067,
+      "step": 3390
+    },
+    {
+      "epoch": 23.901321585903084,
+      "grad_norm": 0.120622366286196,
+      "learning_rate": 0.0001110040266064632,
+      "loss": 0.0075,
+      "step": 3391
+    },
+    {
+      "epoch": 23.908370044052862,
+      "grad_norm": 0.11390563446446086,
+      "learning_rate": 0.00011095836130267212,
+      "loss": 0.0073,
+      "step": 3392
+    },
+    {
+      "epoch": 23.915418502202645,
+      "grad_norm": 0.11763620335658719,
+      "learning_rate": 0.0001109126936858196,
+      "loss": 0.007,
+      "step": 3393
+    },
+    {
+      "epoch": 23.922466960352423,
+      "grad_norm": 0.11766351993452533,
+      "learning_rate": 0.00011086702376554514,
+      "loss": 0.0076,
+      "step": 3394
+    },
+    {
+      "epoch": 23.929515418502202,
+      "grad_norm": 0.11917514496471916,
+      "learning_rate": 0.00011082135155148854,
+      "loss": 0.0076,
+      "step": 3395
+    },
+    {
+      "epoch": 23.93656387665198,
+      "grad_norm": 0.10941884627419622,
+      "learning_rate": 0.00011077567705329018,
+      "loss": 0.0071,
+      "step": 3396
+    },
+    {
+      "epoch": 23.943612334801763,
+      "grad_norm": 0.1146824464673697,
+      "learning_rate": 0.00011073000028059094,
+      "loss": 0.0074,
+      "step": 3397
+    },
+    {
+      "epoch": 23.950660792951542,
+      "grad_norm": 0.1155946881248893,
+      "learning_rate": 0.00011068432124303212,
+      "loss": 0.0071,
+      "step": 3398
+    },
+    {
+      "epoch": 23.95770925110132,
+      "grad_norm": 0.10885314291100824,
+      "learning_rate": 0.00011063863995025557,
+      "loss": 0.0071,
+      "step": 3399
+    },
+    {
+      "epoch": 23.964757709251103,
+      "grad_norm": 0.10673196002349694,
+      "learning_rate": 0.00011059295641190352,
+      "loss": 0.0072,
+      "step": 3400
+    },
+    {
+      "epoch": 23.97180616740088,
+      "grad_norm": 0.1276022159246556,
+      "learning_rate": 0.00011054727063761879,
+      "loss": 0.0076,
+      "step": 3401
+    },
+    {
+      "epoch": 23.97885462555066,
+      "grad_norm": 0.12931654929452616,
+      "learning_rate": 0.00011050158263704456,
+      "loss": 0.0084,
+      "step": 3402
+    },
+    {
+      "epoch": 23.98590308370044,
+      "grad_norm": 0.1142476265659862,
+      "learning_rate": 0.00011045589241982455,
+      "loss": 0.0078,
+      "step": 3403
+    },
+    {
+      "epoch": 23.99295154185022,
+      "grad_norm": 0.11587530846735417,
+      "learning_rate": 0.00011041019999560292,
+      "loss": 0.0082,
+      "step": 3404
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 0.10991261794715332,
+      "learning_rate": 0.00011036450537402432,
+      "loss": 0.0072,
+      "step": 3405
+    },
+    {
+      "epoch": 24.00704845814978,
+      "grad_norm": 0.07632487050049357,
+      "learning_rate": 0.00011031880856473383,
+      "loss": 0.0041,
+      "step": 3406
+    },
+    {
+      "epoch": 24.01409691629956,
+      "grad_norm": 0.0730999752615215,
+      "learning_rate": 0.00011027310957737698,
+      "loss": 0.0042,
+      "step": 3407
+    },
+    {
+      "epoch": 24.02114537444934,
+      "grad_norm": 0.09045417160082123,
+      "learning_rate": 0.00011022740842159989,
+      "loss": 0.0043,
+      "step": 3408
+    },
+    {
+      "epoch": 24.02819383259912,
+      "grad_norm": 0.09938877765732707,
+      "learning_rate": 0.0001101817051070489,
+      "loss": 0.0049,
+      "step": 3409
+    },
+    {
+      "epoch": 24.035242290748897,
+      "grad_norm": 0.09558115710854505,
+      "learning_rate": 0.00011013599964337107,
+      "loss": 0.0055,
+      "step": 3410
+    },
+    {
+      "epoch": 24.04229074889868,
+      "grad_norm": 0.0967278556752667,
+      "learning_rate": 0.00011009029204021373,
+      "loss": 0.0055,
+      "step": 3411
+    },
+    {
+      "epoch": 24.049339207048458,
+      "grad_norm": 0.09051909138848212,
+      "learning_rate": 0.00011004458230722468,
+      "loss": 0.0051,
+      "step": 3412
+    },
+    {
+      "epoch": 24.056387665198237,
+      "grad_norm": 0.07419797018200967,
+      "learning_rate": 0.00010999887045405226,
+      "loss": 0.004,
+      "step": 3413
+    },
+    {
+      "epoch": 24.06343612334802,
+      "grad_norm": 0.07885545034138144,
+      "learning_rate": 0.00010995315649034521,
+      "loss": 0.0043,
+      "step": 3414
+    },
+    {
+      "epoch": 24.070484581497798,
+      "grad_norm": 0.08535589501495561,
+      "learning_rate": 0.00010990744042575269,
+      "loss": 0.0047,
+      "step": 3415
+    },
+    {
+      "epoch": 24.077533039647577,
+      "grad_norm": 0.07418412832413916,
+      "learning_rate": 0.0001098617222699243,
+      "loss": 0.004,
+      "step": 3416
+    },
+    {
+      "epoch": 24.084581497797355,
+      "grad_norm": 0.08725830204801598,
+      "learning_rate": 0.00010981600203251017,
+      "loss": 0.0046,
+      "step": 3417
+    },
+    {
+      "epoch": 24.091629955947138,
+      "grad_norm": 0.09072326476094664,
+      "learning_rate": 0.00010977027972316072,
+      "loss": 0.0045,
+      "step": 3418
+    },
+    {
+      "epoch": 24.098678414096916,
+      "grad_norm": 0.09987404665520483,
+      "learning_rate": 0.00010972455535152697,
+      "loss": 0.005,
+      "step": 3419
+    },
+    {
+      "epoch": 24.105726872246695,
+      "grad_norm": 0.08806530067309719,
+      "learning_rate": 0.00010967882892726025,
+      "loss": 0.0042,
+      "step": 3420
+    },
+    {
+      "epoch": 24.112775330396477,
+      "grad_norm": 0.08648006084965157,
+      "learning_rate": 0.00010963310046001236,
+      "loss": 0.0046,
+      "step": 3421
+    },
+    {
+      "epoch": 24.119823788546256,
+      "grad_norm": 0.09219945642617233,
+      "learning_rate": 0.00010958736995943561,
+      "loss": 0.0048,
+      "step": 3422
+    },
+    {
+      "epoch": 24.126872246696035,
+      "grad_norm": 0.0809616176609884,
+      "learning_rate": 0.0001095416374351826,
+      "loss": 0.0046,
+      "step": 3423
+    },
+    {
+      "epoch": 24.133920704845814,
+      "grad_norm": 0.08491197676725437,
+      "learning_rate": 0.00010949590289690646,
+      "loss": 0.0043,
+      "step": 3424
+    },
+    {
+      "epoch": 24.140969162995596,
+      "grad_norm": 0.08641129835457595,
+      "learning_rate": 0.0001094501663542607,
+      "loss": 0.0043,
+      "step": 3425
+    },
+    {
+      "epoch": 24.148017621145375,
+      "grad_norm": 0.08329904669501577,
+      "learning_rate": 0.0001094044278168993,
+      "loss": 0.0044,
+      "step": 3426
+    },
+    {
+      "epoch": 24.155066079295153,
+      "grad_norm": 0.08550526163021378,
+      "learning_rate": 0.00010935868729447659,
+      "loss": 0.0047,
+      "step": 3427
+    },
+    {
+      "epoch": 24.162114537444936,
+      "grad_norm": 0.08503140876429002,
+      "learning_rate": 0.00010931294479664736,
+      "loss": 0.0046,
+      "step": 3428
+    },
+    {
+      "epoch": 24.169162995594714,
+      "grad_norm": 0.08017266838900862,
+      "learning_rate": 0.00010926720033306684,
+      "loss": 0.0041,
+      "step": 3429
+    },
+    {
+      "epoch": 24.176211453744493,
+      "grad_norm": 0.08381388203691932,
+      "learning_rate": 0.0001092214539133906,
+      "loss": 0.0047,
+      "step": 3430
+    },
+    {
+      "epoch": 24.18325991189427,
+      "grad_norm": 0.09077354899502474,
+      "learning_rate": 0.00010917570554727472,
+      "loss": 0.0047,
+      "step": 3431
+    },
+    {
+      "epoch": 24.190308370044054,
+      "grad_norm": 0.08552954988621134,
+      "learning_rate": 0.00010912995524437559,
+      "loss": 0.0052,
+      "step": 3432
+    },
+    {
+      "epoch": 24.197356828193833,
+      "grad_norm": 0.0843579402860924,
+      "learning_rate": 0.00010908420301435016,
+      "loss": 0.0044,
+      "step": 3433
+    },
+    {
+      "epoch": 24.20440528634361,
+      "grad_norm": 0.0903355444897464,
+      "learning_rate": 0.00010903844886685555,
+      "loss": 0.0045,
+      "step": 3434
+    },
+    {
+      "epoch": 24.211453744493394,
+      "grad_norm": 0.09282323258453713,
+      "learning_rate": 0.00010899269281154953,
+      "loss": 0.0047,
+      "step": 3435
+    },
+    {
+      "epoch": 24.218502202643172,
+      "grad_norm": 0.09063086888921768,
+      "learning_rate": 0.00010894693485809016,
+      "loss": 0.0047,
+      "step": 3436
+    },
+    {
+      "epoch": 24.22555066079295,
+      "grad_norm": 0.09150896731925418,
+      "learning_rate": 0.00010890117501613581,
+      "loss": 0.0046,
+      "step": 3437
+    },
+    {
+      "epoch": 24.23259911894273,
+      "grad_norm": 0.08391088044001513,
+      "learning_rate": 0.00010885541329534543,
+      "loss": 0.0041,
+      "step": 3438
+    },
+    {
+      "epoch": 24.239647577092512,
+      "grad_norm": 0.07956129862326256,
+      "learning_rate": 0.00010880964970537826,
+      "loss": 0.0045,
+      "step": 3439
+    },
+    {
+      "epoch": 24.24669603524229,
+      "grad_norm": 0.08164147583722119,
+      "learning_rate": 0.00010876388425589397,
+      "loss": 0.0043,
+      "step": 3440
+    },
+    {
+      "epoch": 24.25374449339207,
+      "grad_norm": 0.08743372608660288,
+      "learning_rate": 0.00010871811695655257,
+      "loss": 0.0045,
+      "step": 3441
+    },
+    {
+      "epoch": 24.260792951541852,
+      "grad_norm": 0.08199563240602066,
+      "learning_rate": 0.00010867234781701453,
+      "loss": 0.0044,
+      "step": 3442
+    },
+    {
+      "epoch": 24.26784140969163,
+      "grad_norm": 0.08018337036207361,
+      "learning_rate": 0.00010862657684694067,
+      "loss": 0.0043,
+      "step": 3443
+    },
+    {
+      "epoch": 24.27488986784141,
+      "grad_norm": 0.07322540080937147,
+      "learning_rate": 0.0001085808040559922,
+      "loss": 0.0039,
+      "step": 3444
+    },
+    {
+      "epoch": 24.281938325991188,
+      "grad_norm": 0.09634061574341495,
+      "learning_rate": 0.00010853502945383072,
+      "loss": 0.0046,
+      "step": 3445
+    },
+    {
+      "epoch": 24.28898678414097,
+      "grad_norm": 0.08145562543254964,
+      "learning_rate": 0.00010848925305011818,
+      "loss": 0.0043,
+      "step": 3446
+    },
+    {
+      "epoch": 24.29603524229075,
+      "grad_norm": 0.08041807735119501,
+      "learning_rate": 0.000108443474854517,
+      "loss": 0.0044,
+      "step": 3447
+    },
+    {
+      "epoch": 24.303083700440528,
+      "grad_norm": 0.07019684999162604,
+      "learning_rate": 0.00010839769487668985,
+      "loss": 0.0037,
+      "step": 3448
+    },
+    {
+      "epoch": 24.31013215859031,
+      "grad_norm": 0.09149432266588386,
+      "learning_rate": 0.00010835191312629993,
+      "loss": 0.0044,
+      "step": 3449
+    },
+    {
+      "epoch": 24.31718061674009,
+      "grad_norm": 0.07669721485430891,
+      "learning_rate": 0.00010830612961301063,
+      "loss": 0.0039,
+      "step": 3450
+    },
+    {
+      "epoch": 24.324229074889868,
+      "grad_norm": 0.08594258997361302,
+      "learning_rate": 0.00010826034434648587,
+      "loss": 0.0043,
+      "step": 3451
+    },
+    {
+      "epoch": 24.331277533039646,
+      "grad_norm": 0.08673780227500641,
+      "learning_rate": 0.0001082145573363899,
+      "loss": 0.0045,
+      "step": 3452
+    },
+    {
+      "epoch": 24.33832599118943,
+      "grad_norm": 0.08742121185664041,
+      "learning_rate": 0.00010816876859238724,
+      "loss": 0.0044,
+      "step": 3453
+    },
+    {
+      "epoch": 24.345374449339207,
+      "grad_norm": 0.07832148564098501,
+      "learning_rate": 0.00010812297812414294,
+      "loss": 0.0041,
+      "step": 3454
+    },
+    {
+      "epoch": 24.352422907488986,
+      "grad_norm": 0.08687897133050855,
+      "learning_rate": 0.00010807718594132228,
+      "loss": 0.0045,
+      "step": 3455
+    },
+    {
+      "epoch": 24.35947136563877,
+      "grad_norm": 0.08345667826395815,
+      "learning_rate": 0.00010803139205359096,
+      "loss": 0.0043,
+      "step": 3456
+    },
+    {
+      "epoch": 24.366519823788547,
+      "grad_norm": 0.09978420741124983,
+      "learning_rate": 0.00010798559647061506,
+      "loss": 0.0054,
+      "step": 3457
+    },
+    {
+      "epoch": 24.373568281938326,
+      "grad_norm": 0.08153704147955364,
+      "learning_rate": 0.00010793979920206095,
+      "loss": 0.0046,
+      "step": 3458
+    },
+    {
+      "epoch": 24.380616740088104,
+      "grad_norm": 0.09312307967443205,
+      "learning_rate": 0.00010789400025759537,
+      "loss": 0.005,
+      "step": 3459
+    },
+    {
+      "epoch": 24.387665198237887,
+      "grad_norm": 0.08949684393413604,
+      "learning_rate": 0.0001078481996468855,
+      "loss": 0.0044,
+      "step": 3460
+    },
+    {
+      "epoch": 24.394713656387665,
+      "grad_norm": 0.0908851082189725,
+      "learning_rate": 0.00010780239737959879,
+      "loss": 0.0049,
+      "step": 3461
+    },
+    {
+      "epoch": 24.401762114537444,
+      "grad_norm": 0.09247210489056217,
+      "learning_rate": 0.00010775659346540303,
+      "loss": 0.0049,
+      "step": 3462
+    },
+    {
+      "epoch": 24.408810572687223,
+      "grad_norm": 0.07814399818541781,
+      "learning_rate": 0.00010771078791396645,
+      "loss": 0.0043,
+      "step": 3463
+    },
+    {
+      "epoch": 24.415859030837005,
+      "grad_norm": 0.08938210626503584,
+      "learning_rate": 0.00010766498073495748,
+      "loss": 0.0046,
+      "step": 3464
+    },
+    {
+      "epoch": 24.422907488986784,
+      "grad_norm": 0.0819977994311781,
+      "learning_rate": 0.00010761917193804503,
+      "loss": 0.0045,
+      "step": 3465
+    },
+    {
+      "epoch": 24.429955947136563,
+      "grad_norm": 0.09670961149555345,
+      "learning_rate": 0.00010757336153289827,
+      "loss": 0.0047,
+      "step": 3466
+    },
+    {
+      "epoch": 24.437004405286345,
+      "grad_norm": 0.08824842119211172,
+      "learning_rate": 0.00010752754952918678,
+      "loss": 0.0052,
+      "step": 3467
+    },
+    {
+      "epoch": 24.444052863436124,
+      "grad_norm": 0.09166316610771631,
+      "learning_rate": 0.0001074817359365804,
+      "loss": 0.0047,
+      "step": 3468
+    },
+    {
+      "epoch": 24.451101321585902,
+      "grad_norm": 0.08028096841914464,
+      "learning_rate": 0.00010743592076474935,
+      "loss": 0.0045,
+      "step": 3469
+    },
+    {
+      "epoch": 24.458149779735685,
+      "grad_norm": 0.07759122516296652,
+      "learning_rate": 0.00010739010402336415,
+      "loss": 0.0041,
+      "step": 3470
+    },
+    {
+      "epoch": 24.465198237885463,
+      "grad_norm": 0.0869873738605092,
+      "learning_rate": 0.00010734428572209572,
+      "loss": 0.0046,
+      "step": 3471
+    },
+    {
+      "epoch": 24.472246696035242,
+      "grad_norm": 0.08607636703241699,
+      "learning_rate": 0.00010729846587061523,
+      "loss": 0.0045,
+      "step": 3472
+    },
+    {
+      "epoch": 24.47929515418502,
+      "grad_norm": 0.08801020701914312,
+      "learning_rate": 0.00010725264447859417,
+      "loss": 0.0048,
+      "step": 3473
+    },
+    {
+      "epoch": 24.486343612334803,
+      "grad_norm": 0.09192213510932519,
+      "learning_rate": 0.00010720682155570452,
+      "loss": 0.0043,
+      "step": 3474
+    },
+    {
+      "epoch": 24.493392070484582,
+      "grad_norm": 0.08925575031332991,
+      "learning_rate": 0.00010716099711161833,
+      "loss": 0.0049,
+      "step": 3475
+    },
+    {
+      "epoch": 24.50044052863436,
+      "grad_norm": 0.09452781589674057,
+      "learning_rate": 0.00010711517115600817,
+      "loss": 0.0047,
+      "step": 3476
+    },
+    {
+      "epoch": 24.50748898678414,
+      "grad_norm": 0.09652569806796353,
+      "learning_rate": 0.00010706934369854686,
+      "loss": 0.0048,
+      "step": 3477
+    },
+    {
+      "epoch": 24.51453744493392,
+      "grad_norm": 0.09165982910192355,
+      "learning_rate": 0.0001070235147489075,
+      "loss": 0.0047,
+      "step": 3478
+    },
+    {
+      "epoch": 24.5215859030837,
+      "grad_norm": 0.08539563930697,
+      "learning_rate": 0.00010697768431676356,
+      "loss": 0.0049,
+      "step": 3479
+    },
+    {
+      "epoch": 24.52863436123348,
+      "grad_norm": 0.09968261391183567,
+      "learning_rate": 0.0001069318524117888,
+      "loss": 0.005,
+      "step": 3480
+    },
+    {
+      "epoch": 24.53568281938326,
+      "grad_norm": 0.09074944361352764,
+      "learning_rate": 0.00010688601904365731,
+      "loss": 0.0051,
+      "step": 3481
+    },
+    {
+      "epoch": 24.54273127753304,
+      "grad_norm": 0.08522138380458691,
+      "learning_rate": 0.00010684018422204348,
+      "loss": 0.0043,
+      "step": 3482
+    },
+    {
+      "epoch": 24.54977973568282,
+      "grad_norm": 0.09586799220099408,
+      "learning_rate": 0.00010679434795662198,
+      "loss": 0.0049,
+      "step": 3483
+    },
+    {
+      "epoch": 24.556828193832597,
+      "grad_norm": 0.08798102216447769,
+      "learning_rate": 0.00010674851025706777,
+      "loss": 0.0048,
+      "step": 3484
+    },
+    {
+      "epoch": 24.56387665198238,
+      "grad_norm": 0.092263469922929,
+      "learning_rate": 0.00010670267113305623,
+      "loss": 0.0048,
+      "step": 3485
+    },
+    {
+      "epoch": 24.57092511013216,
+      "grad_norm": 0.08544348168019582,
+      "learning_rate": 0.0001066568305942629,
+      "loss": 0.0043,
+      "step": 3486
+    },
+    {
+      "epoch": 24.577973568281937,
+      "grad_norm": 0.09348708254638215,
+      "learning_rate": 0.00010661098865036368,
+      "loss": 0.0047,
+      "step": 3487
+    },
+    {
+      "epoch": 24.58502202643172,
+      "grad_norm": 0.0935206477235683,
+      "learning_rate": 0.00010656514531103483,
+      "loss": 0.0045,
+      "step": 3488
+    },
+    {
+      "epoch": 24.592070484581498,
+      "grad_norm": 0.0884741969348127,
+      "learning_rate": 0.00010651930058595272,
+      "loss": 0.0045,
+      "step": 3489
+    },
+    {
+      "epoch": 24.599118942731277,
+      "grad_norm": 0.08403608559447667,
+      "learning_rate": 0.00010647345448479426,
+      "loss": 0.0042,
+      "step": 3490
+    },
+    {
+      "epoch": 24.606167400881056,
+      "grad_norm": 0.07779147124348769,
+      "learning_rate": 0.00010642760701723642,
+      "loss": 0.0041,
+      "step": 3491
+    },
+    {
+      "epoch": 24.613215859030838,
+      "grad_norm": 0.09247226845460027,
+      "learning_rate": 0.00010638175819295659,
+      "loss": 0.0049,
+      "step": 3492
+    },
+    {
+      "epoch": 24.620264317180617,
+      "grad_norm": 0.10142605133197828,
+      "learning_rate": 0.00010633590802163248,
+      "loss": 0.005,
+      "step": 3493
+    },
+    {
+      "epoch": 24.627312775330395,
+      "grad_norm": 0.08267729289677662,
+      "learning_rate": 0.00010629005651294192,
+      "loss": 0.0047,
+      "step": 3494
+    },
+    {
+      "epoch": 24.634361233480178,
+      "grad_norm": 0.08774906968964097,
+      "learning_rate": 0.00010624420367656319,
+      "loss": 0.005,
+      "step": 3495
+    },
+    {
+      "epoch": 24.641409691629956,
+      "grad_norm": 0.08597429002404762,
+      "learning_rate": 0.00010619834952217475,
+      "loss": 0.0047,
+      "step": 3496
+    },
+    {
+      "epoch": 24.648458149779735,
+      "grad_norm": 0.07677568408988737,
+      "learning_rate": 0.0001061524940594554,
+      "loss": 0.0044,
+      "step": 3497
+    },
+    {
+      "epoch": 24.655506607929514,
+      "grad_norm": 0.09184181124880003,
+      "learning_rate": 0.00010610663729808417,
+      "loss": 0.005,
+      "step": 3498
+    },
+    {
+      "epoch": 24.662555066079296,
+      "grad_norm": 0.08505027929764208,
+      "learning_rate": 0.00010606077924774036,
+      "loss": 0.0046,
+      "step": 3499
+    },
+    {
+      "epoch": 24.669603524229075,
+      "grad_norm": 0.07991202906842473,
+      "learning_rate": 0.00010601491991810361,
+      "loss": 0.0048,
+      "step": 3500
+    },
+    {
+      "epoch": 24.676651982378853,
+      "grad_norm": 0.09688589679240577,
+      "learning_rate": 0.00010596905931885374,
+      "loss": 0.0057,
+      "step": 3501
+    },
+    {
+      "epoch": 24.683700440528636,
+      "grad_norm": 0.08515986937441865,
+      "learning_rate": 0.00010592319745967094,
+      "loss": 0.0047,
+      "step": 3502
+    },
+    {
+      "epoch": 24.690748898678414,
+      "grad_norm": 0.09702534972909982,
+      "learning_rate": 0.00010587733435023553,
+      "loss": 0.0049,
+      "step": 3503
+    },
+    {
+      "epoch": 24.697797356828193,
+      "grad_norm": 0.09307388656481307,
+      "learning_rate": 0.00010583147000022825,
+      "loss": 0.0045,
+      "step": 3504
+    },
+    {
+      "epoch": 24.704845814977972,
+      "grad_norm": 0.08957441669166338,
+      "learning_rate": 0.00010578560441932994,
+      "loss": 0.0047,
+      "step": 3505
+    },
+    {
+      "epoch": 24.711894273127754,
+      "grad_norm": 0.08412927366339007,
+      "learning_rate": 0.00010573973761722187,
+      "loss": 0.0041,
+      "step": 3506
+    },
+    {
+      "epoch": 24.718942731277533,
+      "grad_norm": 0.0865767240639815,
+      "learning_rate": 0.00010569386960358542,
+      "loss": 0.0049,
+      "step": 3507
+    },
+    {
+      "epoch": 24.72599118942731,
+      "grad_norm": 0.08427625731857627,
+      "learning_rate": 0.00010564800038810231,
+      "loss": 0.0045,
+      "step": 3508
+    },
+    {
+      "epoch": 24.733039647577094,
+      "grad_norm": 0.09054092622507807,
+      "learning_rate": 0.00010560212998045451,
+      "loss": 0.0051,
+      "step": 3509
+    },
+    {
+      "epoch": 24.740088105726873,
+      "grad_norm": 0.0917275156515957,
+      "learning_rate": 0.0001055562583903242,
+      "loss": 0.0049,
+      "step": 3510
+    },
+    {
+      "epoch": 24.74713656387665,
+      "grad_norm": 0.09187311772611424,
+      "learning_rate": 0.0001055103856273938,
+      "loss": 0.0051,
+      "step": 3511
+    },
+    {
+      "epoch": 24.75418502202643,
+      "grad_norm": 0.08376980001186864,
+      "learning_rate": 0.00010546451170134605,
+      "loss": 0.0045,
+      "step": 3512
+    },
+    {
+      "epoch": 24.761233480176212,
+      "grad_norm": 0.07594558030510244,
+      "learning_rate": 0.00010541863662186391,
+      "loss": 0.0044,
+      "step": 3513
+    },
+    {
+      "epoch": 24.76828193832599,
+      "grad_norm": 0.1016394917142303,
+      "learning_rate": 0.00010537276039863049,
+      "loss": 0.0052,
+      "step": 3514
+    },
+    {
+      "epoch": 24.77533039647577,
+      "grad_norm": 0.09592116757826281,
+      "learning_rate": 0.00010532688304132934,
+      "loss": 0.0051,
+      "step": 3515
+    },
+    {
+      "epoch": 24.782378854625552,
+      "grad_norm": 0.08739435322498353,
+      "learning_rate": 0.000105281004559644,
+      "loss": 0.0048,
+      "step": 3516
+    },
+    {
+      "epoch": 24.78942731277533,
+      "grad_norm": 0.09569598099711926,
+      "learning_rate": 0.00010523512496325847,
+      "loss": 0.0047,
+      "step": 3517
+    },
+    {
+      "epoch": 24.79647577092511,
+      "grad_norm": 0.09996102356887383,
+      "learning_rate": 0.0001051892442618569,
+      "loss": 0.0049,
+      "step": 3518
+    },
+    {
+      "epoch": 24.80352422907489,
+      "grad_norm": 0.11707505639328585,
+      "learning_rate": 0.00010514336246512355,
+      "loss": 0.006,
+      "step": 3519
+    },
+    {
+      "epoch": 24.81057268722467,
+      "grad_norm": 0.08696556849324863,
+      "learning_rate": 0.00010509747958274315,
+      "loss": 0.0046,
+      "step": 3520
+    },
+    {
+      "epoch": 24.81762114537445,
+      "grad_norm": 0.09003173392465487,
+      "learning_rate": 0.00010505159562440048,
+      "loss": 0.005,
+      "step": 3521
+    },
+    {
+      "epoch": 24.824669603524228,
+      "grad_norm": 0.08844756351986768,
+      "learning_rate": 0.0001050057105997806,
+      "loss": 0.0048,
+      "step": 3522
+    },
+    {
+      "epoch": 24.83171806167401,
+      "grad_norm": 0.08700041156884013,
+      "learning_rate": 0.0001049598245185688,
+      "loss": 0.0046,
+      "step": 3523
+    },
+    {
+      "epoch": 24.83876651982379,
+      "grad_norm": 0.09288537095532752,
+      "learning_rate": 0.00010491393739045061,
+      "loss": 0.0047,
+      "step": 3524
+    },
+    {
+      "epoch": 24.845814977973568,
+      "grad_norm": 0.08440793415295152,
+      "learning_rate": 0.00010486804922511175,
+      "loss": 0.0047,
+      "step": 3525
+    },
+    {
+      "epoch": 24.852863436123346,
+      "grad_norm": 0.09278254883999777,
+      "learning_rate": 0.00010482216003223817,
+      "loss": 0.0047,
+      "step": 3526
+    },
+    {
+      "epoch": 24.85991189427313,
+      "grad_norm": 0.09777305605771278,
+      "learning_rate": 0.00010477626982151603,
+      "loss": 0.0049,
+      "step": 3527
+    },
+    {
+      "epoch": 24.866960352422907,
+      "grad_norm": 0.09723062226014577,
+      "learning_rate": 0.00010473037860263169,
+      "loss": 0.0049,
+      "step": 3528
+    },
+    {
+      "epoch": 24.874008810572686,
+      "grad_norm": 0.10286646353841614,
+      "learning_rate": 0.00010468448638527182,
+      "loss": 0.005,
+      "step": 3529
+    },
+    {
+      "epoch": 24.88105726872247,
+      "grad_norm": 0.10517607718807612,
+      "learning_rate": 0.00010463859317912314,
+      "loss": 0.0053,
+      "step": 3530
+    },
+    {
+      "epoch": 24.888105726872247,
+      "grad_norm": 0.08669231238683825,
+      "learning_rate": 0.00010459269899387272,
+      "loss": 0.0044,
+      "step": 3531
+    },
+    {
+      "epoch": 24.895154185022026,
+      "grad_norm": 0.10633243660246983,
+      "learning_rate": 0.00010454680383920776,
+      "loss": 0.0052,
+      "step": 3532
+    },
+    {
+      "epoch": 24.902202643171805,
+      "grad_norm": 0.09037491908483347,
+      "learning_rate": 0.00010450090772481572,
+      "loss": 0.005,
+      "step": 3533
+    },
+    {
+      "epoch": 24.909251101321587,
+      "grad_norm": 0.09697319166057101,
+      "learning_rate": 0.00010445501066038418,
+      "loss": 0.0057,
+      "step": 3534
+    },
+    {
+      "epoch": 24.916299559471366,
+      "grad_norm": 0.10110259427082091,
+      "learning_rate": 0.00010440911265560096,
+      "loss": 0.005,
+      "step": 3535
+    },
+    {
+      "epoch": 24.923348017621144,
+      "grad_norm": 0.09489341955795834,
+      "learning_rate": 0.00010436321372015414,
+      "loss": 0.005,
+      "step": 3536
+    },
+    {
+      "epoch": 24.930396475770927,
+      "grad_norm": 0.08214238666225875,
+      "learning_rate": 0.00010431731386373192,
+      "loss": 0.0046,
+      "step": 3537
+    },
+    {
+      "epoch": 24.937444933920705,
+      "grad_norm": 0.0976593332681686,
+      "learning_rate": 0.00010427141309602271,
+      "loss": 0.0049,
+      "step": 3538
+    },
+    {
+      "epoch": 24.944493392070484,
+      "grad_norm": 0.09711053812804653,
+      "learning_rate": 0.00010422551142671513,
+      "loss": 0.005,
+      "step": 3539
+    },
+    {
+      "epoch": 24.951541850220263,
+      "grad_norm": 0.09863359489200134,
+      "learning_rate": 0.00010417960886549798,
+      "loss": 0.0049,
+      "step": 3540
+    },
+    {
+      "epoch": 24.958590308370045,
+      "grad_norm": 0.09042859261224843,
+      "learning_rate": 0.00010413370542206023,
+      "loss": 0.0051,
+      "step": 3541
+    },
+    {
+      "epoch": 24.965638766519824,
+      "grad_norm": 0.09367819319318561,
+      "learning_rate": 0.00010408780110609109,
+      "loss": 0.0051,
+      "step": 3542
+    },
+    {
+      "epoch": 24.972687224669603,
+      "grad_norm": 0.09607907673462307,
+      "learning_rate": 0.00010404189592727989,
+      "loss": 0.0049,
+      "step": 3543
+    },
+    {
+      "epoch": 24.979735682819385,
+      "grad_norm": 0.09588056167087011,
+      "learning_rate": 0.00010399598989531616,
+      "loss": 0.0053,
+      "step": 3544
+    },
+    {
+      "epoch": 24.986784140969164,
+      "grad_norm": 0.09599483609145472,
+      "learning_rate": 0.00010395008301988969,
+      "loss": 0.0052,
+      "step": 3545
+    },
+    {
+      "epoch": 24.993832599118942,
+      "grad_norm": 0.09732401544403393,
+      "learning_rate": 0.00010390417531069028,
+      "loss": 0.0058,
+      "step": 3546
+    },
+    {
+      "epoch": 25.00088105726872,
+      "grad_norm": 0.1000302518855856,
+      "learning_rate": 0.00010385826677740809,
+      "loss": 0.0053,
+      "step": 3547
+    },
+    {
+      "epoch": 25.007929515418503,
+      "grad_norm": 0.07446154467155781,
+      "learning_rate": 0.00010381235742973334,
+      "loss": 0.0033,
+      "step": 3548
+    },
+    {
+      "epoch": 25.014977973568282,
+      "grad_norm": 0.07503099634684286,
+      "learning_rate": 0.00010376644727735645,
+      "loss": 0.0034,
+      "step": 3549
+    },
+    {
+      "epoch": 25.02202643171806,
+      "grad_norm": 0.07604220025057823,
+      "learning_rate": 0.00010372053632996802,
+      "loss": 0.0038,
+      "step": 3550
+    },
+    {
+      "epoch": 25.029074889867843,
+      "grad_norm": 0.08009395824394153,
+      "learning_rate": 0.00010367462459725878,
+      "loss": 0.0038,
+      "step": 3551
+    },
+    {
+      "epoch": 25.03612334801762,
+      "grad_norm": 0.07405628768970801,
+      "learning_rate": 0.0001036287120889197,
+      "loss": 0.0034,
+      "step": 3552
+    },
+    {
+      "epoch": 25.0431718061674,
+      "grad_norm": 0.0789495755689891,
+      "learning_rate": 0.00010358279881464182,
+      "loss": 0.0035,
+      "step": 3553
+    },
+    {
+      "epoch": 25.05022026431718,
+      "grad_norm": 0.08703728110606657,
+      "learning_rate": 0.00010353688478411642,
+      "loss": 0.0038,
+      "step": 3554
+    },
+    {
+      "epoch": 25.05726872246696,
+      "grad_norm": 0.07142333559086399,
+      "learning_rate": 0.00010349097000703489,
+      "loss": 0.0035,
+      "step": 3555
+    },
+    {
+      "epoch": 25.06431718061674,
+      "grad_norm": 0.07878951273025298,
+      "learning_rate": 0.00010344505449308886,
+      "loss": 0.0034,
+      "step": 3556
+    },
+    {
+      "epoch": 25.07136563876652,
+      "grad_norm": 0.08666783544268304,
+      "learning_rate": 0.00010339913825196993,
+      "loss": 0.0037,
+      "step": 3557
+    },
+    {
+      "epoch": 25.0784140969163,
+      "grad_norm": 0.0828213980566845,
+      "learning_rate": 0.0001033532212933701,
+      "loss": 0.0034,
+      "step": 3558
+    },
+    {
+      "epoch": 25.08546255506608,
+      "grad_norm": 0.072566905356979,
+      "learning_rate": 0.00010330730362698138,
+      "loss": 0.0033,
+      "step": 3559
+    },
+    {
+      "epoch": 25.09251101321586,
+      "grad_norm": 0.06487027766971411,
+      "learning_rate": 0.00010326138526249587,
+      "loss": 0.0035,
+      "step": 3560
+    },
+    {
+      "epoch": 25.099559471365637,
+      "grad_norm": 0.0628330536399355,
+      "learning_rate": 0.00010321546620960593,
+      "loss": 0.003,
+      "step": 3561
+    },
+    {
+      "epoch": 25.10660792951542,
+      "grad_norm": 0.08260819008359599,
+      "learning_rate": 0.00010316954647800407,
+      "loss": 0.0038,
+      "step": 3562
+    },
+    {
+      "epoch": 25.1136563876652,
+      "grad_norm": 0.07643102778739858,
+      "learning_rate": 0.00010312362607738288,
+      "loss": 0.0035,
+      "step": 3563
+    },
+    {
+      "epoch": 25.120704845814977,
+      "grad_norm": 0.06501440440773402,
+      "learning_rate": 0.0001030777050174351,
+      "loss": 0.003,
+      "step": 3564
+    },
+    {
+      "epoch": 25.12775330396476,
+      "grad_norm": 0.07695044721373856,
+      "learning_rate": 0.00010303178330785364,
+      "loss": 0.0034,
+      "step": 3565
+    },
+    {
+      "epoch": 25.134801762114538,
+      "grad_norm": 0.08218782433734007,
+      "learning_rate": 0.00010298586095833151,
+      "loss": 0.0039,
+      "step": 3566
+    },
+    {
+      "epoch": 25.141850220264317,
+      "grad_norm": 0.06473683760802797,
+      "learning_rate": 0.0001029399379785619,
+      "loss": 0.003,
+      "step": 3567
+    },
+    {
+      "epoch": 25.148898678414096,
+      "grad_norm": 0.06825188852961796,
+      "learning_rate": 0.0001028940143782381,
+      "loss": 0.0034,
+      "step": 3568
+    },
+    {
+      "epoch": 25.155947136563878,
+      "grad_norm": 0.08677708382885128,
+      "learning_rate": 0.0001028480901670535,
+      "loss": 0.0039,
+      "step": 3569
+    },
+    {
+      "epoch": 25.162995594713657,
+      "grad_norm": 0.06906821417084717,
+      "learning_rate": 0.00010280216535470176,
+      "loss": 0.0033,
+      "step": 3570
+    },
+    {
+      "epoch": 25.170044052863435,
+      "grad_norm": 0.0762928031925306,
+      "learning_rate": 0.00010275623995087645,
+      "loss": 0.0033,
+      "step": 3571
+    },
+    {
+      "epoch": 25.177092511013218,
+      "grad_norm": 0.07895548368680413,
+      "learning_rate": 0.00010271031396527145,
+      "loss": 0.0035,
+      "step": 3572
+    },
+    {
+      "epoch": 25.184140969162996,
+      "grad_norm": 0.07903302702094624,
+      "learning_rate": 0.00010266438740758068,
+      "loss": 0.0036,
+      "step": 3573
+    },
+    {
+      "epoch": 25.191189427312775,
+      "grad_norm": 0.07082302106638264,
+      "learning_rate": 0.00010261846028749819,
+      "loss": 0.0036,
+      "step": 3574
+    },
+    {
+      "epoch": 25.198237885462554,
+      "grad_norm": 0.06303676721436335,
+      "learning_rate": 0.00010257253261471815,
+      "loss": 0.0031,
+      "step": 3575
+    },
+    {
+      "epoch": 25.205286343612336,
+      "grad_norm": 0.07706633125501577,
+      "learning_rate": 0.00010252660439893487,
+      "loss": 0.0032,
+      "step": 3576
+    },
+    {
+      "epoch": 25.212334801762115,
+      "grad_norm": 0.08704372586381116,
+      "learning_rate": 0.00010248067564984274,
+      "loss": 0.0039,
+      "step": 3577
+    },
+    {
+      "epoch": 25.219383259911893,
+      "grad_norm": 0.06631075373915345,
+      "learning_rate": 0.00010243474637713626,
+      "loss": 0.0033,
+      "step": 3578
+    },
+    {
+      "epoch": 25.226431718061676,
+      "grad_norm": 0.06396336760423757,
+      "learning_rate": 0.00010238881659051008,
+      "loss": 0.0028,
+      "step": 3579
+    },
+    {
+      "epoch": 25.233480176211454,
+      "grad_norm": 0.08400918436092253,
+      "learning_rate": 0.00010234288629965889,
+      "loss": 0.0038,
+      "step": 3580
+    },
+    {
+      "epoch": 25.240528634361233,
+      "grad_norm": 0.07678732646509458,
+      "learning_rate": 0.00010229695551427763,
+      "loss": 0.0033,
+      "step": 3581
+    },
+    {
+      "epoch": 25.247577092511012,
+      "grad_norm": 0.07261780058248764,
+      "learning_rate": 0.00010225102424406114,
+      "loss": 0.0034,
+      "step": 3582
+    },
+    {
+      "epoch": 25.254625550660794,
+      "grad_norm": 0.0721857785143827,
+      "learning_rate": 0.00010220509249870451,
+      "loss": 0.0034,
+      "step": 3583
+    },
+    {
+      "epoch": 25.261674008810573,
+      "grad_norm": 0.06609019901686526,
+      "learning_rate": 0.00010215916028790293,
+      "loss": 0.0033,
+      "step": 3584
+    },
+    {
+      "epoch": 25.26872246696035,
+      "grad_norm": 0.0798359159904818,
+      "learning_rate": 0.00010211322762135158,
+      "loss": 0.0035,
+      "step": 3585
+    },
+    {
+      "epoch": 25.275770925110134,
+      "grad_norm": 0.07351166409867803,
+      "learning_rate": 0.00010206729450874588,
+      "loss": 0.0034,
+      "step": 3586
+    },
+    {
+      "epoch": 25.282819383259913,
+      "grad_norm": 0.07652531342696146,
+      "learning_rate": 0.00010202136095978117,
+      "loss": 0.0033,
+      "step": 3587
+    },
+    {
+      "epoch": 25.28986784140969,
+      "grad_norm": 0.07484053093162309,
+      "learning_rate": 0.00010197542698415303,
+      "loss": 0.0033,
+      "step": 3588
+    },
+    {
+      "epoch": 25.29691629955947,
+      "grad_norm": 0.07984546771484115,
+      "learning_rate": 0.00010192949259155711,
+      "loss": 0.0037,
+      "step": 3589
+    },
+    {
+      "epoch": 25.303964757709252,
+      "grad_norm": 0.06906166962075887,
+      "learning_rate": 0.0001018835577916891,
+      "loss": 0.0028,
+      "step": 3590
+    },
+    {
+      "epoch": 25.31101321585903,
+      "grad_norm": 0.07626286216642666,
+      "learning_rate": 0.00010183762259424476,
+      "loss": 0.0033,
+      "step": 3591
+    },
+    {
+      "epoch": 25.31806167400881,
+      "grad_norm": 0.07030128207049838,
+      "learning_rate": 0.00010179168700892001,
+      "loss": 0.0032,
+      "step": 3592
+    },
+    {
+      "epoch": 25.325110132158592,
+      "grad_norm": 0.07425416158586651,
+      "learning_rate": 0.00010174575104541078,
+      "loss": 0.0034,
+      "step": 3593
+    },
+    {
+      "epoch": 25.33215859030837,
+      "grad_norm": 0.07904540586177672,
+      "learning_rate": 0.00010169981471341312,
+      "loss": 0.0035,
+      "step": 3594
+    },
+    {
+      "epoch": 25.33920704845815,
+      "grad_norm": 0.06716218990984871,
+      "learning_rate": 0.00010165387802262314,
+      "loss": 0.0033,
+      "step": 3595
+    },
+    {
+      "epoch": 25.34625550660793,
+      "grad_norm": 0.07526800906860615,
+      "learning_rate": 0.00010160794098273702,
+      "loss": 0.0033,
+      "step": 3596
+    },
+    {
+      "epoch": 25.35330396475771,
+      "grad_norm": 0.07164063394055906,
+      "learning_rate": 0.00010156200360345112,
+      "loss": 0.0034,
+      "step": 3597
+    },
+    {
+      "epoch": 25.36035242290749,
+      "grad_norm": 0.08671166521585656,
+      "learning_rate": 0.00010151606589446164,
+      "loss": 0.0041,
+      "step": 3598
+    },
+    {
+      "epoch": 25.367400881057268,
+      "grad_norm": 0.07502220107231757,
+      "learning_rate": 0.00010147012786546506,
+      "loss": 0.0038,
+      "step": 3599
+    },
+    {
+      "epoch": 25.374449339207047,
+      "grad_norm": 0.07214593999840495,
+      "learning_rate": 0.00010142418952615788,
+      "loss": 0.0033,
+      "step": 3600
+    },
+    {
+      "epoch": 25.38149779735683,
+      "grad_norm": 0.07495856307655568,
+      "learning_rate": 0.00010137825088623658,
+      "loss": 0.0033,
+      "step": 3601
+    },
+    {
+      "epoch": 25.388546255506608,
+      "grad_norm": 0.07283430945468423,
+      "learning_rate": 0.0001013323119553978,
+      "loss": 0.0031,
+      "step": 3602
+    },
+    {
+      "epoch": 25.395594713656386,
+      "grad_norm": 0.06344448754640601,
+      "learning_rate": 0.00010128637274333824,
+      "loss": 0.0029,
+      "step": 3603
+    },
+    {
+      "epoch": 25.40264317180617,
+      "grad_norm": 0.0757192965985053,
+      "learning_rate": 0.00010124043325975454,
+      "loss": 0.0036,
+      "step": 3604
+    },
+    {
+      "epoch": 25.409691629955947,
+      "grad_norm": 0.07904076985833804,
+      "learning_rate": 0.00010119449351434353,
+      "loss": 0.0035,
+      "step": 3605
+    },
+    {
+      "epoch": 25.416740088105726,
+      "grad_norm": 0.07942204430705102,
+      "learning_rate": 0.00010114855351680208,
+      "loss": 0.0036,
+      "step": 3606
+    },
+    {
+      "epoch": 25.423788546255505,
+      "grad_norm": 0.07913881670160304,
+      "learning_rate": 0.00010110261327682702,
+      "loss": 0.004,
+      "step": 3607
+    },
+    {
+      "epoch": 25.430837004405287,
+      "grad_norm": 0.08006554103189793,
+      "learning_rate": 0.00010105667280411532,
+      "loss": 0.0035,
+      "step": 3608
+    },
+    {
+      "epoch": 25.437885462555066,
+      "grad_norm": 0.08457064054626044,
+      "learning_rate": 0.00010101073210836399,
+      "loss": 0.0038,
+      "step": 3609
+    },
+    {
+      "epoch": 25.444933920704845,
+      "grad_norm": 0.07598571618345265,
+      "learning_rate": 0.00010096479119927,
+      "loss": 0.0035,
+      "step": 3610
+    },
+    {
+      "epoch": 25.451982378854627,
+      "grad_norm": 0.07409818887767765,
+      "learning_rate": 0.00010091885008653053,
+      "loss": 0.0034,
+      "step": 3611
+    },
+    {
+      "epoch": 25.459030837004406,
+      "grad_norm": 0.08456726460425049,
+      "learning_rate": 0.00010087290877984263,
+      "loss": 0.0037,
+      "step": 3612
+    },
+    {
+      "epoch": 25.466079295154184,
+      "grad_norm": 0.07488654074206524,
+      "learning_rate": 0.0001008269672889035,
+      "loss": 0.0039,
+      "step": 3613
+    },
+    {
+      "epoch": 25.473127753303963,
+      "grad_norm": 0.07112461810886425,
+      "learning_rate": 0.00010078102562341036,
+      "loss": 0.0035,
+      "step": 3614
+    },
+    {
+      "epoch": 25.480176211453745,
+      "grad_norm": 0.07169537448322419,
+      "learning_rate": 0.00010073508379306043,
+      "loss": 0.0039,
+      "step": 3615
+    },
+    {
+      "epoch": 25.487224669603524,
+      "grad_norm": 0.0758893743620292,
+      "learning_rate": 0.00010068914180755098,
+      "loss": 0.0036,
+      "step": 3616
+    },
+    {
+      "epoch": 25.494273127753303,
+      "grad_norm": 0.07210106640376199,
+      "learning_rate": 0.00010064319967657934,
+      "loss": 0.0034,
+      "step": 3617
+    },
+    {
+      "epoch": 25.501321585903085,
+      "grad_norm": 0.07181281490656881,
+      "learning_rate": 0.00010059725740984284,
+      "loss": 0.0036,
+      "step": 3618
+    },
+    {
+      "epoch": 25.508370044052864,
+      "grad_norm": 0.08156644876163656,
+      "learning_rate": 0.00010055131501703886,
+      "loss": 0.0036,
+      "step": 3619
+    },
+    {
+      "epoch": 25.515418502202643,
+      "grad_norm": 0.0773645730975516,
+      "learning_rate": 0.00010050537250786478,
+      "loss": 0.0034,
+      "step": 3620
+    },
+    {
+      "epoch": 25.52246696035242,
+      "grad_norm": 0.08356517883408243,
+      "learning_rate": 0.00010045942989201802,
+      "loss": 0.004,
+      "step": 3621
+    },
+    {
+      "epoch": 25.529515418502204,
+      "grad_norm": 0.08479670409257335,
+      "learning_rate": 0.0001004134871791961,
+      "loss": 0.0045,
+      "step": 3622
+    },
+    {
+      "epoch": 25.536563876651982,
+      "grad_norm": 0.08878280736418565,
+      "learning_rate": 0.00010036754437909636,
+      "loss": 0.0036,
+      "step": 3623
+    },
+    {
+      "epoch": 25.54361233480176,
+      "grad_norm": 0.07944409757845586,
+      "learning_rate": 0.00010032160150141635,
+      "loss": 0.0038,
+      "step": 3624
+    },
+    {
+      "epoch": 25.550660792951543,
+      "grad_norm": 0.06802703141556045,
+      "learning_rate": 0.00010027565855585359,
+      "loss": 0.0032,
+      "step": 3625
+    },
+    {
+      "epoch": 25.557709251101322,
+      "grad_norm": 0.07021057373617141,
+      "learning_rate": 0.00010022971555210552,
+      "loss": 0.0036,
+      "step": 3626
+    },
+    {
+      "epoch": 25.5647577092511,
+      "grad_norm": 0.07337968116795542,
+      "learning_rate": 0.00010018377249986972,
+      "loss": 0.0036,
+      "step": 3627
+    },
+    {
+      "epoch": 25.57180616740088,
+      "grad_norm": 0.07680832731937734,
+      "learning_rate": 0.00010013782940884374,
+      "loss": 0.0038,
+      "step": 3628
+    },
+    {
+      "epoch": 25.57885462555066,
+      "grad_norm": 0.07624770091859126,
+      "learning_rate": 0.00010009188628872508,
+      "loss": 0.0037,
+      "step": 3629
+    },
+    {
+      "epoch": 25.58590308370044,
+      "grad_norm": 0.09027177063218232,
+      "learning_rate": 0.00010004594314921131,
+      "loss": 0.0035,
+      "step": 3630
+    },
+    {
+      "epoch": 25.59295154185022,
+      "grad_norm": 0.07436904228415558,
+      "learning_rate": 0.0001,
+      "loss": 0.0037,
+      "step": 3631
+    },
+    {
+      "epoch": 25.6,
+      "grad_norm": 0.08450494661618857,
+      "learning_rate": 9.995405685078867e-05,
+      "loss": 0.004,
+      "step": 3632
+    },
+    {
+      "epoch": 25.60704845814978,
+      "grad_norm": 0.07967137314195123,
+      "learning_rate": 9.990811371127495e-05,
+      "loss": 0.0037,
+      "step": 3633
+    },
+    {
+      "epoch": 25.61409691629956,
+      "grad_norm": 0.08844040994446216,
+      "learning_rate": 9.986217059115627e-05,
+      "loss": 0.0033,
+      "step": 3634
+    },
+    {
+      "epoch": 25.621145374449338,
+      "grad_norm": 0.0805372101731813,
+      "learning_rate": 9.981622750013029e-05,
+      "loss": 0.0041,
+      "step": 3635
+    },
+    {
+      "epoch": 25.62819383259912,
+      "grad_norm": 0.09098394562011483,
+      "learning_rate": 9.977028444789448e-05,
+      "loss": 0.0047,
+      "step": 3636
+    },
+    {
+      "epoch": 25.6352422907489,
+      "grad_norm": 0.07876341160614576,
+      "learning_rate": 9.972434144414646e-05,
+      "loss": 0.0033,
+      "step": 3637
+    },
+    {
+      "epoch": 25.642290748898677,
+      "grad_norm": 0.08033135053766016,
+      "learning_rate": 9.967839849858368e-05,
+      "loss": 0.0036,
+      "step": 3638
+    },
+    {
+      "epoch": 25.64933920704846,
+      "grad_norm": 0.08259655594978725,
+      "learning_rate": 9.963245562090365e-05,
+      "loss": 0.0042,
+      "step": 3639
+    },
+    {
+      "epoch": 25.65638766519824,
+      "grad_norm": 0.07188026478091643,
+      "learning_rate": 9.958651282080396e-05,
+      "loss": 0.0039,
+      "step": 3640
+    },
+    {
+      "epoch": 25.663436123348017,
+      "grad_norm": 0.08534757855510926,
+      "learning_rate": 9.9540570107982e-05,
+      "loss": 0.0038,
+      "step": 3641
+    },
+    {
+      "epoch": 25.670484581497796,
+      "grad_norm": 0.07340203989514284,
+      "learning_rate": 9.949462749213523e-05,
+      "loss": 0.0036,
+      "step": 3642
+    },
+    {
+      "epoch": 25.677533039647578,
+      "grad_norm": 0.08668412560310082,
+      "learning_rate": 9.944868498296115e-05,
+      "loss": 0.0034,
+      "step": 3643
+    },
+    {
+      "epoch": 25.684581497797357,
+      "grad_norm": 0.07757128470391442,
+      "learning_rate": 9.940274259015721e-05,
+      "loss": 0.0037,
+      "step": 3644
+    },
+    {
+      "epoch": 25.691629955947135,
+      "grad_norm": 0.07898775833495501,
+      "learning_rate": 9.935680032342068e-05,
+      "loss": 0.0035,
+      "step": 3645
+    },
+    {
+      "epoch": 25.698678414096918,
+      "grad_norm": 0.08391348647834077,
+      "learning_rate": 9.931085819244904e-05,
+      "loss": 0.0039,
+      "step": 3646
+    },
+    {
+      "epoch": 25.705726872246697,
+      "grad_norm": 0.07990310576251941,
+      "learning_rate": 9.926491620693958e-05,
+      "loss": 0.0037,
+      "step": 3647
+    },
+    {
+      "epoch": 25.712775330396475,
+      "grad_norm": 0.07092773035361812,
+      "learning_rate": 9.921897437658966e-05,
+      "loss": 0.0036,
+      "step": 3648
+    },
+    {
+      "epoch": 25.719823788546254,
+      "grad_norm": 0.08724594616847199,
+      "learning_rate": 9.917303271109652e-05,
+      "loss": 0.0036,
+      "step": 3649
+    },
+    {
+      "epoch": 25.726872246696036,
+      "grad_norm": 0.081360603522828,
+      "learning_rate": 9.912709122015738e-05,
+      "loss": 0.0037,
+      "step": 3650
+    },
+    {
+      "epoch": 25.733920704845815,
+      "grad_norm": 0.0833423252259002,
+      "learning_rate": 9.90811499134695e-05,
+      "loss": 0.0042,
+      "step": 3651
+    },
+    {
+      "epoch": 25.740969162995594,
+      "grad_norm": 0.07703143315115017,
+      "learning_rate": 9.903520880073003e-05,
+      "loss": 0.0037,
+      "step": 3652
+    },
+    {
+      "epoch": 25.748017621145376,
+      "grad_norm": 0.07356375305465303,
+      "learning_rate": 9.898926789163604e-05,
+      "loss": 0.0036,
+      "step": 3653
+    },
+    {
+      "epoch": 25.755066079295155,
+      "grad_norm": 0.08971802115355816,
+      "learning_rate": 9.894332719588469e-05,
+      "loss": 0.0045,
+      "step": 3654
+    },
+    {
+      "epoch": 25.762114537444933,
+      "grad_norm": 0.0889585415876807,
+      "learning_rate": 9.889738672317303e-05,
+      "loss": 0.0043,
+      "step": 3655
+    },
+    {
+      "epoch": 25.769162995594712,
+      "grad_norm": 0.07510275686567137,
+      "learning_rate": 9.885144648319795e-05,
+      "loss": 0.0033,
+      "step": 3656
+    },
+    {
+      "epoch": 25.776211453744494,
+      "grad_norm": 0.07329666505018603,
+      "learning_rate": 9.880550648565648e-05,
+      "loss": 0.0035,
+      "step": 3657
+    },
+    {
+      "epoch": 25.783259911894273,
+      "grad_norm": 0.07825575077154447,
+      "learning_rate": 9.875956674024551e-05,
+      "loss": 0.0041,
+      "step": 3658
+    },
+    {
+      "epoch": 25.790308370044052,
+      "grad_norm": 0.06893970541503662,
+      "learning_rate": 9.87136272566618e-05,
+      "loss": 0.0037,
+      "step": 3659
+    },
+    {
+      "epoch": 25.797356828193834,
+      "grad_norm": 0.08038071746836753,
+      "learning_rate": 9.86676880446022e-05,
+      "loss": 0.0037,
+      "step": 3660
+    },
+    {
+      "epoch": 25.804405286343613,
+      "grad_norm": 0.07691928519613074,
+      "learning_rate": 9.862174911376342e-05,
+      "loss": 0.0037,
+      "step": 3661
+    },
+    {
+      "epoch": 25.81145374449339,
+      "grad_norm": 0.08101704434611919,
+      "learning_rate": 9.857581047384216e-05,
+      "loss": 0.0039,
+      "step": 3662
+    },
+    {
+      "epoch": 25.81850220264317,
+      "grad_norm": 0.07942887031462267,
+      "learning_rate": 9.852987213453496e-05,
+      "loss": 0.0038,
+      "step": 3663
+    },
+    {
+      "epoch": 25.825550660792953,
+      "grad_norm": 0.08875972585562869,
+      "learning_rate": 9.848393410553836e-05,
+      "loss": 0.0041,
+      "step": 3664
+    },
+    {
+      "epoch": 25.83259911894273,
+      "grad_norm": 0.07624132904430149,
+      "learning_rate": 9.843799639654893e-05,
+      "loss": 0.0034,
+      "step": 3665
+    },
+    {
+      "epoch": 25.83964757709251,
+      "grad_norm": 0.08796814790321227,
+      "learning_rate": 9.839205901726299e-05,
+      "loss": 0.0045,
+      "step": 3666
+    },
+    {
+      "epoch": 25.846696035242292,
+      "grad_norm": 0.09040855720640896,
+      "learning_rate": 9.834612197737689e-05,
+      "loss": 0.0041,
+      "step": 3667
+    },
+    {
+      "epoch": 25.85374449339207,
+      "grad_norm": 0.08814442501850372,
+      "learning_rate": 9.830018528658691e-05,
+      "loss": 0.0041,
+      "step": 3668
+    },
+    {
+      "epoch": 25.86079295154185,
+      "grad_norm": 0.08134422883307434,
+      "learning_rate": 9.825424895458928e-05,
+      "loss": 0.0037,
+      "step": 3669
+    },
+    {
+      "epoch": 25.86784140969163,
+      "grad_norm": 0.08470682425858009,
+      "learning_rate": 9.820831299108003e-05,
+      "loss": 0.0042,
+      "step": 3670
+    },
+    {
+      "epoch": 25.87488986784141,
+      "grad_norm": 0.07995073324319221,
+      "learning_rate": 9.816237740575526e-05,
+      "loss": 0.0036,
+      "step": 3671
+    },
+    {
+      "epoch": 25.88193832599119,
+      "grad_norm": 0.08062457582276027,
+      "learning_rate": 9.811644220831095e-05,
+      "loss": 0.0041,
+      "step": 3672
+    },
+    {
+      "epoch": 25.888986784140968,
+      "grad_norm": 0.08409289231730768,
+      "learning_rate": 9.80705074084429e-05,
+      "loss": 0.0039,
+      "step": 3673
+    },
+    {
+      "epoch": 25.89603524229075,
+      "grad_norm": 0.08328004059897133,
+      "learning_rate": 9.802457301584698e-05,
+      "loss": 0.0041,
+      "step": 3674
+    },
+    {
+      "epoch": 25.90308370044053,
+      "grad_norm": 0.08558152012143883,
+      "learning_rate": 9.797863904021883e-05,
+      "loss": 0.0049,
+      "step": 3675
+    },
+    {
+      "epoch": 25.910132158590308,
+      "grad_norm": 0.08304411305999286,
+      "learning_rate": 9.793270549125417e-05,
+      "loss": 0.0037,
+      "step": 3676
+    },
+    {
+      "epoch": 25.917180616740087,
+      "grad_norm": 0.08276880108297067,
+      "learning_rate": 9.788677237864844e-05,
+      "loss": 0.0041,
+      "step": 3677
+    },
+    {
+      "epoch": 25.92422907488987,
+      "grad_norm": 0.09716680326008327,
+      "learning_rate": 9.784083971209708e-05,
+      "loss": 0.0043,
+      "step": 3678
+    },
+    {
+      "epoch": 25.931277533039648,
+      "grad_norm": 0.07430012151080345,
+      "learning_rate": 9.779490750129548e-05,
+      "loss": 0.0037,
+      "step": 3679
+    },
+    {
+      "epoch": 25.938325991189426,
+      "grad_norm": 0.08428775402572523,
+      "learning_rate": 9.77489757559389e-05,
+      "loss": 0.0044,
+      "step": 3680
+    },
+    {
+      "epoch": 25.94537444933921,
+      "grad_norm": 0.07539037986789966,
+      "learning_rate": 9.770304448572241e-05,
+      "loss": 0.0038,
+      "step": 3681
+    },
+    {
+      "epoch": 25.952422907488987,
+      "grad_norm": 0.08192386723540648,
+      "learning_rate": 9.765711370034111e-05,
+      "loss": 0.0038,
+      "step": 3682
+    },
+    {
+      "epoch": 25.959471365638766,
+      "grad_norm": 0.07689118320148207,
+      "learning_rate": 9.761118340948999e-05,
+      "loss": 0.0039,
+      "step": 3683
+    },
+    {
+      "epoch": 25.966519823788545,
+      "grad_norm": 0.07258683784874598,
+      "learning_rate": 9.756525362286378e-05,
+      "loss": 0.0034,
+      "step": 3684
+    },
+    {
+      "epoch": 25.973568281938327,
+      "grad_norm": 0.09332681805099605,
+      "learning_rate": 9.751932435015729e-05,
+      "loss": 0.0044,
+      "step": 3685
+    },
+    {
+      "epoch": 25.980616740088106,
+      "grad_norm": 0.08013556634230859,
+      "learning_rate": 9.747339560106514e-05,
+      "loss": 0.0043,
+      "step": 3686
+    },
+    {
+      "epoch": 25.987665198237885,
+      "grad_norm": 0.08258244057013438,
+      "learning_rate": 9.742746738528186e-05,
+      "loss": 0.004,
+      "step": 3687
+    },
+    {
+      "epoch": 25.994713656387667,
+      "grad_norm": 0.09526742562889666,
+      "learning_rate": 9.738153971250182e-05,
+      "loss": 0.0047,
+      "step": 3688
+    },
+    {
+      "epoch": 26.001762114537446,
+      "grad_norm": 0.07608445052055464,
+      "learning_rate": 9.733561259241933e-05,
+      "loss": 0.0034,
+      "step": 3689
+    },
+    {
+      "epoch": 26.008810572687224,
+      "grad_norm": 0.06355307377113206,
+      "learning_rate": 9.728968603472857e-05,
+      "loss": 0.0028,
+      "step": 3690
+    },
+    {
+      "epoch": 26.015859030837003,
+      "grad_norm": 0.06568103686371751,
+      "learning_rate": 9.724376004912358e-05,
+      "loss": 0.0025,
+      "step": 3691
+    },
+    {
+      "epoch": 26.022907488986785,
+      "grad_norm": 0.05661765923232585,
+      "learning_rate": 9.719783464529827e-05,
+      "loss": 0.0024,
+      "step": 3692
+    },
+    {
+      "epoch": 26.029955947136564,
+      "grad_norm": 0.0625978622929937,
+      "learning_rate": 9.715190983294651e-05,
+      "loss": 0.0029,
+      "step": 3693
+    },
+    {
+      "epoch": 26.037004405286343,
+      "grad_norm": 0.06717756914697157,
+      "learning_rate": 9.710598562176195e-05,
+      "loss": 0.0026,
+      "step": 3694
+    },
+    {
+      "epoch": 26.044052863436125,
+      "grad_norm": 0.058751757960955955,
+      "learning_rate": 9.706006202143812e-05,
+      "loss": 0.0024,
+      "step": 3695
+    },
+    {
+      "epoch": 26.051101321585904,
+      "grad_norm": 0.06655271350831472,
+      "learning_rate": 9.701413904166852e-05,
+      "loss": 0.003,
+      "step": 3696
+    },
+    {
+      "epoch": 26.058149779735682,
+      "grad_norm": 0.05935427843937188,
+      "learning_rate": 9.696821669214641e-05,
+      "loss": 0.0025,
+      "step": 3697
+    },
+    {
+      "epoch": 26.06519823788546,
+      "grad_norm": 0.06905606882502205,
+      "learning_rate": 9.692229498256492e-05,
+      "loss": 0.0028,
+      "step": 3698
+    },
+    {
+      "epoch": 26.072246696035243,
+      "grad_norm": 0.06421635184130013,
+      "learning_rate": 9.687637392261715e-05,
+      "loss": 0.0027,
+      "step": 3699
+    },
+    {
+      "epoch": 26.079295154185022,
+      "grad_norm": 0.06301116219230378,
+      "learning_rate": 9.683045352199592e-05,
+      "loss": 0.0027,
+      "step": 3700
+    },
+    {
+      "epoch": 26.0863436123348,
+      "grad_norm": 0.06497103817587761,
+      "learning_rate": 9.678453379039409e-05,
+      "loss": 0.0028,
+      "step": 3701
+    },
+    {
+      "epoch": 26.093392070484583,
+      "grad_norm": 0.05636647791467298,
+      "learning_rate": 9.673861473750417e-05,
+      "loss": 0.0025,
+      "step": 3702
+    },
+    {
+      "epoch": 26.100440528634362,
+      "grad_norm": 0.07241710929753986,
+      "learning_rate": 9.669269637301865e-05,
+      "loss": 0.0029,
+      "step": 3703
+    },
+    {
+      "epoch": 26.10748898678414,
+      "grad_norm": 0.07350614641157782,
+      "learning_rate": 9.66467787066299e-05,
+      "loss": 0.0029,
+      "step": 3704
+    },
+    {
+      "epoch": 26.11453744493392,
+      "grad_norm": 0.0677192508512206,
+      "learning_rate": 9.660086174803007e-05,
+      "loss": 0.0028,
+      "step": 3705
+    },
+    {
+      "epoch": 26.1215859030837,
+      "grad_norm": 0.0637010056373325,
+      "learning_rate": 9.655494550691118e-05,
+      "loss": 0.0027,
+      "step": 3706
+    },
+    {
+      "epoch": 26.12863436123348,
+      "grad_norm": 0.06390348024454753,
+      "learning_rate": 9.650902999296512e-05,
+      "loss": 0.0024,
+      "step": 3707
+    },
+    {
+      "epoch": 26.13568281938326,
+      "grad_norm": 0.06331059294193284,
+      "learning_rate": 9.646311521588363e-05,
+      "loss": 0.003,
+      "step": 3708
+    },
+    {
+      "epoch": 26.14273127753304,
+      "grad_norm": 0.06735444823099772,
+      "learning_rate": 9.64172011853582e-05,
+      "loss": 0.0032,
+      "step": 3709
+    },
+    {
+      "epoch": 26.14977973568282,
+      "grad_norm": 0.06530051138327195,
+      "learning_rate": 9.637128791108033e-05,
+      "loss": 0.0026,
+      "step": 3710
+    },
+    {
+      "epoch": 26.1568281938326,
+      "grad_norm": 0.0780686270145651,
+      "learning_rate": 9.632537540274122e-05,
+      "loss": 0.003,
+      "step": 3711
+    },
+    {
+      "epoch": 26.163876651982378,
+      "grad_norm": 0.06383076425008043,
+      "learning_rate": 9.627946367003201e-05,
+      "loss": 0.0025,
+      "step": 3712
+    },
+    {
+      "epoch": 26.17092511013216,
+      "grad_norm": 0.061815159906935924,
+      "learning_rate": 9.623355272264357e-05,
+      "loss": 0.0026,
+      "step": 3713
+    },
+    {
+      "epoch": 26.17797356828194,
+      "grad_norm": 0.06209823954738722,
+      "learning_rate": 9.618764257026665e-05,
+      "loss": 0.0026,
+      "step": 3714
+    },
+    {
+      "epoch": 26.185022026431717,
+      "grad_norm": 0.06523912272773665,
+      "learning_rate": 9.614173322259194e-05,
+      "loss": 0.0026,
+      "step": 3715
+    },
+    {
+      "epoch": 26.1920704845815,
+      "grad_norm": 0.06863670522096867,
+      "learning_rate": 9.609582468930974e-05,
+      "loss": 0.0029,
+      "step": 3716
+    },
+    {
+      "epoch": 26.19911894273128,
+      "grad_norm": 0.06258412812011971,
+      "learning_rate": 9.604991698011033e-05,
+      "loss": 0.0025,
+      "step": 3717
+    },
+    {
+      "epoch": 26.206167400881057,
+      "grad_norm": 0.07279352908954259,
+      "learning_rate": 9.600401010468385e-05,
+      "loss": 0.0027,
+      "step": 3718
+    },
+    {
+      "epoch": 26.213215859030836,
+      "grad_norm": 0.06502785008342495,
+      "learning_rate": 9.595810407272016e-05,
+      "loss": 0.0026,
+      "step": 3719
+    },
+    {
+      "epoch": 26.220264317180618,
+      "grad_norm": 0.06386882157081421,
+      "learning_rate": 9.591219889390895e-05,
+      "loss": 0.0028,
+      "step": 3720
+    },
+    {
+      "epoch": 26.227312775330397,
+      "grad_norm": 0.06666710815195057,
+      "learning_rate": 9.586629457793978e-05,
+      "loss": 0.0034,
+      "step": 3721
+    },
+    {
+      "epoch": 26.234361233480175,
+      "grad_norm": 0.05959264331942103,
+      "learning_rate": 9.582039113450208e-05,
+      "loss": 0.0025,
+      "step": 3722
+    },
+    {
+      "epoch": 26.241409691629954,
+      "grad_norm": 0.06880057856831706,
+      "learning_rate": 9.577448857328489e-05,
+      "loss": 0.0027,
+      "step": 3723
+    },
+    {
+      "epoch": 26.248458149779736,
+      "grad_norm": 0.060319656487487716,
+      "learning_rate": 9.572858690397731e-05,
+      "loss": 0.0028,
+      "step": 3724
+    },
+    {
+      "epoch": 26.255506607929515,
+      "grad_norm": 0.055954965093527265,
+      "learning_rate": 9.568268613626808e-05,
+      "loss": 0.0024,
+      "step": 3725
+    },
+    {
+      "epoch": 26.262555066079294,
+      "grad_norm": 0.06804481544925199,
+      "learning_rate": 9.56367862798459e-05,
+      "loss": 0.0026,
+      "step": 3726
+    },
+    {
+      "epoch": 26.269603524229076,
+      "grad_norm": 0.06473984638599539,
+      "learning_rate": 9.559088734439905e-05,
+      "loss": 0.0024,
+      "step": 3727
+    },
+    {
+      "epoch": 26.276651982378855,
+      "grad_norm": 0.07292924197838009,
+      "learning_rate": 9.554498933961582e-05,
+      "loss": 0.0029,
+      "step": 3728
+    },
+    {
+      "epoch": 26.283700440528634,
+      "grad_norm": 0.07261655221838194,
+      "learning_rate": 9.54990922751843e-05,
+      "loss": 0.0029,
+      "step": 3729
+    },
+    {
+      "epoch": 26.290748898678412,
+      "grad_norm": 0.06957607583085713,
+      "learning_rate": 9.545319616079225e-05,
+      "loss": 0.0027,
+      "step": 3730
+    },
+    {
+      "epoch": 26.297797356828195,
+      "grad_norm": 0.06541086364805515,
+      "learning_rate": 9.54073010061273e-05,
+      "loss": 0.0025,
+      "step": 3731
+    },
+    {
+      "epoch": 26.304845814977973,
+      "grad_norm": 0.06055804819160041,
+      "learning_rate": 9.536140682087686e-05,
+      "loss": 0.0025,
+      "step": 3732
+    },
+    {
+      "epoch": 26.311894273127752,
+      "grad_norm": 0.058948766834017334,
+      "learning_rate": 9.531551361472823e-05,
+      "loss": 0.0027,
+      "step": 3733
+    },
+    {
+      "epoch": 26.318942731277534,
+      "grad_norm": 0.05998374149555603,
+      "learning_rate": 9.526962139736833e-05,
+      "loss": 0.0027,
+      "step": 3734
+    },
+    {
+      "epoch": 26.325991189427313,
+      "grad_norm": 0.06739550613076158,
+      "learning_rate": 9.522373017848401e-05,
+      "loss": 0.0028,
+      "step": 3735
+    },
+    {
+      "epoch": 26.333039647577092,
+      "grad_norm": 0.06393398143222571,
+      "learning_rate": 9.517783996776186e-05,
+      "loss": 0.0027,
+      "step": 3736
+    },
+    {
+      "epoch": 26.34008810572687,
+      "grad_norm": 0.07004581103557578,
+      "learning_rate": 9.51319507748883e-05,
+      "loss": 0.0028,
+      "step": 3737
+    },
+    {
+      "epoch": 26.347136563876653,
+      "grad_norm": 0.07066416407590165,
+      "learning_rate": 9.508606260954941e-05,
+      "loss": 0.0031,
+      "step": 3738
+    },
+    {
+      "epoch": 26.35418502202643,
+      "grad_norm": 0.0760431219466191,
+      "learning_rate": 9.504017548143121e-05,
+      "loss": 0.003,
+      "step": 3739
+    },
+    {
+      "epoch": 26.36123348017621,
+      "grad_norm": 0.07115529229344074,
+      "learning_rate": 9.499428940021945e-05,
+      "loss": 0.0028,
+      "step": 3740
+    },
+    {
+      "epoch": 26.368281938325993,
+      "grad_norm": 0.07132318967512094,
+      "learning_rate": 9.494840437559955e-05,
+      "loss": 0.0027,
+      "step": 3741
+    },
+    {
+      "epoch": 26.37533039647577,
+      "grad_norm": 0.061125058552015075,
+      "learning_rate": 9.490252041725688e-05,
+      "loss": 0.0026,
+      "step": 3742
+    },
+    {
+      "epoch": 26.38237885462555,
+      "grad_norm": 0.06866562882502657,
+      "learning_rate": 9.485663753487644e-05,
+      "loss": 0.0029,
+      "step": 3743
+    },
+    {
+      "epoch": 26.38942731277533,
+      "grad_norm": 0.0735863351880717,
+      "learning_rate": 9.481075573814315e-05,
+      "loss": 0.0028,
+      "step": 3744
+    },
+    {
+      "epoch": 26.39647577092511,
+      "grad_norm": 0.06568575973020087,
+      "learning_rate": 9.476487503674154e-05,
+      "loss": 0.0026,
+      "step": 3745
+    },
+    {
+      "epoch": 26.40352422907489,
+      "grad_norm": 0.07311830261355065,
+      "learning_rate": 9.4718995440356e-05,
+      "loss": 0.0029,
+      "step": 3746
+    },
+    {
+      "epoch": 26.41057268722467,
+      "grad_norm": 0.060469765751126114,
+      "learning_rate": 9.46731169586707e-05,
+      "loss": 0.0024,
+      "step": 3747
+    },
+    {
+      "epoch": 26.41762114537445,
+      "grad_norm": 0.06387114626698402,
+      "learning_rate": 9.462723960136952e-05,
+      "loss": 0.0029,
+      "step": 3748
+    },
+    {
+      "epoch": 26.42466960352423,
+      "grad_norm": 0.0628558373516707,
+      "learning_rate": 9.458136337813613e-05,
+      "loss": 0.0029,
+      "step": 3749
+    },
+    {
+      "epoch": 26.431718061674008,
+      "grad_norm": 0.06799595606928972,
+      "learning_rate": 9.453548829865397e-05,
+      "loss": 0.003,
+      "step": 3750
+    },
+    {
+      "epoch": 26.438766519823787,
+      "grad_norm": 0.07267449560509008,
+      "learning_rate": 9.448961437260624e-05,
+      "loss": 0.0028,
+      "step": 3751
+    },
+    {
+      "epoch": 26.44581497797357,
+      "grad_norm": 0.0742728012880077,
+      "learning_rate": 9.444374160967584e-05,
+      "loss": 0.0028,
+      "step": 3752
+    },
+    {
+      "epoch": 26.452863436123348,
+      "grad_norm": 0.06339621962871721,
+      "learning_rate": 9.439787001954551e-05,
+      "loss": 0.0025,
+      "step": 3753
+    },
+    {
+      "epoch": 26.459911894273127,
+      "grad_norm": 0.083822434057822,
+      "learning_rate": 9.435199961189771e-05,
+      "loss": 0.0029,
+      "step": 3754
+    },
+    {
+      "epoch": 26.46696035242291,
+      "grad_norm": 0.07143186061215757,
+      "learning_rate": 9.430613039641459e-05,
+      "loss": 0.0026,
+      "step": 3755
+    },
+    {
+      "epoch": 26.474008810572688,
+      "grad_norm": 0.06208799067910844,
+      "learning_rate": 9.426026238277815e-05,
+      "loss": 0.0026,
+      "step": 3756
+    },
+    {
+      "epoch": 26.481057268722466,
+      "grad_norm": 0.06668707985807129,
+      "learning_rate": 9.421439558067004e-05,
+      "loss": 0.0032,
+      "step": 3757
+    },
+    {
+      "epoch": 26.488105726872245,
+      "grad_norm": 0.07246811443854369,
+      "learning_rate": 9.41685299997718e-05,
+      "loss": 0.0028,
+      "step": 3758
+    },
+    {
+      "epoch": 26.495154185022027,
+      "grad_norm": 0.0735485704431627,
+      "learning_rate": 9.41226656497645e-05,
+      "loss": 0.0028,
+      "step": 3759
+    },
+    {
+      "epoch": 26.502202643171806,
+      "grad_norm": 0.07394452175992912,
+      "learning_rate": 9.407680254032907e-05,
+      "loss": 0.0027,
+      "step": 3760
+    },
+    {
+      "epoch": 26.509251101321585,
+      "grad_norm": 0.06901182065832377,
+      "learning_rate": 9.403094068114626e-05,
+      "loss": 0.0031,
+      "step": 3761
+    },
+    {
+      "epoch": 26.516299559471367,
+      "grad_norm": 0.0605644426704037,
+      "learning_rate": 9.398508008189644e-05,
+      "loss": 0.0031,
+      "step": 3762
+    },
+    {
+      "epoch": 26.523348017621146,
+      "grad_norm": 0.07329897031448986,
+      "learning_rate": 9.393922075225966e-05,
+      "loss": 0.0029,
+      "step": 3763
+    },
+    {
+      "epoch": 26.530396475770925,
+      "grad_norm": 0.06748512645179935,
+      "learning_rate": 9.389336270191587e-05,
+      "loss": 0.0028,
+      "step": 3764
+    },
+    {
+      "epoch": 26.537444933920703,
+      "grad_norm": 0.06825190125115506,
+      "learning_rate": 9.384750594054465e-05,
+      "loss": 0.003,
+      "step": 3765
+    },
+    {
+      "epoch": 26.544493392070486,
+      "grad_norm": 0.07672838109895501,
+      "learning_rate": 9.380165047782528e-05,
+      "loss": 0.003,
+      "step": 3766
+    },
+    {
+      "epoch": 26.551541850220264,
+      "grad_norm": 0.06251185739282752,
+      "learning_rate": 9.375579632343684e-05,
+      "loss": 0.0025,
+      "step": 3767
+    },
+    {
+      "epoch": 26.558590308370043,
+      "grad_norm": 0.07399391875131957,
+      "learning_rate": 9.370994348705809e-05,
+      "loss": 0.0029,
+      "step": 3768
+    },
+    {
+      "epoch": 26.565638766519825,
+      "grad_norm": 0.06195700319108384,
+      "learning_rate": 9.366409197836757e-05,
+      "loss": 0.0025,
+      "step": 3769
+    },
+    {
+      "epoch": 26.572687224669604,
+      "grad_norm": 0.06497755288615327,
+      "learning_rate": 9.361824180704342e-05,
+      "loss": 0.003,
+      "step": 3770
+    },
+    {
+      "epoch": 26.579735682819383,
+      "grad_norm": 0.08188503135845168,
+      "learning_rate": 9.357239298276359e-05,
+      "loss": 0.0029,
+      "step": 3771
+    },
+    {
+      "epoch": 26.58678414096916,
+      "grad_norm": 0.06969558285541778,
+      "learning_rate": 9.352654551520579e-05,
+      "loss": 0.0029,
+      "step": 3772
+    },
+    {
+      "epoch": 26.593832599118944,
+      "grad_norm": 0.06255590078034234,
+      "learning_rate": 9.34806994140473e-05,
+      "loss": 0.0031,
+      "step": 3773
+    },
+    {
+      "epoch": 26.600881057268722,
+      "grad_norm": 0.07089477507231945,
+      "learning_rate": 9.34348546889652e-05,
+      "loss": 0.0028,
+      "step": 3774
+    },
+    {
+      "epoch": 26.6079295154185,
+      "grad_norm": 0.07433814945082722,
+      "learning_rate": 9.338901134963632e-05,
+      "loss": 0.0029,
+      "step": 3775
+    },
+    {
+      "epoch": 26.614977973568283,
+      "grad_norm": 0.06819929333711325,
+      "learning_rate": 9.334316940573714e-05,
+      "loss": 0.0028,
+      "step": 3776
+    },
+    {
+      "epoch": 26.622026431718062,
+      "grad_norm": 0.0645639192593605,
+      "learning_rate": 9.32973288669438e-05,
+      "loss": 0.0029,
+      "step": 3777
+    },
+    {
+      "epoch": 26.62907488986784,
+      "grad_norm": 0.06047432484238659,
+      "learning_rate": 9.325148974293224e-05,
+      "loss": 0.0027,
+      "step": 3778
+    },
+    {
+      "epoch": 26.63612334801762,
+      "grad_norm": 0.07137987087321561,
+      "learning_rate": 9.320565204337807e-05,
+      "loss": 0.0028,
+      "step": 3779
+    },
+    {
+      "epoch": 26.643171806167402,
+      "grad_norm": 0.07655248421233864,
+      "learning_rate": 9.315981577795655e-05,
+      "loss": 0.003,
+      "step": 3780
+    },
+    {
+      "epoch": 26.65022026431718,
+      "grad_norm": 0.06452207375452122,
+      "learning_rate": 9.31139809563427e-05,
+      "loss": 0.003,
+      "step": 3781
+    },
+    {
+      "epoch": 26.65726872246696,
+      "grad_norm": 0.06596676186205293,
+      "learning_rate": 9.306814758821118e-05,
+      "loss": 0.003,
+      "step": 3782
+    },
+    {
+      "epoch": 26.66431718061674,
+      "grad_norm": 0.08525621472769682,
+      "learning_rate": 9.302231568323646e-05,
+      "loss": 0.0033,
+      "step": 3783
+    },
+    {
+      "epoch": 26.67136563876652,
+      "grad_norm": 0.08194840279758252,
+      "learning_rate": 9.297648525109252e-05,
+      "loss": 0.003,
+      "step": 3784
+    },
+    {
+      "epoch": 26.6784140969163,
+      "grad_norm": 0.0844818421465078,
+      "learning_rate": 9.293065630145316e-05,
+      "loss": 0.0029,
+      "step": 3785
+    },
+    {
+      "epoch": 26.685462555066078,
+      "grad_norm": 0.05941447288224844,
+      "learning_rate": 9.288482884399184e-05,
+      "loss": 0.0026,
+      "step": 3786
+    },
+    {
+      "epoch": 26.69251101321586,
+      "grad_norm": 0.08057071276239601,
+      "learning_rate": 9.28390028883817e-05,
+      "loss": 0.003,
+      "step": 3787
+    },
+    {
+      "epoch": 26.69955947136564,
+      "grad_norm": 0.08190443172569843,
+      "learning_rate": 9.27931784442955e-05,
+      "loss": 0.0037,
+      "step": 3788
+    },
+    {
+      "epoch": 26.706607929515418,
+      "grad_norm": 0.07920195266232898,
+      "learning_rate": 9.274735552140582e-05,
+      "loss": 0.0031,
+      "step": 3789
+    },
+    {
+      "epoch": 26.7136563876652,
+      "grad_norm": 0.08524735787914159,
+      "learning_rate": 9.270153412938482e-05,
+      "loss": 0.0032,
+      "step": 3790
+    },
+    {
+      "epoch": 26.72070484581498,
+      "grad_norm": 0.08917749892413131,
+      "learning_rate": 9.26557142779043e-05,
+      "loss": 0.0032,
+      "step": 3791
+    },
+    {
+      "epoch": 26.727753303964757,
+      "grad_norm": 0.0977795574986954,
+      "learning_rate": 9.260989597663587e-05,
+      "loss": 0.0045,
+      "step": 3792
+    },
+    {
+      "epoch": 26.734801762114536,
+      "grad_norm": 0.07036214492956523,
+      "learning_rate": 9.256407923525066e-05,
+      "loss": 0.0036,
+      "step": 3793
+    },
+    {
+      "epoch": 26.74185022026432,
+      "grad_norm": 0.07685607896058659,
+      "learning_rate": 9.251826406341962e-05,
+      "loss": 0.0033,
+      "step": 3794
+    },
+    {
+      "epoch": 26.748898678414097,
+      "grad_norm": 0.07612310261516586,
+      "learning_rate": 9.247245047081323e-05,
+      "loss": 0.0032,
+      "step": 3795
+    },
+    {
+      "epoch": 26.755947136563876,
+      "grad_norm": 0.08746024233495302,
+      "learning_rate": 9.242663846710171e-05,
+      "loss": 0.0036,
+      "step": 3796
+    },
+    {
+      "epoch": 26.762995594713658,
+      "grad_norm": 0.08043934392470034,
+      "learning_rate": 9.238082806195499e-05,
+      "loss": 0.0034,
+      "step": 3797
+    },
+    {
+      "epoch": 26.770044052863437,
+      "grad_norm": 0.07401547173686407,
+      "learning_rate": 9.233501926504255e-05,
+      "loss": 0.0029,
+      "step": 3798
+    },
+    {
+      "epoch": 26.777092511013215,
+      "grad_norm": 0.06714995523248617,
+      "learning_rate": 9.228921208603357e-05,
+      "loss": 0.0027,
+      "step": 3799
+    },
+    {
+      "epoch": 26.784140969162994,
+      "grad_norm": 0.07260729145380355,
+      "learning_rate": 9.224340653459698e-05,
+      "loss": 0.0031,
+      "step": 3800
+    },
+    {
+      "epoch": 26.791189427312776,
+      "grad_norm": 0.07729616417367238,
+      "learning_rate": 9.219760262040125e-05,
+      "loss": 0.0029,
+      "step": 3801
+    },
+    {
+      "epoch": 26.798237885462555,
+      "grad_norm": 0.07984982101692913,
+      "learning_rate": 9.215180035311452e-05,
+      "loss": 0.0029,
+      "step": 3802
+    },
+    {
+      "epoch": 26.805286343612334,
+      "grad_norm": 0.07973080471298902,
+      "learning_rate": 9.210599974240464e-05,
+      "loss": 0.0028,
+      "step": 3803
+    },
+    {
+      "epoch": 26.812334801762116,
+      "grad_norm": 0.07161977600960741,
+      "learning_rate": 9.206020079793912e-05,
+      "loss": 0.0033,
+      "step": 3804
+    },
+    {
+      "epoch": 26.819383259911895,
+      "grad_norm": 0.07374747899436852,
+      "learning_rate": 9.201440352938497e-05,
+      "loss": 0.0036,
+      "step": 3805
+    },
+    {
+      "epoch": 26.826431718061674,
+      "grad_norm": 0.06990140777047335,
+      "learning_rate": 9.196860794640905e-05,
+      "loss": 0.0035,
+      "step": 3806
+    },
+    {
+      "epoch": 26.833480176211452,
+      "grad_norm": 0.06982103133401135,
+      "learning_rate": 9.19228140586777e-05,
+      "loss": 0.003,
+      "step": 3807
+    },
+    {
+      "epoch": 26.840528634361235,
+      "grad_norm": 0.0727577762017785,
+      "learning_rate": 9.187702187585708e-05,
+      "loss": 0.0029,
+      "step": 3808
+    },
+    {
+      "epoch": 26.847577092511013,
+      "grad_norm": 0.07595812114428536,
+      "learning_rate": 9.183123140761277e-05,
+      "loss": 0.0034,
+      "step": 3809
+    },
+    {
+      "epoch": 26.854625550660792,
+      "grad_norm": 0.06574648331112862,
+      "learning_rate": 9.178544266361012e-05,
+      "loss": 0.003,
+      "step": 3810
+    },
+    {
+      "epoch": 26.861674008810574,
+      "grad_norm": 0.07254457173296723,
+      "learning_rate": 9.173965565351412e-05,
+      "loss": 0.0034,
+      "step": 3811
+    },
+    {
+      "epoch": 26.868722466960353,
+      "grad_norm": 0.09744440532783331,
+      "learning_rate": 9.16938703869894e-05,
+      "loss": 0.004,
+      "step": 3812
+    },
+    {
+      "epoch": 26.87577092511013,
+      "grad_norm": 0.07739285127415774,
+      "learning_rate": 9.16480868737001e-05,
+      "loss": 0.0035,
+      "step": 3813
+    },
+    {
+      "epoch": 26.88281938325991,
+      "grad_norm": 0.07185268978937868,
+      "learning_rate": 9.160230512331014e-05,
+      "loss": 0.0034,
+      "step": 3814
+    },
+    {
+      "epoch": 26.889867841409693,
+      "grad_norm": 0.07436222775050211,
+      "learning_rate": 9.155652514548304e-05,
+      "loss": 0.0029,
+      "step": 3815
+    },
+    {
+      "epoch": 26.89691629955947,
+      "grad_norm": 0.07985375304982831,
+      "learning_rate": 9.151074694988183e-05,
+      "loss": 0.0033,
+      "step": 3816
+    },
+    {
+      "epoch": 26.90396475770925,
+      "grad_norm": 0.07211280189900185,
+      "learning_rate": 9.146497054616931e-05,
+      "loss": 0.0031,
+      "step": 3817
+    },
+    {
+      "epoch": 26.911013215859032,
+      "grad_norm": 0.08142021713658093,
+      "learning_rate": 9.14191959440078e-05,
+      "loss": 0.0037,
+      "step": 3818
+    },
+    {
+      "epoch": 26.91806167400881,
+      "grad_norm": 0.07882439451024129,
+      "learning_rate": 9.137342315305938e-05,
+      "loss": 0.0031,
+      "step": 3819
+    },
+    {
+      "epoch": 26.92511013215859,
+      "grad_norm": 0.07750465875051801,
+      "learning_rate": 9.132765218298549e-05,
+      "loss": 0.0032,
+      "step": 3820
+    },
+    {
+      "epoch": 26.93215859030837,
+      "grad_norm": 0.07650810965870888,
+      "learning_rate": 9.128188304344743e-05,
+      "loss": 0.003,
+      "step": 3821
+    },
+    {
+      "epoch": 26.93920704845815,
+      "grad_norm": 0.07344675300216173,
+      "learning_rate": 9.123611574410608e-05,
+      "loss": 0.0036,
+      "step": 3822
+    },
+    {
+      "epoch": 26.94625550660793,
+      "grad_norm": 0.0662485357447241,
+      "learning_rate": 9.119035029462175e-05,
+      "loss": 0.0031,
+      "step": 3823
+    },
+    {
+      "epoch": 26.95330396475771,
+      "grad_norm": 0.0689279438254466,
+      "learning_rate": 9.114458670465459e-05,
+      "loss": 0.003,
+      "step": 3824
+    },
+    {
+      "epoch": 26.96035242290749,
+      "grad_norm": 0.082248132847139,
+      "learning_rate": 9.10988249838642e-05,
+      "loss": 0.0032,
+      "step": 3825
+    },
+    {
+      "epoch": 26.96740088105727,
+      "grad_norm": 0.07493318261876501,
+      "learning_rate": 9.10530651419099e-05,
+      "loss": 0.003,
+      "step": 3826
+    },
+    {
+      "epoch": 26.974449339207048,
+      "grad_norm": 0.0981155660906076,
+      "learning_rate": 9.10073071884505e-05,
+      "loss": 0.0037,
+      "step": 3827
+    },
+    {
+      "epoch": 26.981497797356827,
+      "grad_norm": 0.08765031443712092,
+      "learning_rate": 9.096155113314444e-05,
+      "loss": 0.0032,
+      "step": 3828
+    },
+    {
+      "epoch": 26.98854625550661,
+      "grad_norm": 0.0662239954117013,
+      "learning_rate": 9.091579698564989e-05,
+      "loss": 0.003,
+      "step": 3829
+    },
+    {
+      "epoch": 26.995594713656388,
+      "grad_norm": 0.07968595952936508,
+      "learning_rate": 9.087004475562443e-05,
+      "loss": 0.0034,
+      "step": 3830
+    },
+    {
+      "epoch": 27.002643171806167,
+      "grad_norm": 0.05651047355532989,
+      "learning_rate": 9.08242944527253e-05,
+      "loss": 0.0026,
+      "step": 3831
+    },
+    {
+      "epoch": 27.00969162995595,
+      "grad_norm": 0.061764502691765776,
+      "learning_rate": 9.077854608660942e-05,
+      "loss": 0.0024,
+      "step": 3832
+    },
+    {
+      "epoch": 27.016740088105728,
+      "grad_norm": 0.05533472673735333,
+      "learning_rate": 9.073279966693323e-05,
+      "loss": 0.0022,
+      "step": 3833
+    },
+    {
+      "epoch": 27.023788546255506,
+      "grad_norm": 0.06819911264127505,
+      "learning_rate": 9.068705520335267e-05,
+      "loss": 0.0024,
+      "step": 3834
+    },
+    {
+      "epoch": 27.030837004405285,
+      "grad_norm": 0.06503559853061354,
+      "learning_rate": 9.064131270552344e-05,
+      "loss": 0.0022,
+      "step": 3835
+    },
+    {
+      "epoch": 27.037885462555067,
+      "grad_norm": 0.062358589182037454,
+      "learning_rate": 9.059557218310075e-05,
+      "loss": 0.0021,
+      "step": 3836
+    },
+    {
+      "epoch": 27.044933920704846,
+      "grad_norm": 0.0505319021373284,
+      "learning_rate": 9.054983364573932e-05,
+      "loss": 0.002,
+      "step": 3837
+    },
+    {
+      "epoch": 27.051982378854625,
+      "grad_norm": 0.09676942751357648,
+      "learning_rate": 9.050409710309356e-05,
+      "loss": 0.0024,
+      "step": 3838
+    },
+    {
+      "epoch": 27.059030837004407,
+      "grad_norm": 0.06421410779570301,
+      "learning_rate": 9.045836256481741e-05,
+      "loss": 0.0023,
+      "step": 3839
+    },
+    {
+      "epoch": 27.066079295154186,
+      "grad_norm": 0.08090932174522342,
+      "learning_rate": 9.041263004056443e-05,
+      "loss": 0.0023,
+      "step": 3840
+    },
+    {
+      "epoch": 27.073127753303964,
+      "grad_norm": 0.06289529220855745,
+      "learning_rate": 9.036689953998766e-05,
+      "loss": 0.0023,
+      "step": 3841
+    },
+    {
+      "epoch": 27.080176211453743,
+      "grad_norm": 0.06619912230324801,
+      "learning_rate": 9.032117107273977e-05,
+      "loss": 0.0021,
+      "step": 3842
+    },
+    {
+      "epoch": 27.087224669603525,
+      "grad_norm": 0.05686219988978407,
+      "learning_rate": 9.027544464847306e-05,
+      "loss": 0.0024,
+      "step": 3843
+    },
+    {
+      "epoch": 27.094273127753304,
+      "grad_norm": 0.06563926950757783,
+      "learning_rate": 9.022972027683932e-05,
+      "loss": 0.0024,
+      "step": 3844
+    },
+    {
+      "epoch": 27.101321585903083,
+      "grad_norm": 0.06728826229396637,
+      "learning_rate": 9.018399796748987e-05,
+      "loss": 0.0028,
+      "step": 3845
+    },
+    {
+      "epoch": 27.10837004405286,
+      "grad_norm": 0.05536580436753439,
+      "learning_rate": 9.013827773007571e-05,
+      "loss": 0.0021,
+      "step": 3846
+    },
+    {
+      "epoch": 27.115418502202644,
+      "grad_norm": 0.058999516879247815,
+      "learning_rate": 9.009255957424736e-05,
+      "loss": 0.002,
+      "step": 3847
+    },
+    {
+      "epoch": 27.122466960352423,
+      "grad_norm": 0.061905786152846344,
+      "learning_rate": 9.004684350965481e-05,
+      "loss": 0.0023,
+      "step": 3848
+    },
+    {
+      "epoch": 27.1295154185022,
+      "grad_norm": 0.058232659118336576,
+      "learning_rate": 9.000112954594775e-05,
+      "loss": 0.0023,
+      "step": 3849
+    },
+    {
+      "epoch": 27.136563876651984,
+      "grad_norm": 0.059381249896848017,
+      "learning_rate": 8.995541769277531e-05,
+      "loss": 0.002,
+      "step": 3850
+    },
+    {
+      "epoch": 27.143612334801762,
+      "grad_norm": 0.07795693669171921,
+      "learning_rate": 8.990970795978632e-05,
+      "loss": 0.003,
+      "step": 3851
+    },
+    {
+      "epoch": 27.15066079295154,
+      "grad_norm": 0.05138405644602429,
+      "learning_rate": 8.986400035662896e-05,
+      "loss": 0.0023,
+      "step": 3852
+    },
+    {
+      "epoch": 27.15770925110132,
+      "grad_norm": 0.0744367555821349,
+      "learning_rate": 8.98182948929511e-05,
+      "loss": 0.0028,
+      "step": 3853
+    },
+    {
+      "epoch": 27.164757709251102,
+      "grad_norm": 0.05760709673666563,
+      "learning_rate": 8.977259157840017e-05,
+      "loss": 0.0023,
+      "step": 3854
+    },
+    {
+      "epoch": 27.17180616740088,
+      "grad_norm": 0.06718547070518165,
+      "learning_rate": 8.972689042262305e-05,
+      "loss": 0.0022,
+      "step": 3855
+    },
+    {
+      "epoch": 27.17885462555066,
+      "grad_norm": 0.06121067266255703,
+      "learning_rate": 8.96811914352662e-05,
+      "loss": 0.0022,
+      "step": 3856
+    },
+    {
+      "epoch": 27.185903083700442,
+      "grad_norm": 0.06306977824722769,
+      "learning_rate": 8.963549462597571e-05,
+      "loss": 0.0021,
+      "step": 3857
+    },
+    {
+      "epoch": 27.19295154185022,
+      "grad_norm": 0.07005315668151649,
+      "learning_rate": 8.958980000439713e-05,
+      "loss": 0.0025,
+      "step": 3858
+    },
+    {
+      "epoch": 27.2,
+      "grad_norm": 0.05807502859541454,
+      "learning_rate": 8.954410758017548e-05,
+      "loss": 0.0022,
+      "step": 3859
+    },
+    {
+      "epoch": 27.207048458149778,
+      "grad_norm": 0.05763231535437903,
+      "learning_rate": 8.949841736295545e-05,
+      "loss": 0.0021,
+      "step": 3860
+    },
+    {
+      "epoch": 27.21409691629956,
+      "grad_norm": 0.06894812998992146,
+      "learning_rate": 8.945272936238125e-05,
+      "loss": 0.0022,
+      "step": 3861
+    },
+    {
+      "epoch": 27.22114537444934,
+      "grad_norm": 0.050540326067831196,
+      "learning_rate": 8.940704358809649e-05,
+      "loss": 0.0021,
+      "step": 3862
+    },
+    {
+      "epoch": 27.228193832599118,
+      "grad_norm": 0.06875904665642935,
+      "learning_rate": 8.936136004974445e-05,
+      "loss": 0.0035,
+      "step": 3863
+    },
+    {
+      "epoch": 27.2352422907489,
+      "grad_norm": 0.06856080097780766,
+      "learning_rate": 8.931567875696788e-05,
+      "loss": 0.0024,
+      "step": 3864
+    },
+    {
+      "epoch": 27.24229074889868,
+      "grad_norm": 0.04956787849311579,
+      "learning_rate": 8.926999971940909e-05,
+      "loss": 0.002,
+      "step": 3865
+    },
+    {
+      "epoch": 27.249339207048457,
+      "grad_norm": 0.07026149719915464,
+      "learning_rate": 8.922432294670985e-05,
+      "loss": 0.0027,
+      "step": 3866
+    },
+    {
+      "epoch": 27.256387665198236,
+      "grad_norm": 0.0692823309653137,
+      "learning_rate": 8.917864844851149e-05,
+      "loss": 0.0027,
+      "step": 3867
+    },
+    {
+      "epoch": 27.26343612334802,
+      "grad_norm": 0.05373530455972185,
+      "learning_rate": 8.913297623445487e-05,
+      "loss": 0.0022,
+      "step": 3868
+    },
+    {
+      "epoch": 27.270484581497797,
+      "grad_norm": 0.05887475177314238,
+      "learning_rate": 8.908730631418041e-05,
+      "loss": 0.0022,
+      "step": 3869
+    },
+    {
+      "epoch": 27.277533039647576,
+      "grad_norm": 0.05905921797015142,
+      "learning_rate": 8.904163869732789e-05,
+      "loss": 0.0026,
+      "step": 3870
+    },
+    {
+      "epoch": 27.284581497797358,
+      "grad_norm": 0.052777804784927766,
+      "learning_rate": 8.899597339353679e-05,
+      "loss": 0.0022,
+      "step": 3871
+    },
+    {
+      "epoch": 27.291629955947137,
+      "grad_norm": 0.058456433617638096,
+      "learning_rate": 8.895031041244603e-05,
+      "loss": 0.0021,
+      "step": 3872
+    },
+    {
+      "epoch": 27.298678414096916,
+      "grad_norm": 0.06783400462458433,
+      "learning_rate": 8.890464976369393e-05,
+      "loss": 0.0023,
+      "step": 3873
+    },
+    {
+      "epoch": 27.305726872246694,
+      "grad_norm": 0.0684930202299146,
+      "learning_rate": 8.88589914569185e-05,
+      "loss": 0.0027,
+      "step": 3874
+    },
+    {
+      "epoch": 27.312775330396477,
+      "grad_norm": 0.06304536824260362,
+      "learning_rate": 8.881333550175713e-05,
+      "loss": 0.0028,
+      "step": 3875
+    },
+    {
+      "epoch": 27.319823788546255,
+      "grad_norm": 0.062828882481563,
+      "learning_rate": 8.876768190784684e-05,
+      "loss": 0.0024,
+      "step": 3876
+    },
+    {
+      "epoch": 27.326872246696034,
+      "grad_norm": 0.05369915486512806,
+      "learning_rate": 8.872203068482397e-05,
+      "loss": 0.0022,
+      "step": 3877
+    },
+    {
+      "epoch": 27.333920704845816,
+      "grad_norm": 0.06556580497545973,
+      "learning_rate": 8.867638184232446e-05,
+      "loss": 0.0024,
+      "step": 3878
+    },
+    {
+      "epoch": 27.340969162995595,
+      "grad_norm": 0.05263954050881244,
+      "learning_rate": 8.863073538998384e-05,
+      "loss": 0.002,
+      "step": 3879
+    },
+    {
+      "epoch": 27.348017621145374,
+      "grad_norm": 0.05972025253616244,
+      "learning_rate": 8.858509133743694e-05,
+      "loss": 0.0022,
+      "step": 3880
+    },
+    {
+      "epoch": 27.355066079295153,
+      "grad_norm": 0.0676119825425627,
+      "learning_rate": 8.85394496943182e-05,
+      "loss": 0.0024,
+      "step": 3881
+    },
+    {
+      "epoch": 27.362114537444935,
+      "grad_norm": 0.07366770116794435,
+      "learning_rate": 8.849381047026159e-05,
+      "loss": 0.0026,
+      "step": 3882
+    },
+    {
+      "epoch": 27.369162995594714,
+      "grad_norm": 0.06390223199373173,
+      "learning_rate": 8.844817367490049e-05,
+      "loss": 0.0027,
+      "step": 3883
+    },
+    {
+      "epoch": 27.376211453744492,
+      "grad_norm": 0.053467755817945745,
+      "learning_rate": 8.840253931786775e-05,
+      "loss": 0.0022,
+      "step": 3884
+    },
+    {
+      "epoch": 27.383259911894275,
+      "grad_norm": 0.048580669965849144,
+      "learning_rate": 8.83569074087958e-05,
+      "loss": 0.0021,
+      "step": 3885
+    },
+    {
+      "epoch": 27.390308370044053,
+      "grad_norm": 0.05409101246751172,
+      "learning_rate": 8.83112779573165e-05,
+      "loss": 0.002,
+      "step": 3886
+    },
+    {
+      "epoch": 27.397356828193832,
+      "grad_norm": 0.06387809911717293,
+      "learning_rate": 8.826565097306112e-05,
+      "loss": 0.0023,
+      "step": 3887
+    },
+    {
+      "epoch": 27.40440528634361,
+      "grad_norm": 0.06339780235109485,
+      "learning_rate": 8.822002646566056e-05,
+      "loss": 0.0021,
+      "step": 3888
+    },
+    {
+      "epoch": 27.411453744493393,
+      "grad_norm": 0.05416019313512758,
+      "learning_rate": 8.817440444474507e-05,
+      "loss": 0.0023,
+      "step": 3889
+    },
+    {
+      "epoch": 27.41850220264317,
+      "grad_norm": 0.05240190877086559,
+      "learning_rate": 8.812878491994449e-05,
+      "loss": 0.0021,
+      "step": 3890
+    },
+    {
+      "epoch": 27.42555066079295,
+      "grad_norm": 0.06363727852793433,
+      "learning_rate": 8.808316790088801e-05,
+      "loss": 0.0023,
+      "step": 3891
+    },
+    {
+      "epoch": 27.432599118942733,
+      "grad_norm": 0.062122384889743636,
+      "learning_rate": 8.803755339720432e-05,
+      "loss": 0.0024,
+      "step": 3892
+    },
+    {
+      "epoch": 27.43964757709251,
+      "grad_norm": 0.06402558890736913,
+      "learning_rate": 8.799194141852168e-05,
+      "loss": 0.0021,
+      "step": 3893
+    },
+    {
+      "epoch": 27.44669603524229,
+      "grad_norm": 0.05620634308684407,
+      "learning_rate": 8.79463319744677e-05,
+      "loss": 0.0026,
+      "step": 3894
+    },
+    {
+      "epoch": 27.45374449339207,
+      "grad_norm": 0.05959465842501195,
+      "learning_rate": 8.79007250746695e-05,
+      "loss": 0.0021,
+      "step": 3895
+    },
+    {
+      "epoch": 27.46079295154185,
+      "grad_norm": 0.06681884507104052,
+      "learning_rate": 8.785512072875369e-05,
+      "loss": 0.0026,
+      "step": 3896
+    },
+    {
+      "epoch": 27.46784140969163,
+      "grad_norm": 0.06710986480208676,
+      "learning_rate": 8.780951894634629e-05,
+      "loss": 0.0026,
+      "step": 3897
+    },
+    {
+      "epoch": 27.47488986784141,
+      "grad_norm": 0.06381154475744516,
+      "learning_rate": 8.776391973707276e-05,
+      "loss": 0.0026,
+      "step": 3898
+    },
+    {
+      "epoch": 27.48193832599119,
+      "grad_norm": 0.0581918912932371,
+      "learning_rate": 8.771832311055814e-05,
+      "loss": 0.0021,
+      "step": 3899
+    },
+    {
+      "epoch": 27.48898678414097,
+      "grad_norm": 0.060832830573274314,
+      "learning_rate": 8.767272907642678e-05,
+      "loss": 0.0025,
+      "step": 3900
+    },
+    {
+      "epoch": 27.49603524229075,
+      "grad_norm": 0.05326306024396483,
+      "learning_rate": 8.76271376443026e-05,
+      "loss": 0.0023,
+      "step": 3901
+    },
+    {
+      "epoch": 27.503083700440527,
+      "grad_norm": 0.058039555225398304,
+      "learning_rate": 8.758154882380888e-05,
+      "loss": 0.002,
+      "step": 3902
+    },
+    {
+      "epoch": 27.51013215859031,
+      "grad_norm": 0.06856284266872749,
+      "learning_rate": 8.753596262456836e-05,
+      "loss": 0.0029,
+      "step": 3903
+    },
+    {
+      "epoch": 27.517180616740088,
+      "grad_norm": 0.07259747981484488,
+      "learning_rate": 8.749037905620334e-05,
+      "loss": 0.0027,
+      "step": 3904
+    },
+    {
+      "epoch": 27.524229074889867,
+      "grad_norm": 0.06122881907455969,
+      "learning_rate": 8.744479812833539e-05,
+      "loss": 0.0024,
+      "step": 3905
+    },
+    {
+      "epoch": 27.53127753303965,
+      "grad_norm": 0.06298462904679507,
+      "learning_rate": 8.739921985058562e-05,
+      "loss": 0.0022,
+      "step": 3906
+    },
+    {
+      "epoch": 27.538325991189428,
+      "grad_norm": 0.06823562594916276,
+      "learning_rate": 8.735364423257461e-05,
+      "loss": 0.0027,
+      "step": 3907
+    },
+    {
+      "epoch": 27.545374449339207,
+      "grad_norm": 0.06231238311619491,
+      "learning_rate": 8.730807128392235e-05,
+      "loss": 0.0023,
+      "step": 3908
+    },
+    {
+      "epoch": 27.552422907488985,
+      "grad_norm": 0.060069874759641234,
+      "learning_rate": 8.726250101424817e-05,
+      "loss": 0.0024,
+      "step": 3909
+    },
+    {
+      "epoch": 27.559471365638768,
+      "grad_norm": 0.053638486625988364,
+      "learning_rate": 8.721693343317101e-05,
+      "loss": 0.0022,
+      "step": 3910
+    },
+    {
+      "epoch": 27.566519823788546,
+      "grad_norm": 0.05233597085048521,
+      "learning_rate": 8.717136855030914e-05,
+      "loss": 0.002,
+      "step": 3911
+    },
+    {
+      "epoch": 27.573568281938325,
+      "grad_norm": 0.07928684158490289,
+      "learning_rate": 8.712580637528019e-05,
+      "loss": 0.0025,
+      "step": 3912
+    },
+    {
+      "epoch": 27.580616740088107,
+      "grad_norm": 0.06390504092266448,
+      "learning_rate": 8.70802469177014e-05,
+      "loss": 0.0025,
+      "step": 3913
+    },
+    {
+      "epoch": 27.587665198237886,
+      "grad_norm": 0.06784595500802881,
+      "learning_rate": 8.703469018718927e-05,
+      "loss": 0.0023,
+      "step": 3914
+    },
+    {
+      "epoch": 27.594713656387665,
+      "grad_norm": 0.05267894198868533,
+      "learning_rate": 8.698913619335988e-05,
+      "loss": 0.002,
+      "step": 3915
+    },
+    {
+      "epoch": 27.601762114537443,
+      "grad_norm": 0.06154473635503828,
+      "learning_rate": 8.694358494582852e-05,
+      "loss": 0.0023,
+      "step": 3916
+    },
+    {
+      "epoch": 27.608810572687226,
+      "grad_norm": 0.057660127590201296,
+      "learning_rate": 8.689803645421011e-05,
+      "loss": 0.0022,
+      "step": 3917
+    },
+    {
+      "epoch": 27.615859030837004,
+      "grad_norm": 0.06415314650483112,
+      "learning_rate": 8.685249072811894e-05,
+      "loss": 0.0022,
+      "step": 3918
+    },
+    {
+      "epoch": 27.622907488986783,
+      "grad_norm": 0.06615951122010083,
+      "learning_rate": 8.680694777716854e-05,
+      "loss": 0.0029,
+      "step": 3919
+    },
+    {
+      "epoch": 27.629955947136565,
+      "grad_norm": 0.058782984080587845,
+      "learning_rate": 8.676140761097211e-05,
+      "loss": 0.0021,
+      "step": 3920
+    },
+    {
+      "epoch": 27.637004405286344,
+      "grad_norm": 0.07415179608286841,
+      "learning_rate": 8.671587023914209e-05,
+      "loss": 0.0028,
+      "step": 3921
+    },
+    {
+      "epoch": 27.644052863436123,
+      "grad_norm": 0.06531885377836137,
+      "learning_rate": 8.667033567129044e-05,
+      "loss": 0.0024,
+      "step": 3922
+    },
+    {
+      "epoch": 27.6511013215859,
+      "grad_norm": 0.0660576018074815,
+      "learning_rate": 8.662480391702844e-05,
+      "loss": 0.0023,
+      "step": 3923
+    },
+    {
+      "epoch": 27.658149779735684,
+      "grad_norm": 0.0678805396516654,
+      "learning_rate": 8.657927498596676e-05,
+      "loss": 0.0025,
+      "step": 3924
+    },
+    {
+      "epoch": 27.665198237885463,
+      "grad_norm": 0.05642648447967276,
+      "learning_rate": 8.653374888771561e-05,
+      "loss": 0.0022,
+      "step": 3925
+    },
+    {
+      "epoch": 27.67224669603524,
+      "grad_norm": 0.06433209195533911,
+      "learning_rate": 8.648822563188452e-05,
+      "loss": 0.0023,
+      "step": 3926
+    },
+    {
+      "epoch": 27.679295154185024,
+      "grad_norm": 0.0560885504731652,
+      "learning_rate": 8.644270522808231e-05,
+      "loss": 0.0022,
+      "step": 3927
+    },
+    {
+      "epoch": 27.686343612334802,
+      "grad_norm": 0.07958399911253791,
+      "learning_rate": 8.63971876859174e-05,
+      "loss": 0.003,
+      "step": 3928
+    },
+    {
+      "epoch": 27.69339207048458,
+      "grad_norm": 0.06680228854498717,
+      "learning_rate": 8.635167301499752e-05,
+      "loss": 0.0028,
+      "step": 3929
+    },
+    {
+      "epoch": 27.70044052863436,
+      "grad_norm": 0.0657566466387974,
+      "learning_rate": 8.630616122492967e-05,
+      "loss": 0.0026,
+      "step": 3930
+    },
+    {
+      "epoch": 27.707488986784142,
+      "grad_norm": 0.06367699884340089,
+      "learning_rate": 8.626065232532047e-05,
+      "loss": 0.0025,
+      "step": 3931
+    },
+    {
+      "epoch": 27.71453744493392,
+      "grad_norm": 0.05813214418999152,
+      "learning_rate": 8.621514632577577e-05,
+      "loss": 0.0024,
+      "step": 3932
+    },
+    {
+      "epoch": 27.7215859030837,
+      "grad_norm": 0.0583146750959745,
+      "learning_rate": 8.61696432359009e-05,
+      "loss": 0.0023,
+      "step": 3933
+    },
+    {
+      "epoch": 27.728634361233482,
+      "grad_norm": 0.08090831272922762,
+      "learning_rate": 8.612414306530046e-05,
+      "loss": 0.0026,
+      "step": 3934
+    },
+    {
+      "epoch": 27.73568281938326,
+      "grad_norm": 0.07193110543690875,
+      "learning_rate": 8.607864582357852e-05,
+      "loss": 0.0025,
+      "step": 3935
+    },
+    {
+      "epoch": 27.74273127753304,
+      "grad_norm": 0.0825066516299263,
+      "learning_rate": 8.603315152033857e-05,
+      "loss": 0.003,
+      "step": 3936
+    },
+    {
+      "epoch": 27.749779735682818,
+      "grad_norm": 0.06762193373121052,
+      "learning_rate": 8.598766016518337e-05,
+      "loss": 0.0026,
+      "step": 3937
+    },
+    {
+      "epoch": 27.7568281938326,
+      "grad_norm": 0.08397145645537993,
+      "learning_rate": 8.594217176771511e-05,
+      "loss": 0.0026,
+      "step": 3938
+    },
+    {
+      "epoch": 27.76387665198238,
+      "grad_norm": 0.0686479044518124,
+      "learning_rate": 8.589668633753538e-05,
+      "loss": 0.0025,
+      "step": 3939
+    },
+    {
+      "epoch": 27.770925110132158,
+      "grad_norm": 0.07344771067921581,
+      "learning_rate": 8.585120388424515e-05,
+      "loss": 0.0027,
+      "step": 3940
+    },
+    {
+      "epoch": 27.77797356828194,
+      "grad_norm": 0.05660008112160205,
+      "learning_rate": 8.580572441744466e-05,
+      "loss": 0.0025,
+      "step": 3941
+    },
+    {
+      "epoch": 27.78502202643172,
+      "grad_norm": 0.07169088034988487,
+      "learning_rate": 8.576024794673368e-05,
+      "loss": 0.0025,
+      "step": 3942
+    },
+    {
+      "epoch": 27.792070484581497,
+      "grad_norm": 0.06572190999689423,
+      "learning_rate": 8.57147744817112e-05,
+      "loss": 0.0023,
+      "step": 3943
+    },
+    {
+      "epoch": 27.799118942731276,
+      "grad_norm": 0.0679987990060795,
+      "learning_rate": 8.566930403197564e-05,
+      "loss": 0.0024,
+      "step": 3944
+    },
+    {
+      "epoch": 27.80616740088106,
+      "grad_norm": 0.05659407019947621,
+      "learning_rate": 8.56238366071248e-05,
+      "loss": 0.0023,
+      "step": 3945
+    },
+    {
+      "epoch": 27.813215859030837,
+      "grad_norm": 0.06661048427276421,
+      "learning_rate": 8.55783722167558e-05,
+      "loss": 0.0026,
+      "step": 3946
+    },
+    {
+      "epoch": 27.820264317180616,
+      "grad_norm": 0.06371299132258669,
+      "learning_rate": 8.553291087046519e-05,
+      "loss": 0.0026,
+      "step": 3947
+    },
+    {
+      "epoch": 27.827312775330398,
+      "grad_norm": 0.07451551702460328,
+      "learning_rate": 8.548745257784878e-05,
+      "loss": 0.0026,
+      "step": 3948
+    },
+    {
+      "epoch": 27.834361233480177,
+      "grad_norm": 0.06036937353343414,
+      "learning_rate": 8.544199734850178e-05,
+      "loss": 0.0025,
+      "step": 3949
+    },
+    {
+      "epoch": 27.841409691629956,
+      "grad_norm": 0.07645195584747952,
+      "learning_rate": 8.539654519201879e-05,
+      "loss": 0.0029,
+      "step": 3950
+    },
+    {
+      "epoch": 27.848458149779734,
+      "grad_norm": 0.05252506272663226,
+      "learning_rate": 8.535109611799375e-05,
+      "loss": 0.0025,
+      "step": 3951
+    },
+    {
+      "epoch": 27.855506607929517,
+      "grad_norm": 0.0623826022767964,
+      "learning_rate": 8.530565013601983e-05,
+      "loss": 0.0026,
+      "step": 3952
+    },
+    {
+      "epoch": 27.862555066079295,
+      "grad_norm": 0.06720919383222575,
+      "learning_rate": 8.526020725568974e-05,
+      "loss": 0.0029,
+      "step": 3953
+    },
+    {
+      "epoch": 27.869603524229074,
+      "grad_norm": 0.06477823332641706,
+      "learning_rate": 8.521476748659544e-05,
+      "loss": 0.0025,
+      "step": 3954
+    },
+    {
+      "epoch": 27.876651982378856,
+      "grad_norm": 0.05937084176931736,
+      "learning_rate": 8.516933083832814e-05,
+      "loss": 0.0024,
+      "step": 3955
+    },
+    {
+      "epoch": 27.883700440528635,
+      "grad_norm": 0.06916211560009473,
+      "learning_rate": 8.512389732047859e-05,
+      "loss": 0.0027,
+      "step": 3956
+    },
+    {
+      "epoch": 27.890748898678414,
+      "grad_norm": 0.09175732696801583,
+      "learning_rate": 8.50784669426367e-05,
+      "loss": 0.0026,
+      "step": 3957
+    },
+    {
+      "epoch": 27.897797356828193,
+      "grad_norm": 0.06317594543540363,
+      "learning_rate": 8.503303971439187e-05,
+      "loss": 0.003,
+      "step": 3958
+    },
+    {
+      "epoch": 27.904845814977975,
+      "grad_norm": 0.07668026721959965,
+      "learning_rate": 8.49876156453327e-05,
+      "loss": 0.0028,
+      "step": 3959
+    },
+    {
+      "epoch": 27.911894273127754,
+      "grad_norm": 0.06153459039400331,
+      "learning_rate": 8.494219474504717e-05,
+      "loss": 0.0026,
+      "step": 3960
+    },
+    {
+      "epoch": 27.918942731277532,
+      "grad_norm": 0.0637133612146345,
+      "learning_rate": 8.489677702312267e-05,
+      "loss": 0.0027,
+      "step": 3961
+    },
+    {
+      "epoch": 27.925991189427315,
+      "grad_norm": 0.07336879785601497,
+      "learning_rate": 8.485136248914576e-05,
+      "loss": 0.0029,
+      "step": 3962
+    },
+    {
+      "epoch": 27.933039647577093,
+      "grad_norm": 0.06767615447521592,
+      "learning_rate": 8.480595115270246e-05,
+      "loss": 0.0025,
+      "step": 3963
+    },
+    {
+      "epoch": 27.940088105726872,
+      "grad_norm": 0.07145795316060376,
+      "learning_rate": 8.476054302337809e-05,
+      "loss": 0.0025,
+      "step": 3964
+    },
+    {
+      "epoch": 27.94713656387665,
+      "grad_norm": 0.06885988402358863,
+      "learning_rate": 8.471513811075729e-05,
+      "loss": 0.0026,
+      "step": 3965
+    },
+    {
+      "epoch": 27.954185022026433,
+      "grad_norm": 0.06291644492396306,
+      "learning_rate": 8.466973642442392e-05,
+      "loss": 0.0024,
+      "step": 3966
+    },
+    {
+      "epoch": 27.96123348017621,
+      "grad_norm": 0.07332653498487546,
+      "learning_rate": 8.462433797396133e-05,
+      "loss": 0.0028,
+      "step": 3967
+    },
+    {
+      "epoch": 27.96828193832599,
+      "grad_norm": 0.0733295173260923,
+      "learning_rate": 8.45789427689521e-05,
+      "loss": 0.0028,
+      "step": 3968
+    },
+    {
+      "epoch": 27.97533039647577,
+      "grad_norm": 0.06603512619793804,
+      "learning_rate": 8.453355081897805e-05,
+      "loss": 0.0027,
+      "step": 3969
+    },
+    {
+      "epoch": 27.98237885462555,
+      "grad_norm": 0.06849891323154982,
+      "learning_rate": 8.448816213362048e-05,
+      "loss": 0.0025,
+      "step": 3970
+    },
+    {
+      "epoch": 27.98942731277533,
+      "grad_norm": 0.07424226112906962,
+      "learning_rate": 8.444277672245984e-05,
+      "loss": 0.0027,
+      "step": 3971
+    },
+    {
+      "epoch": 27.99647577092511,
+      "grad_norm": 0.06401226609023465,
+      "learning_rate": 8.439739459507604e-05,
+      "loss": 0.0027,
+      "step": 3972
+    },
+    {
+      "epoch": 28.00352422907489,
+      "grad_norm": 0.06176708519180175,
+      "learning_rate": 8.435201576104816e-05,
+      "loss": 0.0021,
+      "step": 3973
+    },
+    {
+      "epoch": 28.01057268722467,
+      "grad_norm": 0.05952326919224192,
+      "learning_rate": 8.430664022995461e-05,
+      "loss": 0.0019,
+      "step": 3974
+    },
+    {
+      "epoch": 28.01762114537445,
+      "grad_norm": 0.06648352087328521,
+      "learning_rate": 8.426126801137321e-05,
+      "loss": 0.0022,
+      "step": 3975
+    },
+    {
+      "epoch": 28.024669603524227,
+      "grad_norm": 0.04569908297373411,
+      "learning_rate": 8.421589911488101e-05,
+      "loss": 0.0017,
+      "step": 3976
+    },
+    {
+      "epoch": 28.03171806167401,
+      "grad_norm": 0.03958254521376677,
+      "learning_rate": 8.417053355005428e-05,
+      "loss": 0.0016,
+      "step": 3977
+    },
+    {
+      "epoch": 28.03876651982379,
+      "grad_norm": 0.05722881920110455,
+      "learning_rate": 8.412517132646872e-05,
+      "loss": 0.0023,
+      "step": 3978
+    },
+    {
+      "epoch": 28.045814977973567,
+      "grad_norm": 0.05214196065021283,
+      "learning_rate": 8.407981245369925e-05,
+      "loss": 0.0018,
+      "step": 3979
+    },
+    {
+      "epoch": 28.05286343612335,
+      "grad_norm": 0.03925062989059172,
+      "learning_rate": 8.403445694132006e-05,
+      "loss": 0.0016,
+      "step": 3980
+    },
+    {
+      "epoch": 28.059911894273128,
+      "grad_norm": 0.06296673434575864,
+      "learning_rate": 8.398910479890474e-05,
+      "loss": 0.0021,
+      "step": 3981
+    },
+    {
+      "epoch": 28.066960352422907,
+      "grad_norm": 0.05437667710616035,
+      "learning_rate": 8.394375603602603e-05,
+      "loss": 0.002,
+      "step": 3982
+    },
+    {
+      "epoch": 28.074008810572685,
+      "grad_norm": 0.045915062180974964,
+      "learning_rate": 8.38984106622561e-05,
+      "loss": 0.0018,
+      "step": 3983
+    },
+    {
+      "epoch": 28.081057268722468,
+      "grad_norm": 0.05085032156967527,
+      "learning_rate": 8.385306868716626e-05,
+      "loss": 0.0018,
+      "step": 3984
+    },
+    {
+      "epoch": 28.088105726872246,
+      "grad_norm": 0.05623188137585015,
+      "learning_rate": 8.380773012032719e-05,
+      "loss": 0.002,
+      "step": 3985
+    },
+    {
+      "epoch": 28.095154185022025,
+      "grad_norm": 0.05393191550185485,
+      "learning_rate": 8.376239497130888e-05,
+      "loss": 0.0018,
+      "step": 3986
+    },
+    {
+      "epoch": 28.102202643171807,
+      "grad_norm": 0.05154615622330874,
+      "learning_rate": 8.371706324968049e-05,
+      "loss": 0.0019,
+      "step": 3987
+    },
+    {
+      "epoch": 28.109251101321586,
+      "grad_norm": 0.05864131956442969,
+      "learning_rate": 8.367173496501053e-05,
+      "loss": 0.0019,
+      "step": 3988
+    },
+    {
+      "epoch": 28.116299559471365,
+      "grad_norm": 0.05056188368505456,
+      "learning_rate": 8.362641012686678e-05,
+      "loss": 0.0016,
+      "step": 3989
+    },
+    {
+      "epoch": 28.123348017621144,
+      "grad_norm": 0.0511621716809836,
+      "learning_rate": 8.358108874481634e-05,
+      "loss": 0.0019,
+      "step": 3990
+    },
+    {
+      "epoch": 28.130396475770926,
+      "grad_norm": 0.051450621437951885,
+      "learning_rate": 8.353577082842541e-05,
+      "loss": 0.0018,
+      "step": 3991
+    },
+    {
+      "epoch": 28.137444933920705,
+      "grad_norm": 0.052863144929783716,
+      "learning_rate": 8.349045638725966e-05,
+      "loss": 0.0018,
+      "step": 3992
+    },
+    {
+      "epoch": 28.144493392070483,
+      "grad_norm": 0.04784110002950329,
+      "learning_rate": 8.344514543088394e-05,
+      "loss": 0.0017,
+      "step": 3993
+    },
+    {
+      "epoch": 28.151541850220266,
+      "grad_norm": 0.04551387419491028,
+      "learning_rate": 8.33998379688623e-05,
+      "loss": 0.0017,
+      "step": 3994
+    },
+    {
+      "epoch": 28.158590308370044,
+      "grad_norm": 0.05319851076687466,
+      "learning_rate": 8.335453401075818e-05,
+      "loss": 0.002,
+      "step": 3995
+    },
+    {
+      "epoch": 28.165638766519823,
+      "grad_norm": 0.05155481347011059,
+      "learning_rate": 8.330923356613418e-05,
+      "loss": 0.0019,
+      "step": 3996
+    },
+    {
+      "epoch": 28.172687224669602,
+      "grad_norm": 0.05489686947611893,
+      "learning_rate": 8.326393664455225e-05,
+      "loss": 0.0018,
+      "step": 3997
+    },
+    {
+      "epoch": 28.179735682819384,
+      "grad_norm": 0.05046054025627735,
+      "learning_rate": 8.321864325557348e-05,
+      "loss": 0.0017,
+      "step": 3998
+    },
+    {
+      "epoch": 28.186784140969163,
+      "grad_norm": 0.06196313669662272,
+      "learning_rate": 8.317335340875828e-05,
+      "loss": 0.0019,
+      "step": 3999
+    },
+    {
+      "epoch": 28.19383259911894,
+      "grad_norm": 0.049376312403451646,
+      "learning_rate": 8.312806711366637e-05,
+      "loss": 0.0017,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 250,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.756946984679401e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}