diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7224 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9970788704965923,
+  "eval_steps": 500,
+  "global_step": 1026,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0029211295034079843,
+      "grad_norm": 6.789194746397651,
+      "learning_rate": 9.70873786407767e-08,
+      "loss": 1.0797,
+      "step": 1
+    },
+    {
+      "epoch": 0.005842259006815969,
+      "grad_norm": 6.623744256841628,
+      "learning_rate": 1.941747572815534e-07,
+      "loss": 1.0936,
+      "step": 2
+    },
+    {
+      "epoch": 0.008763388510223954,
+      "grad_norm": 6.814087231706784,
+      "learning_rate": 2.9126213592233014e-07,
+      "loss": 1.0915,
+      "step": 3
+    },
+    {
+      "epoch": 0.011684518013631937,
+      "grad_norm": 6.979864404527602,
+      "learning_rate": 3.883495145631068e-07,
+      "loss": 1.113,
+      "step": 4
+    },
+    {
+      "epoch": 0.014605647517039922,
+      "grad_norm": 7.130506329535252,
+      "learning_rate": 4.854368932038835e-07,
+      "loss": 1.1201,
+      "step": 5
+    },
+    {
+      "epoch": 0.017526777020447908,
+      "grad_norm": 6.457316225305797,
+      "learning_rate": 5.825242718446603e-07,
+      "loss": 1.0589,
+      "step": 6
+    },
+    {
+      "epoch": 0.02044790652385589,
+      "grad_norm": 6.634803833935598,
+      "learning_rate": 6.79611650485437e-07,
+      "loss": 1.0948,
+      "step": 7
+    },
+    {
+      "epoch": 0.023369036027263874,
+      "grad_norm": 6.51924054198785,
+      "learning_rate": 7.766990291262136e-07,
+      "loss": 1.1033,
+      "step": 8
+    },
+    {
+      "epoch": 0.02629016553067186,
+      "grad_norm": 6.172807303107381,
+      "learning_rate": 8.737864077669904e-07,
+      "loss": 1.1025,
+      "step": 9
+    },
+    {
+      "epoch": 0.029211295034079845,
+      "grad_norm": 6.123481829151969,
+      "learning_rate": 9.70873786407767e-07,
+      "loss": 1.0593,
+      "step": 10
+    },
+    {
+      "epoch": 0.03213242453748783,
+      "grad_norm": 5.067066736988861,
+      "learning_rate": 1.0679611650485437e-06,
+      "loss": 1.0514,
+      "step": 11
+    },
+    {
+      "epoch": 0.035053554040895815,
+      "grad_norm": 5.316963931526159,
+      "learning_rate": 1.1650485436893206e-06,
+      "loss": 1.0649,
+      "step": 12
+    },
+    {
+      "epoch": 0.0379746835443038,
+      "grad_norm": 4.8023708760386326,
+      "learning_rate": 1.2621359223300972e-06,
+      "loss": 1.0595,
+      "step": 13
+    },
+    {
+      "epoch": 0.04089581304771178,
+      "grad_norm": 4.307477266760391,
+      "learning_rate": 1.359223300970874e-06,
+      "loss": 1.0045,
+      "step": 14
+    },
+    {
+      "epoch": 0.043816942551119765,
+      "grad_norm": 2.9737946154856254,
+      "learning_rate": 1.4563106796116506e-06,
+      "loss": 0.9864,
+      "step": 15
+    },
+    {
+      "epoch": 0.04673807205452775,
+      "grad_norm": 2.969891758892917,
+      "learning_rate": 1.5533980582524272e-06,
+      "loss": 1.0038,
+      "step": 16
+    },
+    {
+      "epoch": 0.04965920155793573,
+      "grad_norm": 2.6410254985919264,
+      "learning_rate": 1.650485436893204e-06,
+      "loss": 0.9688,
+      "step": 17
+    },
+    {
+      "epoch": 0.05258033106134372,
+      "grad_norm": 2.6767839969522385,
+      "learning_rate": 1.7475728155339808e-06,
+      "loss": 0.9777,
+      "step": 18
+    },
+    {
+      "epoch": 0.055501460564751706,
+      "grad_norm": 2.3845837305117867,
+      "learning_rate": 1.8446601941747574e-06,
+      "loss": 0.9803,
+      "step": 19
+    },
+    {
+      "epoch": 0.05842259006815969,
+      "grad_norm": 2.485668257022799,
+      "learning_rate": 1.941747572815534e-06,
+      "loss": 0.9512,
+      "step": 20
+    },
+    {
+      "epoch": 0.06134371957156767,
+      "grad_norm": 2.9203150216318057,
+      "learning_rate": 2.0388349514563107e-06,
+      "loss": 0.93,
+      "step": 21
+    },
+    {
+      "epoch": 0.06426484907497566,
+      "grad_norm": 2.868722547204036,
+      "learning_rate": 2.1359223300970874e-06,
+      "loss": 0.9171,
+      "step": 22
+    },
+    {
+      "epoch": 0.06718597857838364,
+      "grad_norm": 2.679648431038279,
+      "learning_rate": 2.2330097087378645e-06,
+      "loss": 0.9199,
+      "step": 23
+    },
+    {
+      "epoch": 0.07010710808179163,
+      "grad_norm": 2.459073576199394,
+      "learning_rate": 2.330097087378641e-06,
+      "loss": 0.9157,
+      "step": 24
+    },
+    {
+      "epoch": 0.0730282375851996,
+      "grad_norm": 2.014055908464458,
+      "learning_rate": 2.427184466019418e-06,
+      "loss": 0.9164,
+      "step": 25
+    },
+    {
+      "epoch": 0.0759493670886076,
+      "grad_norm": 1.5955860458303692,
+      "learning_rate": 2.5242718446601945e-06,
+      "loss": 0.8909,
+      "step": 26
+    },
+    {
+      "epoch": 0.07887049659201557,
+      "grad_norm": 1.3790379745407235,
+      "learning_rate": 2.621359223300971e-06,
+      "loss": 0.8631,
+      "step": 27
+    },
+    {
+      "epoch": 0.08179162609542356,
+      "grad_norm": 1.6462939916147095,
+      "learning_rate": 2.718446601941748e-06,
+      "loss": 0.8384,
+      "step": 28
+    },
+    {
+      "epoch": 0.08471275559883155,
+      "grad_norm": 1.7416274692927092,
+      "learning_rate": 2.8155339805825245e-06,
+      "loss": 0.8643,
+      "step": 29
+    },
+    {
+      "epoch": 0.08763388510223953,
+      "grad_norm": 1.4689264445022938,
+      "learning_rate": 2.912621359223301e-06,
+      "loss": 0.8401,
+      "step": 30
+    },
+    {
+      "epoch": 0.09055501460564752,
+      "grad_norm": 1.2970139226424346,
+      "learning_rate": 3.0097087378640778e-06,
+      "loss": 0.8423,
+      "step": 31
+    },
+    {
+      "epoch": 0.0934761441090555,
+      "grad_norm": 1.1674121947058942,
+      "learning_rate": 3.1067961165048544e-06,
+      "loss": 0.837,
+      "step": 32
+    },
+    {
+      "epoch": 0.09639727361246349,
+      "grad_norm": 1.1908632755730892,
+      "learning_rate": 3.2038834951456315e-06,
+      "loss": 0.8203,
+      "step": 33
+    },
+    {
+      "epoch": 0.09931840311587146,
+      "grad_norm": 1.1042928737436872,
+      "learning_rate": 3.300970873786408e-06,
+      "loss": 0.799,
+      "step": 34
+    },
+    {
+      "epoch": 0.10223953261927946,
+      "grad_norm": 1.1603936762022113,
+      "learning_rate": 3.398058252427185e-06,
+      "loss": 0.8303,
+      "step": 35
+    },
+    {
+      "epoch": 0.10516066212268745,
+      "grad_norm": 1.0681312398436076,
+      "learning_rate": 3.4951456310679615e-06,
+      "loss": 0.8052,
+      "step": 36
+    },
+    {
+      "epoch": 0.10808179162609542,
+      "grad_norm": 0.8925175456575719,
+      "learning_rate": 3.592233009708738e-06,
+      "loss": 0.8069,
+      "step": 37
+    },
+    {
+      "epoch": 0.11100292112950341,
+      "grad_norm": 0.8822021824942768,
+      "learning_rate": 3.689320388349515e-06,
+      "loss": 0.7812,
+      "step": 38
+    },
+    {
+      "epoch": 0.11392405063291139,
+      "grad_norm": 0.8935693976115221,
+      "learning_rate": 3.7864077669902915e-06,
+      "loss": 0.7787,
+      "step": 39
+    },
+    {
+      "epoch": 0.11684518013631938,
+      "grad_norm": 0.9243397791705507,
+      "learning_rate": 3.883495145631068e-06,
+      "loss": 0.7936,
+      "step": 40
+    },
+    {
+      "epoch": 0.11976630963972736,
+      "grad_norm": 0.8181188534202998,
+      "learning_rate": 3.980582524271845e-06,
+      "loss": 0.7752,
+      "step": 41
+    },
+    {
+      "epoch": 0.12268743914313535,
+      "grad_norm": 0.723717052367605,
+      "learning_rate": 4.0776699029126215e-06,
+      "loss": 0.7544,
+      "step": 42
+    },
+    {
+      "epoch": 0.12560856864654332,
+      "grad_norm": 0.7268642480625847,
+      "learning_rate": 4.1747572815533986e-06,
+      "loss": 0.7871,
+      "step": 43
+    },
+    {
+      "epoch": 0.12852969814995133,
+      "grad_norm": 0.7154200711509474,
+      "learning_rate": 4.271844660194175e-06,
+      "loss": 0.7617,
+      "step": 44
+    },
+    {
+      "epoch": 0.1314508276533593,
+      "grad_norm": 0.7425189888635864,
+      "learning_rate": 4.368932038834952e-06,
+      "loss": 0.7661,
+      "step": 45
+    },
+    {
+      "epoch": 0.13437195715676728,
+      "grad_norm": 0.6964525874804132,
+      "learning_rate": 4.466019417475729e-06,
+      "loss": 0.7578,
+      "step": 46
+    },
+    {
+      "epoch": 0.13729308666017526,
+      "grad_norm": 0.7077568254698756,
+      "learning_rate": 4.563106796116505e-06,
+      "loss": 0.7466,
+      "step": 47
+    },
+    {
+      "epoch": 0.14021421616358326,
+      "grad_norm": 0.728109048064247,
+      "learning_rate": 4.660194174757282e-06,
+      "loss": 0.7536,
+      "step": 48
+    },
+    {
+      "epoch": 0.14313534566699124,
+      "grad_norm": 0.7073551382013681,
+      "learning_rate": 4.7572815533980585e-06,
+      "loss": 0.7386,
+      "step": 49
+    },
+    {
+      "epoch": 0.1460564751703992,
+      "grad_norm": 0.6810942703152736,
+      "learning_rate": 4.854368932038836e-06,
+      "loss": 0.7771,
+      "step": 50
+    },
+    {
+      "epoch": 0.14897760467380722,
+      "grad_norm": 0.7280450266974076,
+      "learning_rate": 4.951456310679612e-06,
+      "loss": 0.739,
+      "step": 51
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 0.7021483972343961,
+      "learning_rate": 5.048543689320389e-06,
+      "loss": 0.7437,
+      "step": 52
+    },
+    {
+      "epoch": 0.15481986368062317,
+      "grad_norm": 0.7167957676962433,
+      "learning_rate": 5.145631067961165e-06,
+      "loss": 0.764,
+      "step": 53
+    },
+    {
+      "epoch": 0.15774099318403115,
+      "grad_norm": 0.6140023964252928,
+      "learning_rate": 5.242718446601942e-06,
+      "loss": 0.7338,
+      "step": 54
+    },
+    {
+      "epoch": 0.16066212268743915,
+      "grad_norm": 0.7304557287827438,
+      "learning_rate": 5.3398058252427185e-06,
+      "loss": 0.7815,
+      "step": 55
+    },
+    {
+      "epoch": 0.16358325219084713,
+      "grad_norm": 0.6396185255120146,
+      "learning_rate": 5.436893203883496e-06,
+      "loss": 0.7349,
+      "step": 56
+    },
+    {
+      "epoch": 0.1665043816942551,
+      "grad_norm": 0.6558732255969539,
+      "learning_rate": 5.533980582524272e-06,
+      "loss": 0.7212,
+      "step": 57
+    },
+    {
+      "epoch": 0.1694255111976631,
+      "grad_norm": 0.5940933645304805,
+      "learning_rate": 5.631067961165049e-06,
+      "loss": 0.7577,
+      "step": 58
+    },
+    {
+      "epoch": 0.17234664070107109,
+      "grad_norm": 0.6218585913117457,
+      "learning_rate": 5.728155339805825e-06,
+      "loss": 0.7431,
+      "step": 59
+    },
+    {
+      "epoch": 0.17526777020447906,
+      "grad_norm": 0.6908136674947283,
+      "learning_rate": 5.825242718446602e-06,
+      "loss": 0.7368,
+      "step": 60
+    },
+    {
+      "epoch": 0.17818889970788704,
+      "grad_norm": 0.6522752947974526,
+      "learning_rate": 5.9223300970873785e-06,
+      "loss": 0.7126,
+      "step": 61
+    },
+    {
+      "epoch": 0.18111002921129504,
+      "grad_norm": 0.6780346336214896,
+      "learning_rate": 6.0194174757281556e-06,
+      "loss": 0.7367,
+      "step": 62
+    },
+    {
+      "epoch": 0.18403115871470302,
+      "grad_norm": 0.7722253573433183,
+      "learning_rate": 6.116504854368932e-06,
+      "loss": 0.7228,
+      "step": 63
+    },
+    {
+      "epoch": 0.186952288218111,
+      "grad_norm": 0.6399768773236866,
+      "learning_rate": 6.213592233009709e-06,
+      "loss": 0.7244,
+      "step": 64
+    },
+    {
+      "epoch": 0.189873417721519,
+      "grad_norm": 0.6674128228125541,
+      "learning_rate": 6.310679611650487e-06,
+      "loss": 0.7376,
+      "step": 65
+    },
+    {
+      "epoch": 0.19279454722492698,
+      "grad_norm": 0.6775138710637573,
+      "learning_rate": 6.407766990291263e-06,
+      "loss": 0.7286,
+      "step": 66
+    },
+    {
+      "epoch": 0.19571567672833495,
+      "grad_norm": 0.7443299390748634,
+      "learning_rate": 6.50485436893204e-06,
+      "loss": 0.7123,
+      "step": 67
+    },
+    {
+      "epoch": 0.19863680623174293,
+      "grad_norm": 0.6712217876186718,
+      "learning_rate": 6.601941747572816e-06,
+      "loss": 0.7118,
+      "step": 68
+    },
+    {
+      "epoch": 0.20155793573515093,
+      "grad_norm": 0.6173482204736721,
+      "learning_rate": 6.6990291262135935e-06,
+      "loss": 0.7141,
+      "step": 69
+    },
+    {
+      "epoch": 0.2044790652385589,
+      "grad_norm": 0.7586706797857404,
+      "learning_rate": 6.79611650485437e-06,
+      "loss": 0.7182,
+      "step": 70
+    },
+    {
+      "epoch": 0.2074001947419669,
+      "grad_norm": 0.6711328351671544,
+      "learning_rate": 6.893203883495147e-06,
+      "loss": 0.7182,
+      "step": 71
+    },
+    {
+      "epoch": 0.2103213242453749,
+      "grad_norm": 0.6423529972707454,
+      "learning_rate": 6.990291262135923e-06,
+      "loss": 0.7147,
+      "step": 72
+    },
+    {
+      "epoch": 0.21324245374878287,
+      "grad_norm": 0.7175157009799245,
+      "learning_rate": 7.0873786407767e-06,
+      "loss": 0.7078,
+      "step": 73
+    },
+    {
+      "epoch": 0.21616358325219084,
+      "grad_norm": 0.7156449836663106,
+      "learning_rate": 7.184466019417476e-06,
+      "loss": 0.7026,
+      "step": 74
+    },
+    {
+      "epoch": 0.21908471275559882,
+      "grad_norm": 0.591120785534527,
+      "learning_rate": 7.2815533980582534e-06,
+      "loss": 0.6996,
+      "step": 75
+    },
+    {
+      "epoch": 0.22200584225900682,
+      "grad_norm": 0.6568235675952798,
+      "learning_rate": 7.37864077669903e-06,
+      "loss": 0.7098,
+      "step": 76
+    },
+    {
+      "epoch": 0.2249269717624148,
+      "grad_norm": 0.6969907394816692,
+      "learning_rate": 7.475728155339807e-06,
+      "loss": 0.7112,
+      "step": 77
+    },
+    {
+      "epoch": 0.22784810126582278,
+      "grad_norm": 0.674624972312595,
+      "learning_rate": 7.572815533980583e-06,
+      "loss": 0.6905,
+      "step": 78
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6004655479528318,
+      "learning_rate": 7.66990291262136e-06,
+      "loss": 0.6987,
+      "step": 79
+    },
+    {
+      "epoch": 0.23369036027263876,
+      "grad_norm": 0.672439140786889,
+      "learning_rate": 7.766990291262136e-06,
+      "loss": 0.7059,
+      "step": 80
+    },
+    {
+      "epoch": 0.23661148977604674,
+      "grad_norm": 0.6379167803971234,
+      "learning_rate": 7.864077669902913e-06,
+      "loss": 0.6993,
+      "step": 81
+    },
+    {
+      "epoch": 0.2395326192794547,
+      "grad_norm": 0.624979376741808,
+      "learning_rate": 7.96116504854369e-06,
+      "loss": 0.6961,
+      "step": 82
+    },
+    {
+      "epoch": 0.24245374878286272,
+      "grad_norm": 0.5967840653189634,
+      "learning_rate": 8.058252427184466e-06,
+      "loss": 0.7115,
+      "step": 83
+    },
+    {
+      "epoch": 0.2453748782862707,
+      "grad_norm": 0.6448351126797235,
+      "learning_rate": 8.155339805825243e-06,
+      "loss": 0.7058,
+      "step": 84
+    },
+    {
+      "epoch": 0.24829600778967867,
+      "grad_norm": 0.5456246974361787,
+      "learning_rate": 8.25242718446602e-06,
+      "loss": 0.6971,
+      "step": 85
+    },
+    {
+      "epoch": 0.25121713729308665,
+      "grad_norm": 0.5981523190285354,
+      "learning_rate": 8.349514563106797e-06,
+      "loss": 0.6644,
+      "step": 86
+    },
+    {
+      "epoch": 0.25413826679649465,
+      "grad_norm": 0.6094281430676193,
+      "learning_rate": 8.446601941747573e-06,
+      "loss": 0.6822,
+      "step": 87
+    },
+    {
+      "epoch": 0.25705939629990265,
+      "grad_norm": 0.645486530934357,
+      "learning_rate": 8.54368932038835e-06,
+      "loss": 0.6767,
+      "step": 88
+    },
+    {
+      "epoch": 0.2599805258033106,
+      "grad_norm": 0.6225273389721123,
+      "learning_rate": 8.640776699029127e-06,
+      "loss": 0.7152,
+      "step": 89
+    },
+    {
+      "epoch": 0.2629016553067186,
+      "grad_norm": 0.6432259026110182,
+      "learning_rate": 8.737864077669904e-06,
+      "loss": 0.7159,
+      "step": 90
+    },
+    {
+      "epoch": 0.26582278481012656,
+      "grad_norm": 0.6162731782600254,
+      "learning_rate": 8.834951456310681e-06,
+      "loss": 0.7132,
+      "step": 91
+    },
+    {
+      "epoch": 0.26874391431353456,
+      "grad_norm": 0.6542304058964258,
+      "learning_rate": 8.932038834951458e-06,
+      "loss": 0.6809,
+      "step": 92
+    },
+    {
+      "epoch": 0.27166504381694256,
+      "grad_norm": 0.6079172030969984,
+      "learning_rate": 9.029126213592233e-06,
+      "loss": 0.6824,
+      "step": 93
+    },
+    {
+      "epoch": 0.2745861733203505,
+      "grad_norm": 0.6519383743853913,
+      "learning_rate": 9.12621359223301e-06,
+      "loss": 0.6899,
+      "step": 94
+    },
+    {
+      "epoch": 0.2775073028237585,
+      "grad_norm": 0.7282902638094394,
+      "learning_rate": 9.223300970873788e-06,
+      "loss": 0.6922,
+      "step": 95
+    },
+    {
+      "epoch": 0.2804284323271665,
+      "grad_norm": 0.5989786182176935,
+      "learning_rate": 9.320388349514565e-06,
+      "loss": 0.6734,
+      "step": 96
+    },
+    {
+      "epoch": 0.28334956183057447,
+      "grad_norm": 0.7160709546948157,
+      "learning_rate": 9.41747572815534e-06,
+      "loss": 0.6815,
+      "step": 97
+    },
+    {
+      "epoch": 0.2862706913339825,
+      "grad_norm": 0.6813096412081009,
+      "learning_rate": 9.514563106796117e-06,
+      "loss": 0.6885,
+      "step": 98
+    },
+    {
+      "epoch": 0.2891918208373905,
+      "grad_norm": 0.6540975722149734,
+      "learning_rate": 9.611650485436894e-06,
+      "loss": 0.69,
+      "step": 99
+    },
+    {
+      "epoch": 0.2921129503407984,
+      "grad_norm": 0.8050571281257926,
+      "learning_rate": 9.708737864077671e-06,
+      "loss": 0.691,
+      "step": 100
+    },
+    {
+      "epoch": 0.29503407984420643,
+      "grad_norm": 0.6433559989032654,
+      "learning_rate": 9.805825242718447e-06,
+      "loss": 0.6986,
+      "step": 101
+    },
+    {
+      "epoch": 0.29795520934761444,
+      "grad_norm": 0.8412736504392738,
+      "learning_rate": 9.902912621359224e-06,
+      "loss": 0.6926,
+      "step": 102
+    },
+    {
+      "epoch": 0.3008763388510224,
+      "grad_norm": 0.7443726086923518,
+      "learning_rate": 1e-05,
+      "loss": 0.6906,
+      "step": 103
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.6597015006842325,
+      "learning_rate": 9.999971037507608e-06,
+      "loss": 0.675,
+      "step": 104
+    },
+    {
+      "epoch": 0.30671859785783834,
+      "grad_norm": 0.5931435595661035,
+      "learning_rate": 9.99988415036596e-06,
+      "loss": 0.6802,
+      "step": 105
+    },
+    {
+      "epoch": 0.30963972736124634,
+      "grad_norm": 0.7214217523040783,
+      "learning_rate": 9.99973933958164e-06,
+      "loss": 0.7041,
+      "step": 106
+    },
+    {
+      "epoch": 0.31256085686465435,
+      "grad_norm": 0.7234513096207073,
+      "learning_rate": 9.999536606832288e-06,
+      "loss": 0.6872,
+      "step": 107
+    },
+    {
+      "epoch": 0.3154819863680623,
+      "grad_norm": 0.7879752038918911,
+      "learning_rate": 9.999275954466555e-06,
+      "loss": 0.6873,
+      "step": 108
+    },
+    {
+      "epoch": 0.3184031158714703,
+      "grad_norm": 0.6655845938433153,
+      "learning_rate": 9.998957385504103e-06,
+      "loss": 0.6976,
+      "step": 109
+    },
+    {
+      "epoch": 0.3213242453748783,
+      "grad_norm": 0.8522730059744493,
+      "learning_rate": 9.99858090363555e-06,
+      "loss": 0.6719,
+      "step": 110
+    },
+    {
+      "epoch": 0.32424537487828625,
+      "grad_norm": 0.6292219731062122,
+      "learning_rate": 9.998146513222436e-06,
+      "loss": 0.6993,
+      "step": 111
+    },
+    {
+      "epoch": 0.32716650438169426,
+      "grad_norm": 0.7181176135878521,
+      "learning_rate": 9.997654219297176e-06,
+      "loss": 0.6901,
+      "step": 112
+    },
+    {
+      "epoch": 0.33008763388510226,
+      "grad_norm": 0.6962928810640735,
+      "learning_rate": 9.997104027562991e-06,
+      "loss": 0.6951,
+      "step": 113
+    },
+    {
+      "epoch": 0.3330087633885102,
+      "grad_norm": 0.6849230790401417,
+      "learning_rate": 9.996495944393853e-06,
+      "loss": 0.6828,
+      "step": 114
+    },
+    {
+      "epoch": 0.3359298928919182,
+      "grad_norm": 0.7135171270031142,
+      "learning_rate": 9.995829976834402e-06,
+      "loss": 0.6737,
+      "step": 115
+    },
+    {
+      "epoch": 0.3388510223953262,
+      "grad_norm": 0.6814570642632325,
+      "learning_rate": 9.995106132599869e-06,
+      "loss": 0.6875,
+      "step": 116
+    },
+    {
+      "epoch": 0.34177215189873417,
+      "grad_norm": 0.8248995841087691,
+      "learning_rate": 9.99432442007599e-06,
+      "loss": 0.6873,
+      "step": 117
+    },
+    {
+      "epoch": 0.34469328140214217,
+      "grad_norm": 0.6919031766051941,
+      "learning_rate": 9.993484848318899e-06,
+      "loss": 0.6835,
+      "step": 118
+    },
+    {
+      "epoch": 0.3476144109055501,
+      "grad_norm": 0.7748697122331325,
+      "learning_rate": 9.992587427055036e-06,
+      "loss": 0.6933,
+      "step": 119
+    },
+    {
+      "epoch": 0.3505355404089581,
+      "grad_norm": 0.6232356121853884,
+      "learning_rate": 9.99163216668102e-06,
+      "loss": 0.6634,
+      "step": 120
+    },
+    {
+      "epoch": 0.35345666991236613,
+      "grad_norm": 0.8152682417581196,
+      "learning_rate": 9.990619078263543e-06,
+      "loss": 0.6833,
+      "step": 121
+    },
+    {
+      "epoch": 0.3563777994157741,
+      "grad_norm": 0.6829212246748637,
+      "learning_rate": 9.989548173539229e-06,
+      "loss": 0.6904,
+      "step": 122
+    },
+    {
+      "epoch": 0.3592989289191821,
+      "grad_norm": 0.7764702428398512,
+      "learning_rate": 9.988419464914505e-06,
+      "loss": 0.6911,
+      "step": 123
+    },
+    {
+      "epoch": 0.3622200584225901,
+      "grad_norm": 0.792332480824063,
+      "learning_rate": 9.98723296546546e-06,
+      "loss": 0.6817,
+      "step": 124
+    },
+    {
+      "epoch": 0.36514118792599803,
+      "grad_norm": 0.6594623726005864,
+      "learning_rate": 9.985988688937684e-06,
+      "loss": 0.6873,
+      "step": 125
+    },
+    {
+      "epoch": 0.36806231742940604,
+      "grad_norm": 0.8269823477396988,
+      "learning_rate": 9.984686649746119e-06,
+      "loss": 0.693,
+      "step": 126
+    },
+    {
+      "epoch": 0.37098344693281404,
+      "grad_norm": 0.6422126697095933,
+      "learning_rate": 9.983326862974882e-06,
+      "loss": 0.6576,
+      "step": 127
+    },
+    {
+      "epoch": 0.373904576436222,
+      "grad_norm": 0.8200568586438982,
+      "learning_rate": 9.981909344377101e-06,
+      "loss": 0.6929,
+      "step": 128
+    },
+    {
+      "epoch": 0.37682570593963,
+      "grad_norm": 0.7400126500706381,
+      "learning_rate": 9.980434110374725e-06,
+      "loss": 0.6557,
+      "step": 129
+    },
+    {
+      "epoch": 0.379746835443038,
+      "grad_norm": 0.713854878991036,
+      "learning_rate": 9.978901178058333e-06,
+      "loss": 0.6942,
+      "step": 130
+    },
+    {
+      "epoch": 0.38266796494644595,
+      "grad_norm": 0.7483541113835968,
+      "learning_rate": 9.977310565186945e-06,
+      "loss": 0.6781,
+      "step": 131
+    },
+    {
+      "epoch": 0.38558909444985395,
+      "grad_norm": 0.7131870241688308,
+      "learning_rate": 9.975662290187802e-06,
+      "loss": 0.6564,
+      "step": 132
+    },
+    {
+      "epoch": 0.3885102239532619,
+      "grad_norm": 0.7880618680074065,
+      "learning_rate": 9.973956372156166e-06,
+      "loss": 0.6752,
+      "step": 133
+    },
+    {
+      "epoch": 0.3914313534566699,
+      "grad_norm": 0.5977156256261835,
+      "learning_rate": 9.972192830855095e-06,
+      "loss": 0.6763,
+      "step": 134
+    },
+    {
+      "epoch": 0.3943524829600779,
+      "grad_norm": 0.76186235107972,
+      "learning_rate": 9.970371686715205e-06,
+      "loss": 0.7014,
+      "step": 135
+    },
+    {
+      "epoch": 0.39727361246348586,
+      "grad_norm": 0.8438134698760479,
+      "learning_rate": 9.96849296083445e-06,
+      "loss": 0.6902,
+      "step": 136
+    },
+    {
+      "epoch": 0.40019474196689386,
+      "grad_norm": 0.6120514167187466,
+      "learning_rate": 9.966556674977864e-06,
+      "loss": 0.663,
+      "step": 137
+    },
+    {
+      "epoch": 0.40311587147030187,
+      "grad_norm": 0.7891889540366791,
+      "learning_rate": 9.964562851577307e-06,
+      "loss": 0.6458,
+      "step": 138
+    },
+    {
+      "epoch": 0.4060370009737098,
+      "grad_norm": 0.8381022052177086,
+      "learning_rate": 9.962511513731219e-06,
+      "loss": 0.6728,
+      "step": 139
+    },
+    {
+      "epoch": 0.4089581304771178,
+      "grad_norm": 0.7489332182994131,
+      "learning_rate": 9.960402685204347e-06,
+      "loss": 0.6585,
+      "step": 140
+    },
+    {
+      "epoch": 0.4118792599805258,
+      "grad_norm": 0.804869829411989,
+      "learning_rate": 9.958236390427458e-06,
+      "loss": 0.6784,
+      "step": 141
+    },
+    {
+      "epoch": 0.4148003894839338,
+      "grad_norm": 0.7474713341428297,
+      "learning_rate": 9.956012654497073e-06,
+      "loss": 0.6488,
+      "step": 142
+    },
+    {
+      "epoch": 0.4177215189873418,
+      "grad_norm": 0.8115239389486957,
+      "learning_rate": 9.953731503175166e-06,
+      "loss": 0.6894,
+      "step": 143
+    },
+    {
+      "epoch": 0.4206426484907498,
+      "grad_norm": 0.781168520393115,
+      "learning_rate": 9.951392962888868e-06,
+      "loss": 0.6534,
+      "step": 144
+    },
+    {
+      "epoch": 0.42356377799415773,
+      "grad_norm": 0.7722795439014734,
+      "learning_rate": 9.948997060730161e-06,
+      "loss": 0.6504,
+      "step": 145
+    },
+    {
+      "epoch": 0.42648490749756574,
+      "grad_norm": 0.8668435801796398,
+      "learning_rate": 9.946543824455563e-06,
+      "loss": 0.6507,
+      "step": 146
+    },
+    {
+      "epoch": 0.4294060370009737,
+      "grad_norm": 0.8391877814865175,
+      "learning_rate": 9.94403328248581e-06,
+      "loss": 0.6702,
+      "step": 147
+    },
+    {
+      "epoch": 0.4323271665043817,
+      "grad_norm": 0.6805575537389376,
+      "learning_rate": 9.941465463905522e-06,
+      "loss": 0.6744,
+      "step": 148
+    },
+    {
+      "epoch": 0.4352482960077897,
+      "grad_norm": 0.7033559759136386,
+      "learning_rate": 9.938840398462872e-06,
+      "loss": 0.6732,
+      "step": 149
+    },
+    {
+      "epoch": 0.43816942551119764,
+      "grad_norm": 0.8071774421297472,
+      "learning_rate": 9.936158116569231e-06,
+      "loss": 0.6704,
+      "step": 150
+    },
+    {
+      "epoch": 0.44109055501460565,
+      "grad_norm": 0.7788309365778397,
+      "learning_rate": 9.933418649298831e-06,
+      "loss": 0.6697,
+      "step": 151
+    },
+    {
+      "epoch": 0.44401168451801365,
+      "grad_norm": 0.633484330960411,
+      "learning_rate": 9.930622028388388e-06,
+      "loss": 0.6533,
+      "step": 152
+    },
+    {
+      "epoch": 0.4469328140214216,
+      "grad_norm": 0.6677671353777382,
+      "learning_rate": 9.92776828623675e-06,
+      "loss": 0.6323,
+      "step": 153
+    },
+    {
+      "epoch": 0.4498539435248296,
+      "grad_norm": 0.6580113405415786,
+      "learning_rate": 9.924857455904511e-06,
+      "loss": 0.6569,
+      "step": 154
+    },
+    {
+      "epoch": 0.4527750730282376,
+      "grad_norm": 0.7153967279296722,
+      "learning_rate": 9.921889571113629e-06,
+      "loss": 0.6651,
+      "step": 155
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 0.6507727168616313,
+      "learning_rate": 9.918864666247042e-06,
+      "loss": 0.6709,
+      "step": 156
+    },
+    {
+      "epoch": 0.45861733203505356,
+      "grad_norm": 0.6614595067208825,
+      "learning_rate": 9.915782776348263e-06,
+      "loss": 0.6558,
+      "step": 157
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.8137858393206525,
+      "learning_rate": 9.912643937120978e-06,
+      "loss": 0.6756,
+      "step": 158
+    },
+    {
+      "epoch": 0.4644595910418695,
+      "grad_norm": 0.6306212983973334,
+      "learning_rate": 9.909448184928629e-06,
+      "loss": 0.6416,
+      "step": 159
+    },
+    {
+      "epoch": 0.4673807205452775,
+      "grad_norm": 0.6836208681318199,
+      "learning_rate": 9.906195556793996e-06,
+      "loss": 0.6662,
+      "step": 160
+    },
+    {
+      "epoch": 0.47030185004868547,
+      "grad_norm": 0.6141476817991299,
+      "learning_rate": 9.902886090398764e-06,
+      "loss": 0.6774,
+      "step": 161
+    },
+    {
+      "epoch": 0.47322297955209347,
+      "grad_norm": 0.7052661650958192,
+      "learning_rate": 9.899519824083095e-06,
+      "loss": 0.6618,
+      "step": 162
+    },
+    {
+      "epoch": 0.4761441090555015,
+      "grad_norm": 0.5986401947237227,
+      "learning_rate": 9.896096796845172e-06,
+      "loss": 0.6738,
+      "step": 163
+    },
+    {
+      "epoch": 0.4790652385589094,
+      "grad_norm": 0.6916970139226505,
+      "learning_rate": 9.892617048340754e-06,
+      "loss": 0.6588,
+      "step": 164
+    },
+    {
+      "epoch": 0.4819863680623174,
+      "grad_norm": 0.6805676670949773,
+      "learning_rate": 9.889080618882719e-06,
+      "loss": 0.6826,
+      "step": 165
+    },
+    {
+      "epoch": 0.48490749756572543,
+      "grad_norm": 0.6658431348883423,
+      "learning_rate": 9.88548754944059e-06,
+      "loss": 0.6702,
+      "step": 166
+    },
+    {
+      "epoch": 0.4878286270691334,
+      "grad_norm": 0.682259295410329,
+      "learning_rate": 9.881837881640064e-06,
+      "loss": 0.6735,
+      "step": 167
+    },
+    {
+      "epoch": 0.4907497565725414,
+      "grad_norm": 0.5993509164653038,
+      "learning_rate": 9.878131657762535e-06,
+      "loss": 0.6545,
+      "step": 168
+    },
+    {
+      "epoch": 0.4936708860759494,
+      "grad_norm": 0.7534948952289838,
+      "learning_rate": 9.874368920744594e-06,
+      "loss": 0.6812,
+      "step": 169
+    },
+    {
+      "epoch": 0.49659201557935734,
+      "grad_norm": 0.707512873305316,
+      "learning_rate": 9.870549714177538e-06,
+      "loss": 0.6513,
+      "step": 170
+    },
+    {
+      "epoch": 0.49951314508276534,
+      "grad_norm": 0.7200355139629895,
+      "learning_rate": 9.866674082306861e-06,
+      "loss": 0.6438,
+      "step": 171
+    },
+    {
+      "epoch": 0.5024342745861733,
+      "grad_norm": 0.8156295486958224,
+      "learning_rate": 9.86274207003175e-06,
+      "loss": 0.6564,
+      "step": 172
+    },
+    {
+      "epoch": 0.5053554040895814,
+      "grad_norm": 0.6117511012510686,
+      "learning_rate": 9.858753722904552e-06,
+      "loss": 0.6827,
+      "step": 173
+    },
+    {
+      "epoch": 0.5082765335929893,
+      "grad_norm": 0.8219493047877929,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.6718,
+      "step": 174
+    },
+    {
+      "epoch": 0.5111976630963972,
+      "grad_norm": 0.6662487126424073,
+      "learning_rate": 9.850608209565967e-06,
+      "loss": 0.6388,
+      "step": 175
+    },
+    {
+      "epoch": 0.5141187925998053,
+      "grad_norm": 0.7125438710540766,
+      "learning_rate": 9.84645113772032e-06,
+      "loss": 0.6589,
+      "step": 176
+    },
+    {
+      "epoch": 0.5170399221032133,
+      "grad_norm": 0.7487609228616714,
+      "learning_rate": 9.842237919752994e-06,
+      "loss": 0.6544,
+      "step": 177
+    },
+    {
+      "epoch": 0.5199610516066212,
+      "grad_norm": 0.6468420309416394,
+      "learning_rate": 9.8379686044741e-06,
+      "loss": 0.6565,
+      "step": 178
+    },
+    {
+      "epoch": 0.5228821811100293,
+      "grad_norm": 0.7546649500098744,
+      "learning_rate": 9.833643241343642e-06,
+      "loss": 0.6647,
+      "step": 179
+    },
+    {
+      "epoch": 0.5258033106134372,
+      "grad_norm": 0.6258186572488958,
+      "learning_rate": 9.829261880470941e-06,
+      "loss": 0.6392,
+      "step": 180
+    },
+    {
+      "epoch": 0.5287244401168452,
+      "grad_norm": 0.5997447293335689,
+      "learning_rate": 9.82482457261405e-06,
+      "loss": 0.6398,
+      "step": 181
+    },
+    {
+      "epoch": 0.5316455696202531,
+      "grad_norm": 0.6452320212378018,
+      "learning_rate": 9.820331369179166e-06,
+      "loss": 0.6611,
+      "step": 182
+    },
+    {
+      "epoch": 0.5345666991236612,
+      "grad_norm": 0.5580735581912285,
+      "learning_rate": 9.815782322220036e-06,
+      "loss": 0.6548,
+      "step": 183
+    },
+    {
+      "epoch": 0.5374878286270691,
+      "grad_norm": 0.6058080142971995,
+      "learning_rate": 9.811177484437357e-06,
+      "loss": 0.6664,
+      "step": 184
+    },
+    {
+      "epoch": 0.5404089581304771,
+      "grad_norm": 0.7487147947448509,
+      "learning_rate": 9.806516909178161e-06,
+      "loss": 0.665,
+      "step": 185
+    },
+    {
+      "epoch": 0.5433300876338851,
+      "grad_norm": 0.5994532711002538,
+      "learning_rate": 9.801800650435194e-06,
+      "loss": 0.6345,
+      "step": 186
+    },
+    {
+      "epoch": 0.5462512171372931,
+      "grad_norm": 0.6402638834025774,
+      "learning_rate": 9.797028762846305e-06,
+      "loss": 0.6689,
+      "step": 187
+    },
+    {
+      "epoch": 0.549172346640701,
+      "grad_norm": 0.8110120277002857,
+      "learning_rate": 9.792201301693793e-06,
+      "loss": 0.6623,
+      "step": 188
+    },
+    {
+      "epoch": 0.5520934761441091,
+      "grad_norm": 0.6022982451173915,
+      "learning_rate": 9.787318322903784e-06,
+      "loss": 0.642,
+      "step": 189
+    },
+    {
+      "epoch": 0.555014605647517,
+      "grad_norm": 0.8359964918822578,
+      "learning_rate": 9.78237988304557e-06,
+      "loss": 0.6828,
+      "step": 190
+    },
+    {
+      "epoch": 0.557935735150925,
+      "grad_norm": 0.5971974835914099,
+      "learning_rate": 9.77738603933096e-06,
+      "loss": 0.6637,
+      "step": 191
+    },
+    {
+      "epoch": 0.560856864654333,
+      "grad_norm": 0.9007553762322157,
+      "learning_rate": 9.772336849613624e-06,
+      "loss": 0.6489,
+      "step": 192
+    },
+    {
+      "epoch": 0.563777994157741,
+      "grad_norm": 0.6097333173160772,
+      "learning_rate": 9.767232372388406e-06,
+      "loss": 0.6195,
+      "step": 193
+    },
+    {
+      "epoch": 0.5666991236611489,
+      "grad_norm": 0.7430323349181741,
+      "learning_rate": 9.762072666790658e-06,
+      "loss": 0.6602,
+      "step": 194
+    },
+    {
+      "epoch": 0.569620253164557,
+      "grad_norm": 0.698590106062137,
+      "learning_rate": 9.756857792595555e-06,
+      "loss": 0.654,
+      "step": 195
+    },
+    {
+      "epoch": 0.572541382667965,
+      "grad_norm": 0.6098458915248055,
+      "learning_rate": 9.751587810217398e-06,
+      "loss": 0.6571,
+      "step": 196
+    },
+    {
+      "epoch": 0.5754625121713729,
+      "grad_norm": 0.6600018141821303,
+      "learning_rate": 9.746262780708919e-06,
+      "loss": 0.6572,
+      "step": 197
+    },
+    {
+      "epoch": 0.578383641674781,
+      "grad_norm": 0.6601639946678165,
+      "learning_rate": 9.740882765760567e-06,
+      "loss": 0.6593,
+      "step": 198
+    },
+    {
+      "epoch": 0.5813047711781889,
+      "grad_norm": 0.726967683938266,
+      "learning_rate": 9.735447827699798e-06,
+      "loss": 0.6573,
+      "step": 199
+    },
+    {
+      "epoch": 0.5842259006815969,
+      "grad_norm": 0.6076134837821863,
+      "learning_rate": 9.729958029490353e-06,
+      "loss": 0.6495,
+      "step": 200
+    },
+    {
+      "epoch": 0.5871470301850049,
+      "grad_norm": 0.693728259825805,
+      "learning_rate": 9.72441343473153e-06,
+      "loss": 0.6384,
+      "step": 201
+    },
+    {
+      "epoch": 0.5900681596884129,
+      "grad_norm": 0.7278423168034551,
+      "learning_rate": 9.718814107657441e-06,
+      "loss": 0.6584,
+      "step": 202
+    },
+    {
+      "epoch": 0.5929892891918208,
+      "grad_norm": 0.7100671054561837,
+      "learning_rate": 9.713160113136272e-06,
+      "loss": 0.6555,
+      "step": 203
+    },
+    {
+      "epoch": 0.5959104186952289,
+      "grad_norm": 0.6955835624438068,
+      "learning_rate": 9.707451516669533e-06,
+      "loss": 0.6581,
+      "step": 204
+    },
+    {
+      "epoch": 0.5988315481986368,
+      "grad_norm": 0.6862859891275203,
+      "learning_rate": 9.701688384391296e-06,
+      "loss": 0.6471,
+      "step": 205
+    },
+    {
+      "epoch": 0.6017526777020448,
+      "grad_norm": 0.7918106833642026,
+      "learning_rate": 9.695870783067434e-06,
+      "loss": 0.6351,
+      "step": 206
+    },
+    {
+      "epoch": 0.6046738072054528,
+      "grad_norm": 0.762255183423834,
+      "learning_rate": 9.689998780094839e-06,
+      "loss": 0.6464,
+      "step": 207
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.5926349601655899,
+      "learning_rate": 9.684072443500645e-06,
+      "loss": 0.6342,
+      "step": 208
+    },
+    {
+      "epoch": 0.6105160662122687,
+      "grad_norm": 0.863234457455766,
+      "learning_rate": 9.678091841941446e-06,
+      "loss": 0.653,
+      "step": 209
+    },
+    {
+      "epoch": 0.6134371957156767,
+      "grad_norm": 0.7588656251837851,
+      "learning_rate": 9.672057044702492e-06,
+      "loss": 0.6379,
+      "step": 210
+    },
+    {
+      "epoch": 0.6163583252190847,
+      "grad_norm": 0.6108635991637165,
+      "learning_rate": 9.665968121696892e-06,
+      "loss": 0.6605,
+      "step": 211
+    },
+    {
+      "epoch": 0.6192794547224927,
+      "grad_norm": 0.8356230484629192,
+      "learning_rate": 9.659825143464798e-06,
+      "loss": 0.6458,
+      "step": 212
+    },
+    {
+      "epoch": 0.6222005842259006,
+      "grad_norm": 0.6240730332192024,
+      "learning_rate": 9.653628181172596e-06,
+      "loss": 0.6506,
+      "step": 213
+    },
+    {
+      "epoch": 0.6251217137293087,
+      "grad_norm": 0.6689297135107584,
+      "learning_rate": 9.647377306612075e-06,
+      "loss": 0.6299,
+      "step": 214
+    },
+    {
+      "epoch": 0.6280428432327166,
+      "grad_norm": 0.7685374427252067,
+      "learning_rate": 9.641072592199599e-06,
+      "loss": 0.6634,
+      "step": 215
+    },
+    {
+      "epoch": 0.6309639727361246,
+      "grad_norm": 0.6331465350705314,
+      "learning_rate": 9.634714110975263e-06,
+      "loss": 0.6705,
+      "step": 216
+    },
+    {
+      "epoch": 0.6338851022395326,
+      "grad_norm": 0.7142109675799595,
+      "learning_rate": 9.628301936602053e-06,
+      "loss": 0.6539,
+      "step": 217
+    },
+    {
+      "epoch": 0.6368062317429406,
+      "grad_norm": 0.8729837863809322,
+      "learning_rate": 9.62183614336499e-06,
+      "loss": 0.6596,
+      "step": 218
+    },
+    {
+      "epoch": 0.6397273612463485,
+      "grad_norm": 0.5991176634004923,
+      "learning_rate": 9.61531680617027e-06,
+      "loss": 0.6656,
+      "step": 219
+    },
+    {
+      "epoch": 0.6426484907497566,
+      "grad_norm": 0.7145894148388473,
+      "learning_rate": 9.608744000544392e-06,
+      "loss": 0.6643,
+      "step": 220
+    },
+    {
+      "epoch": 0.6455696202531646,
+      "grad_norm": 0.7598912110336243,
+      "learning_rate": 9.602117802633293e-06,
+      "loss": 0.6291,
+      "step": 221
+    },
+    {
+      "epoch": 0.6484907497565725,
+      "grad_norm": 0.6815208811474045,
+      "learning_rate": 9.595438289201453e-06,
+      "loss": 0.6472,
+      "step": 222
+    },
+    {
+      "epoch": 0.6514118792599806,
+      "grad_norm": 0.5760568859368258,
+      "learning_rate": 9.588705537631014e-06,
+      "loss": 0.6563,
+      "step": 223
+    },
+    {
+      "epoch": 0.6543330087633885,
+      "grad_norm": 0.7022056147869815,
+      "learning_rate": 9.581919625920886e-06,
+      "loss": 0.6524,
+      "step": 224
+    },
+    {
+      "epoch": 0.6572541382667965,
+      "grad_norm": 0.67470404537858,
+      "learning_rate": 9.575080632685832e-06,
+      "loss": 0.6436,
+      "step": 225
+    },
+    {
+      "epoch": 0.6601752677702045,
+      "grad_norm": 0.6156654623500659,
+      "learning_rate": 9.568188637155569e-06,
+      "loss": 0.6256,
+      "step": 226
+    },
+    {
+      "epoch": 0.6630963972736125,
+      "grad_norm": 0.6870167927139845,
+      "learning_rate": 9.561243719173844e-06,
+      "loss": 0.628,
+      "step": 227
+    },
+    {
+      "epoch": 0.6660175267770204,
+      "grad_norm": 0.6043043393160271,
+      "learning_rate": 9.554245959197511e-06,
+      "loss": 0.6631,
+      "step": 228
+    },
+    {
+      "epoch": 0.6689386562804285,
+      "grad_norm": 0.6424377779531785,
+      "learning_rate": 9.5471954382956e-06,
+      "loss": 0.6455,
+      "step": 229
+    },
+    {
+      "epoch": 0.6718597857838364,
+      "grad_norm": 0.6747246252989533,
+      "learning_rate": 9.54009223814837e-06,
+      "loss": 0.6482,
+      "step": 230
+    },
+    {
+      "epoch": 0.6747809152872444,
+      "grad_norm": 0.6198749042634925,
+      "learning_rate": 9.532936441046376e-06,
+      "loss": 0.6679,
+      "step": 231
+    },
+    {
+      "epoch": 0.6777020447906524,
+      "grad_norm": 0.6700871145900489,
+      "learning_rate": 9.525728129889505e-06,
+      "loss": 0.6704,
+      "step": 232
+    },
+    {
+      "epoch": 0.6806231742940604,
+      "grad_norm": 0.6486221862584836,
+      "learning_rate": 9.51846738818602e-06,
+      "loss": 0.6533,
+      "step": 233
+    },
+    {
+      "epoch": 0.6835443037974683,
+      "grad_norm": 0.6158430372621209,
+      "learning_rate": 9.511154300051591e-06,
+      "loss": 0.6391,
+      "step": 234
+    },
+    {
+      "epoch": 0.6864654333008764,
+      "grad_norm": 0.6059458334544432,
+      "learning_rate": 9.503788950208324e-06,
+      "loss": 0.6326,
+      "step": 235
+    },
+    {
+      "epoch": 0.6893865628042843,
+      "grad_norm": 0.6630441105155737,
+      "learning_rate": 9.49637142398377e-06,
+      "loss": 0.6419,
+      "step": 236
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.7121010143311686,
+      "learning_rate": 9.48890180730995e-06,
+      "loss": 0.6366,
+      "step": 237
+    },
+    {
+      "epoch": 0.6952288218111002,
+      "grad_norm": 0.6536311654933115,
+      "learning_rate": 9.481380186722354e-06,
+      "loss": 0.6475,
+      "step": 238
+    },
+    {
+      "epoch": 0.6981499513145083,
+      "grad_norm": 0.5870586837084283,
+      "learning_rate": 9.473806649358929e-06,
+      "loss": 0.6664,
+      "step": 239
+    },
+    {
+      "epoch": 0.7010710808179162,
+      "grad_norm": 0.7293444660664181,
+      "learning_rate": 9.466181282959083e-06,
+      "loss": 0.6294,
+      "step": 240
+    },
+    {
+      "epoch": 0.7039922103213242,
+      "grad_norm": 0.6803091119725557,
+      "learning_rate": 9.458504175862665e-06,
+      "loss": 0.6543,
+      "step": 241
+    },
+    {
+      "epoch": 0.7069133398247323,
+      "grad_norm": 0.513109460376802,
+      "learning_rate": 9.450775417008936e-06,
+      "loss": 0.6529,
+      "step": 242
+    },
+    {
+      "epoch": 0.7098344693281402,
+      "grad_norm": 0.6591044352211995,
+      "learning_rate": 9.442995095935542e-06,
+      "loss": 0.6485,
+      "step": 243
+    },
+    {
+      "epoch": 0.7127555988315482,
+      "grad_norm": 0.5639394652214005,
+      "learning_rate": 9.43516330277748e-06,
+      "loss": 0.6354,
+      "step": 244
+    },
+    {
+      "epoch": 0.7156767283349562,
+      "grad_norm": 0.5382276491132706,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.6338,
+      "step": 245
+    },
+    {
+      "epoch": 0.7185978578383642,
+      "grad_norm": 0.5783621915913141,
+      "learning_rate": 9.419345663727805e-06,
+      "loss": 0.6541,
+      "step": 246
+    },
+    {
+      "epoch": 0.7215189873417721,
+      "grad_norm": 0.5457758477722148,
+      "learning_rate": 9.411360001083496e-06,
+      "loss": 0.6649,
+      "step": 247
+    },
+    {
+      "epoch": 0.7244401168451802,
+      "grad_norm": 0.5701118395223765,
+      "learning_rate": 9.403323232846994e-06,
+      "loss": 0.6305,
+      "step": 248
+    },
+    {
+      "epoch": 0.7273612463485881,
+      "grad_norm": 0.5986045250901076,
+      "learning_rate": 9.395235452124239e-06,
+      "loss": 0.6315,
+      "step": 249
+    },
+    {
+      "epoch": 0.7302823758519961,
+      "grad_norm": 0.5915757822980239,
+      "learning_rate": 9.387096752612144e-06,
+      "loss": 0.6563,
+      "step": 250
+    },
+    {
+      "epoch": 0.7332035053554041,
+      "grad_norm": 0.6447044009504002,
+      "learning_rate": 9.378907228597518e-06,
+      "loss": 0.6543,
+      "step": 251
+    },
+    {
+      "epoch": 0.7361246348588121,
+      "grad_norm": 0.6146190085950654,
+      "learning_rate": 9.370666974955973e-06,
+      "loss": 0.6474,
+      "step": 252
+    },
+    {
+      "epoch": 0.73904576436222,
+      "grad_norm": 0.6186340229955254,
+      "learning_rate": 9.362376087150822e-06,
+      "loss": 0.6498,
+      "step": 253
+    },
+    {
+      "epoch": 0.7419668938656281,
+      "grad_norm": 0.6697188050730257,
+      "learning_rate": 9.354034661231976e-06,
+      "loss": 0.629,
+      "step": 254
+    },
+    {
+      "epoch": 0.744888023369036,
+      "grad_norm": 0.7166126450253048,
+      "learning_rate": 9.345642793834825e-06,
+      "loss": 0.6476,
+      "step": 255
+    },
+    {
+      "epoch": 0.747809152872444,
+      "grad_norm": 0.5909733136537622,
+      "learning_rate": 9.337200582179134e-06,
+      "loss": 0.6338,
+      "step": 256
+    },
+    {
+      "epoch": 0.750730282375852,
+      "grad_norm": 0.6690338784895201,
+      "learning_rate": 9.328708124067893e-06,
+      "loss": 0.6425,
+      "step": 257
+    },
+    {
+      "epoch": 0.75365141187926,
+      "grad_norm": 0.6170552061635086,
+      "learning_rate": 9.320165517886207e-06,
+      "loss": 0.649,
+      "step": 258
+    },
+    {
+      "epoch": 0.7565725413826679,
+      "grad_norm": 0.6767600272762853,
+      "learning_rate": 9.31157286260014e-06,
+      "loss": 0.6496,
+      "step": 259
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 0.7264560653599718,
+      "learning_rate": 9.302930257755579e-06,
+      "loss": 0.6583,
+      "step": 260
+    },
+    {
+      "epoch": 0.762414800389484,
+      "grad_norm": 0.5453999254984644,
+      "learning_rate": 9.294237803477076e-06,
+      "loss": 0.648,
+      "step": 261
+    },
+    {
+      "epoch": 0.7653359298928919,
+      "grad_norm": 0.7523663818704205,
+      "learning_rate": 9.285495600466683e-06,
+      "loss": 0.6488,
+      "step": 262
+    },
+    {
+      "epoch": 0.7682570593963,
+      "grad_norm": 0.7485014295621598,
+      "learning_rate": 9.276703750002801e-06,
+      "loss": 0.6255,
+      "step": 263
+    },
+    {
+      "epoch": 0.7711781888997079,
+      "grad_norm": 0.686777506781002,
+      "learning_rate": 9.267862353938988e-06,
+      "loss": 0.6534,
+      "step": 264
+    },
+    {
+      "epoch": 0.7740993184031159,
+      "grad_norm": 0.7655220545156425,
+      "learning_rate": 9.258971514702789e-06,
+      "loss": 0.6439,
+      "step": 265
+    },
+    {
+      "epoch": 0.7770204479065238,
+      "grad_norm": 0.5995690372205543,
+      "learning_rate": 9.250031335294551e-06,
+      "loss": 0.6264,
+      "step": 266
+    },
+    {
+      "epoch": 0.7799415774099319,
+      "grad_norm": 0.6306408667729854,
+      "learning_rate": 9.241041919286227e-06,
+      "loss": 0.633,
+      "step": 267
+    },
+    {
+      "epoch": 0.7828627069133398,
+      "grad_norm": 0.7405970033463782,
+      "learning_rate": 9.232003370820171e-06,
+      "loss": 0.6355,
+      "step": 268
+    },
+    {
+      "epoch": 0.7857838364167478,
+      "grad_norm": 0.5761877077710947,
+      "learning_rate": 9.222915794607942e-06,
+      "loss": 0.6431,
+      "step": 269
+    },
+    {
+      "epoch": 0.7887049659201558,
+      "grad_norm": 0.6417671555725009,
+      "learning_rate": 9.213779295929082e-06,
+      "loss": 0.6302,
+      "step": 270
+    },
+    {
+      "epoch": 0.7916260954235638,
+      "grad_norm": 0.5476966311891922,
+      "learning_rate": 9.204593980629898e-06,
+      "loss": 0.6307,
+      "step": 271
+    },
+    {
+      "epoch": 0.7945472249269717,
+      "grad_norm": 0.6325575856222458,
+      "learning_rate": 9.195359955122244e-06,
+      "loss": 0.6316,
+      "step": 272
+    },
+    {
+      "epoch": 0.7974683544303798,
+      "grad_norm": 0.6487642715668982,
+      "learning_rate": 9.186077326382275e-06,
+      "loss": 0.6324,
+      "step": 273
+    },
+    {
+      "epoch": 0.8003894839337877,
+      "grad_norm": 0.5751549331819923,
+      "learning_rate": 9.176746201949216e-06,
+      "loss": 0.6585,
+      "step": 274
+    },
+    {
+      "epoch": 0.8033106134371957,
+      "grad_norm": 0.6208214920916966,
+      "learning_rate": 9.167366689924116e-06,
+      "loss": 0.6517,
+      "step": 275
+    },
+    {
+      "epoch": 0.8062317429406037,
+      "grad_norm": 0.6579852013112687,
+      "learning_rate": 9.157938898968594e-06,
+      "loss": 0.643,
+      "step": 276
+    },
+    {
+      "epoch": 0.8091528724440117,
+      "grad_norm": 0.5985871969783593,
+      "learning_rate": 9.14846293830358e-06,
+      "loss": 0.6386,
+      "step": 277
+    },
+    {
+      "epoch": 0.8120740019474196,
+      "grad_norm": 0.5776495482063276,
+      "learning_rate": 9.138938917708047e-06,
+      "loss": 0.6367,
+      "step": 278
+    },
+    {
+      "epoch": 0.8149951314508277,
+      "grad_norm": 0.6461097088775256,
+      "learning_rate": 9.129366947517746e-06,
+      "loss": 0.6311,
+      "step": 279
+    },
+    {
+      "epoch": 0.8179162609542356,
+      "grad_norm": 0.6054324701596054,
+      "learning_rate": 9.119747138623925e-06,
+      "loss": 0.6365,
+      "step": 280
+    },
+    {
+      "epoch": 0.8208373904576436,
+      "grad_norm": 0.6046664159813615,
+      "learning_rate": 9.110079602472035e-06,
+      "loss": 0.6549,
+      "step": 281
+    },
+    {
+      "epoch": 0.8237585199610516,
+      "grad_norm": 0.690747949343666,
+      "learning_rate": 9.100364451060457e-06,
+      "loss": 0.6477,
+      "step": 282
+    },
+    {
+      "epoch": 0.8266796494644596,
+      "grad_norm": 0.6352884441285447,
+      "learning_rate": 9.090601796939192e-06,
+      "loss": 0.6315,
+      "step": 283
+    },
+    {
+      "epoch": 0.8296007789678675,
+      "grad_norm": 0.7031617950372325,
+      "learning_rate": 9.080791753208553e-06,
+      "loss": 0.6304,
+      "step": 284
+    },
+    {
+      "epoch": 0.8325219084712756,
+      "grad_norm": 0.6955969158961154,
+      "learning_rate": 9.070934433517872e-06,
+      "loss": 0.6371,
+      "step": 285
+    },
+    {
+      "epoch": 0.8354430379746836,
+      "grad_norm": 0.8508450672946094,
+      "learning_rate": 9.061029952064165e-06,
+      "loss": 0.6392,
+      "step": 286
+    },
+    {
+      "epoch": 0.8383641674780915,
+      "grad_norm": 0.6682143045272909,
+      "learning_rate": 9.05107842359082e-06,
+      "loss": 0.6354,
+      "step": 287
+    },
+    {
+      "epoch": 0.8412852969814996,
+      "grad_norm": 0.730935179569577,
+      "learning_rate": 9.041079963386263e-06,
+      "loss": 0.6365,
+      "step": 288
+    },
+    {
+      "epoch": 0.8442064264849075,
+      "grad_norm": 0.9284592753787645,
+      "learning_rate": 9.031034687282627e-06,
+      "loss": 0.6512,
+      "step": 289
+    },
+    {
+      "epoch": 0.8471275559883155,
+      "grad_norm": 0.5711740420462373,
+      "learning_rate": 9.020942711654404e-06,
+      "loss": 0.6253,
+      "step": 290
+    },
+    {
+      "epoch": 0.8500486854917235,
+      "grad_norm": 0.7521676041243963,
+      "learning_rate": 9.0108041534171e-06,
+      "loss": 0.6346,
+      "step": 291
+    },
+    {
+      "epoch": 0.8529698149951315,
+      "grad_norm": 0.7268819094436544,
+      "learning_rate": 9.000619130025885e-06,
+      "loss": 0.6321,
+      "step": 292
+    },
+    {
+      "epoch": 0.8558909444985394,
+      "grad_norm": 0.6255002973375909,
+      "learning_rate": 8.99038775947422e-06,
+      "loss": 0.6448,
+      "step": 293
+    },
+    {
+      "epoch": 0.8588120740019474,
+      "grad_norm": 0.5654580978730834,
+      "learning_rate": 8.980110160292503e-06,
+      "loss": 0.6546,
+      "step": 294
+    },
+    {
+      "epoch": 0.8617332035053554,
+      "grad_norm": 0.5744283394449378,
+      "learning_rate": 8.969786451546691e-06,
+      "loss": 0.6354,
+      "step": 295
+    },
+    {
+      "epoch": 0.8646543330087634,
+      "grad_norm": 0.5607277367845812,
+      "learning_rate": 8.959416752836915e-06,
+      "loss": 0.6315,
+      "step": 296
+    },
+    {
+      "epoch": 0.8675754625121713,
+      "grad_norm": 0.6883032069319117,
+      "learning_rate": 8.949001184296107e-06,
+      "loss": 0.6284,
+      "step": 297
+    },
+    {
+      "epoch": 0.8704965920155794,
+      "grad_norm": 0.5793320184156081,
+      "learning_rate": 8.938539866588593e-06,
+      "loss": 0.6299,
+      "step": 298
+    },
+    {
+      "epoch": 0.8734177215189873,
+      "grad_norm": 0.6122270930586687,
+      "learning_rate": 8.928032920908709e-06,
+      "loss": 0.6463,
+      "step": 299
+    },
+    {
+      "epoch": 0.8763388510223953,
+      "grad_norm": 0.6718320925279334,
+      "learning_rate": 8.917480468979387e-06,
+      "loss": 0.6405,
+      "step": 300
+    },
+    {
+      "epoch": 0.8792599805258033,
+      "grad_norm": 0.6132181340646334,
+      "learning_rate": 8.906882633050753e-06,
+      "loss": 0.6426,
+      "step": 301
+    },
+    {
+      "epoch": 0.8821811100292113,
+      "grad_norm": 0.6206805066838101,
+      "learning_rate": 8.896239535898702e-06,
+      "loss": 0.65,
+      "step": 302
+    },
+    {
+      "epoch": 0.8851022395326192,
+      "grad_norm": 0.7456156741229341,
+      "learning_rate": 8.885551300823483e-06,
+      "loss": 0.6395,
+      "step": 303
+    },
+    {
+      "epoch": 0.8880233690360273,
+      "grad_norm": 0.6186279743363227,
+      "learning_rate": 8.874818051648267e-06,
+      "loss": 0.6236,
+      "step": 304
+    },
+    {
+      "epoch": 0.8909444985394352,
+      "grad_norm": 0.7192434268898347,
+      "learning_rate": 8.864039912717713e-06,
+      "loss": 0.6427,
+      "step": 305
+    },
+    {
+      "epoch": 0.8938656280428432,
+      "grad_norm": 0.6528244466221247,
+      "learning_rate": 8.853217008896526e-06,
+      "loss": 0.6478,
+      "step": 306
+    },
+    {
+      "epoch": 0.8967867575462513,
+      "grad_norm": 0.6240468261028331,
+      "learning_rate": 8.842349465568018e-06,
+      "loss": 0.6354,
+      "step": 307
+    },
+    {
+      "epoch": 0.8997078870496592,
+      "grad_norm": 0.6086442105860419,
+      "learning_rate": 8.831437408632639e-06,
+      "loss": 0.6175,
+      "step": 308
+    },
+    {
+      "epoch": 0.9026290165530672,
+      "grad_norm": 0.6206978097636743,
+      "learning_rate": 8.820480964506542e-06,
+      "loss": 0.6329,
+      "step": 309
+    },
+    {
+      "epoch": 0.9055501460564752,
+      "grad_norm": 0.67571865721595,
+      "learning_rate": 8.809480260120096e-06,
+      "loss": 0.6302,
+      "step": 310
+    },
+    {
+      "epoch": 0.9084712755598832,
+      "grad_norm": 0.612775453081801,
+      "learning_rate": 8.798435422916425e-06,
+      "loss": 0.6248,
+      "step": 311
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5939423168299965,
+      "learning_rate": 8.787346580849939e-06,
+      "loss": 0.6274,
+      "step": 312
+    },
+    {
+      "epoch": 0.9143135345666992,
+      "grad_norm": 0.6578489213437513,
+      "learning_rate": 8.776213862384838e-06,
+      "loss": 0.6367,
+      "step": 313
+    },
+    {
+      "epoch": 0.9172346640701071,
+      "grad_norm": 0.7439730364901744,
+      "learning_rate": 8.76503739649363e-06,
+      "loss": 0.6584,
+      "step": 314
+    },
+    {
+      "epoch": 0.9201557935735151,
+      "grad_norm": 0.5758626313580656,
+      "learning_rate": 8.753817312655642e-06,
+      "loss": 0.638,
+      "step": 315
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.7372145224476075,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.6391,
+      "step": 316
+    },
+    {
+      "epoch": 0.9259980525803311,
+      "grad_norm": 0.7004119821194315,
+      "learning_rate": 8.73124681158167e-06,
+      "loss": 0.6426,
+      "step": 317
+    },
+    {
+      "epoch": 0.928919182083739,
+      "grad_norm": 0.5393578063337247,
+      "learning_rate": 8.719896655824878e-06,
+      "loss": 0.6326,
+      "step": 318
+    },
+    {
+      "epoch": 0.9318403115871471,
+      "grad_norm": 0.5698772754745327,
+      "learning_rate": 8.708503405076646e-06,
+      "loss": 0.634,
+      "step": 319
+    },
+    {
+      "epoch": 0.934761441090555,
+      "grad_norm": 0.588256486663727,
+      "learning_rate": 8.697067191327748e-06,
+      "loss": 0.6328,
+      "step": 320
+    },
+    {
+      "epoch": 0.937682570593963,
+      "grad_norm": 0.6194561945270076,
+      "learning_rate": 8.685588147066688e-06,
+      "loss": 0.6303,
+      "step": 321
+    },
+    {
+      "epoch": 0.9406037000973709,
+      "grad_norm": 0.5910869397544244,
+      "learning_rate": 8.67406640527816e-06,
+      "loss": 0.6494,
+      "step": 322
+    },
+    {
+      "epoch": 0.943524829600779,
+      "grad_norm": 0.6055073823945398,
+      "learning_rate": 8.662502099441505e-06,
+      "loss": 0.6334,
+      "step": 323
+    },
+    {
+      "epoch": 0.9464459591041869,
+      "grad_norm": 0.5417594595582104,
+      "learning_rate": 8.650895363529172e-06,
+      "loss": 0.6279,
+      "step": 324
+    },
+    {
+      "epoch": 0.9493670886075949,
+      "grad_norm": 0.5878187442362004,
+      "learning_rate": 8.639246332005163e-06,
+      "loss": 0.639,
+      "step": 325
+    },
+    {
+      "epoch": 0.952288218111003,
+      "grad_norm": 0.5233625621707794,
+      "learning_rate": 8.627555139823468e-06,
+      "loss": 0.614,
+      "step": 326
+    },
+    {
+      "epoch": 0.9552093476144109,
+      "grad_norm": 0.5656952115933153,
+      "learning_rate": 8.615821922426517e-06,
+      "loss": 0.6214,
+      "step": 327
+    },
+    {
+      "epoch": 0.9581304771178188,
+      "grad_norm": 0.5806200502649499,
+      "learning_rate": 8.604046815743598e-06,
+      "loss": 0.6424,
+      "step": 328
+    },
+    {
+      "epoch": 0.9610516066212269,
+      "grad_norm": 0.5561534960958242,
+      "learning_rate": 8.592229956189283e-06,
+      "loss": 0.638,
+      "step": 329
+    },
+    {
+      "epoch": 0.9639727361246349,
+      "grad_norm": 0.5340425065767049,
+      "learning_rate": 8.580371480661857e-06,
+      "loss": 0.6238,
+      "step": 330
+    },
+    {
+      "epoch": 0.9668938656280428,
+      "grad_norm": 0.5501524304300768,
+      "learning_rate": 8.568471526541721e-06,
+      "loss": 0.6518,
+      "step": 331
+    },
+    {
+      "epoch": 0.9698149951314509,
+      "grad_norm": 0.571968183721703,
+      "learning_rate": 8.556530231689809e-06,
+      "loss": 0.6588,
+      "step": 332
+    },
+    {
+      "epoch": 0.9727361246348588,
+      "grad_norm": 0.5870934675467651,
+      "learning_rate": 8.544547734445983e-06,
+      "loss": 0.6441,
+      "step": 333
+    },
+    {
+      "epoch": 0.9756572541382668,
+      "grad_norm": 0.5633142366316923,
+      "learning_rate": 8.532524173627438e-06,
+      "loss": 0.6558,
+      "step": 334
+    },
+    {
+      "epoch": 0.9785783836416748,
+      "grad_norm": 0.6080503864783372,
+      "learning_rate": 8.520459688527091e-06,
+      "loss": 0.6239,
+      "step": 335
+    },
+    {
+      "epoch": 0.9814995131450828,
+      "grad_norm": 0.6010534017830508,
+      "learning_rate": 8.508354418911966e-06,
+      "loss": 0.6392,
+      "step": 336
+    },
+    {
+      "epoch": 0.9844206426484907,
+      "grad_norm": 0.6110941694269748,
+      "learning_rate": 8.496208505021572e-06,
+      "loss": 0.6334,
+      "step": 337
+    },
+    {
+      "epoch": 0.9873417721518988,
+      "grad_norm": 0.5748899254330844,
+      "learning_rate": 8.484022087566284e-06,
+      "loss": 0.6213,
+      "step": 338
+    },
+    {
+      "epoch": 0.9902629016553067,
+      "grad_norm": 0.5910067428214469,
+      "learning_rate": 8.471795307725713e-06,
+      "loss": 0.6313,
+      "step": 339
+    },
+    {
+      "epoch": 0.9931840311587147,
+      "grad_norm": 0.6409845798522262,
+      "learning_rate": 8.459528307147066e-06,
+      "loss": 0.6223,
+      "step": 340
+    },
+    {
+      "epoch": 0.9961051606621227,
+      "grad_norm": 0.5798933129205326,
+      "learning_rate": 8.447221227943507e-06,
+      "loss": 0.6375,
+      "step": 341
+    },
+    {
+      "epoch": 0.9990262901655307,
+      "grad_norm": 0.5323356127721257,
+      "learning_rate": 8.434874212692513e-06,
+      "loss": 0.631,
+      "step": 342
+    },
+    {
+      "epoch": 1.0019474196689386,
+      "grad_norm": 1.1893567012708939,
+      "learning_rate": 8.422487404434214e-06,
+      "loss": 1.0195,
+      "step": 343
+    },
+    {
+      "epoch": 1.0048685491723466,
+      "grad_norm": 0.6455689118062985,
+      "learning_rate": 8.41006094666975e-06,
+      "loss": 0.5891,
+      "step": 344
+    },
+    {
+      "epoch": 1.0077896786757545,
+      "grad_norm": 0.6583690984771934,
+      "learning_rate": 8.397594983359591e-06,
+      "loss": 0.5549,
+      "step": 345
+    },
+    {
+      "epoch": 1.0107108081791627,
+      "grad_norm": 0.6011562108844719,
+      "learning_rate": 8.385089658921892e-06,
+      "loss": 0.4942,
+      "step": 346
+    },
+    {
+      "epoch": 1.0136319376825706,
+      "grad_norm": 0.7194629629820957,
+      "learning_rate": 8.372545118230793e-06,
+      "loss": 0.5879,
+      "step": 347
+    },
+    {
+      "epoch": 1.0165530671859786,
+      "grad_norm": 0.7106922347864785,
+      "learning_rate": 8.35996150661476e-06,
+      "loss": 0.6464,
+      "step": 348
+    },
+    {
+      "epoch": 1.0194741966893865,
+      "grad_norm": 0.7004876649145284,
+      "learning_rate": 8.347338969854898e-06,
+      "loss": 0.5635,
+      "step": 349
+    },
+    {
+      "epoch": 1.0223953261927945,
+      "grad_norm": 0.7176313477690597,
+      "learning_rate": 8.334677654183254e-06,
+      "loss": 0.6121,
+      "step": 350
+    },
+    {
+      "epoch": 1.0253164556962024,
+      "grad_norm": 0.6767419736782746,
+      "learning_rate": 8.321977706281135e-06,
+      "loss": 0.5923,
+      "step": 351
+    },
+    {
+      "epoch": 1.0282375851996106,
+      "grad_norm": 0.6606355795156919,
+      "learning_rate": 8.309239273277394e-06,
+      "loss": 0.5375,
+      "step": 352
+    },
+    {
+      "epoch": 1.0311587147030186,
+      "grad_norm": 0.7777491012749531,
+      "learning_rate": 8.296462502746743e-06,
+      "loss": 0.5971,
+      "step": 353
+    },
+    {
+      "epoch": 1.0340798442064265,
+      "grad_norm": 0.6239185790928177,
+      "learning_rate": 8.283647542708026e-06,
+      "loss": 0.6017,
+      "step": 354
+    },
+    {
+      "epoch": 1.0370009737098345,
+      "grad_norm": 0.659621365452411,
+      "learning_rate": 8.27079454162252e-06,
+      "loss": 0.523,
+      "step": 355
+    },
+    {
+      "epoch": 1.0399221032132424,
+      "grad_norm": 0.7314219237543246,
+      "learning_rate": 8.2579036483922e-06,
+      "loss": 0.5992,
+      "step": 356
+    },
+    {
+      "epoch": 1.0428432327166504,
+      "grad_norm": 0.6317967581739343,
+      "learning_rate": 8.244975012358028e-06,
+      "loss": 0.58,
+      "step": 357
+    },
+    {
+      "epoch": 1.0457643622200585,
+      "grad_norm": 0.709552964482946,
+      "learning_rate": 8.232008783298211e-06,
+      "loss": 0.5703,
+      "step": 358
+    },
+    {
+      "epoch": 1.0486854917234665,
+      "grad_norm": 0.5856088171700015,
+      "learning_rate": 8.219005111426472e-06,
+      "loss": 0.5851,
+      "step": 359
+    },
+    {
+      "epoch": 1.0516066212268744,
+      "grad_norm": 0.72335786152197,
+      "learning_rate": 8.205964147390313e-06,
+      "loss": 0.5762,
+      "step": 360
+    },
+    {
+      "epoch": 1.0545277507302824,
+      "grad_norm": 0.6082315528170907,
+      "learning_rate": 8.19288604226926e-06,
+      "loss": 0.6045,
+      "step": 361
+    },
+    {
+      "epoch": 1.0574488802336903,
+      "grad_norm": 0.6904542946382585,
+      "learning_rate": 8.179770947573124e-06,
+      "loss": 0.5649,
+      "step": 362
+    },
+    {
+      "epoch": 1.0603700097370983,
+      "grad_norm": 0.5889683647522176,
+      "learning_rate": 8.166619015240236e-06,
+      "loss": 0.5681,
+      "step": 363
+    },
+    {
+      "epoch": 1.0632911392405062,
+      "grad_norm": 0.5621253187560026,
+      "learning_rate": 8.15343039763569e-06,
+      "loss": 0.6013,
+      "step": 364
+    },
+    {
+      "epoch": 1.0662122687439144,
+      "grad_norm": 0.5587583825073225,
+      "learning_rate": 8.140205247549583e-06,
+      "loss": 0.6026,
+      "step": 365
+    },
+    {
+      "epoch": 1.0691333982473223,
+      "grad_norm": 0.5053935078299595,
+      "learning_rate": 8.126943718195239e-06,
+      "loss": 0.5693,
+      "step": 366
+    },
+    {
+      "epoch": 1.0720545277507303,
+      "grad_norm": 0.5469581161571481,
+      "learning_rate": 8.113645963207432e-06,
+      "loss": 0.6007,
+      "step": 367
+    },
+    {
+      "epoch": 1.0749756572541382,
+      "grad_norm": 0.5430360387939689,
+      "learning_rate": 8.100312136640618e-06,
+      "loss": 0.5754,
+      "step": 368
+    },
+    {
+      "epoch": 1.0778967867575462,
+      "grad_norm": 0.529838189847217,
+      "learning_rate": 8.086942392967131e-06,
+      "loss": 0.5758,
+      "step": 369
+    },
+    {
+      "epoch": 1.0808179162609541,
+      "grad_norm": 0.5537166527008645,
+      "learning_rate": 8.073536887075417e-06,
+      "loss": 0.5596,
+      "step": 370
+    },
+    {
+      "epoch": 1.0837390457643623,
+      "grad_norm": 0.575683387457894,
+      "learning_rate": 8.060095774268217e-06,
+      "loss": 0.6149,
+      "step": 371
+    },
+    {
+      "epoch": 1.0866601752677703,
+      "grad_norm": 0.5212823944156575,
+      "learning_rate": 8.046619210260785e-06,
+      "loss": 0.5376,
+      "step": 372
+    },
+    {
+      "epoch": 1.0895813047711782,
+      "grad_norm": 0.6132577983715938,
+      "learning_rate": 8.03310735117907e-06,
+      "loss": 0.611,
+      "step": 373
+    },
+    {
+      "epoch": 1.0925024342745862,
+      "grad_norm": 0.5741650450345563,
+      "learning_rate": 8.019560353557923e-06,
+      "loss": 0.5428,
+      "step": 374
+    },
+    {
+      "epoch": 1.095423563777994,
+      "grad_norm": 0.6898994227426227,
+      "learning_rate": 8.005978374339264e-06,
+      "loss": 0.6122,
+      "step": 375
+    },
+    {
+      "epoch": 1.098344693281402,
+      "grad_norm": 0.6195159909025971,
+      "learning_rate": 7.992361570870289e-06,
+      "loss": 0.6298,
+      "step": 376
+    },
+    {
+      "epoch": 1.1012658227848102,
+      "grad_norm": 0.5592176863621418,
+      "learning_rate": 7.978710100901617e-06,
+      "loss": 0.527,
+      "step": 377
+    },
+    {
+      "epoch": 1.1041869522882182,
+      "grad_norm": 0.6077274644298606,
+      "learning_rate": 7.965024122585491e-06,
+      "loss": 0.5733,
+      "step": 378
+    },
+    {
+      "epoch": 1.1071080817916261,
+      "grad_norm": 0.5404386498094536,
+      "learning_rate": 7.951303794473926e-06,
+      "loss": 0.5786,
+      "step": 379
+    },
+    {
+      "epoch": 1.110029211295034,
+      "grad_norm": 0.6285033579389189,
+      "learning_rate": 7.937549275516882e-06,
+      "loss": 0.5593,
+      "step": 380
+    },
+    {
+      "epoch": 1.112950340798442,
+      "grad_norm": 0.6136636789874864,
+      "learning_rate": 7.92376072506042e-06,
+      "loss": 0.5887,
+      "step": 381
+    },
+    {
+      "epoch": 1.11587147030185,
+      "grad_norm": 0.5563073653338333,
+      "learning_rate": 7.909938302844856e-06,
+      "loss": 0.637,
+      "step": 382
+    },
+    {
+      "epoch": 1.1187925998052581,
+      "grad_norm": 0.4984935085580574,
+      "learning_rate": 7.896082169002903e-06,
+      "loss": 0.537,
+      "step": 383
+    },
+    {
+      "epoch": 1.121713729308666,
+      "grad_norm": 0.5688731697204966,
+      "learning_rate": 7.882192484057837e-06,
+      "loss": 0.5977,
+      "step": 384
+    },
+    {
+      "epoch": 1.124634858812074,
+      "grad_norm": 0.5409938929481367,
+      "learning_rate": 7.868269408921614e-06,
+      "loss": 0.5477,
+      "step": 385
+    },
+    {
+      "epoch": 1.127555988315482,
+      "grad_norm": 0.6311630545441865,
+      "learning_rate": 7.854313104893014e-06,
+      "loss": 0.5595,
+      "step": 386
+    },
+    {
+      "epoch": 1.13047711781889,
+      "grad_norm": 0.5517528259351719,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.5588,
+      "step": 387
+    },
+    {
+      "epoch": 1.1333982473222979,
+      "grad_norm": 0.515690277716415,
+      "learning_rate": 7.826301457276733e-06,
+      "loss": 0.5767,
+      "step": 388
+    },
+    {
+      "epoch": 1.136319376825706,
+      "grad_norm": 0.5872560884226455,
+      "learning_rate": 7.812246438203905e-06,
+      "loss": 0.5618,
+      "step": 389
+    },
+    {
+      "epoch": 1.139240506329114,
+      "grad_norm": 0.5449550621943328,
+      "learning_rate": 7.798158839264645e-06,
+      "loss": 0.5373,
+      "step": 390
+    },
+    {
+      "epoch": 1.142161635832522,
+      "grad_norm": 0.6192261787272578,
+      "learning_rate": 7.784038823663746e-06,
+      "loss": 0.6117,
+      "step": 391
+    },
+    {
+      "epoch": 1.14508276533593,
+      "grad_norm": 0.5911797274948096,
+      "learning_rate": 7.769886554981549e-06,
+      "loss": 0.5832,
+      "step": 392
+    },
+    {
+      "epoch": 1.1480038948393378,
+      "grad_norm": 0.5393860544929336,
+      "learning_rate": 7.755702197172036e-06,
+      "loss": 0.5511,
+      "step": 393
+    },
+    {
+      "epoch": 1.1509250243427458,
+      "grad_norm": 0.5686675548722897,
+      "learning_rate": 7.741485914560958e-06,
+      "loss": 0.601,
+      "step": 394
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.5965421717249094,
+      "learning_rate": 7.7272378718439e-06,
+      "loss": 0.6142,
+      "step": 395
+    },
+    {
+      "epoch": 1.156767283349562,
+      "grad_norm": 0.5976378598496599,
+      "learning_rate": 7.712958234084395e-06,
+      "loss": 0.532,
+      "step": 396
+    },
+    {
+      "epoch": 1.1596884128529699,
+      "grad_norm": 0.5866514447746561,
+      "learning_rate": 7.698647166712003e-06,
+      "loss": 0.6436,
+      "step": 397
+    },
+    {
+      "epoch": 1.1626095423563778,
+      "grad_norm": 0.5577115650955226,
+      "learning_rate": 7.684304835520395e-06,
+      "loss": 0.5524,
+      "step": 398
+    },
+    {
+      "epoch": 1.1655306718597858,
+      "grad_norm": 0.6320330036922427,
+      "learning_rate": 7.669931406665437e-06,
+      "loss": 0.631,
+      "step": 399
+    },
+    {
+      "epoch": 1.1684518013631937,
+      "grad_norm": 0.5274101108563934,
+      "learning_rate": 7.655527046663254e-06,
+      "loss": 0.5369,
+      "step": 400
+    },
+    {
+      "epoch": 1.1713729308666017,
+      "grad_norm": 0.5563468540404465,
+      "learning_rate": 7.641091922388316e-06,
+      "loss": 0.577,
+      "step": 401
+    },
+    {
+      "epoch": 1.1742940603700098,
+      "grad_norm": 0.5043411412651241,
+      "learning_rate": 7.626626201071494e-06,
+      "loss": 0.5623,
+      "step": 402
+    },
+    {
+      "epoch": 1.1772151898734178,
+      "grad_norm": 0.5658997847969963,
+      "learning_rate": 7.612130050298126e-06,
+      "loss": 0.5613,
+      "step": 403
+    },
+    {
+      "epoch": 1.1801363193768257,
+      "grad_norm": 0.5617715925867288,
+      "learning_rate": 7.597603638006071e-06,
+      "loss": 0.5796,
+      "step": 404
+    },
+    {
+      "epoch": 1.1830574488802337,
+      "grad_norm": 0.5730674124102592,
+      "learning_rate": 7.5830471324837765e-06,
+      "loss": 0.6102,
+      "step": 405
+    },
+    {
+      "epoch": 1.1859785783836416,
+      "grad_norm": 0.5322280548113558,
+      "learning_rate": 7.56846070236831e-06,
+      "loss": 0.5392,
+      "step": 406
+    },
+    {
+      "epoch": 1.1888997078870496,
+      "grad_norm": 0.5636446588596294,
+      "learning_rate": 7.55384451664342e-06,
+      "loss": 0.5805,
+      "step": 407
+    },
+    {
+      "epoch": 1.1918208373904577,
+      "grad_norm": 0.5515192754192108,
+      "learning_rate": 7.539198744637577e-06,
+      "loss": 0.5647,
+      "step": 408
+    },
+    {
+      "epoch": 1.1947419668938657,
+      "grad_norm": 0.649611846767762,
+      "learning_rate": 7.524523556022003e-06,
+      "loss": 0.5804,
+      "step": 409
+    },
+    {
+      "epoch": 1.1976630963972736,
+      "grad_norm": 0.5400876783820088,
+      "learning_rate": 7.5098191208087144e-06,
+      "loss": 0.5295,
+      "step": 410
+    },
+    {
+      "epoch": 1.2005842259006816,
+      "grad_norm": 0.727016701945529,
+      "learning_rate": 7.495085609348549e-06,
+      "loss": 0.6035,
+      "step": 411
+    },
+    {
+      "epoch": 1.2035053554040895,
+      "grad_norm": 0.6474638029028308,
+      "learning_rate": 7.4803231923291905e-06,
+      "loss": 0.5905,
+      "step": 412
+    },
+    {
+      "epoch": 1.2064264849074975,
+      "grad_norm": 0.5322418094330414,
+      "learning_rate": 7.465532040773195e-06,
+      "loss": 0.5696,
+      "step": 413
+    },
+    {
+      "epoch": 1.2093476144109054,
+      "grad_norm": 0.6033615790880061,
+      "learning_rate": 7.45071232603601e-06,
+      "loss": 0.5495,
+      "step": 414
+    },
+    {
+      "epoch": 1.2122687439143136,
+      "grad_norm": 0.5696185950396394,
+      "learning_rate": 7.4358642198039835e-06,
+      "loss": 0.5761,
+      "step": 415
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.58236484440868,
+      "learning_rate": 7.420987894092383e-06,
+      "loss": 0.6225,
+      "step": 416
+    },
+    {
+      "epoch": 1.2181110029211295,
+      "grad_norm": 0.6023725223738031,
+      "learning_rate": 7.406083521243396e-06,
+      "loss": 0.5539,
+      "step": 417
+    },
+    {
+      "epoch": 1.2210321324245375,
+      "grad_norm": 0.5964833664567649,
+      "learning_rate": 7.391151273924135e-06,
+      "loss": 0.5766,
+      "step": 418
+    },
+    {
+      "epoch": 1.2239532619279454,
+      "grad_norm": 0.5579327351939204,
+      "learning_rate": 7.376191325124644e-06,
+      "loss": 0.6037,
+      "step": 419
+    },
+    {
+      "epoch": 1.2268743914313536,
+      "grad_norm": 0.662691054067833,
+      "learning_rate": 7.36120384815588e-06,
+      "loss": 0.5775,
+      "step": 420
+    },
+    {
+      "epoch": 1.2297955209347615,
+      "grad_norm": 0.5457870053427097,
+      "learning_rate": 7.34618901664772e-06,
+      "loss": 0.6032,
+      "step": 421
+    },
+    {
+      "epoch": 1.2327166504381695,
+      "grad_norm": 0.528437147629629,
+      "learning_rate": 7.33114700454694e-06,
+      "loss": 0.5551,
+      "step": 422
+    },
+    {
+      "epoch": 1.2356377799415774,
+      "grad_norm": 0.6396983335282087,
+      "learning_rate": 7.316077986115206e-06,
+      "loss": 0.5546,
+      "step": 423
+    },
+    {
+      "epoch": 1.2385589094449854,
+      "grad_norm": 0.6059608136412865,
+      "learning_rate": 7.300982135927051e-06,
+      "loss": 0.5889,
+      "step": 424
+    },
+    {
+      "epoch": 1.2414800389483933,
+      "grad_norm": 0.5395833042938699,
+      "learning_rate": 7.285859628867851e-06,
+      "loss": 0.5765,
+      "step": 425
+    },
+    {
+      "epoch": 1.2444011684518013,
+      "grad_norm": 0.5938150264550871,
+      "learning_rate": 7.270710640131806e-06,
+      "loss": 0.6058,
+      "step": 426
+    },
+    {
+      "epoch": 1.2473222979552094,
+      "grad_norm": 0.5435408340953944,
+      "learning_rate": 7.255535345219905e-06,
+      "loss": 0.5933,
+      "step": 427
+    },
+    {
+      "epoch": 1.2502434274586174,
+      "grad_norm": 0.5055763559518435,
+      "learning_rate": 7.240333919937893e-06,
+      "loss": 0.5989,
+      "step": 428
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.5862621574043387,
+      "learning_rate": 7.2251065403942355e-06,
+      "loss": 0.5888,
+      "step": 429
+    },
+    {
+      "epoch": 1.2560856864654333,
+      "grad_norm": 0.5234746081250162,
+      "learning_rate": 7.209853382998077e-06,
+      "loss": 0.5537,
+      "step": 430
+    },
+    {
+      "epoch": 1.2590068159688412,
+      "grad_norm": 0.5736429396541277,
+      "learning_rate": 7.1945746244572e-06,
+      "loss": 0.547,
+      "step": 431
+    },
+    {
+      "epoch": 1.2619279454722494,
+      "grad_norm": 0.5662950178895197,
+      "learning_rate": 7.179270441775976e-06,
+      "loss": 0.6502,
+      "step": 432
+    },
+    {
+      "epoch": 1.2648490749756571,
+      "grad_norm": 0.6193942417064999,
+      "learning_rate": 7.163941012253317e-06,
+      "loss": 0.5182,
+      "step": 433
+    },
+    {
+      "epoch": 1.2677702044790653,
+      "grad_norm": 0.5954433533450075,
+      "learning_rate": 7.148586513480614e-06,
+      "loss": 0.5961,
+      "step": 434
+    },
+    {
+      "epoch": 1.2706913339824732,
+      "grad_norm": 0.5407062847357796,
+      "learning_rate": 7.133207123339689e-06,
+      "loss": 0.5475,
+      "step": 435
+    },
+    {
+      "epoch": 1.2736124634858812,
+      "grad_norm": 0.6133121400132443,
+      "learning_rate": 7.117803020000733e-06,
+      "loss": 0.59,
+      "step": 436
+    },
+    {
+      "epoch": 1.2765335929892891,
+      "grad_norm": 0.5272070801107334,
+      "learning_rate": 7.102374381920233e-06,
+      "loss": 0.542,
+      "step": 437
+    },
+    {
+      "epoch": 1.279454722492697,
+      "grad_norm": 0.6014559091825684,
+      "learning_rate": 7.086921387838916e-06,
+      "loss": 0.6136,
+      "step": 438
+    },
+    {
+      "epoch": 1.2823758519961053,
+      "grad_norm": 0.4925204082847554,
+      "learning_rate": 7.071444216779669e-06,
+      "loss": 0.5707,
+      "step": 439
+    },
+    {
+      "epoch": 1.2852969814995132,
+      "grad_norm": 0.5297523991584766,
+      "learning_rate": 7.055943048045476e-06,
+      "loss": 0.5917,
+      "step": 440
+    },
+    {
+      "epoch": 1.2882181110029212,
+      "grad_norm": 0.5742994575633306,
+      "learning_rate": 7.040418061217325e-06,
+      "loss": 0.6161,
+      "step": 441
+    },
+    {
+      "epoch": 1.2911392405063291,
+      "grad_norm": 0.5144128093922836,
+      "learning_rate": 7.024869436152144e-06,
+      "loss": 0.5481,
+      "step": 442
+    },
+    {
+      "epoch": 1.294060370009737,
+      "grad_norm": 0.5651127309438425,
+      "learning_rate": 7.009297352980706e-06,
+      "loss": 0.5789,
+      "step": 443
+    },
+    {
+      "epoch": 1.296981499513145,
+      "grad_norm": 0.46327530559493646,
+      "learning_rate": 6.99370199210555e-06,
+      "loss": 0.5435,
+      "step": 444
+    },
+    {
+      "epoch": 1.299902629016553,
+      "grad_norm": 0.5306445037454199,
+      "learning_rate": 6.978083534198878e-06,
+      "loss": 0.6135,
+      "step": 445
+    },
+    {
+      "epoch": 1.3028237585199611,
+      "grad_norm": 0.6551321587944974,
+      "learning_rate": 6.962442160200484e-06,
+      "loss": 0.6111,
+      "step": 446
+    },
+    {
+      "epoch": 1.305744888023369,
+      "grad_norm": 0.5887843505365383,
+      "learning_rate": 6.9467780513156335e-06,
+      "loss": 0.5428,
+      "step": 447
+    },
+    {
+      "epoch": 1.308666017526777,
+      "grad_norm": 0.5871163724329924,
+      "learning_rate": 6.931091389012983e-06,
+      "loss": 0.5595,
+      "step": 448
+    },
+    {
+      "epoch": 1.311587147030185,
+      "grad_norm": 0.6876137396335796,
+      "learning_rate": 6.915382355022465e-06,
+      "loss": 0.5776,
+      "step": 449
+    },
+    {
+      "epoch": 1.314508276533593,
+      "grad_norm": 0.5978349360901288,
+      "learning_rate": 6.899651131333194e-06,
+      "loss": 0.559,
+      "step": 450
+    },
+    {
+      "epoch": 1.317429406037001,
+      "grad_norm": 0.5600538501243326,
+      "learning_rate": 6.8838979001913454e-06,
+      "loss": 0.5782,
+      "step": 451
+    },
+    {
+      "epoch": 1.3203505355404088,
+      "grad_norm": 0.6670604315541686,
+      "learning_rate": 6.868122844098057e-06,
+      "loss": 0.6669,
+      "step": 452
+    },
+    {
+      "epoch": 1.323271665043817,
+      "grad_norm": 0.6019460261148103,
+      "learning_rate": 6.852326145807302e-06,
+      "loss": 0.6006,
+      "step": 453
+    },
+    {
+      "epoch": 1.326192794547225,
+      "grad_norm": 0.4932938142869286,
+      "learning_rate": 6.836507988323785e-06,
+      "loss": 0.4971,
+      "step": 454
+    },
+    {
+      "epoch": 1.3291139240506329,
+      "grad_norm": 0.6389612562939274,
+      "learning_rate": 6.82066855490081e-06,
+      "loss": 0.5994,
+      "step": 455
+    },
+    {
+      "epoch": 1.3320350535540408,
+      "grad_norm": 0.569391901245458,
+      "learning_rate": 6.804808029038168e-06,
+      "loss": 0.5776,
+      "step": 456
+    },
+    {
+      "epoch": 1.3349561830574488,
+      "grad_norm": 0.5627619450291675,
+      "learning_rate": 6.788926594480001e-06,
+      "loss": 0.5894,
+      "step": 457
+    },
+    {
+      "epoch": 1.337877312560857,
+      "grad_norm": 0.5525138776130833,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.5507,
+      "step": 458
+    },
+    {
+      "epoch": 1.340798442064265,
+      "grad_norm": 0.6243616187017386,
+      "learning_rate": 6.75710173546267e-06,
+      "loss": 0.6052,
+      "step": 459
+    },
+    {
+      "epoch": 1.3437195715676729,
+      "grad_norm": 0.46296070627628405,
+      "learning_rate": 6.741158679694403e-06,
+      "loss": 0.5284,
+      "step": 460
+    },
+    {
+      "epoch": 1.3466407010710808,
+      "grad_norm": 0.556033009188379,
+      "learning_rate": 6.7251954526081294e-06,
+      "loss": 0.6225,
+      "step": 461
+    },
+    {
+      "epoch": 1.3495618305744888,
+      "grad_norm": 0.5425579804205766,
+      "learning_rate": 6.709212239137785e-06,
+      "loss": 0.5919,
+      "step": 462
+    },
+    {
+      "epoch": 1.352482960077897,
+      "grad_norm": 0.5628452971401441,
+      "learning_rate": 6.693209224448853e-06,
+      "loss": 0.6162,
+      "step": 463
+    },
+    {
+      "epoch": 1.3554040895813046,
+      "grad_norm": 0.5283138648819659,
+      "learning_rate": 6.677186593936207e-06,
+      "loss": 0.5022,
+      "step": 464
+    },
+    {
+      "epoch": 1.3583252190847128,
+      "grad_norm": 0.5688480741205626,
+      "learning_rate": 6.661144533221974e-06,
+      "loss": 0.594,
+      "step": 465
+    },
+    {
+      "epoch": 1.3612463485881208,
+      "grad_norm": 0.6112831896983953,
+      "learning_rate": 6.645083228153377e-06,
+      "loss": 0.5803,
+      "step": 466
+    },
+    {
+      "epoch": 1.3641674780915287,
+      "grad_norm": 0.5385606514191121,
+      "learning_rate": 6.629002864800589e-06,
+      "loss": 0.5834,
+      "step": 467
+    },
+    {
+      "epoch": 1.3670886075949367,
+      "grad_norm": 0.5366278285884805,
+      "learning_rate": 6.612903629454568e-06,
+      "loss": 0.5343,
+      "step": 468
+    },
+    {
+      "epoch": 1.3700097370983446,
+      "grad_norm": 0.6713338764048196,
+      "learning_rate": 6.5967857086249065e-06,
+      "loss": 0.5573,
+      "step": 469
+    },
+    {
+      "epoch": 1.3729308666017528,
+      "grad_norm": 0.5538883029206723,
+      "learning_rate": 6.58064928903767e-06,
+      "loss": 0.6164,
+      "step": 470
+    },
+    {
+      "epoch": 1.3758519961051607,
+      "grad_norm": 0.5005412280561692,
+      "learning_rate": 6.56449455763323e-06,
+      "loss": 0.5576,
+      "step": 471
+    },
+    {
+      "epoch": 1.3787731256085687,
+      "grad_norm": 0.5868063901075268,
+      "learning_rate": 6.548321701564099e-06,
+      "loss": 0.5933,
+      "step": 472
+    },
+    {
+      "epoch": 1.3816942551119766,
+      "grad_norm": 0.5499777044713934,
+      "learning_rate": 6.5321309081927665e-06,
+      "loss": 0.562,
+      "step": 473
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5209457126764143,
+      "learning_rate": 6.515922365089524e-06,
+      "loss": 0.5793,
+      "step": 474
+    },
+    {
+      "epoch": 1.3875365141187925,
+      "grad_norm": 0.5045261657801648,
+      "learning_rate": 6.499696260030297e-06,
+      "loss": 0.594,
+      "step": 475
+    },
+    {
+      "epoch": 1.3904576436222005,
+      "grad_norm": 0.5243175331043878,
+      "learning_rate": 6.483452780994459e-06,
+      "loss": 0.534,
+      "step": 476
+    },
+    {
+      "epoch": 1.3933787731256086,
+      "grad_norm": 0.5169962119701911,
+      "learning_rate": 6.467192116162668e-06,
+      "loss": 0.5403,
+      "step": 477
+    },
+    {
+      "epoch": 1.3962999026290166,
+      "grad_norm": 0.5140189723701601,
+      "learning_rate": 6.450914453914674e-06,
+      "loss": 0.6058,
+      "step": 478
+    },
+    {
+      "epoch": 1.3992210321324245,
+      "grad_norm": 0.5330866177837568,
+      "learning_rate": 6.434619982827147e-06,
+      "loss": 0.5882,
+      "step": 479
+    },
+    {
+      "epoch": 1.4021421616358325,
+      "grad_norm": 0.5414814038608168,
+      "learning_rate": 6.418308891671484e-06,
+      "loss": 0.5954,
+      "step": 480
+    },
+    {
+      "epoch": 1.4050632911392404,
+      "grad_norm": 0.5376659643418761,
+      "learning_rate": 6.401981369411626e-06,
+      "loss": 0.5633,
+      "step": 481
+    },
+    {
+      "epoch": 1.4079844206426486,
+      "grad_norm": 0.5438938648229659,
+      "learning_rate": 6.385637605201871e-06,
+      "loss": 0.5677,
+      "step": 482
+    },
+    {
+      "epoch": 1.4109055501460563,
+      "grad_norm": 0.533597042074437,
+      "learning_rate": 6.3692777883846746e-06,
+      "loss": 0.6217,
+      "step": 483
+    },
+    {
+      "epoch": 1.4138266796494645,
+      "grad_norm": 0.5530416412659575,
+      "learning_rate": 6.3529021084884655e-06,
+      "loss": 0.5135,
+      "step": 484
+    },
+    {
+      "epoch": 1.4167478091528725,
+      "grad_norm": 0.6426214488830264,
+      "learning_rate": 6.336510755225447e-06,
+      "loss": 0.6039,
+      "step": 485
+    },
+    {
+      "epoch": 1.4196689386562804,
+      "grad_norm": 0.5321542466408685,
+      "learning_rate": 6.320103918489395e-06,
+      "loss": 0.5615,
+      "step": 486
+    },
+    {
+      "epoch": 1.4225900681596884,
+      "grad_norm": 0.5807878117878842,
+      "learning_rate": 6.303681788353465e-06,
+      "loss": 0.6417,
+      "step": 487
+    },
+    {
+      "epoch": 1.4255111976630963,
+      "grad_norm": 0.6006151574662654,
+      "learning_rate": 6.287244555067984e-06,
+      "loss": 0.5794,
+      "step": 488
+    },
+    {
+      "epoch": 1.4284323271665045,
+      "grad_norm": 0.5126167984608686,
+      "learning_rate": 6.270792409058247e-06,
+      "loss": 0.5242,
+      "step": 489
+    },
+    {
+      "epoch": 1.4313534566699124,
+      "grad_norm": 0.509995152360523,
+      "learning_rate": 6.25432554092232e-06,
+      "loss": 0.5828,
+      "step": 490
+    },
+    {
+      "epoch": 1.4342745861733204,
+      "grad_norm": 0.5960222110178056,
+      "learning_rate": 6.237844141428817e-06,
+      "loss": 0.568,
+      "step": 491
+    },
+    {
+      "epoch": 1.4371957156767283,
+      "grad_norm": 0.5173153132658744,
+      "learning_rate": 6.221348401514703e-06,
+      "loss": 0.5796,
+      "step": 492
+    },
+    {
+      "epoch": 1.4401168451801363,
+      "grad_norm": 0.5671800858222442,
+      "learning_rate": 6.204838512283073e-06,
+      "loss": 0.5506,
+      "step": 493
+    },
+    {
+      "epoch": 1.4430379746835442,
+      "grad_norm": 0.6040430910549448,
+      "learning_rate": 6.188314665000944e-06,
+      "loss": 0.5444,
+      "step": 494
+    },
+    {
+      "epoch": 1.4459591041869522,
+      "grad_norm": 0.5391742690410167,
+      "learning_rate": 6.171777051097037e-06,
+      "loss": 0.5649,
+      "step": 495
+    },
+    {
+      "epoch": 1.4488802336903603,
+      "grad_norm": 0.597749185130725,
+      "learning_rate": 6.155225862159558e-06,
+      "loss": 0.6408,
+      "step": 496
+    },
+    {
+      "epoch": 1.4518013631937683,
+      "grad_norm": 0.5229846439398993,
+      "learning_rate": 6.138661289933981e-06,
+      "loss": 0.5494,
+      "step": 497
+    },
+    {
+      "epoch": 1.4547224926971762,
+      "grad_norm": 0.5920796120572965,
+      "learning_rate": 6.1220835263208256e-06,
+      "loss": 0.5992,
+      "step": 498
+    },
+    {
+      "epoch": 1.4576436222005842,
+      "grad_norm": 0.5775415962111137,
+      "learning_rate": 6.105492763373431e-06,
+      "loss": 0.5911,
+      "step": 499
+    },
+    {
+      "epoch": 1.4605647517039921,
+      "grad_norm": 0.5917654314551983,
+      "learning_rate": 6.088889193295738e-06,
+      "loss": 0.5643,
+      "step": 500
+    },
+    {
+      "epoch": 1.4634858812074003,
+      "grad_norm": 0.5199838430286094,
+      "learning_rate": 6.072273008440052e-06,
+      "loss": 0.5472,
+      "step": 501
+    },
+    {
+      "epoch": 1.4664070107108083,
+      "grad_norm": 0.5522718232673819,
+      "learning_rate": 6.0556444013048265e-06,
+      "loss": 0.546,
+      "step": 502
+    },
+    {
+      "epoch": 1.4693281402142162,
+      "grad_norm": 0.525908351198303,
+      "learning_rate": 6.039003564532423e-06,
+      "loss": 0.6015,
+      "step": 503
+    },
+    {
+      "epoch": 1.4722492697176242,
+      "grad_norm": 0.5105926082337128,
+      "learning_rate": 6.0223506909068875e-06,
+      "loss": 0.5785,
+      "step": 504
+    },
+    {
+      "epoch": 1.475170399221032,
+      "grad_norm": 0.5832929477157227,
+      "learning_rate": 6.005685973351708e-06,
+      "loss": 0.6027,
+      "step": 505
+    },
+    {
+      "epoch": 1.47809152872444,
+      "grad_norm": 0.5447366515820563,
+      "learning_rate": 5.989009604927587e-06,
+      "loss": 0.5833,
+      "step": 506
+    },
+    {
+      "epoch": 1.481012658227848,
+      "grad_norm": 0.5245433114840506,
+      "learning_rate": 5.972321778830202e-06,
+      "loss": 0.4944,
+      "step": 507
+    },
+    {
+      "epoch": 1.4839337877312562,
+      "grad_norm": 0.5377622062276557,
+      "learning_rate": 5.9556226883879685e-06,
+      "loss": 0.5518,
+      "step": 508
+    },
+    {
+      "epoch": 1.4868549172346641,
+      "grad_norm": 0.6351508971275112,
+      "learning_rate": 5.938912527059798e-06,
+      "loss": 0.5982,
+      "step": 509
+    },
+    {
+      "epoch": 1.489776046738072,
+      "grad_norm": 0.5264532680548314,
+      "learning_rate": 5.922191488432857e-06,
+      "loss": 0.5472,
+      "step": 510
+    },
+    {
+      "epoch": 1.49269717624148,
+      "grad_norm": 0.5618733075790842,
+      "learning_rate": 5.90545976622033e-06,
+      "loss": 0.6387,
+      "step": 511
+    },
+    {
+      "epoch": 1.495618305744888,
+      "grad_norm": 0.5042404211846953,
+      "learning_rate": 5.888717554259165e-06,
+      "loss": 0.5188,
+      "step": 512
+    },
+    {
+      "epoch": 1.4985394352482961,
+      "grad_norm": 0.5996827758246558,
+      "learning_rate": 5.871965046507838e-06,
+      "loss": 0.5845,
+      "step": 513
+    },
+    {
+      "epoch": 1.5014605647517039,
+      "grad_norm": 0.513160385214818,
+      "learning_rate": 5.855202437044102e-06,
+      "loss": 0.5407,
+      "step": 514
+    },
+    {
+      "epoch": 1.504381694255112,
+      "grad_norm": 0.49930093522201097,
+      "learning_rate": 5.838429920062734e-06,
+      "loss": 0.5888,
+      "step": 515
+    },
+    {
+      "epoch": 1.50730282375852,
+      "grad_norm": 0.5820994183473087,
+      "learning_rate": 5.8216476898732935e-06,
+      "loss": 0.6026,
+      "step": 516
+    },
+    {
+      "epoch": 1.510223953261928,
+      "grad_norm": 0.5965529197255538,
+      "learning_rate": 5.804855940897866e-06,
+      "loss": 0.5772,
+      "step": 517
+    },
+    {
+      "epoch": 1.5131450827653359,
+      "grad_norm": 0.4891063083201174,
+      "learning_rate": 5.788054867668811e-06,
+      "loss": 0.5383,
+      "step": 518
+    },
+    {
+      "epoch": 1.5160662122687438,
+      "grad_norm": 0.549598064227309,
+      "learning_rate": 5.771244664826512e-06,
+      "loss": 0.5701,
+      "step": 519
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.5201662235289578,
+      "learning_rate": 5.754425527117118e-06,
+      "loss": 0.5221,
+      "step": 520
+    },
+    {
+      "epoch": 1.5219084712755597,
+      "grad_norm": 0.5727695969794172,
+      "learning_rate": 5.737597649390288e-06,
+      "loss": 0.5849,
+      "step": 521
+    },
+    {
+      "epoch": 1.524829600778968,
+      "grad_norm": 0.5387267148344115,
+      "learning_rate": 5.720761226596935e-06,
+      "loss": 0.5899,
+      "step": 522
+    },
+    {
+      "epoch": 1.5277507302823758,
+      "grad_norm": 0.5196417961644373,
+      "learning_rate": 5.703916453786965e-06,
+      "loss": 0.6075,
+      "step": 523
+    },
+    {
+      "epoch": 1.5306718597857838,
+      "grad_norm": 0.5162526391071218,
+      "learning_rate": 5.6870635261070186e-06,
+      "loss": 0.5524,
+      "step": 524
+    },
+    {
+      "epoch": 1.533592989289192,
+      "grad_norm": 0.5712502081281199,
+      "learning_rate": 5.670202638798213e-06,
+      "loss": 0.5639,
+      "step": 525
+    },
+    {
+      "epoch": 1.5365141187925997,
+      "grad_norm": 0.569067876773126,
+      "learning_rate": 5.653333987193876e-06,
+      "loss": 0.5437,
+      "step": 526
+    },
+    {
+      "epoch": 1.5394352482960079,
+      "grad_norm": 0.5925750045532109,
+      "learning_rate": 5.636457766717286e-06,
+      "loss": 0.591,
+      "step": 527
+    },
+    {
+      "epoch": 1.5423563777994158,
+      "grad_norm": 0.5141531017914309,
+      "learning_rate": 5.619574172879405e-06,
+      "loss": 0.5471,
+      "step": 528
+    },
+    {
+      "epoch": 1.5452775073028238,
+      "grad_norm": 0.6597290642553237,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.5901,
+      "step": 529
+    },
+    {
+      "epoch": 1.5481986368062317,
+      "grad_norm": 0.5884715259486929,
+      "learning_rate": 5.585785647588458e-06,
+      "loss": 0.5878,
+      "step": 530
+    },
+    {
+      "epoch": 1.5511197663096397,
+      "grad_norm": 0.5255721650946908,
+      "learning_rate": 5.568881107575353e-06,
+      "loss": 0.5447,
+      "step": 531
+    },
+    {
+      "epoch": 1.5540408958130478,
+      "grad_norm": 0.5091097959641512,
+      "learning_rate": 5.55196997707635e-06,
+      "loss": 0.6276,
+      "step": 532
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.5119232421542075,
+      "learning_rate": 5.53505245200684e-06,
+      "loss": 0.5658,
+      "step": 533
+    },
+    {
+      "epoch": 1.5598831548198637,
+      "grad_norm": 0.556141615427262,
+      "learning_rate": 5.518128728356303e-06,
+      "loss": 0.5951,
+      "step": 534
+    },
+    {
+      "epoch": 1.5628042843232717,
+      "grad_norm": 0.4969794250430925,
+      "learning_rate": 5.501199002186024e-06,
+      "loss": 0.5457,
+      "step": 535
+    },
+    {
+      "epoch": 1.5657254138266796,
+      "grad_norm": 0.5377430121732353,
+      "learning_rate": 5.48426346962683e-06,
+      "loss": 0.587,
+      "step": 536
+    },
+    {
+      "epoch": 1.5686465433300878,
+      "grad_norm": 0.4763376786579273,
+      "learning_rate": 5.467322326876813e-06,
+      "loss": 0.5352,
+      "step": 537
+    },
+    {
+      "epoch": 1.5715676728334955,
+      "grad_norm": 0.5630031793193719,
+      "learning_rate": 5.450375770199063e-06,
+      "loss": 0.652,
+      "step": 538
+    },
+    {
+      "epoch": 1.5744888023369037,
+      "grad_norm": 0.505233569384271,
+      "learning_rate": 5.433423995919383e-06,
+      "loss": 0.5633,
+      "step": 539
+    },
+    {
+      "epoch": 1.5774099318403116,
+      "grad_norm": 0.4918114442897718,
+      "learning_rate": 5.416467200424032e-06,
+      "loss": 0.5422,
+      "step": 540
+    },
+    {
+      "epoch": 1.5803310613437196,
+      "grad_norm": 0.5991334895146722,
+      "learning_rate": 5.399505580157428e-06,
+      "loss": 0.558,
+      "step": 541
+    },
+    {
+      "epoch": 1.5832521908471275,
+      "grad_norm": 0.5363474139996335,
+      "learning_rate": 5.382539331619896e-06,
+      "loss": 0.5699,
+      "step": 542
+    },
+    {
+      "epoch": 1.5861733203505355,
+      "grad_norm": 0.5807115654118472,
+      "learning_rate": 5.365568651365369e-06,
+      "loss": 0.5688,
+      "step": 543
+    },
+    {
+      "epoch": 1.5890944498539437,
+      "grad_norm": 0.5904366045058206,
+      "learning_rate": 5.34859373599913e-06,
+      "loss": 0.5969,
+      "step": 544
+    },
+    {
+      "epoch": 1.5920155793573514,
+      "grad_norm": 0.491932507277215,
+      "learning_rate": 5.33161478217552e-06,
+      "loss": 0.5542,
+      "step": 545
+    },
+    {
+      "epoch": 1.5949367088607596,
+      "grad_norm": 0.5011301426277018,
+      "learning_rate": 5.314631986595669e-06,
+      "loss": 0.569,
+      "step": 546
+    },
+    {
+      "epoch": 1.5978578383641675,
+      "grad_norm": 0.6106059864005979,
+      "learning_rate": 5.297645546005208e-06,
+      "loss": 0.6051,
+      "step": 547
+    },
+    {
+      "epoch": 1.6007789678675755,
+      "grad_norm": 0.5958177689979132,
+      "learning_rate": 5.280655657192003e-06,
+      "loss": 0.5266,
+      "step": 548
+    },
+    {
+      "epoch": 1.6037000973709834,
+      "grad_norm": 0.5395470595824067,
+      "learning_rate": 5.263662516983863e-06,
+      "loss": 0.5657,
+      "step": 549
+    },
+    {
+      "epoch": 1.6066212268743914,
+      "grad_norm": 0.6212970783782318,
+      "learning_rate": 5.246666322246267e-06,
+      "loss": 0.5619,
+      "step": 550
+    },
+    {
+      "epoch": 1.6095423563777995,
+      "grad_norm": 0.6613037032143712,
+      "learning_rate": 5.229667269880078e-06,
+      "loss": 0.6092,
+      "step": 551
+    },
+    {
+      "epoch": 1.6124634858812072,
+      "grad_norm": 0.48332526807644743,
+      "learning_rate": 5.212665556819264e-06,
+      "loss": 0.5401,
+      "step": 552
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.5325857980417065,
+      "learning_rate": 5.195661380028625e-06,
+      "loss": 0.6035,
+      "step": 553
+    },
+    {
+      "epoch": 1.6183057448880234,
+      "grad_norm": 0.5442785714221408,
+      "learning_rate": 5.178654936501492e-06,
+      "loss": 0.5494,
+      "step": 554
+    },
+    {
+      "epoch": 1.6212268743914313,
+      "grad_norm": 0.6437807568653632,
+      "learning_rate": 5.1616464232574635e-06,
+      "loss": 0.6254,
+      "step": 555
+    },
+    {
+      "epoch": 1.6241480038948395,
+      "grad_norm": 0.4492125406814553,
+      "learning_rate": 5.1446360373401125e-06,
+      "loss": 0.5301,
+      "step": 556
+    },
+    {
+      "epoch": 1.6270691333982472,
+      "grad_norm": 0.5622471364004906,
+      "learning_rate": 5.127623975814709e-06,
+      "loss": 0.5715,
+      "step": 557
+    },
+    {
+      "epoch": 1.6299902629016554,
+      "grad_norm": 0.6240851711756873,
+      "learning_rate": 5.110610435765935e-06,
+      "loss": 0.5876,
+      "step": 558
+    },
+    {
+      "epoch": 1.6329113924050633,
+      "grad_norm": 0.4713533254066173,
+      "learning_rate": 5.093595614295599e-06,
+      "loss": 0.4901,
+      "step": 559
+    },
+    {
+      "epoch": 1.6358325219084713,
+      "grad_norm": 0.5796358881475803,
+      "learning_rate": 5.076579708520355e-06,
+      "loss": 0.6325,
+      "step": 560
+    },
+    {
+      "epoch": 1.6387536514118792,
+      "grad_norm": 0.4866837205814519,
+      "learning_rate": 5.059562915569424e-06,
+      "loss": 0.5202,
+      "step": 561
+    },
+    {
+      "epoch": 1.6416747809152872,
+      "grad_norm": 0.5697352103252339,
+      "learning_rate": 5.0425454325822946e-06,
+      "loss": 0.6339,
+      "step": 562
+    },
+    {
+      "epoch": 1.6445959104186954,
+      "grad_norm": 0.5721771538016244,
+      "learning_rate": 5.0255274567064594e-06,
+      "loss": 0.5364,
+      "step": 563
+    },
+    {
+      "epoch": 1.647517039922103,
+      "grad_norm": 0.5605730886016151,
+      "learning_rate": 5.008509185095114e-06,
+      "loss": 0.5717,
+      "step": 564
+    },
+    {
+      "epoch": 1.6504381694255112,
+      "grad_norm": 0.472681806190808,
+      "learning_rate": 4.991490814904888e-06,
+      "loss": 0.555,
+      "step": 565
+    },
+    {
+      "epoch": 1.6533592989289192,
+      "grad_norm": 0.5791761573446532,
+      "learning_rate": 4.974472543293544e-06,
+      "loss": 0.6077,
+      "step": 566
+    },
+    {
+      "epoch": 1.6562804284323271,
+      "grad_norm": 0.5167447517838185,
+      "learning_rate": 4.957454567417708e-06,
+      "loss": 0.5805,
+      "step": 567
+    },
+    {
+      "epoch": 1.6592015579357353,
+      "grad_norm": 0.5561818381815233,
+      "learning_rate": 4.940437084430579e-06,
+      "loss": 0.594,
+      "step": 568
+    },
+    {
+      "epoch": 1.662122687439143,
+      "grad_norm": 0.44095968784824047,
+      "learning_rate": 4.923420291479646e-06,
+      "loss": 0.5235,
+      "step": 569
+    },
+    {
+      "epoch": 1.6650438169425512,
+      "grad_norm": 0.49918372558187907,
+      "learning_rate": 4.906404385704402e-06,
+      "loss": 0.6011,
+      "step": 570
+    },
+    {
+      "epoch": 1.667964946445959,
+      "grad_norm": 0.5067863457643954,
+      "learning_rate": 4.8893895642340665e-06,
+      "loss": 0.5415,
+      "step": 571
+    },
+    {
+      "epoch": 1.6708860759493671,
+      "grad_norm": 0.49737844087035643,
+      "learning_rate": 4.872376024185291e-06,
+      "loss": 0.583,
+      "step": 572
+    },
+    {
+      "epoch": 1.673807205452775,
+      "grad_norm": 0.4659962080708305,
+      "learning_rate": 4.855363962659889e-06,
+      "loss": 0.5627,
+      "step": 573
+    },
+    {
+      "epoch": 1.676728334956183,
+      "grad_norm": 0.4972582768404136,
+      "learning_rate": 4.838353576742538e-06,
+      "loss": 0.5911,
+      "step": 574
+    },
+    {
+      "epoch": 1.6796494644595912,
+      "grad_norm": 0.44592731268572705,
+      "learning_rate": 4.82134506349851e-06,
+      "loss": 0.5459,
+      "step": 575
+    },
+    {
+      "epoch": 1.682570593962999,
+      "grad_norm": 0.506185991053395,
+      "learning_rate": 4.804338619971377e-06,
+      "loss": 0.5306,
+      "step": 576
+    },
+    {
+      "epoch": 1.685491723466407,
+      "grad_norm": 0.5177923049058465,
+      "learning_rate": 4.787334443180735e-06,
+      "loss": 0.6209,
+      "step": 577
+    },
+    {
+      "epoch": 1.688412852969815,
+      "grad_norm": 0.4601309987230691,
+      "learning_rate": 4.7703327301199244e-06,
+      "loss": 0.5145,
+      "step": 578
+    },
+    {
+      "epoch": 1.691333982473223,
+      "grad_norm": 0.4643978865329562,
+      "learning_rate": 4.753333677753734e-06,
+      "loss": 0.5788,
+      "step": 579
+    },
+    {
+      "epoch": 1.694255111976631,
+      "grad_norm": 0.5283219640600681,
+      "learning_rate": 4.736337483016138e-06,
+      "loss": 0.6056,
+      "step": 580
+    },
+    {
+      "epoch": 1.6971762414800389,
+      "grad_norm": 0.5571423915108132,
+      "learning_rate": 4.719344342808e-06,
+      "loss": 0.5791,
+      "step": 581
+    },
+    {
+      "epoch": 1.700097370983447,
+      "grad_norm": 0.49563030247464135,
+      "learning_rate": 4.702354453994794e-06,
+      "loss": 0.5373,
+      "step": 582
+    },
+    {
+      "epoch": 1.7030185004868548,
+      "grad_norm": 0.49902926183967655,
+      "learning_rate": 4.6853680134043345e-06,
+      "loss": 0.5823,
+      "step": 583
+    },
+    {
+      "epoch": 1.705939629990263,
+      "grad_norm": 0.5116289187531037,
+      "learning_rate": 4.668385217824482e-06,
+      "loss": 0.5557,
+      "step": 584
+    },
+    {
+      "epoch": 1.7088607594936709,
+      "grad_norm": 0.5412976872055105,
+      "learning_rate": 4.651406264000871e-06,
+      "loss": 0.5535,
+      "step": 585
+    },
+    {
+      "epoch": 1.7117818889970788,
+      "grad_norm": 0.5297049362432723,
+      "learning_rate": 4.634431348634632e-06,
+      "loss": 0.5872,
+      "step": 586
+    },
+    {
+      "epoch": 1.714703018500487,
+      "grad_norm": 0.46815419228452076,
+      "learning_rate": 4.6174606683801055e-06,
+      "loss": 0.5612,
+      "step": 587
+    },
+    {
+      "epoch": 1.7176241480038947,
+      "grad_norm": 0.5047028920506614,
+      "learning_rate": 4.600494419842573e-06,
+      "loss": 0.5457,
+      "step": 588
+    },
+    {
+      "epoch": 1.720545277507303,
+      "grad_norm": 0.5021848097551853,
+      "learning_rate": 4.58353279957597e-06,
+      "loss": 0.5512,
+      "step": 589
+    },
+    {
+      "epoch": 1.7234664070107109,
+      "grad_norm": 0.5567736723329815,
+      "learning_rate": 4.5665760040806174e-06,
+      "loss": 0.577,
+      "step": 590
+    },
+    {
+      "epoch": 1.7263875365141188,
+      "grad_norm": 0.4982671338092423,
+      "learning_rate": 4.549624229800938e-06,
+      "loss": 0.548,
+      "step": 591
+    },
+    {
+      "epoch": 1.7293086660175268,
+      "grad_norm": 0.49329094074335034,
+      "learning_rate": 4.532677673123188e-06,
+      "loss": 0.5563,
+      "step": 592
+    },
+    {
+      "epoch": 1.7322297955209347,
+      "grad_norm": 0.4890277332079673,
+      "learning_rate": 4.5157365303731705e-06,
+      "loss": 0.5747,
+      "step": 593
+    },
+    {
+      "epoch": 1.7351509250243429,
+      "grad_norm": 0.5600537499346627,
+      "learning_rate": 4.498800997813976e-06,
+      "loss": 0.5816,
+      "step": 594
+    },
+    {
+      "epoch": 1.7380720545277506,
+      "grad_norm": 0.5086959673884651,
+      "learning_rate": 4.481871271643698e-06,
+      "loss": 0.5671,
+      "step": 595
+    },
+    {
+      "epoch": 1.7409931840311588,
+      "grad_norm": 0.5157431177024923,
+      "learning_rate": 4.464947547993162e-06,
+      "loss": 0.5731,
+      "step": 596
+    },
+    {
+      "epoch": 1.7439143135345667,
+      "grad_norm": 0.5137961841228513,
+      "learning_rate": 4.4480300229236525e-06,
+      "loss": 0.544,
+      "step": 597
+    },
+    {
+      "epoch": 1.7468354430379747,
+      "grad_norm": 0.5937636838238008,
+      "learning_rate": 4.431118892424649e-06,
+      "loss": 0.6082,
+      "step": 598
+    },
+    {
+      "epoch": 1.7497565725413826,
+      "grad_norm": 0.5529433304939138,
+      "learning_rate": 4.414214352411544e-06,
+      "loss": 0.5569,
+      "step": 599
+    },
+    {
+      "epoch": 1.7526777020447906,
+      "grad_norm": 0.5159596510951358,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.5455,
+      "step": 600
+    },
+    {
+      "epoch": 1.7555988315481987,
+      "grad_norm": 0.5506758519748508,
+      "learning_rate": 4.3804258271205965e-06,
+      "loss": 0.5505,
+      "step": 601
+    },
+    {
+      "epoch": 1.7585199610516065,
+      "grad_norm": 0.5523849090934733,
+      "learning_rate": 4.363542233282715e-06,
+      "loss": 0.5355,
+      "step": 602
+    },
+    {
+      "epoch": 1.7614410905550146,
+      "grad_norm": 0.6508686455735765,
+      "learning_rate": 4.346666012806126e-06,
+      "loss": 0.602,
+      "step": 603
+    },
+    {
+      "epoch": 1.7643622200584226,
+      "grad_norm": 0.46084404235460724,
+      "learning_rate": 4.329797361201788e-06,
+      "loss": 0.5028,
+      "step": 604
+    },
+    {
+      "epoch": 1.7672833495618305,
+      "grad_norm": 0.5857372990733761,
+      "learning_rate": 4.312936473892984e-06,
+      "loss": 0.6468,
+      "step": 605
+    },
+    {
+      "epoch": 1.7702044790652387,
+      "grad_norm": 0.4935847914878853,
+      "learning_rate": 4.296083546213037e-06,
+      "loss": 0.5578,
+      "step": 606
+    },
+    {
+      "epoch": 1.7731256085686464,
+      "grad_norm": 0.5144278362398599,
+      "learning_rate": 4.279238773403066e-06,
+      "loss": 0.5337,
+      "step": 607
+    },
+    {
+      "epoch": 1.7760467380720546,
+      "grad_norm": 0.46428380216043086,
+      "learning_rate": 4.2624023506097116e-06,
+      "loss": 0.6219,
+      "step": 608
+    },
+    {
+      "epoch": 1.7789678675754625,
+      "grad_norm": 0.4492920362378749,
+      "learning_rate": 4.245574472882882e-06,
+      "loss": 0.5061,
+      "step": 609
+    },
+    {
+      "epoch": 1.7818889970788705,
+      "grad_norm": 0.48814416218826545,
+      "learning_rate": 4.228755335173488e-06,
+      "loss": 0.5045,
+      "step": 610
+    },
+    {
+      "epoch": 1.7848101265822784,
+      "grad_norm": 0.4935317746542341,
+      "learning_rate": 4.21194513233119e-06,
+      "loss": 0.6334,
+      "step": 611
+    },
+    {
+      "epoch": 1.7877312560856864,
+      "grad_norm": 0.5119284354029839,
+      "learning_rate": 4.1951440591021375e-06,
+      "loss": 0.5599,
+      "step": 612
+    },
+    {
+      "epoch": 1.7906523855890946,
+      "grad_norm": 0.4827371863106343,
+      "learning_rate": 4.17835231012671e-06,
+      "loss": 0.5692,
+      "step": 613
+    },
+    {
+      "epoch": 1.7935735150925023,
+      "grad_norm": 0.5212508350905807,
+      "learning_rate": 4.161570079937268e-06,
+      "loss": 0.5947,
+      "step": 614
+    },
+    {
+      "epoch": 1.7964946445959105,
+      "grad_norm": 0.5235515777898554,
+      "learning_rate": 4.1447975629559e-06,
+      "loss": 0.5808,
+      "step": 615
+    },
+    {
+      "epoch": 1.7994157740993184,
+      "grad_norm": 0.5336005328533747,
+      "learning_rate": 4.128034953492163e-06,
+      "loss": 0.5358,
+      "step": 616
+    },
+    {
+      "epoch": 1.8023369036027264,
+      "grad_norm": 0.5597776990519019,
+      "learning_rate": 4.1112824457408375e-06,
+      "loss": 0.5884,
+      "step": 617
+    },
+    {
+      "epoch": 1.8052580331061345,
+      "grad_norm": 0.5036193557727575,
+      "learning_rate": 4.094540233779672e-06,
+      "loss": 0.5688,
+      "step": 618
+    },
+    {
+      "epoch": 1.8081791626095423,
+      "grad_norm": 0.5338396120419266,
+      "learning_rate": 4.077808511567145e-06,
+      "loss": 0.5608,
+      "step": 619
+    },
+    {
+      "epoch": 1.8111002921129504,
+      "grad_norm": 0.4958715717085618,
+      "learning_rate": 4.061087472940204e-06,
+      "loss": 0.5543,
+      "step": 620
+    },
+    {
+      "epoch": 1.8140214216163584,
+      "grad_norm": 0.6029334087991165,
+      "learning_rate": 4.044377311612033e-06,
+      "loss": 0.5691,
+      "step": 621
+    },
+    {
+      "epoch": 1.8169425511197663,
+      "grad_norm": 0.47915777984754265,
+      "learning_rate": 4.0276782211698e-06,
+      "loss": 0.5391,
+      "step": 622
+    },
+    {
+      "epoch": 1.8198636806231743,
+      "grad_norm": 0.5235008297199194,
+      "learning_rate": 4.010990395072414e-06,
+      "loss": 0.5736,
+      "step": 623
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5590747505915074,
+      "learning_rate": 3.9943140266482935e-06,
+      "loss": 0.6075,
+      "step": 624
+    },
+    {
+      "epoch": 1.8257059396299904,
+      "grad_norm": 0.47598042280541497,
+      "learning_rate": 3.977649309093113e-06,
+      "loss": 0.5344,
+      "step": 625
+    },
+    {
+      "epoch": 1.8286270691333981,
+      "grad_norm": 0.48621841809715133,
+      "learning_rate": 3.960996435467577e-06,
+      "loss": 0.6179,
+      "step": 626
+    },
+    {
+      "epoch": 1.8315481986368063,
+      "grad_norm": 0.5017298117872958,
+      "learning_rate": 3.944355598695174e-06,
+      "loss": 0.5443,
+      "step": 627
+    },
+    {
+      "epoch": 1.8344693281402142,
+      "grad_norm": 0.5380565465468942,
+      "learning_rate": 3.9277269915599505e-06,
+      "loss": 0.57,
+      "step": 628
+    },
+    {
+      "epoch": 1.8373904576436222,
+      "grad_norm": 0.5990209575217125,
+      "learning_rate": 3.911110806704265e-06,
+      "loss": 0.5784,
+      "step": 629
+    },
+    {
+      "epoch": 1.8403115871470301,
+      "grad_norm": 0.49509817541993284,
+      "learning_rate": 3.89450723662657e-06,
+      "loss": 0.5628,
+      "step": 630
+    },
+    {
+      "epoch": 1.843232716650438,
+      "grad_norm": 0.550605497861562,
+      "learning_rate": 3.877916473679176e-06,
+      "loss": 0.6147,
+      "step": 631
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.5524680706044535,
+      "learning_rate": 3.861338710066021e-06,
+      "loss": 0.5125,
+      "step": 632
+    },
+    {
+      "epoch": 1.849074975657254,
+      "grad_norm": 0.5494313577644265,
+      "learning_rate": 3.8447741378404436e-06,
+      "loss": 0.5562,
+      "step": 633
+    },
+    {
+      "epoch": 1.8519961051606622,
+      "grad_norm": 0.5058622045652387,
+      "learning_rate": 3.828222948902965e-06,
+      "loss": 0.5967,
+      "step": 634
+    },
+    {
+      "epoch": 1.85491723466407,
+      "grad_norm": 0.4488395706127018,
+      "learning_rate": 3.8116853349990574e-06,
+      "loss": 0.4735,
+      "step": 635
+    },
+    {
+      "epoch": 1.857838364167478,
+      "grad_norm": 0.5073104983823334,
+      "learning_rate": 3.7951614877169285e-06,
+      "loss": 0.6214,
+      "step": 636
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.5020623526859631,
+      "learning_rate": 3.7786515984852977e-06,
+      "loss": 0.5801,
+      "step": 637
+    },
+    {
+      "epoch": 1.863680623174294,
+      "grad_norm": 0.5066558405528695,
+      "learning_rate": 3.762155858571184e-06,
+      "loss": 0.5571,
+      "step": 638
+    },
+    {
+      "epoch": 1.8666017526777021,
+      "grad_norm": 0.5016476079074594,
+      "learning_rate": 3.7456744590776807e-06,
+      "loss": 0.554,
+      "step": 639
+    },
+    {
+      "epoch": 1.86952288218111,
+      "grad_norm": 0.5522667996432958,
+      "learning_rate": 3.729207590941753e-06,
+      "loss": 0.5684,
+      "step": 640
+    },
+    {
+      "epoch": 1.872444011684518,
+      "grad_norm": 0.5063071640192653,
+      "learning_rate": 3.712755444932018e-06,
+      "loss": 0.5536,
+      "step": 641
+    },
+    {
+      "epoch": 1.875365141187926,
+      "grad_norm": 0.47979895935138983,
+      "learning_rate": 3.6963182116465358e-06,
+      "loss": 0.5607,
+      "step": 642
+    },
+    {
+      "epoch": 1.878286270691334,
+      "grad_norm": 0.5466280214439997,
+      "learning_rate": 3.6798960815106065e-06,
+      "loss": 0.5912,
+      "step": 643
+    },
+    {
+      "epoch": 1.881207400194742,
+      "grad_norm": 0.50137359619968,
+      "learning_rate": 3.663489244774555e-06,
+      "loss": 0.6032,
+      "step": 644
+    },
+    {
+      "epoch": 1.8841285296981498,
+      "grad_norm": 0.5176702766961846,
+      "learning_rate": 3.647097891511536e-06,
+      "loss": 0.5441,
+      "step": 645
+    },
+    {
+      "epoch": 1.887049659201558,
+      "grad_norm": 0.45011045640480307,
+      "learning_rate": 3.630722211615328e-06,
+      "loss": 0.5064,
+      "step": 646
+    },
+    {
+      "epoch": 1.889970788704966,
+      "grad_norm": 0.5394614468332969,
+      "learning_rate": 3.614362394798131e-06,
+      "loss": 0.6368,
+      "step": 647
+    },
+    {
+      "epoch": 1.8928919182083739,
+      "grad_norm": 0.4874425981153431,
+      "learning_rate": 3.5980186305883746e-06,
+      "loss": 0.5701,
+      "step": 648
+    },
+    {
+      "epoch": 1.895813047711782,
+      "grad_norm": 0.507766689537816,
+      "learning_rate": 3.5816911083285165e-06,
+      "loss": 0.5798,
+      "step": 649
+    },
+    {
+      "epoch": 1.8987341772151898,
+      "grad_norm": 0.4863274459783003,
+      "learning_rate": 3.565380017172854e-06,
+      "loss": 0.5244,
+      "step": 650
+    },
+    {
+      "epoch": 1.901655306718598,
+      "grad_norm": 0.5805740988868203,
+      "learning_rate": 3.5490855460853275e-06,
+      "loss": 0.6026,
+      "step": 651
+    },
+    {
+      "epoch": 1.904576436222006,
+      "grad_norm": 0.5151920826178348,
+      "learning_rate": 3.5328078838373338e-06,
+      "loss": 0.5578,
+      "step": 652
+    },
+    {
+      "epoch": 1.9074975657254138,
+      "grad_norm": 0.4988300821200546,
+      "learning_rate": 3.516547219005542e-06,
+      "loss": 0.5639,
+      "step": 653
+    },
+    {
+      "epoch": 1.9104186952288218,
+      "grad_norm": 0.5020574341639038,
+      "learning_rate": 3.500303739969704e-06,
+      "loss": 0.5553,
+      "step": 654
+    },
+    {
+      "epoch": 1.9133398247322297,
+      "grad_norm": 0.47286299047516567,
+      "learning_rate": 3.4840776349104755e-06,
+      "loss": 0.5649,
+      "step": 655
+    },
+    {
+      "epoch": 1.916260954235638,
+      "grad_norm": 0.4681307130388586,
+      "learning_rate": 3.4678690918072335e-06,
+      "loss": 0.5682,
+      "step": 656
+    },
+    {
+      "epoch": 1.9191820837390456,
+      "grad_norm": 0.46768532251073486,
+      "learning_rate": 3.451678298435902e-06,
+      "loss": 0.5486,
+      "step": 657
+    },
+    {
+      "epoch": 1.9221032132424538,
+      "grad_norm": 0.4674675744224569,
+      "learning_rate": 3.4355054423667712e-06,
+      "loss": 0.5421,
+      "step": 658
+    },
+    {
+      "epoch": 1.9250243427458618,
+      "grad_norm": 0.5073113729998869,
+      "learning_rate": 3.4193507109623323e-06,
+      "loss": 0.5802,
+      "step": 659
+    },
+    {
+      "epoch": 1.9279454722492697,
+      "grad_norm": 0.47796072007470125,
+      "learning_rate": 3.4032142913750956e-06,
+      "loss": 0.5454,
+      "step": 660
+    },
+    {
+      "epoch": 1.9308666017526777,
+      "grad_norm": 0.44978612526614165,
+      "learning_rate": 3.3870963705454353e-06,
+      "loss": 0.6016,
+      "step": 661
+    },
+    {
+      "epoch": 1.9337877312560856,
+      "grad_norm": 0.49209754086259494,
+      "learning_rate": 3.370997135199413e-06,
+      "loss": 0.5898,
+      "step": 662
+    },
+    {
+      "epoch": 1.9367088607594938,
+      "grad_norm": 0.5261340174516683,
+      "learning_rate": 3.3549167718466245e-06,
+      "loss": 0.5438,
+      "step": 663
+    },
+    {
+      "epoch": 1.9396299902629015,
+      "grad_norm": 0.5215555650313894,
+      "learning_rate": 3.3388554667780272e-06,
+      "loss": 0.59,
+      "step": 664
+    },
+    {
+      "epoch": 1.9425511197663097,
+      "grad_norm": 0.4360310531256464,
+      "learning_rate": 3.322813406063794e-06,
+      "loss": 0.5361,
+      "step": 665
+    },
+    {
+      "epoch": 1.9454722492697176,
+      "grad_norm": 0.4338971494593997,
+      "learning_rate": 3.3067907755511473e-06,
+      "loss": 0.5217,
+      "step": 666
+    },
+    {
+      "epoch": 1.9483933787731256,
+      "grad_norm": 0.5993978224848855,
+      "learning_rate": 3.290787760862215e-06,
+      "loss": 0.6499,
+      "step": 667
+    },
+    {
+      "epoch": 1.9513145082765337,
+      "grad_norm": 0.4277968919618892,
+      "learning_rate": 3.274804547391872e-06,
+      "loss": 0.5059,
+      "step": 668
+    },
+    {
+      "epoch": 1.9542356377799415,
+      "grad_norm": 0.4833235638298786,
+      "learning_rate": 3.258841320305598e-06,
+      "loss": 0.6043,
+      "step": 669
+    },
+    {
+      "epoch": 1.9571567672833496,
+      "grad_norm": 0.4876241950113585,
+      "learning_rate": 3.242898264537331e-06,
+      "loss": 0.5982,
+      "step": 670
+    },
+    {
+      "epoch": 1.9600778967867576,
+      "grad_norm": 0.4287569825122473,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.5291,
+      "step": 671
+    },
+    {
+      "epoch": 1.9629990262901655,
+      "grad_norm": 0.559198431896861,
+      "learning_rate": 3.211073405520001e-06,
+      "loss": 0.5722,
+      "step": 672
+    },
+    {
+      "epoch": 1.9659201557935735,
+      "grad_norm": 0.5529747384120326,
+      "learning_rate": 3.195191970961833e-06,
+      "loss": 0.5766,
+      "step": 673
+    },
+    {
+      "epoch": 1.9688412852969814,
+      "grad_norm": 0.4620747878088523,
+      "learning_rate": 3.1793314450991895e-06,
+      "loss": 0.5911,
+      "step": 674
+    },
+    {
+      "epoch": 1.9717624148003896,
+      "grad_norm": 0.46006255679284874,
+      "learning_rate": 3.1634920116762175e-06,
+      "loss": 0.5622,
+      "step": 675
+    },
+    {
+      "epoch": 1.9746835443037973,
+      "grad_norm": 0.4259271639046577,
+      "learning_rate": 3.1476738541926993e-06,
+      "loss": 0.5287,
+      "step": 676
+    },
+    {
+      "epoch": 1.9776046738072055,
+      "grad_norm": 0.4729000058273843,
+      "learning_rate": 3.1318771559019455e-06,
+      "loss": 0.5863,
+      "step": 677
+    },
+    {
+      "epoch": 1.9805258033106135,
+      "grad_norm": 0.5100371866983662,
+      "learning_rate": 3.1161020998086566e-06,
+      "loss": 0.5687,
+      "step": 678
+    },
+    {
+      "epoch": 1.9834469328140214,
+      "grad_norm": 0.4379626484222088,
+      "learning_rate": 3.1003488686668076e-06,
+      "loss": 0.5155,
+      "step": 679
+    },
+    {
+      "epoch": 1.9863680623174296,
+      "grad_norm": 0.4844416917506815,
+      "learning_rate": 3.0846176449775363e-06,
+      "loss": 0.5782,
+      "step": 680
+    },
+    {
+      "epoch": 1.9892891918208373,
+      "grad_norm": 0.48815074844900985,
+      "learning_rate": 3.0689086109870188e-06,
+      "loss": 0.5709,
+      "step": 681
+    },
+    {
+      "epoch": 1.9922103213242455,
+      "grad_norm": 0.5064903715857387,
+      "learning_rate": 3.0532219486843686e-06,
+      "loss": 0.596,
+      "step": 682
+    },
+    {
+      "epoch": 1.9951314508276532,
+      "grad_norm": 0.492762116678178,
+      "learning_rate": 3.0375578397995178e-06,
+      "loss": 0.5766,
+      "step": 683
+    },
+    {
+      "epoch": 1.9980525803310614,
+      "grad_norm": 0.44675106308795903,
+      "learning_rate": 3.021916465801122e-06,
+      "loss": 0.5206,
+      "step": 684
+    },
+    {
+      "epoch": 2.0009737098344695,
+      "grad_norm": 0.9800800703433351,
+      "learning_rate": 3.0062980078944515e-06,
+      "loss": 0.8991,
+      "step": 685
+    },
+    {
+      "epoch": 2.0038948393378773,
+      "grad_norm": 0.4864022689636578,
+      "learning_rate": 2.990702647019294e-06,
+      "loss": 0.4938,
+      "step": 686
+    },
+    {
+      "epoch": 2.0068159688412854,
+      "grad_norm": 0.47889455472294706,
+      "learning_rate": 2.9751305638478555e-06,
+      "loss": 0.5913,
+      "step": 687
+    },
+    {
+      "epoch": 2.009737098344693,
+      "grad_norm": 0.41456980849801855,
+      "learning_rate": 2.9595819387826753e-06,
+      "loss": 0.4387,
+      "step": 688
+    },
+    {
+      "epoch": 2.0126582278481013,
+      "grad_norm": 0.5149845371997606,
+      "learning_rate": 2.9440569519545258e-06,
+      "loss": 0.5165,
+      "step": 689
+    },
+    {
+      "epoch": 2.015579357351509,
+      "grad_norm": 0.5752113401087547,
+      "learning_rate": 2.9285557832203328e-06,
+      "loss": 0.5373,
+      "step": 690
+    },
+    {
+      "epoch": 2.0185004868549172,
+      "grad_norm": 0.5062618329379694,
+      "learning_rate": 2.9130786121610866e-06,
+      "loss": 0.4947,
+      "step": 691
+    },
+    {
+      "epoch": 2.0214216163583254,
+      "grad_norm": 0.49684882097652555,
+      "learning_rate": 2.897625618079769e-06,
+      "loss": 0.5208,
+      "step": 692
+    },
+    {
+      "epoch": 2.024342745861733,
+      "grad_norm": 0.53299035788322,
+      "learning_rate": 2.88219697999927e-06,
+      "loss": 0.5438,
+      "step": 693
+    },
+    {
+      "epoch": 2.0272638753651413,
+      "grad_norm": 0.5543675569999611,
+      "learning_rate": 2.8667928766603115e-06,
+      "loss": 0.5444,
+      "step": 694
+    },
+    {
+      "epoch": 2.030185004868549,
+      "grad_norm": 0.4572356814121919,
+      "learning_rate": 2.851413486519388e-06,
+      "loss": 0.4634,
+      "step": 695
+    },
+    {
+      "epoch": 2.033106134371957,
+      "grad_norm": 0.48340574114252344,
+      "learning_rate": 2.8360589877466848e-06,
+      "loss": 0.5391,
+      "step": 696
+    },
+    {
+      "epoch": 2.036027263875365,
+      "grad_norm": 0.5141120553439197,
+      "learning_rate": 2.8207295582240248e-06,
+      "loss": 0.4871,
+      "step": 697
+    },
+    {
+      "epoch": 2.038948393378773,
+      "grad_norm": 0.4913089013833701,
+      "learning_rate": 2.8054253755428017e-06,
+      "loss": 0.5574,
+      "step": 698
+    },
+    {
+      "epoch": 2.0418695228821813,
+      "grad_norm": 0.4601265937011902,
+      "learning_rate": 2.7901466170019242e-06,
+      "loss": 0.4977,
+      "step": 699
+    },
+    {
+      "epoch": 2.044790652385589,
+      "grad_norm": 0.5120604434660413,
+      "learning_rate": 2.774893459605766e-06,
+      "loss": 0.56,
+      "step": 700
+    },
+    {
+      "epoch": 2.047711781888997,
+      "grad_norm": 0.47951033771571855,
+      "learning_rate": 2.7596660800621076e-06,
+      "loss": 0.5028,
+      "step": 701
+    },
+    {
+      "epoch": 2.050632911392405,
+      "grad_norm": 0.48501372243432067,
+      "learning_rate": 2.744464654780096e-06,
+      "loss": 0.5174,
+      "step": 702
+    },
+    {
+      "epoch": 2.053554040895813,
+      "grad_norm": 0.4741865276944199,
+      "learning_rate": 2.7292893598681934e-06,
+      "loss": 0.5104,
+      "step": 703
+    },
+    {
+      "epoch": 2.0564751703992212,
+      "grad_norm": 0.48778437702309935,
+      "learning_rate": 2.714140371132149e-06,
+      "loss": 0.5375,
+      "step": 704
+    },
+    {
+      "epoch": 2.059396299902629,
+      "grad_norm": 0.49018615472755583,
+      "learning_rate": 2.69901786407295e-06,
+      "loss": 0.5105,
+      "step": 705
+    },
+    {
+      "epoch": 2.062317429406037,
+      "grad_norm": 0.48782945068406763,
+      "learning_rate": 2.6839220138847966e-06,
+      "loss": 0.4899,
+      "step": 706
+    },
+    {
+      "epoch": 2.065238558909445,
+      "grad_norm": 0.4732567901538321,
+      "learning_rate": 2.6688529954530612e-06,
+      "loss": 0.5388,
+      "step": 707
+    },
+    {
+      "epoch": 2.068159688412853,
+      "grad_norm": 0.4935373768584108,
+      "learning_rate": 2.653810983352282e-06,
+      "loss": 0.5822,
+      "step": 708
+    },
+    {
+      "epoch": 2.0710808179162608,
+      "grad_norm": 0.44379266935434936,
+      "learning_rate": 2.6387961518441223e-06,
+      "loss": 0.4639,
+      "step": 709
+    },
+    {
+      "epoch": 2.074001947419669,
+      "grad_norm": 0.47555161755659076,
+      "learning_rate": 2.6238086748753587e-06,
+      "loss": 0.5153,
+      "step": 710
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.4924498329014108,
+      "learning_rate": 2.6088487260758643e-06,
+      "loss": 0.5245,
+      "step": 711
+    },
+    {
+      "epoch": 2.079844206426485,
+      "grad_norm": 0.4492288702216835,
+      "learning_rate": 2.593916478756605e-06,
+      "loss": 0.4816,
+      "step": 712
+    },
+    {
+      "epoch": 2.082765335929893,
+      "grad_norm": 0.5008247473603998,
+      "learning_rate": 2.579012105907619e-06,
+      "loss": 0.5824,
+      "step": 713
+    },
+    {
+      "epoch": 2.0856864654333007,
+      "grad_norm": 0.479453244836006,
+      "learning_rate": 2.5641357801960186e-06,
+      "loss": 0.4738,
+      "step": 714
+    },
+    {
+      "epoch": 2.088607594936709,
+      "grad_norm": 0.49048895532662706,
+      "learning_rate": 2.5492876739639912e-06,
+      "loss": 0.5299,
+      "step": 715
+    },
+    {
+      "epoch": 2.091528724440117,
+      "grad_norm": 0.5369724976128885,
+      "learning_rate": 2.534467959226806e-06,
+      "loss": 0.5162,
+      "step": 716
+    },
+    {
+      "epoch": 2.094449853943525,
+      "grad_norm": 0.5137260447049424,
+      "learning_rate": 2.519676807670811e-06,
+      "loss": 0.5081,
+      "step": 717
+    },
+    {
+      "epoch": 2.097370983446933,
+      "grad_norm": 0.4930859741487903,
+      "learning_rate": 2.504914390651453e-06,
+      "loss": 0.5576,
+      "step": 718
+    },
+    {
+      "epoch": 2.1002921129503407,
+      "grad_norm": 0.4775125343363975,
+      "learning_rate": 2.4901808791912864e-06,
+      "loss": 0.4998,
+      "step": 719
+    },
+    {
+      "epoch": 2.103213242453749,
+      "grad_norm": 0.5214779418043487,
+      "learning_rate": 2.475476443977996e-06,
+      "loss": 0.5696,
+      "step": 720
+    },
+    {
+      "epoch": 2.1061343719571566,
+      "grad_norm": 0.47977761871037206,
+      "learning_rate": 2.460801255362425e-06,
+      "loss": 0.4764,
+      "step": 721
+    },
+    {
+      "epoch": 2.1090555014605648,
+      "grad_norm": 0.48061705012654404,
+      "learning_rate": 2.446155483356582e-06,
+      "loss": 0.5357,
+      "step": 722
+    },
+    {
+      "epoch": 2.111976630963973,
+      "grad_norm": 0.45629547557990546,
+      "learning_rate": 2.4315392976316923e-06,
+      "loss": 0.4915,
+      "step": 723
+    },
+    {
+      "epoch": 2.1148977604673806,
+      "grad_norm": 0.4780522247938775,
+      "learning_rate": 2.4169528675162256e-06,
+      "loss": 0.4838,
+      "step": 724
+    },
+    {
+      "epoch": 2.117818889970789,
+      "grad_norm": 0.49801842180761746,
+      "learning_rate": 2.40239636199393e-06,
+      "loss": 0.537,
+      "step": 725
+    },
+    {
+      "epoch": 2.1207400194741965,
+      "grad_norm": 0.4846689678826097,
+      "learning_rate": 2.3878699497018763e-06,
+      "loss": 0.549,
+      "step": 726
+    },
+    {
+      "epoch": 2.1236611489776047,
+      "grad_norm": 0.5018761582798917,
+      "learning_rate": 2.373373798928507e-06,
+      "loss": 0.5042,
+      "step": 727
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5072064832309269,
+      "learning_rate": 2.358908077611684e-06,
+      "loss": 0.5795,
+      "step": 728
+    },
+    {
+      "epoch": 2.1295034079844206,
+      "grad_norm": 0.46610002849697485,
+      "learning_rate": 2.344472953336747e-06,
+      "loss": 0.468,
+      "step": 729
+    },
+    {
+      "epoch": 2.132424537487829,
+      "grad_norm": 0.5002554866584938,
+      "learning_rate": 2.3300685933345656e-06,
+      "loss": 0.5022,
+      "step": 730
+    },
+    {
+      "epoch": 2.1353456669912365,
+      "grad_norm": 0.5324744280306557,
+      "learning_rate": 2.3156951644796065e-06,
+      "loss": 0.5877,
+      "step": 731
+    },
+    {
+      "epoch": 2.1382667964946447,
+      "grad_norm": 0.4804432916120911,
+      "learning_rate": 2.3013528332879976e-06,
+      "loss": 0.4814,
+      "step": 732
+    },
+    {
+      "epoch": 2.1411879259980524,
+      "grad_norm": 0.47091232830887136,
+      "learning_rate": 2.287041765915606e-06,
+      "loss": 0.4953,
+      "step": 733
+    },
+    {
+      "epoch": 2.1441090555014606,
+      "grad_norm": 0.4535727155626379,
+      "learning_rate": 2.272762128156101e-06,
+      "loss": 0.5269,
+      "step": 734
+    },
+    {
+      "epoch": 2.1470301850048688,
+      "grad_norm": 0.47112551281102016,
+      "learning_rate": 2.2585140854390432e-06,
+      "loss": 0.5244,
+      "step": 735
+    },
+    {
+      "epoch": 2.1499513145082765,
+      "grad_norm": 0.4545532988740129,
+      "learning_rate": 2.2442978028279634e-06,
+      "loss": 0.5341,
+      "step": 736
+    },
+    {
+      "epoch": 2.1528724440116846,
+      "grad_norm": 0.5024829916260531,
+      "learning_rate": 2.2301134450184535e-06,
+      "loss": 0.5243,
+      "step": 737
+    },
+    {
+      "epoch": 2.1557935735150924,
+      "grad_norm": 0.457679806572502,
+      "learning_rate": 2.215961176336255e-06,
+      "loss": 0.5157,
+      "step": 738
+    },
+    {
+      "epoch": 2.1587147030185005,
+      "grad_norm": 0.41799952326682965,
+      "learning_rate": 2.2018411607353572e-06,
+      "loss": 0.4653,
+      "step": 739
+    },
+    {
+      "epoch": 2.1616358325219083,
+      "grad_norm": 0.4787265660707973,
+      "learning_rate": 2.187753561796097e-06,
+      "loss": 0.5521,
+      "step": 740
+    },
+    {
+      "epoch": 2.1645569620253164,
+      "grad_norm": 0.48532252970802797,
+      "learning_rate": 2.1736985427232684e-06,
+      "loss": 0.5736,
+      "step": 741
+    },
+    {
+      "epoch": 2.1674780915287246,
+      "grad_norm": 0.46670045472968474,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.4856,
+      "step": 742
+    },
+    {
+      "epoch": 2.1703992210321323,
+      "grad_norm": 0.456907097163635,
+      "learning_rate": 2.1456868951069875e-06,
+      "loss": 0.513,
+      "step": 743
+    },
+    {
+      "epoch": 2.1733203505355405,
+      "grad_norm": 0.43480032818372993,
+      "learning_rate": 2.1317305910783886e-06,
+      "loss": 0.5183,
+      "step": 744
+    },
+    {
+      "epoch": 2.1762414800389482,
+      "grad_norm": 0.4646674951011356,
+      "learning_rate": 2.117807515942163e-06,
+      "loss": 0.5169,
+      "step": 745
+    },
+    {
+      "epoch": 2.1791626095423564,
+      "grad_norm": 0.5402952715425572,
+      "learning_rate": 2.1039178309970975e-06,
+      "loss": 0.5256,
+      "step": 746
+    },
+    {
+      "epoch": 2.1820837390457646,
+      "grad_norm": 0.44604300646980855,
+      "learning_rate": 2.090061697155147e-06,
+      "loss": 0.5264,
+      "step": 747
+    },
+    {
+      "epoch": 2.1850048685491723,
+      "grad_norm": 0.4855629840620991,
+      "learning_rate": 2.076239274939582e-06,
+      "loss": 0.497,
+      "step": 748
+    },
+    {
+      "epoch": 2.1879259980525805,
+      "grad_norm": 0.450371819506377,
+      "learning_rate": 2.062450724483118e-06,
+      "loss": 0.492,
+      "step": 749
+    },
+    {
+      "epoch": 2.190847127555988,
+      "grad_norm": 0.4675791607919664,
+      "learning_rate": 2.0486962055260744e-06,
+      "loss": 0.4833,
+      "step": 750
+    },
+    {
+      "epoch": 2.1937682570593964,
+      "grad_norm": 0.4931924675592549,
+      "learning_rate": 2.03497587741451e-06,
+      "loss": 0.5325,
+      "step": 751
+    },
+    {
+      "epoch": 2.196689386562804,
+      "grad_norm": 0.4728333199503108,
+      "learning_rate": 2.021289899098384e-06,
+      "loss": 0.5753,
+      "step": 752
+    },
+    {
+      "epoch": 2.1996105160662123,
+      "grad_norm": 0.44477975654172386,
+      "learning_rate": 2.0076384291297134e-06,
+      "loss": 0.5095,
+      "step": 753
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.42502487470067785,
+      "learning_rate": 1.994021625660737e-06,
+      "loss": 0.5004,
+      "step": 754
+    },
+    {
+      "epoch": 2.205452775073028,
+      "grad_norm": 0.4682554593875959,
+      "learning_rate": 1.9804396464420798e-06,
+      "loss": 0.5377,
+      "step": 755
+    },
+    {
+      "epoch": 2.2083739045764363,
+      "grad_norm": 0.478543398252641,
+      "learning_rate": 1.966892648820932e-06,
+      "loss": 0.5214,
+      "step": 756
+    },
+    {
+      "epoch": 2.211295034079844,
+      "grad_norm": 0.4638445072960299,
+      "learning_rate": 1.953380789739216e-06,
+      "loss": 0.522,
+      "step": 757
+    },
+    {
+      "epoch": 2.2142161635832522,
+      "grad_norm": 0.44455755591464263,
+      "learning_rate": 1.939904225731783e-06,
+      "loss": 0.5279,
+      "step": 758
+    },
+    {
+      "epoch": 2.21713729308666,
+      "grad_norm": 0.43381712587340415,
+      "learning_rate": 1.9264631129245836e-06,
+      "loss": 0.5082,
+      "step": 759
+    },
+    {
+      "epoch": 2.220058422590068,
+      "grad_norm": 0.46180168699161817,
+      "learning_rate": 1.9130576070328695e-06,
+      "loss": 0.5129,
+      "step": 760
+    },
+    {
+      "epoch": 2.2229795520934763,
+      "grad_norm": 0.4554045010755852,
+      "learning_rate": 1.8996878633593829e-06,
+      "loss": 0.499,
+      "step": 761
+    },
+    {
+      "epoch": 2.225900681596884,
+      "grad_norm": 0.4749182830100526,
+      "learning_rate": 1.8863540367925676e-06,
+      "loss": 0.5343,
+      "step": 762
+    },
+    {
+      "epoch": 2.228821811100292,
+      "grad_norm": 0.45836107720097835,
+      "learning_rate": 1.873056281804762e-06,
+      "loss": 0.4869,
+      "step": 763
+    },
+    {
+      "epoch": 2.2317429406037,
+      "grad_norm": 0.45195906937962993,
+      "learning_rate": 1.8597947524504178e-06,
+      "loss": 0.5201,
+      "step": 764
+    },
+    {
+      "epoch": 2.234664070107108,
+      "grad_norm": 0.45112191767699233,
+      "learning_rate": 1.8465696023643115e-06,
+      "loss": 0.5354,
+      "step": 765
+    },
+    {
+      "epoch": 2.2375851996105163,
+      "grad_norm": 0.49025509449815313,
+      "learning_rate": 1.8333809847597644e-06,
+      "loss": 0.558,
+      "step": 766
+    },
+    {
+      "epoch": 2.240506329113924,
+      "grad_norm": 0.4486301410086504,
+      "learning_rate": 1.8202290524268761e-06,
+      "loss": 0.4795,
+      "step": 767
+    },
+    {
+      "epoch": 2.243427458617332,
+      "grad_norm": 0.45776452388300076,
+      "learning_rate": 1.8071139577307416e-06,
+      "loss": 0.5228,
+      "step": 768
+    },
+    {
+      "epoch": 2.24634858812074,
+      "grad_norm": 0.4414886979982915,
+      "learning_rate": 1.7940358526096885e-06,
+      "loss": 0.4903,
+      "step": 769
+    },
+    {
+      "epoch": 2.249269717624148,
+      "grad_norm": 0.44672029042681155,
+      "learning_rate": 1.7809948885735295e-06,
+      "loss": 0.5906,
+      "step": 770
+    },
+    {
+      "epoch": 2.252190847127556,
+      "grad_norm": 0.40774965094554166,
+      "learning_rate": 1.7679912167017922e-06,
+      "loss": 0.4726,
+      "step": 771
+    },
+    {
+      "epoch": 2.255111976630964,
+      "grad_norm": 0.42904714451405873,
+      "learning_rate": 1.7550249876419746e-06,
+      "loss": 0.5263,
+      "step": 772
+    },
+    {
+      "epoch": 2.258033106134372,
+      "grad_norm": 0.46232426979933267,
+      "learning_rate": 1.7420963516078016e-06,
+      "loss": 0.5207,
+      "step": 773
+    },
+    {
+      "epoch": 2.26095423563778,
+      "grad_norm": 0.43964498189412426,
+      "learning_rate": 1.7292054583774809e-06,
+      "loss": 0.5634,
+      "step": 774
+    },
+    {
+      "epoch": 2.263875365141188,
+      "grad_norm": 0.4614885797674806,
+      "learning_rate": 1.7163524572919748e-06,
+      "loss": 0.5167,
+      "step": 775
+    },
+    {
+      "epoch": 2.2667964946445958,
+      "grad_norm": 0.4326167271312955,
+      "learning_rate": 1.7035374972532593e-06,
+      "loss": 0.4734,
+      "step": 776
+    },
+    {
+      "epoch": 2.269717624148004,
+      "grad_norm": 0.477108582339082,
+      "learning_rate": 1.6907607267226079e-06,
+      "loss": 0.5436,
+      "step": 777
+    },
+    {
+      "epoch": 2.272638753651412,
+      "grad_norm": 0.45201769752670545,
+      "learning_rate": 1.6780222937188662e-06,
+      "loss": 0.5207,
+      "step": 778
+    },
+    {
+      "epoch": 2.27555988315482,
+      "grad_norm": 0.48262208131523315,
+      "learning_rate": 1.665322345816746e-06,
+      "loss": 0.5304,
+      "step": 779
+    },
+    {
+      "epoch": 2.278481012658228,
+      "grad_norm": 0.45623936201495924,
+      "learning_rate": 1.6526610301451028e-06,
+      "loss": 0.4604,
+      "step": 780
+    },
+    {
+      "epoch": 2.2814021421616357,
+      "grad_norm": 0.45941621411519823,
+      "learning_rate": 1.6400384933852403e-06,
+      "loss": 0.508,
+      "step": 781
+    },
+    {
+      "epoch": 2.284323271665044,
+      "grad_norm": 0.4640162246141768,
+      "learning_rate": 1.6274548817692088e-06,
+      "loss": 0.5429,
+      "step": 782
+    },
+    {
+      "epoch": 2.2872444011684516,
+      "grad_norm": 0.4517120340093356,
+      "learning_rate": 1.6149103410781086e-06,
+      "loss": 0.5126,
+      "step": 783
+    },
+    {
+      "epoch": 2.29016553067186,
+      "grad_norm": 0.4528239429323758,
+      "learning_rate": 1.6024050166404097e-06,
+      "loss": 0.4995,
+      "step": 784
+    },
+    {
+      "epoch": 2.293086660175268,
+      "grad_norm": 0.4440949852742247,
+      "learning_rate": 1.5899390533302538e-06,
+      "loss": 0.5333,
+      "step": 785
+    },
+    {
+      "epoch": 2.2960077896786757,
+      "grad_norm": 0.3970036577732869,
+      "learning_rate": 1.5775125955657877e-06,
+      "loss": 0.5012,
+      "step": 786
+    },
+    {
+      "epoch": 2.298928919182084,
+      "grad_norm": 0.4484643953022932,
+      "learning_rate": 1.5651257873074898e-06,
+      "loss": 0.5413,
+      "step": 787
+    },
+    {
+      "epoch": 2.3018500486854916,
+      "grad_norm": 0.4679410901074801,
+      "learning_rate": 1.5527787720564946e-06,
+      "loss": 0.5229,
+      "step": 788
+    },
+    {
+      "epoch": 2.3047711781888998,
+      "grad_norm": 0.450714999626221,
+      "learning_rate": 1.5404716928529356e-06,
+      "loss": 0.5079,
+      "step": 789
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.46250776338412736,
+      "learning_rate": 1.5282046922742876e-06,
+      "loss": 0.5056,
+      "step": 790
+    },
+    {
+      "epoch": 2.3106134371957157,
+      "grad_norm": 0.47191060686599,
+      "learning_rate": 1.515977912433717e-06,
+      "loss": 0.5279,
+      "step": 791
+    },
+    {
+      "epoch": 2.313534566699124,
+      "grad_norm": 0.493016587785878,
+      "learning_rate": 1.50379149497843e-06,
+      "loss": 0.5483,
+      "step": 792
+    },
+    {
+      "epoch": 2.3164556962025316,
+      "grad_norm": 0.47810281591953624,
+      "learning_rate": 1.4916455810880358e-06,
+      "loss": 0.4898,
+      "step": 793
+    },
+    {
+      "epoch": 2.3193768257059397,
+      "grad_norm": 0.4478075131535074,
+      "learning_rate": 1.4795403114729095e-06,
+      "loss": 0.5101,
+      "step": 794
+    },
+    {
+      "epoch": 2.3222979552093475,
+      "grad_norm": 0.4875641000416008,
+      "learning_rate": 1.4674758263725614e-06,
+      "loss": 0.5854,
+      "step": 795
+    },
+    {
+      "epoch": 2.3252190847127556,
+      "grad_norm": 0.42525631573701944,
+      "learning_rate": 1.4554522655540176e-06,
+      "loss": 0.4989,
+      "step": 796
+    },
+    {
+      "epoch": 2.3281402142161634,
+      "grad_norm": 0.41706543335088775,
+      "learning_rate": 1.4434697683101928e-06,
+      "loss": 0.516,
+      "step": 797
+    },
+    {
+      "epoch": 2.3310613437195715,
+      "grad_norm": 0.4524630863860165,
+      "learning_rate": 1.4315284734582802e-06,
+      "loss": 0.4936,
+      "step": 798
+    },
+    {
+      "epoch": 2.3339824732229797,
+      "grad_norm": 0.5066460944480237,
+      "learning_rate": 1.4196285193381431e-06,
+      "loss": 0.5665,
+      "step": 799
+    },
+    {
+      "epoch": 2.3369036027263874,
+      "grad_norm": 0.4886439078673485,
+      "learning_rate": 1.4077700438107183e-06,
+      "loss": 0.5191,
+      "step": 800
+    },
+    {
+      "epoch": 2.3398247322297956,
+      "grad_norm": 0.42502615563238366,
+      "learning_rate": 1.3959531842564046e-06,
+      "loss": 0.4785,
+      "step": 801
+    },
+    {
+      "epoch": 2.3427458617332033,
+      "grad_norm": 0.45982217381683227,
+      "learning_rate": 1.3841780775734847e-06,
+      "loss": 0.5643,
+      "step": 802
+    },
+    {
+      "epoch": 2.3456669912366115,
+      "grad_norm": 0.43225051785834356,
+      "learning_rate": 1.3724448601765328e-06,
+      "loss": 0.5629,
+      "step": 803
+    },
+    {
+      "epoch": 2.3485881207400197,
+      "grad_norm": 0.44393943057319624,
+      "learning_rate": 1.3607536679948397e-06,
+      "loss": 0.456,
+      "step": 804
+    },
+    {
+      "epoch": 2.3515092502434274,
+      "grad_norm": 0.44935770423521926,
+      "learning_rate": 1.3491046364708294e-06,
+      "loss": 0.52,
+      "step": 805
+    },
+    {
+      "epoch": 2.3544303797468356,
+      "grad_norm": 0.4573183127076476,
+      "learning_rate": 1.337497900558497e-06,
+      "loss": 0.5231,
+      "step": 806
+    },
+    {
+      "epoch": 2.3573515092502433,
+      "grad_norm": 0.4471873705684597,
+      "learning_rate": 1.325933594721841e-06,
+      "loss": 0.5191,
+      "step": 807
+    },
+    {
+      "epoch": 2.3602726387536515,
+      "grad_norm": 0.46232878806411926,
+      "learning_rate": 1.3144118529333126e-06,
+      "loss": 0.4641,
+      "step": 808
+    },
+    {
+      "epoch": 2.3631937682570596,
+      "grad_norm": 0.46569289026754496,
+      "learning_rate": 1.3029328086722537e-06,
+      "loss": 0.5363,
+      "step": 809
+    },
+    {
+      "epoch": 2.3661148977604674,
+      "grad_norm": 0.4515090725602521,
+      "learning_rate": 1.2914965949233572e-06,
+      "loss": 0.5278,
+      "step": 810
+    },
+    {
+      "epoch": 2.3690360272638755,
+      "grad_norm": 0.45127137221258873,
+      "learning_rate": 1.2801033441751244e-06,
+      "loss": 0.4919,
+      "step": 811
+    },
+    {
+      "epoch": 2.3719571567672832,
+      "grad_norm": 0.45307816143888285,
+      "learning_rate": 1.268753188418329e-06,
+      "loss": 0.5063,
+      "step": 812
+    },
+    {
+      "epoch": 2.3748782862706914,
+      "grad_norm": 0.4681309102712795,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.5185,
+      "step": 813
+    },
+    {
+      "epoch": 2.377799415774099,
+      "grad_norm": 0.4665552455306928,
+      "learning_rate": 1.24618268734436e-06,
+      "loss": 0.5426,
+      "step": 814
+    },
+    {
+      "epoch": 2.3807205452775073,
+      "grad_norm": 0.45997968978470405,
+      "learning_rate": 1.2349626035063705e-06,
+      "loss": 0.4996,
+      "step": 815
+    },
+    {
+      "epoch": 2.3836416747809155,
+      "grad_norm": 0.470113793180687,
+      "learning_rate": 1.2237861376151632e-06,
+      "loss": 0.5392,
+      "step": 816
+    },
+    {
+      "epoch": 2.386562804284323,
+      "grad_norm": 0.47230105708445114,
+      "learning_rate": 1.2126534191500622e-06,
+      "loss": 0.5361,
+      "step": 817
+    },
+    {
+      "epoch": 2.3894839337877314,
+      "grad_norm": 0.4439367141535501,
+      "learning_rate": 1.2015645770835765e-06,
+      "loss": 0.4732,
+      "step": 818
+    },
+    {
+      "epoch": 2.392405063291139,
+      "grad_norm": 0.4656016270422327,
+      "learning_rate": 1.1905197398799074e-06,
+      "loss": 0.5227,
+      "step": 819
+    },
+    {
+      "epoch": 2.3953261927945473,
+      "grad_norm": 0.46417888892808024,
+      "learning_rate": 1.1795190354934587e-06,
+      "loss": 0.5606,
+      "step": 820
+    },
+    {
+      "epoch": 2.398247322297955,
+      "grad_norm": 0.41153069172218604,
+      "learning_rate": 1.168562591367361e-06,
+      "loss": 0.5035,
+      "step": 821
+    },
+    {
+      "epoch": 2.401168451801363,
+      "grad_norm": 0.4591933177826969,
+      "learning_rate": 1.1576505344319843e-06,
+      "loss": 0.5485,
+      "step": 822
+    },
+    {
+      "epoch": 2.4040895813047714,
+      "grad_norm": 0.48641885058814227,
+      "learning_rate": 1.146782991103475e-06,
+      "loss": 0.5122,
+      "step": 823
+    },
+    {
+      "epoch": 2.407010710808179,
+      "grad_norm": 0.452636414728882,
+      "learning_rate": 1.1359600872822879e-06,
+      "loss": 0.528,
+      "step": 824
+    },
+    {
+      "epoch": 2.4099318403115872,
+      "grad_norm": 0.4637194713579063,
+      "learning_rate": 1.1251819483517334e-06,
+      "loss": 0.5335,
+      "step": 825
+    },
+    {
+      "epoch": 2.412852969814995,
+      "grad_norm": 0.46499658066713967,
+      "learning_rate": 1.1144486991765175e-06,
+      "loss": 0.519,
+      "step": 826
+    },
+    {
+      "epoch": 2.415774099318403,
+      "grad_norm": 0.4519765342482912,
+      "learning_rate": 1.1037604641012995e-06,
+      "loss": 0.5488,
+      "step": 827
+    },
+    {
+      "epoch": 2.418695228821811,
+      "grad_norm": 0.4382247666420335,
+      "learning_rate": 1.0931173669492472e-06,
+      "loss": 0.4853,
+      "step": 828
+    },
+    {
+      "epoch": 2.421616358325219,
+      "grad_norm": 0.45306336606053793,
+      "learning_rate": 1.0825195310206132e-06,
+      "loss": 0.5224,
+      "step": 829
+    },
+    {
+      "epoch": 2.424537487828627,
+      "grad_norm": 0.4473053801916244,
+      "learning_rate": 1.0719670790912928e-06,
+      "loss": 0.4971,
+      "step": 830
+    },
+    {
+      "epoch": 2.427458617332035,
+      "grad_norm": 0.44465614500766465,
+      "learning_rate": 1.0614601334114099e-06,
+      "loss": 0.5404,
+      "step": 831
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4251639911624754,
+      "learning_rate": 1.0509988157038952e-06,
+      "loss": 0.5179,
+      "step": 832
+    },
+    {
+      "epoch": 2.433300876338851,
+      "grad_norm": 0.40203118130358384,
+      "learning_rate": 1.0405832471630862e-06,
+      "loss": 0.4465,
+      "step": 833
+    },
+    {
+      "epoch": 2.436222005842259,
+      "grad_norm": 0.49627881222358805,
+      "learning_rate": 1.030213548453311e-06,
+      "loss": 0.5782,
+      "step": 834
+    },
+    {
+      "epoch": 2.439143135345667,
+      "grad_norm": 0.4352033965269781,
+      "learning_rate": 1.019889839707498e-06,
+      "loss": 0.471,
+      "step": 835
+    },
+    {
+      "epoch": 2.442064264849075,
+      "grad_norm": 0.4112365791877552,
+      "learning_rate": 1.0096122405257802e-06,
+      "loss": 0.525,
+      "step": 836
+    },
+    {
+      "epoch": 2.444985394352483,
+      "grad_norm": 0.46113597408698487,
+      "learning_rate": 9.99380869974116e-07,
+      "loss": 0.55,
+      "step": 837
+    },
+    {
+      "epoch": 2.447906523855891,
+      "grad_norm": 0.42894572321057406,
+      "learning_rate": 9.891958465828999e-07,
+      "loss": 0.4883,
+      "step": 838
+    },
+    {
+      "epoch": 2.450827653359299,
+      "grad_norm": 0.4306576340142323,
+      "learning_rate": 9.790572883455974e-07,
+      "loss": 0.5127,
+      "step": 839
+    },
+    {
+      "epoch": 2.453748782862707,
+      "grad_norm": 0.47843377943065934,
+      "learning_rate": 9.689653127173743e-07,
+      "loss": 0.5376,
+      "step": 840
+    },
+    {
+      "epoch": 2.456669912366115,
+      "grad_norm": 0.48286119801655825,
+      "learning_rate": 9.589200366137375e-07,
+      "loss": 0.5108,
+      "step": 841
+    },
+    {
+      "epoch": 2.459591041869523,
+      "grad_norm": 0.4840439892081914,
+      "learning_rate": 9.489215764091813e-07,
+      "loss": 0.5436,
+      "step": 842
+    },
+    {
+      "epoch": 2.4625121713729308,
+      "grad_norm": 0.4360507831964012,
+      "learning_rate": 9.389700479358365e-07,
+      "loss": 0.4733,
+      "step": 843
+    },
+    {
+      "epoch": 2.465433300876339,
+      "grad_norm": 0.449450069287443,
+      "learning_rate": 9.290655664821296e-07,
+      "loss": 0.5599,
+      "step": 844
+    },
+    {
+      "epoch": 2.4683544303797467,
+      "grad_norm": 0.41280059188581175,
+      "learning_rate": 9.192082467914465e-07,
+      "loss": 0.4792,
+      "step": 845
+    },
+    {
+      "epoch": 2.471275559883155,
+      "grad_norm": 0.5093589634384059,
+      "learning_rate": 9.093982030608095e-07,
+      "loss": 0.5397,
+      "step": 846
+    },
+    {
+      "epoch": 2.474196689386563,
+      "grad_norm": 0.45448988250512634,
+      "learning_rate": 8.996355489395442e-07,
+      "loss": 0.4975,
+      "step": 847
+    },
+    {
+      "epoch": 2.4771178188899707,
+      "grad_norm": 0.44480146821328176,
+      "learning_rate": 8.899203975279674e-07,
+      "loss": 0.4914,
+      "step": 848
+    },
+    {
+      "epoch": 2.480038948393379,
+      "grad_norm": 0.4896804183179627,
+      "learning_rate": 8.802528613760775e-07,
+      "loss": 0.5599,
+      "step": 849
+    },
+    {
+      "epoch": 2.4829600778967866,
+      "grad_norm": 0.4529609749837371,
+      "learning_rate": 8.706330524822548e-07,
+      "loss": 0.4598,
+      "step": 850
+    },
+    {
+      "epoch": 2.485881207400195,
+      "grad_norm": 0.4141270566207645,
+      "learning_rate": 8.610610822919546e-07,
+      "loss": 0.5195,
+      "step": 851
+    },
+    {
+      "epoch": 2.4888023369036025,
+      "grad_norm": 0.4413472687231943,
+      "learning_rate": 8.515370616964219e-07,
+      "loss": 0.5181,
+      "step": 852
+    },
+    {
+      "epoch": 2.4917234664070107,
+      "grad_norm": 0.4120555785416349,
+      "learning_rate": 8.420611010314062e-07,
+      "loss": 0.5267,
+      "step": 853
+    },
+    {
+      "epoch": 2.494644595910419,
+      "grad_norm": 0.4661650695044543,
+      "learning_rate": 8.326333100758843e-07,
+      "loss": 0.5086,
+      "step": 854
+    },
+    {
+      "epoch": 2.4975657254138266,
+      "grad_norm": 0.5073361612635612,
+      "learning_rate": 8.232537980507848e-07,
+      "loss": 0.5536,
+      "step": 855
+    },
+    {
+      "epoch": 2.5004868549172348,
+      "grad_norm": 0.4961949890878462,
+      "learning_rate": 8.13922673617727e-07,
+      "loss": 0.5238,
+      "step": 856
+    },
+    {
+      "epoch": 2.5034079844206425,
+      "grad_norm": 0.4720376448013444,
+      "learning_rate": 8.046400448777575e-07,
+      "loss": 0.4823,
+      "step": 857
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.49809468171059407,
+      "learning_rate": 7.954060193701019e-07,
+      "loss": 0.5466,
+      "step": 858
+    },
+    {
+      "epoch": 2.5092502434274584,
+      "grad_norm": 0.4639329084639315,
+      "learning_rate": 7.862207040709191e-07,
+      "loss": 0.5104,
+      "step": 859
+    },
+    {
+      "epoch": 2.5121713729308666,
+      "grad_norm": 0.47993248882289974,
+      "learning_rate": 7.770842053920585e-07,
+      "loss": 0.5379,
+      "step": 860
+    },
+    {
+      "epoch": 2.5150925024342747,
+      "grad_norm": 0.42897728377766037,
+      "learning_rate": 7.679966291798297e-07,
+      "loss": 0.482,
+      "step": 861
+    },
+    {
+      "epoch": 2.5180136319376825,
+      "grad_norm": 0.45778561195288503,
+      "learning_rate": 7.589580807137742e-07,
+      "loss": 0.5032,
+      "step": 862
+    },
+    {
+      "epoch": 2.5209347614410906,
+      "grad_norm": 0.46270673489951974,
+      "learning_rate": 7.4996866470545e-07,
+      "loss": 0.5706,
+      "step": 863
+    },
+    {
+      "epoch": 2.523855890944499,
+      "grad_norm": 0.42072713048943156,
+      "learning_rate": 7.410284852972127e-07,
+      "loss": 0.5204,
+      "step": 864
+    },
+    {
+      "epoch": 2.5267770204479065,
+      "grad_norm": 0.40846266187572267,
+      "learning_rate": 7.321376460610136e-07,
+      "loss": 0.4916,
+      "step": 865
+    },
+    {
+      "epoch": 2.5296981499513143,
+      "grad_norm": 0.4443443498829621,
+      "learning_rate": 7.232962499972002e-07,
+      "loss": 0.5395,
+      "step": 866
+    },
+    {
+      "epoch": 2.5326192794547224,
+      "grad_norm": 0.459368960082765,
+      "learning_rate": 7.145043995333173e-07,
+      "loss": 0.4924,
+      "step": 867
+    },
+    {
+      "epoch": 2.5355404089581306,
+      "grad_norm": 0.47824210909228654,
+      "learning_rate": 7.057621965229267e-07,
+      "loss": 0.513,
+      "step": 868
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.438214605329431,
+      "learning_rate": 6.970697422444228e-07,
+      "loss": 0.5695,
+      "step": 869
+    },
+    {
+      "epoch": 2.5413826679649465,
+      "grad_norm": 0.45798584370168316,
+      "learning_rate": 6.884271373998608e-07,
+      "loss": 0.4925,
+      "step": 870
+    },
+    {
+      "epoch": 2.5443037974683547,
+      "grad_norm": 0.4811217026678713,
+      "learning_rate": 6.798344821137947e-07,
+      "loss": 0.5083,
+      "step": 871
+    },
+    {
+      "epoch": 2.5472249269717624,
+      "grad_norm": 0.4418809012550695,
+      "learning_rate": 6.71291875932108e-07,
+      "loss": 0.5352,
+      "step": 872
+    },
+    {
+      "epoch": 2.5501460564751706,
+      "grad_norm": 0.42576085849329387,
+      "learning_rate": 6.62799417820868e-07,
+      "loss": 0.5262,
+      "step": 873
+    },
+    {
+      "epoch": 2.5530671859785783,
+      "grad_norm": 0.4287609653415205,
+      "learning_rate": 6.543572061651738e-07,
+      "loss": 0.5399,
+      "step": 874
+    },
+    {
+      "epoch": 2.5559883154819865,
+      "grad_norm": 0.45232434071279953,
+      "learning_rate": 6.459653387680248e-07,
+      "loss": 0.4725,
+      "step": 875
+    },
+    {
+      "epoch": 2.558909444985394,
+      "grad_norm": 0.4554887075692202,
+      "learning_rate": 6.376239128491784e-07,
+      "loss": 0.5122,
+      "step": 876
+    },
+    {
+      "epoch": 2.5618305744888024,
+      "grad_norm": 0.46277880760014206,
+      "learning_rate": 6.293330250440277e-07,
+      "loss": 0.4941,
+      "step": 877
+    },
+    {
+      "epoch": 2.5647517039922105,
+      "grad_norm": 0.4962082268099505,
+      "learning_rate": 6.210927714024834e-07,
+      "loss": 0.5592,
+      "step": 878
+    },
+    {
+      "epoch": 2.5676728334956183,
+      "grad_norm": 0.44324064458458867,
+      "learning_rate": 6.129032473878582e-07,
+      "loss": 0.5282,
+      "step": 879
+    },
+    {
+      "epoch": 2.5705939629990264,
+      "grad_norm": 0.43320029335686094,
+      "learning_rate": 6.047645478757635e-07,
+      "loss": 0.4562,
+      "step": 880
+    },
+    {
+      "epoch": 2.573515092502434,
+      "grad_norm": 0.45163259818241996,
+      "learning_rate": 5.966767671530078e-07,
+      "loss": 0.5501,
+      "step": 881
+    },
+    {
+      "epoch": 2.5764362220058423,
+      "grad_norm": 0.4552021870162526,
+      "learning_rate": 5.88639998916506e-07,
+      "loss": 0.5359,
+      "step": 882
+    },
+    {
+      "epoch": 2.57935735150925,
+      "grad_norm": 0.4604883255796744,
+      "learning_rate": 5.806543362721945e-07,
+      "loss": 0.5274,
+      "step": 883
+    },
+    {
+      "epoch": 2.5822784810126582,
+      "grad_norm": 0.41454954591229876,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.4765,
+      "step": 884
+    },
+    {
+      "epoch": 2.5851996105160664,
+      "grad_norm": 0.4325932298835654,
+      "learning_rate": 5.648366972225222e-07,
+      "loss": 0.5282,
+      "step": 885
+    },
+    {
+      "epoch": 2.588120740019474,
+      "grad_norm": 0.433038404211758,
+      "learning_rate": 5.570049040644609e-07,
+      "loss": 0.5137,
+      "step": 886
+    },
+    {
+      "epoch": 2.5910418695228823,
+      "grad_norm": 0.45496347053471886,
+      "learning_rate": 5.492245829910664e-07,
+      "loss": 0.4973,
+      "step": 887
+    },
+    {
+      "epoch": 2.59396299902629,
+      "grad_norm": 0.44790019595001196,
+      "learning_rate": 5.414958241373358e-07,
+      "loss": 0.4881,
+      "step": 888
+    },
+    {
+      "epoch": 2.596884128529698,
+      "grad_norm": 0.4506105888324243,
+      "learning_rate": 5.33818717040917e-07,
+      "loss": 0.5446,
+      "step": 889
+    },
+    {
+      "epoch": 2.599805258033106,
+      "grad_norm": 0.45920237343152387,
+      "learning_rate": 5.261933506410722e-07,
+      "loss": 0.5544,
+      "step": 890
+    },
+    {
+      "epoch": 2.602726387536514,
+      "grad_norm": 0.433968288589714,
+      "learning_rate": 5.186198132776459e-07,
+      "loss": 0.4792,
+      "step": 891
+    },
+    {
+      "epoch": 2.6056475170399223,
+      "grad_norm": 0.45054587006233904,
+      "learning_rate": 5.110981926900488e-07,
+      "loss": 0.5271,
+      "step": 892
+    },
+    {
+      "epoch": 2.60856864654333,
+      "grad_norm": 0.4402130423120147,
+      "learning_rate": 5.036285760162307e-07,
+      "loss": 0.4772,
+      "step": 893
+    },
+    {
+      "epoch": 2.611489776046738,
+      "grad_norm": 0.4560401508104557,
+      "learning_rate": 4.962110497916789e-07,
+      "loss": 0.5078,
+      "step": 894
+    },
+    {
+      "epoch": 2.6144109055501463,
+      "grad_norm": 0.4574934740247989,
+      "learning_rate": 4.888456999484098e-07,
+      "loss": 0.5177,
+      "step": 895
+    },
+    {
+      "epoch": 2.617332035053554,
+      "grad_norm": 0.42797938213430475,
+      "learning_rate": 4.815326118139813e-07,
+      "loss": 0.5068,
+      "step": 896
+    },
+    {
+      "epoch": 2.620253164556962,
+      "grad_norm": 0.4508198643702282,
+      "learning_rate": 4.742718701104965e-07,
+      "loss": 0.4836,
+      "step": 897
+    },
+    {
+      "epoch": 2.62317429406037,
+      "grad_norm": 0.4936637557435792,
+      "learning_rate": 4.670635589536254e-07,
+      "loss": 0.5688,
+      "step": 898
+    },
+    {
+      "epoch": 2.626095423563778,
+      "grad_norm": 0.4446810251135343,
+      "learning_rate": 4.599077618516312e-07,
+      "loss": 0.5213,
+      "step": 899
+    },
+    {
+      "epoch": 2.629016553067186,
+      "grad_norm": 0.4393121918571539,
+      "learning_rate": 4.528045617044019e-07,
+      "loss": 0.515,
+      "step": 900
+    },
+    {
+      "epoch": 2.631937682570594,
+      "grad_norm": 0.4546453714565344,
+      "learning_rate": 4.457540408024896e-07,
+      "loss": 0.5119,
+      "step": 901
+    },
+    {
+      "epoch": 2.634858812074002,
+      "grad_norm": 0.43919005512991516,
+      "learning_rate": 4.387562808261575e-07,
+      "loss": 0.5273,
+      "step": 902
+    },
+    {
+      "epoch": 2.63777994157741,
+      "grad_norm": 0.4491615831317555,
+      "learning_rate": 4.3181136284443204e-07,
+      "loss": 0.4994,
+      "step": 903
+    },
+    {
+      "epoch": 2.6407010710808176,
+      "grad_norm": 0.48434931343594284,
+      "learning_rate": 4.249193673141694e-07,
+      "loss": 0.5186,
+      "step": 904
+    },
+    {
+      "epoch": 2.643622200584226,
+      "grad_norm": 0.41259506287384623,
+      "learning_rate": 4.180803740791156e-07,
+      "loss": 0.5319,
+      "step": 905
+    },
+    {
+      "epoch": 2.646543330087634,
+      "grad_norm": 0.4076433170920647,
+      "learning_rate": 4.112944623689863e-07,
+      "loss": 0.5178,
+      "step": 906
+    },
+    {
+      "epoch": 2.6494644595910417,
+      "grad_norm": 0.421345630456112,
+      "learning_rate": 4.0456171079854833e-07,
+      "loss": 0.4951,
+      "step": 907
+    },
+    {
+      "epoch": 2.65238558909445,
+      "grad_norm": 0.4435092242276153,
+      "learning_rate": 3.978821973667074e-07,
+      "loss": 0.4837,
+      "step": 908
+    },
+    {
+      "epoch": 2.655306718597858,
+      "grad_norm": 0.44521873058867134,
+      "learning_rate": 3.9125599945560866e-07,
+      "loss": 0.5526,
+      "step": 909
+    },
+    {
+      "epoch": 2.6582278481012658,
+      "grad_norm": 0.45505539256105115,
+      "learning_rate": 3.846831938297324e-07,
+      "loss": 0.5158,
+      "step": 910
+    },
+    {
+      "epoch": 2.661148977604674,
+      "grad_norm": 0.433074017412774,
+      "learning_rate": 3.7816385663501097e-07,
+      "loss": 0.4809,
+      "step": 911
+    },
+    {
+      "epoch": 2.6640701071080817,
+      "grad_norm": 0.4502224951078667,
+      "learning_rate": 3.716980633979489e-07,
+      "loss": 0.5363,
+      "step": 912
+    },
+    {
+      "epoch": 2.66699123661149,
+      "grad_norm": 0.43979723124115566,
+      "learning_rate": 3.6528588902473905e-07,
+      "loss": 0.508,
+      "step": 913
+    },
+    {
+      "epoch": 2.6699123661148976,
+      "grad_norm": 0.46901546979749387,
+      "learning_rate": 3.589274078004029e-07,
+      "loss": 0.5489,
+      "step": 914
+    },
+    {
+      "epoch": 2.6728334956183057,
+      "grad_norm": 0.4432062648822634,
+      "learning_rate": 3.5262269338792623e-07,
+      "loss": 0.526,
+      "step": 915
+    },
+    {
+      "epoch": 2.675754625121714,
+      "grad_norm": 0.43511333463780544,
+      "learning_rate": 3.463718188274046e-07,
+      "loss": 0.5332,
+      "step": 916
+    },
+    {
+      "epoch": 2.6786757546251216,
+      "grad_norm": 0.38843912794513447,
+      "learning_rate": 3.401748565352031e-07,
+      "loss": 0.4997,
+      "step": 917
+    },
+    {
+      "epoch": 2.68159688412853,
+      "grad_norm": 0.41986039406023057,
+      "learning_rate": 3.340318783031099e-07,
+      "loss": 0.4804,
+      "step": 918
+    },
+    {
+      "epoch": 2.6845180136319375,
+      "grad_norm": 0.44648799987073357,
+      "learning_rate": 3.279429552975094e-07,
+      "loss": 0.57,
+      "step": 919
+    },
+    {
+      "epoch": 2.6874391431353457,
+      "grad_norm": 0.38625226404630086,
+      "learning_rate": 3.219081580585548e-07,
+      "loss": 0.418,
+      "step": 920
+    },
+    {
+      "epoch": 2.6903602726387534,
+      "grad_norm": 0.46495901890128216,
+      "learning_rate": 3.159275564993558e-07,
+      "loss": 0.562,
+      "step": 921
+    },
+    {
+      "epoch": 2.6932814021421616,
+      "grad_norm": 0.4542095138457356,
+      "learning_rate": 3.100012199051627e-07,
+      "loss": 0.48,
+      "step": 922
+    },
+    {
+      "epoch": 2.6962025316455698,
+      "grad_norm": 0.461034837465003,
+      "learning_rate": 3.0412921693256657e-07,
+      "loss": 0.5802,
+      "step": 923
+    },
+    {
+      "epoch": 2.6991236611489775,
+      "grad_norm": 0.4124404226906978,
+      "learning_rate": 2.9831161560870346e-07,
+      "loss": 0.5124,
+      "step": 924
+    },
+    {
+      "epoch": 2.7020447906523857,
+      "grad_norm": 0.420405602577591,
+      "learning_rate": 2.9254848333046817e-07,
+      "loss": 0.4959,
+      "step": 925
+    },
+    {
+      "epoch": 2.704965920155794,
+      "grad_norm": 0.42637463336396936,
+      "learning_rate": 2.8683988686372956e-07,
+      "loss": 0.529,
+      "step": 926
+    },
+    {
+      "epoch": 2.7078870496592016,
+      "grad_norm": 0.4156040595949887,
+      "learning_rate": 2.811858923425609e-07,
+      "loss": 0.5184,
+      "step": 927
+    },
+    {
+      "epoch": 2.7108081791626093,
+      "grad_norm": 0.4312166608407746,
+      "learning_rate": 2.755865652684703e-07,
+      "loss": 0.5438,
+      "step": 928
+    },
+    {
+      "epoch": 2.7137293086660175,
+      "grad_norm": 0.4369289965460609,
+      "learning_rate": 2.7004197050964744e-07,
+      "loss": 0.4978,
+      "step": 929
+    },
+    {
+      "epoch": 2.7166504381694256,
+      "grad_norm": 0.426344227648129,
+      "learning_rate": 2.645521723002037e-07,
+      "loss": 0.5041,
+      "step": 930
+    },
+    {
+      "epoch": 2.7195715676728334,
+      "grad_norm": 0.4417752376579622,
+      "learning_rate": 2.591172342394349e-07,
+      "loss": 0.5587,
+      "step": 931
+    },
+    {
+      "epoch": 2.7224926971762415,
+      "grad_norm": 0.457814912149392,
+      "learning_rate": 2.537372192910825e-07,
+      "loss": 0.5175,
+      "step": 932
+    },
+    {
+      "epoch": 2.7254138266796497,
+      "grad_norm": 0.45549699443779185,
+      "learning_rate": 2.4841218978260194e-07,
+      "loss": 0.5033,
+      "step": 933
+    },
+    {
+      "epoch": 2.7283349561830574,
+      "grad_norm": 0.44509990954989737,
+      "learning_rate": 2.4314220740444595e-07,
+      "loss": 0.5192,
+      "step": 934
+    },
+    {
+      "epoch": 2.731256085686465,
+      "grad_norm": 0.4542595414389855,
+      "learning_rate": 2.3792733320934348e-07,
+      "loss": 0.5193,
+      "step": 935
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.46788823553431785,
+      "learning_rate": 2.3276762761159588e-07,
+      "loss": 0.5226,
+      "step": 936
+    },
+    {
+      "epoch": 2.7370983446932815,
+      "grad_norm": 0.46548939441743264,
+      "learning_rate": 2.276631503863763e-07,
+      "loss": 0.5234,
+      "step": 937
+    },
+    {
+      "epoch": 2.7400194741966892,
+      "grad_norm": 0.43071296003232673,
+      "learning_rate": 2.2261396066903916e-07,
+      "loss": 0.4968,
+      "step": 938
+    },
+    {
+      "epoch": 2.7429406037000974,
+      "grad_norm": 0.4422590771192652,
+      "learning_rate": 2.176201169544312e-07,
+      "loss": 0.5135,
+      "step": 939
+    },
+    {
+      "epoch": 2.7458617332035056,
+      "grad_norm": 0.43839294000442797,
+      "learning_rate": 2.12681677096217e-07,
+      "loss": 0.5459,
+      "step": 940
+    },
+    {
+      "epoch": 2.7487828627069133,
+      "grad_norm": 0.4296191280132258,
+      "learning_rate": 2.0779869830620692e-07,
+      "loss": 0.4538,
+      "step": 941
+    },
+    {
+      "epoch": 2.7517039922103215,
+      "grad_norm": 0.4119484440139385,
+      "learning_rate": 2.029712371536957e-07,
+      "loss": 0.4839,
+      "step": 942
+    },
+    {
+      "epoch": 2.754625121713729,
+      "grad_norm": 0.4512158763524109,
+      "learning_rate": 1.9819934956480603e-07,
+      "loss": 0.5445,
+      "step": 943
+    },
+    {
+      "epoch": 2.7575462512171374,
+      "grad_norm": 0.4212292286246555,
+      "learning_rate": 1.9348309082184102e-07,
+      "loss": 0.4934,
+      "step": 944
+    },
+    {
+      "epoch": 2.760467380720545,
+      "grad_norm": 0.44496919909800775,
+      "learning_rate": 1.888225155626433e-07,
+      "loss": 0.5456,
+      "step": 945
+    },
+    {
+      "epoch": 2.7633885102239533,
+      "grad_norm": 0.41713723572900646,
+      "learning_rate": 1.8421767777996425e-07,
+      "loss": 0.4462,
+      "step": 946
+    },
+    {
+      "epoch": 2.7663096397273614,
+      "grad_norm": 0.4457624171080001,
+      "learning_rate": 1.7966863082083507e-07,
+      "loss": 0.5427,
+      "step": 947
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.4348007177346548,
+      "learning_rate": 1.7517542738595071e-07,
+      "loss": 0.5568,
+      "step": 948
+    },
+    {
+      "epoch": 2.7721518987341773,
+      "grad_norm": 0.41399848976523823,
+      "learning_rate": 1.7073811952905862e-07,
+      "loss": 0.5083,
+      "step": 949
+    },
+    {
+      "epoch": 2.775073028237585,
+      "grad_norm": 0.43232436981734923,
+      "learning_rate": 1.6635675865635859e-07,
+      "loss": 0.4505,
+      "step": 950
+    },
+    {
+      "epoch": 2.7779941577409932,
+      "grad_norm": 0.4750164143516587,
+      "learning_rate": 1.6203139552590175e-07,
+      "loss": 0.5831,
+      "step": 951
+    },
+    {
+      "epoch": 2.780915287244401,
+      "grad_norm": 0.39972505599387415,
+      "learning_rate": 1.5776208024700702e-07,
+      "loss": 0.4751,
+      "step": 952
+    },
+    {
+      "epoch": 2.783836416747809,
+      "grad_norm": 0.42708956986618357,
+      "learning_rate": 1.5354886227967934e-07,
+      "loss": 0.532,
+      "step": 953
+    },
+    {
+      "epoch": 2.7867575462512173,
+      "grad_norm": 0.4435213719711565,
+      "learning_rate": 1.4939179043403474e-07,
+      "loss": 0.5347,
+      "step": 954
+    },
+    {
+      "epoch": 2.789678675754625,
+      "grad_norm": 0.4416374818886015,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.5459,
+      "step": 955
+    },
+    {
+      "epoch": 2.792599805258033,
+      "grad_norm": 0.45153869521194906,
+      "learning_rate": 1.4124627709544814e-07,
+      "loss": 0.4918,
+      "step": 956
+    },
+    {
+      "epoch": 2.7955209347614414,
+      "grad_norm": 0.4440135260222283,
+      "learning_rate": 1.3725792996825083e-07,
+      "loss": 0.5016,
+      "step": 957
+    },
+    {
+      "epoch": 2.798442064264849,
+      "grad_norm": 0.39484795107770926,
+      "learning_rate": 1.3332591769314006e-07,
+      "loss": 0.5043,
+      "step": 958
+    },
+    {
+      "epoch": 2.801363193768257,
+      "grad_norm": 0.444235962409132,
+      "learning_rate": 1.2945028582246445e-07,
+      "loss": 0.5637,
+      "step": 959
+    },
+    {
+      "epoch": 2.804284323271665,
+      "grad_norm": 0.4583481416669065,
+      "learning_rate": 1.2563107925540774e-07,
+      "loss": 0.5017,
+      "step": 960
+    },
+    {
+      "epoch": 2.807205452775073,
+      "grad_norm": 0.4215520268166574,
+      "learning_rate": 1.2186834223746612e-07,
+      "loss": 0.5136,
+      "step": 961
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.43820018627048524,
+      "learning_rate": 1.1816211835993684e-07,
+      "loss": 0.4964,
+      "step": 962
+    },
+    {
+      "epoch": 2.813047711781889,
+      "grad_norm": 0.4420212484091613,
+      "learning_rate": 1.1451245055941251e-07,
+      "loss": 0.5112,
+      "step": 963
+    },
+    {
+      "epoch": 2.8159688412852972,
+      "grad_norm": 0.4365334218610012,
+      "learning_rate": 1.1091938111728374e-07,
+      "loss": 0.5273,
+      "step": 964
+    },
+    {
+      "epoch": 2.818889970788705,
+      "grad_norm": 0.4377608647741322,
+      "learning_rate": 1.0738295165924783e-07,
+      "loss": 0.5398,
+      "step": 965
+    },
+    {
+      "epoch": 2.8218111002921127,
+      "grad_norm": 0.45046127023142907,
+      "learning_rate": 1.0390320315482982e-07,
+      "loss": 0.5062,
+      "step": 966
+    },
+    {
+      "epoch": 2.824732229795521,
+      "grad_norm": 0.4019879301034501,
+      "learning_rate": 1.0048017591690607e-07,
+      "loss": 0.5114,
+      "step": 967
+    },
+    {
+      "epoch": 2.827653359298929,
+      "grad_norm": 0.4402917529148519,
+      "learning_rate": 9.711390960123634e-08,
+      "loss": 0.5359,
+      "step": 968
+    },
+    {
+      "epoch": 2.8305744888023368,
+      "grad_norm": 0.41904135734811376,
+      "learning_rate": 9.380444320600591e-08,
+      "loss": 0.4684,
+      "step": 969
+    },
+    {
+      "epoch": 2.833495618305745,
+      "grad_norm": 0.4659811321879393,
+      "learning_rate": 9.055181507137245e-08,
+      "loss": 0.5426,
+      "step": 970
+    },
+    {
+      "epoch": 2.836416747809153,
+      "grad_norm": 0.440634928654449,
+      "learning_rate": 8.73560628790232e-08,
+      "loss": 0.5258,
+      "step": 971
+    },
+    {
+      "epoch": 2.839337877312561,
+      "grad_norm": 0.44241019235537177,
+      "learning_rate": 8.421722365173856e-08,
+      "loss": 0.553,
+      "step": 972
+    },
+    {
+      "epoch": 2.842259006815969,
+      "grad_norm": 0.4085933163681831,
+      "learning_rate": 8.113533375295968e-08,
+      "loss": 0.4804,
+      "step": 973
+    },
+    {
+      "epoch": 2.8451801363193767,
+      "grad_norm": 0.42283609192144683,
+      "learning_rate": 7.81104288863721e-08,
+      "loss": 0.5268,
+      "step": 974
+    },
+    {
+      "epoch": 2.848101265822785,
+      "grad_norm": 0.4414452447566111,
+      "learning_rate": 7.514254409549005e-08,
+      "loss": 0.5059,
+      "step": 975
+    },
+    {
+      "epoch": 2.8510223953261926,
+      "grad_norm": 0.477876908869347,
+      "learning_rate": 7.223171376325e-08,
+      "loss": 0.5401,
+      "step": 976
+    },
+    {
+      "epoch": 2.853943524829601,
+      "grad_norm": 0.45667617596842697,
+      "learning_rate": 6.93779716116122e-08,
+      "loss": 0.5179,
+      "step": 977
+    },
+    {
+      "epoch": 2.856864654333009,
+      "grad_norm": 0.43661296400815486,
+      "learning_rate": 6.658135070116978e-08,
+      "loss": 0.479,
+      "step": 978
+    },
+    {
+      "epoch": 2.8597857838364167,
+      "grad_norm": 0.43447801900111893,
+      "learning_rate": 6.384188343076914e-08,
+      "loss": 0.4933,
+      "step": 979
+    },
+    {
+      "epoch": 2.862706913339825,
+      "grad_norm": 0.4567696696725672,
+      "learning_rate": 6.115960153712963e-08,
+      "loss": 0.5225,
+      "step": 980
+    },
+    {
+      "epoch": 2.8656280428432326,
+      "grad_norm": 0.4423237005020546,
+      "learning_rate": 5.8534536094478345e-08,
+      "loss": 0.5649,
+      "step": 981
+    },
+    {
+      "epoch": 2.8685491723466408,
+      "grad_norm": 0.41379887497949286,
+      "learning_rate": 5.596671751419147e-08,
+      "loss": 0.474,
+      "step": 982
+    },
+    {
+      "epoch": 2.8714703018500485,
+      "grad_norm": 0.4382842717392665,
+      "learning_rate": 5.3456175544437936e-08,
+      "loss": 0.529,
+      "step": 983
+    },
+    {
+      "epoch": 2.8743914313534566,
+      "grad_norm": 0.4463860151059245,
+      "learning_rate": 5.100293926984023e-08,
+      "loss": 0.4909,
+      "step": 984
+    },
+    {
+      "epoch": 2.877312560856865,
+      "grad_norm": 0.4057084453405752,
+      "learning_rate": 4.860703711113246e-08,
+      "loss": 0.5029,
+      "step": 985
+    },
+    {
+      "epoch": 2.8802336903602725,
+      "grad_norm": 0.42518495194251527,
+      "learning_rate": 4.626849682483392e-08,
+      "loss": 0.5376,
+      "step": 986
+    },
+    {
+      "epoch": 2.8831548198636807,
+      "grad_norm": 0.45413911287155506,
+      "learning_rate": 4.398734550292716e-08,
+      "loss": 0.5467,
+      "step": 987
+    },
+    {
+      "epoch": 2.8860759493670884,
+      "grad_norm": 0.4142515305933239,
+      "learning_rate": 4.176360957254266e-08,
+      "loss": 0.4568,
+      "step": 988
+    },
+    {
+      "epoch": 2.8889970788704966,
+      "grad_norm": 0.435060369756503,
+      "learning_rate": 3.959731479565465e-08,
+      "loss": 0.5319,
+      "step": 989
+    },
+    {
+      "epoch": 2.8919182083739043,
+      "grad_norm": 0.43929179209673974,
+      "learning_rate": 3.748848626878132e-08,
+      "loss": 0.5782,
+      "step": 990
+    },
+    {
+      "epoch": 2.8948393378773125,
+      "grad_norm": 0.4061515904911967,
+      "learning_rate": 3.543714842269508e-08,
+      "loss": 0.462,
+      "step": 991
+    },
+    {
+      "epoch": 2.8977604673807207,
+      "grad_norm": 0.46420928088916874,
+      "learning_rate": 3.344332502213887e-08,
+      "loss": 0.5218,
+      "step": 992
+    },
+    {
+      "epoch": 2.9006815968841284,
+      "grad_norm": 0.4346617171265348,
+      "learning_rate": 3.150703916554976e-08,
+      "loss": 0.4832,
+      "step": 993
+    },
+    {
+      "epoch": 2.9036027263875366,
+      "grad_norm": 0.45371971197536476,
+      "learning_rate": 2.962831328479465e-08,
+      "loss": 0.5087,
+      "step": 994
+    },
+    {
+      "epoch": 2.9065238558909448,
+      "grad_norm": 0.47344580171350403,
+      "learning_rate": 2.7807169144906108e-08,
+      "loss": 0.5639,
+      "step": 995
+    },
+    {
+      "epoch": 2.9094449853943525,
+      "grad_norm": 0.4343079692320858,
+      "learning_rate": 2.6043627843834185e-08,
+      "loss": 0.534,
+      "step": 996
+    },
+    {
+      "epoch": 2.91236611489776,
+      "grad_norm": 0.41472366849725406,
+      "learning_rate": 2.4337709812199405e-08,
+      "loss": 0.471,
+      "step": 997
+    },
+    {
+      "epoch": 2.9152872444011684,
+      "grad_norm": 0.41847164257771846,
+      "learning_rate": 2.2689434813056856e-08,
+      "loss": 0.5252,
+      "step": 998
+    },
+    {
+      "epoch": 2.9182083739045765,
+      "grad_norm": 0.4540400278681399,
+      "learning_rate": 2.1098821941667457e-08,
+      "loss": 0.5683,
+      "step": 999
+    },
+    {
+      "epoch": 2.9211295034079843,
+      "grad_norm": 0.43916840754064074,
+      "learning_rate": 1.9565889625275945e-08,
+      "loss": 0.4844,
+      "step": 1000
+    },
+    {
+      "epoch": 2.9240506329113924,
+      "grad_norm": 0.4345074027775515,
+      "learning_rate": 1.8090655622899356e-08,
+      "loss": 0.4881,
+      "step": 1001
+    },
+    {
+      "epoch": 2.9269717624148006,
+      "grad_norm": 0.4796907378778938,
+      "learning_rate": 1.6673137025118303e-08,
+      "loss": 0.5576,
+      "step": 1002
+    },
+    {
+      "epoch": 2.9298928919182083,
+      "grad_norm": 0.4751897325081864,
+      "learning_rate": 1.5313350253882143e-08,
+      "loss": 0.5248,
+      "step": 1003
+    },
+    {
+      "epoch": 2.9328140214216165,
+      "grad_norm": 0.4566958532328187,
+      "learning_rate": 1.4011311062316347e-08,
+      "loss": 0.5207,
+      "step": 1004
+    },
+    {
+      "epoch": 2.9357351509250242,
+      "grad_norm": 0.4143592301627743,
+      "learning_rate": 1.2767034534540978e-08,
+      "loss": 0.483,
+      "step": 1005
+    },
+    {
+      "epoch": 2.9386562804284324,
+      "grad_norm": 0.4151441658060871,
+      "learning_rate": 1.158053508549528e-08,
+      "loss": 0.4624,
+      "step": 1006
+    },
+    {
+      "epoch": 2.94157740993184,
+      "grad_norm": 0.44443161265224573,
+      "learning_rate": 1.0451826460772252e-08,
+      "loss": 0.5295,
+      "step": 1007
+    },
+    {
+      "epoch": 2.9444985394352483,
+      "grad_norm": 0.4244569182328679,
+      "learning_rate": 9.38092173645766e-09,
+      "loss": 0.5164,
+      "step": 1008
+    },
+    {
+      "epoch": 2.9474196689386565,
+      "grad_norm": 0.42619195275616983,
+      "learning_rate": 8.367833318980168e-09,
+      "loss": 0.5474,
+      "step": 1009
+    },
+    {
+      "epoch": 2.950340798442064,
+      "grad_norm": 0.44631991928366016,
+      "learning_rate": 7.412572944965335e-09,
+      "loss": 0.4825,
+      "step": 1010
+    },
+    {
+      "epoch": 2.9532619279454724,
+      "grad_norm": 0.4028816094942567,
+      "learning_rate": 6.515151681101284e-09,
+      "loss": 0.507,
+      "step": 1011
+    },
+    {
+      "epoch": 2.95618305744888,
+      "grad_norm": 0.4346618256973241,
+      "learning_rate": 5.675579924011576e-09,
+      "loss": 0.5173,
+      "step": 1012
+    },
+    {
+      "epoch": 2.9591041869522883,
+      "grad_norm": 0.4213345827232603,
+      "learning_rate": 4.89386740013198e-09,
+      "loss": 0.4984,
+      "step": 1013
+    },
+    {
+      "epoch": 2.962025316455696,
+      "grad_norm": 0.40444422520787165,
+      "learning_rate": 4.170023165600001e-09,
+      "loss": 0.5463,
+      "step": 1014
+    },
+    {
+      "epoch": 2.964946445959104,
+      "grad_norm": 0.4158643063044851,
+      "learning_rate": 3.5040556061483043e-09,
+      "loss": 0.4929,
+      "step": 1015
+    },
+    {
+      "epoch": 2.9678675754625123,
+      "grad_norm": 0.45968510533530904,
+      "learning_rate": 2.8959724370092313e-09,
+      "loss": 0.5618,
+      "step": 1016
+    },
+    {
+      "epoch": 2.97078870496592,
+      "grad_norm": 0.45874234552575655,
+      "learning_rate": 2.3457807028248734e-09,
+      "loss": 0.5245,
+      "step": 1017
+    },
+    {
+      "epoch": 2.9737098344693282,
+      "grad_norm": 0.4715342361612812,
+      "learning_rate": 1.8534867775649169e-09,
+      "loss": 0.5129,
+      "step": 1018
+    },
+    {
+      "epoch": 2.976630963972736,
+      "grad_norm": 0.4184866954459166,
+      "learning_rate": 1.4190963644522549e-09,
+      "loss": 0.4844,
+      "step": 1019
+    },
+    {
+      "epoch": 2.979552093476144,
+      "grad_norm": 0.41399082545193333,
+      "learning_rate": 1.0426144958985974e-09,
+      "loss": 0.5588,
+      "step": 1020
+    },
+    {
+      "epoch": 2.982473222979552,
+      "grad_norm": 0.445097303921185,
+      "learning_rate": 7.240455334456276e-10,
+      "loss": 0.4614,
+      "step": 1021
+    },
+    {
+      "epoch": 2.98539435248296,
+      "grad_norm": 0.4214572233508389,
+      "learning_rate": 4.633931677139325e-10,
+      "loss": 0.5183,
+      "step": 1022
+    },
+    {
+      "epoch": 2.988315481986368,
+      "grad_norm": 0.4246075700283406,
+      "learning_rate": 2.60660418360259e-10,
+      "loss": 0.5797,
+      "step": 1023
+    },
+    {
+      "epoch": 2.991236611489776,
+      "grad_norm": 0.4332145327952596,
+      "learning_rate": 1.158496340419868e-10,
+      "loss": 0.5132,
+      "step": 1024
+    },
+    {
+      "epoch": 2.994157740993184,
+      "grad_norm": 0.4486911892623868,
+      "learning_rate": 2.8962492393258546e-11,
+      "loss": 0.5159,
+      "step": 1025
+    },
+    {
+      "epoch": 2.9970788704965923,
+      "grad_norm": 0.4412334672273621,
+      "learning_rate": 0.0,
+      "loss": 0.5245,
+      "step": 1026
+    },
+    {
+      "epoch": 2.9970788704965923,
+      "step": 1026,
+      "total_flos": 883131557314560.0,
+      "train_loss": 0.5971221869800523,
+      "train_runtime": 15429.9437,
+      "train_samples_per_second": 6.384,
+      "train_steps_per_second": 0.066
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1026,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 883131557314560.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}